Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

7.I Introduction

Lipid biosynthesis in plants has been studied for decades and our current molecular understanding of lipid metabolism in plants is substantial. Genes encoding enzymes of glycerolipid biosynthesis and fatty acid desaturation have been identified by genetic and biochemical means (Ohlrogge and Browse, 1995; Joyard et al. 1998; Frentzen, 2004; Benning and Ohta, 2005; Holzl and Dörmann, 2007), and the first examples of components involved in lipid trafficking between the plastid and the endoplasmic reticulum (ER) are being discovered (Jouhet et al. 2007; Benning, 2008). Annotation of the Arabidopsis genome sequence (The Arabidopsis Genome Initiative, 2000) has led to the identification of novel genes, which likely encode proteins involved in lipid biosynthesis, trafficking, and catabolism (Beis-son et al. 2003).

Like Arabidopsis, the eukaryotic green alga Chlamydomonas reinhardtii is a well established model for the study of different processes of general relevance, such as photosynthesis (Niyogi, 1999) and post-transcriptional gene silencing (Wu-Scharf et al. 2000). Beyond these, Chlamydomonas research has provided substantial insights into processes more specific to unicellular algae, e.g., phototaxis and flagellar function (Sil-flow and Lefebvre, 2001), nutrient acquisition (Davies et al. 1994, 1996, 1999), and microalgal metabolism (Grossman et al. 2007). The recent completion of the Chlamydomonas genome sequence (Merchant et al. 2007), as well as the development of insertional mutagenesis (Tam and Lefebvre, 1993), RNA interference (RNAi) methods (Fuhrmann et al. 2001; Sineshchekov et al. 2002), and a molecular map (Kathir et al. 2003) make Chlamydomonas an attractive model to study gene function by genetic or direct molecular analysis. Preliminary annotations of lipid genes present in the genome of Chlamydomonas were recently published (Riekhof et al. 2005b; Riekhof and Benning, 2008). Based on these attributes, Chlamydomonas has great promise for the analysis of the biosynthesis and physiological functions of different lipids.

Availability of a suitable microalgal model system is timely, as microalgae are increasingly discussed as a biomass resource for the production of biofuels that does not have to compete with the agricultural production of food crops (Hu et al. 2008). While Chlamydomonas rein-hardtii itself is not a candidate species for the commercial production of biofuels, it still is the best studied microalga at the genetic and genomic level. Moreover, Chlamydomonas is related to other unicellular green algae that are commercially used, e.g., Dunalliella salina, and Chlamydomonas has been reported to accumulate triacylglycerols (TAGs) under conditions of nutrient deprivation (Weers and Gulati, 1997) or high light (Picaud et al. 1991). Chlamydomonas also synthesizes TAGs from lipids supplied in the medium (Grenier et al. 1991). To fill in the gaps in knowledge, efforts are currently underway in our lab to genetically dissect the biosynthesis of TAGs and its regulation in Chlamydomonas, and to identify genes that might be useful for the engineering microalgal production strains.

7.II General Differences in Lipid Metabolism between Chlamydomonas and Seed Plants

As elaborated below, many aspects of lipid metabolism follow common pathways that were presumably established during the evolution of chloroplasts of green algal and plant ancestors (Reyes-Prieto et al. 2007). However, at least two possibly related aspects of lipid metabolism in Chlamydomonas differ from lipid metabolism in seed plants (Fig. 1). Most prominently, Chlamydomonas is unable to synthesize the otherwise common phosphoglycerolipid phosphatidylcho-line (PC). Instead, it produces the non-phosphorus betaine lipid diacylglyceryl-N,N,N-trimethylho-moserine (DGTS) (Eichenberger and Boschetti, 1977). This lipid is similar in structure and function to PC (Fig. 2) and is thought to substitute for PC in Chlamydomonas (Sato and Murata, 1991; Sato, 1992; Moore et al. 2001). Interestingly, PC is central to lipid metabolism in developing seeds or leaves where it serves as substrate for fatty acid modifying enzymes, such as desaturases (Browse and Somerville, 1991; Ohlrogge and Browse, 1995; Wallis and Browse, 2002), or possibly as the lipid transferred between the ER and the plastid (Jouhet et al. 2007; Benning, 2008).

Fig. 1.
figure 1

Overview of glycerolipid biosynthesis in Chlamydomonas. Endproducts are shown in bold. Abbreviations: ACP, acyl carrier protein; AdoMet, S-adenosylmethionine; ASQD, 2′-O-acyl-sulfoquinovosyldiacylglycerol; CDP, cytidine 5′-diphosphate; CoA, coenzyme A; CTP, cytidine 5′-triphosphate; DAG, diacylglycerol; DGDG, digalactosyldiacylglycerol; DGTS, diacylglyceryl-N,N,N-trimethylhomoserine; Etn, ethanolamine; FA, fatty acid; G3-P, glycerol 3-phosphate; Glc, glucose; Ins-3-P, inositol 3-phosphate; MGDG, monogalactosyldiacylglycerol; P-Etn, phosphoethanolamine; PE, phosphatidylethanolamine; PG, phosphatidylglycerol; PGP, phosphatidylglycerolphosphate; PI, phosphatidylinositol; PA, phosphatidic acid; Ser, serine; SQ, sulfoquinovose; SQDG, sulfoquinovosyldiacylglycerol; TAG, triacylglycerol; UDP, uridine 5′-diphosphate (modified with permission from Fig. 1 in Riekhof et al. 2005b).

Full size image

Precursors of thylakoid lipid biosynthesis in many plants are derived from two pathways (Mongrand et al. 1998), the plastid and the ER pathways. This two pathway hypothesis was formulated by Roughan and coworkers based on labeling experiments (Roughan et al. 1980; Roughan and Slack, 1982) and later confirmed by mutant analysis in Arabidopsis (Browse and Somerville, 1991; Wallis and Browse, 2002). Thylakoid lipid molecular species derived from either of the two pathways can be distinguished based on their fatty acid composition (Heinz and Roughan, 1983), and fluxes through the two pathways have been determined (Browse et al. 1986). While a large number of plant species have lost the ability to de novo assemble thylakoid lipids, such as the dominant galactoglycerolipids in the plastid, nearly all reported plant species derive at least a fraction of their thylakoid lipids from precursors assembled at the ER (Mongrand et al. 1998) requiring import of lipid precursor into the plastid. However, detailed compositional analysis of lipids and labeling studies suggest that in Chlamydomonas all thylakoid lipids are assembled de novo in the plastid (Giroud et al. 1988). Thus, it is possible that the lack of PC and the lack of trafficking of lipid precursors from the ER to the plastid in Chlamydomonas are related if PC is a critical intermediate in ER-to-plastid lipid trafficking. Because in the betaine lipid, DGTS, the head group moiety is ether-linked to the diacylglyceryl moiety (Fig. 2), Chlamydomonas might lack an enzyme to break this ether linkage. This ether linkage is more stable than the phosphate ester linkage in phosphoglycerolipids. Therefore, the conversion of DGTS into the galac-toglycerolipid precursor diacylglycerol might not be possible in Chlamydomonas.

Fig. 2.
figure 2

Structural similarity between phosphatidylcholine (PC) and betaine lipid (DGTS).

Full size image

Aside from the betaine lipid, Chlamydomonas and many other microalgae contain a rich set of polyunsaturated fatty acids (Fig. 3) not present in most seed plants, which will be discussed in detail below.

Fig. 3.
figure 3

Overview of acyl-chain desaturation in Chlamydomonas. Glycerolipid abbreviations are the same as those in Fig. 1. Fatty acids are referred to by the standard abbreviation “carbon atoms:double bonds.” Fatty acids at the sn-1 and sn-2 positions of the glyceryl moiety are indicated. Double bond positions within the fatty acid chain and/or common names of the fatty acids are as follows: 16:0, palmitic acid; 16:1Δ7, plamitoleic acid or 16:1Δ3t in plastidic PG; 16:2Δ7,10; 16:3Δ7,10,13; for 16:4 the double bond position is not known; 18:0, stearic acid; 18:1Δ9, oleic acid or in some lipids 18:1Δ11, vaccenic acid; α18:2Δ9,12, αlinoleic acid; α18:3Δ9,12,15, αlinoleic acid; i18:3Δ5,9,12, pinolenic acid; 18:4Δ5,9,12,15, coniferonic acid. The predominant molecular species of SQDG is 16:0/16:0 (modified with permission from Fig. 2 in Riekhof et al, 2005b).

Full size image

7.III Membrane Glycerolipid Biosynthesis

7.III.1 Fatty Acid Synthesis and Incorporation into Glycerolipids

De novo synthesis of fatty acids is localized to the chloroplast of Chlamydomonas cells (Sirevag and Levine, 1972). The common ancestral origin of green algal and seed plant plastids is particularly apparent in many homologous components of the fatty acid biosynthetic machinery. For example, bioinformatic analysis of the Chlamydomonas genome has identified genes for the full suite of enzymes required for the conversion of acetyl-CoA to acylated-acyl carrier protein (ACP), including the multimeric bacterial-type acetyl-CoA carboxylase and fatty acid synthase complexes (Riekhof et al. 2005b; Riekhof and Benning, 2008). These enzymes are essential for fatty acid biosynthesis in plants (and presumably algae), which predominantly produce 16:0-ACP and 18:1-ACP as the result of desaturation of 18:0-ACP by a soluble stearoyl-ACP Δ9 desaturase (Browse and Somerville, 1991; Shanklin and Somerville, 1991). In plants, fatty acids are incorporated directly into chloroplast membrane glycerolipids by step-wise acylation of glycerol 3-phosphate to form phosphatidic acid (sn1–18:1, sn2–16:0-PA) by glycerol 3-phosphate:acyl-ACP acyltransferase (GPAT), which shows substrate specificity for 18:1-ACP, and then by lysophosphatidate:acyl-ACP acyltransferase (LPAT, 16:0-ACP specific) (Kunst et al. 1988; Browse and Somerville, 1991; Murata and Tasaka, 1997; Kim and Huang, 2004; Xu et al. 2006).

Fatty acids are also assembled into glycer-olipids at the ER where isoforms of the plastid acyltransferases are present and have been characterized in Arabidopsis (Zheng et al. 2003; Kim et al. 2005). Putative orthologs of the plant GPAT and LPAT genes are annotated in the final Chlamydomonas genome draft (Riekhof et al. 2005b; Riekhof and Benning, 2008). Candidates for the plastid PA phosphatase, which produces the diacylglycerol precursors for the biosynthesis of non-phosphorus lipids in the plastid, have been recently identified in Arabidopsis (Nakamura et al. 2007). However, there is currently no good candidate in the Chlamydomonas genome predicted to encode this enzyme (Riekhof et al. 2005b; Riekhof and Benning, 2008).

7.III.2 Chloroplast Membrane Lipids

The overall structural organization of membranes in the chloroplast of Chlamydomonas and seed plant chloroplasts is essentially identical, where the inner and outer envelope membranes enclose an extensive thylakoid membrane system in which the photosynthetic apparatus is embedded. Genetic studies of Arabidopsis have identified many of the genes responsible for the biosynthesis of chloroplast membrane lipids, and have revealed the essential role that lipid composition plays in optimal photosynthetic function (Vijayan et al. 1998; Dörmann and Benning, 2002; Wallis and Browse, 2002; Benning and Ohta, 2005). Though the genetic study of glycerolipid metabolism in Chlamydomonas has far fewer documented examples, detailed biochemical analysis of this alga's lipid composition has long confirmed the presence of the major chloroplast membrane lipids found in land plants including the galactoglyc-erolipids mono- and digalactosyldiacylglycerol (MGDG and DGDG), sulfoquinovosyldiacylg-lycerol (SQDG), and the phosphoglycerolipid phosphatidylglycerol (PG) (Giroud et al. 1988).

As in plants, galactoglycerolipids are the predominant membrane glycerolipid class in Chlamydomonas, where they make up a majority of the chloroplast membrane lipids (Janero and Barrnett, 1981a; Giroud et al. 1988). In Ara-bidopsis, the bulk of galactolipid biosynthesis involves two enzymatic steps, whereby MGDG is formed from diacylglycerol (DAG) and UDP-galactose (UDP-Gal) substrates by MGDG syn-thase (MGD1), and DGDG is formed from MGDG and UDP-Gal by DGDG synthase (DGD1) (Ben-ning and Ohta, 2005). Genes encoding MGDG and DGDG synthases have been identified in the Chlamydomonas genome as orthologs of the Arabidopsis genes MGD1 and DGD1, respectively (Riekhof et al. 2005b; Riekhof and Ben-ning, 2008). MGD1 and DGD1 are single-copy genes in Chlamydomonas, which differs from that of the MGD1, 2, 3 and DGD1, 2 paralogs found in the Arabidopsis genome. Molecular analysis of the Arabidopsis MGD2, 3 and DGD2 genes has revealed their role in a galactolipid biosynthetic pathway that is transcriptionally induced during phosphate deprivation, and is proposed to provide galactolipids for extraplastidic membranes (Härtel et al. 2000; Kelly and Dörmann, 2002; Jouhet et al. 2004). The apparent lack of this induced galactolipid pathway in Chlamydomonas suggests a distinct lipid metabolic response to phosphate limitation, or a lack of need for one; however, to date the galactolipid biosynthetic genes of Chlamydomonas have not been studied in detail at the molecular level to test these hypotheses.

The sulfolipid sulfoquinovosyldiacylglycerol (SQDG) has long been studied in the context of its role in photosynthetic membranes, not only due to its prevalence in photosynthetic eukaryo-tes and prokaryotes, but also because of its association with photosynthetic pigment—protein complexes (Menke et al. 1976; Gounaris and Barber, 1985; Pick et al. 1985; Stroebel et al. 2003). However, the more recent discovery of SQDG and/or the genes and enzymes involved in SQDG biosynthesis in non-photosynthetic bacteria as summarized in (Cedergren and Holl-ingsworth, 1994; Benning et al. 2008), has clearly indicated that the role of sulfolipids is not limited to the function of photosynthetic membranes. The biosynthesis of SQDG is carried out in two enzymatic steps in Arabidopsis by SQD1, which catalyzes the formation of UDP-sulfoquinovose from UDP-Glc and sulfite, and SQD2, which transfers the sulfoquinovose moiety from UDP-sulfoquinovose to DAG, forming SQDG (Essigmann et al. 1998; Sanda et al. 2001; Yu et al. 2002). A single copy ortholog of SQD1 is present in Chlamydomonas, and two possible orthologs of Arabidopsis SQD2 are found in the genome (Yu et al. 2002; Riekhof et al. 2003). Recently, a Chlamydomonas mutant deleted in SQD1 (Δsqd1) and completely lacking sulfolipid has been studied (Riekhof et al. 2003). Phenotypic analysis of Δsqd1 revealed a reduced growth rate during phosphate-limiting conditions, under which the SQDG level was found to double in wild-type cells. This is similar to what has been observed in sulfolipid-deficient mutants in other organisms, such as Arabidopsis, which showed impaired growth after severe phosphate limitation (Yu et al. 2002), and in the photosyn-thetic purple bacterium Rhodobacter sphaeroides (Benning et al. 1993). In both Chlamydomonas and Arabidopsis, the increase in SQDG levels under phosphate-limiting conditions is accompanied by a decrease in PG, resulting in little net change in the amount of anionic glyceroli-pids. These results suggest a role for SQDG in partially replacing PG during phosphate limitation in order to maintain thylakoid membrane function (Riekhof et al. 2003). However, during sulfur (S) limitation a large decrease in SQDG and concomitant increase in PG has been observed in Chlamydomonas (Sugimoto et al. 2008), and SQDG was shown to be a major internal S-source for protein synthesis in the early phases of the S-starvation response (Sugi-moto et al. 2007).

In addition, Δsqd1 showed sensitivity to a photosystem II inhibitor under normal growth conditions (Riekhof et al. 2003). This is consistent with another Chlamydomonas SQDG-defi-cient mutant, hf-2, which was first discovered as a high chlorophyll fluorescence mutant, and was later found to be impaired in photosystem II stability and showed increased sensitivity to a PSII inhibitor, which could be partially restored by SQDG addition (Sato et al. 1995a, b; Minoda et al. 2002, 2003). However, whether the hf-2 mutant is impaired in growth during phosphate limitation has not been reported, nor has the exact molecular defect in this mutant been determined. Interestingly, detailed biochemical analysis of the Δsqd1 mutant also led to the discovery of the novel sulfolipid derivative, 2′-O-acyl-sulfoquinovosyldiacylglycerol (ASQD), which was also not produced in Δsqd1 (Riekhof et al. 2003). Due to the loss of both sulfolipids in Δsqd1, the specific roles played by SQDG and ASQD in Chlamydomonas and phenotypes associated with Δsqd1 can only be fully interpreted after the identification and characterization of the acyltransferase catalyzing ASQD production has been undertaken.

Phosphatidylglycerol (PG) is presumably the only major phospholipid component in thylakoid membranes of seed plants, and biochemical analysis of thylakoid lipid composition has confirmed this to be the case in Chlamydomonas (Janero and Barrnett, 1981b; Mendiola-Morgenthaler et al. 1985). While the gene encoding the final enzyme in PG biosynthesis, phosphatidylglycerolphos-phate (PGP) phosphatase, remains unknown in plants and algae (Beisson et al. 2003), the putative genes encoding the enzymes that catalyze the formation of the two intermediates, CDP-DAG synthetase and phosphatidylglycerolphosphate synthase, have been identified (Riekhof et al. 2005b; Riekhof and Benning, 2008), but not yet confirmed. While neither the single gene encoding the CDP-DAG synthetase or the two putative plastid paralogs encoding phosphatidylglycerol-phosphate synthase have been studied at the molecular/genetic level, PG deficient mutants, mf 1 and mf 2, have been isolated and studied in great biochemical detail (Garnier et al. 1987; Maroc et al. 1987; Garnier et al. 1990; Maanni et al. 1998; Dubertret et al. 2002; Pineau et al. 2004). The mf 1, 2 mutants were first isolated as low fluorescent strains lacking functional Photosys-tem II (PS II), as well as an oligomeric form of the light-harvesting chlorophyll antenna (CPII) (Maroc et al. 1987; Dubertret et al. 1994). It was also shown that both mf 1 and mf 2 contained approximately 30% of wild-type PG levels and lacked Δ3-trans-hexadecenoic acid (16:1Δ3trans [carbons:double bondsΔpositions]) (Maroc et al. 1987; Dubertret et al. 1994), a fatty acid that is specifically esterified to chloroplastic PG in both Arabidopsis and Chlamydomonas (Browse et al. 1985; Garnier et al. 1987; Giroud et al. 1988). Addition of a preparation of spinach leaf PG containing 16:1Δ3trans to mf-2 cells restored the ability to form oligomeric CP II, while 18:0 PG additions did not, and 16:0 PG did so only weakly so (Gar-nier et al. 1990; Dubertret et al. 1994). A PG-deficient mutant in Arabidopsis, pgp1, which is defective in the chloroplastid isoform of PGP synthase has been found to be photosyntheti-cally impaired with decreased quantum yield through PSII, but did not lack 16:1Δ3trans PG (Xu et al. 2002; Hagio et al. 2002; Babiychuk et al. 2003). Similarly, two Synechocystis PG deficient mutants showed altered PSII activity and required exogenous PG for phototropic growth (Hagio et al. 2000; Sato et al. 2000).

Taken together with the contrasting findings from analyses of the Arabidopsis fad4 mutant, which lacks 16:1Δ3trans, but is otherwise not affected in chloroplast PG content and also shows no apparent photosynthetic defects (Browse et al. 1985), it can currently only be concluded that in general PG plays an important role in photosynthetic membrane biogenesis and function, and it seems possible that the 16:1Δ3trans PG form could be essential in some organisms (e.g., Chlamydomonas), but is of conditional importance or dispensable in others. It is certain however, that the elucidation of the exact molecular defects in the Chlamydomonas mf 1, 2 mutants, and the identification of the genes encoding FAD4 activity as well as the elusive plant/algal PGP phosphatase that catalyzes the final step in PG biosynthesis, will be prerequisite to gaining a better understanding of the roles PG plays in the photosynthetic membranes in various species.

7.III.3 Extrachloroplastic Membrane Lipid Metabolism

In eukaryotic photoautotrophs the bulk of extra-plastidic membrane glycerolipids is assembled in the ER from acyl-CoA thioesters, which are formed from free fatty acids after their liberation from acyl-ACPs in the plastid (see Fig. 1). While other extraplastidic sites for lipid synthesis are known (e.g., mitochondria), the ER localized pathway is predominant, and in most plants the ER lipid assembly pathway significantly contributes to thylakoid membrane biogenesis. As such, a discussion of the analogous pathways in Chlamydomonas is merited. As mentioned above, Chlamydomonas lacks the capability for PC biosynthesis (Giroud et al. 1988) and genes predicted to encode enzymes involved in PC biosynthesis are not present in its genome (Riekhof et al. 2005b; Riekhof and Benning, 2008). Instead, it contains the non-phosphorous zwitterionic betaine lipid DGTS (Fig. 2) in its membranes (Eichen-berger and Boschetti, 1977; Janero and Barrnett, 1982), which has similar biophysical properties to PC (Sato and Murata, 1991). DGTS has also been found in other algal species, e.g. (Eichen-berger, 1982), prokaryotes like the purple bacterium Rhodobacter sphaeroides, e.g. (Benning et al. 1995; Hofmann and Eichenberger, 1996), and in non-seed plants, such as ferns, e.g. (Sato and Furuya, 1983; Eichenberger, 1993), but appears to be absent in seed plants. Labeling studies suggest that the biosynthesis of DGTS is similar in all organisms studied (Sato, 1988, 1991; Sato and Kato, 1988; Vogel and Eichenberger, 1992; Hof-mann and Eichenberger, 1996). It begins with the transfer of the 3-amino 3-carboxypropyl residue from S-adenosylmethionine (AdoMet) to DAG catalyzed by AdoMet:DAG 3-amino-3-carboxy-propyltransferase activity followed by successive methylation of the amino group by an AdoMet-dependent N-methyltransferase (Fig. 1). The two genes encoding these catalytic activities, btaA and btaB, were first identified in R. sphaeroides (Klug and Benning, 2001; Riekhof et al. 2005a).

More recently, a single gene sufficient for DGTS biosynthesis in Chlamydomonas, Bta1, was identified in the genome (Riekhof et al. 2005b). The encoded protein Bta1 is similar in its N-terminal domain to bacterial BtaB and in its C-terminal domain to BtaA and the predicted catalytic function of each Bta1 domain was confirmed by mutagenesis (Riekhof et al. 2005b). The bifunctionality observed in Bta1 as a fusion of two prokaryotic enzymes active in the same pathway into a single polypeptide is a common theme in plants (Moore, 2004), and could represent an improvement in DGTS biosynthesis by eliminating the need for coordinated regulation of two independent gene products or permitting substrate channeling. In addition, the presence of DGTS and concomitant lack of PC are perhaps related to the absence of additional galactolipid biosynthetic pathways seen in seed plants, e.g., MGD2, MGD3 and DGD2 in Arabidopsis (Ben-ning and Ohta, 2005). As this alternative galac-tolipid pathway is believed to be involved in replacing phospholipids in extraplastidic membranes with DGDG during phosphate deprivation, the constitutive replacement of PC with the non-phosphorous DGTS has perhaps obviated the need to replace PC by extraplastidic DGDG. Interestingly, DGTS was tentatively identified as a component of purified Chlamydomonas thy-lakoids (Janero and Barrnett, 1981b, 1982) and chloroplast envelope membranes (Mendiola-Morgenthaler et al. 1985). However, whether DGTS is indeed present in the chloroplast membranes of Chlamydomonas or plays a specific function in thylakoids remains to be confirmed.

Phosphatidylserine (PS) is a minor component of extraplastidic membranes of plants and can be decarboxylated to phosphatidylethanolamine (PE) in plants and other organisms (Vance and Steen-bergen, 2005; Nerlich et al. 2007). However, PS is absent in Chlamydomonas membranes (Gir-oud et al. 1988) and genes encoding the phos-phatidylserine synthase or relevant phospholipid base exchange enzymes were not detected in the genome (Riekhof et al. 2005b; Riekhof and Ben-ning, 2008). As such, the biosynthesis of PE, which is a known component of extraplastidic membranes in Chlamydomonas, is likely only carried out by a single pathway (Fig. 1). Genes encoding a serine decarboxylase which produces ethanolamine, an ethanolamine kinase and CTP:phosphoethanolamine cytidylyltransferase the combined activities of which produce CDP-ethanolamine, and a CDP-ethanolamine:DAG ethanolamine phosphotransferase, which produces PE, have been identified in the genome (Riekhof et al. 2005b; Riekhof and Benning, 2008). Recently, the gene encoding the CTP: phospho-ethanolamine cytidylyltransferase has been characterized by heterologous expression in Escherichia coli, and the expression of the respective gene was found to be up-regulated during the reflagella-tion of Chlamydomonas cells (Yang et al. 2004). Phosphatidylinositol (PI) is a minor component of Chlamydomonas membranes, and genes required for its biosynthesis, including inositol-3-phos-phate synthase and CDP-DAG:inositol phospho-transferase, are present in the genome (Riekhof et al. 2005b; Riekhof and Benning, 2008). The PI biosynthetic enzymes of Chlamydomonas have been studied at the biochemical level, and PI synthesis was found to be highest in the microsomal fraction, suggesting its association with the ER (Blouin et al. 2003).

Extraplastidic PG biosynthesis is known to be associated with both the ER and mitochondria, and three isoforms of phosphatidylglycerolphos-phate synthase are encoded in the genome, each with differential targeting prediction probabilities for subcellular localization to the mitochondria, chloroplast or cytosol (Riekhof et al. 2005b; Riekhof and Benning, 2008). However, a detailed analysis of these proteins and their respective genes is not yet available in Chlamydomonas. An extraplastidic candidate for CDP-DAG synthase, which provides one of the substrates for phos-phatidylglycerolphosphate synthase, has been identified in the genome (Riekhof et al. 2005b). However, as in the case of chloroplast PG biosynthesis, no gene encoding an extraplastidic phosphatidylglycerolphosphate phosphatase is currently known.

7.IV Fatty Acid Desaturation

Biochemical studies in Chlamydomonas have indicated that further desaturation of 16:0 and 18:1 acyl groups occurs after the production of the major glycerolipids, in a manner similar to plants (Giroud et al. 1988). Furthermore, the Chlamydomonas fatty acid profile is known to change markedly in response to various environmental conditions, including CO2 concentration, as well as nitrogen and phosphorous limitation (Tsuzuki et al. 1990; Weers and Gulati, 1997). The elucidation of fatty acid desaturase (FAD) genes in Arabidopsis is a classic example of the power of genetic and molecular biological approaches in solving biological problems, which prove to be largely intractable through a strictly biochemical approach (Browse and Somerville, 1991; Wallis and Browse, 2002). The identification of many of the Chlamydomonas desaturase gene candidates by their similarity to Arabidopsis orthologs, combined with a handful of studies of Chlamydomonas desaturase mutants and characterization of cloned FAD genes, has provided a reasonable picture of the fatty acid desaturation pathways in this alga (Fig. 3).

Putative orthologs for the plastidic desaturases encoded by Arabidopsis FAD5 (MGDG palmitate-Δ7-desaturase), FAD6 (ω6-desaturase), and FAD7 or FAD8 (encoding ω3-desaturase isozymes) are present in the Chlamydomonas genome (Riekhof et al. 2005b; Riekhof and Benning, 2008). Of these, only the Chlamydomonas FAD6 gene has been studied to date at the molecular/genetic level (Sato et al. 1995b, 1997). The hf-9 mutant was first isolated by its high chlorophyll fluorescence phenotype, and detailed lipid analysis revealed an apparent defect in ω6-desaturase activity as it showed marked decreases in both 16- and 18-carbon polyunsaturated fatty acyl groups, with concomitant increases in 16:1Δ7 and 18:1Δ9 (Sato et al. 1995b). The hf-9 mutant had an increased doubling time and showed reduced photosynthetic O2 evolution as well as an altered chloroplast ultrastructure (Sato et al. 1995b). The Chlamydomonas FAD6 gene (first described as DES6) was subsequently cloned and found to be highly similar to cyanobacterial Δ12- and seed plant ω6-desaturases, and was also shown to complement the hf-9 mutant desatura-tion defects (Sato et al. 1997). However, it did not restore the photosynthetic defects, suggesting that these phenotypes arose from a mutation in another gene (or possibly from multiple loci). As such, the roles of polyunsaturated fatty acids (PUFAs) in the assembly or maintenance of optimally functioning photosynthetic membranes in Chlamydomonas cannot be easily deduced from analysis of hf-9. The function of PUFAs in this regard have been studied and found to differ in mutants of both plants and cyanobacteria. The Arabidopsis fad6 fad2 double mutant, which lacks both plastidic and ER ω6-desaturases exhibited severe growth and photosynthetic defects (McConn and Browse, 1998). In contrast, a mutant of Synechocystis sp. PC 6,803 lacking PUFAs had no observable photosynthetic defects under normal growth conditions (Gombos et al. 1992). Thus, it still remains to be determined whether the importance of PUFAs in photosyn-thetic membrane function in Chlamydomonas is more similar to that of seed plants or cyanobacteria. Other plastidic desaturases still have no gene candidates in Chlamydomonas. As noted above, the desaturase producing 16:1Δ3trans specifically on plastidic PG is still not identified at the molecular level in plants or algae, although mutants lacking this fatty acid have been obtained in Arabidopsis and Chlamydomonas (Browse et al. 1985; Maroc et al. 1987). Likewise, a gene encoding Δ4 desat-urase activity, which is specific for MGDG-ester-ified 16-carbon acyl groups based on biochemical studies (Giroud et al. 1988), and presumably produces both the 16:3Δ4,7,10 and 16:4Δ4,7,10,13 found in Chlamydomonas is currently unidentified.

The extraplastidic ω6- and ω3-desaturases, which produce 18:2Δ9,12 and 18:3Δ9,12,15, are encoded by FAD2 and FAD3, respectively, in Arabidop-sis, and putative orthologs of these genes have been identified in the Chlamydomonas genome (Riekhof et al. 2005b; Riekhof and Benning, 2008). The extraplastidic Chlamydomonas lipids DGTS and PE have been shown to contain significant amounts of 18:3Δ5,9,12 and 18:4Δ5,9,12,15 esteri-fied to the respective sn-2 positions of the glycerol back bone (Giroud et al. 1988); these Δ5-unsatu-rated fatty acids are also found in gymnosperms (Mongrand et al. 2001; Wolff and Christie, 2002). Recently a Chlamydomonas gene, CrDES, encoding a “front-end” type Δ5-desaturase was identified by a bioinformatics approach through its similarity to a known Δ5-desaturase from the liverwort Marchantia polymorpha (Kajikawa et al. 2006). Heterologous expression of CrDES in Pichia pastoris and analysis of desaturase activity indicated that while the primary substrates were 18:2Δ9,12 and 18:3Δ9,12,15, low but detectable levels of endogenous 18:1Δ9 desaturation were also observed (Kajikawa et al. 2006). Transgenic tobacco plants constitutively expressing the CrDES gene exhibited strikingly high levels of 18:3Δ5,9,12 and 18:4Δ5,9,12,15 (which are normally absent), with the highest combined yield reaching ~ 45% of leaf total fatty acids, and no apparent morphological phenotypes (Kajikawa et al. 2006). The biological roles of these Δ5-unsaturated fatty acids in the organisms which produce them are largely unknown, and the identification of the CrDES gene responsible for 18:3Δ5,9,12 and 18:4Δ5,9,12,15 production in Chlamydomonas will allow for this gene to be targeted for suppression through RNAi technology.

7.V Neutral Lipid Metabolism

To date, little research has been done on neutral glycerolipid synthesis in Chlamydomonas. However, there is an increasing focus on oil production in microalgae due to its potential role as a feedstock for biodiesel or jet fuels (Hu et al. 2008), and as a source of commercial oils and fatty acids (Spolaore et al. 2006). Beginning in the late 1970s, the Department of Energy initiated a two decade-spanning research effort, the Aquatic Species Program, to investigate the possibility of obtaining biodiesel from microalgae (Sheehan et al. 1998). Researchers screened numerous algal strains for oil production, and found many that accumulated oil up to 75% dry weight (Benamotz and Tornabene, 1985; Bigogno et al. 2002; Chisti, 2007). Chlamydomonas also accumulates oil in the form of triacylglycerol under certain conditions, as already mentioned above (Weers and Gulati, 1997).

In most algae species, oil production is triggered by environmental stress, suggesting that triacylglycerol plays a role in microalgae beyond energy storage. One of the main stresses investigated is nutrient deprivation, with nitrogen deprivation being the most common condition used. Growing the chlorophyte Neochloris oleo-abundans in growth media limited for nitrogen resulted in lipids being accumulated up to 56% of the total dry weight, with 80% of that being triacylglycerol (TAG) (Tornabene et al. 1983). The eustigmatophyte Nannochloropsis gave similar results (Suen et al. 1987) as did the chlo-rophyte Parietochloris incise (Merzlyak et al. 2007). Growing the chlorophyte Haematococcus pluvialis in nitrogen-free medium led not only to an increase in total lipid content, but also to a change in fatty acid composition, with an increase in oleic acid (Zhekisheva et al. 2002). Other nutrient deficiencies can trigger lipid accumulation. For example, silicon deficiency leads to increased lipid content (mainly in the form of TAG) in the diatom Cyclotella cryptica (Roessler, 1988). Phosphate limitation leads to an increase in TAG and overall lipid levels in some green algae, and to a decrease in lipid content in others (Khozin-Goldberg and Cohen, 2006).

Other factors, such as light, temperature and growth phase also affect oil accumulation in microalgae. The effect of temperature on lipid accumulation varies between strains, with some reporting increases in lipid levels, and others decreases (Richardson et al. 1983; Dempster and Sommerfeld, 1998; de Swaaf et al. 1999). Inhibition of cell cycle in the chlorophyte Chlorella by high pH leads to an accumulation of TAG similar to that due to nutrient deprivation, suggesting that environmental stress may indirectly trigger TAG synthesis by inhibiting growth, rather than directly (Guckert and Cooksey, 1990). High light intensity has been shown to increase the ratio of TAG to total lipids, although the total lipid level can remain the same or decrease (Zhekisheva et al. 2002; Khotimchenko and Yakovleva, 2004, 2005). In the chlorophyte Dunaliella bardawil, the accumulation of TAG under high-light stress is linked to an accumulation of β-carotene, which suggests that TAG accumulation may help to protect the chloroplasts from photooxidative damage (Benamotz et al. 1989; Rabbani et al. 1998).

The biochemistry of TAGs in microalgae has not been studied, including that in the genetic model Chlamydomonas. Previous research has indicated that lipid synthesis in Chlamydomonas is homologous to that in plants (Fig. 1), and possibly simpler (Riekhof et al. 2005b; Riekhof and Benning, 2008). Therefore, it is likely that many general aspects of TAG synthesis in Chlamydomonas follow that of plants. One common path for TAG synthesis in seed plants is the Kennedy pathway, which involves the step-wise addition of fatty-acyl groups to a glycerol-3-phosphate to form PA, which is converted to DAG by phosphatidic acid phosphatase; DAG is further acylated by diacylglycerol acyltransferase to form TAG, as discussed further below (Kennedy, 1961). Two genes encoding putative extraplastidic phosphatidic acid phosphatases have been identified in the Chlamydomonas genome, but have not yet been studied in molecular detail (Riekhof et al. 2005b). Several studies in many different plant species have also indicated that DAG derived from the PC pool also contributes substantially to TAG biosynthesis (Ohlrogge and Browse, 1995). Clearly, PC plays no role in TAG biosynthesis in Chlamydomonas, and its role in plants is not universal, as studies of mesocarp microsomes in avocado indicated that only the Kennedy pathway was active (Stobart and Stymne, 1985). It may be possible that the assumed functional analog of PC in Chlamyd-omonas, DGTS, is an intermediate in the biosynthesis of TAG. However, no biochemical studies to determine this have been undertaken to date, and utilization of the DGTS pool to provide DAG precursors would require an as yet unidentified enzyme to remove the ether-linked trimethylhomoserine head group. Regardless of the precursors used in forming DAG, the final step is catalyzed by diacylglycerol acyltrans-ferases, or DGATs, which transfer a fatty acid from acyl-CoA to diacylglycerol. DGATs have been isolated and characterized from several plant species, including Arabidopsis (Routaboul et al., 1999; Zou et al. 1999; Hobbs et al. 1999), maize (Zheng et al. 2008) and castor beans (Kroon et al. 2006). The Chlamydomonas genome contains a number of putative DGAT isoforms yet to be studied in molecular detail (Riekhof and Benning, 2008; R. Miller and C. Benning, unpublished, 2009).

Chlamydomonas may not only utilize the Kennedy pathway for TAG synthesis. Indeed, an alternate pathway for TAG synthesis involves phospholipid: diacylglycerol acyltransferases, or PDATs, to generate TAG using a phospholipid as a fatty acid donor, rather than acyl-CoA. PDATs have also been found in plants (Dahlqvist et al. 2000) and represent a possible second type of enzyme that is also present in Chlamydomonas (Riekhof and Benning, 2008). Given the induction of TAG biosynthesis by different stresses, it is likely that the mechanism for the regulation of TAG synthesis differs in Chlamydomonas from that in seed plants, which often produce oil during a specific phase of their life cycle and in specialized tissues.

7.VI Perspectives

The study of lipid metabolism in microalgae is experiencing a renaissance due to their potential for the production of large quantities of biomass in general, and specifically due to their ability to accumulate neutral lipids following nutrient deprivation. However, during the past decade much of basic research on lipid metabolism in photo-synthetic organisms was focused on cyanobacte-ria and seed plants, in particular, the genetic and genomic model Arabidopsis. Chlamydomonas has been developed over the years as an excellent genetic model as well, however, not necessarily for the study of lipid metabolism. The availability of the Chlamydomonas genomic sequence (Merchant et al. 2007) has made the application of knowledge on well studied lipid metabolism in seed plants to this model alga relatively facile using comparative genomics. The result is a testable hypothesis of lipid metabolism in Chlamydomonas (Fig. 1) based on genome annotation (Riekhof et al. 2005b; Riekhof and Benning, 2008), which provides a wealth of opportunities to students of lipid metabolism. Those researchers interested in studying basic lipid metabolism in photosynthetic organisms might wonder what novel concepts research on Chlamydomonas could contribute beyond research on Arabidopsis. The answer lies in the fact that lipid metabolism in Chlamydomonas appears simpler and in some aspects drastically different from that in seed plants as discussed in detail above. The reduced redundancy in Chlamydomonas versus Arabidopsis permits testing of hypotheses on the function of parallel pathways, e.g., galactoglycerolipid biosynthesis (Härtel et al. 2000), present in Arabidop-sis. Moreover, the unicellular organization of Chlamydomonas and its resulting lifestyle requires completely different input for the regulation of TAG biosynthesis (Hu et al. 2008) than the developmental regulation of storage lipid metabolism in developing seeds of Arabidop-sis (Santos-Mendoza et al. 2008). The absence of PC in Chlamydomonas challenges concepts about the role of this lipid as a central metabolite in lipid trafficking and lipid modification in plants. How widespread the replacement of PC by the betaine lipid DGTS in microalgae is not known at this time, but it would be important to explore, if Chlamydomonas is to become the model for the engineering of microalgal biofuel-producing strains.