3 | RESULTS
In control leaves, the pH drift end point was reached after nearly 24 hours at a mean pH of 10.2 (Figure 1, Figure 2a) and a very low final CO2 concentration of ~0.03 µM (about 450-fold below air equilibrium, and at an oxygen concentration of about 353 µM, about 137% of air-equilibrium; Figure 2b) indicating that HCO3- had been used. In leaves treated with AZ or DIDS, the pH drift stopped after 6 to 12 hours and the end point did not exceed pH 9.3; final CO2 concentrations were between 0.8 and 1.6 µM (Figure 1, Figure 2a,b), indicating that HCO3- use had been inhibited. As a consequence of HCO3- use in control leaves, rates of Ci uptake were about 40 µmol g-1 DW h-1 even at the very low CO2concentrations (Figure S2). The slope of Ci uptake vsconcentration of CO2 between 15 and 40 µM in leaves treated with AZ was between 54.1% and 70.6% lower than the control (P<0.05) and in leaves treated with DIDS, it was about 35% lower than the control (P<0.05; Figure 2c). In contrast, the intercept CO2 compensation points increased significantly as a result of the addition of AZ (Figure 2d). The higher AZ concentration treatments had a CO2 compensation concentration close to 20 µM (at an oxygen concentration of 163 µM) suggesting that CCM is absent. These results suggest that AZ not only inhibited CAext but also inhibited the AE protein. The CO2 compensation concentration in the presence of DIDS, at about 5 µM (at an oxygen concentration of 232 µM, about 90% of air-equilibrium), was not significantly different from the control but substantially lower than in the two AZ treatments (Figure 2d). The CT/alkalinity quotient (the remaining total Ci at the end of the drift, CT related to the alkalinity) is a measure of the effectiveness of Ci depletion. A low quotient indicates that a large proportion of the Ci pool is available for acquisition and vice versa. While HCO3- use in control leaves allowed about half of the available inorganic carbon to be accessible, in the AZ and DIDS treated leaves, a high quotient was obtained and only between 11 and 16% of the available inorganic carbon was accessible (Figure 2e).
Figure 3 shows the Ci uptake rates at different CO2concentrations calculated from the pH-drift experiments over a pH range from about 7.7 to 9.3. AZ inhibited Ci-uptake at all the CO2 concentrations (Figure 3a), and both AZ concentrations inhibited Ci uptake by between 70 and 76% when the concentrations of CO2 were between 2.6 and 11 µM. In contrast, DIDS did not affect Ci uptake at CO2concentrations above 4.2 µM but inhibited Ci uptake by about 40% at CO2 concentrations between about 1 and 4 µM (Figure 3b). The inhibitory effect caused by AZ at both concentrations, can be completely reversed by washing since the post-control rates of Ci uptake were not significantly different from the initial control (P>0.05; Figure 4). This confirms that AZ does not penetrate the plasmalemma (Moroney et al., 1985) and thus that the observed effects are linked to inhibition of CAext.
The inhibition of Ci uptake rates in the presence of 0.1 mM AZ and 0.3 mM DIDS were not significantly different in leaves acclimated to HCvs LC, although there was a slightly greater inhibition by 0.2 mM AZ in HC compared to LC leaves (P<0.05; Figure 5a,b). CAext activity was present in both HC and LC leaves but it was greater in LC leaves (P<0.01; Figure 5c). CAext activity was inhibited by AZ: the 0.2 mM treatment caused a greater inhibition than 0.1 mM AZ (Figure 5d). DIDS had no effect on CAext activity neither in HC nor in LC leaves. Ci uptake rates, measured at an initial CO2concentration of 12 µM, were broadly positively related to the activity of CAext (R2 = 0.84 and 0.74 for HC and LC leaves respectively, P<0.01).
The inhibition of Ci uptake in O. alismoides by AZ and DIDS implied that both CAext and anion exchange protein were present. This was characterized further using transcriptomic analysis: mRNA for putative alpha carbonic anhydrase 1 (αCA-1) and HCO3- transporters were expressed. Fifty-three transcripts were functionally annotated to CA according to sequence similarity and translated into 66 peptides. Six of these peptides were homologous with αCA1 based on a comparison of amino acid sequences with the NCBI database and corresponded to four CA isoforms (Figure 6a, Figure S3). Isoform 1 in O. alismoides shows 60% and 61% identity with the chloroplastic isoform X1 and X2 of αCA-1 from the monocot Musa acuminata . Isoforms 2, 3 and 4 show 58%, 55% and 56% identity with the isoform X1 from this species, respectively, as well as 59%, 57% and 58% identity with the isoform X2. However, according to Target P1 software, all the isoforms from O. alismoides were predicted to be localized in the secretory pathway (Figure 6b). The expression of the four isoforms of putative αCA-1, was not significantly different in HC and LC acclimated leaves (P>0.05, Figure 6c).
Unfortunately, transcripts of HCO3-transporters were not detected due to the lower sensitivity of TGS, but were present in the dataset from SGS. Fifteen peptides sequences (Figure S3) were inferred to be homologous to HCO3- transporter family with the following dicot species in the database: Artemisia annua(70.6-78.9%), Corchorus olitorius (73.1-85.7%), Corchorus capsularis (73.1-85.7%), Cynara cardunculus (80.4-85.7%),Lupinus albus (73.22%), Macleaya cordata (76.2-83.5%),Parasponia andersonii (74.0%), Populus alba (77.6%),Prunus dulcis (78.5-82.4%), Striga asiatica(81.6-83.5%), Theobroma cacao (79.2-85.0%) and Trema orientale (75.1-75.8%). This HCO3-transporter family contains Band 3 anion exchange proteins, which also known as anion exchanger 1 or SLC4 member 1. Only partial sequences could be deduced from our analysis and since the peptides for putative HCO3- transporters are membrane proteins, their location could not be predicted. The mRNA expressions of all the transcripts for putative SLC4 HCO3- transporters were not significantly different in HC and LC acclimated leaves (P>0.05, data not shown); the expression-data for the highest expressed transcript for SLC4 HCO3- transporters is presented in Figure 6d.