Characterization of cooperative bicarbonate uptake into chloroplast stroma in the green alga Chlamydomonas reinhardtii

Accumulation of HLA3 and LCIA in Very Low CO₂ Conditions.

First, to define the acclimated states of limiting CO₂ conditions (LC or VLC) of cells grown in liquid culture, total Ci concentration in the culture medium at pH 7.0 was measured, and consequent CO₂ concentrations were calculated (Fig. 1A). CO₂ concentrations supplied with 0.04% CO₂ for 1, 2, 4, 6, and 12 h were estimated as 6.3, 3.1, 2.9, 1.9, and 1.8 µM, respectively, which correspond to the range for VLC (<7 µM CO₂) (16). Thus, we defined the limiting CO₂ conditions of liquid culture as VLC throughout this study. Next, the time course accumulation of HLA3 and LCIA after VLC induction was examined (Fig. 1A). The accumulation of these proteins started within 1 h and reached their maximum levels within 4 h, as was the case for LCI1 used as a control of VLC induction. The molecular masses of HLA3 and LCIA were detected at sizes of ∼133 and 27 kDa, respectively (Fig. S1A and SI Results and Discussion).

Subcellular Localization of HLA3 and LCIA.

To analyze the subcellular localization of HLA3, an indirect immunofluorescence assay was performed (Fig. 1B). Fluorescence signals from an anti-HLA3 antibody were detected peripherally, suggesting the localization of HLA3 to the PM. Fluorescence signals from an anti-LCIA antibody were detected as a single cup-shaped structure (Fig. 1B), as in the previous study (16). To further clarify the localization of HLA3 and LCIA biochemically, protein samples from total cell, PM, and CE fractions were probed with antibodies against HLA3, LCI1, H⁺-ATPase, LCIA, and CCP1 (Fig. 1C and SI Results and Discussion). LCI1 and H⁺-ATPase were enriched in the PM fraction, consistent with the PM localization of these proteins (25, 29). Similarly, a notable enrichment of HLA3 was observed in the PM fraction. LCIA was highly enriched in the CE fraction, where CE protein CCP1 (30) was also enriched. From these results, we concluded that HLA3 and LCIA were localized to the PM and CE, respectively.

Isolation of an HLA3 Insertion Mutant and Photosynthetic Characteristics.

To evaluate the degree of contribution of HLA3 to the CCM, we isolated an HLA3 insertion mutant from our paromomycin resistance gene-tagged mutant library by PCR-based screening, as described previously (31), and designated the strain Hin-1 (Fig. S2 A–C and SI Results and Discussion).

Next, the photosynthetic characteristics were evaluated by measuring the rates of Ci-dependent O₂ evolution of WT, Hin-1, and the complemented strain Hin-1C grown in VLC at different pH. K_0.5 (Ci) values, the Ci concentration required for half maximal O₂-evolving activity, of WT and Hin-1 were similar at pH 6.2 (ratio of HCO₃^–:CO₂ = 0.7:1) and pH 7.8 (HCO₃^–:CO₂ = 28:1), indicating that the difference in photosynthetic Ci affinity between WT and Hin-1 was not significant (Fig. 2A). Because HLA3 KD strains showed retarded growth rates at pH 9.0 (28) where the ratio of HCO₃^–:CO₂ = 446:1 and HCO₃^– was the predominant form of Ci, we evaluated the changes in Ci affinity during acclimation to VLC at pH 9.0 in a time course analysis (Fig. 2B). Both WT and Hin-1 showed a gradual decrease in K_0.5 (Ci) during acclimation to VLC. However, although WT in VLC at 6 h showed almost the same Ci affinity compared with that at 12 h (241 ± 87 µM at 6 h and 290 ± 50 µM at 12 h), Hin-1 still showed much lower Ci affinity especially at 6 h (691 ± 143 µM at 6 h and 405 ± 57 µM at 12 h), and the decreased Ci affinity was restored in Hin-1C (296 ± 78 µM at 6 h and 333 ± 89 µM at 12 h). These results suggested that other Ci uptake systems could compensate for the absence of HLA3 and contribute to the increase in Ci affinity at 12 h and that measuring photosynthetic characteristics at 6 h was appropriate for evaluating the contribution of HLA3 to the CCM.

To evaluate the contribution of HLA3 to actual Ci uptake activity, the accumulation and fixation of [¹⁴C]-labeled Ci in WT, Hin-1, and Hin-1C grown in VLC for 6 h were measured (Fig. 2C). Hin-1 showed significantly lower levels of Ci accumulation of 0.12 mM (0.57-fold of Hin-1C), 0.12 mM (0.32-fold), and 0.06 mM (0.17-fold) after 80, 160, and 240 s of illumination, respectively, and CO₂ fixation of 0.37 nmol⋅μL SIS^–1 (0.59-fold), 0.67 nmol⋅μL SIS^–1 (0.49-fold), and 0.83 nmol⋅μL SIS^–1 (0.44-fold), respectively, compared with that of Hin-1C. These results indicated that HLA3 has a meaningful role in HCO₃^– uptake in VLC conditions.

Isolation of LCIA Insertion Mutants and Photosynthetic Characteristics.

A disruption mutant of LCIA has been characterized, and the contribution of LCIA to the CCM has been reported (16). To compare the degrees of the contributions of HLA3 and LCIA to the CCM, we also isolated two LCIA insertion mutants (Fig. S2 D–G and SI Results and Discussion), designated as Ain (Ain-1 and Ain-2), and compared the photosynthetic characteristics with Hin-1. Interestingly, accumulation of HLA3 was much lower in Ain compared with that in WT, and this decreased accumulation of HLA3 was restored in the complemented strains Ain-1C and Ain-2C (Fig. 3A). This result was in sharp contrast to that of LCI1 and LCIB (32), which were not affected by the impairment of the LCIA (Fig. 3A).

Next, the photosynthetic characteristics of Ain-1, Ain-2, Ain-1C, and Ain-2C were evaluated. As in the case of Hin-1, the K_0.5 (Ci) of Ain-1 and Ain-2 was similar to WT at pH 6.2 (Fig. 3B). However, in contrast to Hin-1, the K_0.5 (Ci) of Ain-1 (57 ± 2 µM at 6 h and 56 ± 3 µM at 12 h) and Ain-2 (57 ± 1 µM at 6 h and 57 ± 2 µM at 12 h) was significantly higher than that of WT (40 ± 3 µM at 6 h and 33 ± 3 µM at 12 h), Ain-1C (38 ± 2 µM at 6 h and 32 ± 2 µM at 12 h), and Ain-2C (37 ± 3 µM at 6 h and 29 ± 2 µM at 12 h), even at pH 7.8 (Fig. 3B). At pH 9.0, although Ain also showed gradual decreases in K_0.5 (Ci) during acclimation to VLC, these cells always showed lower Ci affinity than Hin-1 (Fig. 2B), as well as WT and complemented strains (Fig. 3C). These results suggested a significant contribution of LCIA to increases in Ci affinity and to maintaining HLA3 stability in the CCM.

Isolation of LCIA/HLA3 Double-Insertion Mutants and Photosynthetic Characteristics.

Because Ci affinity in VLC at 12 h was higher than that at 6 h in both HLA3 and LCIA single mutants, either protein could partially complement each other to increase Ci affinity. Thus, we expected that LCIA/HLA3 double-insertion mutants would show an additive decrease in Ci affinity compared with the single-insertion mutants. Thus, we isolated double-insertion mutants by crossing one of the Ain-2 progeny with Hin-1 and designated these as AHin (AHin-1 and AHin-2; Fig. 3D, Fig. S2 H–K, and SI Results and Discussion).

Next, the photosynthetic characteristics of AHin-2 were evaluated (Fig. 3E). As in the case of Hin-1 and Ain, the K_0.5 (Ci) of AHin-2 was similar to WT at pH 6.2. At pH 7.8, the K_0.5 (Ci) of AHin-2 (58 ± 2 µM at 6 h and 61 ± 10 µM at 12 h) was significantly higher than that of WT, but it was similar to Ain. At pH 9.0, AHin-2 showed lower Ci affinity than both Hin-1 and Ain, and Ci affinity was not increased even at 12 h (898 ± 78 µM at 6 h and 901 ± 94 µM at 12 h). Ci accumulation and fixation in AHin-2 grown in VLC at 6 h was also measured (Fig. 3F). After 80, 160, and 240 s of illumination, AHin-2 showed substantially decreased Ci accumulation of 0.05 mM (0.21-fold of WT and 0.41-fold of Hin-1), 0.06 mM (0.16-fold and 0.53-fold), and 0.06 mM (0.15-fold and 1.0-fold), respectively, and CO₂ fixation of 0.07 nmol⋅μL SIS^–1 (0.1-fold and 0.2-fold), 0.12 nmol⋅μL SIS^–1 (0.08-fold and 0.17-fold), and 0.14 nmol⋅μL SIS^–1 (0.07-fold and 0.17-fold), respectively, compared with that of WT and Hin-1.

Finally, the effect of absence of LCIA and/or HLA3 on cell growth was examined. Growth rates were measured in VLC at pH 8.4 (Fig. S2L) because there were no significant differences at pH 7.8, and none of the cell lines could grow at pH 9.0. The doubling time of WT was 7.2 h and that of Hin-1, Ain-1, Ain-2, and AHin-2 increased significantly to 7.6, 9.5, 9.3, and 12.7 h, respectively, reflecting the degree of decreased Ci affinity of each cell line. These results highlighted an additive decrease in Ci affinity/accumulation/growth rates of the double-insertion mutant compared with the HLA3 or LCIA single-insertion mutants.

Isolation of LCIA and/or HLA3 Overexpressing Strains and Photosynthetic Characteristics.

To demonstrate the physiological function of LCIA and HLA3 more directly, the photosynthetic characteristics of cells overexpressing LCIA and/or HLA3 were examined in HC conditions where other VLC-inducible proteins were not induced. For overexpression, two chimeric plasmids, pTY2b-LCIA and pTY2b-HLA3, were constructed (Fig. S3A). These plasmids allowed the induction of LCIA and HLA3 transcripts by switching the nitrogen source from NH₄⁺ to NO₃^– irrespective of the CO₂ conditions. In this study, we cultured the cells with four combinations of nitrogen sources in the medium and CO₂ concentrations, designated as HC-NH₄⁺, HC-NO₃^–, VLC-NH₄⁺, and VLC-NO₃^–.

First, we transformed WT cells with pTY2b-LCIA or pTY2b-HLA3 separately. The transformants showed accumulation of LCIA or HLA3 when grown in HC-NO₃^– conditions and were designated as Aox (Aox-1 and Aox-2) and Hox (Hox-1 and Hox-2), respectively (Fig. S3 B and C and SI Results and Discussion). Next, by introducing pTY2b-HLA3 into Aox-1, we generated two independent transformants expressing LCIA and HLA3 simultaneously and designated these as AHox (AHox-1 and AHox-2; Fig. S3D). Accumulation of HLA3 in AHox-1 and AHox-2 was the same as that of VLC-grown WT. To isolate a strain overexpressing both LCIA and HLA3 with greater abundance, the progeny of Aox-1 was crossed with Hox-1 and a strain designated as AHox-3 was obtained (Fig. S3E).

Next, to evaluate the effect of LCIA and/or HLA3 overexpression on the photosynthetic characteristics, rates of O₂ evoluton at pH 6.2, 7.8, and 9.0 and Ci accumulation at pH 9.0 of these strains were measured. In Aox, there were no differences in Ci affinity at pH 7.8 and pH 9.0, as well as Ci accumulation compared with WT (Fig. 4A and Table S1–S3). In contrast, HC-NO₃^–-grown Hox showed a small but significant increase of Ci accumulation of 0.08 mM (2.5-fold of WT at 80 s), 0.07 mM (1.5-fold at 160 s), and 0.13 mM (2.4-fold at 240 s) in Hox-2, compared with that of HC-NO₃^–-grown WT, but the phenotype led to a slight increase in Ci affinity only at pH 9.0 in Hox-2 (Fig. 4B), suggesting that Ci in the cytosol transported by HLA3 could not efficiently enter the chloroplast stroma in the absence of LCIA. On the other hand, Ci affinity at pH 6.2 was increased in LCIA-overexpressing Aox (Fig. 4A) and AHox (Fig. 4C), but not in Hox (Fig. 4B).

In contrast to Aox and Hox, AHox showed a significant increase in Ci affinity and Ci accumulation compared with WT at alkaline conditions (Fig. 4C and Tables S2 and S3). In particular, HC-NO₃^–-grown AHox-3 showed substantially increased Ci accumulation of 0.21 mM (6.3-fold of WT at 80 s), 0.34 mM (6.8-fold at 160 s), and 0.19 mM (3.6-fold at 240 s) compared with that of HC-NO₃^–-grown WT. Consequently, the respective K_0.5 (Ci) of AHox-1, AHox-2, and AHox-3 decreased to 141 ± 20 (0.61-fold of WT), 174 ± 20 (0.76-fold), and 147 ± 19 µM (0.64-fold) at pH 7.8 and to 1,821 ± 201 (0.68-fold of WT), 1,980 ± 198 (0.75-fold), and 1,626 ± 49 µM (0.61-fold) at pH 9.0. In HC-NH₄⁺ conditions at pH 7.8 where LCIA and HLA3 were not induced, the respective K_0.5 (Ci) of 257 ± 28, 250 ± 30, and 262 ± 29 µM in AHox-1, AHox-2, and AHox-3 was not significantly different from that of 273 ± 31 µM in WT (Table S2). These results indicated that NO₃^–-induced overexpression of LCIA and HLA3 could enhance HCO₃^– accumulation in the chloroplast stroma and increase Ci affinity.

Although PM-localized LCI1 could be associated with Ci uptake (25), the preferred Ci species of LCI1 remained elusive. To evaluate the degree of LCIA/HLA3-driven HCO₃^– uptake activity, we also isolated six transformants expressing LCIA with LCI1 by introducing pTY2b-LCI1 (Fig. S3A) into Aox-1 and designated two representatives as A1ox (A1ox-1 and A1ox-2; Fig. S3F). There were no differences in Ci accumulation and affinity in alkaline conditions compared with WT (Fig. 4D and Table S2 and S3), suggesting that LCI1 was not related to direct HCO₃^– uptake along with LCIA.

A Defect in LCIA Led to a Decrease in HLA3 Accumulation Caused by Suppression of HLA3 mRNA Accumulation.

As described above, accumulation of HLA3 was much lower in Ain compared with that in WT (Fig. 3A). This result suggested two possibilities. First, HLA3 and LCIA undergo a physical interaction where the PM is associated with the CE and the absence of LCIA causes instability of HLA3. Second, the absence of LCIA causes the repression of HLA3 mRNA accumulation.

To examine the former possibility, the molecular masses of LCIA and HLA3 in vivo were estimated by Blue Native-PAGE. We expected that LCIA and HLA3 should be detected with the same molecular mass in nondenaturing conditions if these two proteins interact and form a complex. However, using 1.0% n-dodecyl β-d-maltoside (DDM) as a detergent, LCIA and HLA3 were detected with different sizes of ∼240 and 580 kDa, respectively (Fig. 5A). We also estimated the molecular masses using different DDM concentrations (0.25%, 0.5%, 1.0%, or 2.0%) or using formaldehyde cross-linker, and LCIA and HLA3 were still detected at 240 and 580 kDa, respectively (Fig. S4 A and B). Furthermore, LCIA and HLA3 could form respective complexes with the same molecular masses even in Aox, Hox, and AHox cells grown in HC-NO₃^– conditions (Fig. S4C). These results strongly suggested that LCIA and HLA3 did not interact physically in vivo and at least VLC-inducible proteins other than LCIA and HLA3 were not associated with the formation of the respective protein complexes.

For the latter possibility, HLA3 mRNA accumulation was evaluated by quantitative real-time PCR (Fig. 5B). The sequences of primers used are listed in Table S4. HLA3 mRNA levels were significantly reduced in Ain-1 and Ain-2 grown in VLC, but mRNA accumulation was restored in the complemented strains. In contrast, the mRNA levels of LCIA were not affected in Hin-1 (Fig. 5B), and those of other VLC-inducible genes LCIB and LCI1 were also largely unchanged in Ain-1 and Ain-2, as well as Hin-1 (Fig. S4D). These results suggested that LCIA localized to the CE could affect the mRNA expression level of HLA3 and subsequently caused a decrease in HLA3 protein accumulation.