4 | DISCUSSION
O. alismoides possesses three CCMs, including constitutive abilities to (i) use HCO3- and (ii) operate C4 photosynthesis, and a facultative ability to perform CAM when acclimated to low CO2 concentrations (Zhang et al., 2014; Shao et al., 2017; Huang et al., 2018). We confirm here that this species has a constitutive ability to use HCO3-, and this allows it to exploit a large proportion of the Ci pool and drive CO2 to very low concentrations.
In this study, multiple lines of evidence show that an external CA, putative αCA-1, plays a major role in Ci uptake in O. alismoides : (i) external CA activity was measured, (ii) AZ inhibited Ci uptake with the slope of Ci uptake vs the concentration of CO2 between 15 and 40 µM being about a quarter of the control after treatment with 0.2 mM AZ, (iii) transcripts of putative αCA-1 were detected. The CA was confirmed to be external since (i) washing of leaves treated with AZ, restored CA activity and (ii) its sequence bears a signal peptide consistent with a periplasmic location. External CA is indeed widespread in photoautotrophs from marine and freshwater environments (Moroney et al., 2001; Dimario et al., 2018). The green microalga Chlamydomonas reinhardtii has three αCAs, of which two (Cah1 and Cah2) are localized in the periplasmic space and one (Cah3) in the thylakoid membrane (Fujiwara et al., 1990; Karlsson et al., 1998; Moroney & Chen, 1998). While CAs have the same catalytic activity, their sequence identity could be very low among different classes (Jensen et al., 2019). The αCA-1 from O. alismoides has around 30% sequence identity with the periplasmic Cah1 from C. reinhardtii . Many CAs are regulated by the concentration of CO2. The diatom Phaeodactylum tricornutum does not possess external CA, but the internal CA (β-type CA) is CO2 responsive and crucial for its CCM operation (Satoh et al., 2001; Harada et al., 2005; Harada & Matsuda, 2005; Tsuji et al., 2017). In the marine diatom, Thalassiosira pseudonana , the two external CAs, δ-CA and ζ-CA, as well as a recently identified chloroplastic ι-CA are induced by carbon limitation (Samukawa et al., 2014; Clement et al., 2017; Jensen et al., 2019). In contrast, the putative αCA-1 in O. alismoides is constitutive and its expression was unaffected by the CO2 concentration. This is also true for Cah3 in the thylakoid lumen of C. reinhardtii(Karlsson et al., 1998; Moroney & Chen, 1998), while the expression of the periplasmic CA (Cah1) and the mitochondrial CAs (β-CA1 and β-CA2) are highly CO2-sensitive (Moroney & Chen, 1998).
We show that the anion exchange proteins, one group of the SLC4 family HCO3- transporters (Romero et al., 2013), is involved in HCO3- uptake inO. alismoides . DIDS, a commonly-used inhibitor of AE/SLC-type HCO3- transporters (Romero et al., 2013) significantly decreased the final pH of a drift, and increased the final CO2 concentration to about 0.8 µM which is not substantially less than that expected in the absence of a CCM: a terrestrial C3 plant CO2 compensation point of 36 µL L-1 (Bauer & Martha, 1981) is equivalent to about 1.2 µM. Furthermore, transcripts of putative HCO3- transporter family in O. alismoides were found to contain Band 3 anion exchange proteins (SLC4 member 1), and the peptides shared 70.6-85.7% sequence identity with HCO3- transporters from other terrestrial plant species. Several genes which encode SLC4 family transporters, has been found to be involved in the CCMs in the marine microalgae Phaeodactylum tricornutum and Nannochloropsis oceanica(Nakajima et al., 2013; Poliner et al., 2015), as well as the marine macroalga Ectocarpus siliculosus (Gravot et al., 2010). More broad evidence from the physiological data have demonstrated that anion exchange proteins play a role in HCO3-uptake in green, red and brown marine macroalgae (Drechsler et al., 1993; Granbom & Pedersén, 1999; Larsson & Axelsson, 1999; Fernández et al., 2014). Although HCO3- use by seagrasses is known to involve an anion exchange protein, to our knowledge, this is the first report that provides evidence of the presence of a direct HCO3- uptake via DIDS-sensitive SLC4 HCO3- transporters in an aquatic angiosperm. Whatever, these transporters for direct HCO3- acquisition, appears to be much more restricted in distribution than the widespread external CA.
Three mechanisms of HCO3- use have been proposed in aquatic plants: i) indirect use of HCO3- based on dehydration of HCO3-, facilitated by external CA, to produce elevated CO2 concentrations outside the plasmalemma; ii) direct uptake of HCO3- by an anion exchange transporter in the plasmalemma and iii) direct uptake of HCO3- by a P-type H+-ATPase (Giordano et al., 2005). In this study we provide evidence for the first two mechanisms in O. alismoides . Although we did not specifically check for a P-type H+-ATPase, this process appears to be absent, or of minor importance, in O. alismoides in contrast to Laminaria digitata and L. saccharina (Klenell et al., 2004), because inO. alismoides, HCO3- use was abolished by addition of either AZ or DIDS. An AE is mainly responsible for HCO3- use in the brown marine macroalga Macrocystis pyrifera (Fernández et al., 2014), while in several other brown macroalgae such as Saccharina latissima(formerly Laminaria saccharina ) external CA plays the major role in HCO3- use (Axelsson et al., 2000), though in L. saccharina as in L. digitate , a P-type H+-ATPase has been identified (Klenell et al., 2004). In another brown macroalga, Endarachne binghamiae , HCO3- use was based on an external CA and P-type H+-ATPase with no contribution from an AE (Zhou & Gao, 2010). Another strategy to use HCO3- has been shown in some species of freshwater macrophytes that involves the possession of ‘polar leaves’ (Steemann-Nielsen, 1947). At the lower surface of these leaves, proton extrusion generates low pH and at their upper surface, high pH often generates calcite precipitation (Prins et al., 1980). Consequently, at the lower surface with low pH, the conversion of HCO3- to CO2 near the plasmalemma facilitates the cells to take up Ci. Because of this, there is some evidence for a lower reliance on external CA in macrophytes with polar leaves. For example, in a species with polar leaves,Potamogeton lucens , external CA was absent (Staal et al., 1989) and in the polar leaf species Elodea canadensis , external CA activity was present but not influenced by the CO2concentration (Elzenga & Prins, 1988).
It was initially surprising that AZ completely inhibited HCO3- use. However, Sterling et al. (2001) also found that AZ inhibited AE1-mediated chloride-bicarbonate exchange. This result could be explained by the binding of CA to the AE resulting in the formation of a transport metabolon, where there was a direct transfer of HCO3- from CA active site to the HCO3- transporter (Sowah & Casey, 2011; Thornell & Bevensee, 2015). Thus, when CA is inhibited, then the transport of HCO3-is inhibited.
O. alismoides can perform C4 photosynthesis, however the final CO2 concentration at the end of pH-drift, when HCO3--use was abolished by the inhibitors, was 0.8-1.6 µM, which could be supported by passive entry of CO2 without the need to invoke a CCM. These are slightly higher than the CO2 compensation point in the freshwater C4 macrophyte Hydrilla verticillata at less than 10 ppm (Bowes, 2010), which is equivalent to a dissolved CO2~0.3 µM at 25 °C. If this difference between the species is real and not methodological, it could suggest that in O. alismoides C4 photosynthesis is more important to suppress photorespiration than to uptake carbon.
A simple model of carbon acquisition (Figure 7a) was constructed to quantify the contribution of the three pathways involved in Ci uptake inO. alismoides : passive diffusion of CO2, HCO3--use involving αCA-1 and HCO3--use involving SLC4 HCO3- transporters. Using the Ci uptake rates at different CO2 concentrations in Figure 3, and assuming that 0.3 mM DIDS completely inhibited HCO3- transporters and that 0.2 mM AZ completely inhibited αCA-1 and HCO3-transporters, we calculated: i) passive diffusion of CO2as the rate in the 0.2 mM AZ treatment that inhibited both αCA-1 and SLC4 HCO3- transporters; ii) diffusion of HCO3- and conversion to CO2 by αCA-1 at the plasmalemma as the difference between the rate in the presence of 0.3 mM DIDS and that in the presence of 0.2 mM AZ; and iii) diffusion of HCO3- and transfer across the plasmalemma by SLC4 HCO3- transporters as the difference in the rate between the control and the 0.3 mM DIDS treatment. At a CO2 concentration of about 50 µM, passive diffusion of CO2 contributed 55.7% to total Ci uptake, diffusion of HCO3- and conversion to CO2 by αCA-1 contributed 42.7% and transfer of HCO3- across the plasmalemma by SLC4 HCO3- transporters contributed 1.6% (Figure 7b). At ~9 µM (about 66% of equilibrium with air at 400 ppm CO2) the contribution to total Ci uptake of CO2-diffusion, HCO3- diffusion and conversion to CO2 by αCA-1 and transfer by SLC4 HCO3- transporters was 24.0%, 64.4% and 11.5% respectively and at about 1 µM CO2 (close to a typical C3 CO2 compensation point) diffusion was zero and αCA-1 and SLC4 HCO3- transporters contributed equally to carbon uptake. So, as CO2concentrations fall, passive CO2 diffusion can no longer support Ci uptake and indirect and direct use of HCO3- allows Ci uptake to continue. The stimulation of absolute rates of SLC4 HCO3- transporters-dependent Ci uptake is consistent with patterns seen for a number of freshwater macrophytes during pH-drift experiments, where rates increase as CO2approaches zero before declining as Ci is strongly depleted (Maberly & Spence, 1983). This could be caused by regulation or by direct effects of pH on HCO3- transporters activity.
These results confirm the prevailing notion from seagrasses that external CA plays an important role in contributing to Ci uptake. External CA contributed 25% to Ci uptake in Posidonia australis(James & Larkum, 1996) and ~60% in Zostera marina (approximately 2.2 mM Ci at pH 8.2, equivalent to a dissolved CO2 ~23 µM at 25 °C; Beer & Rehnberg, 1997), albeit in the presence of Tricine buffer that might inhibit the photosynthesis rate. The value reported here for O. alismoides at a CO2 concentration of 23 µM, 56%, is similar toZ. marina .
In conclusion, O. alismoides has developed a jack of trades CCM, the master of which, either external CA or SLC4 HCO3- transporters, depends on the CO2 concentration. There are several future lines of work that need to be pursued. The distribution of HCO3-transporters in freshwater species should be determined. The apparent relationship between polar leaves and low or absent external CA activity could be tested using a range of species, especially within the genusOttelia where calcite precipitation differs among species (Cao et al., 2019). The Ci acquisition mechanisms of more freshwater species should be examined. The cause of the increasing rate of HCO3- transporters-dependent HCO3- uptake as Ci becomes depleted needs to be understood. Finally, production and analysis of genome sequences for freshwater macrophytes will be a powerful tool to answer these and future questions concerning the strategies used by freshwater macrophytes to optimize photosynthesis.