Partitioning of mesophyll conductance for CO₂ into intercellular and cellular components using carbon isotope composition of cuticles from opposite leaf sides

Abstract

We suggest a new technique for estimating the relative drawdown of CO₂ concentration (c) in the intercellular air space (IAS) across hypostomatous leaves (expressed as the ratio c_d/c_b, where the indexes d and b denote the adaxial and abaxial edges, respectively, of IAS), based on the carbon isotope composition (δ¹³C) of leaf cuticular membranes (CMs), cuticular waxes (WXs) or epicuticular waxes (EWXs) isolated from opposite leaf sides. The relative drawdown in the intracellular liquid phase (i.e., the ratio c_c/c_bd, where c_c and c_bd stand for mean CO₂ concentrations in chloroplasts and in the IAS), the fraction of intercellular resistance in the total mesophyll resistance (r_IAS/r_m), leaf thickness, and leaf mass per area (LMA) were also assessed. We show in a conceptual model that the upper (adaxial) side of a hypostomatous leaf should be enriched in ¹³C compared to the lower (abaxial) side. CM, WX, and/or EWX isolated from 40 hypostomatous C₃ species were ¹³C depleted relative to bulk leaf tissue by 2.01–2.85‰. The difference in δ¹³C between the abaxial and adaxial leaf sides (δ¹³C_AB − ¹³C_AD, Δ_b–d), ranged from − 2.22 to + 0.71‰ (− 0.09 ± 0.54‰, mean ± SD) in CM and from − 7.95 to 0.89‰ (− 1.17 ± 1.40‰) in WX. In contrast, two tested amphistomatous species showed no significant Δ_b–d difference in WX. Δ_b–d correlated negatively with LMA and leaf thickness of hypostomatous leaves, which indicates that the mesophyll air space imposes a non-negligible resistance to CO₂ diffusion. δ¹³C of EWX and 30-C aldehyde in WX reveal a stronger CO₂ drawdown than bulk WX or CM. Mean values of c_d/c_b and c_c/c_bd were 0.90 ± 0.12 and 0.66 ± 0.11, respectively, across 14 investigated species in which wax was isolated and analyzed. The diffusion resistance of IAS contributed 20 ± 14% to total mesophyll resistance and reflects species-specific and environmentally-induced differences in leaf functional anatomy.

Estimation of CO2 concentration drawdown across the leaf

Photosynthesis integrated over the leaf profile and ¹³C discrimination

Positive values of net carbon fixation in leaf photosynthesis require CO₂ molecules to enter the leaf along a downward CO₂ concentration gradient from the atmosphere (c_a) to mesophyll chloroplasts (c_c). On the way, CO₂ passes through the laminar boundary layer of air and stomatal pore, diffuses from the substomatal cavity through the IAS of mesophyll, dissolves in water saturating cell walls and finally enters chloroplasts of photosynthesizing cells, where it is assimilated. In hypostomatous leaves where the astomatous leaf surface faces the sun, palisade parenchyma adjoining the upper (adaxial) epidermis is the ultimate end of the CO₂ pathway. However, portions of net CO₂ influx are consumed by mesophyll cells adjoining the stomatous epidermis and in ‘deeper’ mesophyll layers before reaching the palisade cells. Assuming in first approximation that the CO₂ carboxylation takes place at the sun lit adaxial leaf side, net photosynthesis is proportional to the CO₂ concentration difference between the IAS adjoining abaxial and adaxial leaf sides (c_b − c_d) and to the mean value of IAS conductance for CO₂ diffusion between the stomatal cavity and adaxial leaf side (g_IAS):

$$A={g_{{\text{IAS}}}} \cdot \left( {{c_{\text{b}}} - {c_{\text{d}}}} \right)$$

(7)

(for more rigorous integration of A across the leaf on a volume basis see Appendix 1 in Lloyd et al. 1992). Assimilates synthesized in chloroplasts located at opposite leaf sides serve as precursors for long-chain aliphatics presumably deposited in the adjacent cuticle. Here we search for the relationship between CO₂ concentration drawdown across the leaf and carbon isotopic composition of cuticles at the opposite leaf sides. Two parts of the overall CO₂ concentration drawdown in mesophyll may be distinguished: the transversal one across the IAS (c_d/c_b) and one in the cellular liquid phase (c_c/c_d or c_c/c_b where c_c is the CO₂ concentration in chloroplast stroma).

Plants discriminate against ¹³CO₂ during photosynthetic carbon assimilation. This phenomenon was quantified as the deviation (Δ) in isotopic compositions between CO₂ in ambient air (δ_a), and leaf assimilates or dry mass (δ_L) as the product of photosynthetic CO₂ fixation: Δ = (δ_a − δ_L)/(1 + δ_L). The isotopic composition δ shows a relative shift in the ¹³C/¹²C ratio of the sample (R = [¹³C/¹²C]) from the isotopic ratio of PDB carbonate standard (R_s): δ = (R − R_s)/R_s (for details see for example Farquhar et al. 1989). ¹³C discrimination (Δ) in C₃ photosynthesis is related to CO₂ concentrations inside the leaf by a relationship that partitions the overall discrimination into two components: discrimination due to (i) diffusion through stomata and (ii) carboxylation by Rubisco (Farquhar et al. 1982):

$$\varDelta =a \cdot \frac{{({c_{\text{a}}} - {c_{\text{b}}})}}{{{c_{\text{a}}}}}+b \cdot \frac{{{c_{\text{b}}}}}{{{c_{\text{a}}}}},$$

(8)

where a is the carbon isotope fractionation factor during diffusion of CO₂ across the stomata (4.4‰), and b is the net discrimination due to Rubisco carboxylation, CO₂ dissolution and diffusion in water (29–30‰). The relationship assumes no CO₂ drawdown from the substomatal cavities to the chloroplasts (c_b = c_d = c_c).

Carbon isotope composition of CO₂ in the leaf transection

Photosynthetic assimilation of CO₂ penetrating the hypostomatous leaf results in CO₂ concentration drawdown (the difference between c_b and c_d). Further, diffusion through and assimilation by mesophyll change ¹³CO₂ abundance in the IAS and create the difference between δ_b and δ_d. Two opposite effects on δ_b − δ_d can be anticipated: (i) diffusion across the IAS depletes ¹³CO₂ due to kinetic fractionation in the same way as diffusion from the ambient atmosphere adjoining the leaf into the leaf [the first term in Eq. (8)], and (ii) discrimination against ¹³CO₂ during RuBP carboxylation by Rubisco [the second term in Eq. (8)] enriches the IAS gas in ¹³CO₂ diffusing back out of chloroplasts into the IAS. The isotopic depletion due to (i), shown schematically by the line ‘D’ in Fig. 10, can be expressed as

$${\left. {({\delta _{\text{b}}} - {\delta _{\text{d}}})} \right|_{\text{D}}}=a\frac{{{c_{\text{b}}} - {c_{\text{d}}}}}{{{c_{\text{b}}}}},$$

(9)

where the notation |_D indicates the partial effect of diffusion in the IAS on δ_b − δ_d. The partial effect due to (ii), (δ_b − δ_d)|_C (line ‘C’ in Fig. 10), can be approximated using the relationship developed by Evans et al. (1986) for on-line measurements of ¹³CO₂ discrimination. The authors related discrimination in photosynthesizing tissue against ¹³CO₂ in surrounding air (Δ_L) to accumulation of ¹³CO₂ in air passing the leaf (Δ_a) as

$${\varDelta _{\text{a}}}= - \frac{{{\varDelta _{\text{L}}}}}{{\xi \cdot (1+{\varDelta _{\text{L}}})}},$$

(10)

where ξ = c_b/(c_b − c_d). In Evans et al.’s notation, c_b and c_d were the CO₂ concentrations at the chamber inlet (c_i in Evans et al.) and outlet (c_o), respectively, Δ_L showed the discrimination of leaf carbon against ¹³CO₂ in ambient air [and was defined as the isotopic ratio of the source CO₂ in air, R_a, to that of carbon in leaf assimilates, R_L: Δ_L = (R_a/R_L) − 1], and Δ_a was the discrimination in well-mixed air inside the leaf chamber relative to the source air at the leaf chamber inlet [Δ_a = (R_i/R_o) − 1 with the original inlet, outlet notation]. Therefore, Δ_a had negative values as the air around the leaf becomes enriched (R_o > R_i). Our case of CO₂ diffusion across the leaf resembles the on-line discrimination experiment and the relation (10) can be derived using similar steps as in Evans’s original work. The main difference is that CO₂ entering the leaf in our case is not transported by turbulent mass flow of air with a flow rate u [mol s^− 1, see Eq. (7) in Evans et al. 1986] but by diffusion with diffusion conductance g_IAS (mol m^− 2 s^− 1). Therefore, in analogy to Evans’s Eq. (A1), we rearrange our Eq. (7) and write for the balance of carbon (i) entering the leaf by diffusion on the one side and (ii) leaving the leaf by backward diffusion plus incorporation in leaf matter by assimilation: c_b·g_IAS = c_d·g_IAS + A. Since isotopic mass conservation remains valid for both mass flow and diffusion, the derivation leads to the relationship (10), identical to A10 in Evans et al. (1986). At the leaf mesophyll scale, we consider the “input” as identical to the abaxial IAS (_i = _b), and “output” denotes the isotopically modified and CO₂ concentration-reduced adaxial IAS (_o = _d). However, to accept the analogy, we must take the intercellular air in bulk IAS as being isotopically homogeneous, with its CO₂ concentration averaging (c_b + c_d)/2. Then, the discrimination of the averaged IAS against “input” air (Δ_a = Δ_bd) relates to isotopic compositions δ as Δ_bd = [1/2(δ_b − δ_d)]/(1 + δ_d), and the ‘xi’ parameter in fractionation due to carboxylation, ξ_C, is twice the previously defined ξ, i.e., ξ_C = 2c_b/(c_b − c_d). Similarly, discrimination of bulk leaf tissue against ¹³CO₂ in IAS, Δ_L, can be expressed as Δ_L = [1/2 (δ_b + δ_d)] − δ_L)/(1 + δ_L). Substituting Δ_bd for Δ_a as well as the new Δ_L into Eq. (10) and rearrangement yields

$${\left. {\left( {{\delta _{\text{b}}} - {\delta _{\text{d}}}} \right)} \right|_{\text{C}}}=\frac{{\left[ {{\delta _{\text{L}}} - \frac{1}{2}({\delta _{\text{b}}}+{\delta _{\text{d}}})} \right] \cdot (2+{\delta _{\text{b}}}+{\delta _{\text{d}}})}}{{{\xi _{\text{C}}}\left[ {1+\frac{1}{2}({\delta _{\text{b}}}+{\delta _{\text{d}}})} \right]}}.$$

(11)

The values of δ_b + δ_d as well as of 1/2(δ_b + δ_d) are much smaller than 2 and 1, respectively, and δ_L − 1/2(δ_b + δ_d) is − Δ_L. This allows us to simplify the relation (11):

$${\left. {\left( {{\delta _{\text{b}}} - {\delta _{\text{d}}}} \right)} \right|_{\text{C}}}= - \frac{1}{\xi }{\varDelta _{\text{L}}}.$$

(12)

The total effect of diffusion and carboxylation on the isotopic change of CO₂ in the IAS (δ_b − δ_d) is obtained by adding up the contributions of diffusion (9) and carboxylation (12):

$$\left( {{\delta _{\text{b}}} - {\delta _{\text{d}}}} \right)=\frac{1}{\xi }\left( {a - {\varDelta _{\text{L}}}} \right).$$

(13)

It is shown by the thick dot-dash line in Fig. 10.

Fig. 10

Schematic of ¹³C discrimination processes occurring during CO₂ transport and assimilation across the leaf. CO₂ concentration c, indicated on the left-hand vertical axis, which also represents the abaxial stomatous epidermis, decreases as CO₂ diffuses across the leaf and is assimilated in photosynthesis. This process is approximated by the full line connecting the CO₂ concentrations at the inner side of abaxial and adaxial epidermis c_b and c_d, respectively. The isotopic composition of leaf internal CO₂, scaled on the right-hand vertical axis, which also represents the adaxial astomatous epidermis, changes across the mesophyll due to convolution of two processes indicated by the thin dot-dash lines: a slower diffusion of ¹³CO₂ than ¹²CO₂ depleting the intercellular air close to the adaxial epidermis in ¹³C (line D) and b preferential assimilation of the lighter ¹²CO₂ molecules by Rubisco, leaving the inorganic carbon in chloroplasts close to the adaxial side enriched in ¹³CO₂ compared to those near the abaxial side (line C). Combination of these processes eventually results in different isotopic composition of CO₂ in chloroplasts near the abaxial (δ_b) and adaxial (δ_d) leaf sides, represented by the bold dot-dash line. ¹³C discrimination during photosynthetic CO₂ fixation, Δ, is side-specific due to different c_b and c_d and δ_b and δ_d. We hypothesize that isotopic composition of cuticles and waxes from opposite leaf sides, δ_AB and δ_AD, mirrors the transectional differences in chloroplastic CO₂ concentration and isotopic composition. The lines showing the transectional course of c and δ indicate qualitative changes only, not the quantitative pattern

Comparison of cuticles from opposite leaf sides as a proxy for changes in δ of CO₂ across the leaf

The relationship (13) does not provide any feasible way to estimate the relative CO₂ drawdown c_d/c_b (c_d/c_b = − [(1/ξ) −  1]) because the isotopic compositions of abaxial and adaxial intercellular air (δ_b and δ_d) cannot be measured directly. However, as mentioned above, δ_b and δ_d and the relevant concentrations are imprinted in δ of the assimilates synthesized in chloroplasts located in the respective regions of the leaf: c_b and δ_b in assimilates from near the stomatous (abaxial) epidermis, incorporated in the abaxial cuticle, and c_d and δ_d in assimilates synthetized close to the adaxial leaf side and sent to the adaxial cuticle. δ_L derives from the leaf bulk tissue. Therefore, we have tried to find a relationship between the isotopic composition of abaxial and adaxial cuticles or their waxes (δ_AB, δ_AD) and the isotopic composition and concentration of CO₂ in substomatal and palisade intercellular air (δ_b, δ_d and c_b, c_d, respectively).

¹³C discriminations imprinted in abaxial and adaxial cuticles (Δ_AB and Δ_AD) are defined as:

$${\varDelta _{{\text{AB}}}}=\frac{{{\delta _{\text{b}}} - {\delta _{{\text{AB}}}}}}{{1+{\delta _{{\text{AB}}}}}}$$

(14a)

and

$${\varDelta _{{\text{AD}}}}=\frac{{{\delta _{\text{d}}} - {\delta _{{\text{AD}}}}}}{{1+{\delta _{{\text{AD}}}}}}.$$

(14b)

Given that the denominators are close to one, the ¹³C depletions of cuticles are approximated by the differences in the numerators. Mechanistically, the ¹³C content in cuticles can be attributed to fractionation effects during (i) CO₂ transport from the IAS to chloroplast stroma (a_w), (ii) carboxylation by Rubisco (b) and (iii) post-photosynthetic fractionation during fatty acid and wax synthesis (h). The fractionation effects (i)–(iii) are additive and the first two components depend on the CO₂ drawdown from IAS at cell wall (c_b or c_d) to the chloroplast interior (c_cb or c_cd):

$${\delta _{\text{b}}} - {\delta _{{\text{AB}}}}={a_{\text{w}}}\frac{{{c_{\text{b}}} - {c_{{\text{cb}}}}}}{{{c_{\text{b}}}}}+b\frac{{{c_{{\text{cb}}}}}}{{{c_{\text{b}}}}}+h$$

(15a)

and

$${\delta _{\text{d}}} - {\delta _{{\text{AD}}}}={a_{\text{w}}}\frac{{{c_{\text{d}}} - {c_{{\text{cd}}}}}}{{{c_{\text{d}}}}}+b\frac{{{c_{{\text{cd}}}}}}{{{c_{\text{d}}}}}+h.$$

(15b)

Equations 15a, b require the assumption that ¹³C fractionation during synthesis of cuticular compounds (h) is identical for the adaxial and abaxial leaf sides. Similarly, we will assume that the relative drawdown of CO₂ from the IAS to chloroplast stroma is not leaf side-specific (c_cb/c_b = c_cd/c_d, i.e., the CO₂ gradient in the cells across the leaf is homogeneous). Isotopic fractionation due to respiration and photorespiration was assumed to be negligible (Farquhar et al. 1982) and not included in Eqs. (15a), (15b); here, we may alleviate the assumption and allow this fractionation to be non-zero but identical in magnitude at opposite leaf sides. These assumptions simplify the estimation of the isotopic difference between abaxial and adaxial cuticles obtained by subtraction of Eqs. (15b) and (15a):

$${\delta _{\text{b}}} - {\delta _{\text{d}}}={\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}}$$

(16)

which yields an expression for δ_b − δ_d alternative to that shown in Eq. (13).

Relative drawdown of CO₂ concentration in the gas phase

Substitution of Eq. (16) into (13) and rearrangement allows to factor out the δ_b − δ_d term:

$${\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}}=\frac{1}{\xi }\left( {a - {\varDelta _{\text{L}}}} \right).$$

(17)

Rearrangement and expression of ξ in terms of CO₂ concentrations yield

$$\xi =\frac{{{c_{\text{b}}}}}{{{c_{\text{b}}} - {c_{\text{d}}}}}=\frac{{a - {\varDelta _{\text{L}}}}}{{{\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}}}}$$

(18)

and the relative drawdown of CO₂ in the IAS, c_d/c_b, is:

$$\frac{{{c_{\text{d}}}}}{{{c_{\text{b}}}}}=1 - \frac{{{\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}}}}{{a - {\varDelta _{\text{L}}}}}.$$

(19)

Typically, leaf tissue is depleted in ¹³C compared to the source CO₂ (Δ has a positive value) by more than the value of a, so the term a − Δ_L is negative. We expect that the abaxial cuticle should be more depleted than the adaxial one, thus (δ_AB − δ_AD) < 0. Therefore, the second term on the right side of Eq. (19) should be positive and less than 1 and, thus, the drawdown c_d/c_b range between zero and one. The variable Δ_L represents ¹³C discrimination in bulk leaf assimilates against ¹³CO₂ in air inside the leaf. Assuming in first approximation that the IAS air is isotopically identical with ‘adaxial air’ (Δ_L ≅ δ_d − δ_L) and substituting δ_d with δ_b from (16), relation (19) can be reformulated as:

$$\frac{{{c_{\text{d}}}}}{{{c_{\text{b}}}}}=1 - \frac{1}{{1+\frac{{{\delta _{\text{L}}} - {\delta _{\text{b}}}+a}}{{{\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}}}}}}.$$

(20)

In a more rigorous alternative, if Δ_L is defined against ‘average’ IAS air [Δ_L = (δ_bd − δ_L)/(1 + δ_L), where δ_bd is the isotopic composition of the average IAS air: δ_bd = δ_b − 1/2(δ_b − δ_d)=δ_b − 1/2(δ_AB − δ_AD)], Eq. (19) is transformed to the following formula:

$$\frac{{{c_{\text{d}}}}}{{{c_{\text{b}}}}}=1 - \frac{{({\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}})(1+{\delta _{\text{L}}})}}{{(a+{\delta _{\text{L}}})(1+{\delta _{\text{L}}})+\frac{1}{2}({\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}}) - {\delta _{\text{b}}}}}.$$

(21)

Relative drawdown of CO₂ concentration in the liquid phase

In analogy to Eq. (15), we can write for the discrimination in mesophyll cells against ¹³CO₂ in the IAS with the mean concentration c_bd = 1/2(c_b + c_d) and ¹³C abundance δ_bd = 1/2(δ_b + δ_d):

$${\delta _{{\text{bd}}}} - {\delta _{\text{L}}}={a_{\text{w}}} \cdot \left( {1 - \frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}}} \right)+b\frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}},$$

(22)

where a_w is the combined fractionation factor for CO₂ dissolution (1.2‰) and diffusion (0.6‰) in cell walls and cell interior (a_w = 1.8‰). Rearrangement yields the expression that we have used in assessment of the CO₂ drawdown in the cellular liquid phase, c_c/c_bd, in C₃ plants:

$$\frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}}=\frac{{{\varDelta _L} - {a_{\text{w}}}}}{{b - {a_{\text{w}}}}},$$

(23)

where Δ_L = (δ_bd − δ_L)/(1 + δ_L) and δ_bd = δ_b − 1/2(δ_AB − δ_AD). Finally, c_c/c_bd can be expressed in terms of δ_L, (δ_AB − δ_AD) and δ_b as:

$$\frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}}=\frac{{{\delta _{\text{b}}} - {\delta _{\text{L}}} - {a_{\text{w}}}(1+{\delta _{\text{L}}}) - 0.5({\delta _{{\text{AB}}}} - {\delta _{{\text{AD}}}})}}{{(b - {a_{\text{w}}})(1+{\delta _{\text{L}}})}}.$$

(24)

The magnitude of δ_b in Eqs. (21) and (24) may be approximated by evaluating the isotopic effect of CO₂ diffusion from ambient air at the leaf surface into substomatal cavities as δ_b = δ_a − a(1 − c_i/c_a). Values of − 8.0‰, 4.4‰ and 0.75 were used for δ_a, a and c_i/c_a, respectively, which yields δ_b = − 9.1‰ as the value typical for C₃ plants. It should be noted that the relations (21) and (24) are derived for δ and isotope fractionation factors expressed in fractional notation (values add up to 1). For calculations in per mille, the term (1 + δ_L) has to be substituted by (1 + δ_L/1000).

Partitioning of mesophyll conductance

Net photosynthesis rate integrated over the intracellular pathways across the leaf profile can be expressed in an alternative form to Eq. (7) as

$$A={g_{{\text{liq}}}}\left( {{c_{{\text{bd}}}} - {c_{\text{c}}}} \right).$$

(25)

Resistances (the inverse values of conductances) for diffusion of CO₂ in the IAS (r_IAS), the cell interior (r_liq) and the whole mesophyll (r_m) are

$${r_{{\text{IAS}}}}=\frac{{{c_{\text{b}}} - {c_{\text{d}}}}}{A},$$

(26a)

$${r_{{\text{liq}}}}=\frac{{{c_{{\text{bd}}}} - {c_{\text{c}}}}}{A},$$

(26b)

$${r_{\text{m}}}={r_{{\text{IAS}}}}+{r_{{\text{liq}}}}$$

(26c)

which, provided that c_b and c_bd are not far apart, yields the relative portion of IAS in total mesophyll diffusion resistance or conductance (r_IAS/r_m and g_IAS/g_m):

$$\frac{{{r_{{\text{IAS}}}}}}{{{r_{\text{m}}}}} \cong \frac{1}{{1+\frac{{1 - \frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}}}}{{1 - \frac{{{c_{\text{d}}}}}{{{c_{\text{b}}}}}}}}},$$

(27a)

$$\frac{{{g_{{\text{IAS}}}}}}{{{g_{\text{m}}}}} \cong 1+\frac{{1 - \frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}}}}{{1 - \frac{{{c_{\text{d}}}}}{{{c_{\text{b}}}}}}}.$$

(27b)

The analog expressions for the cellular (liquid) path are:

$$\frac{{{r_{{\text{liq}}}}}}{{{r_{\text{m}}}}} \cong \frac{1}{{1+\frac{{1 - \frac{{{c_{\text{d}}}}}{{{c_{\text{b}}}}}}}{{1 - \frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}}}}}},$$

(28a)

$$\frac{{{g_{{\text{liq}}}}}}{{{g_{\text{m}}}}} \cong 1+\frac{{1 - \frac{{{c_{\text{d}}}}}{{{c_{\text{b}}}}}}}{{1 - \frac{{{c_{\text{c}}}}}{{{c_{{\text{bd}}}}}}}}.$$

(28b)