Expression of a low CO₂-inducible protein, LCI1, increases inorganic carbon uptake in the green alga Chlamydomonas reinhardtii.

Aquatic photosynthetic organisms can modulate their photosynthesis to acclimate to CO₂-limiting stress by inducing a carbon-concentrating mechanism (CCM) that includes carbonic anhydrases and inorganic carbon (Ci) transporters. However, to date, Ci-specific transporters have not been well characterized in eukaryotic algae. Previously, a Chlamydomonas reinhardtii mutant (lcr1) was identified that was missing a Myb transcription factor. This mutant had reduced light-dependent CO₂ gas exchange (LCE) activity when grown under CO₂-limiting conditions and did not induce the CAH1 gene encoding a periplasmic carbonic anhydrase, as well as two as yet uncharacterized genes, LCI1 and LCI6. In this study, LCI1 was placed under the control of the nitrate reductase promoter, allowing for the induction of LCI1 expression by nitrate in the absence of other CCM components. When the expression of LCI1 was induced in the lcr1 mutant under CO₂-enriched conditions, the cells showed an increase in LCE activity, internal Ci accumulation, and photosynthetic affinity for Ci. From experiments using indirect immunofluorescence, LCI1-green fluorescent protein fusions, and cell fractionation procedures, it appears that LCI1 is mainly localized to the plasma membrane. These results provide strong evidence that LCI1 may contribute to the CCM as a component of the Ci transport machinery in the plasma membrane.


INTRODUCTION
Aquatic photosynthetic organisms acclimate to different environmental changes, such as light intensity, light quality, temperature, and nutrient availability. The supply of inorganic carbon (Ci; CO 2 and HCO 3 2 ) is one of the environmental factors that can limit photosynthetic performance because the environmental Ci concentration is usually lower than that required for the optimal carboxylase activity of ribulose-1,5-bisphosphate carboxylase/ oxygenase. To overcome the difficulty in acquiring Ci at low CO 2 concentrations, cyanobacteria and green algae, including Chlamydomonas reinhardtii, induce a transport system called the carbon-concentrating mechanism (CCM) to increase Ci uptake (Badger et al., 1980;Giordano et al., 2005;Spalding, 2008;Yamano and Fukuzawa, 2009). In cyanobacteria, five types of cytoplasmic Ci transporters have been identified (Price et al., 2008), including three bicarbonate transporters and two NAD(P) H dehydrogenase-dependent CO 2 uptake systems. In the model eukaryotic algae C. reinhardtii, only one possible HCO 3 2 transporter, HLA3, has been characterized in some detail (Duanmu et al., 2009b), although several candidate genes responsible for Ci uptake have been identified in transcriptomic analyses Moroney and Ynalvez, 2007;Spalding, 2008;Yamano et al., 2008;Yamano and Fukuzawa, 2009).
The LC-inducible protein LCIA, also called NAR1.2, shows significant sequence similarity to a chloroplast nitrite transporter, NAR1.1, and is also induced under LC conditions and may function as a bicarbonate transporter whose subcellular location is not identified . Although it was reported that expression of LCIA in oocytes from Xenopus laevis facilitated the uptake of bicarbonate as well as nitrite (Mariscal et al., 2006), the function of LCIA in vivo remains to be elucidated. Recently, it was shown that the pmp1-1 mutation was complemented by the LCinducible gene LCIB (Wang and Spalding, 2006). Although the protein deficiency in pmp1-1 was expected to be in a Ci transporter, LCIB does not have any transmembrane domains and colocalized with a homologous protein, LCIC, in the vicinity of the pyrenoid (Yamano et al., 2010). Moreover, LCIB is genetically shown to function downstream of CAH3 in the CCM. Thus, this protein is assumed to act after the accumulated stromal HCO 3 2 is dehydrated to CO 2 rather than as a Ci transporter (Duanmu et al., 2009a). In addition, HLA3, which is an LC-inducible ABCtype transporter, was shown to play a role in Ci uptake in C. reinhardtii (Duanmu et al., 2009b). Although other CO 2 -responsive membrane proteins, including LIP36 (Ramazanov et al., 1993), CemA (previously designated as Ycf10; Rolland et al., 1997), and Rh1 (Soupene et al., 2004), might also be involved in Ci transport, further biochemical evidence is needed to elucidate their functions in the CCM.
In a previous study, we isolated the regulatory mutant lcr1, which lacks a Myb-DNA binding transcription factor, LCR1. The loss of LCR1 greatly reduces the expression of three LC-inducible genes, namely, CAH1, LCI1, and LCI6 . Whereas CAH1 is known to encode a periplasmic carbonic anhydrase, LCI1 and LCI6 have not been characterized biochemically. In addition, the lcr1 mutant showed a 30% decrease in growth rate and a higher apparent K 1/2 (Ci) value compared with that of photosynthetically wild-type strain Q304P3 under LC conditions. Because it has been shown that a loss of CAH1 does not result in reduced growth under LC conditions ( Van and Spalding, 1999) and that LCI6 encodes a soluble protein , the putative membrane protein LCI1 was assumed to play a significant role in the CCM (Burow et al., 1996;Yoshioka et al., 2004).
In this study, we provide evidence supporting the hypothesis that LCI1 increases Ci uptake and intracellular Ci accumulation under LC conditions. To reach this conclusion, we measured the light-dependent CO 2 gas exchange (LCE) activity, the rate of Ci uptake, and the Ci affinity in transgenic C. reinhardtii cells that artificially induce LCI1 expression under the control of the nitrogen species-specific responsive DNA elements of the nitrate reductase gene. We also provide evidence that LCI1 is localized to the plasma membrane. Together, these observations support the idea that LCI1 is, in part, responsible for Ci uptake in C. reinhardtii.

LCE Activity of Low and High CO 2 -Grown Cells
To evaluate the effects of mutations in LCR1 and CAH1 on photosynthesis, we compared the LCE activities of cells of the photosynthetically wild-type transgenic strain Q304P3  and mutants lcr1  and cah1 using an open gas exchange system with an air supply containing 50 ppm CO 2 (Yamano et al., 2008) (Figure 1, Table 1). Previously, it was shown that LCE activity mirrors total Ci consumption, which includes Ci uptake and CO 2 fixation (Spalding and Ogren, 1985) and provides a reasonably good indirect estimate of the CCM in cells (Yamano et al., 2008). The photosynthetically wild-type strain Q304P3 showed a LCE activity of 72.6 mmol CO 2 mg Chl 21 h 21 under LC conditions and 13.1 mmol CO 2 mg Chl 21 h 21 in high-CO 2 (HC) conditions. The cah1 mutant lacking the LC-inducible periplasmic carbonic anhydrase (Van and Spalding, 1999) showed a similar LCE activity under LC conditions ( Figure 1, Table 1). On the other hand, the lcr1 mutant showed a decreased level of LCE activity, but the activity of the LCR1-complemented strain (lcr1+G)  was equivalent to that of Q304P3 ( Figure 1, Table 1). Although the lcr1 mutant cannot induce 13 genes in LC conditions, the induction of only three genes, namely, LCI1, LCI6, and CAH1, were recovered in the LCR1-complemented strain . Together, these results suggest that LCR1-regulated genes other than CAH1 are necessary for high LCE activity under CO 2limiting conditions.

Induction of LCI1 by Changing Nitrogen Source
To determine whether the lcr1 mutant can induce the synthesis of the LCI1 protein, we performed protein immunoblot analysis using an affinity-purified antibody against synthetic polypeptides corresponding to parts of LCI1 ( Figure 2). A 17-kD protein band was detected by the antibody against LCI1 more abundantly in LC-grown Q304P3 cells compared with HC-grown Q304P3 cells. Although LCI1 encodes a putative 21-kD protein with four hydrophobic domains, the first domain in the N-terminal region was predicted to be a signal peptide by the prediction program SignalP (http://www.cbs.dtu.dk/services/SignalP/) (Bendtsen et al., 2004). Thus, the detected 17-kD band was assumed to correspond to the mature LCI1 protein without the first hydrophobic domain, although it is not clear which amino acid residue is the N terminus of the mature LCI1 protein in vivo. In the HCand LC-grown lcr1 mutant, a faint signal was observed near 17 kD, but its level was equivalent to that in HC-grown Q304P3 and was not affected by changes in the level of CO 2 supplied ( Figure  2). Because the faint bands at 17 kD were detected by both oligopeptide affinity-purified antibody and antiserum (Figures 2 and 3C, respectively), it is possible that a trace amount of LCI1 Four strains, Q304P3, lcr1, cah1, and lcr1, transformed with genomic DNA containing the LCR1 gene (lcr1+G), grown under HC conditions were subsequently aerated with ordinary air containing 0.04% CO 2 to establish LC growth conditions. The maximum difference in the CO 2 concentration from the basal level within 5 min, shown as "v," was defined as the LCE activity. The LCE activities of cells corresponding to 5 mg mL À1 of chlorophyll are indicated below the profiles and are summarized in Table 1. Arrowheads indicate when actinic light was turned on (open arrowheads) and off (closed arrowheads), respectively. accumulates in HC-grown Q304P3 and even in both HC-and LCgrown lcr1 mutant cells, although these results do not rule out a possibility that the antibody used in this study may recognize a 17-kD protein other than LCI1.
To analyze whether LCI1 is involved in LCE activity, we generated a chimeric gene in which tandemly duplicated NIA1 regulatory elements (Loppes and Radoux, 2002) and the minimal promoter of TUB2 encoding b 2 -tublin (Davies and Grossman, 1994) were fused to the LCI1 coding region followed by the 39untranslated region (UTR) of RBCS2 ( Figure 3A). This construct allowed us to induce the expression of LCI1 by switching the nitrogen source from NH 4 + to NO 3 2 irrespective of the CO 2 conditions. Although Q304P3 is photosynthetically a wild-type strain, this strain cannot use NO 3 2 as a nitrogen source because of a deficiency in the NIA1 gene encoding nitrate reductase. By contrast, the lcr1 mutant can use NO 3 2 as a nitrogen source because this mutant was generated by introducing the NIA1 gene into Q304P3 as a tag , and it does not induce endogenous LCI1 protein in the presence of NO 3 2 ( Figure  2). Thus, this lcr1 mutant is appropriate to use as the host strain for this gain-of-function analysis of LCI1. After transformation of the chimeric plasmid into the lcr1 mutant, Zeocin-resistant colonies were selected and used for RNA gel blot analyses with a probe specific to the LCI1 coding region ( Figure 3B). The level of endogenous LCI1 mRNAs of 1.3 kb (Burow et al., 1996;Yoshioka et al., 2004) increased in LC-grown Q304P3. In the lcr1 mutant, there was no significant accumulation of the LCI1 mRNA either under LC conditions or in the presence of NO 3 2 , although this mutant previously was shown to accumulate a trace amount of LCI1 mRNA under LC conditions . The LCI1 mRNA in the LC-grown lcr1 mutant could not be detected at this exposure level because the exposure period of the autoradiograph was too short to detect such a low amount of mRNA. On the other hand, in the LCI1 transformants designated as C2 and E4, the level of the 0.8-kb mRNA, which corresponds to the predicted LCI1-encoding mRNA from the transcription initiation site of TUB2 to the predicted poly(A) signal (Davies and Grossman, 1994) in the 39-UTR of RBCS2, increased in the presence of NO 3 2 irrespective of changes in the level of CO 2 supplied to the cells ( Figure 3B).
We also analyzed the accumulation of LCI1 protein in the LCI1 transformants by protein immunoblot analysis ( Figure 3C). Both LCI1 transformants grown in high salt minimal medium containing NO 3 2 , referred to as HSM(NO3), accumulated LCI1 protein under both HC and LC conditions. Although the levels of LCI1 protein in HC-and LC-grown LCI1 transformant cells cultured in HSM(NO3) media were ;25 and 50% of that in LC-grown Q304P3, respectively, the level of LCI1 protein was 10-to 20-fold higher than in those cells grown with NH 4 + as the nitrogen source. Because the combination of LC conditions and NO 3 2 did not affect the level of LCI1 protein in the lcr1 mutant (Figure 2), the accumulation of LCI1 protein seen in the transformants should be ascribed to the exogenously introduced LCI1 gene fused to NIA1 regulatory elements.

LCE Activity of the LCI1 Transformants
To evaluate the apparent photosynthetic capacity of the lcr1 mutant and the LCI1 transformants, we measured the LCE activity of the cells grown in high salt minimal medium containing either NH 4 + or NO 3 2 , referred to as HSM(NH4) or HSM(NO3), respectively ( Figure 4, Table 2). Under HC conditions, there were no meaningful differences among the LCE activities of the lcr1 mutant, C2, and E4 grown in HSM(NH4) ( Table 2). When the cells were grown in HSM(NO3), it was notable that the LCE activity of the LCI1 transformants C2 and E4 increased 2.2-and 3.6-fold, respectively, in the presence of NO 3 2 , whereas the lcr1 mutant showed only a 1.3-fold increase in activity.
LCE activity was also measured for cells grown under LC conditions in the presence of NH 4 + or NO 3 2 . Comparisons of LCgrown lcr1 mutant cultured in HSM(NH4) and HSM(NO3) indicated that a change of nitrogen source had no effect on LCE activity in the lcr1 mutant (Table 2). On the other hand, transformants C2 and E4 cultured in HSM(NH4) had almost the same LCE activities as the lcr1 mutant but showed a slight increase in LCE activity ( Figure 4, Table 2) when they were grown in HSM (NO3). Whereas artificial induction of LCI1 alone does not completely restore LCE activity to wild-type levels, these results reveal that the expression of LCI1 increases the LCE activity in both HC and LC conditions.
To analyze whether the expression of LCI1 preferentially increased CO 2 or HCO 3 2 uptake, we measured the LCE activity Accumulation of LCI1 protein in the photosynthetically wild-type strain Q304P3 and the lcr1 mutant strain was analyzed by protein immunoblot analysis. Cells were grown under HC or LC conditions in HSM(NH4) or HSM(NO3). Aliquots of total cell protein corresponding to 2 mg of chlorophyll were solubilized and separated by 15% PAGE containing 6 M urea. After transferring onto polyvinylidene difluoride membranes, the blots were probed with an affinity-purified antibody against LCI1 oligopeptides. High CO 2 13.1 6 1.7 11.6 6 0.9 15.2 6 1.2 n.d. Low CO 2 72.6 6 2.2 36.9 6 6.7 70.5 6 1.7 69.7 6 5.1 The data are shown as mmol CO 2 mg Chl À1 h À1 , 6SD, obtained from three independent experiments. lcr1+G, the LCR1-complemented strain; n.d., not determined.
LCI1 Enhances Ci Uptake in C. reinhardtii 3 of 13 of the lcr1 mutant and C2 cells grown in HC conditions at two other external pH values: pH 6.2 and 7.8 (Table 3). At pH 6.2, HSM(NH4)-grown lcr1 mutant cells showed comparable LCE activity to that of HSM(NO3)-grown cells, indicating that changing the nitrogen species in the medium did not affect LCE activity in the parent strain. By contrast, HSM(NO3)-grown C2 cells showed a 2.1-fold increase in the LCE activity compared with that in HSM(NH4)-grown cells. At pH 7.8, a similar increase in the LCE activity of the C2 cells was observed when the culture medium was changed from HSM(NH4) to HSM(NO3). These results mean that the increase in the LCE activity in HSM(NO3)grown C2 cells was observed in both acidic and alkaline pH conditions. Interestingly, although the LCE activities at pH 7.8 in HSM(NH4)-grown cells decreased by ;50% compared with those at pH 6.2 in both strains, the increase in the LCE activities of HSM(NO3)-grown C2 cells compared with that of HSM(NO3)grown lcr1 mutant cells was almost constant (shown as [C2 Table 3). Unfortunately, because the cells were easily broken by bubbling during the measurements under more alkaline conditions (pH 8.6 and 9.0), we could not evaluate the effect of the expression of LCI1 on LCE activity under more alkaline conditions.

Dissolved Ci-Dependent Photosynthesis of the LCI1 Transformants
We also analyzed the photosynthetic affinity for Ci in the LCI1 transformants by measuring the rate of Ci-dependent O 2evolving activity at two different pH values, namely, pH 7.0 and 7.8. At pH 7.0, the K 1/2 value of the HC-grown lcr1 mutant was comparable to that in the HC-grown Q304P3 cells, the photosynthetically wild-type strain, whereas that of LC-grown lcr1 mutant was higher than that of LC-grown Q304P3 ( Figure 5A; see The LCE activities of cells grown under HC or LC conditions in HSM(NH4) or HSM(NO3) were measured at pH 7.0. Arrowheads indicate when actinic light was turned on (open arrowheads) and off (closed arrowheads), respectively. The dotted lines represent the light-dependent decrease in the CO 2 levels of the lcr1 mutant, which was used as the host strain for the LCI1 transformants. The LCE activities are summarized in Table 2. (A) A schematic illustration of the chimeric construct for artificial expression of LCI1. The coding region of LCI1 was placed at the downstream of NIA1 short promoter, which was followed by 39-UTR of the RBCS2 gene encoding ribulose-1,5-bisphosphate carboxylase. NIA1 short promoter consists of tandemly duplicated NIA1 regulatory elements and the b 2tublin (TUB2) promoter. For analysis of the subcellular localization of LCI1, the CrGFP coding region was inserted in frame just upstream of the translation termination codon of the LCI1 gene (shown as a closed triangle) to express an LCI1-CrGFP chimeric protein (see Figure 8). The gene filled in black is the Ble r gene (Lumbreras et al., 1998), which allows for growth in the presence of zeocin.
(B) Accumulation of LCI1 mRNA in cells of Q304P3, the lcr1 mutant, and transformants C2 and E4 carrying LCI1 with the NIA1 short promoter, grown in either HSM(NH4) or HSM(NO3), was examined by RNA gel blot analyses using a PCR fragment of the LCI1 cDNA as a probe. Five micrograms of total RNA were separated on a 1% denaturing agarose gel, stained with ethidium bromide to show rRNAs (25S and 18S), and blotted onto nylon membranes.
(C) Expression of LCI1 protein in cells grown in either HSM(NH4) or HSM (NO3) was determined by protein immunoblot analysis using an anti-LCI1 antiserum, AS-F. SDS-PAGE and immunoblot analysis were performed as described in Figure 2 and in Methods.

of 13 The Plant Cell
Supplemental Table 1 online). The LC-grown lcr1 mutant showed a lower Ci affinity compared with that of Q304P3, as described previously . In addition, the K 1/2 values of both HC-and LC-grown lcr1 mutant cells were not affected by the nitrogen source in the culture media ( Figure 5A; see Supplemental Table 1 online). By contrast, the affinities of C2 and E4 cells for Ci increased when the cells were grown in HSM(NO3). The K 1/2 values of C2 and E4 cells grown in HSM(NO3) under HC conditions were decreased compared with those of the cells grown in HSM(NH4), indicating that the Ci affinity of the LCI1 transformants increased when exogenous LCI1 was expressed. However, the K 1/2 values of all the strains grown in LC conditions, except for Q304P3, were ;40 mM. This value was not affected by the nitrogen source, although the increase in LCE activity was observed in both LCI1 transformants (Figure 4). At pH 7.8, the photosynthetic affinities for Ci decreased in all the examined strains compared with those at pH 7.0. The K 1/2 value of the HC-grown lcr1 mutant cells was not affected by the nitrogen source ( Figure 5B; see Supplemental Table 2 online). By contrast, the K 1/2 values of C2 and E4 cells grown in HSM(NO3) under HC conditions were apparently decreased when compared with those of the cells grown in HSM(NH4), indicating that the Ci affinity of the LCI1 transformants increased in the presence of NO 3 2 . These results were consistent with the increases observed in LCE activities ( Figure 4). However, in LC conditions, the K 1/2 values of all the strains except for Q304P3 were ;100 mM and were not affected by the nitrogen source ( Figure 5B; see Supplemental Table 2 online).

Inorganic Carbon Uptake of the LCI1 Transformants
To evaluate the contribution of LCI1 to Ci transport activity, we compared the accumulation and fixation of [ 14 C]-labeled Ci by the transformants with those of the lcr1 mutant using a silicone oil layer centrifugation method ( Figure 6). We performed the analysis at pH 7.8 using HC-grown cells because the effects of the artificial expression of LCI1 on Ci affinity were more marked under those conditions ( Figure 5; see Supplemental Tables 1 and 2 online). The lcr1 mutant showed almost the same levels of Ci accumulation (;0.1 mM) and CO 2 fixation (;1.0 nmol mL SIS 21 ) after 80 s with either nitrogen source in the media. By contrast, the levels of Ci accumulation of HSM(NO3)-grown LCI1 transformants C2 and E4 were more than twice the levels observed in the transformants cultured in HSM(NH4). In addition, both HSM(NO3)-grown LCI1 transformants showed an ;1.3-fold increase in CO 2 fixation, although the lcr1 mutant cells did not show any apparent effects of nitrogen source on Ci accumulation and fixation. These results reveal that nitrateinduced expression of the LCI1 protein enhanced the Ci uptake and/or CO 2 fixation of the transformants grown under HC conditions.

Growth of the LCI1 Transformants
The effects of the expression of LCI1 on the growth of cells were also analyzed at both pH 7.0 and 7.8. At pH 7.0, Q304P3 cells showed a similar growth rate in both HC and LC conditions, where the doubling time was 6.5 h (Figure 7). By contrast, at pH 7.0, the growth of the lcr1 mutant was slower in LC conditions than in HC conditions, which is consistent with the deficiency in the CCM of this mutant, as described previously . The growth of the lcr1 mutant at pH 7.0 was not affected by the nitrogen source in the media in either HC or LC conditions. At pH 7.0, the growth of the LCI1 transformant C2 was similar to that of the lcr1 mutant and was not affected by the nitrogen source, although increases in LCE activity ( Figure 4) and Ci affinity ( Figure 5) were observed at pH 7.0. The data are shown as mmol CO 2 mg Chl À1 h À1 , 6SD, obtained from three independent experiments. The data are shown as mmol CO 2 mg Chl À1 h À1 , 6SD, obtained from three independent experiments. a HCO 3 À /CO 2 ratio was calculated using the equation {pH = pK a +log 10 [HCO 3 À ]/[CO 2 ]}, where the pK a is an acid dissociation constant of 6.35. LCI1 Enhances Ci Uptake in C. reinhardtii 5 of 13 At pH 7.8, Q304P3 cells grew more slowly, even in HC conditions, with a doubling time of 9 h, which was longer than that of the growth at pH 7.0 ( Figure 7). A similar increase in doubling time was observed in both the lcr1 mutant and the LCI1 transformant C2. The lcr1 mutant and C2 cells were grown in both HSM(NH4) and HSM(NO3) in HC conditions. Although the apparent increases in the Ci affinity and the Ci uptake in HSM (NO3)-grown C2 cells were observed at pH 7.8 ( Figures 5 and 6), the growth rate for those cells was almost the same as that of cells cultured in HSM(NH4) (Figure 7).

Subcellular Localization of LCI1 Protein
To analyze the localization of the LCI1 protein, three methods were applied: visualization using indirect immunofluorescence, a C. reinhardtiis-adapted green fluorescent protein (CrGFP) fused to LCI1 (Fuhrmann et al., 1999), and a cell fractionation method, in which LCI1 was detected using a specific antibody. Indirect immunofluorescence analyses with anti-LCI1 were applied to photosynthetically grown Q304P3 cells. As expected, signals from fluorescently labeled anti-LCI1 antibodies were detected in LC-grown cells but not in HC-grown cells (see Supplemental Figure 1 online). In LC-grown cells, fluorescence was detected around the nucleus as well as at the cell periphery. Similar fluorescence signals were observed in the HSM(NO3)-grown LCI1 transformants. Because C. reinhardtii has a single, large, cup-shaped chloroplast that closely appresses to the plasma membrane (Ohad et al., 1967), it was difficult to determine whether the fluorescence arose from the chloroplast envelope or the plasma membrane. To resolve this question, CrGFP was fused to the C terminus of LCI1 and further studies were conducted.
To express LCI1 fused with CrGFP, the genomic region of LCI1 was fused in frame with the CrGFP coding region, which was also placed under the control of regulatory elements of the NIA1 gene ( Figures 3A and 8A). The generated plasmid was transformed into the wild-type strain C9 and then Zeocin-resistant colonies were screened for fluorescent signals using a microscope. One typical transformant was selected and analyzed using an anti-LCI antibody to examine whether the cell expressed the LCI1-CrGFP fusion protein in the presence of NO 3 2 ( Figure 8B). A positive band of ;45 kD, which corresponded to the predicted size of the LCI1-CrGFP fusion protein, was detected when the transgenic cells were cultured in HSM(NO3) but not when they were cultured in HSM(NH4). Although the level of the LCI1-CrGFP fusion protein in cells grown in HSM(NO3) was only a small fraction of that of the endogenous LCI1 protein in LC grown cells ( Figure 8B), green fluorescence signals were observed ( Figure 8C). In these experiments, GFP fluorescence was detected in two regions. In many cells, a prominently labeled region surrounding the nucleus was observed. The second region of the cell showing prominent fluorescence was the cell periphery. Because the chloroplast envelope is appressed to the plasma membrane, it was not possible to determine whether the LCI1-CrGFP fluorescence is associated with the chloroplast envelope or the plasma membrane. However, we found that the label was observed surrounding the entire cell, including the region of the cell with the open end of the cup-shaped chloroplast (indicated by arrows in Figure 8C). In addition, no fluorescence was detected in the region of the chloroplast envelope furthest from the plasma membrane.
To clarify further the localization of LCI1, we isolated both chloroplast envelope and plasma membrane fractions. We probed the proteins in the total cell, chloroplast envelope, and plasma membrane fractions with antibodies against LCI1, CCP1, and D1 (Figure 9). Using equal protein loading, we detected all the proteins in the total protein fraction from LC-grown cells. CCP1 was enriched in the chloroplast envelope, which was consistent with the chloroplast envelope localization of this protein (Ramazanov et al., 1993). In contrast with CCP1, a notable enrichment of LCI1 was observed in the plasma membrane fraction (Figure 9). D1 protein was detected only in the total cell protein fraction, indicating very little thylakoid membrane contamination was present in the chloroplast envelope and plasma membrane fractions. Considering the results from fractionation studies, as well as the GFP fluorescence, it is reasonable to conclude that LCI1 is mainly localized to the plasma membrane.

LCI1 Protein Is Involved in Light-Dependent Ci Uptake
In this study, we demonstrated that the level of LCE activity as well as the rate of Ci accumulation increased in transgenic C. reinhardtii cells expressing LCI1 grown in HSM(NO3) compared with cells grown in HSM(NH4) even under HC conditions (Figures  4 and 6). Because NO 3 2 did not affect LCE activity or Ci accumulation in control cells, it is reasonable to conclude that the increases in LCE activity and the rate of Ci accumulation observed in the LCI1 transformants are caused by the LCI1 protein accumulating in response to the change of the nitrogen source from NH 4 + to NO 3 2 in the culture medium. Although the increase in the Ci accumulation in the cells grown in HSM(NO3) was ;2-fold compared with cells grown in HSM(NH4) at the 80 s time point, CO 2 fixation increased ;1.3-fold in transformants grown in HSM(NO3) (Figure 6). These results suggest that LCI1 is involved in the active uptake of Ci rather than directly in CO 2 fixation, although our results do not rule out the possibility that some components required for CO 2 fixation are activated after 80 s when LCI1 is expressed in the absence of other CCM components. The localization data showing that LCI1 is present on the plasma membrane are also consistent with a hypothesis that LCI1 plays a role in the uptake of Ci but not directly in CO 2 fixation. The positive effects of artificial expression of LCI1 can be seen in the LCE activities shown in Figure 4 where the expression of LCI1 under HC conditions in the C2 and E4 transformants increased LCE activity 2-fold. Similarly, Ci affinity showed an apparent increase in HC-grown cells cultured in HSM (NO3) ( Figure 5). However, when these transformants were grown under LC conditions, almost no effect of the artificial expression of LCI1 was observed on Ci-dependent O 2 evolution. This discrepancy between the LCE and Ci-dependent O 2  Cells were grown in Tris-acetate-phosphate medium and then diluted with HSM(NH4) or collected by centrifugation and resuspended in HSM (NO3) to an OD730 of 0.03. Subsequently, the cells were bubbled with air containing 5% CO 2 (HC) or ordinary air (LC). The pH of the media was buffered by supplementing with 20 mM MOPS, pH 7.0, or 20 mM HEPES, pH 7.8. The optical density at 730 nm was measured at the indicated time points. Closed (circles, triangles, and squares) and open symbols (circles, triangles, and squares) show the growth in HSM(NH4) and HSM (NO3), respectively. The circles and triangles represent the growth at pH 7.0 in HC and LC conditions, respectively. Squares show the growth in HC conditions at pH 7.8. LCI1 Enhances Ci Uptake in C. reinhardtii 7 of 13 evolution measurements could be caused by the difference in the conditions and sensitivities of the two methods. The measurement of LCE activity is performed in an open system bubbled with a continuous supply of air containing 50 ppm CO 2 , which provides a more reliable estimation of the activity. On the other hand, the measurement of O 2 evolution is performed in a closed system, where the Ci concentration is constantly decreasing as it is assimilated by the cells. Therefore, at low Ci concentrations, this results in highly variable and underestimated rates of O 2 evolution in the closed system. By contrast, more reliable and sensitive rates of LCE activity are obtained compared with that of O 2 evolution measurement for detecting changes in the characteristics of the CCM. At pH 7.0, LCE activity and Ci affinity increased in all the strains examined compared with the same strains at pH 7.8. Although an increase in the LCE activity of LC-grown LCI1 transformant cells was observed (Figure 4), the K 1/2 value was not affected by the nitrogen source in the media. Because C. reinhardtii cells prefer to use CO 2 rather than HCO 3 2 (Spalding, 1998), the high affinity of the cells for CO 2 possibly increased the background, obscuring the effects of the expression of LCI1. On the other hand, the difference in K 1/2 values was only 10 mM between Q304P3 and the lcr1 mutant at pH 7.0 (see Supplemental Table 1 online), while the LCE activity of the lcr1 mutant was ;60% of that of wild-type Q304P3 cells in LC conditions at the same pH (Table 1). Considering the results for Q304P3 and the lcr1 mutant, the measurement of LCE activity with an air supply of 50 ppm CO 2 could be a more sensitive method than Ci-dependent O 2 evolution for detecting changes in the characteristics of the CCM as described above.
In a previous study, it was shown that the lcr1 mutant had reduced Ci affinity and failed to induce the expression of three genes, namely, CAH1, LCI1, and LCI6, under LC conditions . The cah1 mutant is missing only the periplasmic carbonic anhydrase. We found that while the loss of CAH1 in the cah1 mutant resulted in some kinetic changes to the LCE activity (Figure 1), the overall extent of LCE activity remained Total cell protein samples were isolated from cells of CC503 grown under HC or LC conditions (HC/total and LC/total). Chloroplast envelope and plasma membrane fractions were isolated from a cell wall-less strain grown under LC condition. Protein samples corresponding to 10 mg were separated by SDS-PAGE and then subjected to immunoblot analysis with antibodies against LCI1 (AS-M), the chloroplast envelope protein CCP1, and the thylakoid membrane protein D1. Total, total cell protein; ce, chloroplast envelope; pm, plasma membrane. (A) Schematic illustration of CrGFP protein, which was fused to LCI1 protein at the C terminus. The amino acid sequence GGSGGR represents a synthetic linker between LCI1 and GFP. (B) Accumulation of the LCI1-CrGFP chimeric protein in transformants grown in either HSM(NH4) or HSM(NO3) revealed by protein immunoblot analysis using anti-LCI1 antiserum AS-F. Endogenous LCI1 protein is also indicated. (C) Live image observation of one of the transformants carrying an LCI1-CrGFP fusion construct. Cells were grown in HSM(NO3) and then observed by confocal microscopy using different optical conditions. White arrows show positions corresponding to the open end of the cupshaped chloroplasts. DIC, differential interference contrast; GFP, at 410 to 470 nm; chlorophyll, at 600 to 700 nm. Bars = 5 mm.

of 13
The Plant Cell unchanged. Because the reduction in LCE activity in the cah1 mutant was minimal, this implied that LCI1 and/or LCI6 significantly contributed to LCE activity under normal conditions. Therefore, LCI1 was placed under the control of the NIA1 promoter and put back into the lcr1 mutant. This allowed us to express LCI1 selectively: under HC conditions when most of the CCM components are absent and also under LC conditions, when most of the components of the CCM are present. The effect of the selective expression of LCI1 could be observed more clearly under HC conditions, even though the level of LCI1 protein was not high in the transformants. Although the upregulation of LCE activity, Ci affinity, and Ci uptake/accumulation in the LCI1 transformants grown in HSM(NO3) were apparent (Figures 4 to 6), the levels of LCE activity and Ci affinity were still lower than those of LC-grown wild-type cells (Tables 1 and 2). This may be because the level of LCI1 protein accumulation in these transformants ( Figure 3C) was ;50% of the wild-type level, which may not have been sufficient to support the full levels of LCE activity and Ci affinity. Another possibility is that other LCinducible proteins, including CAH1 and LCI6, were needed to achieve similar levels of CCM activity compared with that of wildtype cells. The upregulation of LCE activity, Ci affinity, and Ci uptake/accumulation in the LCI1 transformants grown in HSM (NO3) was apparent as described above, but the growth of these transformants was not affected by the nitrogen source in this study (Figure 7). Measurements of LCE activity, O 2 evolution, and Ci uptake were performed with a severely limited supply of Ci (e.g., air containing 50 ppm CO 2 for LCE activity measurements) and relatively strong illumination of actinic light at intensities of 1000, 750, and 300 mmol photons m 22 s 21 , respectively. On the other hand, during growth assays, cells were illuminated at 80 mmol photons m 22 s 21 with bubbling of ordinary air containing 370 ppm CO 2 . These differences in light intensity and level of CO 2 supplied for the measurement may be important for observing the effects of the artificial expression of LCI1 resulting from the change in nitrogen source. The absolute increase in the LCE activity of LC-grown LCI1 transformants was ;15 mmol CO 2 mg Chl 21 h 21 compared with the HSM(NO3)-grown lcr1 mutant ( Figure 4, Table 2), which was slightly higher than that in HC-grown cells (10 mmol CO 2 mg Chl 21 h 21 ), although the level of NO 3 2 -induced LCI1 protein in LCgrown cells was 2-fold compared with that in HC-grown cells ( Figure 3C). Because only the LCI1 gene was introduced into the transformants, LCI1 could act alone, but its function is possibly limited in the absence of other LC-inducible proteins, including CAH1 and LCI6. Considering the fact that the cah1 mutant showed very little decrease in LCE activity (Figure 1), LCI6 could also play some role in the CCM. LCI6 contains a Pro-rich region that is involved in protein-protein interactions (Harauz and Libich, 2009). This protein might interact with LCI1 through this region and support its function. The decrease in growth rates of Q304P3 at pH 7.8 compared with that at pH 7.0 might be due not only to Ci utilization but also other factors controlling cell growth. Thus, the expression of LCI1 alone was unable to enhance the growth of cells, indicating that the contribution of LCI1 by itself to the CCM may not be especially large. Alternatively, the growth of the lcr1 mutant might be already partially compensated for by other Ci transporters, obscuring the effects of the expression of LCI1. Interestingly, the mRNA expression level of the chimeric LCI1 gene was quite high in the presence of NO 3 2 ( Figure 3B), but the level of LCI1 protein accumulation was lower than that in LC-grown Q304P3 ( Figure 3C). Therefore, the accumulation of LCI1 protein might possibly be regulated at the level of either translation or stability, in addition to normal transcriptional control. This regulatory pattern seen in the artificially expressing LCI1 transformants may be caused by the exchange of the 59-and 39-UTRs of the LCI1 or lack of other proteins, including LCI6, regulated by LCR1. To understand the detailed molecular mechanism of the regulation of LCI1, it will be important to reveal whether it is mediated by LCR1 or not. It is also possible that LCI1 participates in a protein complex of which the stable accumulation is dependent on other components. In the LCI1 transformants grown in HSM(NO3) under LC conditions, most LC-inducible proteins, other than LCR1regulated proteins, such as CAH1 and LCI6, should accumulate. CAH1 is a periplasmic carbonic anhydrase that does not appear critical to the function of the CCM ( Van and Spalding, 1999). However, the potential role of LCI6 in the CCM and its possible interaction with LCI1 merits further detailed analyses.

LCI1 Is Localized Mainly to the Plasma Membrane
In this study, we showed that fluorescent signals from CrGFPfused LCI1 were detected on the periphery of the cell, in the vicinity of the chloroplast, and on the outer perimeter of the nucleus (Figure 8). The GFP fluorescence that originated from the cell periphery could be from the plasma membrane. The internal signals consisted of a ring-like arrangement around the outer perimeter of the nuclei and patches in the vicinity of the chloroplast. When the chloroplast envelope and the plasma membrane were separated, LCI1 was strongly enriched in the plasma membrane fraction but not in the chloroplast envelope fraction ( Figure  9). These fluorescent images of CrGFP-fused LCI1 and the immunobloting data using separated membrane fractions strongly suggest that LCI1 is localized mainly to the plasma membrane. LCI1 localized at the plasma membrane could be involved in LCE activity to support the CCM. On the other hand, it is unlikely that LCI1 molecules localized around the nuclear envelope and cytosolic vesicles contribute to the CCM. Therefore, it is possible that the internal GFP signals originated from a precursor of LCI1, which underwent transportation to the plasma membrane. The fluorescence signals from either GFP-fused proteins or immunolabeled proteins are not specifically quantitative; thus, the precursor may be undetectable or undistinguishable from the mature CrGFPfused LCI1 in immunoblot analyses ( Figure 8A).

Possible Role of LCI1
It remains to be elucidated how LCI1 is involved in Ci transport. When the LCE activity was compared between HSM(NO3)grown lcr1 and C2 in HC conditions, the increase in the LCE activity in C2 cells grown in HC conditions in HSM(NO3) was constant, at ;10 mmol CO 2 mg Chl 21 h 21 , between the examined pH conditions despite obvious changes in the basic LCE activity in the HSM(NH4)-grown cells (see Table 3, [C2 (NO3)]-[lcr1 (NO3)]). From this observation of the pH independence of LCE activity, it appeared that LCI1 could be involved in the LCI1 Enhances Ci Uptake in C. reinhardtii 9 of 13 transport of both Ci species. On the other hand, the decreases in the K 1/2 value in HSM(NO3)-grown LCI1 transformants were ;30 and 250 mM at pH 7.0 and 7.8, respectively (see Supplemental  Tables 1 and 2 online), indicating a more obvious increase in the Ci affinity in the HSM(NO3)-grown LCI1 transformants at pH 7.8 than at pH 7.0. Therefore, it also appeared that LCI1 functions well at alkaline pH. However, at pH 7.0, the high affinity of C. reinhardtii cells for CO 2 (Spalding, 1998) possibly obscured the effects of the expression of LCI1 on the Ci affinity as described above. Considering these results from the LCE activity and the Ci affinity, it is difficult to conclude whether or not LCI1 increased the flux of a specific species of Ci. According to prediction programs, such as SMART (http:// smart.embl-heidelberg.de/), LCI1 does not contain any domains typical of known transporters. However, the Chlamydomonas, Chlorella, Cocomyxa, and Volvox genomes, as described in the Joint Genome Initiative database (http://genome.jgi-psf.org/), contain several genes and hypothetical genes that are predicted to encode membrane proteins with three or four transmembrane helices, of which the domain structure and hydrophobicity are similar to LCI1 (see Supplemental Figure 2 online). Four regions of LCI1 (Leu-18 to  seemed to be conserved between these proteins, although it is also possible that the similarity is due to convergent evolution. The functions of most of these proteins are unknown, but one of them encodes CST1, which is annotated as a putative transporter protein (Merchant et al., 2007) containing four hydrophobic domains, as does LCI1 (see Supplemental Figure 3 online). Based on the similarity of the domain structure, LCI1 might function as a small transporter or permease for Ci. In addition, LCI1 also shows a low level of similarity to the transmembrane domains of Sho1 in yeast, which plays a crucial role in the osmotic stress response (Saito and Tatebayashi, 2004) by interacting with Hkr1/Msb2 through its transmembrane domains and activating the MAPK cascade (Tatebayashi et al., 2007). Alternatively, LCI1 might enhance Ci uptake by interacting with unknown Ci transporter(s) through the transmembrane helices. In any case, the silicone oil centrifugation experiments clearly showed an increase in the Ci pool sizes of the LCI1 transformants grown in HSM(NO3) (Figure 6).
It is generally expected that multiple Ci uptake and/or accumulation systems function in C. reinhardtii, as can be seen in cyanobacteria. Recently, it was reported that the LC-inducible proteins HLA3 (Duanmu et al., 2009b) and LCIB (Wang and Spalding, 2006) play crucial roles in Ci transport and/or accumulation and that HLA3-dependent and LCIB-dependent processes act partially in overlapping, complementary systems. Whereas the Ci transport and/or accumulation processes involving either LCIB or HLA3 work mainly at alkaline pH, the LCI1dependent mechanism appeared to function at both acidic and alkaline pH, at least between pH 6.2 and 7.8. Therefore, the LCI1dependent Ci transport and/or accumulation process may interact with, or act in parallel with, either LCIB or HLA3.
As proposed previously (Giordano et al., 2005;Spalding, 2008), Ci needs to be transported from the outside of the cells across at least three kinds of membranes, namely, the plasma membrane, chloroplast envelope, and thylakoid membranes, because ribulose-1,5-bisphosphate carboxylase/oxygenase and CAH3 are localized mainly in the pyrenoid and on the lumenal side of pyrenoid tubules, respectively (Borkhsenious et al., 1998;Moroney and Ynalvez, 2007;Duanmu et al., 2009a). Therefore, distinct Ci transporters and accumulation machinery should be localized to each membrane and subcellular compartment. Ci transport and/or accumulation related to LCI1 should be a part of the mechanism. Unfortunately, we could not determine whether LCI1 contributes mostly to CO 2 or HCO 3 2 transport/accumulation. This is likely due to the fact that the amount of the LCI1 protein was lower than in LC-grown wild-type cells. It also remains unclear whether LCI1 is involved in Ci transport or accumulation. However, our results strongly suggest that LCI1 contributes to Ci transport or accumulation at the plasma membrane. Further biochemical analyses on the structure and functional relationships of LCI1 could provide new perspectives on Ci transport and/or accumulation mechanisms in eukaryotic photosynthetic organisms.

Strains and Culture Conditions
Chlamydomonas reinhardtii strains C9 (wild-type strain), CC503 (photosynthetically wild type, wall-less strain), Q304P3 (photosynthetically wild type, cell wall-less strain; Yoshioka et al., 2004), lcr1 (generated from Q304P3; Yoshioka et al., 2004), lcr1+G (lcr1 mutant complemented with LCR1; Yoshioka et al., 2004), and cah1 ( Van and Spalding, 1999) were grown in Tris-acetate-phosphate medium and diluted with modified HSM (NH4) supplemented with 20 mM MOPS, pH 7.0, to an OD 730 of ;0.01 (;7 3 10 4 cells mL 21 ) for photoautotrophic growth. Subsequently, the cells were cultured at 288C with illumination at 80 mmol photons m 22 s 21 under HC conditions for 24 h until the cell density reached an OD 730 of ;0.1. The cells were further cultured for 12 h under HC or LC conditions until they reached logarithmic growth phase. To induce the gene expression of the transgene, cells grown in HSM(NH4) containing 9.35 mM NH 4 Cl for 24 h were collected by centrifugation and then resuspended in fresh HSM(NO3) containing 9.35 mM KNO 3 . After the change in media, cells for RNA analysis were grown for an additional 6 h in HC or LC conditions, while cells used for physiological measurements or protein immunoblot analyses were grown for 12 h under HC or LC conditions. The cell density (OD 730 ) per 5 mg Chl mL 21 was ;0.6 for all the examined strains, justifying normalization on the basis of chlorophyll content in the physiological measurements.

LCE Activity with an Open Gas Exchange System
The cells grown as described above were collected by centrifugation at 700g for 10 min and resuspended in fresh HSM(NH4) or HSM(NO3) media. To analyze the effects of pH, the collected cells were also resuspended in HSM media of pH 6.2 and 7.8, which were prepared using 20 mM MES, pH 6.2, or 20 mM HEPES, pH 7.8. The light-dependent CO 2 gas exchange activity of cells (5 mg mL 21 of chlorophyll) was measured with bubbling air containing 50 ppm CO 2 under actinic light illumination of 1000 mmol photons m 22 s 21 at 288C using an open infrared gas analysis system that records the rate of CO 2 exchange as a function of time as described previously (Yamano et al., 2008).

Intercellular Concentration of Dissolved Ci
The intercellular concentration of dissolved Ci was measured by the silicone oil centrifugation method (Fukuzawa et al., 1998). Cells grown as 10 of 13 The Plant Cell described above were collected by centrifugation (700g at 288C) and subsequently suspended at a cell density of 25 mg mL 21 chlorophyll in 50 mM HEPES-NaOH buffer, pH 7.8, that had been bubbled with N 2 gas. One milliliter of cell suspension was bubbled with N 2 gas for 10 min in an O 2 electrode until they no longer produced O 2 . First, a 60-mL silicone oil layer (SH550/SH556 = 4/7 [v/v]) was overlaid on a 20-mL layer of the termination solution containing 1 M glycine-NaOH, pH 10.0, and 0.75% SDS (w/v). Then, the cells (300 mL) were further overlaid on the silicone oil layer. SH550 and SH556 were purchased from Toray Dow Corning Silicone and Nacalai Tesque, respectively. Ci uptake was initiated by the addition of 10 mL of NaH 14 CO 3 at 50 mM final concentration, which was immediately followed by 20, 40, or 80 s of illumination with an actinic light source of 300 mmol photons m 22 s 21 , and the reaction was terminated by centrifugation. Control cells that had been supplied with nonradioactive NaHCO 3 were collected as the zero time point. After centrifugation, the labeled cells were immediately frozen using liquid nitrogen and then suspended in 0.1 N NaOH. The alkaline cell suspension was divided into two aliquots. One was directly subjected to a liquid scintillation counting, which was described as total Ci consumption. The other aliquot was added to 0.5 N HCl, desiccated to liberate 14 C except for the fixed 14 CO 2 and then subjected to a liquid scintillation counting to analyze the 14 C level, which was described as CO 2 fixation. Ci uptake/accumulation was calculated as the difference between the total Ci consumption and CO 2 fixation. The Ci uptake was corrected by estimating the cell volume as sorbitol impermeable space (SIS) using [ 14 C]-sorbitol and 3 H 2 O as described previously (Heldt, 1980).

Ci-Dependent Photosynthetic Oxygen Evolution
The affinity of cells for Ci was evaluated by measuring the rate of dissolved Ci-dependent photosynthetic O 2 evolution. Cells were collected by centrifugation and then resuspended in 50 mM HEPES-KOH buffer, pH 7.8, or 50 mM MOPS-KOH buffer, pH 7.0, at 20 mg mL 21 chlorophyll. Photosynthetic oxygen evolution was measured using a Clark-type oxygen electrode (Hansatech Instruments) (Yamano et al., 2008). The maximum O 2 -evolving activity defined as V max was measured in the presence of 5 mM NaHCO 3 .

Plasmid Construction and Transformation of C. reinhardtii Cells
The LCI1 coding region was amplified by PCR using Pyrobest polymerase (TaKaRa Bio) using genomic DNA from Q304P3 as a template and the primers 59-AAAtcgcgaATGTCGGGCTTGAACAAGTTC-39 containing a NruI site and 59-CGCggatccCTTACACCTTGGTGGCGC-39 including a BamHI site, and then inserted into an expression vector pTY2b (DDBJ/ NCBI accession number AB447355). To obtain the plasmid for expression of LCI1-CrGFP fusion protein, the CrGFP coding region was amplified from the plasmid pMF59 (Fuhrmann et al., 1999) as a template with primers 59-CGgaattcCgcggccgcATGGCCAAGGGCGAGGAGCTGTTC-39 and 59-CGgaattcTTACTTGTACAGCTCGTCC-39 and then cloned into the EcoRI site of pTY2b. The resultant vector was designated as pTY2b-GFP. The LCI1 gene flanked with 59-and 39-UTRs was amplified by PCR using genomic DNA from the wild-type strain C9 as a template using primers 59-GGAagatctTCCAACTAGTGTGTTGCATCGTTTCTCGCCTT-TACAGTCGAGTACC-39 and 59-ATTTgcggccgcCGGAGCCGCCCGC-GGTGGCGGGCTTGGC-39 and inserted between the Bglll and NotI sites of pTY2b-GFP. The resultant vector was designated as pTY2b-Lci1-GFP. C. reinhardtii cells in early logarithmic phase were collected and transformed by electroporation (Shimogawara et al., 1998) using a 4-mm gap electroporation cuvette with a Gene Pulser II (Bio-Rad) with a modified 25-mF capacitor and a shunt resistor of 330 V. The transformants were incubated at 288C for 24 h with gentle agitation under illumination at 10 mmol photons m 22 s 21 and then screened on Trisacetate-phosphate plates containing 10 mg mL 21 of Zeocin.

Protein Immunoblot Analysis
Whole-cell proteins corresponding to 2 mg of chlorophyll were suspended in 50 mM Tris-HCl, pH 8.0, buffer containing 25% glycerol (w/v) to a volume of 10 mL. The cells were solubilized by incubation at 1008C for 1 min in the presence of 2% SDS and 0.1 M DTT and subsequently centrifuged at 13,000g for 3 min. The supernatant containing solubilized proteins was subjected to SDS-PAGE with a 15% polyacrylamide gel that contained 6 M urea (Ohnishi and Takahashi, 2001). In the case of protein immunoblot analysis for the cell fractionation experiments, samples containing equal amounts of protein solubilized in a buffer containing 5% SDS, 0.1 mM DTT, 0.1 M Na 2 CO 3 , 7 M urea, 0.02% mercaptoethanol, and 25 mM Tris were run on 12% SDS-PAGE gels. Separated polypeptides were blotted electrophoretically onto polyvinylidene difluoride membranes and probed with polyclonal antibodies against LCI1 (1 3 10 23 dilution), CCP1 (1 3 10 23 dilution; Ramazanov et al., 1993), and D1 protein of photosystem II (1 3 10 24 dilution; product number AS05 084 from Agrisera). The antibodies against LCI1 (AS-F and AS-M) were raised in rabbits using synthetic oligopeptides as the antigens, which were FDTQEGIDKYTPVY (from Phe-43 to Tyr-56) or DAEE-SHAMPNVHVTSDGATKV (from Asp-172 to , respectively. The former oligopeptide was also used for affinity purification of the antibody against LCI1. The rabbit antibody against CCP1 was raised against the peptide CFKQVMSKHGIKGLYRGFTST. Signals were visualized using an enhanced chemiluminescence reagent (ECL; GE Healthcare) and detected using a LAS Image analyzer (FUJI FILM). Three typical blots were subjected to densitometric quantification. ECL-detectable recombinant protein mixture (The MagicMark XP Western Protein Standard; Invitrogen) was used as a size marker.

GFP Fluorescence Imaging
For GFP fluorescence imaging, live cells were immobilized in 1% lowmelting-point agarose and observed using a laser scanning microscope LSM710 system at an excitation of 405 nm and emission at 490 to 550 nm for GFP image capture and at an excitation of 488 nm and emission at 600 to 700 nm for the chlorophyll image capture.

Isolation of Chloroplast Envelope and Plasma Membrane
Intact chloroplasts were isolated from cells of CC503 according to Mason et al. (2006). Isolated chloroplasts were then used for chloroplast membrane fractionation by the method of Clemetson and Boschetti (1988). The chloroplast envelope proteins were solubilized in buffer containing detergent. The procedure followed that for the isolation of plasma membranes adopted from Norling et al. (1996). Protein concentration was determined using the bichinconic acid assay (Pierce) before loading on SDS-PAGE gels.

Accession Numbers
Sequence data from this article can be found in the DDBJ/NCBI data libraries under accession numbers AB447355 (pTY2b), U31976 (LCI1), and XP_001693802 (CST1).

Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure 1. Indirect Immunofluorescence Analyses of LCI1 Protein in C. reinhardtii Cells.
Supplemental Figure 2. Comparison of Amino Acid Sequences between LCI1 and Proteins That Contain Similar Domain Structure to LCI1. LCI1 Enhances Ci Uptake in C. reinhardtii