A Transgenic Approach to Understanding the Influence of Carbonic Anhydrase on C¹⁸OO Discrimination during C₄ Photosynthesis

Abstract

The oxygen isotope composition of atmospheric CO₂ is an important signal that helps distinguish between ecosystem photosynthetic and respiratory processes. In C₄ plants the carbonic anhydrase (CA)-catalyzed interconversion of CO₂ and bicarbonate (HCO₃⁻) is an essential first reaction for C₄ photosynthesis but also plays an important role in the CO₂-H₂O exchange of oxygen as it enhances the rate of isotopic equilibrium between CO₂ and water. The C₄ dicot Flaveria bidentis containing genetically reduced levels of leaf CA (CA_leaf) has been used to test whether changing leaf CA activity influences online measurements of C¹⁸OO discrimination (Δ¹⁸O) and the proportion of CO₂ in isotopic equilibrium with leaf water at the site of oxygen exchange (θ). The Δ¹⁸O in wild-type F. bidentis, which contains high levels of CA relative to the rates of net CO₂ assimilation, was less than predicted by models of Δ¹⁸O. Additionally, Δ¹⁸O was sensitive to small decreases in CA_leaf. However, reduced CA activity in F. bidentis had little effect on net CO₂ assimilation, transpiration rates (E), and stomatal conductance (g_s) until CA levels were less than 20% of wild type. The values of θ determined from measurements of Δ¹⁸O and the ¹⁸O isotopic composition of leaf water at the site of evaporation (δ_e) were low in the wild-type F. bidentis and decreased in transgenic plants with reduced levels of CA activity. Measured values of θ were always significantly lower than the values of θ predicted from in vitro CA activity and gas exchange. The data presented here indicates that CA content in a C₄ leaf may not represent the CA activity associated with the CO₂-H₂O oxygen exchange and therefore may not be a good predictor of θ during C₄ photosynthesis. Furthermore, uncertainties in the isotopic composition of water at the site of exchange may also limit the ability to accurately predict θ in C₄ plants.

The oxygen isotope composition (δ¹⁸O) of atmospheric CO₂ is an important tool for monitoring variations in the global exchange of CO₂ (Farquhar et al., 1989, 1993; Farquhar and Lloyd, 1993; Yakir and Wang, 1996; Flanagan and Ehleringer, 1998; Yakir and Sternberg, 2000). The use of isotopes in this context relies on the fact that photosynthesis and ecosystem respiration generally have different effects on the isotope composition of atmospheric CO₂ and the δ¹⁸O of CO₂ in the atmosphere has been used to distinguish between the photosynthetic CO₂ uptake and CO₂ release during respiration (Farquhar et al., 1993; Gillon and Yakir, 2001). Dissolved CO₂ exchanges oxygen molecules with water allowing the CO₂ to take on the isotopic signature of the water that is in a much higher molar concentration. The water in leaves is generally highly enriched in ¹⁸O relative to ground water due to the preferential evaporation of H₂¹⁶O during leaf transpiration (Yakir and Wang, 1996; Yakir and Sternberg, 2000). Furthermore the exchange of ¹⁸O between CO₂ and water is facilitated in leaves by the presence of carbonic anhydrase (CA), which catalyzes the interconversion of CO₂ and bicarbonate (HCO₃⁻). Thus, the retrodiffusion of CO₂ out of a leaf during photosynthesis drives the ¹⁸O enrichment of atmospheric CO₂. Root respiration and decomposition of organic material by soil microbes work in the opposite direction, releasing CO₂ depleted in ¹⁸O because it is in equilibrium with unenriched water (Flanagan and Ehleringer, 1998; Yakir and Sternberg, 2000).

To accurately interpret the δ¹⁸O of atmospheric CO₂ requires an understanding of the isotopic fractionation steps associated with specific processes during leaf gas exchange (Yakir and Sternberg, 2000). The δ¹⁸O of the water at the site CO₂-H₂O oxygen exchange within a leaf (δ_ex) is the primary factor influencing the δ¹⁸O of CO₂ diffusing out of a leaf (Farquhar et al., 1993). Additionally, the proportion of CO₂ in isotopic equilibrium with the water at the site of oxygen exchange (θ) will also influence the δ¹⁸O of the CO₂. The value of θ is determined by the balance between the gross flux of CO₂ into the leaf and the activity of CA at the site of CO₂-H₂O oxygen isotope exchange as this influences the residence time of CO₂ within a leaf and thus the number of hydration reactions per CO₂ molecule (Gillon and Yakir, 2000a, 2000b).

The CA-catalyzed hydration of CO₂ to HCO₃⁻ is the first enzymatic step of the C₄ photosynthetic pathway, a biochemical CO₂ concentrating mechanism that concentrates CO₂ around Rubisco in bundle sheath cells (BSC; Hatch, 1987; Kanai and Edwards, 1999). The HCO₃⁻ is subsequently fixed via phosphoenol pyruvate carboxylase into a four-carbon acid that diffuses to the BSC for decarboxylation (Kanai and Edwards, 1999). The majority of CA in a C₄ plant occurs within the mesophyll cytoplasm and the CA-catalyzed oxygen exchange between CO₂-H₂O occurs throughout the cytoplasmic space (Ku and Edwards, 1975; Burnell and Hatch, 1988; Hatch and Burnell, 1990). This is in contrast to C₃ plants where the chloroplasts, which contain the majority of the leaf's CA, are appressed against the cell walls adjacent to the intercellular air space and the sites of water evaporation during transpiration. In C₃ plants the majority of leaf CA activity is generally thought to be in close proximity to the water at the site of evaporation (Williams et al., 1996; Gillon and Yakir, 2000a). Under such conditions the isotopic signature of water at the site of evaporation (δ_e) is likely similar to the site of CO₂-H₂O oxygen exchange δ_ex; however, this may not be the case in C₄ species.

Online measurements of C¹⁸OO discrimination (Δ¹⁸O) during gas exchange of leaves have found Δ¹⁸O to be much lower in two C₄ monocots (Zea mays and Sorghum bicolor) than what is commonly observed for C₃ species (Gillon and Yakir, 2000a, 2000b). This lower discrimination can be explained by a combination of lower stomatal conductances and low CA activities assuming that the CO₂-H₂O oxygen exchange takes place in water with the same isotopic signature as the water at the site of evaporation (Gillon and Yakir, 2000b). In a survey of leaf CA activity of species belonging to different functional types it was found that C₄ dicots contain more CA than their C₄ monocot counterparts (Gillon and Yakir, 2001). The large amount of CA relative to net rates of CO₂ assimilation in C₄ dicots suggests that Δ¹⁸O ought to be higher in these plants, but this has not yet been tested (Gillon and Yakir, 2001). In this article we examine the influence changing CA activity has on Δ¹⁸O and the proportion of CO₂ in isotopic equilibrium with leaf water at the site of oxygen exchange (θ) during C₄ photosynthesis in a C₄ plant with high levels of CA activity.

The efficient genetic transformation of the C₄ dicot Flaveria bidentis provides the opportunity to manipulate various aspects of the C₄ photosynthetic pathway, allowing this plant to be used as a model system for understanding the controls and limitation of C₄ photosynthesis and isotope exchange (Chitty et al., 1994; Furbank et al., 1997; von Caemmerer et al., 1997a; Ludwig et al., 1998; Cousins et al., 2006). Additionally, under steady-state conditions F. bidentis has higher stomatal conductance than most other C₄ plants allowing fractionation factors other than stomatal limited diffusion to influence Δ¹⁸O. These attributes make F. bidentis an excellent model system to study how changes in C₄ photosynthesis and CA activity influence the isotopic exchange of CO₂. Here we present data from online measurements of Δ¹⁸O, leaf CA activity and the isotope composition of transpired water in F. bidentis with genetically modified levels of leaf CA. The antisense suppression of CA activity provides us with the unique opportunity to asses the role of CA in determining Δ¹⁸O without modifying environmental conditions. We show that Δ¹⁸O in F. bidentis leaves, which contain high levels of CA activity, is nevertheless sensitive to small changes in CA activity even when stomatal conductances and CO₂ assimilation rates are unaffected. These results are discussed in relationship to the isotopic signature of water at the site of exchange and the internal conductance of CO₂ diffusion from the intercellular airspaces to the site of phosphoenol pyruvate carboxylase reaction.

ISOTOPE THEORY

The δ¹⁸O of water at the sites of evaporation within a leaf (δ_e) can be estimated from the Craig and Gordon model of evaporative enrichment (Craig and Gordon, 1965; Farquhar and Lloyd, 1993)

(1)

where e_a and e_i are the vapor pressures in the atmosphere and the leaf intercellular spaces. δ_a and δ_t are the isotopic composition of water vapor in the air and transpired by the leaf, respectively. The kinetic fractionation during diffusion of water from leaf intercellular air spaces to the atmosphere (ɛ_k) can be calculated as (Cernusak et al., 2004)

(2)

where r_s and r_b are the stomatal and boundary layer resistance and 32 and 21 are the fractionation factors in parts per mil. The equilibrium fractionation between liquid water and water vapor (ɛ⁺) is calculated as (Cernusak et al., 2004)

(3)

where T is leaf temperature in degrees Kelvin. Under steady-state conditions the value of δ_t is equal to the isotopic composition of source water (δ_s), the water taken up by the plant (Harwood et al., 1998). A summary of the symbols used in the text are listed in Table I.

Table I.

Symbols used in the text

Symbol	Description
A	Net CO2 assimilation
a	Fractionation during diffusion of CO2 from the mesophyll to the atmosphere (7.7‰)
α	Fractionation factor for 18O between water and CO2 at 25°C (41.1‰)
Δea	18O enrichment of CO2 at the site of exchange compared to the atmosphere when the CO2 is in full isotopic equilibrium with the water
Δca	18O enrichment of CO2 at the site of exchange compared to the atmosphere
Δ13C	13CO2 isotope discrimination
Δ18O	C18OO isotope discrimination
δ18O	Isotopic ratio of oxygen (‰ relative to VSMOW)
δa	18O isotopic composition of water vapor in the air (‰ relative to VSMOW)
δc	The dilution corrected value of δ18O
δe	18O isotopic composition of water at the site of evaporation in the leaf (‰ relative to VSMOW)
δex	18O isotopic composition of water at the site of CO2-H2O oxygen exchange (‰ relative to VSMOW)
δg	δ18O value of the equilibrating CO2
δm	Measured δ18O value
δt	18O isotopic composition of water transpired by the leaf (‰ relative to VSMOW)
	18O isotopic composition of water sample (‰ relative to VSMOW)
	Dilution corrected 18O isotopic composition of the water sample
	Dilution corrected 18O isotopic composition of reference water no. 1
	Dilution corrected 18O isotopic composition of reference water no. 2
	The 18O isotopic composition of reference water no. 1 relative to VSMOW (−6.44‰)
	The 18O isotopic composition of reference water no. 2 relative to VSMOW (−22.83‰)
E	Transpiration rate
ea	Vapor pressure in the atmosphere
ei	Vapor pressure in the leaf intercellular air spaces
ɛ	pm/(pa − pm)
ɛk	Kinetic fractionation during diffusion of water from leaf intercellular air spaces to the atmosphere
ɛw	The equilibrium 18O fractionation between CO2 and water (40.17‰ at 30°C)
ɛ+	The equilibrium 18O fractionation between liquid and vapor water (8.77‰ at 30°C)
Fin	Gross influx of CO2 into a leaf
gs	Stomatal conductance
gw	The internal conductance to the diffusion of CO2 between the intercellular air space and the site of carboxylation in the mesophyll cytoplasm
k	The ratio of oxygen atoms in the water to the oxygen atoms in CO2
kCA	The hydration rate constant of CA
kt	The number of hydration reactions per CO2 molecule
pCO2	Partial pressure of CO2
pa	pCO2 of dry air in the atmosphere
pe	pCO2 of dry air entering the leaf chamber
pm	pCO2 of the mesophyll cytoplasm and site of CO2-H2O oxygen exchange
po	pCO2 of dry air leaving the leaf chamber
Re	13C/12C of the air entering the leaf chamber
Ro	13C/12C of the air leaving the leaf chamber
rb	Boundary layer resistance to water vapor diffusion (m2 s mol−1)
rs	Stomatal resistance to water vapor diffusion (m2 s mol−1)
θ	The proportion of CO2 in isotopic equilibrium with water at the site of oxygen exchange
ξ	pe/(pe − po)
τ	Residence time of CO2 in the leaf cytosol

Discrimination against C¹⁸OO (Δ¹⁸O) when water at the site of exchange and CO₂ are at full isotopic equilibrium (θ = 1) can be predicted (Farquhar and Lloyd, 1993) as

(4)

where a is the diffusional discrimination (7.7‰) and ɛ is calculated as p_m/(p_{a −} p_m) where p_a is the CO₂ partial pressure (pCO₂) in air and p_m is the pCO₂ at the site of leaf CO₂-H₂O oxygen exchange in the mesophyll cytosol. The ¹⁸O enrichment of CO₂ compared to the atmosphere at the site of exchange in full oxygen isotope equilibrium with the water was calculated as (Cernusak et al., 2004)

(5)

where the equilibrium fractionation between water and CO₂ (ɛ_w) can be calculated as (Cernusak et al., 2004)

(6)

where T is leaf temperature in degrees Kelvin. The proportion of CO₂ in isotopic equilibrium with water at the site of oxygen exchange can be calculated from (Gillon and Yakir, 2000a)

(7)

where Δ_ca is the oxygen isotope composition of CO₂ at the site of exchange during photosynthesis determined by

(8)

where Δ¹⁸O is the discrimination against C¹⁸OO as defined above.

There has been much interest in the role CA plays in enhancing isotopic equilibrium (θ) because it has a significant influence on the atmospheric CO₂ isotope signature. It has been suggested that the extent of θ in a leaf can be determined by in vitro CA assays coupled with the unidirectional flux of CO₂ into the leaf (Gillon and Yakir, 2000a, 2000b, 2001) from the equation initially developed by Mills and Urey (1940)

(9)

where CA_leaf/F_in represents the mean number of hydration reactions for each CO₂ molecule inside the leaf (Gillon and Yakir, 2000a). Leaf CA activity (CA_leaf) is determined as the product of the CA hydration rate constant (k_CA, μmol m⁻² s⁻¹ Pa⁻¹) and the mesophyll pCO₂ (p_m). The rate constant k_CA is calculated from in vitro measurements of CA activity in leaf extracts (see “Materials and Methods”). The gross influx of CO₂ into a leaf (F_in = g_t p_a) as well as p_m determine the residence time (τ = p_m/F_in) of CO₂ within the leaf. The parameter g_t is defined as the total conductance of CO₂ from the atmosphere to the site of CO₂-H₂O oxygen exchange (Gillon and Yakir, 2000a). The relationship of CA_leaf/F_in indicates that conditions that influence p_a, p_m, g_t, or k_CA can alter the value of θ.

RESULTS

Light Response and Isotopic Equilibrium

The online Δ¹⁸O values were determined by directly coupling a mass spectrometer to the outlet of a LI-6400 gas exchange system via a gas permeable silicone membrane (Cousins et al., 2006). This allowed the measurements of the C¹⁸OO/C¹⁶OO ratio of the CO₂ in the air stream without prior purification of CO₂. We obtained a range of Δ¹⁸O values and rates of net CO₂ assimilation by manipulating the light conditions that F. bidentis and tobacco (Nicotiana tabacum) leaves were exposed to during the gas exchange measurements (Table II). Leaf CA activity (CA_leaf), determined as the product of the rate constant (k_CA, μmol m⁻² s⁻¹ Pa⁻¹) and the mesophyll pCO₂ (p_m), increased as rates of net CO₂ assimilation decreased under limiting light conditions because p_m increased (Table II). The k_CA was determined from leaf extracts using mass spectrometry to measure the rates of ¹⁸O₂ exchange from labeled ¹³C¹⁸O₂ to H₂¹⁶O (von Caemmerer et al., 2004; Cousins et al., 2006). The CA_leaf activity for tobacco was similar to earlier published values (Williams et al., 1996; Gillon and Yakir, 2000a), whereas CA_leaf in Z. mays was higher than previously reported by Gillon and Yakir (2001) but similar to a more recent publication (Affek et al., 2006). F. bidentis plants with reduced levels of Rubisco (caused by antisense RNA constructs targeted to the nuclear-encoded gene for the small subunit of Rubisco [SSU; anti-SSU plants]) had lower photosynthetic rates than wild-type plants and higher CA_leaf activity at similar light levels because of higher values of p_m (Table II).

Table II.

Gas exchange, CA activity, Δ¹⁸O, and isotopic equilibrium measurements

Net CO₂ assimilation rate (A), the pCO₂ at the site of CO₂-H₂O oxygen exchange in the mesophyll cytosol (p_m), the ratio of mesophyll cytosolic to ambient CO₂ partial pressure (p_m/p_a), the residence time of CO₂ in the aqueous phase within the leaf (τ), stomatal conductance (g_s), leaf CA activity (CA_leaf), online Δ¹⁸O discrimination, and proportion of chloroplast CO₂ in isotopic equilibrium with chloroplast water (θ) determined from online Δ¹⁸O measurements (Eq. 7) and predicted from in vitro CA assays and gas exchange (Eq. 9) in F. bidentis and tobacco measured at various irradiances (μmol quanta m⁻² s⁻¹) and the F. bidentis with reduced levels of Rubisco (anti-SSU) and wild-type Z. mays measured at 2,000 (μmol quanta m⁻² s⁻¹). Other conditions are as described in the legend to Figure 1. g_w = 10 mol m⁻² s⁻¹ Pa⁻¹ for all plants except tobacco where g_w = 5 mol m⁻² s⁻¹ Pa⁻¹. n = 3 to 5.

Plant	Irradiance	A				gs	CAleaf	Δ18O	Measured	Predicted
	μmol m−2 s−1	μmol m−2 s−1	pm	pm/pa	τ	mol m−2 s−1	μmol m−2 s−1	‰	θ	θ
F. bidentis wild type	150	6.2 ± 0.2	410 ± 19	0.77 ± 0.03	11.5 ± 1.6	0.12 ± 0.03	2,855 ± 602	182 ± 38	1.06 ± 0.04	1.00 ± 0.01
	300	12.2 ± 0.2	363 ± 32	0.70 ± 0.06	7 ± 1	0.21 ± 0.07	2,558 ± 652	103 ± 29	0.89 ± 0.02	1.00 ± 0.01
	800	25.9 ± 0.9	283 ± 28	0.59 ± 0.06	3.8 ± 0.4	0.33 ± 0.09	2,000 ± 532	43 ± 10	0.65 ± 0.04	0.99 ± 0.01
	1,400	33.7 ± 1.3	252 ± 25	0.56 ± 0.05	2.8 ± 0.2	0.41 ± 0.07	1,844 ± 262	31 ± 5	0.54 ± 0.04	0.99 ± 0.01
	2,000	37.6 ± 1.2	234 ± 15	0.53 ± 0.03	2.6 ± 0.1	0.42 ± 0.04	1,868 ± 180	23 ± 2	0.45 ± 0.04	0.99 ± 0.01
Tobacco	300	15.9 ± 0.9	161 ± 14	0.45 ± 0.04	5.0 ± 0.3	0.18 ± 0.03	1,197 ± 194	37 ± 4	0.89 ± 0.08	0.99 ± 0.01
	500	21.1 ± 1.5	142 ± 13	0.41 ± 0.04	3.6 ± 0.2	0.25 ± 0.04	1,060 ± 182	29 ± 2	0.81 ± 0.10	0.99 ± 0.01
	800	24.7 ± 1.9	130 ± 11	0.39 ± 0.04	2.9 ± 0.2	0.30 ± 0.06	976 ± 183	23 ± 1	0.67 ± 0.04	0.99 ± 0.01
F. bidentis anti-SSu	2,000	23.6 ± 3.1	323 ± 5	0.68 ± 0.02	4.0 ± 0.7	0.40 ± 0.09	2,285 ± 231	51 ± 3	0.63 ± 0.02	0.99 ± 0.01
Z. mays	2,000	33.9 ± 1.5	86 ± 8	0.19 ± 0.02	1.9 ± 0.1	0.18 ± 0.01	296 ± 32	13 ± 1	0.62 ± 0.10	0.88 ± 0.03

In both wild-type F. bidentis and tobacco the predicted isotopic equilibrium (θ) determined from the in vitro CA assays (Eq. 9) did not vary with irradiance, whereas θ determined from the measured Δ¹⁸O (Eq. 7) decreased dramatically with increasing irradiance (Table II). The residence time (τ = p_m/F_in) of CO₂ within the leaf increased in both F. bidentis and tobacco at the low irradiances (Table II). This was due to an increase in p_m as well as a decrease in F_in caused by a reduction in g_s (Table II). In both species, the values of θ predicted from in vitro CA assays (Eq. 9) were substantially higher than the values measured from Δ¹⁸O (Eq. 7), except at the low light levels where the predicted and measured values of θ started to converge. In F. bidentis we used the low light measurement, when isotopic equilibrium is expected to be complete (i.e. θ = 1), to estimate an internal conductance to the diffusion of CO₂ between the intercellular air space and the site of CO₂-H₂O oxygen exchange (g_w) of 10 mol m⁻² s⁻¹ Pa⁻¹ (method described by Gillon and Yakir, 2000a). The value of g_w for tobacco (5 mol m⁻² s⁻¹ Pa⁻¹) was determined from Δ¹³C measurements as described by von Caemmerer and Evans (1991). Both the predicted θ values from in vitro CA assays (Eq. 9) and the measured values determined from Δ¹⁸O (Eq. 7) in Z. mays were lower than in wild-type F. bidentis plants at similar light levels (Table II). The F. bidentis anti-SSU plants with reduced levels of Rubisco had higher online determined values of θ and τ than wild-type F. bidentis at similar light conditions because of higher values of p_m (Table II).

Gas Exchange and CA Activity

We obtained plants with a range of CA activity from the progeny of F. bidentis plants containing the antisense RNA constructs targeted toward the putative cytosolic CA (anti-CA plants, see Fig. 1). As previously reported, decreases in CA activity in F. bidentis had little effect on net CO₂ assimilation until CA levels were less than 20% of wild type (Fig. 1A; see also von Caemmerer et al., 2004; Cousins et al., 2006). The transpiration rate (E) and stomatal conductance (g_s) recorded after approximately 1.5 h in the leaf chamber under steady-state conditions were similar in plants with a large range of leaf CA activities (Fig. 1, B and C). However, both E and g_s were lower (5.1 versus 7.5 for E and 0.36 versus 0.17 for g_s) in the anti-CA plants with dramatically reduced CA activity and net CO₂ assimilation rates compared to wild-type plants (Fig. 1).

Figure 1.

Net CO₂ assimilation rate, transpiration rate (E), and stomatal conductance (g_s) as a function of the rate constant of leaf CA (k_CA mol m⁻² s⁻¹ Pa⁻¹). Each point represents a measurement made on a different plant grown in a glass house at ambient CO₂ or in a growth cabinet at 0.96 kPa CO₂: wild-type plants grown at ambient CO₂ (□); anti-CA plants grown at ambient CO₂ (▪); wild type grown at 0.96 kPa CO₂ (○); and anti-CA plants grown at 0.96 kPa CO₂ (•). Gas exchange measurements were made at an irradiance of 2,000 μmol quanta m⁻² s⁻¹, leaf temperature of 30°C and an inlet CO₂ concentration of 52 Pa in 90.5 kPa of N₂ and 4.8 kPa of O₂ gas mixture.

δ¹⁸O of Water at the Site of Evaporation

The oxygen isotope composition of water at the evaporative site in the leaves (δ_e) determined from the model of Craig and Gordon (1965) and further developed by Farquhar and Lloyd (1993) ranged between 18‰ and 27‰ (Eq. 1). We waited a minimum of 1.5 h under steady conditions to measure Δ¹⁸O, and subsequently collected transpired water to calculate δ_e. Under steady-state conditions the isotopic composition of transpired water vapor (δ_t) should be equal to the isotopic composition of source water (δ_s, the water take up by the root system). The values of δ_t and δ_s were similar for plants grown under 1% CO₂, −6.2 ± 0.3‰ and −5.1 ± 0.4‰, respectively. However, δ_t and δ_s in the glass house-grown plants were −2.6 + 0.5‰ and −5.3 ± 0.3‰, respectively. Values of δ_e, calculated from Equation 1 using values of δ_t, were similar between anti-CA and wild-type plants grown under ambient CO₂ in the glass house (Fig. 2). However, δ_e in wild-type plants grown in 1% CO₂ tended to be more depleted in ¹⁸O, whereas δ_e in the anti-CA plants grown under similar conditions was more enriched in ¹⁸O (Fig. 2). The difference in δ_e between the high CO₂-grown wild-type and low anti-CA plants is attributed to the differences in E and g_s, which caused ɛ_k (Eq. 2) to increase from 30 ± 0.1‰ to 31 ± 0.1‰ and the ratio of e_a/e_i (used in Eq. 1) to decrease from 0.45 ± 0.01‰ to 0.32 ± 0.02‰. These changes in E and g_s as well as the shift in p_m/p_a contribute to the large shift in Δ¹⁸O in the low anti-CA plants (see below).

Figure 2.

The isotopic composition (‰) of the water at the site of evaporation (δ_e) and the oxygen isotope discrimination (Δ¹⁸O) as a function of the rate constant of leaf CA (k_CA mol m⁻² s⁻¹ Pa⁻¹). Values of δ_e were calculated as described in the text using Equation 1 and source water (δ_s) was −5.3 ± 0.3‰. The line represents a linear regression (R² = 0.87) for all data except the plants with high Δ¹⁸O values. Symbols and gas exchange conditions are as in Figure 1 and δ values are expressed in reference to VSMOW.

Measured C¹⁸OO Discrimination and Isotopic Equilibrium

Although there was no detectable change in the net CO₂ assimilation rate in anti-CA plants with greater than 20% of wild-type CA activity, Δ¹⁸O decreased with reduced CA activity (Fig. 2B). There was a strong relationship between Δ¹⁸O and CA activity, where Δ¹⁸O decreased with CA activity until low levels of CA dramatically altered the rates of net CO₂ assimilation. Plants with low photosynthetic rates had Δ¹⁸O values markedly higher than the other anti-CA plants corresponding to the high p_m/p_a in these plants (Figs. 2B and 3).

Figure 3.

Oxygen isotope discrimination (Δ¹⁸O) as a function of the ratio of mesophyll cytosolic to ambient CO₂ partial pressure (p_m/p_a). Where p_m was calculated with g_w = 10 mol m⁻² s⁻¹ Pa⁻¹ for the C₄ plants and 5 mol m⁻² s⁻¹ Pa⁻¹ for tobacco. The line represents the theoretical relationship of Δ¹⁸O and p_m/p_a at full isotopic equilibrium where a = 7.7‰ and Δ_ea = 33.7‰ (Eq. 4) and the CO₂ supplied to the leaf had a δ¹⁸O of 24‰ relative to VSMOW. Each point represents a measurement made on a different plant grown under ambient conditions in a glass house or at 0.96 kPa CO₂ in growth chambers. Measurement conditions are as in Table II and Figure 1. Each point represents a measurement made on a different plant or different light level: wild-type F. bidentis (□); anti-CA F. bidentis (▪); anti-SSU F. bidentis (▴); wild-type Z. mays (⋄); and wild-type tobacco (▵). Plants grown in ambient and 0.96 kPa CO₂ were grouped together in this figure.

The values of Δ¹⁸O increased with p_m/p_a as predicted from Equation 4, but the measured values for F. bidentis were always less than those predicted at full isotopic equilibrium with an assumed constant Δ_ea of 33.7‰ (Fig. 3). The value of Δ_ea was taken as the average value determined from all wild-type plants at high irradiance (2,000 μmol quanta m⁻² s⁻¹). At similar p_m/p_a values the anti-CA plants, with wild-type rates of net CO₂ assimilation, had lower Δ¹⁸O and were thus further away from the predicted values of Δ¹⁸O. The values of Δ¹⁸O and p_m/p_a were high in the anti-CA plants with extremely low CA activity compared to wild-type plants due to the lack of net CO₂ assimilation (Fig. 3). However, the measured values of Δ¹⁸O were even further away than wild type from the predicted Δ¹⁸O values. F. bidentis plants with reduced amounts of Rubisco (anti-SSU plants) have a high p_m/p_a, due to the low rates of net CO₂ assimilation, but wild-type levels of extractible CA activity. Their measured values of Δ¹⁸O are closer to the predicted values of Δ¹⁸O, compared to wild-type plants (Fig. 3). The values of Δ¹⁸O for both Z. mays and tobacco correspond to lower p_m/p_a and were closer to the predicted values of Δ¹⁸O compared to wild-type F. bidentis (Fig. 3).

Using Equation 9 and in vitro CA assays to predict θ indicated that wild-type plants should be at complete isotopic equilibrium with water at the site of oxygen exchange (Fig. 4A). This is in contradiction to the relatively low θ values determined from measurements of Δ¹⁸O and Equation 7 (Fig. 4B), highlighting the discrepancy in θ determined from either Equation 7 or 9. The values of θ predicted from in vitro CA assays (Eq. 9) were substantially higher in both the wild-type and anti-CA F. bidentis plants than the values measured from Δ¹⁸O (Eq. 7), as shown in Figure 4 and Table II. Both methods of determining θ indicated a decrease in θ as the amount of CA decreased in the leaves of anti-CA plants (Fig. 4), although values estimated from in vitro CA assays were more sensitive to changes in CA (Fig. 4A).

Figure 4.

The extent of isotopic equilibrium (θ) as a function of the rate constant of leaf CA (k_CA mol m⁻² s⁻¹ Pa⁻¹) in wild-type and anti-CA F. bidentis plants. The predicted values of θ were determined from in vitro CA assays using Equation 9 (A) and the measured values of θ were determined from Δ¹⁸O using Equation 7 (B). Symbols and gas exchange conditions are as in Figure 1. Calculations were made with g_w = 10 mol m⁻² s⁻¹ Pa⁻¹.

The Water at the Site of Exchange, Internal CO₂ Conductance, and the Extent of Isotopic Equilibrium

The value of θ calculated from the online isotope measurements (Eq. 7) is influenced not only by Δ¹⁸O but also by the partial pressure of CO₂ and the isotopic signature of leaf water at the site of CO₂-H₂O oxygen exchange, p_m and δ_ex, respectively. The Craig-Gordon model describes the water enrichment at the site of evaporation (δ_e), but does not provide information on the isotopic gradient of water within a leaf. It is generally assumed that δ_e is equal to δ_ex, but theoretically δ_ex can vary between δ_e (δ_e = 18‰ to 25‰) and source water (δ_s= −5.3 ± 0.3‰). Figure 5 shows that modeled changes in δ_ex shifts θ from around 0.4 to 1.4, demonstrating that δ_ex is crucial for calculating θ when using Equation 7. The model is also used to vary g_w, which is used to estimate ɛ in Equation 4, to see how g_w influences the value of θ estimated from Δ¹⁸O measurements (Fig. 5).

Figure 5.

The predicted isotopic equilibrium (θ) determined with various isotopic compositions of water at the site of CO₂-H₂O oxygen exchange (δ_ex). Calculations were made with g_w values ranging from 10 to 5 mol m⁻² s⁻¹ Pa⁻¹. θ was calculated using Equation 7 where Δ_ea was determined using Equation 5 and substituting δ_e with various values of δ_ex. Δ_ca was calculated from parameters taken from the high light wild-type F. bidentis measurements in Table II. δ_ex values are presented relative to VSMOW.

DISCUSSION

The Influence of CA on C¹⁸OO Discrimination during C₄ Photosynthesis

We measured Δ¹⁸O concurrently with gas exchange in transgenic F. bidentis with a range of leaf CA activity due to an antisense construct targeted to a putative cytosolic CA to explore the role CA plays in Δ¹⁸O in a C₄ species. These plants have been previously described by von Caemmerer et al. (2004) and Cousins et al. (2006). Our measurements show that even small reductions in CA activity resulted in decreased Δ¹⁸O (Fig. 2) and reductions in the extent of isotopic equilibrium, θ, estimated from Δ¹⁸O (Fig. 4). However, these plants showed no differences in the rate of net CO₂ assimilation or stomatal conductance (Fig. 1). It has been reported that low levels of leaf CA, particularly in C₄ monocots, limits the extent of θ (Gillon and Yakir, 2000a, 2000b, 2001). The limited amount of CA activity in these leaves is thought to be insufficient to fully equilibrate the exchange of oxygen between CO₂ and water. However, the C₄ dicot F. bidentis has high levels of leaf CA, compared to these C₄ monocots, with levels comparable to many C₃ plants (Table II; Gillon and Yakir, 2000a, 2000b, 2001; von Caemmerer et al., 2004; Cousins et al., 2006). It was therefore intriguing to measure low Δ¹⁸O and θ values in wild-type F. bidentis and observe the responsiveness of Δ¹⁸O to small reductions in CA activity.

The value of θ is related to the mean number of hydration reactions a CO₂ molecule experiences inside a leaf. This in turn is the product of residence time (τ = p_m/F_in) and the hydration constant of leaf CA, k_CA (as described in the theory section). We therefore also experimentally thought to increase θ by using transgenic F. bidentis with reduced Rubisco activity. This increased τ, Δ¹⁸O, and the measured θ values because reduced assimilation rates are not matched by reductions in stomatal conductance, thus leading to an increased p_m and p_m/p_a (Table II; Fig. 3). This further suggests that CA activity is not saturating for isotopic equilibrium in wild-type F. bidentis plants.

CA Activity as a Predictor of Isotopic Equilibrium

In both Z. mays and F. bidentis the value of θ determined from Δ¹⁸O measurements was considerably less than that estimated from in vitro measurements of CA activity that predicted θ values close to one for all species (Table II; Fig. 4). As previously reported and shown in Table II, θ in the C₃ plant tobacco determined from Δ¹⁸O measurements was higher than in C₄ plants (Williams et al., 1996; Gillon and Yakir, 2000a, 2000b). As noted above, the extent of θ is in part determined by the CA activity associated with the CO₂-H₂O oxygen exchange, but knowing the isotopic signature of the water at the site of exchange (δ_ex), as well as knowing the pCO₂ at the site of exchange, are also both important factors (see below).

The discrepancy between our estimates of θ in F. bidentis from online measurements and the in vitro CA activity (Fig. 4) could be due to the fact that measurements of total leaf CA activity are not representative of the CA activity associated with the CO₂-H₂O oxygen exchange that influences Δ¹⁸O. Multiple forms of CA have been reported from leaf tissue of both C₃ and C₄ species (Badger and Price, 1994; Badger, 2003). In C₄ plants the majority of CA is located within the mesophyll cytoplasm (Ku and Edwards, 1975; Burnell and Hatch, 1988; Hatch and Burnell, 1990) and in F. bidentis approximately only 5% of the total leaf CA activity is specifically located within the BSC cells (Ludwig et al., 1998). In C₃ plants the chloroplasts, which contain the majority of the leaf's CA, are appressed against the cell walls adjacent to the intercellular air space, thus positioning CA in close proximity to the water at the site of evaporation. Under such conditions it is likely that the isotopic signature of water at the site of evaporation (δ_e) is similar to δ_ex and this may explain why in tobacco we found quite close agreement between θ estimated from Δ¹⁸O and in vitro CA activity (Gillon and Yakir, 2000a, 2000b). However, in C₄ plants, because the CA activity is distributed throughout the mesophyll cytoplasm, the majority of CA may not be positioned adjacent to the intercellular air spaces and the site of evaporation. This suggests that whole leaf in vitro CA assays likely overestimate the amount of CA important for determining Δ¹⁸O in C₄ plants. The different localization of CA activity in C₃ and C₄ species may be in part responsible for the different relationship between in vitro CA activity and Δ¹⁸O in F. bidentis and tobacco.

The Water Isotopic Signature and the Internal Conductance to CO₂

The Craig-Gordon model (Eq. 1) is used for predicting the isotopic composition of water at the site of evaporation, δ_e, but there are a number of studies showing that there can be a large isotopic gradient within a leaf between the water supplied via the vascular tissue and water at the site of evaporation (Farquhar and Gan, 2003; Barbour and Farquhar, 2004; Barbour et al., 2004). As shown in Figure 5, hypothetically varying the isotopic signature of the water at the site of exchange (δ_ex) has a large influence on the calculated values of θ. The values of δ_ex presented in Figure 5 are a reasonable isotopic gradient to consider when δ_e in the F. bidentis plants varied between 19‰ and 25‰ and source water was −5.3 ± 0.3‰. Due to the fact that the majority of CA occurs within the mesophyll cytoplasm in C₄ plants it is possible that the CO₂-H₂O oxygen exchange occurs in a range of δ_ex values. Indeed the average δ_ex values may be closer to the isotopic composition of source water (Fig. 5). Thus in C₄ plants water at the site of evaporation may not be the only water that influences Δ¹⁸O.

The difference between the measured and predicted values of Δ¹⁸O is also due to uncertainty in determining the pCO₂ within the mesophyll cytoplasm (p_m) that is used to calculate ɛ in Equation 4. Knowing the internal conductance to CO₂ (g_w) is necessary to accurately calculate values of p_m; however, in C₄ plants it is difficult to determine g_w with the traditional Δ¹³C measurements because of the low Δ¹³C values (Evans and von Caemmerer, 1996). We have used Δ¹⁸O, as described by Gillon and Yakir (2000a), under low light conditions when θ = 1 would be expected to estimate values of g_w in F. bidentis (Table II). However, Figure 5 shows that dramatically varying g_w has only a small influence on the calculated θ values and although g_w is important for estimating p_m, uncertainty in this value would not be sufficient to raise the measured θ values to the values predicted by the in vitro CA assays.

The Influence of Light on CA Activity and Isotopic Equilibrium

The increase in CA_leaf activity at low photosynthetic rates is attributed to the increase in substrate availability for CA due to the lack of photosynthetic CO₂ drawdown at the low irradiances. The increase in p_m causes CA_leaf to increase, assuming k_CA remains constant with changing light conditions, because CA activity is generally limited by CO₂. Additionally, at low light the residence time of CO₂ (τ = p_m/F_in) is greater as p_m increases and the gross flux of CO₂ into the leaf (F_in = g_t p_a) decreases due to the reduction in g_t caused by the drop in g_s (Table II). Therefore, the number of hydration reactions per CO₂ increases at low light as demonstrated by CA_leaf/F_in in Equation 9. It is unknown if CA activity in C₄ leaves is modulated with changes in irradiances, but potentially the redox statues of mesophyll chloroplast could regulate CA activity (Lee et al., 2004). Under high light conditions, stomatal conductance (g_s) in F. bidentis and tobacco is high compared to Z. mays (0.42 ± 0.04, 0.30 ± 0.06, and 0.18 ± 0.01, respectively; see Table II). In Z. mays the low g_s causes the Δ¹⁸O to be largely influenced by the stomatal fractionation whereas in F. bidentis and tobacco the higher values of g_s allow for the greater potential influence of CA on Δ¹⁸O (Fig. 4). Therefore, in Z. mays the low g_s influences the values of θ more so than in F. bidentis or tobacco.

Using Equation 9, C₄ plants with less CA (e.g. grasses) would be predicted to have a much larger shift in θ due to the changing light conditions. This is in contrast to F. bidentis where the high CA_leaf activity predicted that θ would be close to 1 under all light conditions (Table II). The variation in θ under different light conditions may have important implications for predicting the influence of plant communities on the δ¹⁸O of atmospheric CO₂.

CONCLUSION

The amount of CA activity in a leaf plays an important role in determining C¹⁸OO discrimination during C₄ photosynthesis. We have shown that Δ¹⁸O and the extent of CO₂-H₂O isotopic equilibrium (θ) in F. bidentis leaves, which contain high levels of in vitro CA activity relative to rates of net CO₂ assimilation, is sensitive to small changes in CA activity even when rates of net CO₂ assimilation are unaffected. We conclude that the cytoplasmic localization of the majority of CA in C₄ species has important implications for predicting Δ¹⁸O and the estimation of θ under different environmental conditions. It appears that in vitro CA activity on whole leaf extract of F. bidentis may not represent the CA activity associated with the CO₂-H₂O oxygen exchange and therefore may not be a good predictor of θ. The ability to accurately predict θ in C₄ species will be limited by the uncertainties in CA activity involved in the oxygen exchange and knowing the isotopic composition of water at the site of exchange.

MATERIALS AND METHODS

Growth Conditions

Flaveria bidentis plants were previously transformed with antisense RNA constructs targeted to either the nuclear-encoded gene for the SSU (anti-SSU plants) or a putative cytosolic CA (anti-CA plants; Furbank et al., 1996; von Caemmerer et al., 1997b, 2004). The segregating T1 generations of anti-CA primary transformants with photosynthetic rates similar to wild type were grown during the summer months in a glass house under natural light conditions (27°C day and 18°C night temperatures). The wild-type tobacco (Nicotiana tabacum) plants were also grown in a glass house under similar conditions. Anti-CA (segregating T2 generation from primary transformants 15, 12, and 8) and anti-SSU plants (segregating T2 generation from primary transformant 136-13) with low photosynthetic capacities and wild-type plants were grown under 0.96 kPa of CO₂ in a controlled environment growth cabinet at an irradiance of 400 μmol quanta m⁻² s⁻¹ at plant height and air temperature of 27°C during the day and 18°C at night with a day length of 14 h. Zea mays plants were grown in similar growth cabinet conditions at ambient CO₂ concentrations. All plants were grown in 5 L pots in garden mix with 2.4 to 4 g Osmocote/L soil (15/4.8/10.8/1.2 N/P/K/Mg plus trace elements: B, Cu, Fe, Mn, Mo, and Zn; Scotts Australia Pty) and watered daily.

Gas Exchange Measurements

The uppermost fully expanded leaves were placed into the leaf chamber of the LI-6400 (LI-COR) and equilibrated under measurement conditions for a minimum of 1.5 h (Cousins et al., 2006). Air entering the leaf chamber was prepared by using mass flow controllers (MKS Instruments) to obtain a gas mix of 90.5 kPa dry N₂ and 4.8 kPa O₂ (Cousins et al., 2006). A portion of the nitrogen/oxygen air was used to zero the mass spectrometer to correct for N₂O and other contaminates contributing to the 44 and 46 peaks. Pure CO₂ (δ¹⁸O = 24‰ VSMOW) was added to the remaining air stream to obtain a CO₂ partial pressure of approximately 52 Pa. Low oxygen (4.8 kPa) was used to minimize contamination of the 46 (mass-to-charge ratio) signal caused by the interaction of O₂ and N₂ to produce NO₂ with the mass spectrometer source.

The gas mixtures were fed to the inlet of the LI-6400 console and a flow rate of 200 μmol s⁻¹ was maintained over the leaf. The remaining air stream was vented or used to determine the isotopic composition of air entering the leaf chamber (Cousins et al., 2006). The efflux from the leaf chamber was measured by either replacing the match valve line with a line connected directly to the mass spectrometer or by placing a tee in the match valve line allowing flow to both the mass spec and the match valve simultaneously. Gas exchange parameters were determined by the LI-6400 and pCO₂ leaving the chamber was subsequently corrected for the dilution of CO₂ by water vapor (von Caemmerer and Farquhar, 1981).

Online Isotopic Measurements

The efflux from the leaf chamber and the gas mix supplied to the LI-6400 system was linked to a mass spectrometer through an ethanol/dry ice water trap and a thin, gas permeable silicone membrane that was housed in a temperature-controlled cuvette. The masses (mass-to-charge ratio) 44 and 46 were monitored continuously and the oxygen isotope discrimination during CO₂ exchange, Δ¹⁸O, was calculated from the ratio of mass 46 to 44 in the reference air, determined before and after each sample measurement, entering the chamber (R_e), and the composition of the sample air leaving the leaf chamber (R_o) as (Evans et al., 1986)

(10)

where ξ = p_e/(p_e − p_o), and p_e and p_o are the pCO₂ of dry air entering and leaving the leaf chamber, respectively. A summary of the symbols used in the text are listed in Table I. Zero values for the 44 and 46 peaks were determined before and after the sample measurements were subtracted from both the sample and reference measurements prior to determining the mass ratios. The zero values were typically 1% of the 44 and 6% of the 46 peak.

Collection and ¹⁸O Isotopic Measurements of Water Vapor

A line connected directly to the exhaust port of the LI-6400 was used to cryogenically trap transpired water in a modified glass collection line submerged in an ethanol/dry ice bath. Water vapor was collected for 45 min (netting 50–100 μL of water depending on transpiration rates) and allowed to warm to room temperature in a sealed collection tube. The tube was subsequently centrifuged and placed into an ice water bath for 30 min. The liquid water was removed from the tube and stored at 4°C in screw-top sample vials (Alltech) until measured. To measure the δ¹⁸O of the collected water, 10 mL head space vials (Alltech) with crimped tops containing butyl septa (Alltech) were flushed with 1.9 kPa CO₂ (in a N₂ background) at 2 L min⁻¹ for 3 min and allowing the pressure inside the vials to remain at atmospheric levels. Water samples (25 μL) were injected through the septa with a gas-tight syringe (SGE) and the CO₂-H₂O was allowed to equilibrate for 48 h on a shaker at room temperature (Fessenden et al., 2002).

Prior to the isotopic measurements the vials were placed for a minimum of 2.5 h on a temperature block set a 25°C. The CO₂ samples were analyzed by injecting 200 μL of the headspace gas into a 500 μL N₂ purged gas-tight temperature-controlled cuvette containing a Teflon gas permeable membrane linked to a mass spectrometer (micromass ISOPRIME, Micromass). Masses 44 and 46 were monitored continuously and the zero values determined before and after the sample measurements were subtracted from the values prior to determining the mass ratios. The zero values were typically 3% to 4% of the 44 and 46 peak. Two standard laboratory waters were measured during each measurement to calibrate our measured values against known standards. Our standard waters (S1 = −6.44 and S2 = −22.83) were calibrated by the Stable Isotope Facilities in the Earth Environment Group within the Research School of Earth Sciences at The Australian National University.

The measured δ¹⁸O value (δ_m) was corrected for the contribution of oxygen from the CO₂ used for equilibration and normalized against VSMOW as outlined in Scrimgeour (1995). The dilution corrected value was obtained by

(11)

where δ_g is the signal of the equilibrating gas, α is the fractionation factor for ¹⁸O between water and CO₂ (41.1‰ at 25°C), and k is the ratio of oxygen atoms in the water to the oxygen atoms in the CO₂ (k = 78). The δ_c values were then normalized to VSMOW using the two standard waters (−6.44‰ and −22.83‰) and the following equation:

(12)

where the superscripts SAM, S1, and S2 refer to the sample in question and the reference waters 1 and 2, respectively. The subscripts c and VSMOW refer to the dilution corrected values and those calibrate against VSMOW, respectively. The precision of analyses, based on the repeated measurements of gas samples sealed in vials was 0.1‰ (1 sd, n = 8).

CA Activity Measurements

CA activity was measured on leaf extracts using mass spectrometry to measure the rates of ¹⁸O₂ exchange from labeled ¹³C¹⁸O₂ to H₂¹⁶O (Badger and Price, 1989; von Caemmerer et al., 2004; Cousins et al., 2006). Measurements of leaf extracts were made at 25°C with a subsaturating total carbon concentration of 1 mm. The hydration rates were calculated from the enhancement in the rate of ¹⁸O loss over the uncatalyzed rate and the nonenzymatic first order rate constant was applied (pH 7.4 appropriate for the mesophyll cytosol). The CA activity was reported as a first order rate constant k_CA (mol m⁻² s⁻¹ Pa⁻¹) and k_CAp_m gives the in vivo CA activity at that particular cytosolic pCO₂. Leaf samples were collected after the gas exchange measurements on the same leaf material and subsequently frozen in liquid nitrogen and stored at −80°C.