In a typical herbaceous dicotyledon (Figure 1.1) lower leaf surfaces are covered with epidermal outgrowths, known to impede movement of small insects, but also contributing to formation of a boundary layer. This unstirred zone of air immediately adjacent to upper and lower epidermes varies in thickness according to surface relief, area and wind speed. Boundary layers are significant in leaf heat budgets and feature in the calculation of stomatal and mesophyll conductances from measurements of leaf gas exchange.  

The diffusion of CO₂ into leaves can be modelled using an analogue with electrical resistance (R) and conductance (the reverse of resistance), as in Figure 1.1, right hand side. This shows a series of resistances (r) that would be experienced by CO₂ molecules diffusing from outside (ambient) air, through the boundary layer (b), the stomata (s), the intercellular airspaces (i), the cell walls and liquid phase (w) to fixed sites inside chloroplasts. These values emphasise the prominence of stomatal resistance within the series.

Corresponding values for CO₂ concentration in ambient air (a), the leaf surface (s), the substomatal cavity (i), the mesophyll cell wall surface (w) to the sites of carboxylation with the chloroplasts (c) reflect photosynthetic assimilation within leaves generating a gradient for inward diffusion.

In transverse fracture as shown below in Figure 1.2(A) the bifacial nature of leaf mesophyll is apparent with columnar cells in the palisade layer beneath the upper surface and irregular shaped cells forming the spongy mesophyll below. Large intercellular airspaces, particularly in the spongy mesophyll, facilitate gaseous diffusion. The lower surface of this leaf is shown in Figure 1.2(B). On the left-hand side, the epidermis is present with its irregular array of stomata. Diagonally through the centre is a vein with broken-off hair cells and on the right the epidermis has been fractured off revealing spongy mesophyll cells beneath. Light micrographs of sections cut parallel to the leaf surface (paradermal) through palisade (C) and spongy (D) tissue reveal chloroplasts lying in a single layer and covering most of the internal cell wall surface adjacent to airspaces. Significantly, chloroplasts are rarely present on walls that adjoin another cell. Despite the appearance of close packing, mesophyll cell surfaces within the palisade layer are generally exposed to intercellular airspace. Inward diffusion of CO₂ to chloroplasts is thereby facilitated. 

Leaves that develop in sunny environments and have high photosynthetic capacities are generally thicker than leaves from shaded environments. This is achieved with more elongate cells within the palisade layer and/or several layers of cells forming the palisade tissue. Thicker leaves in a sunny environment enable more Rubisco to be deployed which confers a higher photosynthetic capacity. Fitting more Rubisco into a unit of leaf area with good access to intercellular airspace requires an increase in mesophyll cell surface which is possible by increasing the thickness of the mesophyll tissue and hence leaf thickness. A thicker leaf in sunny environments is energy effective because enough photons reach chloroplasts in lower cell layers to keep their Rubisco gainfully employed. By contrast, in a shaded habitat, less Rubisco is required for a leaf with lower photosynthetic capacity and this can be fitted into thinner leaves.

 

Pigments in thylakoid membranes of individual chloroplasts (Figure 1.7) are ultimately responsible for strong absorption of wavelengths corresponding to blue and red regions of the visible spectrum (Figure 1.3). Irradiated with red or blue light, leaves appear dark due to this strong absorption, but in white light leaves appear green due to weak absorption around 550 nm, which corresponds to green light. Ultraviolet (UV) light (wavelengths below 400 nm) can be damaging to macromolecules, and sensitive photosynthetic membranes also suffer. Consequently, plants adapt by developing an effective sunscreen in their cuticular and epidermal layers. 

Overall, absorption of visible light by mesophyll tissue is complex due to sieve-effects and scattering. Sieve-effect is an outcome from packaging pigments into discrete units, in this case chloroplasts, while remaining leaf tissue is transparent. This increases the probability that light can bypass some pigment and penetrate more deeply. A regular, parallel arrangement of columnar cells in the palisade tissue with chloroplasts all vertically aligned means that about 80% of light entering a leaf initially bypasses the chloroplasts, and measurements of absorption in a light integrating sphere confirm this. Scattering occurs by reflection and refraction of light at cell walls due to the different refractive indices of air and water. Irregular-shaped cells in spongy tissues enhance scattering, increasing the path length of light travelling through a leaf and thus increasing the probability of absorption. Path lengthening is particularly important for those wavelengths more weakly absorbed and results in nearly 80% absorption, even at 550 nm (Figure 1.3). Consequently, leaves typically absorb about 85% of incident light between 400 and 700 nm; only about 10% is reflected and the remaining 5% is transmitted. These percentages do of course vary according to genotype × environment factors, and especially adaptation to aridity and light climate.

Sunlight entering leaves is attenuated with depth in much the same way as light entering a canopy of leaves shows a logarithmic attenuation with depth that follows Beer’s Law (Section 12.4). Within individual leaves, the pattern of light absorption is a function of both cell anatomy and distribution of pigments. An example of several spatial profiles for a spinach leaf is shown in Figure 1.4. Chlorophyll density peaks in the lower palisade layer and decreases towards each surface. The amount of light declines roughly exponentially with increasing depth through the leaf. Light absorption is then given by the product of the chlorophyll and light profiles. Light absorption initially increases from the upper surface, peaking near the base of the first palisade layer, then declines steadily towards the lower surface. Because light is the pre-eminent driving variable for photosynthesis, CO₂ fixation tends to follow the light absorption profile (see ¹⁴C fixation pattern in Figure 1.4). However, the profile is skewed towards the lower surface because of a non-uniform distribution of photosynthetic capacity. Chloroplasts near the upper surface have ‘sun’-type characteristics which include a higher ratio of Rubisco to chlorophyll and higher rate of electron transport per unit chlorophyll. Chloroplasts near the lower surface show the converse features of ‘shade’ chloroplasts. Similar differences between ‘sun’ and ‘shade’ leaves are also apparent. Chloroplast properties do not change as much as the rate of absorption of light. Consequently, the amount of CO₂ fixed per quanta absorbed increases with increasing depth beneath the upper leaf surface. The lower half of a leaf absorbs about 25% of incoming light, but is responsible for about 31% of a leaf’s total CO₂ assimilation.

Leaves are covered with a barrier or ‘cuticle’ on the outer walls of epidermal cells that is impermeable to both water and CO₂. To enable CO₂ entry into the leaf for photosynthesis, the epidermis is perforated by pores called stomata (Figure 1.5). As CO₂ molecules diffuse inwards they encounter an opposite flux of H₂O molecules rushing outwards that is three to four orders of magnitude stronger. This problem of transpirational water loss is a particular problem for plants in hot, dry climates, such as in most of Australia. Leaves control this gas exchange by adjusting the aperture of stomata which can vary within minutes in response to changes in several environmental variables including light, humidity and CO₂ concentration (see Chapter 15 for more details). Air-spaces inside leaves are effectively saturated with water vapour (equivalent to 100% relative humidity at that leaf temperature) and because air surrounding illuminated leaves is almost universally drier, water molecules diffuse outwards down this concentration gradient from leaf to air. 

The diffusion pathway for H₂O out of a leaf is usually divided into two parts, namely the boundary layer of still air at the leaf surface and stomatal pores (Figure 1.5). Boundary layer thickness depends on windspeed, leaf dimensions and the presence of surface structures (e.g. hairs in Figure 1.1). Positioning of stomata also varies between species. Leaves of terrestrial plants always have stomata on their lower (abaxial) surface but many species have stomata on both surfaces, especially if they have high photosynthetic rates and are in sunny locations such as pendulant leaves of eucalypts. Adaptations for arid environments include having surface structures like hairs and waxes, which increase the thickness of the boundary layer, and leaf rolling and encryption of stomata by placing them in crevices in the leaf surface. While these features restrict water loss, they also impose an increased resistance (decreased conductance) to CO₂ uptake. 

The flux of water escaping from a leaf, called transpiration rate, can be understood from Fick’s law. It depends on the product between conductance and the gradient in water vapour from the inside of the leaf to the surrounding air. The vapour pressure gradient depends on both the humidity of the surrounding air and leaf temperature. Dry air (low humidity), or hotter leaf temperatures will result in greater transpiration rates for a given conductance. Maximum leaf conductance depends on the number and size of stomata per unit leaf area which is a leaf property that becomes fixed during development. However, the aperture of stomata can be varied, so stomatal conductance can vary over the timescale of minutes. Stomatal conductance responds to light, CO₂ and humidity. The sensitivity of a leaf to these variables is not fixed but can change over time in response to, for example, drought. Transpiration rate can be measured by a variety of means. With the availability of portable instruments, it is now most commonly obtained by measuring the increase in water vapour content of air from a leaf enclosed in a chamber. Stomatal conductance can then be calculated from Fick’s law by dividing the transpiration rate by the vapour pressure gradient between the leaf and the air.

CO₂ molecules diffusing inwards from ambient air to chloroplasts encounter restrictions additional to boundary layer and stomata (Figure 1.5). CO₂ must also diffuse from substomatal cavities throughout the mesophyll, dissolve in wet cell walls, cross the plasma membrane to enter the cytosol, diffuse into chloroplasts across a double membrane (outer envelope in Figure 1.7) and finally reach fixation sites within the stroma of those chloroplasts. The combination of these restrictions from intercellular airspace to the sites of fixation within chloroplasts has been termed mesophyll conductance.

There is considerable variation in leaf anatomy and hence potential restriction to CO₂ diffusion, but in general leaves with high rates of photosynthesis tend to have more permeable leaves (e.g. tobacco in Figure 1.2) and this complex anatomy ensures a greatly enlarged surface area for diffusion across interfaces. Indeed the total mesophyll cell wall area can be 20 times that of the projected leaf surface.

Chloroplasts tend to be appressed against cell walls adjacent to intercellular spaces (Figure 1.2 C, D) which improves access to CO₂, and they contain carbonic anhydrase which speeds up diffusion of CO₂ by catalysing interconversion of CO₂ and bicarbonate within the stroma of chloroplasts. Although CO₂ rather than HCO₃^– is the substrate species for Rubisco, the presence of carbonic anhydrase enables bicarbonate ions, which are more abundant under the alkaline conditions (pH 8.0) that prevail inside chloroplasts, to diffuse to Rubisco in concert with diffusion of CO₂. By sustaining a very rapid equilibration between CO₂ and HCO₃^– immediately adjacent to active sites on Rubisco, carbonic anhydrase enhances inward diffusion of inorganic carbon.

Light impinging on plants arrives as discrete particles we term photons, so that a flux of photosynthetically active photons can be referred to as ‘photon irradiance’. Each photon carries a quantum of electromagnetic (light) energy. In biology the terms photon and quantum (plural quanta) tend to be used interchangeably.

CO₂ assimilation varies according to both light and CO₂ partial pressure. At low light (low photon irradiance in Figure 1.6) assimilation rate increases linearly with increasing irradiance, and the slope of this initial response represents maximum quantum yield (mol CO₂ fixed per mol quanta absorbed). Reference to absorbed quanta in this expression is important. Leaves vary widely in surface characteristics (hence reflectance) as well as internal anatomy and chlorophyll content per unit leaf area. Therefore, since absorption of photosynthetically active quanta will vary, quantum yield expressed in terms of incident irradiance does not necessarily reflect the photosynthetic efficiency of the mesophyll. In the case of comparisons between sun and shade leaves, it has led to a widely held but mistaken belief that shade leaves (thinner and with higher chlorophyll content) are more efficient. Expressed in terms of absorbed quanta, sun and shade leaves have virtually identical quantum efficiencies for CO₂ assimilation. 

Assimilation rate increases more slowly at higher irradiances until eventually a plateau is reached where further increases in irradiance do not increase the rate of CO₂ assimilation (Figure 1.6). Chloroplasts are then light saturated. Absolute values for both quantum yield and light-saturated plateaux depend on CO₂ concentration. Quantum yield increases as CO₂ concentration increases as it competes more successfully with other species such as oxygen, at the binding site on Rubisco. Leaf absorptance has a hyperbolic dependence on chlorophyll content. For most leaves, 80–85% of 400–700 nm light is absorbed and it is only in leaves produced under severe nitrogen deficiency where there is less than 0.25 mmol Chl m^–2 that absorptance falls below 75%. 

The plateau in Figure 1.6 at high irradiance is set by maximum Rubisco activity. With increasing CO₂ partial pressure, the rate of carboxylation increases. The transition from light-limited to Rubisco-limited CO₂ assimilation as irradiance increases becomes progressively more gradual at higher CO₂ partial pressures. In part, this gentle transition reflects the fact that a leaf is a population of chloroplasts which have different photosynthetic properties depending on their position within that leaf. As discussed above, the profile of photosynthetic capacity per chloroplast changes less than the profile of light absorption per chloroplast (Figure 1.4). This results in an increase in CO₂ fixed per quanta absorbed with increasing depth. A transition from a light to a Rubisco limitation therefore occurs at progressively higher incident irradiances for each subsequent layer and results in a more gradual transition in the irradiance response curve of a leaf compared to that of a chloroplast. 

Photosynthetic capacity of leaves varies widely according to light, water and nutrient availability and these differences in capacity usually reflect Rubisco content. Leaves in high light environments (‘sun’ leaves) have greater CO₂ assimilation capacities than those in shaded environments and this is reflected in the larger allocation of nitrogen-based resources to photosynthetic carbon reduction (PCR cycle; Section 2.1). Sun leaves have a high stomatal density, are thicker and have a higher ratio of Rubisco to chlorophyll in order to utilise the larger availability of photons (and hence ATP and NADPH). Shade leaves are larger and thinner, but have more chlorophyll per unit leaf dry weight than sun leaves. They can have a greater quantum yield per unit of carbon invested in leaves, but with a relatively greater allocation of nitrogen-based resources to photon capture, shade leaves achieve a lower maximum rate of assimilation.

Despite such differences in leaf anatomy and chloroplast composition, leaves sustain energy transduction and CO₂ fixation in an efficient and closely coordinated fashion. Processes responsible are discussed below (Section 1.2).

Chloroplasts are easily recognised under a light microscope in leaf sections as distinctive green organelles suspended in the cytoplasm and usually appressed against cell walls. Chloroplasts are abundant in mesophyll tissue (commonly 200–300 in each palisade cell) and functional organelles can be isolated from homogenates of leaf tissue. 

Chloroplasts are surrounded by a double membrane, or envelope, just visible in transmission electron micrographs (Figure 1.7). This envelope encapsulates a soluble (gel-like) stroma which contains all the enzymes necessary for carbon fixation, many enzymes of nitrogen and sulphur metabolism and the chloroplast’s own genetic machinery. 

The inner membrane of a chloroplast envelope is an effective barrier between stroma and cytoplasm, and houses transporters for phosphate and metabolites (Section 2.1.8) as well as some of the enzymes for lipid synthesis. By comparison, the outer membrane of the chloroplast envelope is less complex and more permeable to both ions and metabolites. 

Suspended within the stroma, and entirely separate from envelope membranes, is an elaborately folded system of photosynthetic membranes or ‘thylakoids’ (literally ‘little sacs’). Embedded within these membranes are the complexes that enable light harvesting and electron flow from H₂O molecules to NADP+, thereby converting light energy into chemically usable forms. There are four basic complexes comprising two types of photosystem (with interlinked protein and pigment molecules), cytochrome b/f complexes (pivotal for photosynthetic electron transport) and ATP synthase complexes (responsible for proton egress from thylakoid lumen to stroma, and consequent ATP generation). These complexes are densely packed within the thylakoids. This remarkable transduction of energy, with such profound implications for life as we know it, starts with selective absorption of incoming light by chlorophylls and accessory pigments (certain carotenoids) that operate within both photosystems.

Chlorophylls are readily extracted from (soft) leaves into organic solvent and separated chromatographically into constituent types, most notably chlorophyll a (Chl a) and chlorophyll b (Chl b). These two chemical variants of chlorophyll are universal constituents of wild vascular plants and express highly characteristic absorption spectra (Figure 1.8, upper curves). Both chlorophylls show absorption maxima at wavelengths corresponding to blue and red, but chlorophyll assay in crude extracts, which inevitably contain carotenoids as well, is routinely based on absorption maxima in red light to avoid overlap with these accessory pigments that show strong absorption below 500 nm. Absorption maxima at 659 and 642 for Chl a and Chl b respectively would thus serve for assay in diethylether, but these peaks will shift slightly according to solvent system, and such shifts must be taken into account for precise measurement (see Porra et al. 1989 for details). Additional chlorophylls have been discovered that exist in cyanobacteria which extends their absorption spectrum into the infrared (Figure 1.9).

Chl a and Chl b differ with respect to both role and relative abundance in higher plants. Chl a/b ratios commonly range from 3.3 to 4.2 in well-nourished sun-adapted species, but can be as low as 2.2 or thereabouts in shade-adapted species grown at low light. Such variation is easily reconciled with contrasting functional roles for both Chl a and Chl b. Both forms of chlorophyll are involved in light harvesting, whereas special forms of only Chl a are linked into energy-processing centres of photosystems. In weak light, optimisation of leaf function calls for greater investment of leaf resources in light harvesting rather than energy processing. As a result the relative abundance of Chl b will increase and the Chl a/b ratio will be lower compared with that in strong light. Conversely, in strong light, photons are abundant and require greater capacity for energy processing by leaves (hence the higher Chl a/b ratio).  As a further subtlety, the two photosystems of higher plant chloroplasts (discussed later) also differ in their Chl a/b ratio, and this provided Boardman and Anderson (1964) with the first clue that they had achieved a historic first in the physical separation of those two entities. 

Carotenoids also participate in photosynthetic energy transduction. Photosystems have an absolute requirement for catalytic amounts of these accessory pigments, but their more substantive involvement is via dissipation of potentially harmful energy that would otherwise impact on delicate reaction centres when leaves experience excess photon irradiance (further details in Chapter 12). Carotenoids are thus regarded as ‘accessory’ to primary pigments (chlorophylls) and in molar terms are present in mature leaves at about one-third the abundance of Chl (a + b). 

Chlorophyll in leaves is not free in solution but is held in pigment-protein complexes, each with a different absorption spectrum (see Evans and Anderson 1987). In particular, light-harvesting Chl a, b–protein complexes (LHC in Figure 1.8, lower curves) develop a secondary absorption peak at 472 nm with a shoulder at 653 nm, while the Chl a of photosystem II reaction centres shows absorption peaks at 437 and 672 nm (compared with 429 and 659 nm for purified Chl a in ether; Figure 1.8, upper curves).

Subtle alterations in the molecular architecture of chlorophyll molecules according to the particular protein to which they bind in either light-harvesting or energy-processing centres are responsible for these shifts in absorption peaks, and for a general broadening of absorption spectra (compare lower and upper curves in Figure 1.8). Such effects are further accentuated within intact leaves by accessory pigments and greatly lengthened absorption pathways resulting in about 85% of visible wavelengths being absorbed (Figure 1.10). Any absorbed quanta at wavelengths below 680 nm can drive one electron through either reaction centre. Maximum quantum yield (Figure 1.10) occurs when both reaction centres absorb equal numbers of such quanta. When one photosystem population (PSII) absorbs more quanta than the other (PSI), excess quanta cannot be used to drive whole-chain (linear) electron flow. Quantum yield is reduced as a consequence, and leads to a slight discrepancy between in vivo absorption maxima (Figure 1.8) and quantum yield (Figure 1.10).

Although UV wavelengths are absorbed by leaves and would be capable of driving photosynthesis, such short wavelengths are damaging to biological systems and plants have adapted by developing a chemical sunscreen. Consequently, the quantum yield from these wavelengths drops off markedly below about 425 nm. Beyond 700 nm (infrared band) absorption drops to near zero, and forestalls leaf heating from this source of energy. However, quantum yield falls away even faster, and this ‘red drop’, though puzzling at first, led subsequently to a comprehensive model for photosynthetic energy transduction, outlined below.

Plants and many algae contain two distinct protein complexes for trapping and processing photons of light; photosystems I and II (PSI and PSII). These two systems can be separated and identified using a combination of biochemical and chemical techniques. Within the chloroplast, however, these two systems must work cooperatively and sequentially to absorb photons and convert their quantum energy into a flow of electrons. Interestingly, although PSI was discovered first, in cyanobacteria, photosynthetic electron flow is initiated in PSII and then proceeds to PSI. In PSII electrons are provided through the splitting of water molecules. PSI is responsible for finally delivering these electrons to NADH+.

This section presents a historical account of the discovery of the two photosystems and how they work together to split water and produce NADH+.

Prior to the advent of high-precision leaf gas exchange methods (as employed for Figure 1.10), O₂ evolution was taken as a measure of photosynthetic activity. Action spectra were measured on a number of plants and algae over the range of visible radiation. A crucial and consistent observation was that O₂ evolution dropped off much faster in the long-wavelength red region (>690 nm) than did absorption. Put another way, more quanta were being absorbed at longer wavelengths than could be used for photosynthesis. It seemed at these longer wavelengths as though a light absorber was being robbed of energy-processing capacity.

Anticipating that bimodal absorption implied a two-step process, and knowing that chlorophyll also absorbed photons at shorter wavelengths, Robert Emerson (working at Urbana in the mid-1950s) supplemented far-red light with shorter wavelength red irradiance and demonstrated that the relatively low photosynthetic rate in far-red light could be significantly increased. In fact the photosynthetic rate achieved with the two light qualities combined could be 30–40% higher than the sum of the rates in far-red or shorter red when measured separately (Emerson et al. 1957). This phenomenon became known as the ‘Emerson Enhancement Effect’ and contributed to a working hypothesis for photosynthetic energy conversion based upon two photochemical acts (proposed by Duysens et al. 1961), but additional lines of evidence were impacting on this outcome.

At about the same time as Emerson was establishing his enhancement effect, Myers and French observed ‘sequential enhancement’; that is, a disproportionate increase in photosynthetic rate or efficiency when the two light qualities were separated in time. The upper limits of dark intervals between two flashes of different light quality were 6 s for far-red after green and 1 min for green after far-red. Clearly, the ‘product’ of photochemical act 1 was stable for 1 min, that of act 2 for only 6 s. This discovery implied that chemical intermediates, rather than an altered physical state, were involved in a two-step cooperation (see Clayton 1980).

According to physical laws of photochemical equivalence, there should be a 1:1 yield in converting light energy to chemical energy by a perfect system. Quantum requirement for such events would be 1. However in photosynthesis the absolute quantum requirement for O₂ is much greater than I. In the 1950s, Robert Emerson (at Urbana) and co-workers determined that 8-10 quanta were required. Hill and Bendell (1960) suggested a 'Z' scheme that was consistent with a requirment of 8-10 quanta, the cooperation of 2 quanta in the separation of one strong reducing and one strong oxidising equivalent, and the operation of two sequential photochemical acts. Figure 1.11 is a greatly developed version of their original model.

In linear flow, water molecules are split in PSII, liberating O₂ and providing a source of electrons. M is the manganese—containing cluster which oxidises water, Z is tyrosine-161 of the D1 protein which in turn represents the primary electron donor to P680⁺ (a special pair of Chl a molecules with an absorption peak at 680 nm). Pheo is the primary electron acceptor pheophytin a, a chlorophyll molecule lacking magnesium; Q_A is the first stable and permanently bound plastoquinone electron acceptor; Q_B is the second, temporarily bound, plastoquinone electron acceptor which actually leaves PSII in a reduced form (PQH₂). Further along, FeS = Rieske iron—sulphur centre; Cyt f = cytochrome f; PC = plastocyanin; P700 = reaction centre chlorophyll a of PSI; A₀, A₁, F_X, F_B and F_A are electron acceptors of PSI; Fd = ferredoxin; Cyt b559 = cytochrome b559; Cyt b563 = cytochrome b563. Also shown as tapered arrows is H⁺ accumulation in the lumen associated with water and plastoquinol oxidations.

The original version of this ‘Z’ scheme was further validated by unequivocal evidence from Australia that the two (inferred) photosystems were indeed separate physical entities. Using sophisticated biochemical chloroplast purification and subfractionation methods, coupled with detergent solubilisation of membranes, Boardman and Anderson (1964) achieved the first physical separation of photosystem II (PSII) and photosystem I (PSI), thus confirming the separate identities of those complexes. 

A source of electrons had long been recognised as basic to the operation of this ‘Z’ scheme, with H₂O molecules an obvious source, but were photosynthetic membranes capable of photolysis? Early experiments by Robin Hill and colleagues at Cambridge had established this capability. They used isolated thylakoid membrane preparations and showed that O₂ could be evolved in the absence of CO₂ as long as external electron acceptors were present (Hill reaction). Intact leaves or whole chloroplasts have no need for an artificial acceptor because electron flow is directed to NADP⁺ and subsequent reduction of CO₂ (first demonstrated with intact chloroplasts; see Arnon 1984). The O₂-evolving function of photosynthesis was found to be associated with PSII in experiments with isolated thylakoids using external (artificial) electron donors and acceptors and specific electron transport inhibitors. As one outcome of those early Cambridge experiments, O₂ evolution is now measured routinely in vitro (and in vivo on leaves) with O₂ electrodes (Walker 1987). 

Chloroplast structure and function is by now sufficiently well defined to consider photosynthetic electron flow in detail. Figure 1.11 applies equally well to vascular plants or to algae with oxygenic photosynthesis, where in either case two photosystems work cooperatively and sequentially in absorbing photons and converting their quantum energy into a flow of electrons. Paradoxically, convention has it that photosynthetic electron flow initiates in PSII and proceeds to PSI. PSII was so named because PSI had already been described in single-celled (prokaryotic) organisms and, owing to the rules of nomenclature, was accorded priority. 

Both photosystems are large multi-subunit complexes, quite different structurally and functionally, and operating in series. In PSII, electrons are provided from a water-splitting apparatus via a manganese complex which undergoes oxidation from a valency state of +2 to +4. These oxidation states are made possible by P680⁺ (a special form of Chl a with an absorption peak at 680 nm). P680⁺ is a powerful oxidant generated by absorption of energy from a photon. P680 is referred to as a ‘special pair’ because it is a pair of Chl a molecules. Electrons from P680 pass to pheophytin (Pheo in Figure 1.11) and on to a bound quinone molecule, Q_A. From there a second transiently bound quinone, Q_B, receives two electrons in succession and requires protonation. The entire, fully reduced, quinone molecule leaves PSII and enters a plastoquinone pool (2PQ).

In PSI, absorption of quantum energy from a photon causes oxidation of P700, the PSI reaction centre equivalent of P680. In contrast to PSII, where electrons are drawn from a water-splitting apparatus, P700 accepts electrons from PC (reduced form PC^– in Figure 1.12). Electrons then pass through three iron–sulphur (FeS) centres and out of PSI to ferredoxin (Fd). The reaction centre of PSI contains several proteins, but most of the electron transfer cofactors are bound to large heterodimeric proteins which in turn bind the inner Chl a antenna. The LHCI complex consists of possibly eight polypeptides of between 24 and 27 kDa which carry Chl a and Chl b plus carotenoids.

A chemiosmotic coupling mechanism is responsible for ATP synthesis. Protons are ‘pumped’ across the thylakoid membrane from outside (stroma) to inside (lumen) by a complex arrangement of electron carriers embedded within the membrane. A prodigious concentration of protons builds up within the lumen, partly from photolysis of water molecules (water-splitting apparatus on PSII) and partly from oxidation of plastoquinone (PQ) on the inner face of the membrane. Hence, energy originally carried by incident photons is transduced into energy stored within an electrochemical gradient acrosss the thylakoid membrane. The protonmotive force from inside (lumen) to outside (stroma) is used to generate ATP within the stroma via an ATP synthase complex (CF₀ and CF₁) that straddles the thylakoid membrane. OEC = oxygen-evolving complex; Pheo = pheophytin a.These two photosystems are juxtaposed across thylakoid membranes in such a way that linear electron transport is harnessed for charge separation, leading to a massive accumulation of H⁺ ions within the lumen of illuminated thylakoids, which is then employed in ATP generation.

Combining concepts of photolysis and photosynthetic electron flow outlined earlier (Figure 1.11) and putting that conceptual framework into a thylakoid membrane system (Figure 1.12), a picture emerges where electrons generated from splitting H₂O molecules on the inner surface of PSII are transferred from plastoquinol (PQH₂) to the Rieske iron– sulphur centre (Rieske FeS) of the cytochrome b₆/f complex (Cyt b₆/f) and further to cytochrome f (Cyt f). The pivotal importance of Cyt f in facilitating electron transport from PSII to PSI was demonstrated by Duysens and colleagues (see Levine 1969), who showed that preferential energisation of PSII (light at <670 nm) caused reduction, whereas preferential energisation of PSI (light at >695 nm) caused oxidation. This elegant ‘push–pull’ experiment confirmed the cooperative and sequential nature of PSII and PSI, as well as indicating overall direction of photosynthetic electron flow. 

Proteins which bind the Rieske FeS centre and Cyt f together with cytochrome b₅₆₃ (Cyt b₆) form a large electron transfer complex. This complex (Figure 1.12) spans the membrane and is located between the two photosystems. Electrons are transferred to PC (forming PC^–), a copper-containing soluble protein extrinsic to the thylakoid membrane and located in the lumen. On the other side of the membrane, attached to the stromal side, is ferredoxin (Fd) which accepts electrons from PSI and passes them on to ferredoxin–NADP reductase, an enzyme, also extrinsic to thylakoids, and attached on the stromal side of the thylakoid membrane. This enzyme accomplishes the final electron transfer in an overall linear chain and reduced NADP is then protonated.

While linear electron transport from water to NADP⁺ is the main and most important path, electrons can also be transferred to O₂ in a so-called pseudocyclic or Mehler reaction (Figure 1.11). This pathway probably operates in vivo as a sink for electrons when synthetic events call for more ATP than NADPH. Electrons can also be cycled around both PSII and PSI. Electrons cycling around PSI will produce ATP but with no accompanying NADPH. Cyclic electron flow around PSII may have a completely different role and may be related to the downregulation of this photosystem during photoinhibition (Chapter 12).

According to this multistage scheme, electrons are transferred from donor (reductant) to acceptor (oxidant). The direction of that transfer depends upon a difference in oxidation–reduction potential between a given donor and a given acceptor (as indicated on the ordinate in Figure 1.11). A more positive potential implies stronger oxidative power (i.e. capacity to accept electrons); a more negative potential implies stronger reducing power (i.e. capacity to donate electrons). P680* thus has a strong capacity to donate electrons (a strong reductant); P700* has an even stronger capacity to donate electrons (an even stronger reductant). 

Molecules which accept electrons are immediately protonated. In aqueous systems, such as chloroplasts in vivo, hydrogen ions (H⁺) are ubiquitous, and these ions combine with electron acceptors to generate hydrogen atoms (i.e. H⁺ ion + electron → H atom). In Figure 1.11, some events involve electron transfer, while others include transfer of hydrogen atoms. As a simplifying convention, all such events are referred to as electron transfers. Ironically, the end result of all these reactions is a net transfer of hydrogen atoms!

During photosynthetic electron transfer from water to NADP⁺, energy captured in two photoacts is stored as an electrochemical potential gradient of protons. First, such reduction of Q_B requires protonation with protons drawn from the stromal side of the membrane. Reoxidation (and deprotonation) occurs towards the thylakoid lumen. In addition, protons are lost from the stromal side via protonation of reduced NADP and they are also generated in the lumen during photolysis. A massive ΔpH, of approximately 3–4 pH units, equivalent to an H⁺ ion concentration difference of three to four orders of magnitude, develops across the thylakoid membrane. This immense gradient drives ATP synthesis (catalysed by ATP synthase) within a large energy-transducing complex embedded in the thylakoid membrane (Figure 1.12). 

ATP synthesis in chloroplasts (photophosphorylation) proceeds according to a mechanism that is basically similar to that in mitochondria. Chemiosmotic coupling (Mitchell 1961) which links the movement of protons down an electro-chemical potential gradient to ATP synthesis via an ATP synthase applies in both organelles. However, the orientation of ATP synthase is opposite. In chloroplasts protons accumulate in thylakoid lumen and pass outwards through the ATP synthase into the stroma. In mitochondria, protons accumulate within the intermembrane space and move inwards, generating ATP and oxidising NADH within the matrix of these organelles (Figure 2.24).

In chloroplasts, ATP synthase is called the CF₀CF₁ complex. The CF₀ unit is a hydrophobic transmembrane multiprotein complex which contains a water-filled proton conducting channel. The CF₁ unit is a hydrophilic peripheral membrane protein complex that protrudes into the stroma. It contains a reversible ATPase and a gate which controls proton movement between CF₀ and CF₁. Entire CF₀CF₁ complexes are restricted to non-appressed portions of thylakoid membranes due to their bulky CF₁ unit. 

Direct evidence for ATP synthesis due to a transthylakoid pH gradient can be adduced as follows. When chloroplasts are stored in darkness in a pH 4.0 succinic acid buffer (i.e. a proton-rich medium), thylakoid lumen equilibrate to this pH. If the chloroplasts, still in the dark, are rapidly transferred to a pH 8.0 buffer containing ADP and P_i, ATP synthesis then occurs. This outcome confirms a central role for the proton concentration difference between thylakoid lumen and stroma for ATP synthesis in vitro; but does such a process operate on that scale in vivo? 

Mordhay Avron, based in Israel, answered this question in part during the early 1970s via a most elegant approach (Rottenberg et al. 1972). Working with thylakoid preparations, Avron and colleagues established that neutral amines were free to exchange between bathing medium and thylakoid lumen, but once protonated in illuminated preparations they became trapped inside. By titrating the loss of such amines from the external medium when preparations where shifted from dark to light, they were able to infer the amount retained inside. Knowing that the accumulation of amine depended upon H⁺ ion concentration in that lumen space, the difference in H⁺ ion concentration and hence ΔpH across the membrane were established. 

At saturating light, chloroplasts generate a proton gradient of approximately 3.5 pH units across their thylakoid membranes. Protons for this gradient are derived from the oxidation of water molecules occurring towards the inner surface of PSII and from transport of four electrons through the Cyt b/f complex, combined with cotranslocation of eight protons from the stroma into the thylakoid space for each pair of water molecules oxidised. Electrical neutrality is maintained by the passage of Mg²⁺ and Cl^– across the membrane, and as a consequence there is only a very small electrical gradient across the thylakoid membrane. The electrochemical potential gradient that yields energy is thus due almost entirely to the concentration of intrathylakoid H⁺ ions. 

For every three protons translocated via ATP synthase, one ATP is synthesised. Linear electron transport therefore generates about four molecules of ATP per O₂ evolved. Thus eight photons are absorbed for every four ATP molecules generated or for each O₂ generated. Cyclic electron transport is slightly more efficient at producing ATP and generates about four ATP per six photons absorbed. However, linear electron transport also generates NADPH, which is equivalent, in energy terms, to six ATP per O₂ released.

As implied in Figure 1.12, the four thylakoid complexes, PSII, PSI, Cyt b/f and ATP synthase, are not evenly distributed in plant thylakoid membranes but show a lateral heterogeneity. This distribution is responsible for the highly characteristic structural organisation of the continuous thylakoid membrane into two regions, one consisting of closely appressed membranes or granal stacks, the other of non-appressed stroma lamellae where outside surfaces of thylakoid membranes are in direct contact with the stroma. This structural organisation is shown on a modest scale in Figure 1.7, but extreme examples are evident in chloroplasts of shade-adapted species grown in low light (Chapter 12). Under such conditions, membrane regions with clusters of PSII complexes and Cyt b/f complexes become appressed into classical granal stacks. Cyt b/f complexes are present inside these granal stacks as well as in stroma lamellae, but PSI and ATP synthase are absent from granal stacks. Linear electron transport occurs in granal stacks from PSII in appressed domains to PSI in granal margins. Nevertheless, shade plants have only a low rate of linear electron transport because they have fewer Cyt b/f and to a lesser extent fewer PSII complexes compared to PSI, a consequence of investing more chlorophyll in each PSII to enhance light harvesting (see Anderson (1986) and Chapter 12 for more detail).

A dilute solution of leaf chlorophyll in organic solvent appears green when viewed in white light. Wavelengths corresponding to bands of blue and red have been strongly absorbed (Figure 1.8), whereas mid-range wavelengths corresponding to green light are only weakly absorbed, hence the predominance of those wavelengths in transmitted and reflected light. However, when viewed at right angles to the light source, the solution will appear deep red due to energy re-emitted as fluorescence (Figure 1.13). The red colour is evident regardless of the colour of the source light.

Chlorophyll within the two photosystems can absorb energy from incident photons. This absorbed energy can be dissipated by driving the processes of photosynthesis, as heat, or re-emitted as fluorescence radiation. These are all complementary processes so that fluorescence provides an important tool in the study of photosynthesis. The normal processes of photochemistry and electron transport within intact leaves typically reduce the amount of fluorescence, a process referred to as quenching. In the demonstration shown in Figure 1.13 the chlorophyll has been isolated from the plant these processes are disrupted, minimizing the quenching effects.

Fluorescence spectra are invariate, and the same spectrum will be obtained (e.g. Figure 1.8 inset) regardless of which wavelengths are used for excitation. This characteristic emission is especially valuable in identifying source pigments responsible for given emission spectra, and for studying changes in their photochemical status during energy transduction. 

Fluorescence emission spectra (Figure 1.8 inset) are always displaced towards longer wavelengths compared with corresponding absorption spectra (Stoke’s shift). As quantum physics explains, photons intercepted by the chromophore of a chlorophyll molecule cause an instantaneous rearrangement of certain electrons, lifting that pigment molecule from a ground state to an excited state which has a lifetime of c. 10^–9 s. Some of this excitation energy is subsequently converted to vibrational energy which is acquired much more ‘slowly’ by much heavier nuclei. A non-equilibrium state is induced, and molecules so affected begin to vibrate rather like a spring with characteristic periodicity, leading in turn to energy dissipation as heat plus remission of less energetic photons of longer wavelength.

Apart from their role in photon capture and transfer of excitation energy, photosystems function as energy converters because they are able to seize photon energy rather than lose as much as 30% of it through fluorescence as do chlorophylls in solution. Moreover, they can use the trapped energy to lift an electron to a higher energy level from where it can commence a ‘downhill’ flow via a series of electron carriers as summarised in Figure 1.11.

Protein structure confers very strict order on bound chlorophylls. X-ray crystallographic resolution of the bacterial reaction centre has given us a picture of the beautiful asymmetry of pigment and cofactor arrangements in these reaction centres, and electron diffraction has shown us how chlorophylls are arranged with proteins that form the main light-harvesting complexes of PSII. This structural constraint confers precise distance and orientation relationships between the various chlorophylls, as well as between chlorophylls and carotenoids, and between chlorophylls and cofactors enabling the photosystems to become such effective photochemical devices. It also means that only 2–5% of all the energy that is absorbed by a photosystem is lost as fluorescence. 

If leaf tissue is held at liquid nitrogen temperature (77 K), photosynthetic electron flow ceases and chlorophyll fluorescence increases, including some emission from PSI (Figure 1.14). Induction kinetics of chlorophyll fluorescence at 77 K have been used to probe primary events in energy transduction, and especially the functional state of photosystems. Present discussion is restricted to room temperature fluorescence where even the small amount of fluorescence from PSII is diagnostic of changes in functional state. This is because chlorophyll fluorescence is not emitted simply as a burst of red light following excitation, but in an ordered fashion that varies widely in flux during continuous illumination. These transient events (Figure 1.15) are referred to collectively as fluorescence induction kinetics, fluorescence transients, or simply as a Kautsky curve in honour of its discoverer Hans Kautsky (Kautsky and Franck 1943). 

At room temperature and under steady-state conditions, in vivo Chl a fluorescence from leaves show a characteristic emission spectrum with two distinct peaks around 680–690 nm and 750 nm, both of which mainly originate from photosystem II (Figure 1.14). Because  other chlorophyll molecules can reabsorb fluorescence emitted at 680–690 nm within a leaf, the spatial origin of fluorescence can differ between the 680 and 750nm fluorescence that is detected. The fluorescence waveband measured by room temperature fluorometers differs between instruments. 

(a) Fluorescence induction kinetics

Strength of emission under steady-state conditions varies according to the fate of photon energy captured by LHCII, and the degree to which energy derived from photosynthetic electron flow is gainfully employed. However, strength of emission fluctuates widely during induction (Figure 1.15) and these rather perplexing dynamics are an outcome of some initial seesawing between photon capture and subsequent electron flow. Taking Figure 1.11 for reference, complexities of a fluorescence transient (Figure 1.15) can be explained as follows. At the instant of excitation (zig-zag arrow), signal strength jumps to a point called \(F_0\) which represents energy derived largely from chlorophyll molecules in the distal antennae of the LHCII complex which fail to transfer their excitation energy to another chlorophyll molecule, but lose it immediately as fluorescence. \(F_0\) thus varies according to the effectiveness of coupling between antennae chlorophyll and reaction centre chlorophyll, and will increase due to high-temperature stress or photodamage. Manganese-deficient leaves show a dramatic increase in \(F_0\) due to loss of functional continuity between photon-harvesting and energy-processing centres of PSII (discussed further in Chapter 16). 

Returning to Figure 1.15, the slower rise subsequent to \(F_0\) is called \(I\), and is followed by a further rise to \(F_m\). These stages reflect a surge of electrons which fill successive pools of various electron acceptors of PSII. Significantly, F_m is best expressed in leaves that have been held in darkness for at least 10–15 min. During this dark pretreatment, electrons are drawn from Q_A, leaving this pool in an oxidised state and ready to accept electrons from PSII. An alternative strategy is to irradiate leaves with far-red light to energise PSI preferentially, and so draw electrons from PSII via the Rieske FeS centre. The sharp peak (\(F_m\)) is due to a temporary restriction on electron flow downstream from PSII. This constraint results in maximum fluorescence out of PSII at about 500 ms after excitation in Figure 1.15(a). That peak will occur earlier where leaves contain more PSII relative to electron carriers, or in DCMU-treated leaves. 

Photochemistry and electron transport activity always quench fluorescence to a major extent unless electron flow out of PSII is blocked. Such blockage can be achieved with the herbicide 3-(3,4-dichlorophenyl)-1,1-dimethyl urea (DCMU) which binds specifically to the D1 protein of PSII and blocks electron flow to Q_B. DCMU is a very effective herbicide because it inhibits photosynthesis completely. As a consequence, signal rise to \(F_m\) is virtually instantaneous, and fluorescence emission stays high. 

Variation in strength of a fluorescence signal from \(F_0\) to \(F_m\) is also called variable fluorescence (\(F_v\)) because scale and kinetics of this rise are significantly influenced by all manner of environmental conditions. \(F_0\) plus \(F_v\) constitute the maximal fluorescence (\(F_m\)) a leaf can express within a given measuring system. The \(F_v/F_m\) ratio, measured after dark treatment, therefore reflects the proportion of efficiently working PSII units among the total PSII population. Hence it is a measure of the photochemical efficiency of a leaf, and correlates well with other measures of photosynthetic effectiveness (discussed further in Chapter 12).

(b) Fluorescence relaxation kinetics

Both the patterns of initial induction of fluorescence, and its subsequent decay once the light has ceased, are important indicators of the underlying structure and function of photosynthetic systems. The latter is referred to as the relaxation kinetics of a fluorescence event. In a typical experiment the chlorophyll is exposed to repeated pulses of light and the relaxation kinetics measured (Figure 1.16).

Excellent fluorometers for use in laboratory and field such as the Plant Efficiency Analyser (Hansatech, King’s Lynn, UK) make accurate measurements of all the indices of the Kautsky curve and yield rapid information about photochemical capacity and response to environmental stress. Conventional fluorometers (e.g. Figure 1.15) use a given source of weak light (commonly a red light-emitting diode producing only 50–100 µmol quanta m^–2 s^–1) for both chlorophyll excitation and as a source of light for photosynthetic reactions.

Even more sophisticated is the Pulse Amplitude Modulated (PAM) fluorometer (Walz, Effeltrich, Germany) which employs a number of fluorescence- and/or photosynthesis-activating light beams and probes fluorescence status and quenching properties. These fluorimeters measure fluorescence excited by a weak source of light that is modulated: that is a beam that applies short, square pulses of saturating light for chlorophyll excitation on top of a constant beam of light that sustains photosynthesis (actinic light). A combination of optical filters plus sophisticated electronics is used to tune the detector to detect only fluorescence excited by the modulated light beam.

In this way, most of the continuous background fluorescence and reflected long-wavelength light is disregarded. Most significantly, relative fluorescence can be measured in full sunlight in the field. The functional condition of PSII in actively photosynthesising leaf tissue is thus amenable to analysis. This instrument also reveals the relative contributions to total fluorescence quenching by photochemical and non-photochemical processes and will help assess any sustained loss of quantum efficiency in PSII. Photosynthetic electron transport rates can be calculated concurrently. These techniques have revolutionised the application of chlorophyll fluorescence to the study of photosynthesis.

Photochemical quenching (\(q_p\)) varies according to the oxidation state of electron acceptors on the donor side of PSII. When Q_A is oxidised (e.g. subsequent to dark pretreatment), quenching is maximised. Equally, \(q_p\) can be totally eliminated by a saturating pulse of excitation light that reduces Q_A, so that fluorescence yield will be maximised, as in a PAM fluorometer. Concurrently, a strong beam of actinic light drives photosynthesis (maintaining linear electron flow) and sustaining a pH gradient across thylakoid membranes for ATP synthesis. Those events are a prelude to energy utilisation and contribute to non-photochemical quenching (\(q_n\)). This \(q_n\) component can be inferred from a combination of induction plus relaxation kinetics.

In Figure 1.16, a previously darkened radish leaf (Q_A oxidised and ready to receive an electron from P680; 'traps open') initially receives weak modulated light (<1 µmol quanta m^–2 s^–1) that is insufficient to close traps but sufficient to establish a base line for constant yield fluorescence (\(F_0\)). This value will be used in subsequent calculations of fluorescence indices. The leaf is then pulsed with a brief (0.8 s) saturating flash (9000 µmol quanta m^–2 s^–1) to measure \(F_m\). Pulses follow at 10 s intervals to measure \(F_m^\prime\). Actinic light (500 µmol quanta m^–2 s^–1) starts with the second pulse and pH starts to build up in response to photosynthetic electron flow. Photosynthetic energy transduction comes to equilibrium with these conditions after a minute or so, and fluorescence indices \(q_n\) and \(q_p\) can then be calculated as follows:

\[ q_n=\frac{F_m - F_m^\prime}{F_m - F_0} \text{, and } q_p=\frac{F_m^\prime-F}{F_m^\prime - F_0} \tag{1.1} \]
Under these steady-state conditions, saturating pulses of excitation energy are being used to probe the functional state of PSII, and by eliminating \(q_p\) the quantum efficiency of light-energy conversion by PSII (\(\Phi_{PSII}\)) can be inferred:

\[ \Phi_{PSII} = \frac{F_m^\prime - F}{F_m^\prime} \tag{1.2} \]
If overall quantum efficiency for O₂ evolution is taken as 10 (discussed earlier), then the rate of O₂ evolution by this radish leaf will be: 

\[ \Phi_{PSII} \times \text{photon irradiance}/10 \;(\mu\text{mol O}_2 m^{-2}s^{-1}) \tag{1.3} \]

In summary, chlorophyll fluorescence at ambient temperature comes mainly from PSII. This photosystem helps to control overall quantum efficiency of electron flow and its functionality changes according to environmental and internal controls. In response to establishment of a ΔpH across thylakoid membranes, and particularly when irradiance exceeds saturation levels, some PSII units become down-regulated, that is, they change from very efficient photochemical energy converters into very effective energy wasters or dissipators (Chapter 12). Large amounts of the carotenoid pigment zeaxanthin in LHCII ensure harmless dissipation of this energy as heat (other mechanisms may also contribute). PSII also responds to feedback from carbon metabolism and other energy-consuming reactions in chloroplasts, and while variation in pool size of phosphorylated intermediates has been implicated, these mechanisms are not yet understood.

The biochemical pathway of CO₂ fixation was discovered by feeding radioactively labelled CO₂ in the light to algae and then extracting the cells and examining which compounds accumulated radioactivity. Figure 2.1(b) shows a typical labelling ‘pattern’ for a C₃ plant. Here, a short burst of labelled CO₂ was given to the plants, then the label was ‘chased’ through the photosynthetic pathway by flushing with unlabelled air. Atmospheric CO₂ is initially incorporated into a five-carbon sugar phosphate (ribulose-1,5-bisphosphate or RuBP) to produce two molecules of the phosphorylated three-carbon compound 3-phosphoglycerate, often referred to as the acidic form 3-phosphoglyceric acid (3-PGA). Hence, plants which use Rubisco as their primary enzyme of CO₂ fixation from the air are called C₃ plants. Consequently, in C₃ plants, 3-PGA is the first labelled sugar phosphate detected after a pulse of ¹⁴CO₂ has been supplied (Figure 2.1b). In the PCR cycle, 3-PGA is phosphorylated by the ATP produced from thylakoid electron transport (see Chapter 1) and then reduced by NADPH to produce triose phosphate. Triose phosphates are the carbon backbones, produced by the PCR cycle, for the synthesis of critical carbohydrate for the maintenance of plant growth and the productive yield of stored carbohydrate in seed.

Newly synthesised triose phosphate faces three options. It can be (1) exported to the cytosol for sucrose synthesis and subsequent translocation to the rest of the plant, (2) recycled within the chloroplast to produce more RuBP or (3) diverted to produce starch (Figure 2.1a). This is shown by the time-course of the appearance of radioactivity in starch and sucrose after it has passed through 3-PGA (Figure 2.1b). The energy requirements of the PCR cycle are three ATP and two NADPH per CO₂ fixed, in the absence of any other energy-consuming processes.

Sucrose and starch synthesis

Most of the triose phosphate synthesised in chloroplasts is converted to either sucrose or starch. Starch accumulates in chloroplasts, but sucrose is synthesised in the surrounding cytosol, starting with the export of dihydroxyacetone phosphate and glyceraldehyde phosphate from the chloroplast. A condensation reaction, catalysed by aldolase, generates fructose-1,6-bisphosphate, and this is converted to fructose-6-phosphate after an hydrolysis reaction catalysed by fructose-1,6-phosphatase. Sucrose-6-phosphate synthase then generates sucrose-6-phosphate from the reaction of fructose-6-phosphate and UDP-glucose. The phosphate group is removed by the action of sucrose-6-phosphatase. This P_i is transported back into the chloroplast where it is available for ATP synthesis. For each molecule of triose phosphate exported from a chloroplast, one P_i is translocated inwards.

Sucrose synthesised within the cytosol of photosynthesising cells is then available for general distribution and is commonly translocated to other carbon-demanding centres via the phloem (see Chapter 5).

By contrast, starch synthesis occurs within chloroplasts. The first step is a condensation of glucose-1-phosphate with ATP. Starch synthase then transfers glucose residues from this molecule to the non-reducing end of a pre-existing molecule of starch. Starch consists of two types of glucose polymer, namely amylose and amylopectin. Amylose is a long, unbranched chain of D-glucose units connected via (α1–4) linkages. Amylopectin is a branched form, with (α1–6) linkages forming branches approximately every 24–30 glucose residues.

Ribulose bisphosphate (RuBP) is consumed in the carboxylating step of carbon fixation. If such fixation is to continue, RuBP must be regenerated, and in this case via the PCR cycle. The PCR cycle operates within the stroma of chloroplasts, and consists of a sequence of 11 steps where a three-carbon compound (3-phosphoglycerate) is phosphorylated, reduced to glyceraldehyde 3-phosphate and isomerised to dihydroxyacetone phosphate. Condensation of this three-carbon compound with glyceraldehyde 3-phosphate yields a six- carbon compound (fructose bisphosphate). Following a series of carbon shunts, involving four-, five- and seven-carbon compounds, RuBP is regenerated.

Important features of the PCR cycle include: (1) for every step of the cycle to occur once, three carboxylations must occur via ribulose bisphosphate carboxylase thus generating six moles of phosphoglycerate (18 carbons); (2) for one turn of the cycle, three molecules of RuBP participate (15 carbons) and thus a net gain of three carbons has occurred for the plant; (3) in regenerating three molecules of RuBP, nine ATP and six NADPH are consumed.

Photosynthetic carbon fixation in air is constrained by the kinetic properties of Rubisco. Form I Rubisco in higher plants is a large protein (approximately 550 kDa) comprised of eight large (approx. 50-55 kDa) and eight small subunits (approx. 13-18 kDa) to form an L₈S₈ hexadecamer. Rubisco synthesis and assembly in higher plants is a complex process whereby the large subunit gene (rbcL) is encoded in the chloroplast genome, while the small subunit genes (rbcS) are encoded as a multi-gene family in the nucleus. The Rubisco small subunits are translated as precursors in thc cytosol and are equipped with a transit-peptide to target them to the chloroplast. Upon import in the chloroplast the transit-peptide is cleaved by a stromal peptidase and the N-terminus modified by methylation of the n-terminal methionine. The large subunits  are synthesised within the chloroplast and also post-translationally modified through the removal of the the N-terminal methionine and serine amino acids and the subsequent acetylation of proline at the N-terminus and the methylation of lysine at position 14. The assembly of large and small subunits into functional hexadecameric Rubisco is reliant on the coordination of chloroplast-localised chaperones.

Despite selection pressure over evolutionary history, Rubisco remains an inefficient catalyst (Spreitzer and Salvucci 2002). Therefore, to achieve a productive maximum CO₂ assimilation rate (A_max), plants must compensate for catalytic inefficiency by investing large amounts of nitrogen in Rubisco. Consequently, Rubisco comprises more than 50% of leaf soluble protein in C₃ plants. On a global scale, this investment equates to around 10 kg of nitrogen per person!

More than 1000 million years of evolution has still not resulted in a ‘better’ Rubisco adapted for the current and future concentrations of CO₂. Such a highly conserved catalytic protein is an outcome of thermodynamic and mechanistic difficulties inherent to this reaction. Rubisco requires carbamylation of the absolutely-conserved residue K201 that is then stabilised by the binding of Mg²⁺. Without this activation step Rubisco is unable to function. The fixation of CO₂ to RuBP to form two molecules of 3-PGA is a five step catalytic process that produces highly reactive transition state intermediates that bind CO₂. The highly reactive transition states make Rubisco prone to generating misfiring products, which generate inhibitors within the active site. Therefore, Rubisco requires its own catalytic protection enzyme Rubisco activase. Plants devoid of this enzyme fail to grow properly in air as the activation and subsequent activity of Rubisco is impeded (Portis and Salvucci 2002). Rubisco activase is an ATP-dependent process that removes inhibitors from the active site of Rubisco allowing for activation and catalysis to proceed. Recently, the crystal structure of Rubisco activase has been solved, which will provide key insight into the molecular interaction between Rubisco and Rubisco activase (reviewed by Portis et al. 2008).

Rubisco first evolved when the earth’s atmosphere was rich in CO₂, but virtually devoid of O₂. With the advent of oxygen-producing photosynthesis by land plants, and the resulting increases in atmospheric O₂, one key deficiency of this enzyme became apparent. Rubisco would not only catalyse fixation of CO₂ but would also permit incorporation of O₂ into RuBP to produce, instead of two molecules of 3-PGA, just one molecule of 3-PGA with one molecule of a two-carbon compound, 2-phosphoglycolate (Section 2.3). Indeed, CO₂ and O₂ compete directly for access to the active sites of Rubisco. So feeble is Rubisco’s ability to distinguish between these two substrates that in air (20% O₂) approximately one molecule of O₂ is fixed for every three molecules of CO₂.

Fixation of O₂ and subsequent photorespiration (Section 2.3) is an energy-consuming process, due to competition between O₂ and CO₂ for RuBP, plus the energy cost of converting the phosphoglycolate product to a form which can be recycled in the PCR cycle. This energy cost is increased at higher temperatures because O₂ competes more effectively with CO₂ at the active site of Rubisco. Such sensitivity to temperature × O₂ explains why CO₂ enrichment, which reduces photorespiration, has a proportionally larger effect upon net carbon gain at higher temperatures than at lower temperatures (Section 13.3).

Notwithstanding a meagre catalytic effectiveness in present day Rubisco, more efficient variants would still have had a selective advantage, and especially during those times in the earth’s geological history when atmospheric CO₂ concentration was decreasing. Indeed there has been some improvement (Figure 2.3) such that specificity towards CO₂ as opposed to O₂ has improved significantly. Recently evolved angiosperms show a relative specificity almost twice that of 'older' organisms such as photosynthetic bacteria.

Despite such improvement, Rubisco remains seemingly maladapted to its cardinal role in global carbon uptake, and in response to selection pressure for more efficient variants of CO₂ assimilation, vascular plants have evolved with photosynthetic mechanisms that alleviate an inefficient Rubisco. One key feature of such devices is a mechanism to increase CO₂ concentration at active sites within photosynthetic tissues. Some of these photosynthetic pathways are dealt with below.

One hundred million years ago (Mid-Cretaceous), atmospheric CO₂ was between 1500 and 3000 µL L^–1, or four to ten times post-industrial levels. Atmospheric CO₂ declined during the Oligocene (20-30 million years ago) from the high Tertiary levels (>1000 µL L^-1), and oscillated between 180 and 300 µL L^-1 for the last 1-3 million years. The Oligocene was also a time when the Earth was dry and the tropics were relatively hot. The earliest origins of C₄ photosynthesis date back to this period. Curiously, C₄ plants remained in low abundance for a long period of time. According to stable carbon isotopic data, a worldwide expansion of C₄ grasslands and savannas occurred during the Late Miocene and Pliocene (3 to 8 million years ago), most probably through the displacement of C₃ vegetation (Edwards et al. 2010).

Under the early high concentration of CO₂, photorespiration of C₃ plants was inhibited (Section 2.3) so that photosynthetic efficiency was higher than it is now. In addition, maximum photosynthetic rates were double twentieth century values, and the energy cost of photosynthesis would have been around three ATP and two NADPH per molecule of CO₂ fixed. As atmospheric CO₂ concentrations declined to approximately 250–300 µL L^–1, photosynthetic rates were halved, photorespiration increased substantially, photosynthetic efficiency declined and the energetic costs of photosynthesis increased to approximately five ATP and 3.2 NADPH per CO₂ molecule fixed. Such events would have generated a strong selection pressure for genetic variants with increased carboxylation efficiency and increased photosynthetic rates.

Angiosperms have a higher relative specificity of Rubisco for CO₂ than ferns and mosses (see examples of other less evolutionarily advanced species in Figure 2.3). Such differences imply minor evolution in this highly conserved molecule of Rubisco and there is little variation between species of vascular plants. Consequently, alteration of Rubisco in response to a changing atmospheric CO₂ concentration has not been an option.

By contrast, evolution of a new photosynthetic pathway (C₄) has occurred independently and on many occasions in diverse taxa over 25 to 30 million years as CO₂ levels declined. Despite its complexity, C₄ photosynthesis evolved more than 60 independent times in 19 distantly related flowering families. About 50% of C₄ species are grasses (Poaceae) with ~18 distinct origins distributed over 370 genera and ~4600 species (Sage et al. 2011). The oldest identifiable fossils with pronounced bundle sheath layers are seven million years old, although necessary metabolic pathways could have evolved earlier, prior to this adaptation in anatomy. C₄ plants are known to differ from C₃ plants in their discrimination against atmospheric ¹³CO₂, and shifts in the stable carbon isotope signature of soil carbonate layers that reflect emergence of C₄ plants have been dated at 7.5 million years bp. Modern evidence from molecular phylogeny places the origin of the main C₄ taxa at 25-30 million years ago (Christin et al. 2009). By inference, C₄ photosynthesis evolved in response to a significant decline in atmospheric CO₂ concentration, from 1500–3000 µL L^–1 to about 300 µL L^–1. By evolving a CO₂-concentrating mechanism, C₄ plants presented their Rubisco with an elevated partial pressure of CO₂ despite lower atmospheric CO₂. As a consequence, photorespiration was inhibited, maximum photosynthetic rates increased and energetic costs reduced.

The C₄ pathway (Figure 2.3) is ‘a unique blend of modified biochemistry, anatomy and ultra-structure’ (Hatch 1987). The classical C₄ syndrome in most terrestrial plants consists of two photosynthetic cycles (C₃ (or PCR) and C₄) operating across two photosynthetic cell types (mesophyll and bundle sheath), which are arranged in concentric layers around the vascular bundle, also known as the kranz anatomy.

Initial and rapid fixation of CO₂ within mesophyll cells results in the formation of a four-carbon compound which is then pumped to bundle sheath cells for decarboxylation and subsequent incorporation into the PCR cycle in that tissue. This neat division of labour hinges on specialised anatomy and has even resulted in evolution of distinct classes of chloroplasts in mesophyll compared with bundle sheath cells. Three biochemical variants of C₄ photosynthesis (termed subtypes) are known to have evolved from C₃ progenitor and in all cases with a recurring theme where the C₄ cycle of mesophyll cells is complemented by a PCR cycle in bundle sheath cells, where Rubisco is exclusively localised. In effect, a biochemical ‘pump’ concentrates CO₂ at Rubisco sites in bundle sheath cells thereby sustaining faster net rates of CO₂ incorporation and virtually eliminating photorespiration. For this overall mechanism to have evolved, a complex combination of cell specialisation and differential gene expression was necessary. Figure 2.3a shows a low-magnification electron micrograph of a C₄ leaf related to a generalised scheme for the C₄ pathway.

By analogy with Calvin’s biochemical definition of the C₃ pathway at Berkeley in the 1950s, the C₄ pathway was also delineated with radioactively labelled CO₂ (see Feature Essay 2.1). Significantly, and unlike C₃ plants, 3-PGA is not the first compound to be labelled after a ¹⁴C pulse (Figure 2.3b). Specialised mesophyll cells carry out the initial steps of CO₂ fixation utilising the enzyme phosphoenolpyruvate (PEP) carboxylase. The product of CO₂ fixation, oxaloacetate, is a four-carbon organic acid, hence the designation ‘C₄’ photosynthesis (or colloquially, C₄ plant). A form of this four-carbon acid, either malate or aspartate depending on the C₄ subtype, migrates to the bundle sheath cells which contain Rubisco and the PCR cycle. In the bundle sheath cells, CO₂ is removed from the four-carbon acid by a specific decarboxylase and a three-carbon product returns to the mesophyll to be recycled to PEP for the carboxylation reaction. Thus, label first appears in the four-carbon acid after ¹⁴C feeding, followed by 3-PGA and, finally, in sucrose and starch (Figure 2.1b).

A physical barrier to CO₂ diffusion exists in the thickened walls of the bundle sheath cells (lined with suberised in some C₄ species), preventing CO₂ diffusion back to the mesophyll and allowing CO₂ build up to levels at least 10 times those of ambient air. Build up of CO₂ in the bundle sheath is also facilitated by the higher activity ratio (2-4 times) of PEP carboxylase relative to Rubisco in C₄ plants. Rubisco is thus exposed to a saturating concentration of CO₂ which both enhances carboxylation due to increased substrate supply, and forestalls oxygenation of RuBP (hence photorespiration) by outcompeting O₂ for CO₂ binding sites on Rubisco (Figure 2.5).

In leaves of C₃ plants, the PCR cycle operates in all mesophyll chloroplasts, but in C₄ plants the PCR cycle is restricted to bundle sheath cells (Figure 2.3). Rubisco is pivotal in this cycle, and can be used as a marker for sites of photosynthetic carbon reduction. Rubisco was visualised by localising this photosynthetic enzyme with antibodies via indirect immunofluorescent labelling (Hattersley et al. 1977; Figure 2.6). In this pioneering method, ‘primary’ rabbit anti-Rubisco serum (from rabbits injected with purified Rubisco) is first applied to fixed transverse sections of leaves. Rabbit antibodies to Rubisco bind to the enzyme in situ. Then ‘secondary’ sheep anti-rabbit immunoglobulin tagged with a fluorochrome (fluorescein isothiocyanate) is applied to the preparation. This fluorochrome binds specifically to the rabbit antibodies and fluoresces bright yellow wherever Rubisco is located (blue light excitation using an epifluorescence light microscope).

In the C₃ grass Microlaena stipoides (Figure 2.6a), all chloroplasts are fluorescing bright yellow and this indicates wide distribution of Rubisco throughout mesophyll tissue. By contrast, only bundle sheath cells are equipped with Rubisco in the C₄ grass Digitaria brownii (Figure 2.6b). These two native Australian grasses co-occur in the ACT but contrast in relative abundance. M. stipoides (weeping grass) is common in dry sclerophyll woodlands throughout southeast temperate Australia, whereas D. brownii (cotton panic grass) in the ACT is at the southern end of its distribution, being far more abundant in subtropical Australia and, in keeping with its C₄ physiology, especially prevalent in semi-arid regions.

One disadvantage of the C₄ pathway is that an energy cost is incurred by C₄ plants to run the CO₂ ‘pump’. This is due to the ATP required for recycling PEP from pyruvate by the chloroplastic enzyme pyruvate, Pi dikinase in the mesophyll cells (Figure 2.4 and Hatch 1987). Under ideal conditions, five ATP and two NADPH are required for every CO₂ fixed in C₄ photosynthesis (two ATP are required to run the CO₂ pump, i.e., regenerate PEP). In addition, a proportion (20-30%) of CO₂ fixed by PEP carboxylase in the mesophyll is not fixed by Rubisco in the bundle sheath, and subsequently leaks back to the mesophyll. This leaked (or overcycled) CO₂ represents an additional, inherent energetic cost of the C₄ pathway.

From the previous section, the C₄ pathway is obviously energetically more expensive than the C₃ pathway in the absence of photorespiration. However, at higher temperatures the ratio of RuBP oxygenation to carboxylation is increased and the energy requirements of C₃ photosynthesis can rise to more than five ATP and three NADPH per CO₂ fixed in air (for these calculations see Hatch 1987).

Representative light response curves for photosynthesis in C₃ cf. C₄ plants (Figure 2.7) can be used to demonstrate some of these inherent differences in photosynthetic attributes. At low temperature (10°C in Figure 2.7) a C₃ leaf shows a steeper initial slope as well as a higher value for light-saturated photosynthesis. By implication, quantum yield is higher and photosynthetic capacity is greater under cool conditions. In terms of carbon gain and hence competitive ability, C₃ plants will thus have an advantage over C₄ plants at low temperature and especially under low light.

By contrast, under warm conditions (35°C, upper curves in Figure 2.7) C₄ photosynthesis in full sun greatly exceeds that of C₃, while quantum yield (inferred from initial slopes) remains unaffected by temperature. Significantly, C₃ plants show a reduction in quantum yield under warm conditions (compare 10°C and 35°C curves; right side of Figure 2.7). At 35°C C₃ plants also show lower rates of light-saturated assimilation compared with C₄ plants. Increased photorespiratory losses from C₃ leaves at high temperature are responsible (Section 2.3). C₄ plants will thus have a competitive advantage over C₃ plants under warm conditions at both high and low irradiance.

C₄ photosynthesis calls for metabolic compartmentation which is in turn linked to specialised anatomy (Figure 2.4). Three biochemical subtypes of C₄ photosynthesis have evolved which probably derive from subtle differences in the original physiology and leaf anatomy of their C₃ progenitors.

CO₂ assimilation by all three C₄ subtypes (Figure 2.8) involves five stages:

1. carboxylation of PEP in mesophyll cells, thereby generating four-carbon acids (malate and/or aspartate);
2. transport of four-carbon acids to bundle sheath cells;
3. decarboxylation of four-carbon acids to liberate CO₂;
4. re-fixation of this CO₂ via Rubisco within the bundle sheath, using the C₃ pathway;
5. transport of three-carbon acid products following decarboxylation back to mesophyll cells to enable synthesis of more PEP.

Recognising some systematic distinctions in whether malate or aspartate was transported to bundle sheath cells, C₄ plants were further subdivided into three subtypes according to their four-carbon acid decarboxylating systems and ultrastructural features (Hatch et al. 1975). Members of each subtype contain high levels of either NADP-malic enzyme (NADP-ME), phosphoenolpyruvate carboxykinase (PCK) or NAD-malic enzyme (NAD-ME) (so designated in Figure 2.8). High NADP-malic enzyme activity is always associated with higher NADP-malate dehydrogenase activity, while those species featuring high activities of either of the other two decarboxylases always contain high levels of aminotransferase and alanine aminotransferase activities. As a further distinction, each of the decarboxylating enzymes is located in bundle sheath cells; NAD-malic enzyme is located in mitochondria but PEP carboxykinase is not.

In all three subtypes, the primary carboxylation event occurs in mesophyll cytoplasm with PEP carboxylase acting on HCO₃^– to form oxaloacetate. However, the fate of this oxaloacetate varies according to subtype (Table 2.1; Figure 2.8). In NADP-ME species, oxaloacetate is quickly reduced to malate in mesophyll chloroplasts using NADPH. By contrast, in NAD-ME and PCK species, oxaloacetate is transaminated in the cytoplasm, with glutamate donating the amino group, to generate aspartate. Thus, malate is transferred to bundle sheath cells in NADP-ME species and aspartate is transferred in NAD-ME and PCK species. The chemical identity of three-carbon acids returned to mesophyll cells varies accordingly.

In NADP-ME species, only chloroplasts are involved in decarboxylation and subsequent carboxylation via the PCR cycle (Figure 2.8). By contrast, in NAD-ME and PCK species, chloroplasts, cytoplasm and mitochondria are all involved in moving carbon to the PCR cycle of bundle sheath chloroplasts. In NAD-ME and PCK species, aspartate arriving in bundle sheath cells is reconverted to oxaloacetate in either mitochondria (NAD-ME) or cytoplasm (PCK) (Table 2.2). Reduction and decarboxylation of oxaloacetate occurs in mitochondria of NAD-ME species and CO₂ is thereby released for fixation by chloroplasts of bundle sheath cells. In PCK species, oxaloacetate in the cytoplasm is decarboxylated by PCK, thereby releasing CO₂ for fixation in bundle sheath chloroplasts (Figure 2.8).

Transport of metabolites to bundle sheath cells

A rapid transfer of malate and aspartate to bundle sheath cells from mesophyll cells is required if the CO₂ concentration in bundle sheath cells is to stay high. A very high density of plasmodesmata linking bundle sheath cells to mesophyll cells facilitates this traffic. Consequently, the permeability coefficient of C₄ bundle sheath cells to small metabolites such as four-carbon acids is about 10 times larger than that of C₃ mesophyll cells (Table 2.2). However, coupled with this need for a high permeability to metabolites moving into bundle sheath cells is a low permeability to CO₂ molecules so that CO₂ released through decarboxylation in the bundle sheath does not diffuse rapidly into mesophyll air spaces. For some species, a layer of suberin in the cell wall of bundle sheath–mesophyll junctions (suberin lamella) significantly reduces CO₂ efflux (Table 2.2).

Centrifugal versus centripetal chloroplasts

Not all species contain a suberin layer, but all C₄ plants have a need to prevent CO₂ from diffusing quickly out of bundle sheath cells, so that the location of chloroplasts of bundle sheath cells becomes critical in those species lacking a suberin layer (Figure 2.9). Where species have a suberin layer, chloroplasts are located in a centrifugal position, that is, on the wall furtherest away from the centre of the vascular bundle lying in the middle of the bundle sheath (Figure 2.9E, F). In those C₄ species lacking a suberin layer, chloroplasts are located centripetally, that is, on the wall closest to the centre of the vascular bundle lying within the bundle sheath (Figure 2.9A, B). Such a location would help restrict CO₂ diffusion from bundle sheath to mesophyll cells.

Regulation of C₄ photosynthesis

Fixation of CO₂ by C₄ plants involves the coordinated activity of two cycles in separate anatomical compartments (Figure 2.8). The first cycle is C₄ (carboxylation by PEP carboxylase), the second is C₃ (carboxylation by Rubisco). Given this biochemical and anatomical complexity, close regulation of enzyme activities is a prerequisite for efficient coordination.

PEP carboxylase, NADP-malate dehydrogenase and pyruvate orthophosphate dikinase are all light-regulated and their activities vary according to irradiance. NADP-malate dehydrogenase is regulated indirectly by light via the thioredoxin system.

PEP carboxylase in C₄ plants exists in the same homo-tetramer in light- and dark-acclimated leaves. This is in marked contrast to CAM species where different forms exist in light- and dark-acclimated leaves. In C₄ plants, PEP carboxylase has extremely low activity at night, thus preventing uncontrolled consumption of PEP. Such complete loss of activity in darkness is mediated via divalent metal ions, pH plus allosteric activators and inhibitors. As a consequence, and over a period of days, C₄ plants can increase or decrease PEP carboxylase in response to light regime.

Rubisco is characterised by its low affinity for its productive substrate, CO₂ and slow catalytic turnover rate (i.e., 1-3 cycles per sec). Importantly, Rubisco reacts with O₂ (photorespiration), and this culminates in loss of CO₂ and energy. In C₃ plants, photorespiration can drain more than 25% of fixed CO₂ under non-stressful conditions. The ratio of photorespiration to photosynthesis increases with increasing temperature and decreasing intercellular CO₂ such as occurs when stomatal conductance is reduced under water stress. C₃ plants compensate for Rubisco’s inefficiencies by (i) opening their stomata to increase CO₂ diffusion into chloroplasts, which increases water loss and lowers leaf-level water use efficiency, WUE; and (ii) investing up to 50% of leaf nitrogen in Rubisco, which lowers their leaf-level nitrogen use efficiency, NUE.

The C₄ pathway supercharges photosynthesis and suppresses photorespiration by operating a CO₂ concentrating mechanism which elevates CO₂ around Rubisco. Although C₄ photosynthesis incurs additional energy, the energy cost of photorespiration exceeds that of the CO₂ concentrating mechanism above 25^oC. Hence, higher radiation use efficiencies (i.e., efficiency of converting absorbed radiation into biomass) have been recorded for C₄ than C₃ crops. High bundle sheath CO₂ concentration saturates C₄ photosynthesis at relatively low intercellular CO₂, allowing C₄ plants to operate with lower stomatal conductance. Thus, leaf-level WUE is usually higher in C₄ than C₃ plants. Relative to C₃ plants, Rubisco of C₄ plants is faster (higher turnover rate) and operates under saturating CO₂. Thus, C₄ plants typically achieve higher photosynthetic rates with about 50% less Rubisco and less leaf nitrogen. Hence, photosynthetic NUE is higher in C₄ than C₃ plants. Accordingly, C₄ plants are advantaged relative to C₃ plants in hot and nitrogen-poor environments with short growing seasons, hence their great abundance in wet/dry tropics such as Northern Territory savannas.

As mentioned earlier, more than 50% of C₄ plants are grasses. C₄ grasses are confined to low latitudes and altitudes, whereas C₃ species dominate at higher latitudes and altitudes. Generally, C₄ species frequently occur in regions of strong irradiance. Ehleringer and colleagues (Ehleringer et al. 1997) proposed that these distribution patterns are best explained by the different responses of photosynthetic quantum yields to temperature between C₃ and C₄ plants.

C₄ photosynthesis suppresses photorespiration by operating a CO₂ concentrating mechanism that comes at additional energetic cost. This cost is independent of ambient CO₂ and temperatures. In contrast, photorespiration (and its associated energy cost) increases steeply with temperature in C₃ plants and is highly dependent on CO₂ concentrations. Under saturating irradiance and current ambient atmospheric CO₂ concentration, the threshold temperature where the cost of photorespiration in C₃ plants exceeds that of the CO₂ concentrating mechanism in C₄ plants is estimated around 25^oC. This model provides a physiological basis for understanding today’s contrasting geographic distribution between C₃ and C₄ grasses.

As an example, the C₄ grasses of the northern Australian savannas are relatively un-shaded because of the low tree density and sparse canopy. Light is abundant and since the CO₂ concentration inside C₄ leaves is high, a potentially high rate of light-saturated assimilation can be exploited. Most C₃ species reach light saturation in the range of one-eight to one-half full sunlight (Figure 2.7). In C₄ species, canopy assimilation might not become light saturated even in full sunlight. C₄ plants thus maintain a competitive advantage over C₃ plants in tropical locations, where average daily light receipt is much larger than in temperate zones, and associated with warmer conditions that also favour C₄ photosynthesis (Figure 2.6). Given strong sunlight, warmth and seasonally abundant water, biomass production by C₄ plants is commonly double the rate for C₃ plants. Typically, C₃ plants produce 15–25 t ha^–1 but C₄ plants easily produce 35–45 t ha^–1.

Physiological characteristics of the C₄ subtypes

As outlined in previous sections, characteristic biochemical, anatomical and physiological traits are associated with each of the three “classical” C₄ subtypes (Table 2.3). However, it should be noted that many C₄ plants have leaf structures that fall outside the “classical” subtype division (eg, NADP-ME tribes, Arundinelleae and Neurachneae). As many as 11 anatomical-biochemical suites have been identified in C₄ grasses. A curious aspect about the subtypes of C₄ grasses is their biogeography. In Australia and elsewhere, NADP-ME grasses are more frequent at higher rainfall, NAD-ME grasses predominate at lower rainfall, while the distribution of PCK grasses is even across rainfall gradients (Hattersley 1992).

The fundamental paradigm underpinning the efficiency of C₄ photosynthesis in terrestrial plants is the ‘division of labour’ between the initial fixation of CO₂ into C₄ acids, and their subsequent utilisation to generate high concentrations of CO₂ for ultimate fixation by Rubisco. The basic model for C₄ plants with classical kranz anatomy consists of two photosynthetic cycles (C₃ and C₄) operating across two photosynthetic cell types (mesophyll and bundle sheath), with strict cell- and organelle-specific localisation of key enzymes and with sufficient resistance to CO₂ back-diffusion. Indeed, the discovery of the kranz anatomy by Haberlandt preceded that of C₄ biochemistry by a century. The prevailing consensus has been that efficient C₄ photosynthesis necessitates the collaboration of two cell types.

Recently, this notion has been challenged by the discovery of non-kranz or single-cell C₄ photosynthesis in shrubs (Borszczowia aralocaspica and Bienertia cycloptera; Chenopodiaceae family) found in the salt deserts of Central Asia (Voznesenskaya et al. 2002, 2003). These plants show CO₂ and O₂ responses typical of C₄ photosynthesis but lack the kranz anatomy. They perform C₄ photosynthesis through the spatial localisation of dimorphic chloroplasts (as well as other organelles and photosynthetic enzymes) in distinct positions within a single chlorenchyma cell. Yet, the details of the partitioning differ between the two species (Edwards et al. 2004).

In Bienertia, the central cytoplasmic compartment of the chlorenchyma cell plays the role of bundle sheath cells in kranz-type C₄ (NAD-ME) plants; it is filled with mitochondria surrounded by chloroplasts. The peripheral cytoplasm lacks mitochondria and plays the role of the mesophyll cell in kranz-type C₄ plants. Accordingly, chloroplastic Rubisco and mitochondrial NAD-ME and glycine decarboxylase are restricted to the central compartment; chloroplastic pyruvate, Pi dikinase is restricted to the peripheral compartment, which is highly enriched with cytosolic PEP carboxylase. In Borszczowia, the compartmentation occurs at the distal (mesophyll equivalent) and proximal (bundle sheath equivalent) ends of the elongated, cylindrical chlorenchyma cell. The inter-connecting cytoplasm between the two intra-cellular compartments provides a liquid diffusion path, thus replacing the role of the bundle sheath cell wall in kranz-type C₄ plants (Edwards et al. 2004).

A low conductance for CO₂ diffusion out of the bundle sheath cells (or its equivalent cellular compartment) is critical for the efficient operation of C₄ photosynthesis. The total diffusive resistance to CO₂ has multiple components with different levels of contribution. These components include bundle sheath walls, membranes, bundle sheath chloroplast position, the site of C₄ acid decarboxylation, and the liquid-phase diffusion path. For kranz-type C₄ plants, calculated total bundle sheath resistance on a leaf area basis can range from 50 to 150 m² s^-1 mol^-1 (von Caemmerer and Furbank 2003). Evidently, single-cell C₄ plants have sufficient resistance to CO₂ back-diffusion which is essentially made of the cytoplasmic liquid phase and the special localisation of the (Rubisco-containing) chloroplasts surrounding the mitochondria (site of C₄ acid decarboxylation).

Thus, single-cell C₄ plants have efficient photosynthesis which is not inhibited by O₂, and their carbon isotope values are similar to kranz-type C₄ plants. Although, single-cell C₄ photosynthesis breaks away from the classical kranz anatomy, it remains within the general ‘division of labour’ paradigm.

More than 40 eudicot and monocot species distributed over 21 lineages have been reported to possess intermediate C₃ and C₄ photosynthetic characteristics and CO₂ compensation points. These intermediate species are likely remnants of the complex processes that led to the evolution of C₄ plants from C₃ ancestors, although reversions from the C₄ condition have also been suggested. Moreover, a number of identified C₃-C₄ species occur in taxa that are not closely related to any C₄ lineage, raising the possibility that the C₃-C₄ photosynthetic pathway may be a distinct adaptation. This, in addition to the small number of intermediate species found so far cast doubts over their physiological and ecological fitness, and whether they represent living fossils of evolutionary paths or evolutionary dead-ends (Rawsthorne 1992; Sage et al. 2011).

Leaves of all C₃-C₄ intermediates have partial or full kranz anatomy, with prominent bundle sheath cells containing chloroplasts and other organelles, and intermediate interveinal distances. Bundle sheath chloroplasts contain Rubisco and functional PCR cycle in both mesophyll and bundle sheath cells. Intermediate leaves also have CO₂ compensation points that are lower than what is observed for C₃ leaves and can be indistinguishable from C₄ leaves, due to reduced photorespiration. Biochemically, C₃-C₄ intermediates differ in the level of activity of the C₄ cycle and the extent to which CO₂ is concentrated in bundle sheath cells.

C₃-C₄ intermediate plants reduce photorespiration (and hence, CO₂ compensation point) using a ‘photorespiratory pump’ based on modified localisation of the mitochondrial photorespiratory enzyme, glycine decarboxylase (Figure 2.10). In these plants, glycine decarboxylase activity is restricted to bundle sheath cells and excluded from mesophyll cells. Consequently, photorespired CO₂ is released in the bundle sheath where it is largely refixed by Rubisco and the bundle sheath PCR before it diffuses back to the mesophyll. Such a system may weakly elevate CO₂ in bundle sheath cells. Intermediate species that rely on the ‘photorespiratory pump’ are termed C₃-like or Type I intermediates (e.g., Panicum milioides, Flaveria pubescens) and have intermediate CO₂ compensation points and negligible C₄ cycle activity.

In Type II and C₄-like intermediates (e.g., Flaveria brownii), up to 70% of atmospheric CO₂ may be first fixed into C₄ acids. These plants have C₄-like CO₂ compensation points but are not classified as C₄ plants because they lack the strict localisation of photosynthetic enzymes (e.g., Rubisco is present in mesophyll cells) and their bundle sheath cell walls have high CO₂ permeability, resulting in only a partial CO₂ concentrating mechanism (Brown 1980; Ku et al. 1991; Rawthorne 1992; Vogan and Sage 2011).

The physiological advantages of the intermediate photosynthetic pathway in all its naturally occurring forms remain unclear. It may be hypothesised that lowered photorespiration may lead to reduced CO₂ limitation of photosynthesis, and thus allow C₃-C₄ plants to operate with lower stomatal conductance, thus conferring higher water use efficiency relative to C₃ counterparts. Moreover, increased nitrogen cost associated with ‘building’ another set of photosynthetic cells (bundle sheath) may reduce nitrogen use efficiency if the gains in CO₂ uptake are not substantial.

Work conducted with C₃-C₄ species yielded inconclusive evidence on the likely advantages of C₃-C₄ photosynthesis relative to the ancestral C₃ mode. Generally, these studies demonstrated that, short of substantial C₄ cycle activity and advanced cell-specific localisation of C₃ and C₄ cycle enzymes between the mesophyll and bundle sheath cells, C₃-C₄ photosynthesis does not improve photosynthetic efficiency (Bolton and Brown 1980; Pinto et al. 2011, Vogan and Sage 2011). Therefore, partial recycling of photorespired CO₂ or a partial CO₂ concentrating mechanism reduce photorespiratory loss normally associated with C₃ photosynthesis, without leading to significant gains in plant fitness or productivity.

Joseph Holtum¹, Klaus Winter² and Barry Osmond³

¹Centre for Tropical Biodiversity and Climate Change, James Cook University, Australia; ²Smithsonian Tropical Research Institute, Balboa, Ancón, Republic of Panama; ³School of Biological Sciences, University of Wollongong, and Research School of Biology, Australian National University, Australia

Crassulacean acid metabolism (CAM) is a water-conserving mode of photosynthesis that, like C₄ photosynthesis, is a modification of the C₃ photosynthetic pathway fitted with a CO₂ concentrating mechanism (CCM) that can increase the [CO₂] around ribulose bisphosphate carboxylase/oxygenase (Rubisco) by more than 10-fold and suppress photorespiration. The overall energy demand of the CAM pathway is only about 10% more than that of C₃ photosynthesis, as costs of the CCM machinery are partially offset by reducing photorespiration.

In C₄ plants, as explained earlier in Section 2.2.2, this CCM is most commonly achieved by an “in-line turbocharger” based on initial CO₂ fixation by phosphoenolpyruvate carboxylase (PEPC) into C₄ acids in the cytoplasm of outer mesophyll cells. These acids diffuse rapidly to adjacent relatively CO₂-tight bundle-sheath cells (Figure 2.31 right) where CO₂ is released again. High [CO₂] builds up in this spatially separated compartment where it is refixed by Rubisco.

In CAM plants enzyme systems analogous to those in C₄ plants achieve the same result through a “battery-like” dark accumulation of CO₂ into the 2^nd carboxyl group of malic acid (acidification phase) in the vacuole of large mesophyll cells (Figure 2.31 left). Malic acid can accumulate to very high concentrations, attaining concentrations of greater than 1 mole acid per litre in mesophyll cells of tropical tree-CAM plants (Clusia spp.). Indeed one can sometimes taste the acid, with acid taste-testing for the presence of CAM being possibly first recorded in Aloe sp. by Nehemiah Grew in 1682 and in field reports from India by Benjamin Heyne in 1815.

In the light, malic acid returns to the cytoplasm where it is rapidly decarboxylated (deacidification phase). The CO₂ released, which accumulates to high internal [CO₂] as stomata close, is refixed by Rubisco in chloroplasts of the same mesophyll cell where it is further assimilated by the photosynthetic carbon reduction (PCR) cycle (Figure 2.32).  

Ultimately three of the four carbons recovered from the malic acid must be stored as starch and/or sugars in order to provide to PEPC the C₃ substrate required for CO₂ uptake during the following night. The fourth carbon, effectively that obtained from the atmosphere, is available for growth. Deacidification may generate high [CO₂] behind closed stomata but photorespiration is not completely abolished (Lüttge 2002) since photosynthesis also generates high internal [O₂]. While exploring Lake Valencia in Venezuela in 1800, Alexander von Humboldt measured elevated [O₂] in bubbles streaming from the cut base of presumably CAM Clusia leaves standing in water in the light.

Of course by closing stomata in the light CAM plants minimize water loss when evaporative demand is highest (von Caemmerer and Griffiths 2009). The biomass production per unit water utilized in CAM was 6 times higher than for C₃ plants and 2 times higher than for C₄ plants when plants exhibiting all three photosynthetic pathways were grown together in a garden outdoors (Winter et al. 2005). The attribute of water-use efficiency undoubtedly contributes significantly to the success of CAM photosynthesis in nature, with CAM species outnumbering C₄ species by about two to one. The paradoxes of CAM, a mode of photosynthesis that involves stomatal opening and CO₂ uptake during the dark, continue to inform many aspects of plant biochemistry, physiology, ecology and evolution. This article draws heavily on two recent reviews (Borland et al. 2011; Winter et al. 2015).

Vascular plants often inhabit regions subject to tidal submergence while others carry out their entire life cycle under water. Examples of common submerged aquatic macrophytes are pond weeds and seagrasses. Once again, an evolutionary selective pressure for these plants has been the availability of CO₂. Low levels of dissolved CO₂ are common in both inland and marine waters, particularly at more alkaline pH. In more productive inland lakes, CO₂ content can vary enormously, requiring considerable flexibility in the actual mode of carbon acquisition. At high pH, HCO₃^– becomes the more abundant form of inorganic carbon, whereas dissolved CO₂ will predominate at low pH (Section 18.2). Consequently, when SAM plants evolved from their C₃ progenitors on land, there was some adaptive advantage in devices for CO₂ accumulation because CO₂ rather than HCO₃^– is substrate for Rubisco. The nature of this ‘CO₂ pump’ and the energetics of carbon assimilation are not fully characterised in SAM plants but considerable CO₂ concentrations do build up within leaves, enhancing assimilation and suppressing photorespiration.

In summary:

Regardless of photosynthetic mode, and despite catalytic limitations, Rubisco is ubiquitous and remains pivotal to carbon gain in our biosphere. As a corollary, carbon loss via photorespiration is an equally universal feature of C₃ leaves, and the evolution of devices that overcome such losses have conferred significant adaptive advantages to C₄, CAM and SAM plants.

 

Bolton JK, Brown RH (1980) Photosynthesis of grass species differing in carbon dioxide fixation pathways. V. Response of Panicum maximum, Panicum milioides, and tall fescue (Festuca arundinacea) to nitrogen nutrition. Plant Physiol 66: 97-100

Brown DA (1980). Photosynthesis of grass species differing in carbon dioxide fixation pathways. IV. Analysis of reduced oxygen response in Panicum milioides and Panicum schenckii. Plant Physiol 65: 346-349

Christin PA, Samaritani E, Petitpierre B et al. (2009) Evolutionary insights on C₄ photosynthetic subtypes in grasses from genomics and phylogenetics. Genom Biol Evol 1: 221–230

Edwards GE, Franceschi VR, Voznesenskaya EV (2004) Single-cell C₄ photosynthesis versus the dual-cell (Kranz) paradigm. Annu Rev Plant Biol 55: 173–196

Ehleringer JR, Cerling TE, Helliker BR (1997) C₄ photosynthesis, atmospheric CO₂, and climate. Oecologia 112: 285-299

Ghannoum O, von Caemmerer S, Conroy JP (2002) The effect of drought on plant water use efficiency of nine NAD–ME and nine NADP–ME Australian C₄ grasses. Funct Plant Biol 29: 1337-1348

Ghannoum O, Evans JR, Chow WS et al. (2005) Faster rubisco is the key to superior nitrogen-use efficiency in NADP-malic enzyme relative to NAD-malic enzyme C₄ grasses. Plant Physiology 137: 638-650

Hatch MD (1987) C₄ photosynthesis: a unique blend of modified biochemistry, anatomy and ultrastructure. Biochim Biophys Acta 895: 81–106

Hatch MD, Kagawa T, Craig S (1975) Subdivision of C₄-pathway species based on differing C₄ acid decarboxylating systems and ultrastructural features. Aust J Plant Physiol 2: 111-128.

Hattersley PW, Watson L, Osmond CB (1977) In situ immunofluorescent labelling of ribulose-1,5-bisphosphate carboxylase in leaves of C₃ and C₄ plants. Aust J Plant Physiol 4: 523-539

Hattersley PW (1992) In ‘Desertified Grasslands: their Biology and Management’ (ed. Chapman GP) pp 181-212. Academic Press: London

Ku MSB, Wu JR, Dai ZY et al. (1991) Photosynthetic and photorespiratory characteristics of Flaveria species. Plant Physiol 96: 518-528

Pinto H, Tissue DT, Ghannoum O (2011) Panicum milioides (C₃-C₄) does not have improved water or nitrogen economies relative to C₃ and C₄ congeners exposed to industrial-age climate change. J Exp Bot 62: 3223-3234

Portis AR Jr, Salvucci ME (2002). The discovery of Rubisco activase – yet another story of serendipity. Photosyn Res 73: 257–264

Portis AR Jr, Li C, Wang D, Salvucci ME (2008) Regulation of Rubisco activase and its interaction with Rubisco. J Exp Bot 59: 1597-1604

Rawsthorne S (1992) C₃-C₄ intermediate photosynthesis - Linking physiology to gene expression. Plant J 2: 267-274

Sage RF (2004). The evolution of C₄ photosynthesis. New Phytol 161: 341-370

Sage RF, Christin PA, Edwards EJ (2011) The C₄ plant lineages of planet Earth. J Exp Bot 62, 3155-3169

Spreitzer RJ, Salvucci ME (2002) Rubisco: structure, regulatory interactions, and possibilities for a better enzyme. Annu Rev Plant Biol 53: 449-475

Vogan PJ, Sage RF (2011) Water-use efficiency and nitrogen-use efficiency of C₃-C₄ intermediate species of Flaveria. Plant Cell Environ 34: 1415–1430.

von Caemmerer S, Furbank RT (2003) The C₄ pathway: an efficient CO₂ pump. Photosyn Res 77: 191–207

Voznesenskaya EV, Franceschi VR, Kiirats O et al. (2002) Proof of C₄ photosynthesis without Kranz anatomy in Bienertia cycloptera. Plant J 31: 649–662

Voznesenskaya EV, Edwards GE, Kiirats O et al. (2003) Development of biochemical specialization and organelle partitioning in the single celled C₄ system in leaves of Borszczowia aralocaspica. Amer J Bot 90: 1669-1680

(a) Historical evidence for photorespiration

The first line of evidence for photorespiration came from the different compensation points of C3 and C4 plants. When air is recirculated over an illuminated leaf in a closed system, photosynthesis will reduce CO₂ concentration to a low level where fixation of CO₂ by photosynthesis is just offset by release from respiration. For many C₃ plants this compensation point is around 50 µL L^–1 but is markedly affected by oxygen, photon irradiance and leaf temperature (Tregunna et al. 1966; Zelitch 1966). In low concentrations (1–2% O₂) the CO₂ compensation point of C3 plants is near zero. Significantly, early researchers in this area had already noted that some tropical grass species appeared to have a compensation point at or close to zero CO₂, even in normal air (20% O₂). This was first reported for corn (Zea mays) (Meidner 1962) and raised a very perplexing question as to whether these species even respired in light. However, we now know that C₄ photosynthesis is responsible for the low evolution of CO₂ (Section 2.2) and that C₄ plants have a CO₂ concentrating mechanism that forestalls photorespiration, resulting in a CO₂ compensation point close to zero.

A second line of evidence for leaf respiration in light was provided by a transient increase in release of CO₂ when leaves are transferred from light to dark. This ‘post-illumination CO₂ burst’ was studied extensively during the early 1960s by Gleb Krotkov and colleagues at Queens University (Kingston, Ontario). The intensity of this burst increased with the photon irradiance during the preceding period of photosynthesis. Understandably, the Queens group regarded this post-illumination burst as a ‘remnant’ of respiratory processes in light, and coined the term ‘photorespiration’. A functional link with the CO₂ compensation point was inferred, because the burst was abolished in low O₂ (Figure 2.12). A competitive inhibition by O₂ on CO₂ assimilation was suspected and was subsequently proved to be particularly relevant in defining Rubisco’s properties. Nevertheless, for many years a biochemical explanation for interaction between these two gases remained elusive.

(b) Biochemistry

Significant progress came when Ludwig and Krotkov designed an open gas exchange system in which ¹⁴CO₂ was used to separate the fixing (photosynthetic) and evolving (respiratory) fluxes of CO₂ for an illuminated leaf (Ludwig and Canvin 1971). Results using this steady-rate labelling technique were particularly revealing and provided the first direct evidence that respiratory processes in light were qualitatively different from those in darkness. They were able to show that CO₂ evolved during normal high rates of photosynthesis by an attached sunflower leaf was derived from currently fixed carbon. The specific activity of evolved CO₂ (ratio of ¹⁴C to ¹²C) was essentially the same as that of the CO₂ being fixed, indicating that photorespiratory substrates were closely related to the initial products of fixation. Ludwig and Krotkov concluded that ¹⁴CO₂ supplied to a photosynthesising leaf was being re-evolved within 28–45 s! Furthermore the rate of CO₂ evolution in light was as much as three times the rate in darkness, and while early fixed products of photosynthesis (intermediates of the PCR cycle) were respired in light, this was not the case in darkness.

The radiolabelling method of Ludwig and Krotkov had, for the first time, provided measurements of what could be regarded as a true estimate of light-driven respiration which was not complicated by transient effects (as the post-illumination burst had been), or by changes in CO₂ concentration (as was the case for measurements in closed gas exchange systems or in CO₂-free air) or by difficulties associated with detached organs. Ludwig and Canvin (1971) subsequently concluded that processes underlying photorespiration re-evolved 25% of the CO₂ which was being fixed concurrently by photosynthesis. Such a rate of CO₂ loss was not a trivial process so a biochemical basis for its operation had to be established, and particularly when photorespiration seemed to be quite different from known mechanisms of dark (mitochondrial) respiration.

The search for the substrates of ‘photorespiration’ occupied many laboratories worldwide for many years. Much work centred on synthesis and oxidation of the two-carbon acid glycolate because as early as 1920 Warburg had reported that CO₂ fixation by illuminated Chlorella was inhibited by O₂, and under these conditions the alga excreted massive amounts of glycolic acid (Warburg and Krippahl 1960). Numerous reports on the nature of the ¹⁴C-labelled products of photosynthesis showed that glycolate was a prominent early-labelled product. A very wide variety of research with algae and leaves of many higher plants established two significant features of glycolate synthesis: formation was enhanced in either low CO₂ or high O₂. Both of these features had been predicted from Ludwig’s physiological gas exchange work and eventually proved a key to understanding the biochemistry of photorespiration.

(c) Source of glycolate production

The photosynthetic carbon reduction (PCR) cycle for CO₂ fixation (Section 2.1) involves an initial carboxylation of ribulose-1,5-bisphosphate (RuBP) to form 3-PGA, but makes no provision for glycolate synthesis. However, Wang and Waygood (1962) had described the ‘glycolate pathway’, namely a series of reactions in which glycolate is oxidised to glyoxylate and aminated, first to form glycine and subsequently the three-carbon amino acid serine. The intracellular location of this pathway in leaves was established in a series of elegant studies by Tolbert and his colleagues who also established that leaf microbodies (peroxisomes) were responsible for glycolate oxidation and the synthesis of glycine. Kisaki and Tolbert (1969) suggested that the yield of CO₂ from the condensation of two molecules of glycine to form serine could account for the CO₂ evolved in photorespiration. This idea was incorporated in later formulations of the pathway.

What remained elusive was the source of photosynthetically produced glycolate. Many studies had suggested that the sugar bisphosphates of the PCR cycle could yield a two-carbon fragment which, on the basis of short-term ¹⁴CO₂ fixation, would have its two carbon atoms uniformly labelled (if the two carbons were to be derived directly from 3-PGA this would not be the case as PGA was asymmetrically labelled in the carboxyl group). The mechanism was likened to the release of the active ‘glycolaldehyde’ transferred in the thiamine pyrophosphate (TPP)-linked transketolase-catalysed reactions of the PCR cycle. In some cases significant glycolate synthesis from the sugar bisphosphate intermediates of the cycle were demonstrated in vitro; however, the rates were typically too low to constitute a viable mechanism for glycolate synthesis in vivo.

A more dynamic approach to carbon fixation was needed to resolve this impasse. In particular, the biochemical fate of early products would have to be traced, and using a development of the open gas exchange system at Queens, Atkins et al. (1971) supplied ¹⁴CO₂ in pulse–chase experiments to sunflower leaf tissue under conditions in which photorespiration was operating at high rates (21% O₂) or in which it was absent (1% O₂). A series of kinetic experiments showed that synthesis of ¹⁴C-glycine and ¹⁴C-serine was inhibited in low O₂ and that the ¹⁴C precursor for their synthesis was derived from sugar bisphosphates of the PCR cycle, especially RuBP. Indeed RuBP was the obvious source of glycine carbon atoms and the kinetics of glycine turnover closely matched those of RuBP. As these authors concluded, ‘we can no longer view this (glycolate) pathway as an adjunct to the Calvin cycle but must incorporate it completely into the carbon fixation scheme for photosynthesis’ (Atkins et al. 1971).

The question was finally and most elegantly resolved by Ogren and Bowes (1971) who demonstrated that the carboxylating enzyme of the PCR cycle, RuBP carboxylase, was both an oxygenase and a carboxylase! During normal photosynthesis in air, this enzyme thus catalysed formation of both P-glycolate (the precursor of glycolate) and 3-PGA from the oxygenation of RuBP as well as two molecules of PGA from carboxylation of RuBP. In effect, CO₂ and O₂ compete with each other for the same active sites for this oxygenation/carboxylation of RuBP, at last providing a biochemical mechanism which had confused and perplexed photosynthesis researchers since the 1920s. This primary carboxylating enzyme of the PCR cycle, which had hitherto rejoiced in a variety of names (carboxydismutase, fraction 1 protein, RuDP carboxylase and RuBP carboxylase), was renamed Rubisco (ribulose-1,5-bisphosphate carboxylase/oxygenase) to reflect its dual activity.

A scheme for the PCR cycle and photosynthetic carbon oxidation (PCO) pathways then represents the synthesis of almost 70 years of research effort, and integrates the metabolism of P-glycolate with the PCR cycle. This is shown in the following section. 

 

A scheme for photorespiration involving the photosynthetic carbon reduction (PCR) cycle and the photosynthetic carbon oxidation (PCO) pathways represents the synthesis of almost 70 years of research. It integrates the metabolism of P-glycolate with the PCR cycle. This scheme is shown in Figure 2.13.

Specialised reactions within three classes of organelles in leaf cells are required, namely chloroplasts, peroxisomes (originally called microbodies) and mitochondria. Their close proximity in leaf cells (Figure 2.14 below) plus specific membrane transporters facilitate the exchange of metabolites.

Within choroplasts, oxygenase activity by Rubisco results in formation of phosphoglycolate which then enters a PCO cycle, and is responsible for loss of some of the CO₂ just fixed in photosynthesis.

Within peroxisomes, O₂ is consumed in converting glycolate to glyoxylate, and aminated to form glycine.

Within mitochondria, CO₂ is released during conversion of glycine to serine. Subsequently, serine is recovered by peroxisomes where it is further metabolised, re-entering the PCR cycle of chloroplasts as glycerate. About 75% of carbon skeletons channelled into photorespiration are eventually recovered as carbohydrate.

Transport of glycerate and glycolate across the inner membrane of chloroplasts may involve separate translocators as shown in Figure 2.13, or it may involve a single translocator that exchanges two glycolate molecules for one molecule of glycerate. Transport of metabolites across the peroxisomal membrane most likely occurs through unspecific channel proteins, similar to those in the outer membranes of mitochondria and chloroplasts. These outer membranes are not included in this diagram. Mitochondria take up two molecules of glycine and release one molecule of serine. A specific translocator most probably mediates the exchange of these amino acids.

Not only does the photosynthetic oxidation pathway consume O₂ and release CO₂ but ammonia is also produced by mitochondria during synthesis of serine from glycine (Figure 2.13). This ammonia would be extremely toxic if it were not metabolised by either cytosolic or chloroplastic glutamine synthetase. A very effective herbicide that blocks glutamine synthetase has been developed (phosphinothricin, also known as glufosinate or Basta), and when it is applied to (or expressed in) actively growing plants they are killed by their photorespiratory ammonia release.

In summary, participation of photorespiration in leaf gas exchange, and thus dry matter accumulation by plants, reflects kinetic properties of Rubisco, and in particular a relatively high affinity for CO₂ (K_m = 12 µM) compared with a much lower affinity for O₂ (K_m = 250 µM). That contrast in affinity is, however, somewhat offset by the relative abundance of the two gases at catalytic sites of the enzyme where the ratio of O₂:CO₂ partial pressures approaches 1000:1! Thus, CO₂ assimilation always prevails over CO₂ loss in photorespiration. 

At around the same time as the nature of photorespiration was becoming clearer Hatch and Slack (1966) demonstrated that in tropical grasses (initially sugar cane) the first-formed products of photosynthetic CO₂ fixation were the four-carbon acids oxalacetate, malate and aspartate, rather than the 3-PGA formed in the PCR cycle. Furthermore, the carboxylation reaction involved PEP carboxylase and carbon was subsequently transferred to PCR cycle intermediates. As noted earlier (Section 2.2) C₄ plants show no apparent CO₂ release in light. The explanation lies in their anatomy and multiple carboxylation reactions rather than in the absence of the pathway of photorespiration. Bundle sheath cells are equipped with a CO₂-concentrating mechanism that favours carboxylation over oxygenation reactions due to increased partial pressure of CO₂, while photorespiratory release of CO₂ is further prevented through the activity of PEP carboxylase which refixes any respired CO₂ formed from the oxygenase function of Rubisco.

Unicellular green algae also posed a problem for the simple extrapolation of early models for photorespiratory metabolism in C₃ leaves. Although organisms such as Chlorella had been used to establish the PCR cycle, and indeed provided much early evidence for effects of O₂ on photosynthesis and formation of glycolate in light, they also appeared to lack CO₂ evolution in light (Lloyd et al. 1977). In this case the explanation lies in a CO₂-concentrating mechanism which effectively increases the internal pool of inorganic carbon (CO₂ and HCO₃^–) thereby favouring the carboxylase function of Rubisco over its oxygenase function.

Such a substantial loss of carbon concurrent with CO₂ fixation raises the question of whether eliminating or minimising photorespiration in C₃ plants could enhance their yield, and specifically that of major crop plants such as rice, wheat, grain legumes, oil seeds and trees, none of which are C₄ species. Faced with an expanding world human population and an increasing demand for food and animal feed, enhanced agricultural productivity is a global necessity. In its most obvious form a scenario which alters or removes the oxygenase function of Rubisco could achieve such a goal. In an early review of the process of photorespiration in plants, Ogren (1984) noted that ‘the sequence of reactions constituting the photorespiratory pathway in C₃ plants appears to be firmly established’ and he went on to suggest that, although reducing the loss of fixed carbon as CO₂ in the process may be a valid goal to improve the yield of crop plants, it is not clear whether or not this can be achieved by specific changes to the kinetic and catalytic properties of Rubisco alone.

Photorespiration may be loosely considered as a wasteful process because previously fixed carbon is lost and energy is dissipated. Ideal destinations for photoassimilates include synthetic pathways leading to fixed biomass and respiratory pathways for re-release of fixed energy in a controlled sequence of reactions leading to ATP and NAD(P)H for use in other synthetic events.

However, situations do exist where energy dissipation via photorespiration can be beneficial. For example, photo-oxidative damage can be alleviated in shade-adapted plants that experience strong irradiance if photorespiratory processes are allowed to proceed. Depriving such plants of an external oxygen supply, and hence preventing photosynthetic carbon oxidation, will exacerbate chloroplast lesions due to strong irradiance. Photosynthetic variants which obviate photo-respiratory loss, and most notably C₄ plants, integrate structure and function in a way that forestalls photo-oxidative damage and leads to their outstanding performance under warm conditions. Environmental factors that constitute selection pressure for such photosynthetic adaptation are reviewed by Sage et al (2012). 

Atkins CA, Canvin DT, Foch H (1971) Intermediary metabolism of photosynthesis in relation to carbon dioxide evolution in sunflower. In MD Hatch, CB Osmond, RO Slatyer, eds, Photosynthesis and Photorespiration. John Wiley, New York, pp 497-505

Hatch MD, Slack CR (1966) Photosynthesis by sugarcane leaves. A new carboxylation reaction and the pathway of sugar formation. Biochem J 101: 103-111

Kisaki T, Tolbert NE (1969) Glycolate and glyoxylate metabolism by isolated peroxisomes and chloroplasts. Plant Physiol 44: 242-250

Lloyd ND, Canvin DT, Culver DA (1977) Photosynthesis and photorespiration in algae. Plant Physiol 70: 1637-1640

Ludwig LJ, Canvin DT (1971) The rate of photorespiration during photosynthesis and the relationship of the substrate of light respiration to the products of photosynthesis in sunflower leaves. Plant Physiol 48: 712-719

Meidner HA (1962) The minimum intercellular-space CO₂ concentration of maize leaves and its influence on stomatal movements. J Exp Bot 13: 284-293

Ogren WL (1984) Photorespiration: pathways, regulation and modification. Annu Rev Plant Physiol 35: 415-442

Ogren WL, Bowes G (1971) Ribulose diphosphate carboxylase regulates soybean photosynthesis. Nature New Biol 230: 159-160

Sage RF, Sage TL, Kocacinar F (2012) Photorespiration and the evolution of C₄ photosynthesis. Annu Rev Plant Biol 63: 19-47.

Tregunna EB, Krotkov G, Nelson CD (1966) Effect of oxygen on the rate of photorespiration in detached tobacco leaves. Physiol Plant 19: 723-733

Warburg O, Krippahl G (1960) Glykolsaurebildung in Chlorella. Z Naturforsch 15b: 197-199

Wang D, Waygood ER (1962) Carbon metabolism of ¹⁴CO₂ labelled amino acids in wheat leaves. I. A pathway of glyoxylate-serine metabolism. Plant Physiol 37: 826-832

Zelitch I (1966) Increased rate of net photosynthetic carbon dioxide uptake caused by the inhibition of glycolate oxidase. Plant Physiol 41: 1623-1631

Starch is the principal storage carbohydrate in plants and this carbon reserve plays a number of important roles in plants. It is composed of two polymers of glucose, amylose and amylopectin and is stored in the plastid (chloroplast in leaves, amyloplasts in non-photosynthetic tissues) as insoluble, semi-crystalline granules. Starch is accumulated during rapid growth in the day and is almost completely degraded at night to mostly glucose and maltose, which is exported from the chloroplast and metabolised in the cytosol (Figure 2.19). Starch degradation is initiated by the addition of phosphate groups at the C6-position and C3-position of individual glucosyl residues that act to disrupt the packing of the glucans at the granule surface. These phosphate additions are catalysed by two enzymes, glucan water dikinase (GWD) and phosphoglucan water dikinase (PWD) respectively. The hydrolysis of the resulting glucan and phosphoglucan chains is carried out by a suite of enzymes including the phosphoglucan phosphatases (SEX4/LSF2), β-amylases (BAM1/BAM3), debranching enzymes (DBE; ISA3/LDA), α-amylase (AMY3), α-glucan phosphorylase and the disproportionating enzyme 1 (D-enzyme 1; an α-1,4-glucanotransferase).). The resulting maltose and glucose are exported to the cytosol by the glucose transporter (pGlcT) and maltose transporter (MEX1) and glucose-1-phosphate is thought to be exported by a similar but yet unknown mechanism. Once in the cytosol the maltose and glucose are converted to substrates for either sucrose synthesis, glycolysis or the oxidative pentose phosphate pathway by a number of enzymes including the disproportionating enzyme 2 (D-enzyme 2; an α-1,4-glucanotransferase), α-glucan phosphorylase, hexokinase and phosphoglucomutase.

Sucrose is the world’s most abundant disaccharide, it is only produced by photosynthetic organisms and serves a role as a transportable carbohydrate and sometimes as a storage compound. The reactions in plant tissues leading to degradation of sucrose to hexose monophosphates are outlined in Figure 2.20.

The first step is cleavage of the glycosidic bond by either invertase (Equation 2.1) or sucrose synthase (Equation 2.2).

\[\text{Sucrose + H}_2\text{O} \rightarrow \text{D-Glucose + D-Fructose}\tag{2.1}\]

\[\text{Sucrose + UDP} \rightarrow \text{UDP-Glucose + D-Fructose}\tag{2.2}\]

Plant tissues contain distinct invertases located in the vacuole, cell wall (acid invertases) cytosol, mitochondria, nucleus, and cholorplast (neutral/alkaline invertases) which hydrolyse sucrose to glucose and fructose in an irreversible reaction. The invertases are differentially regulated by a number of mechanisms including pH to allow them to function in cell expansion, supply of carbon skeletons and energy metabolism. Multiple isoforms of sucrose synthase are located in the cytosol or cytosolic membranes that catalyse a thermodynamically reversible reaction, but this reaction probably acts only to breakdown sucrose in vivo. Their activity is developmentally regulated and they have functions in the supply of activated glucose for starch and cellulose biosynthesis. While both invertase and sucrose synthase can both breakdown sucrose, research using knockouts of multiple isoforms of both enzymes has shown that sucrose synthase is not required for normal growth in Arabidopsis, whereas invertase is indispensable. However this does not rule out the requirement of sucrose synthase in certain tissues of crop plant tubers, seeds and fruits where it has been shown to be crucial. Glucose and fructose are metabolised further by phosphorylation to the corresponding hexose-6-P by hexokinase. Hexokinase in plant tissues is associated with the outer surface of mitochondria.

The glycolytic pathway involves the oxidation of the hexoses and hexose phosphates molecules produced from the breakdown of starch or sucrose to generate ATP, reductants and pyruvate (Figure 2.21).

Regulation of PFK and PPi-PFK is achieved by a combination of mechanisms, including pH, the concentration of substrates and effector metabolites and changes in subunit association. Phosphoenolpyruvate (PEP) is a potent inhibitor of both of PFK and PPi-PFK, inhibiting at µM concentrations and Pi can activate the cytoplasmic PFK, whereas the plastidic form is slightly inhibited by Pi. A number of other effectors of PFK have been identified including ADP, 3-phosphoglycerate and phosphoglycolate as well as it ability to accept ribonucleoside triphosphates other than ATP as the phosphate donor. PP_i-PFK, has a catalytic potential higher than that of PFK and is strongly activated by Fructose-2,6-bisphosphate, but has no effect on PFK.

Fructose-1,6-bisphosphate is cleaved by fructose bisphosphate aldolase to form glyceraldehyde-3-phosphate and dihydroxyacetone phosphate, and these triose phosphates can be interconverted in a reaction catalysed by triose phosphate isomerase. Glyceraldehyde-3-phosphate is oxidised to 1,3-bisphosphoglycerate by a nicotinamide adenine dinucleotide (NAD⁺)-dependent glyceraldehyde 3-P dehydrogenase. Glyceraldehyde 3-P dehydrogenase is sensitive to inhibition by the reduced pyridine nucleotide cofactor (NADH), which must be reoxidised to maintain the flux through the glycolytic pathway. A phosphate group is then transferred from 1,3-bisphosphoglycerate to ADP forming ATP and 3-phosphogylcerate by phosphoglycerate kinase. In the cytosol a bypass is present that can convert glyceraldehyde-3-phosphate directly to 3-phosphoglycerate without phosphorylation by a non-phosphorylating NADP dependent glyceraldehyde 3-phosphate dehydrogenase. The resulting 3-phosphoglycerate is then converted to phosphenolpyruvate (PEP) by the action of phosphoglycerate mutase and then enolase.  

The end-products of glycolytic reactions in the cytosol are determined by the relative activities of the two enzymes that can utilise PEP as a substrate: pyruvate kinase, which forms pyruvate and a molecule of ATP, and PEP carboxylase, which forms oxaloacetate (Figure 2.21). Both of these reactions are essentially irreversible and there are fine controls that regulate the partitioning of PEP between these reactions. Pyruvate kinase is controlled post translationally by a partial C-terminal truncation which may yield altered regulatory properties and a phosphorylation and ubiquitin conjugation that targets the protein to the 26S proteasome for complete degradation, it is also inhibited by ATP. Whereas PEP carboxylase is inhibited by malate and thus its regulation is independent of cell energy status. The sensitivity of PEP carboxylase to malate is regulated by phosphorylation of a N-terminal serine of the enzyme, with the phosphorylated form less sensitive to malate inhibition. Oxaloacetate is then reduced by malate dehydrogenase to malate which, along with pyruvate, can be taken up into mitochondria and metabolised further in the TCA cycle (see below). The reduction of oxaloacetate in the cytosol could provide a cytosolic mechanism for oxidising NADH formed by glyceraldehyde 3-P dehydrogenase (Figure 2.21).

Another level of regulation of components of glycolysis is their physical location within the plant cell. Under conditions of high respiratory activity, a greater proportion of the cytosolic enzymes of glycolysis are present on the surface of mitochondria. In contrast when respiration is experimentally inhibited, a decrease in the association of glycolytic enzymes with the mitochondria is observed. It is likely that the glycolytic enzymes associate dynamically with mitochondria to support respiration and that this association restricts the use of glycolytic intermediates by competing metabolic pathways.

An alternative route for the breakdown of glucose-6-phosphate is provided by the oxidative pentose phosphate pathway (OPP) (Figure 2.22). This pathway functions mainly to generate reductant (i.e. NADPH) for biosynthetic processes including the assimilation of inorganic nitrogen and fatty acid biosynthesis and to maintain redox potential to protect against oxidative stress. In addition, the reversible oxidative section of the pathway is the source of carbon skeletons for the synthesis of a number of compounds. For example ribose-5-phosphate provides the ribosyl moiety of nucleotides and is a precursor for the biosynthesis of purine skeletons and erythrose-4-phosphate, which is the precursor for the biosynthesis of aromatic amino acids by the shikimic acid pathway.

 

The pathway begins with the dehydrogenation of glucose-6-phosphate catalyzed by glucose-6-phosphate dehydrogenase to produce 6-phosphoglucolactone and is the first step of the oxidative phase of the pathway. The 6-phosphoglucolactone is then hydrolysed to 6-phosphogluconate by 6-phosphogluconolactonase and then undergoes oxidative decarboxylation by 6-phosphogluconate dehydrogenase to produce ribulose-5-phosphate in the final step of the oxidative phase. Overall this phase of the pathway produces two molecules of NADPH from the conversion of glucose-6-phosphate to ribulose-5-phosphate. The non-oxidative phase begins with the reaction of ribulose-5-phosphate with either ribulose-5-phosphate isomerase or ribulose-5-phosphate epimerase followed by a series of reactions catalyzed by transaldolase and transketolase. These reactions result in the production of two molecules of fructose-6-phosphate and one glyceraldehyde 3-phosphate. The glyceraldehyde 3-phosphate and fructose-6-phosphate in the oxidative pentose phosphate pathway may be exchanged with enzymes of glycolysis.

As with glycolysis, reactions of the pentose phosphate pathway are catalysed by different isoforms of the enzymes that occur either in the cytosol or in plastids. Although transketolase and transaldolase may be absent from the cytosol of some species, the activity is maintained by phosphate translocator proteins on the plastid inner-envelope membrane that have the capacity to translocate sugar phosphates.

Organic acids such as pyruvate and malate produced in the cytosol by processes described above are further oxidised in mitochondria by the tricarboxylic acid (TCA) cycle and subsequent respiratory chain. Energy released by this oxidation is used to synthesise ATP which is then exported to the cytosol for use in biosynthesis and growth.

(a)  Mitochondrial structure

Plant mitochondria (Figure 2.23) are typically double-membrane organelles where the inner membrane is invaginated to form folds known as cristae to increase the surface area of the membrane. The outer membrane contains relatively few proteins (<100) and is permeable to most small compounds (< Mr=5 kDa) due to the presence of the pore-forming protein VDAC (voltage dependent anion channel) which is a member of the porin family of ion channels. The inner membrane is the main permeability barrier of the organelle and controls the movement of molecules by means of a series of carrier proteins many of which are members of mitochondrial substrate carrier family (MSCF). The inner membrane also houses the large complexes that carry out oxidative phosphorylation and encloses the soluble matrix which contains the enzymes of the TCA cycle and many other soluble proteins involved in a myriad of mitochondrial functions.

Mitochondria are semi-autonomous organelles with their own DNA and protein synthesis machinery. However, the mitochondrial genome encodes only a small portion of the proteins which make up the mitochondrion; the rest are encoded on nuclear genes and synthesised in the cytosol. These proteins are the transported into the mitochondrion by the protein import machinery and assembled with the mitochondrially synthesised subunits to form the large respiratory complexes. The number of mitochondria per cell varies with tissue type (from a few hundred in mature differentiated tissue to some thousands in specialised cells). Understandably, more active cells, with high energy demands, such as those in growing meristems are generally equipped with larger numbers of mitochondria per unit cell volume, and consequently show faster respiration rates.

(b)  Mitochondrial substrates

Two substrates are produced from glycolytic PEP for oxidation in mitochondria: malate and pyruvate (Figure 2.21). These compounds are thought to be the most abundant mitochondrial substrates in vivo. However, amino acids may also serve as substrates for mitochondrial respiration in some tissues, particularly in seeds rich in stored protein or under conditions of sugar depletion such as extended darkness, shading and senescence. β-oxidation of fatty acids typically does not occur in plant mitochondria, this oxidation is principally carried out in peroxisomes in plants.

(c) Carbon metabolism in mitochondria

Malate and pyruvate enter the mitochondrial matrix across the inner membrane via separate carriers. Malate is then oxidised by either malate dehydrogenase (a separate enzyme isoform from that in the cytosol), which yields oxaloacetate (OAA) and reduced nicotinamide adenine dinucleotide (NADH), or NAD⁺-linked malic enzyme, which yields pyruvate and NADH and releases CO₂ (Figure 2.24). Cytosolic pyruvate carboxylase is an alternative means of providing substrate to mitochondria by combining pyruvate with HCO₃ to yield OAA that can then be imported into mitochondria.

 

Pyruvate formed either from malate and malic enzyme or transported directly from the cytosol is oxidised inside mitochondria by the pyruvate dehydrogenase complex (PDC) to form CO₂, acetyl-CoA and NADH. This enzyme, which requires coenzyme A, thiamine pyrophosphate and lipoic acid as cofactors, effectively links the TCA cycle to glycolysis. PDC comprises three enzymes E1 (2-oxo acid dehydrogenase), E2 (acyltransferase) and E3 (lipoamide dehydrogenase). This complex is regulated by phosphorylation of the E1 subunit, lowering PDC activity in the day and increasing PDC activity at night. Pyruvate dehydrogenase is also subject to feedback inhibition from acetyl-CoA and NADH.

(d) Tricarboxylic acid (TCA) cycle

The TCA cycle begins with the condensation of acetyl-CoA and OAA, to form the six-carbon molecule citrate and the release coenzyme A (CoA) (Figure 2.24) in a reaction catalysed by citrate synthase. Aconitase catalyses the next step, converting citrate to isocitrate in a two-step reaction (dehydration/hydration) with cis-aconitate as an intermediate.

NAD-linked isocitrate dehydrogenase then oxidatively decarboxylates isocitrate to form CO₂ and 2-oxoglutarate, and reduce NAD⁺ to NADH. The 2-oxoglutarate formed is also oxidatively decarboxylated to succinyl-CoA in a reaction catalysed by the enzyme 2-oxoglutarate dehydrogenase. This enzyme complex has similarities to pyruvate dehydrogenase and its reaction is analogous to the formation of acetyl-CoA from pyruvate by pyruvate dehydrogenase. The reaction mechanisms are also very similar, with 3 subunit enzymes, but 2-oxoglutarate dehydrogenase is not subject to the phosphorylation control that regulates pyruvate dehydrogenase. Succinyl-CoA synthase then catalyses the conversion of succinyl-CoA to succinate, with the concomitant phosphorylation of ADP to ATP, the only substrate-level phosphorylation step in the mitochondrion. This enzyme in plants differs from its mammalian counterpart in that it is specific for ADP rather than GDP.

Succinate dehydrogenase (SDH, Complex II), which catalyses the oxidation of succinate to fumarate, is the only membrane-bound enzyme of the TCA cycle and is part of the respiratory electron transfer chain (Figure 2.25). SDH is a large complex consisting of four core subunits, as well as number other associated subunits.

Fumarase catalyses the hydration of fumarate to malate followed by malate dehydrogenase that catalyses the final step of the TCA cycle, oxidising malate to OAA and producing NADH. The reaction is freely reversible, although the equilibrium constant strongly favours the reduction of OAA, necessitating rapid turnover of OAA and NADH to maintain this reaction in a forward direction.

Overall, during one turn of the cycle, three carbons of pyruvate are released as CO₂, one molecule of ATP is formed directly, and four NADH and one FADH₂ are produced. The strong reductants are oxidised in the respiratory chain to reduce O₂ and produce ATP. Although most of the TCA cycle enzymes in plant mitochondria are NAD linked, NADP-dependent isoforms of isocitrate and malate dehydrogenases also exist, and these may play a role in a protective reductive cycle in the matrix.

Regulation of carbon flux through the TCA cycle probably occurs via phosphorylation/dephosphorylation of pyruvate dehydrogenase, which will depend in turn on mitochondrial energy status and feedback inhibition of various enzymes by NADH and acetyl-CoA. The rate of cycle turnover thus depends on the rate of electron flow through the respiratory chain (to reoxidise NADH) and the utilisation of ATP in the cell to provide ADP for substrate level and oxidative phoshorylation. TCA cycle turnover will also depend on the rate of substrate provision by reactions in chloroplasts and cytosol. A number of studies of TCA cycle mutants have demonstrated the wide impact these enzymes have not simply on TCA cycle function but as steps for the delivery of organic acids for other processes in plant cells such a photosynthetic performance, plant biomass, root growth, photorespiration, nitrogen assimilation, amino acid metabolism, and stomatal function.

The respiratory electron transfer chain (ETC) of mitochondria consists of a series of large membrane-bound protein complexes (Complexes I, II, III, IV) which together with a small lipid ubiquinone (UQ) and the small protein cytochrome c catalyse the transfer of electrons from NADH and succinate to O₂, forming H₂O (Figure 2.25). Electron flow from NADH and succinate to oxygen is coupled to proton translocation out of the matrix to the intermembrane space which establishes a proton electrochemical gradient (Dµ_H⁺) across the inner membrane that is used to drive phosphorylation of ADP to form ATP by the F₁F_O ATP synthase (Complex V, Figure 2.25).

(a) Complex I (CI)

 

NADH-UQ oxidoreductase, is responsible for the oxidation of matrix NADH and reduction of ubiquinone (UQ) in the inner mitochondrial membrane (Figure 2.25). In plants it is a large multi-subunit complex composed of 49 subunits, up to ten of which are synthesised in the mitochondrion whilst the others are imported from the cytosol. One of the subunits, a 50 kDa protein, contains flavin mononucleotide as a cofactor and is the dehydrogenase which oxidises NADH and passes electrons to iron-sulphur containing subunits of the complex, and eventually to ubiquinone. The passage of electrons through the complex is accompanied by H⁺ translocation across the membrane. Complex I is inhibited specifically by the flavonoid rotenone and its analogues. The NADH-binding site is exposed to the matrix and the complex oxidises NADH produced by the TCA cycle and other NAD-linked enzymes (Figure 2.25). Studies of mutations of CI subunits have shown that plants can survive without CI due to the activity of alternative NAD(P)H dehydrogenases (see below). Such mutants have a variety of interesting phenotypes including viral infection tolerance, prolonged hydration under water-deficient conditions and altered organic and amino acid concentrations. Using a series of Arabidopsis CI subunit knockout mutants, a number membrane arm subcomplexes (of 200, 400, 450 and 650 kDa) have been identified using BN-PAGE and antibodies. It is proposed that at least some of these subcomplexes may be assembly intermediates during CI formation, and these are seen to accumulate when specific subunits are absent.

(b) Complex II (CII)

Succinate dehydrogenase, is an enzyme of both the respiratory ETC and the TCA cycle (see above) (Figure 2.25). It is composed of four core subunits: a flavoprotein (SDHI), an iron-sulphur subunit (SDH2) and two membrane anchor subunits (SDH3 and SDH4) in most organisms. In plants, the purification of the complex has revealed the common core subunits, but also additional proteins of unknown function that co-migrate with the complex. All SDH subunits are encoded in the nuclear genome in Arabidopsis. SDH contains FeS and flavin centres which participate in electron transfer from succinate to ubiquinone. Unlike complex I, complex II does not pump H⁺ and succinate oxidation is therefore linked to the synthesis of less ATP per O₂ reduced (see below). Malonate, an analogue of succinate, is a strong competitive inhibitor of succinate dehydrogenase. Knockout mutants of the SDH1 gene have been shown to be embryo lethal, but knockdown of SDH1 and SDH2 leads a array of phenotypes including altered stomatal aperture, mitochondrial ROS production and nitrogen use efficiency. SDHAF1 and SDHAF2 assist in CII assembly in plants and knockdown of the SDHAF2 homolog lowers SDH assembly and reduces root growth.

(c) Complex III (CIII)

Ubiquinone-cytochrome c oxidoreductase or the cytochrome b/c₁ complex as it is sometimes known, contains 10 subunits including a number of bifunctional core proteins. These proteins act both in CIII function and as a matrix processing peptidase, removing targeting presequences from imported matrix synthesized proteins (Figure 2.25). A single subunit of this complex, cytochrome b, is encoded by the plant mitochondrial genome, whilst the others are encoded by the nuclear genome. It contains two b-type cytochromes, b₅₆₆ and b₅₆₂, cytochrome c₁, an iron-sulpher protein named the Rieske iron–sulphur protein and several other polypeptides. Electron flow from ubiquinol to cytochrome c is accompanied by the translocation H⁺ across the membrane, via the so-called Q cycle. Various inhibitors of complex III have been discovered, with antimycin A and myxothiazol most widely used in research. The assembly of CIII is modular and includes an early core subcomplex, a late core subcomplex and the final dimeric CIII. Approximately 13 assembly factors implicated in aiding one or more of the different stages of CIII assembly in yeast, however little is known about CIII assembly or functional assembly factors in plants.

(d) Complex IV (CIV)

Cytochrome c oxidase is the final step of electron transfer of the classical ETC. As the name implies, cytochrome c oxidase accepts electrons from cytochrome c and transfers them to O₂ which is reduced to form H₂O. Purification of CIV in plants has identified a complex containing 14 protein subunits (Figure 2.25). Eight of these proteins are homologous to known CIV subunits from other organisms, together with a further six proteins that are probably plant specific. Two cytochrome haem centres, a and a₃, and two copper atoms make up its redox active components and like complex I, cytochrome oxidase is a proton pump. Cytochrome c oxidase is sensitive to a number of inhibitors, the best known of which are carbon monoxide and cyanide. Plants, however, show resistance to both carbon monoxide and cyanide because they are equipped with an alternative oxidase (see below). Studies of human and yeast CIV has shown an assembly pathway that involves the sequential incorporation of CIV subunits, initiated by subunit 1 and assisted by over 40 assembly factors. Research investigating the plant homolog of the yeast assembly factor COX19 has found it is capable of complementing the yeast cox19 null mutant. This suggests it might play a role in the biogenesis of plant cytochrome c oxidase or in the replacement of damaged forms of the enzyme. However, our knowledge of the assembly of CIV in plants is still incomplete.

(e) Ubiquinone and Cytochrome c

These large multi-subunit complexes (I, II, III, IV) of the respiratory ETC chain are embedded in the inner mitochondrial membrane by virtue of their hydrophobic subunits, and interact with one another via two smaller molecules: ubiquinone and cytochrome c. The lipid-soluble ubiquinone also known as coenzyme Q₁₀ is a small mobile electron carrier which moves rapidly along and across the membrane, and participates in H⁺ transport across the membrane via the Q-cycle as well as shuttling electrons from complexes I and II to complex III. Cytochrome c is a small haem-containing protein located on the outer surface of the inner membrane, which shuttles electrons between complexes III and IV. In this respect, the respiratory chain is similar in layout to the photosynthetic electron transport chain: three large complexes which communicate by a quinone and a small mobile protein (Cyt c or plastocyanin). However, orientation of components in the membrane is inverted and the net reaction catalysed is opposite to that in chloroplasts (Figure 1.12).

When electrons are transferred from NADH to O₂, a large release of redox energy enables ATP formation in complex V of the respiratory chain (Figure 2.24). Energy release associated with electron transport is conserved by H⁺ translocation across the membrane to form a proton electrochemical gradient (Δµ_H⁺) that has both an electrical membrane potential (Δψ) and a pH component (Δµ_H⁺ = Δψ + ΔpH). This is known as the chemiosmotic theory and was originally proposed by Peter Mitchell in 1960s. In plant mitochondria, Δµ_H⁺ exists mainly as a Δψ of ~150–200 mV, with a pH gradient (ΔpH) of ~0.2–0.5 units. ATP synthesis occurs as H⁺ move from a compartment of high potential (the intermembrane space) to one of low potential (the mitochondrial matrix) through the ATP synthase complex. Oxidation of NADH via the cytochrome pathway has three associated H⁺ translocation sites and is linked to synthesis of up to three ATP molecules for each molecule of NADH oxidized. By contrast, both succinate and alternative NADH oxidation by the rotenone-insensitive NADH dehydrogenases (see below) are linked to the synthesis of only two ATP molecules per NADH or succinate, as these events are associated with only two H⁺ pumping sites.

 

(a) Complex V (CV)

ATP synthase is a membrane-bound F₁F₀ type H⁺-ATP synthase that harnesses the Δµ_H⁺ generated by the ETC to produce ATP. It is composed of a hydrophobic F₀ component which channels protons through the inner mitochondrial membrane and also anchors the complex to the membrane and a hydrophilic F₁ component which catalyses ATP formation and protrudes into the matrix. The core subunits of the enzyme are highly conserved in both prokaryotic and eukaryotic organisms. In plants, the majority of mitochondrial F₁ subunits are encoded in the nucleus and translated in the cytosol before being imported into the mitochondria (including  α, β, γ and ε subunits), while most of the F₀ subunits are encoded in the plant mitochondrial genome and translated in the mitochondrial matrix (including a, b, c and A6L subunits).  The reaction mechanism of the ATP synthase is known as the three-site alternating binding site mechanism. According to this model, F₁ has three nucleotide-binding sites which can exist in three configurations: one with loosely bound nucleotides, one with tightly bound nucleotides and the third in a nucleotide-free state. H⁺ movement through F₀ results in rotation of F₁, causing a conformational change during which the site with loosely bound ADP and P_i is converted to one which binds them tightly in a hydrophobic pocket in which ATP synthesis occurs. Further H⁺ movement then causes another rotation of F₁ and the ATP binding site is exposed and releases the nucleotide. In the meantime, the other nucleotide-binding sites are undergoing similar changes, with ADP and P_i being bound and converted to ATP. Thus H⁺ translocation drives the three sites through three different configurations and the main expenditure of energy is in the induction of a conformational change that releases tightly bound ATP, rather than in ATP synthesis itself. The F₀ complex also contains a protein known as the oligomycin-sensitivity-conferring protein (OSCP) because it binds the antibiotic oligomycin that prevents H⁺ translocation through F₀ and inhibits ATP synthesis. Therefore, adding oligomycin to mitochondria oxidising a substrate in the presence of ADP restricts O₂ uptake. Generally knockouts of ATP synthase core subunits are lethal in plants, however inducible knockdowns have enabled investigations into the tissue-specific phenotypes incurred by slowing the rates of mitochondrial ADP:ATP cycling at a number of different developmental stages. It has been proposed that the assembly of plant CV comprises of three steps, the first being the formation of a rapidly turned over F1 subcomplex in the matrix, followed by an intermediate stage where F1 associates with the inner membrane and still turns over at a fast rate, and then finally a union of F₁ with F_O to form functional CV. A number of assembly factors (Atp10, Atp11, Atp12, Atp22, Atp23 and Fmc1) have been discovered for yeast ATP synthase, however, a detailed study of the presence and conservation of CV assembly factors in plants has not been undertaken.

(b) Respiratory control

Electron transport through the respiratory chain, and therefore rate of O₂ uptake, is controlled by availability of ADP and P_i, a phenomenon described as ‘respiratory control’. In the absence of ADP or P_i, the proton pore of ATP synthase is blocked and a Δµ_H⁺ builds up to a point where it restricts further H⁺ translocation across the inner membrane. Since electron transport is functionally linked to H⁺ translocation, this elevated Δµ_H⁺ will also restrict O₂ consumption. That outcome is easily seen with isolated mitochondria (Figure 2.26) where O₂ uptake is stimulated by adding ADP (‘State 3’ respiration). When all of the added ADP has been consumed, O₂ uptake decreases again (‘State 4’). In steady state, the rate of electron flow is determined by the rate of flow of H⁺ back across the membrane: when ADP and P_i are available the backflow is rapid and occurs via ATP synthase; in the absence of these compounds, backflow is by slow diffusion through the membrane.

The ratio of State 3 to State 4 (the respiratory control ratio) is thus an indication of coupling between ADP phosphorylation and electron transport. Larger values represent tighter coupling. The proton leak can be dramatically stimulated by some compounds which act as protonophores or proton channels; these compounds collapse the Δµ_H⁺ and increase O₂ uptake up to the State 3 rate (Figure 2.26). However, no ATP is formed and these compounds are called uncouplers because they uncouple the linked processes of electron transport and phosphorylation.

Plant mitochondria have a respiratory chain which is more complicated than that of animals and contains alternative NADH dehydrogenases, alternative oxidases which catalyse cyanide-insensitive O₂ consumption and uncoupling proteins that acts to dissipate the Δµ_H⁺. The alternative NADH dehydrogenases and alternative oxidase do not translocate protons and therefore are not linked to ATP synthesis; they are often referred to as the non-phosphorylating bypasses of the plant respiratory chain. These pathways were initially identified in plant mitochondria as they are able to continue to respire in the presence of the CIV inhibitor, cyanide and the CI inhibitor, rotenone and by their ability to exhibit natively uncoupled respiration in the absence of an ADP source.

(a) Alternative oxidase

Cyanide-insensitive respiration is catalysed by the alternative oxidase (AOX). This alternative terminal oxidase is a diiron quinol oxidase that branches from the classical respiratory chain at UQ and reduces oxygen to water without an associated proton translocation. The oxidase exists in mitochondria as a dimer which can be inactivated by covalent linkage via disulphide bonds. The reduced enzyme is stimulated allosterically by pyruvate and some other 2-oxo acids (such as glyoxylate), which interact directly with the oxidase. The exact role of AOX continues to be debated but it appears to play an antioxidant role in plant mitochondria. Research has shown it is actively induced by oxidative stress and the different genes for the oxidase have been shown to be both development- and tissue-specific. Knockout of AOX leads to reactive oxygen species and anthocyanin accumulation in the leaves exposed to a combination of high light and drought stress. AOX can be inhibited by hydroxamic acids such as n-propylgallate (nPG) and salicyl hydroxamic acid (SHAM).

(b) Alternative NAD(P)H dehydrogenases

Alternative NAD(P)H dehydrogenases have been shown to be present on both sides of the inner mitochondrial membrane. These type II NAD(P)H dehydrogenases oxidise external or cytosolic and matrix NADH and NADPH and are insensitive to the classical CI inhibitor rotenone. As with AOX, these enzymes do not translocate protons and therefore are not linked to ATP synthesis. The Arabidopsis genome contains seven genes encoding NAD(P)H dehydrogenases, although it appears that some of these isoforms are present in multiple subcellular compartments in addition to mitochondria.

(c) Uncoupling proteins

Uncoupling proteins (UCPs) are members of the mitochondrial carrier family of proteins. They act to dissipate the Δµ_H⁺ built up the ETC by transporting H⁺ back across the inner membrane uncoupling proton and electron transport. The reactive oxygen species superoxide activates UCPs and this suggests a possible mechanism for the engagement of this enzyme in vivo. Analysis of knockouts of UCP (AtUCP1) showed that its absence led to localized oxidative stress but did not impair the ability of the plant to withstand a wide range of abiotic stresses. However, knockout of UCP1 did limit the photorespiration rate of plants and led to a reduction in photosynthetic carbon assimilation. This suggests that the main role of UCP1 in leaves is to maintain the redox poise of the mitochondrial ETC to facilitate photosynthesis.

(a)  Efficiency

Respiration represents a substantial loss of carbon from a plant, and under adverse conditions can be as high as two-thirds of the carbon fixed daily in photosynthesis. Both the rate and the efficiency of respiration will therefore affect plant growth significantly. The overall process of respiration results in the release of a substantial amount of energy which may be harnessed for metabolic work. In theory, the energy released from the complete oxidation of one molecule of glucose to CO₂ and H₂O in respiratory reactions leads to the synthesis of 36 molecules of ATP. However, in plants, because there are alternative routes for respiration, this yield can be greatly reduced. Mechanisms for regulating respiration rates in whole plants remain unclear. Convention has it that the rate of respiration is matched to the energy demands of the cell through feed-back regulation of glycolysis and electron transport by cytosolic ATP/ADP. However, since plants have non-phosphorylating bypasses in their respiratory chain that are insensitive to ATP levels, and since PEP carboxylase and PFP might be involved in sucrose degradation, the situation in vivo is not so simple. For example, the rotenone-insensitive alternative NADH dehydrogenases requires high concentrations of NADH in the matrix before it can operate and seems to be active only when substrate is plentiful and electron flow through complex I is restricted by lack of ADP. Alternative oxidase activity also depends on carbon and ADP availability and its flux is very dependent on the degree of environmental stress of the plant. In other words, non-phosphorylating pathways act as carbon or reductant ‘overflows’ of the main respiratory pathway and will only be active in vivo when sugar levels are high and the glycolytic flux rapid, when the cytochrome chain is inhibited, or when the bypasses have been induced significantly during stress. In glycolysis, the interaction between environmental signals and key regulatory enzymes, as well as the role of PFP and its activator fructose-2,6-P₂, will be important.

(b)  Allocation of respiratory energy to process physiology

One way of viewing respiratory cost for plant growth and survival is by subdividing measured respiration into two components associated with (1) growth and (2) maintenance. This distinction is somewhat arbitrary, and these categories of process physiology must not be regarded as discrete sets of biochemical events. Such energy-dependent processes are all interconnected because ATP represents a universal energy currency for both, while a common pool of substrates is drawn upon in sustaining production of that ATP. Nevertheless, cells do vary in their respiratory efficiency, while genotype × environment interactions are also evident in both generation and utilisation of products from oxidative metabolism. The benefit of a high respiration rate is that more ATP is produced, which provides vital energy for growth of new tissue and defence processes, such as antioxidant activation, metabolite transport or production of resistant protein isoforms. However, the cost of high respiration rates is that carbon is expended on respiration instead of being allocated to synthesis of new tissue, therefore limiting growth capacity. Variation in respiration rate has implications for growth and resource use efficiency in plants during drought, temperature and salinity responses of plants.

Araujo WL, Nunes-Nesi A, Nikoloski Z, Sweetlove LJ, Fernie AR (2012) Metabolic control and regulation of the tricarboxylic acid cycle in photosynthetic and heterotrophic plant tissues. Plant Cell Environ 35: 1-21

Atkin OK, Macherel D (2009) The crucial role of plant mitochondria in orchestrating drought tolerance. Ann Bot 103: 581-597

Atkin OK, Tjoelker MG (2003) Thermal acclimation and the dynamic response of plant respiration to temperature. Trends Plant Sci 8: 343-351

Jacoby RP, Li L, Huang S, Pong Lee C, Millar AH, Taylor NL (2012) Mitochondrial composition, function and stress response in plants. Journal of Integrative Plant Biology 54: 887-906

Jacoby RP, Taylor NL, Millar AH (2011) The role of mitochondrial respiration in salinity tolerance. Trends Plant Sci 16: 614-623

Kruger NJ, von Schaewen A (2003) The oxidative pentose phosphate pathway: structure and organisation. Curr Opin Plant Biol 6: 236-246

Mannella CA (2008) Structural diversity of mitochondria functional implications. Mitochondria and Oxidative Stress in Neurodegenerative Disorders 1147: 171-179

Millar AH, Whelan J, Soole KL, Day DA (2011) Organization and regulation of mitochondrial respiration in plants. Annual Review of Plant Biology 62: 79-104

Moller IM (2001) Plant mitochondria and oxidative stress: electron transport, NADPH turnover, and metabolism of reactive oxygen species. Annu Rev Plant Physiol Plant Mol Biol 52: 561-591

Patrick JW, Botha FC, Birch RG (2013) Metabolic engineering of sugars and simple sugar derivatives in plants. Plant Biotechnol J 11: 142-156

Plaxton WC (1996) The organization and regulation of plant glycolysis. Annu Rev Plant Physiol Plant Mol Biol 47: 185-214

Stitt M (2013) Progress in understanding and engineering primary plant metabolism. Curr Opin Biotechnol 24: 229-238

Stitt M, Zeeman SC (2012) Starch turnover: pathways, regulation and role in growth. Curr Opin Plant Biol 15: 282-292

Streb S, Zeeman SC (2012) Starch metabolism in Arabidopsis. Arabidopsis Book 10: e0160

Sweetlove LJ, Beard KF, Nunes-Nesi A, Fernie AR, Ratcliffe RG (2010) Not just a circle: flux modes in the plant TCA cycle. Trends Plant Sci 15: 462-470

Sweetlove LJ, Fernie AR (2013) The spatial organization of metabolism within the plant cell. Annu Rev Plant Biol 64: 723-746

Taylor NL, Day DA, Millar AH (2004) Targets of stress-induced oxidative damage in plant mitochondria and their impact on cell carbon/nitrogen metabolism. Journal of Experimental Botany 55: 1-10

Tcherkez G, Mahe A, Hodges M (2011) (12)C/(13)C fractionations in plant primary metabolism. Trends Plant Sci 16: 499-506

Zeeman SC, Kossmann J, Smith AM (2010) Starch: its metabolism, evolution, and biotechnological modification in plants. Annu Rev Plant Biol 61: 209-234

Well-watered plants are turgid, and their leaves and stems are upright and firm, even without woody tissue to support them. If water is lost from leaves through the stomates at a faster rate than it is resupplied by roots, then plants wilt (Figure 3.3)

Well-watered plants are turgid because their cells are distended by large internal hydrostatic pressures (Figure 3.4a). This internal hydrostatic pressure (also called turgor pressure) is typically 0.5 MPa or more. Lack of water causes cells to shrink until the pressure inside equals that of the atmosphere (zero), and the cells thus have zero turgor (Figure 3.4b). The initial shrinkage while turgor drops from 0.5 to zero MPa is determined by the properties of the cell wall: cell walls are slightly elastic, and the relation between volume change and turgor pressure depends on the “elastic modulus” of the wall. This involves little change in whole cell volume for a drop in turgor pressure to zero. However, further water loss causes the wall to shrink and deform inwards, and the whole cell volume decreases markedly.

The turgor pressure of a fully turgid cell may even exceed 1 MPa, about five times the pressure in a car tyre, and ten times the pressure in the atmosphere. In a physically unconstrained cell, the turgor pressure is borne by the cell wall, which develops a large tension within it. But in cells that are physically constrained, such as those of a tree root whose growth becomes hampered by the presence of a slab of concrete, the tension in the cell walls is relieved and the pressure is applied directly to the constraint (Figure 3.5).

It is easy to see how a constrained tree root could eventually lift a slab of concrete: 1 MPa applied over 100 cm² is equivalent to a weight of one tonne. Pressure is force/area, and 1 MPa is approximately equal to 10 kg weight per cm².

A definition of all these terms is summarised at the end of this section (Section 3.1.7).

How is it that plant cells can have such large turgor pressures? The essential reason is that the cells contain large concentrations of solutes. These solutes attract water into the cells through a process known as osmosis, which involves water flowing in through semipermeable membranes that prevent the passage of solutes but not of water. The inflow of water swells the cells until a hydrostatic pressure is reached at which no more water will flow in. In cells bathed in fresh water, such as algal cells in a pond, this equilibrium hydrostatic pressure is known as the osmotic pressure (\( \pi \)) of the cell contents, and is commonly about 500 kPa or 0.5 MPa.

This osmotic pressure can be measured directly with an osmometer, or it can be calculated from the solute concentration in the cell (\( C \)) from the van‘t Hoff relation:

\[ \pi = RTC \tag{1} \]

where \( R \) is the gas constant, \( T \) is the absolute temperature (in degrees Kelvin) and \( C \) is the solute concentration in Osmoles L^-1. At 25 ºC, \( RT \) equals 2.5 litre-MPa per mole, and \( \pi \) is in units of MPa. Hence a concentration of 200 mOsmoles L^-1 has an osmotic pressure of 0.5 MPa.

However, land plants are different from algae in a pond. Their leaves are in air, and the water in their cell walls, unlike the water in a pond, is not free. It has a negative hydrostatic pressure (discussed further in the next section). Thus, for a given osmotic pressure (\( \pi \)) within a cell, the hydrostatic pressure, \( P \), will be lower than if the cell were bathed in free water. This difference is known as the water potential (\( \psi \)) of the cell. It is zero in an algal cell in fresh water, but it is always negative in land plants. Its value is the difference between \( P \) and \( \pi \), that is:

\[ \psi=P-\pi \tag{2} \]

An alternative notation for equation (2) used commonly by plant physiologists is:

\[ \psi_w = \psi_p + \psi_s \tag{3} \]

In this case, \( \psi_w \) is the total water potential, \( \psi_s \) is the solute potential and \( \psi_p \) is the pressure potential. Thus \( \psi_s \) is equal, but opposite in sign, to \( \pi \).

The notion of water potential can be applied to any sample of water, whether inside a cell, in the cell wall, in xylem vessels, or in the soil. Water will flow from a sample with a high water potential to one with a low water potential provided the samples are at the same temperature and provided that no solutes move with the water. Water potential thus defined is always zero or negative, for by convention it is zero in pure water at atmospheric pressure.

Positive values of hydrostatic pressure occur in the living cells of plants, in the symplast, and as explained above are induced by high solute concentrations and the resultant osmotic pressure. However, large negative values are common in the apoplast of plants and the soil they are growing in. These negative values arise because of capillary effects - the attraction between water and hydrophilic surfaces at an air/water interface, the effects of which can be seen in the way that water wicks into a dry dishcloth. This attraction reduces the pressure in the water, and does so more intensely the narrower are the water-filled pores. It accounts for how cell walls, which have very narrow pores, can remain hydrated despite very low water potentials in the tissue they are part of. For a geometrically simple cylindrical pore, the relation between the induced pressure and the radius of the pore can be derived as follows:

Take a glass capillary tube with a radius \( r \) (m) and place it vertically with one end immersed in water. Water will rise in the tube against the gravitational force until an equilibrium is reached at which the weight of the water in the tube is balanced by the force of attraction between the water and the glass. A full, hemispherical, meniscus will have now developed, i.e. one with a radius of curvature equal to that of the tube (Figure 3.6).

The meniscus is curved because it is supporting the weight of the water - much as a trampoline sags when several people are sitting on it. There is a difference in pressure, \( \Delta P \) (Pa), across the meniscus, with the pressure in the water being less than that of the air. The downward acting force (N) on the meniscus is the difference in pressure multiplied by the cross-sectional area of the tube, i.e \(\pi r^2 \Delta P\). The upward acting force is equal to the perimeter of contact between water and glass (\( 2 \pi r \)) multiplied by the surface tension, \( \gamma \) (N m^-1), of water, namely \( 2 \pi r \gamma \) (provided the glass is perfectly hydrophilic, when the contact angle between the glass and the water is zero, otherwise this expression has to be multiplied by the cosine of the angle of contact). Thus, because these forces are equal at equilibrium, we have \(\pi r^2 \Delta P = 2 \pi r \gamma \), whence

\[ \Delta P = 2\gamma/r \tag{4}\]

The surface tension of water is 0.075 N m^-1 at about 20°C, so \( \Delta P \) (Pa) equals 0.15 divided by the radius \( r \) (m):

\[ \Delta P = 0.15/r \tag{5}\]

Thus a fully-developed meniscus in a cylindrical pore of radius 0.15 mm would have a pressure drop across it of 1.0 MPa. The pressure, \( P \), in the water would therefore be -1.0 MPa if referenced to normal atmospheric pressure, or -0.9 MPa absolute pressure (given that standard atmospheric is approximately 100 kPa).  

This argument applies not only to cylindrical pores. It is the curvature of the meniscus that determines the pressure drop, and this curvature is uniform over a meniscus occupying a pore of any arbitrary shape. It is such capillary action that generates the low pressures (large suctions) in the cell walls of leaves that induce the long-distance transport of water from the soil through a plant to the sites of evaporation. The pores in cell walls are especially small (diameters of the order of 15 nm), and are therefore able to develop very large suctions, as they do in severely water-stressed plants. Such pores can hold water against a suction of 10 MPa. (Table 3.1)

In plants, other water-filled pores vary in size from large xylem vessels with diameters of 100 mm or more down to a few mm, so for them to remain water-filled requires that they have no air/water interfaces.

When a cell in an intact plant growing in soil loses water, turgor declines and solute concentrations increase. As explained before (3.1.1), at turgor loss point, when turgor becomes zero, the hydrostatic pressure in the cell sap is equal to the atmospheric pressure, meaning that no net force is exerted on the cell wall, and the plant is wilting. If water continues to be lost from the cell, the pressure within the cytoplasm drops below atmospheric pressure, resulting in a force imbalance that collapses the cell wall. The deformation of living cells upon desiccation is called cytorrhysis. Note that the plasma membrane remains in close contact with the cell wall throughout desiccation ie plasmolysis does not occur, because the hydrostatic pressure in the cytoplasm remains greater than the hydrostatic pressure in the apoplast.

Plasmolysis only occurs in cells that are completely immersed in solution and have no air spaces around them, as in epidermal strips floating on water. Plasmolysis starts when the osmotic pressure of the solution is increased above that of the cells, causing the protoplast to shrink, and the plasma membrane separates from the wall (Fig. 3.7). Large gaps created between the plasma membrane and the wall fill with the bathing solution. This cannot occur in normal tissues as the cells have air spaces between them that are not filled with water. This includes root cells of intact plants growing in hydroponic solution or in waterlogged soil, as they still have air spaces.

Air cannot enter the cell through the cell walls as the small pore size, about 15 nm, would need a suction of 20 MPa to drain the pores (Table 3.1, in previous section) which is impossible.        

During plasmolysis, the plasma membrane is stretched into strands that remain tethered to the wall at particular sites (Figure 3.8). Plasmolysis has been used by microscopists to demonstrate the tethering of the plasma membrane to specific sites on cell walls, by floating tissue such as epidermal peels of onion bulbs on high concentration of solution of sucrose (Figure 3.8). When the protoplast shrinks away from the cell wall, and solution with small molecular weight solutes pentrates the cell wall and floods into the space between the wall and the proplasts, some parts of the plasma membrane stay tethered and the rest become pulled into very fine strands.

The difference between cytorrhysis versus plasmolysis is most easily seen in leaves with single cell layers like mosses. Figure 3.9 shows that in the hydrated leaflet of Physcomytrella, when the central vacuole is distended, the chloroplasts line the cell wall. Rapid water loss causes a general shrinkage and eventually a collapse at the central parts of the cells. In cytorrhysis, the plasma membrane always remains in close contact with the cell wall. In contrast, when cells are bathed in a solution of small molecules like sucrose, glycerol, or low molecular weight polyethylene glycol, PEG, the solutes pass through the cell wall but not the plasma membrane, causing shrinkage of the protoplast and detachment of the plasma membrane from the cell wall. In plasmolysis, the gaps between the cell wall and the plasma membrane are filled with plasmolytic solution (Figure 3.9).

Cytorrhysis also occurs during freezing, when water is withdrawn from cells (Buchner and Neuner 2010).

Plasmolysis is a laboratory phenomenon and does not occur in nature. It is an experimental artifact.

Water flows throughout the plant in three different ways: (a) in bulk, (b) by diffusion in a liquid, and (c) by diffusion as a vapour. Different mechanisms are involved in these three types of flow.

Bulk flow is driven by gradients in hydrostatic pressure. It is much faster than diffusive flow because the molecules are all travelling in the same direction and hence their movement is cooperative. This is the flow that occurs in xylem vessels, in the interstices of cell walls, and in water-filled pores in soil. The resistance to such flow depends very strongly on the size of the flow channels.

Tall trees and fast-growing cereal crops like maize have large xylem vessels, of 100 µm in diameter or more. Flow rates are fast because the rate of volume flow increases in proportion to the fourth power of the radius for a given pressure gradient. Volume flow rate (m³s^-1) in a cylindrical tube of radius \( r \) is proportional to \( r^4 \) and to the gradient in pressure along the tube, and inversely proportional to the viscosity \( η \) (Pa s) (Poiseuille’s Law)

\[ \text{Volume flow rate} = \left(\frac{\pi r^4}{8\eta}\right) * \left(\frac{\Delta P}{\Delta x}\right) \tag{6}\]

Where \( η \) is the solution viscosity and \( ΔP/Δx \) (Pa m^-1)is the gradient in hydrostatic pressure. From equation (5) we understand that wide tubes are enormously more effective than narrow tubes. The importance of the relationship between tube radius and conductive efficiency becomes apparent when we examine long distance transport of water through the xylem (Section 3.2).

Diffusive flow in the liquid phase is driven by gradients in osmotic pressure. It is much slower than bulk flow because the net flows of solute and water molecules are in opposite directions and therefore impede each other. Where two liquid phases are separated by a semi-permeable membrane the flow of water across the membrane to the phase with the higher osmotic pressure is essentially diffusive, and the flow is driven by the difference in water potential across the membrane.

Vapour flow, for example through the stomata, is driven by gradients in vapour concentration, which are usually expressed in terms of partial pressure, but are nevertheless mechanistically concentrations.

As water in the transpiration stream moves from the soil to the roots, through the plant, and out through the stomata, all three types of flow are involved at various stages.

The potential energy of water is affected by gravity: unconstrained water runs down hill. In most plants the effect of gravity is small relative to common values of the water potential, but in tall trees it can dominate. Where the effect is important it is convenient to introduce the notion of a total water potential, \( Φ \), which is the sum of the water potential, \( ψ \), and a gravitational term, thus

\[\Phi = \psi + \rho gh = P - \pi + \rho gh  \tag{7}\]

where \( ρ \) (kg m^-3) is the density of water, \( g \) (m s^-2) is the acceleration due to gravity, and \( h \) (m) is the height (relative to some reference) in the gravitational field. \( Φ \) is constant in a system at equilibrium with respect to water even when height varies. The value of \( g \) is approximately 10 m s^-2, so the gravitational term, \( ρgh \), increases by 10 kPa for each metre increase in height. Hence, at equilibrium, when \( Φ \) and \( π \) are uniform (at least, in a system without semipermeable membranes) the hydrostatic pressure falls by 10 kPa for each metre increase in height.

In the tallest trees, for example a Eucalyptus regnans 100 m tall, equation (6) predicts that, even when the tree is not transpiring, the water potential at the top is about 1.0 MPa lower than at the base.

Definition of Terms

Pressure is force per unit area, Newtons per square meter, or N m^-2. Its unit is the Pascal (Pa). 1 MPa is approximately equal to 10 kg weight per cm².

Hydrostatic pressure is the pressure in a stationary fluid. (Note that hydrostatic pressure is usually quoted as the difference from atmospheric pressure, and is therefore taken to be zero when it equals atmospheric pressure).

Turgor pressure is the term used for the hydrostatic pressure in the cells’ contents.

Osmotic pressure (\( π \)) is the hydrostatic pressure in a compartment containing an aqueous solution that will just prevent pure water at atmospheric pressure flowing into that compartment through its membrane that is permeable to the water but not to the solutes within.

Water potential is the difference between \( P \) and \( π \).

Further Reading

Kramer PF, Boyer JS (1995) Water relations of plants and soils.   http://udspace.udel.edu/handle/19716/2830

Nobel PS (2005) Physicochemical and environmental plant physiology (3^rd edition). Elsevier Academic Press, Burlington, MA

Passioura JB (1980) The meaning of matric potential. J Exp Bot 31:1161-1169

Passioura JB (2010) Plant–Water Relations. In: Encyclopedia of Life Sciences. Wiley, Chichester. DOI: 10.1002/9780470015902.a0001288.pub2

Around 1905, great plans were made to resolve the mystery of the ascent of sap in trees by Professor E.J. Ewart in Melbourne, using eucalypts as a model plant. At that time, Australian mountain ashes (Eucalyptus regnans) vied with American coast redwoods (Sequoia sempervirens) as the tallest trees in the world, being well over 100 m tall. Using special scaffolding, Ewart climbed eucalypt trees, removed lengths of branch and measured the pressures required to push water through these stems. These investigations led him to conclude ‘The ascent of water is, therefore, a vital problem in so far as it depends upon conditions which hitherto can only be maintained in living wood’. If water transport required living cells, it could not be supported by discovery of a pump akin to that in animals. Even roots, which sometimes could pump water by root pressure, lacked the necessary positive pressures to push water so far aloft, especially around midday when water was most needed.

Suction from the shoots was an alternative explanation, but manmade suction pumps cannot do this without inducing formation of air bubbles (embolisms) in the xylem and blocking flow. One clue to the solution came from Dixon and Joly (1894) who claimed that very pure water molecules would be held together by powerful cohesive forces provided the water was especially clean (much cleaner than in manmade pumps).

Ewart did not agree with the unorthodox proposal that the suction of pure water through xylem vessels underpinned transpiration. However, Dixon (1914) ultimately postulated the Cohesion Theory, based on those properties of water which distinguish it as an ideal biological solvent. Cohesion (due to hydrogen bonding between molecules of water), adhesion to walls of the vessels, and surface tension, are central features. In short, in the absence of microscopic gas bubbles water could withstand quite enormous tensions.

Evaporation from wet cell walls of substomatal cavities in leaves creates a large tension (also called negative pressure or suction), which is transmitted via xylem conduits, pulling more sap from roots to leaves. Fine pores in cell walls provide sufficient suction to draw water to the crown of even a lofty tree: a curved interface in a 10 nm pore can store a pressure of -30 MPa. This value can be derived from equation (5) in the previous section: DP = 0.15/r where P is in Pa and r in this case is 5 x 10^-9 m.

Through the evaporative power of the atmosphere, a continuous ‘chain’ or ‘catena’ of water, well below atmospheric pressure, could be drawn up to a leaf canopy. The tensions created in this way could even suck water from the surrounding soil. We now recognise that the evaporative energy is supplied as the latent heat of vaporisation ultimately derived from solar radiation. This cohesive property of water gave rise to the ‘Cohesion Theory for the Ascent of Sap’.

Two other properties of water are also essential for long-distance water transport: surface tension, and the adhesion of water to solid surfaces such as the xylem vessels within trees. Dixon claimed that if water could ‘hang together’, the enormous evaporative energy of the air (the same power which dries the washing hanging on a line) could be harnessed to lift sap, which is mainly water, vertically. This would entail no metabolic energy on the part of the plant. This theory of sap flow accorded with earlier experiments by Professor E. Strasburger in 1893 showing that a tall oak tree trunk, severed at the base, could draw poisons and dyes up to the leaves by some wick-like action. If metabolism energised sap flow, poison should have inhibited it. This was well illustrated in later experiments (Figure 3.11a) in which mercury was drawn through fine tracheids of pine stems purely through the suction created by transpirational water loss from the shoot above. The tension required to achieve this is about 2 MPa.

However, the physical properties of plants had to be more complex than those of simple pipes conducting water. As mentioned, manmade pumps failed through embolisms if used to suck water higher than 10 m, whereas hundreds of litres of water reaches the canopies of tall trees daily. Even overlapping sawcuts in tree trunks, which should allow a massive quantity of air to flow into xylem vessels when under suction and cause trees to die from embolisms, did not stop all water flow to leaves. If water was under such suction, how could trees keep air bubbles out of the sap when the trunk was cut? This additional problem was not resolved until the very complex anatomical structures of trunks were much better understood. The highly compartmentalized, extensively redundant structure of the xylem network performs the critical role of isolating gas voids while water transport continues in adjacent conduits. In reality, the complex structure of the xylem network is what makes reliable water transport under tension possible. 

Xylem is not composed merely of pipes: it is made up of partially sealed units (technically vessels, tracheids and fibres, called collectively conduits), which most effectively limit the spread of introduced gases and thus, maintain water flow in some conduits despite very severe disruption from embolisms in others.

The diameter of xylem vessels can be as small as 10 µm as in Arabidopsis, 60-100 µm in the roots of wheat and rapid growing annuals like maize, to over 100 µm in trees. Remembering that trees can be over 100 m in height, the conductive efficiency of xylem conduits is essentials for plants to move water to the canopy at rates that satisfy the transpirational water loss at the leaf surface. These dimensions are for the vessels with maximum diameter, the late metaxylem in the central part of the stele (e.g. Figure 3.12).

The diameter of xylem vessels in a given species varies greatly with root type (Watt et al. 2008). For example, in wheat and barley, the diameter varies from 10 to 60 µm depending on position within the stele (central or peripheral), and the type of root (seminal or nodal). Figure 3.12 shows a section of a nodal root from barley.

A wider xylem diameter translates to an increase in conductive efficiency that can be appreciated by revisiting equation (6). From the Hagen-Poiseuille Law, which shows that flow increases with the fourth power of the radius, we can see that a four-fold increase in the radius of a tube leads to a 256 fold increase in the volumetric flow rate.

In addition to allowing for high rates of water flow, the xylem must also protect the plant against formation and spread of gas bubbles. For xylem sap to sustain tensions required in tall trees, there must be no gas bubbles in the system. Cohesion breaks down if there is a single ‘nucleation site’ on which bubbles can form and enlarge. On the other hand, sap normally contains dissolved gases which, surprisingly, do not disrupt the system provided there are no nucleation sites available. Even the rigid walls of xylem vessels are compatible with high xylem tensions, attracting water by adhesion, which is essential for transport.

Surface tension acts as an interfacial water–air stopper, preventing air from being sucked into the many millions of tiny pores present in all plant cell walls. For example, water delivered to leaf cells by xylem vessels passes through these tiny menisci, which act effectively as non-return valves, so preventing air from being sucked into the xylem (Section 3.3). Surface tension also explains how water in leaves remains under strain within an essentially porous system through which water flows.

Vascular transport systems have evolved to become amazingly reliable despite the metastable condition of the sap (existing as a liquid below its vapour pressure). From primitive, thickened, hollow cells, increasing specialisation has produced greater elongation and thickening of the tubes (Figure 3.13a,b). Xylem walls contain pits, in which zones of the primary wall known as ‘pit membranes’ allow water to be transmitted between vessels efficiently, while preventing a gas phase spreading through the interconnected system of vessels and blocking transport through embolisation (the blockage of a fluid channel with a bubble of gas) (Figure 3.13d). No living membrane is present in these wall structures. The efficiency with which pit membranes isolate adjacent vessels is shown in Figure 3.11b (in the previous section) where mercury, a highly cohesive liquid, is drawn into specialised bordered pits of pine tracheids without being able to exit into neighbouring tracheids.

Vascular systems have evolved from plant species possessing only fibres and tracheids, for example the more primitive Tasmannia, to more advanced plants possessing vessels which resemble the unicellular tracheids in structure but are much wider and longer and originate from a number of cell initials fused together. Lignin thickening patterns have also evolved. Some thickening designs, such as annular and spiral, allow the tubes to extend longitudinally while supplying growing organs.

When elongation growth has ceased, an organ can be provided with more efficient pipes of larger bore and with stronger thickenings, in reticulate and scalariform patterns (Figure 3.13c). Pit fields which allow water transport across vessel walls can also be simple, unreinforced structures (simple pits) or more elaborate bordered pits in which secondary cell walls mechanically support the pit membrane. All these forms of pits can prevent air in an air-filled conduit from spreading to adjacent conduits which are conducting water under strong suction. Reinforcement of the walls around pits allows pit membranes to be as large as possible and thereby maximise water exchange between vessels.

The late metaxylem carries the bulk of the water to the shoot because of its greater diameter - four-fold increase in the radius of a tube lead to a 256-fold increase in the volumetric flow rate (3.1.5). However, it does not mature until well back from the root tip, and so the younger part of the root is not functional in providing water to the rest of the plant.

Root tips take up enough water for their own cell expansion but they cannot pass this onto the rest of the plant as their xylem vessels are still alive and not able to function as a conduit. Only when they die, mature, and lose the integrity of their plasma membranes, and the cell walls in the transverse plane disintegrate, can they function as conduits.

Xylem conduits (vessels and tracheids) are dead at maturity. They do not mature until sometime after they are fully elongated, and so they remain alive long after they leave the growing zone of the root. The last-formed xylem vessels in angiosperms, the late metaxylem, may be found alive for some distance from the root tip. In Arabidopsis they remain alive until the root hair zone, but in most species they can only be seen further than 5 or even 10 cm from the root tip.

In soybeans, immature vessels were found as far as 150 mm from the root tip (McCully 1994). In barley the late metaxylem vessel (LMX) was still differentiating 100-150 mm from the tip (Huang and van Steveninck, 1988). Light microscopy of hand-cut sections showed the presence of cytoplasmic strands and intact cross walls in LMX up to 100 mm from the tip (Figure 3.16).

Living/immature xylem vessels can be recognised by their high K⁺ concentrations, 100 mM or more.  In contrast, xylem sap that flows through the roots has  K⁺ concentrations of only 5-10 mM, as shown in Table 3.2 in the following section.

When the vessels mature, and their end walls disintegrate, their cellular contents are carried away by the transpiration stream. This leaves hollow tubes that greatly increase the conductance of water flow through the xylem.

Figure 3.17 shows the effect of xylem differentiation and maturity on axial conductance in roots of two crop species, wheat and lupin. Xylem are continually differentiating in lupin within a short region of a young root, which results in dramatic increase in axial conductance. In contrast, once the central metaxylem of wheat has matured there is little change in the root’s axial conductance.

Xylem sap contains all the inorganic nutrients needed for plant growth, in the proportions in which they are needed. The concentrations of some nutrients are dilute when compared to phloem sap, in particular potassium and nitrogen-containing solutes (Table 3.2). The concentrations also vary at different times of day, being lowest in the middle of the day when transpiration is highest, and quite high at night when stomates are closed so there is very little flow of sap to the shoots. The osmotic pressure of xylem sap therefore ranges from less than 0.05 MPa during the day to about 0.15 MPa (60 mOsmol L^–1) during the night. The flux of solutes (the concentration multiplied by the flow rate of the sap) is maintained at a steady rate over the whole 24 hours.

Most solutes in xylem sap are inorganic ions (e.g. nitrate, potassium, magnesium and calcium), but organic solutes are also present (Table 3.2). Organic acids and amino acids in xylem sap can be present in substantial concentrations, and sugars (but not sucrose) reaching 5 mM in some perennial species. Many trees including eucalypts are host to boring insects at particular times of the year, when sugar and nitrogen content of the sap is nutritionally valuable. Even though sugar concentrations in xylem sap are much lower than in phloem sap, the high nitrogen to sugar ratios and low osmotic pressures make it a good substrate for many herbivores. More extreme examples of the carbohydrate content of xylem sap are temperate deciduous trees such as maple, which have traditionally been tapped to yield a sugary solution in the period prior to budburst. This indicates that xylem can be a conduit for carbon remobilisation in addition to its central role as a pathway for water and nutrient transport. Xylem parenchyma cells load ions into xylem vessels in roots (Section 3.6) and also contribute to modification of ion levels along the xylem pathway. In secondary tissues, rapid transfer of solutes into and out of the xylem is partly achieved through close association of living ray cells and xylem vessels.

Other organic molecules act to transport inorganic nutrients to the shoots. Nitrate and ammonium are assimilated into organic forms, such as amino acids, in the roots of many plants. In legumes, nodules deliver an even wider selection of nitrogenous compounds to the xylem, including ureides and amides. These often constitute the dominant form of nitrogen reaching shoots and are therefore, a major component of the sap. Other examples of complexed forms of inorganic nutrients in xylem sap are metal ions such as zinc, copper and iron, which are almost exclusively chelated to organic acids.

Phytohormones such as abscisic acid and cytokinins are present in xylem sap and serve as signals from roots to shoots affecting growth and development, and other physiological responses.

The simplest vein architecture is found in conifer needles where a single unbranched strand of xylem and phloem is surrounded by mesophyll (Figure 3.22).

Vascular strands are enclosed by an endodermis that separates them from the mesophyll, and are embedded in a mixture of parenchyma cells and tracheids called transfusion tissue. Water from the xylem permeates radially outward through transfusion tissue, endodermis and mesophyll to evaporate below lines of stomata in the epidermis.

Leaves of angiosperms have much more complicated venation than conifer needles. If you look at a grass leaf with your hand lens, parallel veins run the length of the leaf, but they are not all the same size. A few large veins have several small veins lying between them. On closer inspection with a light microscope, all these parallel veins are connected at intervals by very small transverse veins (Figure 3.23).

There are in fact two vein systems with different functions: large veins supply water rapidly to the whole length of a leaf blade while small veins and their transverse connections distribute water locally, drawing it from the large veins. Water in large veins flows only towards the tip, but in small veins it can flow either forwards or backwards along the leaf blade or transversely between adjacent parallel veins. The distinction of flow patterns in large and small veins arises as a result of different vessel sizes. Large veins have wide vessels (about 30 µm diameter), while small veins have narrow vessels (about 10 µm diameter). As the volume of water flowing along pipes is proportional to the fourth power of the radius (Equation 6; Poiseuille’s Law), volume flow in the larger vessels will be 3 to the fourth power (= 81) times greater than the flow in the smaller vessels. Put another way, pressure gradients along the leaf in large vessels will be very slight, but steep pressure gradients can develop locally in narrow vessels that will direct local flows into the mesophyll. Large veins supply water rapidly over the whole lamina while small veins distribute it locally and slowly. The slower flux along minor veins is compensated by their far greater number, which results in them having a greater length per unit leaf area than major veins.

With a hand lens, you do not see the vessels, instead you see the sheaths that surround xylem and phloem, much as an endodermis surrounds vascular strands of a conifer needle. Grass leaf veins have two sheaths of cells surrounding the parallel veins and containing the xylem and phloem tissues. In the leaf anatomical development typical of C₃ species (e.g. wheat), both sheaths are parenchymatous and lack chloroplasts (Figure 3.24).

The large veins of a C₃ species is surrounded by two sheaths of living cells, the inner mestome sheath and outer parenchyma sheath (Figure 3.24). The mestome sheath of these large veins is impermeable to water. There is no apoplastic path for water through the mestome sheath of large veins, except through a connecting transverse vein. In small veins, by contrast, two or three mestome sheath cells next to the xylem permit flow of water and solutes through the cell wall apoplasm.

In C₄ species (e.g. maize), only the inner (mestome) sheath is without chloroplasts. The outer ring of sheath cells contains large chloroplasts and is known as the bundle sheath (Figure 3.25). This is the cell layer in which CO₂ fixation in the Calvin cycle takes place in C₄ plants. This is an essential part of the carbon concentrating mechanism, and the special anatomy of C₄ photosynthesis, as detailed in Chapter 2, Section 2.2.2

A dicotyledonous leaf contains the same two vein systems as a grass leaf, but these are differently arranged. Large supply veins are prominent, comprising a midrib and two orders of branches off it, often standing out from the surface of the lamina. These contain wide vessels and carry water rapidly to the leaf margins. Between them lie distribution veins, another two branch orders of small veins dividing the mesophyll up into islets about 1–2 mm across (Figure 3.26), and within these islets a fifth and final order of branches of the finest veins. The fourth- and fifth-order branches have only narrow vessels (Figure 3.26).

In dicotyledenous leaves, as in grass leaves, these vascular tissues are enclosed by bundle sheath cells through which water and solutes must pass when leaving the xylem.

Any distribution network such as the branching vessels of decreasing size in leaves is found to obey Murray’s Law. This states that the cube of the radius of a parent vessel is equal to the sum of cubes of the radii of the daughter vessels (e.g. a 50 µm vessel would branch into five 30 µm vessels). Such a pattern of branching produces optimal flow in several senses: minimum energy cost of driving that flow, minimum energy cost of maintaining the pipeline, constant shear stress at the walls of pipes, and rapid flow in supply pipes with slow flow in distributing pipes to permit exchange through the pipe walls (LaBarbera 1990).

The ring of cells forming the sheath around the xylem and phloem tissue (Figure 3.27) acts both as a mechanical barrier that may confine pressure within the vein, and a permeability barrier that can control rates and places of entry and exit of materials. Exceptions are to be found at the ultimate ends of some dicotyledonous fine veins where tracheids or sieve elements may be unaccompanied by other cells, in the transverse veins of grasses which have no sheath, and in some special veins at leaf margins where the sheath is absent on the xylem side. There is evidence of a suberised layer in walls of these sheaths cells in some species, but not in others. It is uncertain whether xylem sap must traverse the cell membranes, as in the suberised endodermis of roots, or if it can travel along the cell walls.

Transpiration operates by suction, but leaves are especially liable to damage by grazing, mechanical forces, and extreme negative pressures in the xylem water column due to high rates of transpiration. Leaf vessels therefore need special protection against air embolisms spreading in the vessel network that would block liquid flow. This is achieved at all points in the leaf distal to the node (i.e. petiole, large veins, small veins) by the vessels being very short. That is, files of vessel elements joined to make a single pipe with a terminal end-wall are much shorter in the leaf than in the rest of the plant. Water flows through vessel end-walls with little extra resistance, but an air–water interface cannot be pulled through an end-wall or pit membrane because the pores are so small. The force needed to curve the interface into a meniscus small enough to pass through the end-wall is similar to that generated by evaporation from wet cell walls (Section 3.1.3). From equation (4), (DP = 0.15/r) we see that a pressure of roughly 0.3/d (MPa) is required to pull the interface through a hole of diameter d (µm). While 0.1 MPa can pull an interface through a 3 µm hole, 6 MPa is required to pull air through a 0.05 µm hole. Suctions of 6 MPa are not known to occur in transpiring plants, and cell walls have pores much smaller than 0.05 µm - in the order of 0.003 µm. So an embolism formed from cavitating water fills one vessel but does not spread beyond it.

The length of xylem vessels is demonstrated by allowing a leaf to transpire in a fine colloidal suspension that cannot pass end-walls. Latex paint, diluted 100 times with water and allowed to settle for a week or two, provides such a suspension. Leaves that have drawn up this suspension for an hour or so during transpiration can be cleared by dissolving out the chlorophyll and soaking in lactic acid. Progress of the paint is then readily seen (Figure 3.28). Very few vessels exceed 1 cm in length.

Xylem vessels (and tracheids) are not just pipes to carry water, they are pipes with holes in them (pits) through which water can leak out, fulfilling a principal leaf function of water distribution through the transpiration stream to places where it will evaporate. The branching network of vessels is beautifully adapted to achieve this.

Think about flow in a leaky vessel. As explained earlier, the rate of volume flow varies as the fourth power of the radius (Poiseuille’s Law). The frequency of leaks through vessel walls varies with surface area of the walls, that is, as the first power of the radius (2πr). So if the vessel is wide, forward flow is much larger than leakage. The wide vessels in large veins supply water all over the leaf without losing much on the way. As the width of a vessel becomes smaller, the forward flow (a function of r⁴) is reduced much more strongly than the leaks (a function of r). The proportion of water lost to leakage therefore increases as vessels become smaller. Indeed, for a fixed pressure gradient there is a critical radius at which all water entering the vessels supplies leaks, and there is no forward flow at all. The finest veins of dicotyledenous leaves have vessels of a radius that is close to this critical value. As sap disperses into the fine ramifications of the network, it moves more and more slowly forward, and leaks increasingly outwards through the sheath to the mesophyll. This is the rationale of the distributing networks of the small branched veins of both dicotyledons and grasses.

Extraction of water from fine veins can be readily demonstrated. Stand a cut leaf in an aqueous solution of dye, such as 0.1% sulphorhodamine G, and allow it to transpire for an hour. The solution moves rapidly through large veins all over the leaf in a few minutes. Then it moves increasingly slowly into the network of distributing small veins. By the end of an hour it has reached the ends of the finest veins and, as water is extracted from them, the dye becomes more and more concentrated (Figure 3.29). Movement of dyes from the finest veins to leaf surfaces 100 µm away takes about 30 min, suggesting that diffusion rather than mass flow is responsible for solute distribution to cells.

(a) Phloem export

Potassium is known to be extracted from the xylem and re-exported via the phloem (Section 5.1). In leaves, the concentrated sap flowing through the narrow xylem vessels of fine veins is separated from sieve tubes of the phloem by only two or three parenchyma cells, frequently enhanced by transfer cells. Exchange from xylem to phloem is probably made by this direct path. Solutes travel in the phloem back to stems, either distally to younger developing leaves and the stem apex or proximally towards roots.

(b)  Scavenging cells  

Recycling of solutes out of leaves can also be fed by a variety of special structures adapted for processing rather larger volumes of sap, and so suited for collecting solutes present in the stream at quite low concentrations. The transfusion tissue of conifer needles is one of these structures. As sap leaves the xylem of a vascular strand it moves through a bed of tracheids mixed intimately with transfusion parenchyma cells (Figure 3.22 shown earlier). The endodermis acts as the ultrafiltration barrier, allowing water to pass through while leaving dilute solutes to accumulate in transfusion tracheids. Transfusion parenchyma cells have very active H⁺-ATPases in their cell membranes which accumulate selected solutes (certainly some amino acids) back into the symplasm for return to the phloem and re-export. Such actively accumulating cells are called scavenging cells.

A tissue that acts in the same way is a special layer of cells in the central plane of many legume leaves (extended bundle sheath system or paraveinal mesophyll). It consists of scavenging cells with active H⁺-ATPases and accumulates amino acids, stores them and forwards them to developing seeds via the phloem.

Jagged ‘teeth’ on the margins of many leaves also contain scavenging cells (Figure 3.33). Veins carry large volumes of the xylem sap to these points, where evaporation is especially rapid. Within a ‘tooth’, xylem strands end in a spray of small vessels among a bed of scavenging cells. Scavenging cells can thereby collect amino acids and load them into the phloem.

(c) Solute excretion

Not all the solutes of the transpiration stream are welcome back in the plant body. Some, such as calcium, are immobilised in insoluble compounds (calcium oxalate crystals) and shed when leaves fall. Others are excreted through the surface of living leaves. A striking excretion system is found along the margin of maize leaves. Here the outermost vein has a single very wide vessel. Rapid evaporation from the exposed leaf edge cooperates with this low-resistance vessel to draw to the leaf margin all residual solutes that have not been taken out of the stream by other veins. Thus foreign material (like dyes) accumulates in this outermost vessel. The vein sheath is missing from the outer part of this vein so that vessels abut directly the airspace at leaf margins (Canny 1990). Solutes are excreted from this marginal vein, dissolved out by rain and dew, and, more actively, at night time by guttation fluid when there is positive pressure in the xylem. A similar accumulation of dye from the transpiration stream is shown along the margin of a eucalyptus leaf in Figure 3.34.

 

Water drains through soil due to gravity, leaving medium-sized pores still filled with water, and with thin films around soil particles. Soil is porous and holds water in its pores by capillary forces. As a soil dries, the larger pores drain and the remaining pores hold water ever more tenaciously. Water in these pores is under suction (negative hydrostatic pressure, P) and this suction typically ranges from about 10 kPa to about 1.5 MPa in soils supplying water to plants. At suctions of less than 10 kPa, water is held in such large pores that it is likely to drain quickly away; at suctions greater than 1.5 MPa, most plants are at their limit for exerting sufficient suction to extract the water.

Field capacity

When substantial rain falls on a freely draining soil, it may fill all the pores, even to the point of run-off. Thereafter, water quickly drains out of the largest pores under the influence of gravity into the drier soil below. This process of redistribution continues for days, even weeks, but it does so ever more slowly as the remaining largest pores continue to drain. During this time the hydraulic conductivity of the soil falls by several orders of magnitude because the resistance to viscous flow through water-filled pores depends inversely on the 4^th power of their diameter.

After about 2 days of redistribution, the rate of drainage typically drops to only a few mm per day, which in the field is similar to other forms of water loss from the soil: uptake by roots, or evaporation from the soil surface. At this point there will be a suction (negative pressure) in the soil water of somewhere between 10 and 30 kPa, and the water content then is known as “field capacity” or “drained upper limit”.

Soil types

Soil scientists classify soils according to particle size as:

Gravel:	> 2 mm
Sand:	0.05 – 2 mm
Silt:	0.002 – 0.05 mm
Clay:	< 0.002 mm (2 µm)

The structure and texture of the soil determines how much water can be held in it. Sand particles are large (50 mm - 2 mm) and hold onto little water, but a sandy soil provides good aeration. Sand holds little water because the pores between the grains are large. Therefore, its field capacity, the amount of water in the soil remaining a day or two after irrigation, can be as low as 10% (weight per volume; or 10 mL water per 100 mL soil volume).

Clay particles are much smaller (less than 2 mm) and easily compacted. That makes clay a good material for building bricks, but not so good for allowing water, air, and plant roots through. Clay particles have stacked platelets which provide large amounts of surface area. Clay soils store large amounts of water and have a field capacity above 40%, though much of that may be in pores too small for plants to extract water, as shown in the table below.

Silt is the medium size particle (2-50 mm), with better water retention than sand, but less nutrients than clay. Loam is a soil comprised of roughly equal amounts of sand and silt and a little less clay.

Different types of soil are comprised of various proportions of these particle sizes, as illustrated in Figure 3.37.

The diameter of pores in soil is roughly equal to one fifth of the diameter of the soil particles. Thus, at the boundaries between these classes, the approximate diameters of the pores would be as shown in Table 3.3 with the suction needed to drain them.

This table shows that many pores in a gravelly soil would drain at a suction of only about 1 kPa, but that a clay soil would need ten times this suction, close to 1 MPa, to drain many pores. Transpirational pull by plants can provide suctions above this, but there is a limit at which plants can maintain suction without the risk of embolism.

Wilting point

This term is used by crop physiologists, and often called the “lower limit of water extraction” by agronomists and crop modellers. When the water content of a soil is below the wilting point, plants have difficulty accessing it, as they cannot generate a sufficiently low water potential. In practice, this critical water potential has been found to be about -1.5 MPa, which effectively marks the lower limit of the available soil water. 

Soil at wilting point is not dry. Different soil types hold very different amounts of moisture at their wilting points. As plant remove water and the soil starts to dry, the small pores make it difficult for roots to extract all the water, so clay soils hang onto their water more strongly than sandy or loamy soils, and results in the plant available water being less for clay soils than for loamy soils which have a lower clay content (Table 3.4).  Silt contains pore sizes intermediate to clay and sand, and makes up about 50% of the composition of loamy soils (Figure 3.37) and give it the favourable property of retaining more water than sand as the soil but releasing it more easily than clay.

The relationship between soil water content and soil water potential (Ψ_soil) is shown in Figure 3.38 for the three different soil types. Water is held with progressively stronger suctions as the soil dries, and at the permanent wilting point of -1.5 MPa clay soils retain much more water than sands or sandy loams that is not available to plants.

Another feature of clay soils that makes them less desirable for agriculture or horticulture is that root length density (length of roots per unit volume of soil) tends to be lower than in the lighter textured loamy and sandy soils. Despite their substantial moisture reserves, fine-textured clay soils are generally less hospitable to plant roots than loamy or sandy soils. As a result, plant-extractable moisture in clay soils is somewhat less than the soil’s physical properties alone would imply.

Field capacity is not a term relevant to pot experiments, unless they are very tall. After irrigation, a drained pot will contain much more water than soil in the field.

The soil at the bottom of a freshly watered and drained pot is inevitably saturated with water; if the pot is short the whole of the medium may have such little air-filled porosity that it becomes hypoxic. This is a special problem with field soils used in pots, for these typically do not contain many pores large enough to be drained at the small water suctions that prevail. Such suctions are zero at the bottom of a freshly drained pot and rise by 1 kPa for every 100 mm increment in height, leading to suctions at the top of a typical pot that are much less than the 10–30 kPa that occurs in soils in the field that have drained to ‘field capacity’ (Passioura 2006).

When a pot has finished draining, the soil at the bottom is saturated: it has zero water suction. The suction varies linearly with height, as illustrated in Fig. 3.39, from zero at the bottom of the pot, to 10 H Pa, where H is the height of the pot in mm. Thus, a freshly watered and drained pot that is 100 mm high has a water suction at its top of 1 kPa, very much less than the 10–30 kPa that occurs in a freshly watered and drained soil in the field. Similarly, a pot that is 200 mm high has a suction of 2 kPa at the top. Thus the average suction and thence average soil water content of a freshly watered and drained pot depends on its height.

It is worth noting that the size and number of drainage holes in the bottom of the pot have no bearing on the distribution of suction with height at equilibrium. The rate at which equilibrium is attained will also depend little on the number and size of the drainage holes — just one hole of a few mm diameter is adequate.

The implication of these small suctions is that much larger pores contain water than in a drained soil in the field. For example, at the top of a pot that is 100mmhigh, and where the suction is 1 kPa, all pores less than 0.3 mm wide will contain water. This contrasts greatly with the field soil, in which the suction will be at least 10 kPa and the diameter of the largest water-filled pore will be 0.03 mm or even less.

This has considerable implications for doing pot experiments on the availability of soil water to plants using field soils. Standard watering techniques, which result in adding excess water that then drains from pots, lead to initial soil water contents that are much more than they would be in the freshly drained soil in the field. To match what happens in the field, it would be necessary to water the pots by weight to reproduce a water content consistent with a suction somewhere between 10 and 30 kPa. Further, the large initial water content can produce problems with aeration.

The plant nursery industry is well aware of the difficulties in adequately aerating pots. That is why their potting mixes contain peat or vermiculite or other bulky materials that create large pores of 1mm or more in diameter, which contain air at heights greater than 30mm above the bottom of a freshly drained pot. These large pores protect the roots from hypoxia in a material that might otherwise have a dangerously small air-filled porosity. Nevertheless, despite these large air-filled pores giving the roots growing in them access to oxygen, the interior of aggregates in the soil or potting mix will still be essentially saturated and possibly hypoxic.

Data for three different types of potting mix are shown in Figure 3.29: horticultural topsoil, commercial potting mix is designed for growing plants for sale in garden nurseries, and fine potting mix designed for growing Arabidopsis thaliana in small pots. The topsoil is the worst aerated, with the bottom 150mm of a freshly drained pot in danger of becoming hypoxic. It is notable that in the small pots (70 mm tall) often used for growing Arabidopsis, all of the medium is in danger of becoming hypoxic, for the porosity ranges from 0 to only 7%within the pot (Figure 3.40). With a frequent watering regime, say twice daily, the medium could be permanently hypoxic. Commercial potting mixes overcome this problem by using coarse materials in the mix, which create many large pores (>1 mm diameter) that drain at small suctions.

Field capacity should therefore not be confused with the water content of a drained pot, which should be called “pot capacity”.

Flow of water through soil is induced by gradients in hydrostatic pressure, P. The rate of flow, F (m s^–1), depends on both the gradient in P and on hydraulic conductivity, K  (m² MPa^–1 s^–1), of the soil:

\[F=K \frac{dP}{dx} \tag{8}\]

where x (m) is distance. This equation is Darcy’s Law. Conductivity, K, varies enormously, by about a thousand-fold, over the range of available water content. The reason for this large range is that water flows much more easily in a large pore than in a small one. As explained earlier (Section 3.1), the flow rate varies according to Poiseuille’s Law — the flow in capillary tubes depends on the fourth power of the diameter. As the soil dries, the large pores are empty and water is drawn from smaller and smaller pores.

Water removed by transpiration results in drier soil around roots compared with bulk soil. As soil dries near the root surface, water flows radially from bulk soil to replenish it. Calculated distributions of water content and pore water pressure with radial distance from an absorbing root predict a pronounced increase in suction adjacent to the root.

(a) Theory of water uptake

Roots are cylindrical, so the flow of water to them is radial. This has considerable consequences: the flow lines converge as they approach the root, which requires increasing gradients of pressure in the soil water to keep the water flowing (Figure 3.41).

A radial flow of water from the bulk soil towards roots of transpiring plants is generated by suction at the root surface. However, because K decreases with falling water content, there is a limit to how fast roots can extract water from soil. Once this limit has been reached, increasing suction by roots simply steepens the gradient in P to match the fall in K close to root surfaces so that the product of the two (Equation 3.1) remains the same.

Although soil water is driven by gradients of pressure, it is necessary when water content is changing to describe this flow in terms of gradients in volumetric water content, θ (m³ m^–3). The coefficient relating flow rate to the gradient in water content is known as diffusivity, D (m² s^–1), and the appropriate equation is formally analogous to Fick’s First Law of diffusion:

\[F=D \frac{d\theta}{dx} \tag{9}\]

This equation can be elaborated to allow for the cylindrical flow, and then solved to derive an approximate relationship for the uptake of water by a population of roots:

\[Q \approx 2D*\Delta \theta * L \tag{10}\]

where Q is the flow rate of water through the soil, (m³ m^–3 s^–1), now expressed as the overall rate of change of θ in the sample soil, and L is the rooting density, average length of absorbing root per unit volume of soil (m m^–3).

Like K, D varies with the soil water content, although not so widely.  Laboratory measurements of  D, which are so far the only ones that have been made with some accuracy, show that D is about 10^-7 m²s^-1 when the soil is fairly wet, and falls to about 10^-9 m²s^-1 once the soil has dried far enough for the soil water suction to exceed about 300 kPa.  Where D is of the order of 10^-9 m² s^-1 one can calculate from eqn 3.3 that wherever the density of roots exceeds about 1 cm per cm³ of soil (10⁴ m m^-3) the flow of water through the soil is not likely to limit uptake until there is almost no available water left in the soil. But where there is less than about 1 mm of root per cm³ (10³ m m^-3) of soil, uptake is likely to be sufficiently limited by flow through the soil that it would take several days for the roots to extract most of the available water.

 (b) The uptake of water by roots - in practice

Many experiments have been done to explore this theory, with mixed results.  Most have involved growing plants in repacked soil in controlled laboratory conditions. Some have confirmed the theory, many have not. The most likely reasons for the discrepancies is that interfacial effects - at the junction between the surface of the root and the soil - come into play.  These are discussed in the next section.

In field soil, the concentration of roots in the topsoil is usually so high that the local rate of uptake of water is unlikely to be ever limited by the properties of the soil.  It is in the subsoil that the rooting density drops to a level at which flow through the soil may limit the rate of uptake. Further, the rate of uptake is often much lower than what the simple theory would predict, even when the often sparse rooting density is taking into account.

In subsoil, there are other possible reason for the discrepancy between theory and observation.  One is that, in contrast to disturbed soil, either in ploughed topsoil or repacked in lab experiments, roots in the subsoil do not ramify more or less randomly through the soil. Subsoils are usually dense and difficult to penetrate.  Roots grow predominantly in pre-existing fissures or in continuous large pores, biopores, made by previous roots or soil fauna (Figure 3.42). 

A second reason is that it may be wrong to extrapolate laboratory measurements of D in repacked soil to undisturbed soil in the field. The undisturbed structure of the soil may inhibit the flow of water.  For example, the formation of aggregates of particles in the soil often results in particles of clay (which are usually in the form of small plates) becoming oriented parallel to the surface. Such orientation is likely to increase greatly the flow path for the water, but so far no reliable measurements of D have been yet made on undisturbed subsoil.

In summary, the dense root systems common in topsoils extract water effectively from surface soil layers. Extracting water from subsoil layers is more difficult. Australian subsoils are typically inhospitable to roots. They are dense, have a large resistance to penetration. They are often sodic, that is, sodium dominates the exchange complexes on soil particles, altering the soil structure so that it sets like concrete when dry, and becomes impermeable to water when wet. Moreover, subsoils can be acutely deficient in some nutrients that are required locally by roots. Native vegetation overcomes these difficulties by forming deep biopores in the subsoil. For example, roots of jarrah trees can create and maintain a path to water deep in the subsoil, possibly even as far as a water table 20 m below the surface.

The junction between root and soil is more than just two surfaces touching.  It is often marked by a mucilaginous layer, rich in bacteria and fungi, that may be a few hundred micrometres thick, and in which adjacent soil particles are half or even wholly embedded. This is called the rhizosphere (Chapter 4). The hydraulic conductivity of this layer is not known, but there are two properties of this interfacial zone that may influence the flow of water across it.  The first is that extensive gaps may occur within it, which strongly impede the flow.  The second is that any exclusion of ions from the soil solution may result in large increases in the concentrations of these ions at the surface of the roots and this, because of the osmotic effects, would also impede the flow of water.

Water usually flows from soil to root as liquid, carrying with it dissolved nutrients.  But there may be times when the hydraulic continuity between root and soil is so poor that a substantial proportion of the flow of water is as vapour.  So far, nobody has managed to measure such flows, or even the relative importance of liquid and vapour flow, but their possible importance can be appreciated with the help of a few calculations.

The flow of vapour through soil is by diffusion, in contrast to that of liquid water, which moves in bulk.  The consequence of this, and the fact that the vapour pressure of water is close to saturated in moist soil so that gradients in it are slight, is that water flows as vapour many times more slowly, in response to a given difference in water potential, than it does as liquid.

There are two main reasons why a root may be in poor hydraulic contact with the soil.  The first is that it may be growing in a pre-existing pore larger than itself, so that it makes only glancing contact with the wall of the pore (see Figure 3.42 above).  The second is that it has shrunk within a pore that it had itself made and which it had previously fitted snugly.  What would induce a root to shrink?  A fall in the water potential of the cortex would do so, for the cells there are thin-walled and separated in part by intercellular gas spaces, so they are likely to buckle.  Observations made in rhizotrons (glass-walled tunnels for observing the behaviour of roots in the field) clearly show a diurnal shrinkage in cotton roots of up to 40% where the roots are growing in large pores, but we still do not know, because they are so difficult to observe, whether roots growing in intimate contact with the soil are similarly prone to shrink.  A few observations made using neutron autoradiography have shown no shrinkage in roots growing in the field.

Because of the small size of the interfacial zone, it is very difficult to measure concentrations of ions within it.  Rough estimates have been made by carefully separating the roots from the bulk of the soil, leaving only tenacious soil attached, then shaking off this remaining soil and analysing it.  Such analyses do show somewhat higher concentrations (compared with those in the bulk soil) of ions such as sodium that we would expect to be largely excluded by the roots.  But because the gradients in concentration are likely to be very large close to the root, these analyses are hard to assess for their consequences on the water relations of the plant.  The gradients are likely to be especially large when the soil is becoming fairly dry, for the diffusion coefficient for the solutes falls by a few orders of magnitude as the soil dries, so that any excluded solutes will diffuse away from the root surface only very slowly.

 

Water and ions may move through root tissues along one of three routes: symplastic, apoplastic, or transmembrane, the latter also known as the “transcellular pathway” (Figure 3.47). An update on the anatomical and molecular bases of these pathways in Arabidopsis is given by Barberon and Geldner (2014).

The proportion of water flowing through the different pathways undoubtedly varies with the plant’s rate of transpiration.

Apoplastic Flow

Water and soil solutes might move through the cell walls of the cortical cells until the endodermis, where they would cross the plasma membrane of the endodermal cells. Water would cross via aquaporins, and solutes via ion channels or transporters (see Section 3.6). This is possible in roots without a suberised hypodermis and when the endodermis is in the primary stage of development, as described in the following page. Entry of anions to deep layers of the cortex is likely to be restricted by charge repulsion from dissociated, negative carboxyl groups in cell walls (Donnan Free Space). In general, cations also pass through cell walls more readily than anions, particularly if many of the carboxyl groups in cell walls are not occupied by Ca²⁺ ions. Nonetheless, apoplastic flow of water through roots can sustain large ion fluxes during periods of high transpiration.

It is likely that the bulk of the soil water taken up by the plant moves in the apoplast across the cortex, but that solutes are taken up by epidermal or outer cortical cells and then move in the symplasm across the cortex.

When the endodermis is suberised, water and solutes cannot enter from the apoplast and instead are taken up into the adjacent cortical cells and move via plasmodesmata into the endodermis.

Symplastic Flow

Symplastic flow occurs when water or solutes are taken up by the epidermis or an outer cortical cell, and then subsequently cross the rest of the cortex via plasmodesmata. Plasmodesmata are extensions of the plasma membrane that cross cell walls and provide cytoplasmic continuity between cells, allowing neighbouring cells to communicate and exchange materials more freely. Water and ions can move through the cortex via a series of plasmodesmal connections, thereby remaining in the cytoplasm until reaching the stele. Rate of water and nutrient transport is largely regulated by plasmodesmal resistance, which depends on size and frequency of plasmodesmata in the cell membranes. Alternatively, water and some ions enter vacuoles and are therefore, subject to transport properties of the tonoplast.

Transmembrane (transcellular) Flow

A third possibility is the transmembrane or transcellular route, in which water and nutrients cross membranes, passing repeatedly between symplasm and apoplasm as they are transported from one cell to the next. For nutrients, this form of transport is facilitated by carriers that are distributed in a polar fashion, with influx carriers located on the outer walls and efflux carries on the inner walls. Water transport across cell membranes is via aquaporins for which there is no “rectification”, or intrinsic directionality of transport. They are therefore most likely to be distributed evenly along the plasma membrane.

This pathway may increase resistance to flow of water, but also provides a more effective mechanism for controlling the flow of solutes and water across the root than the purely symplastic route via plasmodesmata when only the first membrane crossing is available to control ion and water transport.

The epidermis and root hairs probably mediate most of the selective uptake of solutes from the soil solution. Nutrients are selectively taken up, and potential toxins excluded.  The cortical cell layers continue this selective uptake if solutes travel apoplastically towards the stele.

The endodermis is the innermost cortical layer that surrounds the central vasculature, and forms a barrier preventing water flow and free diffusion of solutes in the apoplast because of its Casparian strip and suberin lamellae. It thus, prevents soil solution moving into the stele without crossing a membrane, and also prevents water from within the stele being lost back to the soil at night.  Although, in some regions of the endodermis that are opposite xylem poles, water and solutes may bypass the endodermis through unsuberised “passage cells”. The endodermis also functions in structural support for the stele, particularly in drying soil, and minimises shrinkage or swelling of the cells of the stele. Its role in ion selectivity is minor compared to the cells of the cortex and epidermis.

In many plants, the cortical layer under the epidermis (the  hypodermis) develops Casparian bands and suberin lamellae, and develops into the exodermis. This forms another barrier for the apoplastic movement of water and solutes. Just like in the endodermis, some passage cells in the exodermis may be free of suberin and can take up water and solutes.

(a) Endodermis

An endodermis with a Casparian strip is a feature of roots of all land plants from ferns upwards. As the innermost cortical layer that surrounds the central vasculature of roots, the endodermis acts as a barrier to the free diffusion of solutes from the soil into the stele, and the stele into the soil. The protective functions range from efficient water and nutrient transport to defence against soil-borne pathogens. The genes and regulation mechanisms that drive the differentiation of this intricately structured barrier have been reviewed by Geldner (2013).

The development of the endodermis has three stages: the primary stage in which the Casparian strip forms, the second stage when suberin lamellae encases the entire endodermal cell, and the tertiary stage when the inner tangential walls thicken and a layer of cellulose is deposited over the suberin lamellae.

The Casparian strip is made of lignin and suberin and deposited as a ring in the radial walls of the endodermal cells, like a hoop around a barrel of beer. It reaches from the plasma membrane to the outermost part of the wall and adjoins adjacent cells so there are no air spaces between the cells at this point. It therefore blocks the flow of water through the cell walls. It differentiates as root cells mature about 5-10 mm from the tip, and so an entirely apoplastic pathway from soil to central stele can occur only in very young root parts or at sites where the Casparian strip is disrupted such as site of lateral root development, which starts at the pericycle, the cell layer beneath the endodermis.

In the first stage of endodermis development, the Casparian strip in the primary wall prevents flow of water and solutes through the wall from inner cortex to stele. Any solutes remaining in the apoplastic water must enter the endodermis via membrane transporters. Water in the apoplast must enter the endodermis via aquaporins as illustrated in the lower pathway shown in Figure 3. 50. 

In the second stage of development, the endodermis becomes suberised as the secondary walls develop, and suberin lamellae are formed all over the cell, underneath the primary wall that contains the Casparian strip. This occurs at varying distances from the root tip depending on species and is often induced under drought to form in younger parts of the root.

Suberin is a hydrophobic polymer, deposited in the secondary cell wall in lamellae. It therefore seals off the plasma membrane from solutes, as water and ion channels are sealed. So the transcellular and apoplastic pathways are curtailed, and water and solutes enter the endodermis only through plasmodesmata from neighbouring cells (Figure 3.51). Any function in selective control of particular nutrients into or out of the endodermal cell through the suberin lamellae is unclear.

Some endodermal cells remain at the primary stage even late in root development, and are called passage cells.

Stage 3 of the endodermis involves further deposition of cellulosic wall material, further impeding flow of solution through walls.

(b) Exodermis

Exodermis is the name given to a hypodermis with Casparian strips and suberised lamellae.

Occurrence

Roots of most species form an exodermis with time. First, a Casparian band is laid down in the primary wall of the hypodermis, then all the wall is suberised especially the inner wall, as for the endodermis. Some cells in this hypodermal layer (‘passage cells’) remain unsuberised.

The development of the exodermis is very common in the plant kingdom (Perumalla et al. 1990). In a study of 180 angiosperm species, the great majority (89 %) showed a clearly suberised exodermis with Casparian strip. It was found in roots of primitive and advanced plant families, from hydrophytic, mesophytic and xerophytic habitats, but was lacking in some Poaceae. It was notably absent in oat, barley and wheat (Perumalla et al. 1990), but generalizations within cereals cannot be made as subsequently maize was found to have an exodermis (Hose et al. 2001),

The Casparian band can develop close to the root tip. For example, in aeroponically grown maize a complete exodermal layer formed 30 mm above the root tip. In roots elongating more slowly due to abiotic stress or low temperature, it can be found closer to the tip. The extent at which apoplastic barriers form depends on the stage of development of the root system and also the habitat: drought, waterlogging, salinity, nutrient deficiency or toxicity may strongly influence the degree of suberisation (Hose et al. 2001).

Permeability to water and solutes

When an exodermis forms, it also imposes a restriction to radial transport. Complete layers of suberin constrain water and solute flow only via plasmodesmata from the epidermis and inner cortical cells.  Whether or not it can be considered as a barrier depends on the degree of suberisation and the number of passage cells within it. Its properties as a barrier are variable (Hose et al. 2001).

The exodermis represents a resistance to the radial flow of both water and solutes, much like the endodermis in Stage 1 development. It restricts radial apoplasmic movement and may also restrict transmembrane transport of nutrients. Exodermal layers become functionally mature 20–120 mm from the apex, where lateral roots are initiated, and therefore constitute a barrier to apoplastic ion flow only in root zones where an endodermis is already present. In a similar way to the endodermis, maturation of an exodermis involves further deposition of suberin and cellulosic wall material, further impeding flow of solution through walls.

The exodermis is not totally impermeable to water or nutrients and may have a differential selectivity that varies with the environment. How much the root is sealed off depends on the type of exodermis: (1) uniform exodermis where the cells are uniform in shape (suberin deposition is patchy and develops late) or (2) dimorphic exodermis, which consists of long and short passage cells (the former of which are suberised). Suberin lamellae enclosing the long cells disrupt the cytoplasmic continuity through plasmodesmata and the cell dies, but suberin lamellae in the uniform-type exodermis do not affect plasmodesmata. Individual passage cells allow passage of solutes and water via uptake carriers, not via the apoplast as flow there is blocked by the Casparian strip. They have an active role in ion uptake and often become the only plasmalemma facing the soil solution especially when the epidermis dies. Some families (e.g. irises) have large numbers of passage cells while others have very few.

The diagram of a root to the right (Figure 3.54) indicates the stage of development at which the various structures are functional.

The anatomy of roots and the alternative pathways of water and solute flow across roots and into the xylem indicates that the relationship between water and nutrient flow to the shoots could be quite complex. Water may take different pathways than nutrients. The distribution of water in the different pathways depends not only on the age of the root, but may also vary with the flux of water moving through, that is, the rate of transpiration.

 

The various pathways taken by water and nutrients, and the environmental factors that influence them, are not well understood. The fundamental processes governing the relationship between water and nutrient flow through roots are complex. One thing is clear: nutrients do not move passively with the transpiration stream. But neither is their movement entirely independent of it.

This section looks at the relation between water and solute fluxes to the shoot: how the rate of transpiration affects solute concentrations in the xylem, and how active solute transport affects water flow rates at night (root pressure).

(a) Effect of transpiration on solute flux in the xylem

Solute fluxes in the xylem usually respond to changes in transpiration rate, although the relationship is not proportional. The figure below (Figure 3.55) indicates an apparently strong relation between water flux and  K⁺ flux, but it must be noted that the flux of  K⁺ is already high at very low flow rates, increases linearly with moderate increases in flow rate, but then tapers off at high flow rates.

A similar relation has been found for other nutrients like nitrate.

The importance of transpiration in carrying nutrients to the shoot has long been debated. However, experimental evidence showed that transpiration was not necessary to get nutrients to the shoot, as growth rates and net nutrient transport rates were unaffected by humidity and other environmental conditions that would reduce transpiration (Tanner and Beevers 2001). Transpiration was not a prerequisite for long-distance transport of nutrients, as root pressure (see below) in the absence of transpiration can supply the shoot with the required nutrients.

(b) Effect of transpiration on solute concentration in the xylem

Transpiration rate (the volume flux in the xylem) has a marked effect on the concentration of solutes in the xylem sap. Plants which are transpiring rapidly have a low concentration of nutrients in their xylem sap compared with slowly transpiring plants.  This relationship between concentration and volume flux holds for most nutrients, see example with K⁺ (Figure 3.55 above) in which increasing flow rates decreased the concentration of  K⁺ in xylem sap.

The data also shows that an apoplastic or diffusive uptake of K⁺ is not important because the concentration of K⁺ in the xylem was greater than the external K⁺ over a range of volume fluxes which produced an increasing flux (Figure 3.54). These data strongly suggest that ion concentrations in xylem sap are a result of either a single variable pump, or more likely of several sequential processes. The composition of the transpiration stream could be modified after the first passage across a membrane as it flows towards the stele or upwards through the stele by either active or passive ion fluxes from cortical or xylem parenchyma cells.  K⁺ could be pumped in (or diffuse in) at rates which increase with increasing volume flux.

Similar relationships occur between most solutes in the sap and transpiration rate, as shown in Figure 3.56.

This shows how much influence the transpiration rate has on the concentration of solutes in the xylem, and that concentrations in sap collected from exuding cut stumps are not typical of concentrations in transpiring plants.

(c) Root pressure

Plants that have had shoots removed have very concentrated xylem fluid, which exudes from the cut stumps under positive hydrostatic pressure from the roots (‘root pressure’). Root pressure is also responsible for the droplets of water seen on the margins of guttating leaves early in the morning (Figure 3.57).

This occurs because nutrient uptake is an active process, independent of water uptake from the soil, and continues during the night – the rate of nutrient uptake varies little over 24 h. When nutrients are pumped into the stele, water flows in by osmosis, and the pressure builds up. Positive pressures of 30-300 kPa can be achieved in this way.

The Casparian strip and the suberisation of the endodermis is important as it provides a barrier to prevent back-flow of water and also structural support so that the root can contain a positive pressure at night. The pathways of nutrient and water flow across the root cortex may change at night, for instance the apoplastic pathway may be of lesser importance at low rates of transpiration. At zero transpiration induced by a root pressure probe and when the shoot is removed, pathways may be different again.

Root pressure is not just a phenomenon, it is an essential process responsible for moving nutrients to the shoot during the night when transpiration is low. It may also have a function in dissolving gas bubbles that might have caused cavitation in the xylem during the day.

This section covers diffusion of molecules, and permeability of cell membranes, essential to the process of osmosis. Cell membranes are bilayers of phospholipids in which transport proteins are embedded (Figure 3.65).

A plant membrane is often described as semi-permeable, meaning permeable to water but not biological solutes. However the membrane is not 100% permeable to water, as water can enter cells only by being transported through aquaporins, neither is it 100% impermeable to solutes, as solutes can slowly permeate the membrane particularly through specific transport proteins.

Unrestricted movement of water relative to solutes is the basis of osmosis, and in plants the generation of turgor pressure, \(P\). The principles of diffusion and selectivity, which are used to describe differential rates of molecular movement, provide a physical rationale for osmosis.

(a) Diffusion

In a homogeneous medium, net movement of molecules down their concentration gradient is described by Fick’s First Law of diffusion. The molecule and medium may be a solute in water, a gas in air or a molecule within the lipid bilayer. Fick’s Law holds when the medium is homogeneous in all respects except for the concentration of the molecule. If there was an electric field or a pressure gradient then Fick’s Law may not apply. Considering the case of a solute in water, say, sugar, Fick’s Law states that net movement of this solute, also called the net flux \(J_s\), is proportional to the concentration gradient of the solute \( \Delta C_s / \Delta x \):

\[ J_s = -D_s \frac{\Delta C_s}{\Delta x} \tag{12}\]

The diffusion coefficient (\(D_s\)) is a constant of proportionality between flux, \(J_s\), and concentration gradient (mol m^-4), where solute concentration (\(C_s\), mol m^-3) varies over a distance (\( \Delta x\), m). Flux is measured as moles of solute crossing a unit area per unit time (mol m^–2 s^–1), so \(D_s\) has the units m² s^–1. \(D_s\) has a unique value for a particular solute in water which would be quite different from \(D_s\) for the same solute in another medium, for example the oily interior of a lipid membrane.

Across a membrane, the movement of a molecule from one solution to another can be described by Fick’s Law applied to each phase (solution 1–membrane–solution 2). However, flux across the membrane also depends on the ability of the molecule to cross boundaries (i.e. to partition) from solution into the hydrophobic membrane and then from the membrane back into solution. Another difficulty is that the thickness of membranes is relatively undefined and we need to know this for Fick’s equation above (\( \Delta x \)). The two solutions might differ in pressure and voltage and these can change steeply across a membrane; however, if for simplicity we consider a neutral solute at low concentration, these factors are not relevant (see below for charged molecules). A practical quantitative description of the flux of neutral molecules across membranes uses an expression intuitively related to Fick’s Law stating that flux across a membrane (\( J_s \)) of a neutral molecule is proportional to the difference in concentration \(\Delta C_s\):

\[ J_s = P_s \Delta C_s \tag{13}\]

The constant of proportionality in this case is the permeability coefficient (\( P_s \)), expressed in m s^–1. When \( P_s \) is large, solutes will diffuse rapidly across a membrane under a given concentration gradient. \( P_s \) embodies several factors: partitioning between solution and membrane, membrane thickness and diffusion coefficient of the solute in the membrane which may be largely depend on specific transporters. It can be used to compare different membranes and to compare treatments that might alter the ability of a solute to move across the membrane.

Note that Equation 13 assumes that the concentration gradient is only across the membrane, and that when the permeability coefficient is measured, concentration gradients leading to diffusion in solutions adjacent to the membrane will not be significant. If the two solutions are stirred rapidly then this will help to justify this assumption. However, there is always an unstirred layer adjacent to the membrane through which diffusion occurs, and for molecules that can permeate the membrane very rapidly the unstirred layer can be a problem for the correct measurement of permeability.

(b) Permeabilities

Solute movement across membranes can be across the lipid phase of the membrance, depending on size, charge and polarity, and it can be assisted by transport proteins embedded in the membrane.

The range of permeability coefficients for various ions, solutes and water in plant membranes is shown in Figure 3.66, along with the permeability of artificial phospholipid bilayers. The permeability of ions as they cross membranes is higher than that through the artificial lipid bilayer, especially for potassium, indicating the presence of specialised permeation mechanisms, ion transporters, in plant membranes. Water permeabilities are high in both plant and artificial membranes but can range over an order of magnitude in plant membranes. This variability may be partially accounted for by the activity of aquaporins.

A comparison of artificial lipid membranes with biological membranes supports this notion because it shows that many molecules and ions permeate biological membranes much faster than would be predicted on the basis of oil solubility and size (Figure 3.66). For these solutes there are transport proteins in biological membranes that increase solute permeability.

(c) Reflection coefficient - water versus solute permeability

Plant membranes are ideally semipermeable, that is, water permeability is much larger than solute permeability.

The degree of semi-permeability that a membrane shows for a particular solute is measured as the reflection coefficient, \( \sigma \):

\[ \sigma = 1 - \frac{\text{Solute Permeability}}{\text{Water Permeability}} \tag{15}\]

If a plant cell or an epidermal strip is bathed in solution, the reflection coefficient for a particular solute can also be considered as the ratio of the effective osmotic pressure versus the actual osmotic pressure in the bathing solution.

The reflection coefficient usually ranges between zero and one, being zero for molecules with properties similar to water, like methanol, to one for large non-polar molecules like sucrose.

Figure 3.67 shows turgor pressure (\( P \)) in a Tradescantia virginiana epidermal cell as a function of time after the external osmotic pressure was changed with different test solutes. The initial decrease in \( P \) is due to water flow out of the cell and is larger for solutes with a reflection coefficient near one (sucrose and urea). Propanol induces no drop in \( P \), indicating that its reflection coefficient is zero. Subsequent increase in \( P \) is due to penetration of particular solutes such as alcohol across the cell membrane. Water flows osmotically with the solute thereby increasing \( P \) to its original value. Removing solutes reverses osmotic effects. That is, a decrease in \( P \) follows the initial inflow of water as solutes (e.g. alcohols) diffuse out of cells.

The pressure probe apparatus is illustrated in Figure 3.68(a).

Using the pressure probe to measure turgor pressure, \( P \), the membrane is found to be ideally semipermeable for sucrose (\( \sigma = 1 \)); that is, the membrane almost totally ‘reflects’ sucrose. Over long periods, sucrose is taken up slowly but permeability relative to water is negligible. In this case, the change in \( P \) would be equivalent to the change in \( \pi \). If \( \sigma \) is near zero, then water and the solute (say, propanol) are equivalent in terms of permeability. No change in \( P \) can be generated across a cell wall if \( \sigma \) is zero.

Diffusion of neutral molecules at low concentrations is driven by differences in concentrations across membranes, as explained above. There are other forces that may influence solute diffusion, including the voltage gradient when considering movement of charged molecules (ions) and the hydrostatic pressure when considering movement of highly concentrated molecules (such as water in solutions). These forces can be added to give the total potential energy of a particular molecule (\( \mu_s \)) relative to a reference value (\( \mu_s^* \)):

\[ \mu_s = \mu_s^* + RT\ln C_s + z_s FE + V_sP \tag{16}\]

Gravitational potential energy could also be added to this equation if we were to examine the total potential over a substantial height difference, but for movement of molecules across membranes this is not relevant.

The concentration term (\( RT\ln C_s \)) is a measure of the effect on chemical potential of the concentration of solutes (actually the activity of the solutes which is usually somewhat less than total concentration). The gas constant, \(R\) (8.314 joules mol^–1K^–1), and absolute temperature, \(T\) (in degrees Kelvin, which is equals 273 plus temperature in degrees Celsius), account for the effects of temperature on chemical potential.

Incidentally, from this term the well-known van‘t Hoff relation is derived for osmotic pressure \(\pi\), as given at the beginning of this chapter:

\[ \pi = RTC \tag{1} \]

where \(R\) is the gas constant (8.31 joules mol^–1K^–1), \(T\) is the absolute temperature in degrees Kelvin (273 plus degrees Celsius), and \( C \) is the solute concentration (Osmoles L^-1). At 25 ºC, \(RT\) equals 2.48 liter-MPa per mole, and (\( \pi \)) is in units of MPa. Hence a concentration of 200 mOsmoles L^-1 has an osmotic pressure of 0.5 MPa.

The electrical term (\( z_s FE \) ) is a measure of the effect of voltage (\( E \) ) on chemical potential. The charge on a solute (\( z \)) determines whether an ion is repelled or attracted by a particular voltage. Electrical charge and concentration are related by the Faraday constant (\( F\) ) which is 96,490 coulombs mol^–1. The electrical and concentration terms form the basis of the Nernst equation (see below).

The pressure term measures the effect of hydrostatic pressure on chemical potential, where \(P = \text{Pressure}\) and \( \overline{V}_s \) is the partial molar volume of the solute.

Molecules diffuse across a membrane down a chemical potential gradient, that is, from higher to lower chemical potential. Diffusion continues until the difference in chemical potential equals zero, when equilibrium is reached. The direction of a chemical potential gradient relative to transport of a molecule across that membrane is important because it indicates whether energy is or is not added to make transport proceed (Figure 3.69). Osmotic ‘engines’ must actively pump solutes against a chemical potential gradient across membranes to generate \(P\) in a cell. Sometimes ions move against a concentration gradient even when the flux is entirely passive (no energy input) because the voltage term dominates the concentration term in Equation 16. In this case, ions flow according to gradients in electrical and total chemical potential. For this reason, the chemical potential of ions is best referred to as the electrochemical potential.

Ions such as potassium and chloride (K⁺ and Cl^–) are major osmotic solutes in plant cells. Deficiency of either of these two nutrients can increase a plant’s susceptibility to wilting. Most other inorganic nutrients are acquired as ions and some major organic metabolites involved in photosynthesis and nitrogen fixation bear a charge at physiological pH. For example, malic acid is a four-carbon organic acid that dissociates to the divalent malate anion at neutral pH. Calcium (Ca²⁺) fluxes across cell membranes are involved in cell signalling and although not osmotically significant they play a crucial role in the way cells communicate and self-regulate. Finally, some ions are used to store energy but need not occur at osmotically significant concentrations. Cell membranes from all kingdoms use hydrogen (H⁺) ions (protons) in one way or another to store energy that can be used to move other ions or to manufacture ATP (Chapter 2). The highest concentration of H⁺ that occurs is only a few millimoles per litre yet H⁺ plays a central role in energy metabolism.

To understand ion movement across membranes, two crucial points must be understood: (1) ionic fluxes alter and at the same time are determined by voltage across the membrane; (2) in all solutions bounded by cell membranes, the number of negative charges is balanced by the number of positive charges. Membrane potential is attributable to a minute amount of charge imbalance that occurs on membrane surfaces. So at constant membrane potential the flux of positive ions across a membrane must balance the flux of negative ions. Most biological membranes have a capacitance of about 1 microFarad cm^–2 which means that to alter membrane voltage by 0.1 V, the membrane need only acquire or lose about 1 pmol of univalent ion cm^–2 of membrane. A univalent ion is one with a single positive (e.g. K⁺) or negative (e.g. Cl^–) atomic charge. In a plant cell of about 650 pL, this represents a change in charge averaged over the entire cell volume of 12 nmol L^–1!

The membrane voltage or membrane potential difference, as it is sometimes called, can be measured by inserting a fine capillary electrode into a plant cell (Figure 3.68b). Membrane voltage is measured with respect to solution bathing the cell and in most plant cells the voltage is negative across the plasma membrane. That is, the cytoplasm has a charge of –0.1 to –0.3 V (–100 to –300 mV) at steady state with occasional transients that may give the membrane a positive voltage. The tonoplast membrane that surrounds the central vacuole is generally 20 to 40 mV more positive than the cytoplasm (still negative with respect to the outside medium).

Cell membrane voltages can be affected by ion pumps, diffusion potential and fixed charges on either side of the membrane.

Special mention needs to be made of one such fixed charge which arises from galacturonic acid residues in cell walls. Although cations move to neutralise this fixed negative change, there is still a net negative potential associated with cell walls (Donnan potential). In spite of being external to the plasma membrane, the Donnan potential is in series with it and probably adds to what we measure as the membrane potential with electrodes.

Most charge on macromolecules in the cytoplasm is also negative (e.g. nucleic acids, proteins) and because of their size it can be regarded as a fixed negative charge. This has consequences on the water relations of the cytoplasm in that they exert a significant osmotic potential, even though not in solution, as do the clay particles in soil (Passioura 1980).

Different ions have different permeabilities in membranes. Potassium, for example, is usually the most permeable ion, entering under most conditions about 10 to 100 times faster than Cl^– Since ions diffuse at different rates across membranes, a slight charge imbalance occurs and gives rise to a membrane voltage (Figure 3.69). This voltage in turn slows down movement of the rapidly moving ion so that the counter-ion catches up. The result is that when net charge balance is achieved, a diffusion potential has developed that is a function of the permeabilities (\( P_\text{ion} \)) of all diffusible ions present and concentrations of each ion in each compartment.

The Goldman equation describes this phenomenon and gives the membrane voltage (\( \Delta E \) ) that would develop due to diffusion of ions. The Goldman equation for the ions that mostly determine this diffusion potential (K⁺, Na⁺ and Cl^–) is given by:

\[ \Delta E = \frac{RT}{F} \ln \frac{P^{}_K C_K^o + P^{}_{Na}C_{Na}^o + P^{}_{Cl}C_{Cl}^i}{P^{}_K C_K^i + P^{}_{Na}C_{Na}^i + P^{}_{Cl}C_{Cl}^o} \tag{17}\]

The superscripts refer to the inside (i) or outside (o) of the membrane and \( R \), \( T \), \( F \) and \( C \) are defined elsewhere (Equation 4.5). Note that the concentration terms for Cl^– are reversed in the numerator and denominator compared to the cations. This is because Cl^– is the only anion represented. Many texts do not include H⁺ in the Goldman equation because, in spite of high permeability of H⁺, diffusion of H⁺ is unlikely to have a strong effect on ΔE at such low (micromolar) concentrations. However, membrane potential is occasionally dominated by the diffusion of H⁺, indicating that H⁺ permeability must be exceedingly high. For example, local variations in pH cause alkaline bands to form on Chara corallina cells and in the leaves of aquatic plants at high pH.

The Nernst equation

When one ion has a very high permeability compared to all other ions in the system the membrane will behave as an ion-sensitive electrode for that ion (e.g. Figure 4.7). A pH electrode which is sensitive to H⁺ flux across a glass membrane serves as an analogy. In the case of a single ion, the Goldman equation can be reduced to the simpler Nernst equation that yields the equilibrium membrane potential which would develop for a particular concentration gradient across a membrane.

\[ \Delta E = \frac{RT}{zF} \ln \frac{C_o}{C_i} \tag{18}\]

where \( R \) and \( T \) are the gas constant and temperature (degrees Kelvin) and \( F \) is the Faraday constant. Typical charges on ions (\( z \)) would be –1 for Cl^-, +1 for K⁺) and so on. This term in the Nernst equation gives the correct sign for the calculated membrane potential.

The Nernst equation is routinely used by electro-physiologists to calculate the equilibrium potential for each ion. Theoretical equilibrium potentials can then be compared with the actual membrane potential in order to decide whether the membrane is highly permeable to one particular ion. For example, in many plant cells there are K⁺ channels that open under particular circumstances. When this occurs, the membrane becomes highly permeable to K⁺ and the measured membrane potential very nearly equals the Nernst potential for K⁺. The Nernst equation can also be used as a guide in deciding whether there is active transport through a membrane. For example, when the measured membrane potential is less negative than the most negative Nernst potential, an electrogenic pump must be engaged for K⁺ to enter the cell (Table 4.1). If the membrane potential is less negative than the Nernst potential and if a K⁺ channel were open then K⁺ would leak out of the cell. For K⁺ uptake to occur with such a gradient for passive efflux then energy would need to be generated.

Equation 4.7 can be rewritten with constants solved and log₁₀ substituted for the natural logarithm. This yields a useful form as follows,

\[ \Delta E = \frac{58}{z} \log_{10} \frac{C_o}{C_i} \tag{19} \]

showing that 10-fold differences in concentration across a membrane are maintained by a 58 mV charge separation for monovalent ions. For example, -58 mV will keep K⁺ concentrations 10 times higher inside a cell than in the external medium and Cl^– concentrations 10 times lower. Plasma membranes are normally about -116 mV, which would keep K⁺ concentrations inside a cell 100 times higher and Cl^- concentrations 100 times lower than in the external solution.

Internal membranes have a different electrical potential, the mitochondria being more negative than the plasma membrane (around -180 mV) and the chloroplast and tonoplast being slightly positive (around +50 mV).

The concentration of an ion inside a cell membrane (\( C_i \)) that would occur at equilibrium for any \( C_o \) and \( \Delta E \) can be calculated by rearranging the above equation as:

\[ \log_{10} C_i = \log_{10} C_o - \frac{z\Delta E}{58} \tag{20} \]

remembering that \( z \) and \( \Delta E \) can be positive or negative, depending on the ion and the particular cell membrane being considered.

Plant and animal membranes have much higher permeability to water than can be explained by diffusion rates through a lipid bilayer. Furthermore, the activation energy for diffusion of water across a plant membrane is lower than would be expected across a lipid bilayer, where water has to overcome the high-energy barrier of partitioning into a very hydrophobic oily layer. Some reports put the activation energy for water flow across membranes as low as the value for free diffusion of water. In other words, water enters the membrane about as readily as it diffuses through a solution. This suggests that water is moving across the membrane through a pathway other than the lipid, perhaps some kind of water pore or water channel. Since the discovery of water channel proteins in animal cell membranes, molecular biologists discovered that similar proteins exist in plants.

Water channels, like ion channels, are proteins embedded in membranes that facilitate the passive transport (non-energised flow) of water or ions down their respective energy gradients. Movement of a solute or water through these transport proteins is not coupled to movement of any other solute, and does not require ATP. The proteins that facilitate passive transport are diverse; some are specific for particular ions and allow high transport rates per protein molecule (ion channels), some are specific for water (water channels or aquaporins) and some are specific for neutral solutes and may have slower transport rates per protein molecule.

Why are there water channels in membranes when the lipid itself is already somewhat permeable to water? There are several rationales for the presence of water channels in plant membranes. One is that specialised transport proteins can control water flow. That is, a water channel protein may be turned on and off, for example by phosphorylation, while water permeability of the lipid is essentially constant. In animal cells, such as in the kidney, water channels are controlled by antidiuretic hormones. Plant hormones could also influence the function of water channels. A second rationale for the presence of water channnels is to balance water flow and prevent bottlenecks. In the root, water channels are most abundent in the endodermis and inner stele where water flow across membranes is rapid. 

The approach to studying water channels has been to inject genetic material from plants into Xenopus oocytes (a particular type of frog’s egg). The Xenopus oocyte is particularly useful because it is large, enabling observations of cell response to foreign proteins. It is one of several expression systems along with Chara (giant algal cells) and yeast cells. cDNA arising from screens of cDNA libraries can be injected into the Xenopus nucleus, or poly (A)⁺-RNA can be injected into the oocyte cytoplasm where it is translated. Plant water transport proteins expressed in the oocyte plasma membrane result in physiological changes; for example, the oocyte swells rapidly when the external osmotic pressure in the bathing medium is lowered (Figure 3.71a). The first plant aquaporin gTIP (Tonoplast Integral Protein) that was discovered occurs in the tonoplast and probably accounts for its high water permeability. Provided that the increase in water permeability is not a consequence of some other change or a side effect of other types of transport, it can be concluded that the protein catalyses transport of water across membranes.

Water channels can be inhibited by mercuric chloride when expressed in Xenopus occytes (Figure 3.71b), and also in plant cells (Figure 3.71c). Inhibition is reversed by applying mercaptoethanol to block the action of HgCl₂. Under mercury inhibition, the activation energy of water flow increases markedly indicating that water flow is now restricted to diffusional flow across the lipid bilayer, that is, aquaporins are blocked.

Plants have many aquaporin genes. For example, Arabidopsis thaliana has 35 and rice (Oryza sativa) has 33. Proteins encoded by aquaporin genes are localized in the plasma membrane, tonoplast and endo membranes and classed as Plasma membrane Intrinsic Proteins (PIPs), Tonoplast Intrinsic Proteins (TIPs), Nodulin26-like Intrinsic Proteins (NIPS), Small basic Intrinsic Proteins (SIPs) or X Intrinsic Proteins (XIPs) (Luu and Maurel 2013). Within membranes, clustering of PIPs in membrane rafts has been observed, and there is variation in the lateral mobility for different aquaporins. Of the many aquaporin proteins in plants, some transport water only (up to 10⁹ molecules per second) and others are permeable to a range of neutral solutes such as gases (carbon dioxide, ammonia), metalloids (boron, silicon, arsenic), or reactive oxygen species (hydrogen peroxide). The transcription, translation, trafficking, and gating of PIPs are regulated by environmental and developmental factors involving signalling molecules, phytohormones and the circadian clock (Chaumont and Tyerman 2014). The transcription, translation, trafficking, and gating of PIPs are regulated by environmental and developmental factors involving signalling molecules, phytohormones and the circadian clock (Chaumont and Tyerman 2014).

When PIP genes are transcribed, their mRNA is translated in the rough endoplasmic reticulum (ER), and the proteins targeted to the plasma membrane. PIPs belonging to the PIP2 group form homo-oligomers, or hetero-oligomers by associating with PIP1 isoforms. Some PIP2s contain a diacidic motif in their N terminus that acts as an ER export signal. It may be recognized by Sec24 which is the main “cargo” selection protein of the coat protein complex COPII that mediates vesicle formation at the ER export sites. PIP oligomers then transit through the Golgi apparatus and trans-Golgi network and are then loaded into secretory vesicles and routed to the plasma membrane (Figure 3.72). Insertion of PIPs into the plasma membrane is mediated by a protein that regulates vesicle fusion (the “syntaxin” SYP121). The plasma-membrane localized PIPs can be recycled internally: once internalized in vesicles, PIPs are delivered to the trans-Golgi network before being routed back to the plasma membrane or directed into lytic vacuoles for degradation (Figure 3.72). Salt stress causes dephosphorylation and internalization of PIPs, and drought stress induces ubiquitylation of PIPs, which are then degraded in the proteasome.

Aquaporins assemble as homo- or hetero-tetramers, each monomer acting as an independent water channel. The structure of an aquaporin monomer (Figure 3.72) consists of six membrane-spanning α-helices connected by five loops, with both N and C termini facing the cytosol. Two loops form two short hydrophobic α-helices dipping halfway into the membranes, which, together with the membrane-spanning helices, create a pore with high specificity (Murata et al. 2000).

Aquaporins play a central role in regulating plant water relations. Water diffusion across cell membranes is facilitated by aquaporins that provide plants with the means to rapidly and reversibly modify water permeability. This is done by changing aquaporin density and activity in the membrane, including post-translational modifications and protein interaction that act on their trafficking and gating. At the whole organ level aquaporins modify water conductance and gradients at key “gatekeeper” cell layers that impact on whole plant water flow and plant water potential. In this way they may act in concert with stomatal regulation.

PIP and TIP expression is higher during the day than the night, and correlates with diurnal changes in transpiration. It is likely under circadian regulation. Expression can correlate with changes in hydraulic conductivity, Lp, so that aquaporins are more abundant or have higher activity at times when stomatal conductance is higher. Over a diurnal cycle, Lp can change 2-5 fold and so can PIP transcript activity or protein abundance (Chaumant and Tyerman 2014). This occurs in both roots and shoots.

This section has shown how the flow of water and ions across membranes are linked. The function of membrane transport in regulating nutrient supply is covered in Chapter 4.

Although the absolute quantities of nutrients required vary between plant species, genotypes and growth environments, essential nutrients are categorised into so-called macronutrients (N, K, Ca, P, Mg, S) that plants require in larger quantities, and micronutrients (Fe, Cl, Mn, B, Zn, Cu, Mo, Ni) that are needed in small amounts (Figure 4.2). Additional beneficial elements include Si (e.g. for grasses) and Na (for many sea-shore plants).

Macronutrients form the structural components of proteins, cell walls, membranes, nucleotides and chlorophyll, and have roles in energy and water maintenance. The macronutrient potassium has a special function in regulating the osmotic potential of plant cells. Under saline or dry conditions, Cl (and for some plants Na) is important in plant water relations.

Micronutrients mainly provide functional groups in enzymes (BOX 1 shows how Ni forms the active site in urease, as an example).

BOX 1 – Nickel (Ni) at the centre of Urease 4.1-Ch-Box-1.png In 1926 James B. Sumner from Cornell University studied the structure of Urease from Jack Bean plants and demonstrated that the enzyme is also a protein. This work led to the recognition that most enzymes are in fact proteins and in 1946 Sumner was awarded the Nobel Prize in Chemistry. Urease is an enzyme that breaks down urea to ammonia and carbon dioxide in plants, bacteria and fungi and contains a nickel active site. (NH2)2CO + H2O -> CO2 + 2NH3 For 3D structure see http://www.proteopedia.org/wiki/index.php/Urease Further reading: Follmer 2008; Carter et al. 2009.

In terrestrial ecosystems and in agriculture, the availability of nitrogen (N) and phosphorus (P) are often limiting and so affect plant growth and productivity most strongly. However, other elements can also be limiting. Plants showing nutrient deficiencies will exhibit symptoms such as stunted growth, leaf or shoot tip chlorosis, and defoliation, and will die if supplements are not provided. Fertilisers are applied to supply essential elements in agriculture to maximise plant growth and enhance yields. Along with the discovery of ‘dwarfing genes’ and development of short stature crop varieties, it was the use of synthetic nitrogen fertilisers that played a significant role in the Green Revolution of the 1930s-1960s.

The acquisition of mineral nutrients starts with their movement from the surrounding soil to root surfaces. The movement of nutrients from bulk soil towards plant roots occurs via diffusion or mass flow (Figure 4.3). Root interception occurs as the root comes in contact with, and displaces the soil through which it is growing. Nutrient availability in soils and the physical and chemical factors influencing their movement from soils to the root surface is comprehensively described in a review by Marschner and Rengel (2012). Nutrients are taken up into roots by active or passive transport across cell membranes, which is described later in this chapter (Section 4.5).

Diffusion occurs along a concentration gradient, over relatively short distances (in the order of 1 cm). As roots take up nutrients and ions from the soil a depletion zone can be established allowing diffusion to occur into the depletion zone. The rate of diffusion depends on how fast the roots are taking up the nutrient, how much of the nutrient is present in the soil (this determines the steepness of the concentration gradient that forms) and also on the mobility of the ions by diffusion. Soluble ions would take about a day to diffuse 1 cm; ions bound to the soil matrix would take longer. For examples, Marscher and Rengel (2012) show that nitrate by diffusion in a ‘typical’ soil travels 3 mm in a day, potassium about 1 mm in a day, and phosphate moves only about 0.1 mm in a day. This illustrates the importance of root hairs in intercepting and accessing phosphate.

Mass flow is driven by the uptake of water caused by the transpiration rate of the plants and can occur over long distances. Many soluble nutrients such as nitrate are dissolved in the soil water and as the plant pulls the water from the soil, the nutrients move too. Some nutrients move by mass flow faster than their uptake rate so they build up on the surface of the root during daylight hours (Marschner and Rengel, 2012). The rate of movement by mass flow of solution depends basically on the rate of transpiration of the plant, so there is little movement at night. It is also influenced by soil water content and soil texture (see Chapter 3, Section 3.4).

Nutrients are unevenly distributed in the soil, generally being concentrated in the topsoil due to decomposition of leaf litter, but also dispersed elsewhere in pockets. Uneven surface enrichment arises from diverse sources such as dead fauna, urine patches from grazing animals, and localised application of fertiliser. Phosphorus and all cations are relatively immobile as they bind to the soil while nitrate and other anions (except phosphorus) are soluble and can readily be leached to deeper soil layers.

Because the soil is so heterogeneous, plants have developed adaptable (plastic) root systems so that the roots proliferate close to the nutrients for uptake.

The root system architecture is the arrangement of different roots in solid space. Just like a building has walls, roof, and floors, plant root systems also contain different structures including root types (primary, lateral, adventitious), root hairs, and specialized features such as nodules and cluster roots (see case study). In contrast to a fixed structure like a building, the root system is dynamic with new structures forming as needed to explore the soil and old structures breaking down when their use has expired. This four dimensional architecture within soil can now be visualized using technology such as X-ray microscale computed tomography (microCT) and magnetic resonance imaging (MRI).

In order to understand root architecture it is important to understand the different structures that make up the root system. This section will focus on the different root types while cluster roots are explained in the case study that follows. Root hairs are described in the section on the rhizosphere (Section 4.2) and the formation of N2-fixing nodules in Section 4.4.

Primary root: the first root to emerge from a germinating seed (starts as the radicle). Since the primary root is present within the embryo, this class of root is embryonic.

Lateral roots: roots that form from other roots. The lateral roots that form from the primary root are first order lateral roots; the lateral roots that form from the first order laterals are second order laterals and so on. This class of root is post-embryonic.

Seminal roots: form adjacent to the radicle and dominate the early root growth in monocots. This root type is embryonic.

Adventitious roots: any root that forms from anything other than another root. This includes roots that form on the base of stem cuttings, from leaf explants, from stems in flooded plants and also from nodes of cereal crops (often called crown roots). These root types are very diverse (Steffens and Rasmussen, 2015) so can include both embryonic and post-embryonic roots.

Root hairs: single-cell, hair-like extrusions from the epidermis which increase root surface area for nutrient uptake (Jones and Dolan, 2012) and are important for nodulation (Section 4.4).

The combination of different root types present in the root system differs across species. In particular the root systems of cereal crops (monocots) differ dramatically to the root systems of tree crops (eudicots) (Figure 4.8).

Eudicots typically develop a primary (tap) root from a single radicle that emerges from a seed. This primary root, plus the first order lateral roots which emerge from it, provide a framework on which higher-order lateral roots are formed. Such a framework strengthens due to secondary thickening when division of the cambium gives rise to more cell layers, leading to massive roots that are often seen radiating from the base of a tree trunk (Figure 4.9).

Monocots such as grasses and cereal crops do not have a cambium for secondary thickening and develop a fibrous root system. This root system begins with the radicle which grows into the primary root. Adjacent to the radicle, several seminal roots also emerge and combined with the primary root these roots dominate the young root systems of monocots. Next nodal adventitious roots (often called crown roots) emerge from lower stem nodes and these thicker roots gradually dominate the root system. Finally in some monocots, such as maize (corn), nodal adventitious roots emerge above the soil level (brace roots) to provide additional structural support. Stems of monocots are typically anchored by the nodal roots, which are more numerous than seminal roots (Hochholdinger et al. 2004; Hochholdinger 2009).

Despite these structural differences between monocot and eudicot root systems, they can all vary the soil volume which they explore depending on water and nutrient availability. In this way the root distribution in the soil can vary both vertically and locally depending on available resources.

Root distribution

The amount of roots present in a volume of soil varies both vertically and locally depending on resource availability and physical restrictions. This is often measured as the total length of all roots present per unit volume of soil (root length density, L, expressed in km m^–3).

Vertically, the root length density is often large in surface layers of the soil and typically decreases with increasing depth. Commonly, hundreds of kilometres of root per cubic metre of soil are observed near the soil surface.

Figure 4.10 shows root length density, L, as a function of depth in a wheat crop in early spring, and under a jarrah forest, also in spring. Both have a dense population of roots near the surface but wheat roots barely penetrate below 1 m, whereas jarrah roots penetrate to well below the 2.5 m shown here, often to 20 m. Dense root proliferation near the soil surface probably reflects an adaptation of plants to acquire phosphorus, potassium and other cations such as the micronutrients zinc and copper. These nutrients do not move readily in soil as they are bound to the soil surfaces, hence roots branch prolifically to ensure close proximity (a few millimetres) between adsorbing surfaces and these soil-immobile ions. Roots of jarrah are also concentrated near the soil surface (Figure 4.10) to access phosphate and nutrients released by litter decomposition, but some roots penetrate very deeply to tap subsoil moisture.

Nutrients are distributed unevenly in the soil. Root systems respond to enriched zones of nutrients by high levels of branching. Figure 4.11 shows an example of such a proliferation; the dense roots in the centre of the figure are a response by the row of wheat plants to application of a large pellet of nitrogen fertiliser (see arrow).

Such proliferations around bands of fertilizer ensure plants maximise nutrient uptake with the minimum cost to plant development. This efficiency fits within optimal partitioning theory which states that plants respond to environmental variation by partitioning biomass among the plant organs to optimize the acquisition of nutrients, light, water and carbon to maximize plant growth (Reich, 2002). This means that in low nutrient conditions the plants will put more energy into growing roots and less into shoot growth (Reich, 2002). Likewise when light is limiting, plants will invest more energy in leaf area and less in root development (Weaver and Himmel, 1929; Reich, 2002). Maintaining the balance between root and shoot is important as the roots must be extensive enough to supply nutrients and water in proportion to the demand and hydraulic pull from the leaves and vice versa the leaves must produce enough sugars to continue the growth of the root system (Weaver and Himmel, 1929). Consistent with this, Butler et al. (2010) found in Sitka spruce forest, the root absorbing area was correlated with the tree stem diameter and to the transpiring leaf area index. This highlights the link in hydraulics between leaf and root areas.

Young roots absorb nutrients more rapidly than old roots. New roots supply annual plants with abundant sites for nutrient uptake, especially during establishment. A feature of the roots of perennials is that they have a large turnover of the fine, high-order lateral roots that emerge from the secondarily thickened framework each year. This turnover draws heavily on photoassimilate, equivalent to half the CO₂ fixed in annuals and up to 90% of the standing biomass of temperate forests. Production of fine (and often ephemeral) roots ensures uptake of nutrients over many years.

Because many soils are deficient in key nutrients, plants have developed a special relationship with certain fungi called mycorrhizae (Section 4.4). In this symbiosis the fungi obtain fixed carbon from the host plant, and in turn supply the host with poorly mobile nutrients, especially phosphorus. This is achieved by proliferating their hyphae to provide a much greater surface area for nutrient uptake than could be provided by roots alone. Another adaptation, common in the Proteaceae, and also occurring in some species of lupin, is proteoid roots, clusters of tiny rootlets that greatly enlarge the available surface area for ion uptake and which are inducible by low levels of phosphorus (see Case study 4.1).

M. Watt

Cluster or proteoid roots (Figure 1) are found in many species originating from nutrient-deficient soils (Dinkelaker et al. 1995). They enhance uptake of nutrients, especially phosphate. Species which develop these “dense clusters of rootlets of limited growth” include members of the Australian family Proteaceae, where they were first described by Purnell (1960). Other families such as the Casuarinaceae, Cyperaceae, Mimosaceae and Restionaceae also contain species with heavily branched root systems (Lamont 1993). Significantly, few species with cluster roots are mycorrhizal, implying that root clusters fulfil a similar role to mycorrhizal fungi.

Australian soils generally contain low concentrations of plant-available phosphate, much of it bound with iron–aluminium silicates into insoluble forms or concentrated in the remains of decaying plant matter. Because very little of this phosphate is soluble, most roots extract it only slowly. Plants with cluster roots gain access to fixed and organic phosphate through an increase in surface area and release of phosphate-solubilising exudates. Hence plants with cluster roots grow faster on phosphate-fixing soils than species without clusters.

Cluster roots have a distinct morphology. Intense proliferation of closely spaced, lateral ‘rootlets’ occurs along part of a root axis to form the visually striking structures. Root hairs develop along each rootlet and result in a further increase in surface area compared to regions where cluster roots have not developed.

In the Proteaceae, clusters generally form on basal laterals so that they are abundant near the soil surface where most nutrients are found. For example, Banksia serrata produces a persistent, dense root mat capable of intercepting nutrients from leaf litter and binding the protecting underlying soil from erosion (Figure 1a). New clusters differentiate on the surface of this mat after fires and are well placed to capture nutrients. In contrast, Banksia prionotes forms ephemeral clusters which export large amounts of nutrients during winter. Lupinus albus has more random clusters which appear on up to 50% of roots (Figure 2).

Rootlets not only represent an increase in surface area but also exude protons and organic acids, solubilising phosphate and making it available for uptake (Watt and Evans 1999a). Exudates from cluster roots represent up to 10–23% of the total weight of an L. albus plant, suggesting that they constitute a major sink for photoassimilates. However, not all this additional carbon comes from photosynthesis because approximately 30% of the carbon demand of clusters is met by dark CO₂ fixation via phosphoenolpyruvate carboxylase. Because cluster roots form on roots of L. albus even when phosphate supply is adequate, growth of L. albus in soils with low phosphate availability is not restricted by an additional carbon ‘drain’ to roots. On the other hand, the great many species which produce cluster roots in response to environmental cues like phosphate deficiency might experience a carbon penalty to support these roots.

Cluster roots on L. albus are efficient with respect to carbon consumption by generating citrate on cue. Most of the citrate exuded by clusters is released during a two to three day period when the cluster is young (Watt et al. 1999b). A large root surface area in clusters works in concert with this burst of exudation to solubilise phosphate before it is re-fixed to clay surfaces or diffuses away.

Cluster roots can mine a pocket of phosphate-rich soil which would otherwise not yield its nutrients. They are an elegant adaptation of root structure and biochemistry to nutrient-poor soils.

References

Dinkelaker B, Hengeler C, Marschner H (1995) Distribution and function of proteoid roots and other root clusters. Bot Acta 108: 183–200

Lamont BB (1993) Why are hairy root clusters so abundant in the most nutrient impoverished soils of Australia? Plant Soil 156: 269–272

Purnell HM (1960) Studies of the family Proteaceae 1. Anatomy and morphology of the roots of some Victorian species. Aust J Bot 8: 38–50

Watt M, Evans JR (1999a) Proteoid Roots. Physiology and Development. Plant Physiol 121: 317–323

Watt M, Evans JR (1999b) Linking development and determinacy with organic acid efflux from proteoid roots of Lupinus albus L. grown with low phosphorus and ambient or elevated atmospheric CO₂ concentration. Plant Physiol 120: 705–716

The rhizosphere is the narrow zone of soil surrounding plant roots that is characterised by root exudation and an abundance of micro-oganisms which can be beneficial or harmful to plants, or have no effect on root growth and function. These microbes are saprophytic, pathogenic or symbiotic bacteria and fungi, including rhizobia forming nodules and arbuscular mycorrhizal fungi (Figure 4.14).

Many root phenomena suggest specific roles for the rhizosphere. For example, roots have long been thought to find a relatively frictionless path through soils because of exudation of organic substances and cell sloughing, but the chemical and physical processes that underpin this phenomenon are still quite unclear (McKenzie et al. 2012). Production of new roots around local zones of enrichment (Section 4.1) is made far more effective through rhizosphere activity associated with these young roots. Phosphate availability is particularly likely to be improved by the presence of a rhizosphere. Potential mechanisms will be discussed below.

Enhancement of root growth under conditions which favour high root:shoot ratios and the attendant rhizosphere surrounding those roots (rhizosheath) require a substantial input of organic carbon from shoots. Some is used in structural roles, while roots and microbes also require large amounts of carbon to sustain respiration. Even in plants growing in nutrient-adequate, moist soils, 30–60% of net photosynthate finds its way to roots (Marschner 1995). Carbon allocation to roots can be even greater in poor soils or during drought. The rhizosphere accounts for a large amount of the root carbon consumption (Jones et al. 2009). Barber and Martin (1976) showed that 7–13% of net photosynthate was released by wheat roots over three weeks under sterile conditions while 18–25% was released when roots were not sterile. This difference might be considered carbon released because of microbially-induced demand in the rhizosphere, and therefore made unavailable for plant growth.

Rhizosphere chemistry and physics differ from the adjacent soil matrix and root tissues. Gradients in solutes, water and gases combine with microbial activity to produce a unique compartment through which roots perceive bulk soil. This zone of influence typically extends not more than 3 mm from the root axis (Figure 4.15), partly due to the low diffusion coefficients of most solutes that move through the rhizosphere (10^–12 to 10^–15 m² s^–1 for ions such as orthophosphates). Even a relatively mobile ion such as nitrate, with a diffusion coefficient (D) of around 10^–9 m² s^–1 in soil solution, diffuses through about 1 cm of soil in a day. Because the time required (t) for diffusion of ions is a function of the square of distance traversed (l), where t = l²/D a nitrate ion would take four days to travel 2 cm, nine days to travel 3 cm and so on. Similarly, organic carbon diffuses away from roots only slowly, sustaining a microbial population as it is consumed in the rhizosphere.

Roots advancing through soil perceive a wide range of chemical and biological environments: a rhizosphere simultaneously fulfils buffering, extraction and defence roles allowing roots to exploit soils. A rhizosphere is thus a dynamic space, responding to biological and environmental conditions and often improving acquisition of soil resources. New roots develop an active rhizosphere which matures rapidly as the root axis differentiates.

Photoassimilate diffuses from roots into the rhizosphere where it is either respired by microorganisms, volatilized, or deposited as organic carbon (‘rhizodeposition’). Some of this photoassimilate loss is in the form of soluble metabolites, but polymers and cells sloughed off the root cap also provide carbon substrates. Grasses undergo cortical cell death as a normal developmental process, providing further carbon substrates to support a rhizosphere microflora. Nitrogen and some other inorganic nutrients which are co-released with plant carbon are often reabsorbed by roots. Extraction of minerals from bulk soil also relies strongly on rhizosphere processes, especially near the root apices. Compounds exuded from roots interact with soil components in direct chemical reactions (e.g. adsorption reactions), through microbially mediated events (e.g. immobilisation reactions) and volatilisation. In addition, complex polysaccharides and glycoproteins of microbial and root origin give rise to a gelatinous mucilage which associates with soil particles to form a rhizosheath.

The rhizosheath is known as the soil that adheres to roots when they are removed from the pot or field Figure 4.16). The amount of soil can vary depending on how gently or roughly the roots are removed. For wheat, at least, the size of the rhizosheath correlates with root hair length. Mutants without root hairs have no rhizosheath. The distinction between the terms rhizosheath and rhizosphere are that the first term refers to the soil that physically adheres, and the second term the volume of soil influenced by the root. Mutants without root hairs would still have a rhizosphere of sorts since the root would still chemically influence its surrounding soil.

Rhizosheaths have physical and chemical implications for root function. Hydraulic continuity between soil and roots is, for example, thought to be enhanced by the hydrated mucigel, which facilitates water uptake by roots in dry soils. Negatively charged groups on side-chains of mucilagenous polysaccharides attract cations like Ca²⁺, providing exchange sites from which roots might absorb nutrients. The mucigel between the sloughed root cap and root border cells also acts as a lubricant for reducing penetration resistance of the expanding root tip in soil (McKenzie et al. 2012). For example, root elongation through hard soil is greatly reduced if the root cap cells are removed. Once the soil and the mucigel dry up, this lubrication effect is significantly reduced.

Such a diversity of chemical reactions in the rhizosphere is largely an outcome of the array of root-derived exudates. For example, phenolic compounds can be released by root cells in large amounts (Marschner 1995), both as a result of degradation of cell walls and from intracellular compartments. Flavonoids are a group of phenolics that can be specifically exuded into the rhizosphere as signal molecules to attract rhizobia (See section 4.4 on nitrogen fixation). Release of organic acids (principally citric, fumaric and malic acids) solubilises phosphate from surfaces to which they are adsorbed in many species, including those of the family Proteaceae. A modest release of organic acids accounting for about 0.1% of the root mass each week is sufficient to enhance phosphate acquisition in a selection of annual legumes (Ohwaki and Hirata 1992). In more extreme cases, up to a quarter of the dry weight of Lupinus albus plants is released from cluster roots, mostly as citrate (see Case study 4.1). Even the fungal hyphae of mycorrhizal eucalypt and pine roots can secrete photoassimilates, in the form of oxalic acid, causing phosphorus to be solubilised from insoluble calcium apatite (Malajczuk and Cromack 1982).

The main families of low molecular weight compounds which react with inorganic ions are phenolics, amino acids and organic acids. Heavy metals such as aluminium, cadmium and lead are complexed by phenolics, affecting the mobility and fate of these ions in contaminated soils. Flavonoids can chelate iron and make iron oxides available to plants. Manganese is complexed by organic acids, as are ferric ions, which also interact chemically with phenolic compounds and amino acids. For example, highly specialised amino acids (phytosiderophores) can complex ferric ions and enhance uptake from soils by rendering iron soluble. Low iron status actually stimulates release of phytosiderophores into the rhizosphere (Marschner 1995). Other metals such as zinc and copper might also be made more available to the plant through the chelating action of phytosiderophores. Chemical processing by chelating agents is dependent on plant perception of nutrient deficiencies, leading to an ordered change in rhizosphere chemistry. A significant demand on photoassimilates is required to sustain chelation of nutrient ions.

Enzymes are also released from roots, particularly phosphatases, which cleave inorganic phosphate from organic sources. The low mobility of orthophosphates means that phosphatases can be an important agent in phosphorus acquisition, especially in heathland soils where the native phosphorus levels are low relative to the phosphorus-rich remnants of decaying plant material.

pH is another important rhizosphere property. Roots can acidify the rhizosphere by up to two pH units compared to the surrounding bulk soil through release of protons, bicarbonate, organic acids and CO₂ (Figure 4.17). In contrast, the rhizosphere of roots fed predominantly with nitrate was more alkaline than bulk soil. A distinct rhizospheric pH arises because of the thin layer of intense biological activity close to roots, especially young roots. In addition to proton fluxes, release of CO₂ by respiring roots and microbes is likely to cause stronger acidification of the rhizosphere near root apices where respiration is most rapid.

Rhizosphere acidification affects nutrient acquisition by liberating cations from negative adsorption sites on clay surfaces and solubilising phosphate from phosphate-fixing soils. Furthermore, micronutrients present as hydroxides can be released at low pH, conferring alkalinity tolerance on those species with more acidic rhizospheres. So, the rhizosphere is a space which ensheathes particularly the youngest, most active parts of a root in a chemical milieu of the root’s making. In this way, acquisition of soil resources is strongly controlled by processes within roots. Local variations within soil are buffered by rhizosphere chemistry, enabling roots to exploit heterogeneous soils effectively.

Microbial activity, sustained by photoassimilates secreted from roots, contributes substantially to rhizosphere properties. The level of microbial activity is also influenced by availability of nitrogen as a substrate for microbial growth. Soils with high fertility and biological activity have microbial densities 5–50 times greater in the rhizosphere than in bulk soil. The diversity of rhizosphere microflora is spectacular (Figure 4.18) and still incompletely described. An initial hurdle in the identification of rhizosphere microbes was the fact that most of them are unculturable on most known growth media. Metagenomics allows species to be sequenced from soil samples without culturing, largely overcoming this bottleneck. Next Generation sequencing studies of microbes present on root surfaces and in the rhizosphere soil have discovered thousands of different bacterial and fungal species living in close association with plant roots (e.g. Bakker et al. 2013; Bulgarelli et al. 2012; Lundberg et al. 2012). Some of these are very abundant and found in association with many plant species, others are less abundant and highly variable (Figure 4.19).

These microbes can almost be viewed as an extension of the plant into the soil. Like the human gut microbiome, the plant rhizosphere microbiome appears to be an essential part of the plant with multiple functions in nutrition and pathogen defense; it is inseparable from the plant and has been dubbed the plants second genome. The rhizosphere community is highly structured and not a random collection of species – it is strongly influenced by plant species and even ecotypes, by the type of the soil, availability of nutrients and the exudation of chemicals from the root (Bakker et al 2013). Plant mutants with altered chemical composition of root exudates have been found to attract significantly altered microbial communities. It will be fascinating to discover to what extent this is an active strategy of the plant to attract the most appropriate rhizosphere microbiome to help the plant survive in a given environment.

Rhizosphere microorganisms are also not uniformly distributed along roots. Apices are almost free of microbes but densities can increase dramatically in subapical zones. Very mature root axes with lateral branches are sparsely populated with microbes. Even within these zones, there are large variations in distribution, with radial epidermal walls of roots secreting exudates which can support huge microbial populations, up to 2 × 10¹¹ microbes cm^–3. Composition of microbial communities varies with their distribution along the root as well, likely reflecting different nutrient sources along the root. Fluorescence in situ hybridization (FISH) can be used to visualise different taxa of bacteria on the root surface (rhizoplane; Figure 4.20).

Roots do of course influence adjacent soil throughout their length by setting up gradients of water, gases and ions. For example, in waterlogged soils leakage of O₂ from aerenchymatous roots leads to oxidation of metal ions and local build up of aerobic microflora around roots of agricultural plants (Chapter 18). In general, however, the most active microbial populations and rates of chemical transformation in the rhizosphere occur in the subapical zones of the root. In supporting these processes, root-associated microbes metabolise inorganic nitrogen, depositing protein nitrogen in the process of immobilisation. Microbial activity also produces plant growth regulators such as auxin, cytokinins and gibberellins, sometimes in amounts sufficient to influence root morphogenesis. Ethylene can also be produced by rhizospheric fungi, potentially influencing root morphological changes such as lateral root initiation. Some bacteria have been found to promote plant growth by reducing ethylene levels around roots through production of an enzyme degrading an ethylene precursor, 1-aminocyclopropane-1-carboxylate (ACC) deaminase.

Root function and overall plant performance can benefit conspicuously from processes in the rhizosphere. Infection by rhizobia (Section 4.4) and mycorrhizal fungi (Section 4.3) improve the nutritional status of many species. Rhizobial strains have even been used to manipulate rhizosphere biology. A significant proportion of photoassimilate is used to support a rhizosphere, reflecting the high cost of microbial activity and polymer exudation. This pattern is repeated in many species with up to 20% of plant carbon consistently lost by roots, however, this value can vary substantially with the biotic and abiotic conditions. Relative rates of microbial and root respiration are almost impossible to estimate in roots growing in undisturbed soils because of the intimacy of roots and microbes. In addition to consuming large amounts of plant carbon, some microbes can produce phytotoxins, which can impose further restrictions on root function. Some microbes also contribute to nutrient depletion in the rhizosphere, for example by converting usable forms of nitrogen, i.e. nitrate or ammonium, into unusable forms like N₂.

Mechanisms describing how a rhizosphere benefits its host are even more elusive because of the diversity of reactions in such a small space. Chelation is identified as a major influence on nutrient acquisition and might also help ameliorate ion toxicities. Physical properties of the rhizosphere are even less well understood, with questions such as root lubrication, root–mucilage shrinkage and interfacial water transport not yet resolved. Physical properties of mucilage do not suggest it is an ideal lubricant. Whether the dynamic properties of a rhizosphere bring constant benefits to a plant or simply passively coexist with growing roots remains a critical question.

One demonstrated benefit of the rhizosphere microbiome is the protection of the plant from diseases. Several mechanisms have been suggested for this effect (Bakker et al. 2013; Berendsen et al. 2012): Disease suppressiveness, the ability of the microbial inhabits of the rhizosphere to suppress the infection of plants by soil-swelling pathogens, has been ascribed to the production of antimicrobial substances by bacteria, to competition between beneficial and pathogenic microbes and to the induction of systemic resistance by beneficial bacteria. An intriguing example is the colonization of plants by pathogens, which can lead to changes in the exudation of organic acids that then attract beneficial bacteria that induce systemic resistance in the plant, reducing pathogen infection. The induction of systemic resistance to pathogens can also be triggered by specific signaling molecules of bacteria, quorum sensing signals, which bacteria used to ‘talk’ to each other to coordinate multicellular-like behaviours of bacterial colonies. Perception of quorum sensing signals from rhizosphere bacteria by plants can increase systemic resistance to pathogens in the shoot, and can also enhance symbiosis with nitrogen-fixing bacteria. Quorum sensing signal perception also triggers the production of so called quorum sensing mimic compounds – signals that interfere with bacterial communication in the rhizosphere (Teplitski et al. 2011). While we still need to identify most of the signals, signal mimics and exudate components in the interaction of roots with their microbiome, it is clear that plants actively create the rhizosphere, and that this is likely to benefit the plant in its environment.

Mycorrhizal associations are classified according to the way in which the fungi interact with the host plant root, in particular, the structure of the interface that forms between host cells and fungal hyphae. This classification leads to a number of distinct types of mycorrhizal association, as defined in Table 4.1. However, only two of these are widely distributed in the plant kingdom: arbuscular mycorrhizas (AM) and ectomycorrhizas. Orchid and ericoid mycorrhizas are confined to genera within the Orchidaceae and Ericaceae families, respectively.

Mycorrhizal types generally form with a characteristic group of plant species, but there are occasional examples of overlap, such as many Australian plants in the families Fabaceae and Myrtaceae, which have both arbuscular mycorrhizas and ectomycorrhizas. Arbuscular mycorrhizas occur in a vast array of herbaceous genera. In fact, as shown in Figure 4.21 above, some 75% of all plant species form arbuscular mycorrhizas, including most major crop species, that is, all cereals and most grain legumes and pasture legumes.

Table 4.1 shows that the main types of mycorrhizas differ in host preference and in the structures they form during association with the host root, but they are similar in the ways by which they enhance host plant nutrition. Each type of mycorrhiza can be formed by many species of fungi and a single root may often be colonised by more than one species.

Plants with arbuscular mycorrhizas are common in most ecosystems, but are more likely to be dominant in regions of relatively high mean annual temperatures and rates of evapotranspiration, where phosphorus availability is often the major limiting factor for plant growth. However, in some soils with very low phosphorus availability, nonmycorrhizal plant species with cluster roots may be locally dominant and these plants seem to be more efficient at obtaining phosphorus from these soils than mycorrhizal species (e.g. south west Western Australia).

Ectomycorrhizas are most common in tree species, but also occur in some shrubs. In the Northern hemisphere, ectomycorrhizal associations are typically dominant in boreal forests where temperatures and evapotranspiration are relatively low, leading to slow rates of decomposition and accumulation of plant litter in soil and low nitrogen availability. However, ectomycorrhizal plants are also dominant or co-dominant in many other temperate forests, as well as some tropical and subtropical areas, where soil properties are not substantially different from habitats where only arbuscular mycorrhizal plants occur.

Each type of mycorrhizal association has evolved separately to enhance growth and survival of both the host plants and the mycorrhizal fungi. While the primary role of these associations is to increase nutrient supply to the host plant, mycorrhizas have also been shown under some circumstances to enhance plant water status, confer protection against root pathogens, contribute to soil structure through hyphal binding of soil particles and other processes, and render plants less susceptible to toxic elements. The relative importance of these secondary roles is very difficult to determine since they are difficult to separate from nutritional benefits to plants in experiments and they will not be considered in detail here.

Arbuscular mycorrhizal associations are formed by fungi from division Glomeromycota, class Glomeromycetes. DNA sequence evidence shows they are closely related to the Zygomycota. These fungi simultaneously exist in both the soil and roots, with different forms of hyphae in each environment, as shown in the diagram below (Figure 4.23).

During the infection process, fungal hyphae penetrate the epidermal cell layer, often forming distinctive large hyphae within the root at the point of entry (Figure 4.24). From the entry point, hyphae then spread through the root cortex by growing either through the intercellular spaces or from cell to cell by penetrating the cell walls. Hyphae do not, however, penetrate the endodermis or enter the stele.

Within individual cortex cells, hyphae may form a distinctive structure called an arbuscule. From the base of each arbuscule, hyphae repeatedly branch, becoming thinner and thinner as they do so (Figure 4.25). The host cell plasma membrane is never penetrated by the fungus.

Thus in a cell with an arbuscule, the host cell plasma membrane remains intact and functional, but proliferates to surround the arbuscular branches (Figure 4.26). The highly-branched nature of arbuscules is thought to increase the surface area to volume ratio of the host plant plasma membrane by up to 20-fold, relative to unoccupied root cells, thus providing an extensive interface across which nutrient exchange can take place (Figure 4.26).

When we consider that all fungal biomass was built using plant carbon, it becomes evident that considerable carbon is needed to maintain the symbiosis.

Arbuscular mycorrhizal fungi store the carbon they obtain from the host plant root primarily in the form of lipids. Lipids are particularly dense in vesicles and spores, which also act as inoculum. Vesicles and spores may form within or outside of roots and often develop most prolifically when roots begin to senesce. The morphology of the fungal infection, particularly the vesicles and spores, differs with the species of fungi.

Ectomycorrhizal symbioses are formed primarily by higher fungi in the Basidiomycotina and Ascomycotina, which form mycorrhizas with the short lateral roots of trees (Table 4.1). Unlike arbuscular mycorrhizas and ericoid mycorrhizas, hyphae of ectomycorrhizal fungi do not normally penetrate host cell walls. Rather, they form an entirely extracellular interface, with highly branched hyphae growing between epidermal or cortical cells, forming a network known as the Hartig net (Figure 4.27).

In Gymnosperms such as Pinus, the Hartig net may extend through most of the root cortex, but in most Angiosperms it is confided to the epidermis (Figure 4.28). In both cases, the highly branched hyphae of the Hartig net provide a substantial surface area for nutrient exchange between the fungus and the plant. Ectomycorrhizas are further differentiated from the other mycorrhizal types by the fact that the fungus usually forms a dense hyphal mantle around each short lateral root, greatly reducing its contact with the soil (e.g. Figure 4.28).

Orchid mycorrhizal associations consist of coiling hyphae within cells of a root or stem (Figure 4.29). The most common fungi involved are members of the Rhizoctonia alliance, but ectomycorrhizal fungi are also found in some orchids, especially in achlorophyllous species lacking photosynthesis. A key feature of orchid mycorrhizas is the capacity of fungi to germinate the tiny seeds of orchids to form tiny protocorms which lack roots or leaves. It is thought that orchids start out by exploiting fungi, but then may develop more mutualistic associations as they grow larger and develop leaves.

Ericoid mycorrhizal fungi (largely ascomycetes) form an interface within cells, consisting of dense hyphal coils which are surrounded by host plasma membrane which is similar to orchid mycorrhizas (Figure 4.30). Many members of the Ericaceae host these associations in very fine lateral roots called hair roots, which are only a few cells wide (Figure 4.30).

The association between fungus and plant delivers nutrients to the host plant via: (a) mobilisation and absorption by fungal mycelia in the soil; (b) translocation to the fungus–root interface within the root and (c) transfer across the fungus–root interface into the cytoplasm of root cells. As shown in Figure 4.31, both roots and mycorrhizas can absorb nutrients such as phosphorus from the soil, so plants with highly branched fine roots and long root hairs are less likely to benefit substantially from mycorrhizal associations.

The majority of plants in natural ecosystems have relatively thick and unbranched roots without long root hairs so are likely to be highly dependent on mycorrhizal associations in the soils where they normally grow. Some crop and garden plants, such as many grasses and members of the Brassicaceae have long root hairs so tend to benefit less from mycorrhizas or have nonmycorrhizal roots (e.g. Figure 4.32).

(a) Nutrient uptake from the soil by the fungi

In addition to hyphae in direct contact with the root surface, all mycorrhizal fungi produce soil hyphae (extramatrical mycelium) which extend into the surrounding soil. Both arbuscular mycorrhizal and ectomycorrhizal fungi can produce copious soil hyphae, that extends well beyond the nutrient depletion zone for immobile nutrients around individual roots and display a complex architecture that renders them an efficient nutrient-collecting and transport network (Figure 4.33). The soil hyphae of many ectomycorrhizal fungi form hyphal aggregations, known as mycelial strands and rhizomorphs, that play a major role in transport of inorganic nutrients or photoassimilates.

Extramatrical mycelium is also primarily responsible for spread of the association to new roots and translocation of energy from the plants for fungal reproduction. Fungi forming mycorrhizal associations can also spread by germination of wind or animal dispersed spores and, in some cases, from old root pieces.

In arbuscular mycorrhizas, fine highly branched soil hyphae (diameter 1–5 µm) provide surface area for nutrient absorption, while larger diameter hyphae (up to 10 µm) form a transport network in the soil for moving solutes from bulk soil to the root (Figure 4.34). Absorption of phosphate by the fungus is maximised by the action of a high-affinity transporter which is expressed only in the soil hyphae of arbuscular mycorrhizal fungi during symbiosis with the plant. The fungi take up inorganic phosphate and quickly convert it to polyphosphate, a macromolecule where the charge of the phosphate ions is balanced by cations including those of potassium and magnesium. Polyphosphate allows phosphorus to be transported to the plant without affecting hyphal osmotic balance. For instance, within the root, concentrations of phosphorus in hyphae may be up to 350 mM, but plant cell vacuoles generally have < 10 mM. Once the polyphosphate reaches an arbuscule in the plant root it is converted back to phosphate and released into the peri-arbuscular space where it is absorbed by the host plant. Active transporters in the host cell plasma membrane maintain a concentration gradient across the plant-fungus interface.

Many experiments have demonstrated a relationship between arbuscular mycorrhizal infection and improved plant phosphorus status, particularly under glasshouse and laboratory conditions (Figure 4.35). Arbuscular mycorrhizal fungi do not appear to have access to sources of soil phosphorus that are otherwise unavailable to nonmycorrhizal roots. Thus, increased plant absorption in the presence of arbuscular mycorrhizal fungi of phosphorus, nitrogen and other macronutrients such as calcium and sulphur, and micronutrients including zinc and copper, seems to primarily reflect the increased absorptive surface of the soil hyphae. However, soil hyphae also provide a conduit for rapid transport of carbon from plants into soil and there is evidence that hyphal exudation may promote breakdown of organic nutrient sources by other microorganisms (see below). Note that the effects of individual species or strains of fungi on plant nutrition will vary, in part due to different morphology of soil hyphae. Under some circumstances, the presence of arbuscular mycorrhizal fungi may decrease plant growth especially in heavily fertilised crop plants.

The soil hyphae of ectomycorrhizal fungi increase the absorptive area of a root system substantially, extending the volume of soil explored by the host plant and consequently the quantity of minerals available. Ectomycorrhizal fungi, however, use additional strategies to enhance nutrient acquisition. Some secrete extracellular proteinases, peptidases, phosphomonoesterases and phosphodiesterases that effectively hydrolyse organic nitrogen and phosphorus sources to liberate some nitrogen and phosphorus compounds which can be absorbed by the fungi. Some ectomycorrhizal fungi produce hydrolytic enzymes within the cellulase, hemicellulase and lignase families that may facilitate hyphal entry to moribund plant material in soil and access to mineral nutrients sequestered therein. In these ways, ectomycorrhizal fungi short-circuit conventional nutrient cycles, releasing nutrients from soil organic matter with little or no involvement of saprotrophic organisms. Ectomycorrhizal fungi also release siderophores capable of complexing iron and oxalate to improve potassium uptake and have also been implicated in promoting weathering of rocks to release mineral nutrients for plants.

(b) Carbon uptake from the plant by the fungi

Arbuscular mycorrhizal fungi depend completely on the host plant for carbon and are unable to grow without being associated with a host plant. This has made the culture of these fungi difficult and proved a significant barrier to development of cheap technologies for inoculation (as might be desirable in land rehabilitation or agriculture) on a large scale. Transfer of carbon from the host to an arbuscular mycorrhizal fungus likely takes place in the arbuscule where the plant releases simple sugars (hexoses) which are absorbed by the fungi. These sugars are rapidly converted into trehalose, glycogen and lipids. The lipids and, to a lesser extent, glycogen are transported to the soil hyphae. Once in the soil hyphae, lipids are progressively broken down into hexoses and trehalose and used to fuel the growth of the fungus. As the lifecycle of the fungus progresses, large amounts of lipid are stored, particularly in vesicles and spores which may be inside or outside of the roots.

In contrast to arbuscular mycorrhizal fungi, ectomycorrhizal fungi can utilise carbon substrates other than those provided by the host plant. It seems that most ectomycorrhizal fungi have some ability to use lignin and cellulose, along with various other substrates including starch, glycogen and sugars such as glucose. Ability to utilise various substrates differs among fungal species. As a result of these abilities, ectomycorrhizal fungi are able to be isolated and grown in culture. For ectomycorrhizal fungi associated with host roots, sucrose is thought to be hydrolysed in root cell walls and glucose to be then absorbed by hyphae from the interface apoplasm.

It has been estimated that 20-50% of plant photosynthate is allocated to mycorrhizal fungi, much of which is allocated to soil hyphae. The soil hyphae of the fungi exude carbon compounds which will influence soil processes including the growth, composition and function of the soil microbial community. Recent research suggests that roots and mycorrhizas may differentially affect soil carbon pools. Thus, overall, the fungi provide a significant pipeline for the movement of carbon from the plant shoot into the soil and may greatly influence soil processes and microbial activity both within and away from the rhizosphere. Indeed, it is now thought that the fungi may significantly influence the global carbon cycle (e.g. their cell walls include some components that are very slow to decompose in soils). In addition, some compounds exuded by the soil hyphae of mycorrhizas, such as glomalin, play an important role in maintenance of soil structure through gluing together soil particles, especially in sandy soils. In addition, colonisation can change the amount and composition of compounds exuded by roots. For instance, the presence of arbuscular mycorrhizal fungi can result in the amount of carboxylates in the rhizosphere being reduced by 50% or more (Figure 4.36). Carboxylates are low molecular weight organic anions which are thought to play a role in release of highly sorbed phosphorus into forms that plant roots or the hyphae of arbuscular mycorrhizal fungi can absorb. Hence, the presence of the fungi enhances the ability of the host to access orthophosphate, but perhaps at the expense of its ability to release phosphorus from sorbed sources.

Overall, there are many fascinating complexities to the relationships among host plants, fungi, nutrients such as phosphorus and nitrogen, carbon and the soil microbial community. For instance, plants make “trade-offs” among nutrient acquisition strategies, probably due to the carbon costs of each strategy. For example, colonisation by arbuscular mycorrhizal fungi often results in a reduction in root / shoot ratio, and root hair production and fine root production tends to be greater in nonmycorrhizal plants. While there are substantial costs to plants in supporting mycorrhizal associations, the cost of producing roots that function well without them seems to be even greater. Overall, the best evidence that mycorrhizal roots are more efficient than nonmycorrhizal roots is provided by the data in Figures 4.21 and 4.22. This global dataset shows that mycorrhizal plants are normally dominant in ecosystems, with the exception of habitats where conditions are likely to suppress fungal activity (e.g. waterlogged, saline, or very cold soils and epiphytic habitats).

Plant growth is frequently limited by nitrogen. Plants generally obtain nitrogen from soil reserves of nitrate or ammonium (so-called mineral nitrogen) but these reserves are often scarce.

Natural ecosystems can ‘run down’ with respect to nitrogen through soil leaching and fire. Relative abundance of nitrogen-fixing species will then increase. For example, a walk from east to west across Fraser Island, Queensland, will take you across progressively older and more nitrogen deficient sand dunes, and from rainforest to heathland.

Agriculture and horticulture, with its harvest of nitrogen-rich grains or leaves, removes site nitrogen that must be replaced by further mineralisation of soil nitrogen, import of mineral nitrogen (fertiliser) or fixation of atmospheric nitrogen (N₂).

The earth’s atmosphere is rich in N₂ (about 78% N₂) which is very unreactive, due to its stable triple bond. No known eukaryotes have the ability to fix nitrogen, i.e. to reduce atmospheric nitrogen into a usable form like ammonia (NH₃). Hydrogen (H₂) will react with N₂ at very high temperatures and pressures on a catalyst by the Haber-Bosch process, where the pressures needed are 10–100 MPa, the temperatures are 400–550 ºC, and the catalyst is Fe. Then:

\[ N_{2} + 3H_{2} \rightarrow 2NH_{3} \tag{1} \]

Large quantities of ammonia are produced by this method for industrial and agricultural use. Some nitrogen is also fixed in the atmosphere during lightning strikes (Fowler et al. 2013).

Amazingly, some bacteria have the ability to catalyse this reaction with an enzyme complex called nitrogenase donating at least four pairs of electrons to every N₂ molecule to effect reduction to two NH₄⁺ and at least one H₂. The reaction takes place at ambient conditions, catalysed by the Fe–Mo-containing enzyme, nitrogenase.

\[N_{2} + 16ATP + 8e^– + 10H^+ \rightarrow 2NH_{4}^+ + H_{2} + 16ADP + 6P_{i} \tag{2}\]

Biological N₂ fixation is energetically expensive even though it occurs at ambient conditions — estimates fall between 3 and 7 g carbon respired gram of nitrogen fixed (Layzell 1992). Photoassimilate consumed to support N₂ fixation is unavailable for other processes such as growth. Consider a crop fertilised with 140 kg N ha^–1. An N₂ fixer could replace this fertiliser, but only at a cost of at least 420 kg C ha^–1. As most plant dry matter contains 40% carbon, this is equivalent to a loss of one tonne of dry matter per hectare! However, in natural ecosystems where no nitrogen fertiliser is applied, and with rising costs of synthetic nitrogen fertiliser, which consumes about 2% of global fossil fuels annually, biological nitrogen fixation confers a distinct advantage to plants that associate with N₂-fixing bacteria (Figure 4.39). In fact, our use of nitrogen fertilisers in agriculture has expanded so enormously that pollution of aquatic habitats by nitrogen run-off has moved beyond safe limits for our planet (Steffen et al. 2015). During the last 30 years, research into biological nitrogen fixation, especially in the symbiosis of legumes with rhizobia, has advanced to a point where transfer of this symbiosis to non-legume crops is now becoming a serious goal (Rogers and Oldroyd 2014).

Many bacterial species from diverse phyla have the ability to fix nitrogen, including many (but not all!) cyanobacteria, actinobacteria and proteobacteria. The nitrogen fixation genes are thought to have spread between distantly related bacteria by horizontal gene transfer of clusters of nitrogen-fixation (nif) genes. Nitrogen-fixing bacteria, also called ‘diazotrophs’, can be free-living in water or on solid substrates like soil or rocks. More than half of the biological nitrogen fixation on earth stems from nitrogen-fixing marine bacteria, the rest from terrestrial sources (Fowler et al. 2013). Terrestrial N₂-fixing bacteria are found in the soil, in aquatic, and often extreme, habitats, such as hot springs, and nutrient-poor areas. Much of the action of N₂-fixing bacteria happens in soils. Plants can then use nitrogen released by decay of such organisms.

However, some plants have evolved a tighter relationship with N₂-fixing bacteria, involving an exchange of carbon and nitrogen between plant host and bacterial partner (Santi et al. 2013). Several different symbioses of this type have evolved independently (Table 4.2), from primitive algae to higher vascular plants. These associations are characterised by active attraction of the symbiotic partner through metabolites or signals from the host, as well as a physical housing of the symbiont inside the host (Delaux et al. 2015). For example, algae can form a symbiosis with N₂-fixing cyanobacteria to form lichens, and Bryophytes like Anthoceros harbor N₂-fixing cyanobacteria in cavities of their thalli. Roots or leaves of some plants form a loose association with N₂-fixing bacteria, with plant exudates used as a carbon source by the bacteria. In the water fern Azolla, the cyanobacterium Anabaena is located in cavities on the underside of modified leaves, with a secretory trichome delivering sugars and absorbing fixed nitrogen. This symbiosis is especially important in rice paddy fields, where Azolla densely covers the water surface and can then be incorporated into the field soil. In other plants, the N₂ fixers are located in intercellular spaces of the host plant, as reported for sugarcane. These less intimate associations supply host plants with substantial amounts of nitrogen. While such associations have been explored for other grasses, the likely limitation for substantial nitrogen fixation in these loose associations is the limited amount of carbon that the plant provides for its symbionts.

In more highly developed associations, plants localise the symbiotic association within a modified root or ‘nodule’. In cycads, the microsymbiont Anabaena is located in intercellular spaces of the mid-cortex of short, highly branched, modified roots (Figure 4.40a and b). In another class of symbioses, the actinorhizal plants, the micro-symbiont Frankia (an actinomycete, or filamentous bacterium) is located within the cortical cells of a modified root. This group includes the genera Casuarina, Allocasuarina, Alnus, Datisca and Myrica from eight plant families, all belonging to the Rosid I class, as do legumes (Figure 4.40c and d). These plants often grow in nutrient-poor habitats where nitrogen fixation provides a nutritional advantage. Parasponia, a tropical tree native in New Guinea, is the only non-legume known to form a symbiosis with the rod-shaped bacterium Rhizobium. Unlike legumes, the Parasponia nodule has a central vascular bundle and the microsymbiont is always encapsulated within cellulosic material (termed a ‘persistent infection thread’). In legumes, nodules typically have a central infected zone, and rhizobia are enveloped by plasma membrane-derived vesicles called symbiosomes (Figure 4.40e and f).

The symbiosis of legumes with rhizobia is the most effective and agriculturally the most important nitrogen-fixing symbiosis. Rhizobia are soil bacteria from the α- and β-proteobacteria (including Burkholderia sp.) that can occur free-living in the soil, but benefit greatly from symbiosis with legume partners, which provide a carbon source, shelter inside a nodule as well as a niche outside the species-rich and competitive zone of the rhizosphere (See previous section on the Soil-root interface). Many species of rhizobia only effectively fix nitrogen inside a nodule (Figure 4.41).

The ‘precursor’ of symbiosis evolved approximately 100 million years ago, and led to the similar symbioses of actinorhizal plants with Frankia, and the legume symbiosis with rhizobia. Nitrogen fixing symbioses in legumes evolved at a time of relatively high atmospheric CO₂. Nitrogen fixation is likely to have conferred an advantage in using this increased CO₂ for photosynthesis (Sprent 2007). Phylogenetic reconstruction of symbiotic nitrogen fixation suggests that it evolved once, and was subsequently gained and lost several times in legumes (Figure 4.42; Werner et al. 2014).

Nodules formed by members of the family Leguminoseae have a central zone of infected cells, surrounded by a cortex of uninfected cells. A root vascular strand branches within the cortex of the nodule. This structure is quite distinct from nodules of the cycads or actinorhizal plants, which have a central vascular bundle and an infected cortex (a typical root vascular anatomy). Different legume species display various nodule growth patterns, but they can be roughly classified as either of indeterminate growth (i.e. with an apical meristem and consequent elongated shape) or determinate growth (i.e. a spherical meristem which ceases activity at nodule maturity).

The association between rhizobia and legumes is a controlled infection. Typically, the bacterial partner infects the plant through root hairs, and is then encapsulated by polysaccharide material produced by the host plant, forming infection threads. Infection threads then grow into the root cortex, while bacteria multiply within each thread. Finally, bacteria are released from the infection threads and engulfed by plant cells in a form of phagocytosis. This process results in a bacterium (sometimes several) encapsulated by a plant cell membrane. Encapsulating membranes control the delivery of photoassimilate to bacteria, thus ensuring a symbiotic rather than a parasitic relationship. These units are termed ‘symbiosomes’ (see also later Section 4.4.5)

Evolution of this partnership might be similar to that of other endosymbiotic organelles such as mitochondria and chloroplasts. Perhaps a future step in the evolution of a legume–rhizobium symbiosis will be retention of bacteria within plant cells to create a new organelle! If this were to happen, the legume would no longer be dependent on the presence of a microsymbiont for infection. Cells could maintain a low resident population of the new organelle, like plastids in non-photosynthetic tissue, and allow proliferation under set conditions within nodule structures.

In some legume symbioses, bacteria are not released from infection threads. This character is one of several that distinguish each of the three legume subfamilies Caesalpinoideae, Mimosoideae and Papilionoideae (e.g. cassia, acacia and soybean, respectively). The Caesalpinoideae are largely trees or shrubs, and the few species which nodulate have little nodule mass proportional to plant biomass (Sprent and Raven 1985). In most of the caesalpinoid species that do nodulate, the microsymbiont remains encapsulated in an infection thread throughout the life of a nodule. In some species the infection threads are thin walled, while in others bacteria are released into the cytoplasm. The Papilionoideae is considered the most advanced of the legume subfamilies.

Biological interactions between host plant and bacterium are subtle. Just as legumes vary genetically, so do the rod-shaped bacteria (rhizobia) that infect various legumes. Not all rhizobia are equally infective (able to infect and form nodules) or effective (able to fix N₂) on all legumes. An appropriate bacterial partner must therefore be matched genetically with each legume for optimal N₂ fixation. Pure cultures of rhizobia are produced commercially, generally in a peat-moss-based medium or as a seed coating, for inoculating legume seed prior to planting.

Over the last 30 years, the molecular, cellular and genetic analysis of the association between legumes and rhizobia has given us a detailed insight into the specificity and regulation of this symbiosis. As a prerequisite for the symbiosis, the symbionts have to first recognise each other in the soil; the plant host then attracts specific rhizobial symbionts that only nodulate one or a few specific hosts; in response, the rhizobia initiate infection and nodule development. On a whole-plant level, the plant strictly regulates the number of nodules on the root system depending on carbon availability relative to nitrogen demand (See ‘Case Study 4.3 - Regulation of nodule numbers’).

The attraction of rhizobia to the root system of its appropriate host is initiated by the exudation of flavonoids by the plant roots. Flavonoids are a diverse class of phenolic compounds that share a common 15-carbon skeleton consisting of two phenyl rings and a heterocyclic ring, but differ in their exact final structure, for example through modification of the flavonoid ‘backbone’ through hydroxylation, methylation or glycosylation. Over ten thousand different flavonoid molecules have been identified from different plant species. Each legume species exudes a specific mixture of flavonoids into the soil that chemotactically attract their rhizobial symbionts (Figure 4.44). Flavonoids have evolved into specific signalling molecules in the symbiosis by binding specifically to a transcription factor, NodD, inside the rhizobia. This binding of the correct flavonoid activates NodD, and allows it to bind to and activate the promoter regions of a large number of nodulation genes. These nodulation genes of rhizobia are responsible for synthesising a specific signalling molecule, the Nod factor, as well as other genes necessary for symbiosis.

Nod factors are lipochitin-oligosaccharides that are composed of a short ‘backbone’ of N-acetylglucosamine units (the same building blocks that are polymerised to chitin in fungal cell walls, insect exoskeletons and crustacean shells). This backbone typically varies in length between 3-6 units, and is additionally ‘decorated’ with chemical substitutions like acetylation, sulfation or glycosylation (Figure 4.45). Importantly, an acyl side chain of specific length and saturation pattern gives each Nod factor molecule its specificity. Different strains of rhizobia produce specific (mixtures of) Nod factor molecules, which are only recognised by their specific hosts. The Nod factor molecule is crucial for nodulation; mutants defective in Nod factor synthesis cannot nodulate their hosts. Nod factor molecules are also partially sufficient for many of the nodulation steps: application of purified Nod factors to the correct hosts initiates root hair curling and nodule development, but will not lead to infection threads or fully formed nodules. For the complete symbiosis to succeed, other signalling molecules from rhizobia are necessary. For example, surface molecules like exopolysaccharides are necessary to evade host defence responses (Figure 4.45), and communication signals like quorum-sensing signals are necessary to coordinate bacteria behaviours during nitrogen fixation.

Nod factors are recognised by their host by Nod factor receptors. Binding of the correct Nod factor structure activates a signalling cascade in the infected root hairs. A critical part of the signalling cascade is a periodic spiking of calcium concentrations in the root hair (‘calcium spiking’), which starts within minutes of Nod factor perception (Figure 4.45). Calcium spiking activates downstream transcription factors that lead to the activation of cytokinin signaling. The plant hormone cytokinin is necessary, and in some cases sufficient to induce cell divisions in the root cortex, the first step of nodule initiation.

One of the most significant findings of recent years has been the discovery that the same early machinery for Nod factor signal transduction is also required to signal a successful interaction between mycorrhizal fungi and plants (Figure 4.45), and both symbioses share similar mechanisms of bacterial or fungal invasion (Figure 4.46). Mycorrhizal fungi have evolved symbioses with the majority of plant plants much earlier (~400 MYA) than the emergence of the Rhizobium-legume symbiosis (~100 MYA) (see previous section 4.3 on mycorrhiza). Mycorrhizal fungi produce signals that are structurally closely related to Nod factors, and these have been termed ‘Myc factors’. Myc factors are thought to be perceived by different receptors to the Nod factor receptors. Despite some of the shared signals, including calcium spiking, activated by both Myc factors and Nod factors, the later responses in the root differ, leading either to nodulation or to mycorrhization (Figure 4.45). It is still not known what enables legumes to form nodules in response to Nod factors, but other plants not; however, the knowledge that mycorrhizal signalling is part of the response machinery present in all plants that form mycorrhizal symbioses might make it easier to engineer non-legume crops with the ability to form rhizobial symbioses (Rogers and Oldroyd 2014).

Having established the initial contact with the legume host, rhizobia invade the young, just emerging root hairs of their legume partners. In some species, infection can occur intercellularly, and this may have been also the initial mode of infection at the early evolutionary stage of the symbiosis (Held et al. 2014). The infected root hairs curl around small colonies or single cells of rhizobia to form a typical ‘Sheppard’s crook’. The rhizobia locally digest the cell wall of the root hair and induce the formation of ‘infection threads’, tubular invaginations of the root hair in which the rhizobia multiply and travel towards the cortex of the root. This infection step requires correct Nod factor and exopolysaccharide structures from rhizobia, otherwise infection threads are aborted by the plant through defence responses (Figure 4.47). From the infection threads, rhizobia are released into cortical cells as small vesicles, in which rhizobia remain separated from the plant cytoplasm by the plasma membrane. In these vesicles, called symbiosomes, rhizobia differentiate into bacteroids, the nitrogen-fixing forms of rhizobia (Figure 3.49).

Concurrent with the growth of the infection thread towards the root cortex, Nod factors produced by the invading rhizobia trigger the re-initiation of cell division in the cortex. These divisions are activated through cytokinin and auxin gradients that the plant forms in response to Nod factors (Desbrosses and Stougaard 2011). In some legumes and actinorhizal plants, the invading rhizobia do not initiate nodules de novo from cortical cells, but target an emerging lateral root primordium that is then ‘converted’ into a nodule.

One surprising recent discovery has been the identification of rhizobia that nodulate legumes in the absence of Nod factors. Sequencing of a large number of Rhizobium species uncovered rhizobial genomes that do not contain any of the canonical nodulation genes that would be required to make Nod factors. It is still unclear what the signal is that these rhizobia use to regulate infection and nodule formation. Similarly, Frankia symbionts do not produce Nod factors, yet use some of the same symbiotic (SYM) signal transduction cascade as legumes. Unravelling the identity and mode of action of these unknown nodulation signals will be a rewarding future challenge.

(a) Protecting nitrogenase from O₂

A basic conflict arises in biological N₂ fixation: nitrogenase is destroyed by O₂, yet aerobic respiration is essential to sustain the high energy demand of N₂ fixation. Nitrogen-fixing bacteria must be protected from O₂, while a level of aerobic respiration occurs in the host cell cytoplasm. In cycads, cyanobacteria provide their own O₂ protection. Nitrogenase is located in specialised cells (heterocysts) which have an O₂-impermeable lining of glycolipid. An analogous structure (a vesicle) affords protection to nitrogenase in the microsymbiont Frankia within most actinorhizal nodules. In Parasponia and most caesalpinoid nodules the persistent infection threads provide O₂ protection to nitrogenase (Sprent and Raven 1985).

There is one major problem with structures of ‘fixed’ resistance. As respiration rate varies (i.e. O₂ flux), O₂ concentration inside the structure must also vary: following Fick’s Law of diffusion, O₂ flux into the nodule will change in proportion to the O₂ concentration gradient at constant resistance. Free water bathing the nodule is equilibrated with the atmosphere (20.8% O₂), therefore containing approximately 360 µM O₂ in solution, while nitrogenase is destroyed by submicromolar concentrations of dissolved O₂. For practical purposes then, an O₂ gradient of 360 (outside) to 0 (inside) µM must be maintained. If respiration rate was halved without a change in resistance, the O₂ concentration gradient would also halve from 360 to 180 µM. An O₂ concentration of 180 µM O₂ inside would destroy nitrogenase. So, resistance must vary too.

A legume-nodule cortex copes with variations in respiration rate by providing a variable level of O₂ protection (Layzell and Hunt 1990). According to their model, a layer of cells adjacent to the infected zone either lacks radial intercellular spaces (preventing inflow of O₂) or has intercellular spaces filled with water. The thickness of this layer could vary under osmotic control to set nodule permeability. Diffusivity of O₂ through water is about 10,000 times slower than through air, so flooding of radial air spaces in the nodule cortex would be an effective way of decreasing O₂ diffusion into infected tissue.

Any O₂ leaking through this cortex can diffuse freely in the intercellular airspaces of infected tissue and dissolve in the cytoplasm of infected cells. O₂ gradients which might be expected within infected cells because of rapid bacterial respiration are largely avoided by the presence of leghemo-globin (Lb) (a molecule similar to the hemoglobin in mammalian blood). O₂ diffuses to Lb molecules where it is bound to form high concentrations of oxygenated Lb (estimated at 0.7 mM by Bergersen 1982). Effective nodules are pink because of oxygenated Lb (Figure 4.48); indeed this colour change can be used to estimate free O₂ concentrations. Soybean nodules seem to regulate the free O₂ in infected cells at between 5–60 nM (e.g. Layzell and Hunt 1990). Finally, residual O₂ diffuses through the symbiosome to the bacteroids, supporting a level of aerobic respiration.

‘Conventional’ chemistry may not be appropriate when describing O₂ movement in cells because O₂ molecules in cellular compartments are so scarce. A sphere of 1 µm radius — roughly the size of a mitochondrion or bacterium — containing a solution with 10 nM O₂ will contain only 24 molecules of O₂.

(b) Carbon supply and nitrogen export

Nodules are metabolically highly active. A typical maximum rate of nitrogenase activity in soybean nodules, as measured by gas exchange (discussed below) is 300 µmol electron pairs g^–1 (nodule) h^–1. This value is useful to bear in mind when reading the literature about N₂ fixation, with low values possibly indicating unhealthy or disturbed plants. As nitrogenase is at best 75% efficient, with respect to N₂ fixation (Equation 2), this rate is equivalent to the fixation of some 150 µmol N g^–1 (dry weight) h^–1. Reduced nitrogen is exported from nodules to the host plant while carbon is imported into the nodule, supporting energy needs of fixation (through respiration) and providing carbon skeletons for packaging nitrogen as an organic molecule.

Photoassimilate (host to nodule) and nitrogen-based resources (nodule to host) must pass through the endodermis of nodule vascular bundles. Radial walls of this endodermis have Casparian bands and tangential walls have relatively few plasmodesmata, so this cell layer restricts apoplasmic and symplasmic flow of carbon into nodules and nitrogen out of nodules.

The transport of solutes, including C and N metabolites but also metal ions, e.g. Fe and Mo required for nitrogenase, are exchanged via both plant and bacterial transporters on the symbiosome and bacterial membrane, and this transport is carefully controlled to balance supply and demand (Figure 4.49; Udvardi and Poole 2013). Not all rhizobia that invade and inhabit legume nodules fix nitrogen efficiently, with the legume host able to restrict C flow to those nodules that do not fix sufficient N.

The concentration of nitrogenous solutes in the xylem apoplasm causes a hydrostatic pressure to develop, and this results in a mass flow of nodule xylem sap to adjacent roots. The water that accompanies sucrose entering the nodule as phloem sap is re-exported with assimilated nitrogen in the xylem. Nodules are thus analogous to ‘glands’ that secrete nitrogenous compounds.

Rates of N₂ fixation can be measured by a number of techniques to address questions of nodule efficiency and nitrogen cycling in agricultural and natural plant systems. Nitrogenase is pivotal for initial reduction of N₂ but this same enzyme will also reduce acetylene (C₂H₂) to ethylene (C₂H₄). Acetylene is an effective competitor with N₂ for nitrogenase so the rate of C₂H₄ synthesis is proportional to nitrogenase activity. Acetylene reduction gives an instantaneous estimate of the N₂ fixation rate. Another instantaneous technique requires flushing nodulated roots with an argon : oxygen gas mixture (79:21) to displace all N₂. All electron flux through nitrogenase is then diverted to the reduction of protons to H₂ rather than N₂ to NH₄⁺ (Equation 2). The rate of H₂ evolution by roots can thus be used to estimate nitrogenase activity.

Alternative approaches to ‘instantaneous’ estimates of N₂ fixation provide an integrated rate of fixation over periods of hours or days. The proportions of inorganic and organic nitrogen compounds in xylem sap are affected by the ratio of inorganic nitrogen taken up to symbiotic N₂ fixation; this can be exploited in genera of legumes in which amides and ureides are major products of N₂ fixation. Soybean, for example, exports less than 10% of nitrogen to shoots in the form of ureides when supplied nitrate but more than 80% when all nitrogen is biologically fixed. Thus, relative ureide levels in sap give an estimate of N₂ fixation.

Many experiments now rely on ¹⁵N-based techniques to obtain an integral of fixation over the life of a plant. These techniques rely on a difference in ratio of the stable isotopes of nitrogen (¹⁵N and ¹⁴N) in soil and atmosphere (Figure 4.50). The soil must be enriched in ¹⁵N relative to the atmosphere — either naturally (the process of denitrification causes a fractionation of the two isotopes, leaving the soil enriched in ¹⁵N) or by artificial ¹⁵N addition. The N₂-fixing plant of interest is sampled, together with an adjacent non-N₂-fixing plant (e.g. grass) whose ¹⁵N enrichment represents that of soil nitrogen. ¹⁵N enrichment in digested plant material and soil is analysed isotopically in a mass spectrometer and contribution of biological N₂ fixation calculated.

A typical ‘good’ rate of fixation for a (non-irrigated) field of subtropical legumes in northern Australia is c. 60–100 kg N ha^–1 year^–1. About the same amount of nitrogen is harvested as seed from a crop of cowpea, soybean or chickpea, so growing these legumes does not add net nitrogen to the soil; it does, however, spare nitrogen which would otherwise be removed at harvest. Irrigated legume-based pastures in temperate Australia or New Zealand fix 250–300 kg N ha^–1 year^–1 and make a substantial contribution to the low energy costs of agriculture in these regions. Selection of appropriate biological N₂ fixers could greatly improve N₂ fixation in tropical legume crops.

(a) Source processes

 

Net export of photoassimilates occurs from fully expanded leaves (Figure 5.2) and long-term storage pools located along the axial transport pathway. Chloroplasts of C₃ plants (Chapters 1 and 2) partition photoassimilates between the photosynthetic oxidative cycle and starch biosynthesis or release them immediately to the cytosol as triose phosphate for sucrose synthesis. In non-starch-forming leaves, high concentrations of sugars can be accumulated in the vacuoles of mesophyll cells or made available for immediate loading into the phloem and export. Leaves also serve as secondary sources for nutrients and amino acids previously delivered in the transpiration stream. Nutrients and amino acids can be exported in the phloem immediately, or after accumulation in short-term storage pools.

An additional source of photoassimilates is located along the axial phloem path (petioles, stems, peduncles, pedicels and roots) as a result of leakage from the vascular tissues. Leaked photoassimilates accumulate in short- or long-term storage pools which serve as secondary sources to buffer photo-assimilate supplies to the sinks against shifts in export rates from the primary photoassimilate sources.

(b) Path processes

Assimilates including sucrose, amino acids are transferred into sieve elements of fully expanded leaves against significant concentration and electrochemical gradients. This process is referred to as phloem loading. The cellular pathways of phloem loading, and hence transport mechanisms and controls, vary between plant species. Longitudinal transport of assimilates through sieve elements is achieved by mass flow and is termed phloem translocation. Mass flow is driven by a pressure gradient generated osmotically at either end of the phloem pathway, with a high concentration of solutes at the source end and a lower concentration at the sink end. At the sink, assimilates exit the sieve elements and move into recipient sink cells where they are used for growth or storage. Movement from sieve elements to recipient sink cells is called phloem unloading. The cellular pathway of phloem unloading, and hence transport mechanisms and controls, vary depending upon sink function.

(c) Sink processes

Many sink organs are characterised by low rates of transpiration (an exception is a developing leaf) so that most assimilates are delivered by the phloem. Having reached the sink cell cytoplasm through the post-sieve-element transport pathway, assimilates are either metabolised to satisfy the energy, maintenance and growth requirements of sink cells or are compartmented into polymer or vacuolar storage. Collectively, metabolism and compartmentation create a demand for assimilates which is ultimately responsible for driving phloem import.

(a) Whole-plant growth

Sink and source strength must be in balance at a whole-plant level. Thus, an increase in whole-plant sink strength must be matched by an equal increase in source strength, either through increases in source activity or source size. Prior to canopy closure in a crop, much of the increase in source strength comes from increased source size, source activity remaining relatively constant. Significantly, until a leaf has reached some 30% of its final size, photoassimilates for leaf production are exclusively imported through the phloem from fully expanded leaves (Figure 5.2).

(b) Photoassimilate transport and crop yield

During domestication of crop plants, plant breeders selected for crop yield via maximum investment into harvested organs (mostly seeds). Total plant biomass production of advanced wheat is the same as its wild progenitors yet grain yield has increased some 30-fold through breeding. That is, whole-plant source and sink strength have not changed. Increases in wheat yield are associated with a diversion of photoassimilates from vegetative organs to the developing grain, as illustrated by the relative accumulation of ¹⁴C photoassimilates exported from the flag leaf.

Final grain yield is not only determined by partitioning of current photoassimilates, but also depends upon remobilisation of non-structural carbohydrates stored in stems, particularly under conditions where environmental stress impairs leaf photosynthesis (Wardlaw 1990). In fact, remobilisation of reserves affects yield in many food plants. For example, deciduous fruit trees depend entirely on remobilised photo-assimilates to support flowering and fruit set as do early stages of pasture regrowth following grazing.

Photoassimilate transport to harvestable organs plays a central role in crop yield brought about by greater harvest indices. This raises questions about transport and transfer processes that collectively influence photoassimilate partitioning between competing sinks.

Historically, these questions were elucidated by observing partitioning patterns of photoassimilates exported from specified source leaves labelled with ¹⁴C supplied as a pulse of ¹⁴CO₂. Following a chase period, in which ¹⁴C photoassimilates are transported to and accumulated by recipient sink organs, the plant is harvested. The pattern of photoassimilate partitioning operating during the pulse is deduced from ¹⁴C activity accumulated by sinks (Figure 5.3).

Photoassimilates are partitioned from source leaves to sinks in characteristic and reproducible patterns. For instance, in a vegetative plant, lower leaves are the principal suppliers of photoassimilate to roots, whereas upper leaves are the principal suppliers to the shoot apex. Leaves in an intermediate position export equal quantities of photoassimilates in either direction. However, the pattern of photoassimilate partitioning is not static, it changes with plant development. In vegetative plants, the direction of flow from a leaf changes as more leaves above it become net exporters. Furthermore, at the onset of reproductive development, growing fruits or seeds become dominant shoot sinks for photoassimilates at the expense of vegetative apices.

Photoassimilate partitioning patterns can be altered experimentally by removal of selected sources (e.g. leaves) or sinks (e.g. fruits). These manipulative experiments demonstrate that photoassimilate partitioning reflects the relative strengths of individual sources and sinks. Properties of the phloem pathway connecting sources with sinks are shown in the following Section 5.2.

(a) Phloem structure

In most plant species, phloem is made up of phloem fibres, phloem parenchyma, sieve cells (sieve elements) and their accompanying companion cells (Figure 5.5a). Sieve elements are ideally suited for rapid transport of substances at high rates over long distances. They are elongated and are arranged end to end in files referred to as sieve tubes (Figure 5.16b). Abutting sieve elements are interconnected through membrane-lined pores (sieve pores) with large diameters (1 to 15µm). These pores collectively form sieve plates (Figure 5.16c). The transport capacity of sieve tubes is dependent on a developmentally programmed degeneration of the sieve element protoplasm (cell contents) leaving an open, membrane-bound tube. In mature conducting sieve elements, the protoplast is limited to a functional plasma membrane enclosing a sparse cytoplasm containing low densities of plastids, mitochondria and smooth endo-plasmic reticulum distributed along the lateral walls (Figure 5.16d). These relatively empty sieve tubes provide a longitudinal network which conducts phloem sap (Figure 5.5b).

Sieve elements are closely associated with one or more companion cells, forming a sieve element–companion cell (se–cc) complex (Figure 5.5d) that plays an important role in transport. These distinct cell types result from division of a common procambial mother cell. In mature se–cc complexes, relatively open sieve elements contrast with adjacent companion cells containing dense, ribosome-rich cytoplasm with a prominent nucleus and abundant mitochondria and rough endoplasmic reticulum (Figure 5.5d). High densities of extensively branched plasmodesmata in contiguous walls of sieve elements and companion cells (Figure 5.6) account for intense intercellular coupling in se–cc complexes (van Bel 1993). Thus, companion cells are considered to perform the metabolic functions surrendered by, but required for, maintenance of viable sieve elements. This functional coupling has led to the concept of se–cc complexes being responsible for phloem transport.

(b) Visualising the translocation stream

Transport of radioactively labelled substances through phloem has been demonstated using microautoradiography (Figure 5.7), providing irrefutable evidence that sieve elements are conduits for transport of phloem sap. Experimentally, a pulse of ¹⁴CO₂ is fixed photosynthetically and ¹⁴C-labelled sugars are given time to reach the stem, which is then excised and processed for microautoradiography. As ¹⁴C first moves through the stem, most of the isotope is confined to the transport pathway and very little has had time to move laterally into storage pools. High densities of ¹⁴C-labelled sugars are found in sieve elements (Figure 5.7), demonstrating that these cells constitute a transport pathway.

(c) Phloem sealing mechanisms

Herbivory or environmental factors causing physical damage could pose a threat to transport through sieve tubes and has undoubtedly imposed strong selection pressure for the evolution of an efficient and rapid sealing mechanism for damaged sieve tubes. Since sieve tube contents are under a high turgor pressure (P), severing would cause phloem contents to surge from the cut site, incurring excessive assimilate loss in the absence of a sealing mechanism. For dicotyledonous species, an abundant phloem-specific protein (P-protein) provides an almost instantaneous seal. P-protein is swept into sieve pores where it becomes entrapped, thus sealing off the damaged sieve tubes. Production of callose (β-1,3 glucan) in response to wounding or high-temperature stress is another strategy to seal off damaged sieve tubes. Callose also seals off sieve pores during overwintering in deciduous plants. Callose is deposited between the plasma membrane and cell wall, eventually blocking sieve pores. Whether deposited in response to damage or overwintering, callose can be degraded by β-1,3 glucanase, allowing sieve tubes to regain transport capacity.

Since phloem translocation is confined to sieve elements embedded within a tissue matrix, it is difficult to obtain uncontaminated samples of translocated sap. The least equivocal approach has been to take advantage of the high P of sieve tube contents. Puncturing or severing sieve tubes should cause exudation of phloem sap provided a sealing mechanism is not activated.

For some plant species, sieve-pore sealing develops slowly, or can be experimentally down-regulated by massage or repeated excisions (Milburn and Kallarackal 1989) or slowed by puncturing the vasculature while it is snap frozen in liquid N₂ (Pate et al 1984). Carefully placed incisions that do not disturb the underlying xylem, which in any case is more likely to be under tension, permit collection of relatively pure phloem exudate through the severed sieve tubes. Nevertheless, contamination with the contents of cells other than sieve tubes damaged at the site of incision is inevitable.  For the major solutes of phloem such as sugars or amino acids that are present in high concentrations this problem is minimal but for less abundant molecules like hormones or other signals, particularly proteins or nucleic acids, conclusions about the origin and functions of these must be made with caution. The ‘natural hemophiliacs’ of the plant world are few and include a number of cucurbits, some brassicas, castor bean, species of the genus Yucca and some species of lupin (Lupinus albus, L. angustifolius, L. mutabilis and L. cosentinii). The excision technique has been expanded to plant species that do not readily exude, by chemically inhibiting the sealing mechanism. Callose production is blocked when wounded surfaces are exposed to the chelating agent ethylenediaminetetraacetic acid (EDTA) by complexing with calcium, a cofactor for callose synthase. Immersing whole, excised organs in EDTA solution, which is essential to inhibit blockage, risks contaminating sap with solutes lost from the apoplast as well as non-conducting cells. This is not an ideal technique.

Enlisting sap-sucking aphids or leaf hoppers to sample sap has been more successful. Aphids can guide a long syringe-like mouthpart (a stylet) into conducting sieve elements (Figure 5.8). Pressure normally forces sieve-tube sap through the stylet into the aphid’s gut where it becomes food or is excreted as ‘honeydew’. By detaching the aphid from its mouthpart pure phloem contents can be collected from the cut end of the implanted stylet. Detaching the aphid body can be achieved by surgery following rapid anesthesia in high CO₂ or by severing the stylet using a laser.  While stylectomy has been successful with a number of monocotyledons (rice, wheat and barley) the technique has proved more difficult to use with dicotyledons, yielding at best a few microlitres of phloem contents. On the other hand collection of milliliter volumes of exudate from one of the natural hemophiliacs is possible permitting extensive analysis of solutes and macromolecules. In the case of lupins, exudation occurs readily at many sites on the plant so that solutes translocated from source tissues as well as entering sinks can be collected and analysed (Figure 5.9).

(a) Chemical analysis of phloem exudate

Chemical analyses of phloem exudate collected from a wide range of plant species have led to a number of generalisations (e.g. Milburn and Baker 1989) about the contents of sieve tubes. Phloem exudate is a concentrated solution (10–12% dry matter), generating an osmotic pressure (Π) of 1.2 to 1.8MPa. pH is characteristically alkaline (pH 8.0 to 8.5). The principal organic solutes are non-reducing sugars (sucrose), amides (glutamine and asparagine), amino acids (glutamate and aspartate) and organic acids (malate). Of these solutes, non-reducing sugars generally occur in the highest concentrations (300–900 mM). Nitrogen is transported through the phloem as amides and amino acids; nitrate is absent and ammonium only occurs in trace amounts. Calcium, sulphur and iron are scarce in phloem exudate while other inorganic nutrients are present, particularly potassium which is commonly in the range of 60–120 mM. Physiological concentrations of auxins, gibberellins, cytokinins and abscisic acid have been detected in phloem exudate along with nucleotide phosphates. The principal macromolecule group is protein but low levels of peptides and nucleic acids are also present.  While in cucurbits the protein in exudate is comprised largely of P-protein, a diverse array of proteins, many of them enzymes, have also been detected.

(b) Significance of the chemical forms translocated 

Phloem sap provides most inorganic and all organic substrates necessary to support plant growth. Non-transpiring tissues are particularly dependent on resources delivered in the phloem (Section 5.1). That translocated sugars represent the major chemical fraction of the phloem sap is consistent with the bulk of plant dry matter (90%) being composed of carbon, hydrogen and oxygen. Carbon transport is further augmented by transport of nitrogen in organic forms.

Carbohydrate is translocated as non-reducing sugars in which the metabolically reactive aldehyde or ketone group is reduced to an alcohol (mannitol, sorbitol) or combined with a similar group from another sugar to form an oligosaccharide. Apart from sucrose, transported oligosaccharides belong to the raffinose series. In this series, sucrose is bound with increasing numbers of galactose residues to form raffinose, stachyose and verbascose respectively. However, sucrose is the most common sugar species transported. In a small number of plant families, other sugar species predominate. For example, the sugar alcohol sorbitol is the principal transport sugar in the Rosaceae (e.g. apple) and stachyose predominates in the Cucurbitaceae (e.g. pumpkin and squash). Exclusive transport of non-reducing sugars probably reflects packaging of carbohydrate in a chemical form which protects it from being metabolised. Metabolism of these transported sugars requires their conversion to an aldehyde or ketone by enzymes which are thought to be absent from sieve-tube contents.

Plant physiologists have long regarded the two long distance translocation streams of xylem and phloem as having functions additional to the distribution of nutrients and assimilates. Specifically, each serves as a means of communication between the source and sink organs such that systemic signals are thought to transmit molecular responses to endogenous and environmental cues. Furthermore, evidence is accumulating that some of these signals regulate gene expression as a consequence of their translocation (see below). 

(c) Macromolecule composition of phloem

Proteomic and transcriptomic analyses have demonstrated a widely diverse composition of proteins, peptides and nucleic acids, including mRNA and small RNAs, in phloem exudates. While the origin of each individual protein or nucleic acid remains to be verified the limited compositional data available from stylectomy confirms that indeed each group of macromolecules is present in phloem. In cucurbit phloem exudate some 1110 different proteins have been detected along with a large number of mRNAs and similar data have been obtained for exudates from other species (Brassica napus, Ricinus communis and Lupinus albus). Compositional data for phloem proteins of these species show a common complement that includes phloem-specific P proteins together with proteins involved in sugar metabolism and transport, protein turnover and transport, detoxification of reactive oxygen species, as well as proteins that provide defence against insect herbivores and pathogens (Figure 10).  Some undoubtedly play a role in maintenance of the SE system while others, such as the Flowering Locus T (FT) protein associated with the flowering response (‘florigen’), appear to be systemic ‘signals’ (Rodriguez-Medina et al 2011) and there may be many more. Because sieve tubes are enucleate and lack ribosomes (5.2.2 a), proteins in the translocation stream are not formed in situ but are transported from sites of synthesis in the companion cells.

Functional analysis of the cDNA identified in transcriptome studies of phloem exudates revealed transcripts involved in a wide range of processes that include metabolism, plant responses to stresses, transport, DNA/RNA binding and protein turnover.  The presence of transcripts in phloem exudate supports the idea of an RNA-based signalling network that is thought to function in control of processes associated with plant growth and development (Lough and Lucas, 2006).  However, the functional role of transcripts in the contents of sieve tubes as well as their actual translocation is yet to be determined.

Small RNA molecules (18-25 nt) have been identified in phloem exudate collected from rape, white lupin, pumpkin, castor bean and Yucca filamentosa as well as in aphid stylet exudate collected from apple stems. The population includes both microRNAs (miRNA) and small interfering RNAs (siRNA) a large number of which target mRNA of transcription factors that themselves regulate genes expressions. miRNAs are also involved in mediating environmental responses, including responses to salinity, drought, nutrient limitations, as well as hormone interactions. Their small size and powerful functions in targeting mRNAs to regulate expression suggest that those in phloem exudate are likely to be systemic signals.

An important question that relates to the significance of macromolecules in the contents of sieve tubes is proof that they are translocated and that translocation is essential for their function at a sink. A diversity of studies that have exploited cucurbit root stocks and grafted scions has provided clear evidence that P proteins among others are graft transmissible. In a series of elegant experiments Aoki et al. (2005) labelled and injected two isolated pumpkin phloem proteins (CmPP16‑1 and CmPP16-2) into the vasculature of intact rice plants through severed leaf hopper stylets and showed their translocation as well as some evidence for specificity in protein translocation.  The Flowering Locus T (FT) protein formed in leaves mediates the flowering transition of shoot apical meristems and the evidence that it is translocated is compelling. The long distance movement of RNA molecules was first demonstrated for plant viruses and there is now good evidence for phloem translocation of a number of transcripts (Lough and Lucas 2006). A recent compilation identified 13 miRNAs involved in plant responses to drought/salt stress (Covarrubias and Reyes 2010).  Eight of these were identified in lupin phloem exudate (Rodriguez-Medina et al. 2011) and, importantly, six were also recovered from PCR amplification of apple stylet exudate (Varkonyi-Gasic et al 2010).  There is thus a possibility that the responses to drought and salinity are mediated through miRNAs translocated from sites where the stress is sensed to sites where a response is initiated. 

The most convincing case for a translocated miRNA in phloem regulating gene expression relates to Pi homeostasis. While both local and systemic signals are involved, miR399 is phloem mobile and acts directly in roots to down regulate the expression of PHO2 (a ubiquitin conjugating enzyme) that results in greater expression of Pi transporters to increase Pi uptake under conditions of deficiency. Systemic signaling has also been implicated in homeostasis of other nutrients, including N, S and Cu with, in each case, miRNAs involved.