Introduction

Microalgal photosynthesis plays a significant role in global biomass production by contributing a large portion of carbon capture and sequestration. Living in an aquatic ecosystem with frequent CO2 limitation, many microalgae species have developed a CO2 concentrating mechanism (CCM) to enhance their photosynthetic performance and to mitigate the stress caused by the inadequate CO2 supply (Spalding 2008, 2009; Badger and Price 2003; Raven et al. 2008). The operation of the CCM results in a large intracellular inorganic carbon (Ci) pool that elevates the CO2 concentration at the site of ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco), the enzyme that catalyzes the carboxylation of ribulose-1,5-bisphosphate (RuBP), which is the primary step to assimilate CO2 into the photosynthetic carbon cycle. Since Rubisco also catalyzes the oxygenation of RuBP to divert the carbon into the “wasteful” photorespiratory pathway, and is also known for its inherently slow turnover rate, elevated intracellular Ci effectively increases the apparent CO2/O2 specificity for Rubisco, decreases the carbon flux into the photorespiratory pathway and increases the rate of carboxylation by raising the CO2 substrate concentration closer to saturation.

Diverse forms of the microalgal CCM have been discovered in prokaryotic species and eukaryotic species (Badger and Price 2003; Giordano et al. 2005; Spalding 2008, 2009; Raven et al. 2008). Even though the CCMs vary in complexity in different organisms, three major components are common and essential in these microalgal CCMs: (1) Ci uptake systems responsible for accumulating high levels of intracellular Ci; (2) enzymatic systems, including various carbonic anhydrases (CAs), catalyzing rapid inter-conversion between different Ci species and acting in concert with Ci transporters to facilitate Ci uptake and accumulation; and (3) sequestered Rubisco in a specified micro-compartment where CO2 can be elevated, as exemplified by cyanobacterial carboxysomes or pyrenoids in many eukaryotic algae. Since the discovery of the microalgal CCMs, significant effort has been devoted in understanding the molecular mechanism of Ci uptake systems. To date, many fundamental aspects of the Ci transport systems in prokaryotic algae have been revealed at the biochemistry and gene level by studying several cyanobacteria model species. In eukaryotic microalgae, understanding of the Ci transport has been gained largely from Chlamydomonas reinhardtii (hereafter, Chlamydomonas), a unicellular green alga used as a model system to study the eukaryotic CCM for decades because of its well-established genetic background. Only in recent years, with the completion of the Chlamydomonas genome sequencing and development of new molecular tools, have many gene products been discovered as candidates for the functional components of the Chlamydomonas Ci uptake systems, and a few among them have been recently confirmed as being directly involved in the Ci accumulation. In this review, we will focus on the recent advances in identifying and characterizing Ci transport systems and a putative Ci recapturing system in Chlamydomonas.

Chlamydomonas CCM overview

In aqueous environments, green algae, like Chlamydomonas, obtain carbon dioxide from dissolved Ci sources (CO2, HCO3 −, and CO3 2−). These Ci species must cross two barriers, the plasma membrane and the chloroplast inner envelope to reach the stromal space, which is believed to be the primary location for the accumulated Ci pool. It has been confirmed that active Ci transport occurs at both the plasma membrane and the chloroplast envelope in low CO2 acclimated cells (Sültemeyer et al. 1988, 1989; Palmqvist et al. 1988). Regardless of which Ci species being transported into chloroplasts, it is reasonable to assume that HCO3 − is the major form accumulated there, because HCO3 − is much less permeable to lipid membranes and the alkaline stromal pH favors HCO3 − formation. The rapid conversion of CO2 diffusing into the chloroplast to HCO3 − in the stroma may require CA, and CAH6, a plastid CA, has been suggested for this role (Mitra et al. 2004; Moroney and Ynalvez 2007). Accumulated HCO3 − cannot be directly used by Rubisco, and must be dehydrated to CO2, because Rubisco can only use CO2 as substrate. CAH3, a thylakoid lumen CA, catalyzes this critical step and releases CO2 in the vicinity of Rubisco. This has been demonstrated in cah3 mutants defective in the CAH3 gene. In these mutants, a large amount of Ci is apparently accumulated intracellularly but not available to Rubisco (Spalding et al. 1983a; Moroney et al. 1985; Funke et al. 1997; Karlsson et al. 1998). In such context, HCO3 − also must be transported across the thylakoid membranes into the thylakoid lumen, although this has not been demonstrated so far. The central site of CO2 fixation is in the pyrenoid, a specialized structure within the chloroplast stroma retaining high levels of Rubisco. It has been shown that some thylakoid tubules penetrate into pyrenoid bodies and are enriched with CAH3 (Mitra et al. 2005; Markelova et al. 2009 ), and CO2 is assumed to be released from these tubules for Rubisco in the pyrenoid.

Ci transporters in plasma membranes

Photosynthetically driven net uptake of Ci across the plasma membrane represents the largest nutrient flux that green algae encounter. On the basis of evidence from isotope disequilibrium and mass spectrometric studies, acclimated Chlamydomonas cells can take up both CO2 and HCO3 − during steady state photosynthesis (Sültemeyer et al. 1988, 1989; Palmqvist et al. 1988), although it is still questionable whether or not the observed CO2 uptake at the plasma membrane is an active transport or rapid diffusion in response to active consumption of CO2 internally (Amoroso et al. 1998). So far only two candidate plasmalemma Ci transporters have been identified in Chlamydomonas, HLA3/MRP1 and LCI1, as discussed below.

Im and Grossman (2002) identified the HLA3 gene, encoding a putative ATP-binding cassette (ABC) type transporter of the multidrug-resistance-related proteins (MRP) subfamily, whose expression is activated by both high light (under conditions in which the cells were becoming CO2 limited) and low CO2, for which it was shown to be under the control of CIA5 (=CCM1), the Chlamydomonas master regulator for the responses to limiting CO2 conditions (Miura et al. 2004; Xiang et al. 2001; Fukuzawa et al. 2001). HLA3 was predicted to be targeted to the secretory pathway and possibly to be located in the plasma membrane. As HLA3 is not induced in high light under high CO2 conditions, and not expressed in the cia5 mutant, the putative transport protein encoded by this gene was speculated as a vital component of the CCM, possibly even a Ci transporter. Interestingly, evidence shows that the whole cell Ci uptake can be inhibited by the H+-ATPase inhibitor vanadate, and MRP-associated transport is often sensitive to vanadate (Karlsson et al. 1994).

BCT1, a limiting Ci inducible high affinity HCO3 − transporter of the ABC type, also was identified from Synechococcus PCC7942 and several other cyanobacterial species (Omata et al. 1999). BCT1 is a four subunit complex encoded by the cmpABCD operon: CmpA protein was shown to bind HCO3 − specifically with a KD around 5 μM; CmpB is a membrane protein and may form a dimer as transport path; both CmpC and CmpD have conserved ATP-binding sites. In contrast, HLA3/MRP1 contains a full size ABC-MRP domain arrangement and may function as single protein transporter.

The involvement of HLA3 in active Ci uptake was experimentally confirmed by efficient RNAi-mediated knockdown in both wild type and mutant Chlamydomonas strains (Duanmu et al. 2009b). When HLA3 expression alone was inhibited, substantial decreases in both photosynthetic Ci affinity and Ci uptake were observed at alkaline pH, at which the predominant form of Ci is HCO3 −. When combined with either LCIB mutations or simultaneous knockdown of LCIA mRNA, the cells demonstrated more severe growth defects, together with further decreased photosynthetic Ci affinity and Ci uptake. These results strongly support the hypothesis that HLA3 is directly or indirectly required for HCO3 − transport into the cells. HLA3 homologs also were found in several other microalgae, such as Volvox carteri, Chlorella sp. NC64A, and Ostreococcus RCC809 (Yamano and Fukuzawa 2009). However, further biochemical characterization of HLA3/MRP1 is still needed to elucidate the mechanism of this putative plasma membrane HCO3 − transporter.

The LCI1 gene was first identified as a limiting-CO2-inducible gene (Burow et al. 1996) and the expression of LCI1 is controlled by CIA5 as well as a MYB type transcription factor LCR1, both of which are implicated in regulation of Chlamydomonas CCM genes (Im et al. 2003; Miura et al. 2004; Yoshioka et al. 2004). The LCI1 gene encodes a novel protein of 192 amino acids showing very little homology to any other proteins in the available databases. The location of LCI1 and its involvement in Ci uptake was recently demonstrated by artificially over-expressing LCI1 in the LCR1 mutant (Ohnishi et al. 2010). Increased Ci affinity and Ci uptake were observed in the over-expression strains when compared to the LCR1 mutant, in which only trace amount of LCI1 could be detected. Changing external pH from 6.2 to 7.8 appeared to have little impact on the LCI1-stimulated increase in Ci consumption measured as the light-dependent CO2 gas-exchange activity, which led Ohnishi et al. (2010) to conclude that LCI1-involved Ci uptake system may not differentiate CO2 and HCO3 −. It should be noted, however, that the host strain LCR1 does not express CAH1, a major periplasmic CA, so the hydration of CO2 to HCO3 − at the cell surface in the LCI1 over-expression lines may not occur as rapidly as that in the wild type strains. Thus, the actual CO2/HCO3 − ratio may be under-estimated in the CO2 gas-exchange measurement at higher pH. Predicted to contain three transmembrane helices, LCI1 also was confirmed to be associated with the plasma membrane. Therefore, regardless of the Ci species transported, LCI1 appears to represent a novel plasma membrane Ci transporter or an essential component of a novel Ci uptake system.

Ci transport into chloroplasts

It is commonly accepted that chloroplasts are essential in Ci concentrating and active uptake of Ci at the chloroplast envelope has been demonstrated by using isolated intact chloroplasts (Sültemeyer et al. 1988, 1989; Amoroso et al. 1998). Some promising candidate genes and proteins have been proposed, including LCIA (NAR1.2), CCP1, CCP2, LCIB, and the unique LCIB protein family LCIC/LCID/LCIE.

LCIA (also named NAR1.2) encodes a membrane protein belonging to the Formate/Nitrite Transporter (FNT) family (Galvan et al. 2002; Miura et al. 2004). LCIA was proposed to be a putative Ci transporter because the expression of LCIA is induced under limiting-CO2 conditions and is under the control of CIA5, a transcription factor required for induction of most other CCM genes. LCIA is predicted to be targeted to the chloroplast and has six transmembrane helices. When expressed in Xenopus oocytes, LCIA reportedly increased the HCO3 − transport activity in the oocytes but exhibited a very low affinity for HCO3 − and high affinity to nitrite, leaving its importance as Ci transporter in question (Mariscal et al. 2006). It has recently been shown that LCIA and HLA3 co-knockdown in wild type Chlamydomonas cells caused more severe phenotypic impact than HLA3 knockdown alone, suggesting that LCIA and HLA3 are key synergistic or complementary components of the active Ci transport pathway in limiting Ci acclimated cells (Duanmu et al. 2009b).

Two other genes, CCP1 and CCP2, encode two similar putative plastid Ci transport proteins with 96% identical amino acid sequences (Spalding and Jeffrey 1989; Ramazanov et al. 1993; Chen et al. 1997). With six transmembrane domains, CCP1 and CCP2 are associated with the chloroplast envelope, and are among the most abundant low-CO2-inducible proteins. They exhibit sequence similarity to the mitochondrial carrier protein superfamily, which consists of many small proteins involved in transporting various metabolites across the mitochondrial inner membrane. However, RNAi-mediated knockdown of CCP1 and CCP2 showed little impact on photosynthetic Ci affinity, although the RNAi strains grew poorly in low CO2 (Pollock et al. 2004). It was suggested that these two putative transporters may be involved with transport of metabolic intermediates important in limiting CO2 acclimation. It is also possible that, as was observed with the HLA3 RNAi knockdown, which exhibited no clear phenotype unless paired with decreased function of another transporter or of a Ci accumulation system, any Ci transporter functions of CCP1/CCP2 may have been masked by some other Ci transport or accumulation systems under the conditions in which the Ci affinity was tested.

LCIB associated Ci accumulation in chloroplasts

The pmp1 mutant and its allelic strain ad1 (air dier 1) are among the very few single gene mutants identified so far that fail to accumulate Ci when grown in atmospheric CO2, and thus the pmp1 mutant was long proposed as a mutant defective in active Ci transport (Spalding et al. 1983b). Interestingly, pmp1 and ad1 mutants exhibit a distinct air dier phenotype, i.e., they can survive in either a high concentration (5%; HC) or a very low concentration (<0.01%; VLC) of CO2, but die in an intermediate concentration of CO2 (atmosphere level ~0.04%; LC), indicating the existence of multiple acclimation states associated with different levels of CO2 concentration (Van et al. 2001; Vance and Spalding 2005). These allelic mutations directly affecting the internal Ci accumulation and the growth defects in pmp1 and ad1 have been linked to lesions in LCIB (Wang and Spalding 2006), a gene previously identified among the low-CO2-inducible transcripts (Miura et al. 2004).

The LCIB gene belongs to a novel gene family. LCIB and three homologous genes in Chlamydomonas, LCIC, LCID and LCIE are all responsive to limiting CO2, and the transcripts of LCIB and LCIC are among the most abundant transcripts upon the limiting CO2 induction (Wang and Spalding 2006; Yamano et al. 2008). LCIB orthologs can also be found in a few other green algae and diatom species (given that their genome sequences are available), but show a very limited phylogenetic distribution outside these groups, only in a very small number of cyanobacteria and bacteria species (Yamano et al. 2010). In contrast to its long postulated role as a Ci transporter, LCIB is a soluble protein lacking any obvious transmembrane domain or any recognizable motifs. LCIB has been localized in the chloroplasts, with two distinct localizations observed, either dispersed throughout the entire chloroplast stroma or concentrated mainly in a discrete region surrounding the pyrenoid (Duanmu et al. 2009a; Yamano et al. 2010). It has been recently reported that LCIB and LCIC form a 350 KD hexameric complex, and that the LCIB/LCIC complex could re-localize from the stroma to the region surrounding the pyrenoid when CO2 concentration changed from high CO2 to low CO2, whereas the pyrenoid associated LCIB/LCIC complex in low CO2 rapidly diffused from around the pyrenoid to the stroma when the cells were transferred to high CO2 or to the dark (Yamano et al. 2010). Yamano et al. (2010) argue that the re-localization of the LCIB/LCIC complex is associated with the acclimation status under different CO2 conditions and that the pyrenoid associated location is essential for the function of LCIB (see discussion below).

The demonstration that CAH3 loss-of-function mutations suppress the air-dier phenotype of LCIB mutants has shed some light on the puzzled function of LCIB (Duanmu et al. 2009a). The LCIB/cah3 double mutants exhibit a similar growth phenotype as cah3 mutants (slow growth in low CO2 but no growth in very low CO2), as well as over-accumulated intracellular Ci concentrations when grown in either low or very low CO2. The over-accumulation of Ci in the cah3/LCIB double mutant suggests that LCIB must not directly participate in Ci uptake because the LCIB mutation, when combined with cah3 mutation, has little effect on Ci uptake and accumulation. The epistatic effect of cah3 on the LCIB mutation also argues that LCIB probably functions downstream of CAH3. As the function of CAH3 is to dehydrate HCO3 − and release CO2 from the lumen (Duanmu et al. 2009a), this suggests that LCIB functions to recapture the excess CO2 released by CAH3, thus preventing the CO2 leakage.

Based on this hypothesis, without LCIB, a futile cycle would be created between the active Ci uptake and the rapid CO2 release by CAH3 through dehydration of HCO3 −. The chloroplast stromal CA, CAH6 also has been proposed to prevent CO2 leakage (Mitra et al. 2004; Moroney and Ynalvez 2007), and a possible interaction between CAH6 and LCIB, physically or functionally, has been suggested for the proposed Ci-recapture mechanism (Duanmu et al. 2009a). A logical extension of this hypothesis would be to speculate that since the LCIB/LCIC complex apparently can trap CO2 generated by CAH3-mediated dehydration of accumulated HCO3 −, this complex might also be responsible for active accumulation of CO2 diffusing into the stroma from outside the cell; i.e., that the LCIB/LCIC complex may be responsible for the apparent active uptake of CO2 in Chlamydomonas.

Since the LCIB/LCIC complex was reported to localize in the vicinity of the pyrenoid in low CO2, an alternative hypothesis also was proposed in which the LCIB/LCIC complex may act as a structural barrier to avoid the leakage of CO2 and maintain CO2 concentrations in the pyrenoid matrix for efficient CO2 fixation (Yamano et al. 2010). It should be possible to distinguish experimentally between the two hypotheses regarding the function of the LCIB/LCIC complex, i.e., either trapping CO2 into the HCO3 − pool or physically blocking loss of CO2 from the pyrenoid.

CemA (YCF10) dependent system

Disruption of the plastid YCF10 open reading frame in Chlamydomonas resulted in high light sensitivity and decreased Ci uptake (Rolland et al. 1997). The product of YCF10 displays sequence homology with the plant chloroplast protein CemA and the cyanobacterial PxcA (CotA) protein. CemA proteins are integral membrane proteins localized in the inner chloroplast envelope. The cyanobacteria PxcA protein has been implicated in Ci uptake, but it was suggested that it functions in light-induced Na+-dependent proton extrusion to maintain electrical and pH homeostasis during Ci uptake, rather than acting as a Ci transporter itself (Katoh et al. 1996a, b). The Chlamydomonas CemA protein has been proposed to play a similar role as PxcA during chloroplast Ci uptake, so is thought to pump protons out of the stroma (Rolland et al. 1997; Kaplan and Reinhold 1999). This proton extrusion may be required to maintain the alkaline stromal pH for functional Ci accumulation in chloroplasts, as hydration of transported or diffused CO2 in chloroplast stroma should generate protons.

Putative CO2 channel

The Chlamydomonas genome also encodes two Rhesus-like proteins (RHP1 and RHP2), which have similarity to the Rh proteins in the human red blood membrane (Soupene et al. 2002, 2004; Kustu and Inwood 2006). Both RHP1 and RHP2 are predicted to have 12 transmembrane domains, and RHP1 has recently been localized to plasma membranes (Yoshihara et al. 2008). The expression of RHP1 is relatively low in low CO2 acclimated cells, but is highly induced by high CO2. Since the RNAi knockdown of RHP1 expression resulted in a growth defect in high CO2 but not in low CO2, the role of RHP1 was proposed as a bidirectional CO2 channel that allows rapid equilibration of CO2 under high CO2 growth conditions to provide sufficient CO2 for photosynthesis (Soupene et al. 2004). No evidence so far supports a role for RHP1 and/or RHP2 in the facilitated CO2 diffusion in low CO2 conditions, although, because both are expressed to some extent in low CO2, such a possibility cannot be ruled out.

Conclusions and future challenges

Recent advances in genetic and molecular approaches have unraveled some unique candidate components for the Ci uptake and accumulation systems in Chlamydomonas. As illustrated in Fig. 1, several putative Ci transporters, as well as other candidate components for Ci accumulation can be integrated into a model for the Chlamydomonas CCM.

Fig. 1
figure 1

Hypothetical schematic model of the Chlamydomonas CCM with emphasis on inorganic carbon (Ci, CO2, and HCO3 −) transport. The proposed transport of Ci from the extracellular space to the thylakoid lumen via the cytosol and stroma is illustrated. The carbonic anhydrases (red ovals) catalyze the interconversion of CO2 and HCO3 − in respective compartments. Non-membrane-bound subcellular structure pyrenoid (boundary labeled as dashed line) is within the stroma and is the site of Rubisco localization. Blue ovals indicate HCO3 − transporters and these include HLA3, LCI1, and LCIA. Additional potential HCO3 − transporters (light blue circles) or HCO3 − facilitated diffusion channels (light blue barrels) are indicated with question marks. Putative CO2 gas channel and proton extrusion from stroma are operated by RHP1 (purple barrel) and CemA (purple oval), respectively. Stroma localized LCIB/LCIC complex is proposed to be involved in trapping of the CO2 released from thylakoid lumen via rehydration into the HCO3 − pool in the stroma, possibly including involvement of the putative stromal CA, CAH6

Full size image

However, even though a number of candidate proteins have been identified and implicated in Ci uptake, their precise roles still remain undefined. For example, are HLA3, LCIA, and LCI1 bona fide Ci transporters? How does LCIB function to either capture or prevent leakage of CO2? Answering these questions requires more extensive biochemical, biophysical, and molecular studies.

One of the biggest challenges in studying the CCM is identifying the essential Ci transport components, clarifying their substrate specificities and elucidating the regulatory mechanisms. A large number of CO2 responsive genes have been revealed by global gene expression analysis (Miura et al. 2004; Yamano et al. 2008; Yamano and Fukuzawa 2009), and there is no doubt that the number on this gene list will get even larger with the recent development of high-throughput sequencing technologies. Interestingly, but not surprisingly, many Chlamydomonas CO2 responsive genes have been revealed as novel genes. Some of these novel gene products may be associated with still unidentified Ci uptake systems or Ci accumulation mechanisms in eukaryotes. It will be rewarding to link these genes to their physiological functions, as elucidation of these novel gene functions will provide potential genetic engineering approaches for improving biomass productivity and bioproducts.

Even with devoted efforts, only a small number of Chlamydomonas CCM mutants have been generated through classic forward genetic approaches, and none of these mutants has been demonstrated to be defective in genes that encode Ci transport proteins, raising a doubt as to whether phenotypical screening will lead to identification of defects in Ci uptake systems. As has been seen in cyanobacteria (Price et al. 2002, 2008), the complexity and diversity of active Ci uptake systems may have complementary functions, thus preventing identification of putative Ci transporters by simple phenotype screening. However, a potential Ci transport system can be revealed when multiple genetic defects are combined, as evidenced by the phenotypes in HLA3/LCIA RNAi co-knockdown or HLA3/LCIB double mutant strains (Duanmu et al. 2009b), and as demonstrated in cyanobacteria (Price et al. 2002, 2008). As many novel candidate genes associated with the Ci uptake or Ci transport systems are available, characterization of their function may require a more reliable reverse genetic approach. Whereas the RNA silencing approaches (RNAi or miRNA) have been proven successful in many cases (Schroda 2006; Zhao et al. 2009; Molnar et al. 2009), future advances in directed mutation provided by, for example, zinc finger nuclease technology (Durai et al. 2005) or TAL nuclease technology (Li et al. 2011), should provide more powerful tools to characterize these potential components in Ci transport or Ci uptake systems.