genomic and biochemical analysis of lipid biosynthesis in ... · istry to reveal similarities and...

EUKARYOTIC CELL, June 2007, p. 1006–1017 Vol. 6, No. 61535-9778/07/$08.00�0 doi:10.1128/EC.00393-06Copyright © 2007, American Society for Microbiology. All Rights Reserved.

Genomic and Biochemical Analysis of Lipid Biosynthesis in theUnicellular Rhodophyte Cyanidioschyzon merolae: Lack of

a Plastidic Desaturation Pathway Results in theCoupled Pathway of Galactolipid Synthesis�†

Naoki Sato* and Takashi MoriyamaDepartment of Life Sciences, Graduate School of Arts and Sciences, The University of Tokyo,

Komaba 3-8-1, Meguro-ku, Tokyo 153-8902, Japan

Received 8 December 2006/Accepted 20 March 2007

The acyl lipids making up the plastid membranes in plants and algae are highly enriched in polyunsaturatedfatty acids and are synthesized by two distinct pathways, known as the prokaryotic and eukaryotic pathways,which are located within the plastids and the endoplasmic reticulum, respectively. Here we report the resultsof biochemical as well as genomic analyses of lipids and fatty acids in the unicellular rhodophyte Cyan-idioschyzon merolae. All of the glycerolipids usually found in photosynthetic algae were found, such as mono- anddigalactosyl diacylglycerol, sulfolipid, phosphatidylglycerol, phosphatidylcholine, phosphatidylethanolamine,and phosphatidylinositol. However, the fatty acid composition was extremely simple. Only palmitic, stearic,oleic, and linoleic acids were found as major acids. In addition, 3-trans-hexadecanoic acid was found as a veryminor component in phosphatidylglycerol. Unlike the case for most other photosynthetic eukaryotes, polyenoicfatty acids having three or more double bonds were not detected. These results suggest that polyunsaturatedfatty acids are not necessary for photosynthesis in eukaryotes. Genomic analysis suggested that C. merolaelacks acyl lipid desaturases of cyanobacterial origin as well as stearoyl acyl carrier protein desaturase, both ofwhich are major desaturases in plants and green algae. The results of labeling experiments with radioactiveacetate showed that the desaturation leading to linoleic acid synthesis occurs on phosphatidylcholine locatedoutside the plastids. Monogalactosyl diacylglycerol is therefore synthesized by the coupled pathway, usingplastid-derived palmitic acid and endoplasmic reticulum-derived linoleic acid. These results highlight essentialdifferences in lipid biosynthetic pathways between the red algae and the green lineage, which includes plantsand green algae.

The acyl lipids that make up the plastid membranes arehighly enriched in polyunsaturated fatty acids in most plantsand algae. In flowering plants, linoleic (18:2) and linolenic(18:3) acids are the most common unsaturated acids, whereasin many algae, highly unsaturated long-chain acids, such asarachidonic (20:4), eicosapentaenoic (20:5), and docosahexae-noic (22:6) acids, are found as major fatty acid components,depending on the species (4, 15, 17, 38, 50). (Fatty acids areexpressed by a combination of the number of carbon atoms [X]and the number of double bonds [Y], such as X:Y. The posi-tions of double bonds are specified in parentheses.) Unsatur-ated fatty acids, especially polyunsaturated ones, are importantin maintaining membrane fluidity during the cold acclimationof photosynthesis (37). Linolenic acid was also found to beimportant in tolerance to high temperatures (32). The plastidlipids consist of two types of molecular species, namely, 1-C18-2-C16 species, or prokaryotic molecular species, and 1-C18-2-C18 species, or eukaryotic species. The relative abundance ofthese two types of molecular species varies with plants and

algae. In flowering plants, the prokaryotic and eukaryotic typesof lipid molecular species are synthesized by two distinct path-ways, known as the prokaryotic and eukaryotic pathways,which are located within the plastids and the endoplasmicreticulum (ER), respectively (for a review, see reference 10).However, lipid biosynthesis in red algae, which constitute an-other major lineage of photosynthetic eukaryotes (11, 30, 42,56), remains largely unknown.

Cyanidioschyzon merolae, a unicellular rhodophyte isolatedfrom an Italian hot spring, has a very simple cell structureconsisting of one mitochondrion, plastid, and microbody percell (25). Its normal habitat is warm (up to 50°C) and acidic(pH 1.5 to 2.5) water containing sulfuric acid. The size of thenuclear genome is 16.5 Mbp (28). The cell proliferates bybinary fission. These characteristics, as well as phylogeneticanalyses (34), suggested that C. merolae is one of the mostprimitive red algae, probably diverged from near the root ofthe red lineage. The red lineage includes red algae, whereas thegreen lineage includes green algae and land plants (42). Thesingle origin of plastids in the red and green lineages is be-lieved to be highly probable (27, 42), and the single origin ofplastid-harboring cells in these two lineages is gaining support-ing evidence (30, 34). In addition, the chromists (brown algae,diatoms, cryptophytes, etc.) are believed to originate from sec-ondary endosymbiosis by an ancestral red algal cell (11, 42, 56).C. merolae is therefore a good target of comparative biochem-

* Corresponding author. Mailing address: Department of Life Sci-ences, Graduate School of Arts and Sciences, University of Tokyo,3-8-1 Komaba, Meguro-ku, Tokyo 153-8902, Japan. Phone: 81-3-54546631. Fax: 81-3-54546998. E-mail: [email protected].

† Supplemental material for this article may be found at http://ec.asm.org/.

� Published ahead of print on 6 April 2007.

1006

on August 16, 2019 by guest

http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

istry to reveal similarities and differences in the red and greenlineages.

There is a short report on the total fatty acids of C. merolae(31). Among thermoacidophilic red algae, Cyanidium cal-darium and C. merolae contained no detectable �-linolenic acid(18:3), while this acid was abundant in Pleurococcus sulfurarius(currently called Galdieria sulfuraria) at a low temperature.There was a report on �-18:3 in Cyanidium caldarium (23), butthe current understanding is that several different algae werecalled Cyanidium in the past. No further analysis of the com-position and biosynthesis of C. merolae has been attemptedsince the paper by Moretti and Nazzaro (31).

Apart from experimental analyses, one postgenomic study isto find all possible candidate enzymes of a metabolic pathway.For Arabidopsis thaliana, an attempt to make functional an-notations for all proteins involved in lipid metabolism is inprogress (7). For Chlamydomonas reinhardtii, draft sequencedata were used to predict proteins that might be involved inlipid biosynthesis (36). The functional annotation of C. merolaehas also been started (28). We have been trying to compare thegenome contents of A. thaliana and C. merolae as well ascyanobacteria, and many proteins that are conserved in all ofthese photosynthetic organisms are being identified (45). Com-parative genomics of these photosynthetic organisms are nowconveniently analyzed through a web interface called theGclust server (43; http://gclust.c.u-tokyo.ac.jp/). Based on suchinformatics, we are now able to predict probable proteins thatare involved in acyl lipid metabolism in C. merolae.

In this report, we try to find distinct features of the redlineage with respect to lipid biosynthesis, using C. merolae as amodel organism. We present results on the analysis of lipidsand fatty acids in C. merolae. Next, lipid biosynthetic pathwaysof this alga are summarized based on the genomic data. Theintracellular localization of some key enzymes was confirmedby green fluorescent protein (GFP) experiments. The origin ofthe minimal set of desaturases of this alga was established bycomprehensive phylogenetic analysis. Finally, tracer experi-ments were conducted to validate the lipid biosynthetic path-way estimated by the genomic analysis. The coupled pathwayfor the synthesis of galactolipids is proposed, involving plastid-derived palmitic acid and extraplastidically synthesized linoleicacid. All of these data indicate distinct differences between thered and green lineages in lipid biosynthesis, in spite of theirmonophyletic origin.

MATERIALS AND METHODS

Growth of organism. Cells of Cyanidioschyzon merolae strain 10D (55) weregrown under continuous white light (about 20 �E m�2 s�1) in the AA mediumdescribed by Allen (2), with aeration in 1% CO2, at 38°C or 25°C.

Analysis of lipids and fatty acids. All analytical methods used were essentiallythe same as those described in previous publications (44, 48). Briefly, lipids wereextracted from the cells by the method of Bligh and Dyer (8) and were separatedby two-dimensional thin-layer chromatography (TLC). Lipids were quantified bymeasuring the amounts of fatty acids, determined as their methyl esters, by gaschromatography. A fused silica capillary column (0.25-mm internal diameter by50 m) coated with SS-10 (equivalent to Silar 10C; Sinwa Kako, Kyoto, Japan) wasused. The following temperature program was used: 0.5 min at 180°C, a linearincrease to 230°C at a rate of 3°C min�1, and then 10 min at 230°C. Under theseconditions, most commonly occurring isomers of fatty acid methyl esters wereclearly separated. Fatty acid methyl esters from total lipids of Adiantum capillus-veneris (44) and those from monogalactosyl diacylglycerol (MGDG) of Anabaenavariabilis (46) were used as references. The positional distribution of fatty acids

within individual classes of lipids (including phospholipids) was analyzed byspecific hydrolysis of the C-1 acyl ester linkage with the lipase from Rhizopusdelemar (16) or the C-2 acyl ester linkage with phospholipase A2.

Radiolabeling of lipids. C. merolae cells (25-ml culture) that had been grownat 38°C were incubated with [2-14C]acetate (2.0 MBq) at 38°C for 1 h in the light,with vigorous shaking, in a tightly closed 100-ml flask. Unlabeled acetate (3 mM)was then added, and the cells were harvested by centrifugation. They werewashed once with fresh medium and then resuspended in fresh medium. Thecells were allowed to grow under normal growth conditions for 20 h. Aliquotswere withdrawn at 0, 2, 6, 10, and 20 h, and lipids were extracted. Lipids wereseparated by two-dimensional TLC, and then the lipid spots were detected withprimuline under UV light. The analysis of radioactive lipids was performedessentially as described previously (39). Radioactivity was located by autoradiog-raphy. Radioactive lipid spots were scraped off, and the radioactivity was mea-sured by liquid scintillation counting.

For detailed analysis, MGDG, digalactosyl diacylglycerol (DGDG), and phos-phatidylcholine (PC) were recovered from the TLC plate. Lipid molecular spe-cies were analyzed by argentation TLC (48), and radioactivity was detected byautoradiography. A precoated silica gel plate (Merck) was impregnated withAgNO3 by immersing it in 5% AgNO3 in acetonitrile for 30 min and then wasdried at 60°C for 30 min. The developing solvents were acetone-benzene-water(90:30:8, by volume) for MGDG and chloroform-methanol-water (60:30:5, byvolume) for DGDG and PC. For fatty acid analysis of individual lipid classes, theisolated lipids were subjected to methanolysis. The resultant fatty acid methylesters were analyzed by reversed-phase argentation TLC (26), a technique re-cently developed for the analysis of small amounts of radioactive fatty acids.An RP-18 HPTLC plate (5 by 10 cm; Merck) was used. The developingsolvent was 10% AgNO3 in acetonitrile–1,4-dioxane–acetic acid (80:20:1, byvolume). This TLC method clearly resolves 18:1 and 16:0, which comigrate inordinary argentation TLC.

Incorporation of radioactive galactose was performed by incubating isolatedplastids of C. merolae, which were prepared according to a published protocol(51), with modifications (T. Moriyama, K. Terasawa, M. Fujiwara, and N. Sato,unpublished data), with 37 kBq UDP-[U-14C]galactose (GE Healthcare/Amer-sham) in 400 �l plastid isolation medium at 38°C for 1 h. Lipids were extractedand separated by two-dimensional TLC. MGDG and DGDG were recovered,and the molecular species were analyzed by argentation TLC.

Genomic data and computational sequence analysis. Genomic data for C.merolae were generated and annotated by the Cyanidioschyzon Genome Project(28), in which genes related to lipid metabolism were estimated by BLAST2 (3)searches, with known genes as queries. The sequences of the seed genes wereretrieved from GenomeNet (ftp://ftp.genome.ad.jp/). The Gclust database, re-cently made publicly accessible in the Gclust server (43), was also used to findphylogenetically conserved proteins. Sequence manipulation was performed withthe SISEQ package, version 1.30 (40). The sequence alignment of desaturaseswas prepared by the Clustal X program (12) after trimming of poorly conservedN and C termini. Phylogenetic analysis was done with MEGA2 software, version2.1 (24), PAUP software, version 4 beta 10 (Sinauer Associates, Sunderland,MA), and the MOLPHY package, version 2 beta 3 (1).

Targeting of GFP fusion proteins. DNA constructs consisting of the 35Spromoter, a 5� part of the putative �12 desaturase gene, and the NOS terminatorwere made by successive PCRs. The sGFP plasmid (13) was used as the templatefor the 5� and the 3� parts, while the PCR fragment corresponding to variousparts of the C. merolae CMK291C gene was obtained by using the C. merolaegenome as a template. The following three constructs were made: Met1-GFP,Met2-GFP, and Met3-GFP, containing residues 1 to 531, 202 to 531, and 364 to531, respectively (numbers refer to the nucleotide count beginning from the mostupstream ATG). The 5� part includes the 35S promoter until the multiple cloningsite, while the 3� part includes the NOS terminator sequence beginning from themultiple cloning site in the sGFP plasmid. The constructs were introduced intoa scaly leaf of an onion bulb by particle bombardment using a PDS-1000/Heparticle delivery system (Bio-Rad). Rupture disks for 1,100 lb/in2 and tungstenparticles with a diameter of 1.1 �m were used. After 24 h of continued growth,the epidermis was peeled and examined under a fluorescence microscope(Olympus model BX-60) with an IB filter cube.

RESULTS

Compositions of lipids and fatty acids. We started by ana-lyzing the compositions of lipids and fatty acids in C. merolaecells (Table 1). Major lipids included common chloroplast lip-ids, such as MGDG, DGDG, and sulfoquinovosyl diacylglyc-

VOL. 6, 2007 CYANIDIOSCHYZON LIPIDS 1007


http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

erol (SQDG), as well as common phospholipids, namely, PG,phosphatidylcholine (PC), phosphatidylethanolamine (PE),and phosphatidylinositol (PI). The level of MGDG was higherat 25°C than at 38°C, whereas the level of PC was lower.Cardiolipin (diphosphatidylglycerol) and phosphatidylserinewere not detected by the methods employed (see Discussion).

Figure 1 shows representative results of capillary gas chro-matographic analysis of fatty acid methyl esters. The compo-sition of fatty acids in the total lipids (Fig. 1A) was strikinglysimple. Palmitic (16:0; peak 1) and linoleic (18:2; peak 11)acids were the major acids. Stearic (18:0; peak 8) and oleic(18:1; peak 9) acids were also found as abundant components.PG contained the unusual fatty acid 3-trans-hexadecanoic acid(3-trans-16:1; peak 3), as do the PGs from all plants and algae.It is worth noting that no peak corresponding to either 18:3(9,12,15) (Fig. 1A, arrow) or 18:3(6,9,12) was detected in theanalysis of fatty acids of C. merolae (Fig. 1A), even at a lowtemperature (Table 1). No peak corresponding to 18:3 isomerswas detected in the analysis of any individual classes of lipids.This was also confirmed by direct methanolysis of the driedwhole cells. Long-chain fatty acids, such as 20:4 and 22:6, werenot detected.

Table 1 summarizes the fatty acid compositions of individualclasses of lipids. Cells grown at 38°C and 25°C were analyzed.All classes of lipids contained 16:0 and 18:2 as the major acids,although 16:0 accumulated to about 70% in SQDG. 18:0 and18:1 were also abundant in PE, PC, and PI. 16:1(9), which iscommonly found in lipids of cyanobacteria (Fig. 1D) (22, 46)and algae (17), was detected only in PC, at a very low level. Thelevel of 3-trans-16:1 in PG (Fig. 1C) was significantly lower inC. merolae (ca. 5%) than in land plants and algae analyzed todate (15 to 40%). Effects of growth temperature were notedfor the fatty acid compositions of MGDG, PE, PC, and PI. InMGDG, the level of 18:2 was higher, while that of 16:0 waslower, at 25°C than at 38°C. For the three classes of phospho-lipids analyzed, the level of 18:2 was higher, while that of 18:1was lower, at 25°C. The level of 18:0 was also lower in PE andPC. However, the growth temperature did not change thequalitative compositions of fatty acids.

Positional distribution of fatty acids. An analysis of thepositional distribution of fatty acids in individual classes oflipids revealed marked differences between the lipids knownas chloroplast lipids and other phospholipids (Table 2). InMGDG, DGDG, SQDG, and PG (chloroplast or plastid lip-ids), 16:0 was primarily bound to the C-2 position, whereas18:2 was attached to the C-1 position. This is the prokaryotictype of distribution (1-C18-2-C16) found in the chloroplast lip-ids of plants and algae. In addition, a significant level of 18:2was also found at the C-2 position in MGDG and DGDG. Thispoints to the presence of the eukaryotic type of molecularspecies (1-C18-2-C18). SQDG contained 1-C16-2-C16 molecularspecies as well. The 3-trans-16:1 species was exclusively boundto the C-2 position of PG, as in the case of plants and algae. Atotally different distribution was found in PE and PC. 16:0 and18:0 were bound to the C-1 position, whereas the C-2 positionwas occupied mainly by 18:1 and 18:2. A low level of 18:2 alsobound to the C-1 position. Therefore, PE and PC consistedmainly of the 1-saturated-2-unsaturated type of molecular spe-cies, as in many eukaryotes. These results suggest that the

TA

BL

E1.

Fat

tyac

idco

mpo

sitio

nof

indi

vidu

allip

idcl

asse

sat

two

diffe

rent

grow

thte

mpe

ratu

res

Fat

tyac

id

Com

posi

tion

(%)

atin

dica

ted

grow

thte

mpa

MG

DG

DG

DG

SQD

GPG

PEPC

PI

38°C

25°C

38°C

25°C

38°C

25°C

38°C

25°C

38°C

25°C

38°C

25°C

38°C

25°C

14:0

0.2

�0.

20.

00.

00.

1�

0.1

0.4

�0.

10.

4�

0.1

0.0

0.0

0.0

0.0

0.7

�0.

20.

4�

0.4

0.0

0.0

16:0

32.5

�1.

624

.4�

1.9

31.3

�0.

828

.8�

1.1

67.4

�1.

371

.3�

1.1

41.6

�1.

638

.2�

3.2

35.7

�1.

339

.0�

2.0

28.7

�2.

231

.2�

2.4

31.9

�1.

835

.8�

2.1

16:1

(9)

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.0

0.8

�0.

20.

5�

0.4

0.0

0.0

16:1

(3-tr

ans)

0.0

0.0

0.0

0.0

0.0

0.0

4.8

�1.

35.

7�

0.5

0.0

0.0

0.0

0.0

0.0

0.0

17:0

0.0

0.0

0.0

0.0

1.1

�0.

10.

5�

0.4

0.0

0.0

1.9

�0.

81.

0�

0.3

1.8

�0.

51.

0�

0.2

2.9

�0.

90.

018

:00.

9�

0.2

1.9

�0.

42.

1�

0.0

1.2

�0.

49.

9�

4.4

7.0

�1.

53.

0�

1.2

2.8

�0.

411

.6�

0.6

7.0

�0.

98.

3�

0.9

2.4

�0.

319

.8�

0.4

16.2

�0.

918

:1(9

)1.

5�

0.5

1.6

�0.

41.

4�

0.0

2.2

�0.

31.

0�

0.1

1.4

�0.

15.

4�

2.6

5.2

�1.

022

.8�

0.7

10.5

�0.

314

.1�

2.4

4.8

�0.

818

.7�

2.4

10.1

�1.

118

:2(9

,12)

65.0

�2.

472

.1�

2.7

65.2

�0.

967

.7�

0.6

20.4

�5.

519

.5�

2.3

45.0

�6.

648

.1�

3.9

28.0

�1.

842

.4�

1.3

45.6

�1.

860

.0�

2.3

26.9

�3.

638

.0�

3.7

Am

ount

oflip

ids

14.0

�0.

318

.8�

1.3

21.3

�0.

724

.2�

3.4

14.5

�1.

211

.3�

1.5

5.5

�0.

68.

5�

0.5

17.3

�0.

417

.0�

1.6

23.0

�1.

716

.6�

1.5

4.5

�0.

83.

6�

1.0

aE

ach

valu

eis

anav

erag

epe

rcen

tage

for

thre

edi

ffere

ntde

term

inat

ions

.Sta

ndar

dde

viat

ions

are

show

nfo

rno

nzer

ova

lues

.For

lipid

com

posi

tion,

each

valu

ere

pres

ent

anav

erag

epe

rcen

tage

for

two

(25°

C)

orth

ree

(38°

C)

diffe

rent

dete

rmin

atio

ns.

1008 SATO AND MORIYAMA EUKARYOT. CELL


http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

plastid lipids of C. merolae consist of prokaryotic and eukary-otic molecular species, as occurs in plants.

Genes involved in the synthesis of lipids and fatty acids. Wethen searched the genomic data of C. merolae to infer genesinvolved in the synthesis of fatty acids and lipids (Table 3). Allcommon genes for the synthesis of long-chain saturated fattyacids were detected. Two types of acetyl-coenzyme A (acetyl-CoA) carboxylases, namely, a prokaryotic multisubunit enzymeof the plastid and a multifunctional cytoplasmic enzyme, werefound. The plastid localization of the nuclear accC gene(CMS299C) product was confirmed by GFP experiments (Ta-ble 3; see Fig. S12 in the supplemental material). Two copies ofthe condensing enzymes, CMM286C and CML329C, were de-tected, but the former was confirmed as the plastid enzyme(see Fig. S12 in the supplemental material). Only three genesfor putative desaturases were found. The gene for the stearoylacyl carrier protein (ACP) desaturase, which is essential inproducing oleate in flowering plants, was not detected. Accord-ing to the Gclust database (data set CZ35, cluster 4024), pu-tative genes for this enzyme were detected in the green algaChlamydomonas reinhardtii (36) and the diatom Thalassiosirapseudonana (5).

Genes encoding the enzymes involved in the synthesis ofcomplex lipids were also identified by homology searches(Table 3). The eukaryotic MGDG synthase (encoded byCMI271C) (9) was found, and its intracellular localization wasconfirmed with a GFP fusion protein (Table 3; see Fig. S12 inthe supplemental material). Curiously, a homolog of the Syn-echocystis enzyme Sll1377, which was recently found to be theenzyme involved in the synthesis of monoglucosyl diacylglyc-erol (GlcDG) (6), was also detected (CMT267C) (Gclust dataset CZ20x0, cluster 2792). However, we could not detect a spotof putative GlcDG migrating slightly faster than MGDG inTLC after a short (10 min) pulse-labeling period (data notshown), which normally can detect a significant amount ofGlcDG (equivalent to the amount of MGDG) in Anabaena(47). The function of CMT267C therefore still has to be de-termined.

Another point is the lack of a plant-type DGDG synthesisenzyme (DGD1/DGD2). No homolog of DGD1/2 (Gclust dataset CZ35, cluster 5665) has been detected in cyanobacteria.Until now, a cyanobacterial DGDG synthesis enzyme has not

FIG. 1. Gas chromatographic separation of fatty acid methyl esters.A capillary column (50 m long) coated with SS-10 was used. (A) Fattyacid methyl esters prepared from the total lipids of C. merolae cellsgrown at 25°C. (Inset) Enlargement (16-fold) to show the absence of18:3(9,12,15) (arrow). (B) Fatty acid methyl esters prepared from thetotal lipids of a fern, Adiantum capillus-veneris. All peaks were identi-fied previously (44) and served as a reference. The peaks before peak1 were degradation products of pigments. (C) Fatty acid methyl estersprepared from the PG of C. merolae cells. (D) Reference fatty acidmethyl esters prepared from MGDG of a cyanobacterium, Anabaenavariabilis (46). Peaks: 1, 16:0; 2, 16:1(9); 3, 16:1(3-trans); 4, 16:2(9,12);5, 17:0; 6, 17:1(9); 7, 16:3(7,10,13); 8, 18:0; 9, 18:1(9); 10, 18:1(11); 11,18:2(9,12); 12, 18:3(6,9,12); 13, 18:3(9,12,15); and 14, 18:4(6,9,12,15).

TABLE 2. Positional distribution of fatty acids inmajor classes of lipids

Fatty acid

% at positiona

MGDG DGDG SQDG PG PE PC

C-1 C-2 C-1 C-2 C-1 C-2 C-1 C-2 C-1 C-2 C-1 C-2

14:0 0 0 0 0 0 0 0 0 0 0 0 016:0 2 30 4 29 22 45 3 39 33 2 25 216:1(9) 0 0 0 0 0 0 0 0 0 0 1 016:1(3-trans) 0 0 0 0 0 0 0 4 0 0 0 017:0 0 0 0 0 1 0 0 0 2 0 2 018:0 0 0 0 1 9 1 2 0 10 1 8 118:1(9) 1 0 1 1 1 0 4 0 0 23 3 1218:2(9,12) 47 20 45 19 17 4 41 7 5 23 11 35

a The sum of fatty acids at each position was set to 50%.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

been reported. This raises the possibility that C. merolae has acyanobacterial DGDG synthesis enzyme (see Discussion).

Cardiolipin was not detected (see above), but genomic datasuggested the presence of an enzyme for cardiolipin biosyn-thesis (CMN196C) (Gclust data set CZ20x0, cluster 4414). Thefamily of cluster 4414 was originally annotated as a phosphati-dylglycerophosphate synthase, but Katayama et al. (21) showedthat the Arabidopsis homolog is involved in the synthesis ofcardiolipin. To increase the sensitivity of detection of phos-pholipids, C. merolae cells were incubated with [32P]phosphate,and the lipids were analyzed by two-dimensional TLC (resultsnot shown). However, the putative spot of cardiolipin was stillobscure, and we cannot definitively confirm the presence ofthis lipid at a very low level.

The genomic analysis suggested that C. merolae has thecapability of synthesizing some sterols (Table 3), although thisis not the main topic of the present study. In Table 3, lanosterolsynthase is listed because of its homology, but the exact spec-ificity of the enzyme must be determined experimentally.

Sterol methyltransferases with unknown specificity are alsopredicted (not listed in Table 3).

Phylogenetic analysis of desaturases. To obtain further in-formation on the biosynthesis of fatty acids, a phylogeneticanalysis of desaturases was performed (Fig. 2A). An uncom-pressed version of the identical tree is available upon request.The �9 desaturases are divergent from the �12 and �3 desatu-rases, and all of these groups of desaturases, though still sig-nificantly homologous, diverged from each other before theseparation of prokaryotes and eukaryotes (Fig. 2A). Othertypes of desaturases, such as �6 desaturases, were also in-cluded in the published large phylogenetic tree of desaturases(54), but they are too divergent to allow construction of areliable tree. Among the acyl-CoA �9 desaturases, one groupof enzymes have an extra cytochrome b5 domain (29, 33). Oneof the C. merolae enzymes (CMM045C) belongs to this type(20). Acyl lipid �9 desaturases are typically found in cyanobac-teria (DesC), and homologs are also found in plants, which areknown to function as �7 desaturases acting on MGDG (19).

TABLE 3. Genes for synthesis of lipids in the C. merolae genome

Substrate Product Enzyme or gene name Locus tag(s) and/or gene name(s)a

NA NA ACP acpP (cp)Acetyl-CoA Molonyl-CoA Acetyl-CoA carboxylase (AccABCD) accABD (cp), accC (cp), CMS299C*Acetyl-CoA Molonyl-CoA Acetyl-CoA carboxylase (single

subunit type)CMM188C

Molonyl-CoA Molonyl ACP Transacylase CMT420CMolonyl ACP Acyl ACP Fatty acid synthase complex CMM286C*/CML329C (cytosol),

CMS393C*, CMI240C, CMT381C*Acyl ACP Fatty acid Thioesterase NoneFatty acid Acyl CoA Acyl-CoA synthetase CME186CAcyl-CoA 1-LPA G3P acyltransferase CMJ027C1-LPA Phosphatidic acid LPA acyltransferase CMJ021C, CMF185C, CME109CDiacylglycerol Phosphatidic acid DG kinase CMR054CDiacylglycerol MGDG MGDG synthase CMI271C*Diacylglycerol GlcDG Sll1377 (Synechocystis) CMT267C?b

Diacylglycerol SQDG SQD1/SQD2 CMR012C/CMR015CEthanolamine phosphate CDP ethanolamine MUQ1 CMS052CDiacylglycerol PE Ethanolamine phosphotransferase

(EPT1)CMF133C

Diacylglycerol Triacylglycerol DG acyltransferase CMQ199CPE PC Methyltransferases CMF090C/CMA134CPhosphatidic acid CDP-DG CDS1 CMN215CCDP-DG PGP PGS1 CMJ134CPGP PG Unknown UnknownCDP-DG PI PIS1 CMM125CMGDG DGDG DGD1/DGD2 NoneMGDG DGDG DGDG synthase ycf82 (cp)?c

PG DPG CLS CMN196C?d

Stearyl ACP Oleoyl ACP Stearoyl ACP desaturase NoneStearoyl-CoA Oleoyl-CoA Stearoyl-CoA desaturase

(cytochrome b fusion)CMM045Ce (ER)

Stearoyl lipid (?) Oleoyl lipid (?) Stearoyl lipid (?) desaturase CMJ201C (particulate)Oleoyl lipid (?) Linoleoyl lipid (?) Oleoyl lipid (?) desaturase CMK291C (ER)Farnesyl diphosphate Squalene Squalene synthase CMG178CSqualene 2,3-Epoxysqualene Squalene monooxygenase CMH256C2,3-Epoxysqualene Lanosterol Lanosterol synthase CMJ009C

a cp, genes carried by the chloroplast genome. Genes were identified by sequence similarity and annotations are included in the Cyanidioschyzon Genome Project(28), except where relevant references are given. Asterisks indicate that the plastid localization was confirmed by GFP experiments, while other localizations are shownin parentheses.

b See reference 6.c Sakurai, Mizusawa, Wada, and Sato, unpublished data.d See reference 21.e See reference 20.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

Cyanobacterial DesC and plant DesC-like proteins are sistergroups (Fig. 2A). This suggests that the DesC-like enzymes inplants originated from the cyanobacterial endosymbiont. How-ever, the second �9 desaturase (CMJ201C) in C. merolae isoutside the DesC group and close to bacterial enzymes.

The phylogeny of the �12 (and �3) desaturases is compli-cated (Fig. 2A). The cluster that diverges from the root in-cludes bacterial and some marine cyanobacterial enzymes ofunknown specificity (cluster I). Cluster II includes cyanobac-terial DesA and plant FAD6 localized in the chloroplast. An-other large cluster includes plant FAD2, which is localized inthe ER (cluster IV). The �12 enzymes in marine cyanobacte-ria, C. merolae, and Phaeodactylum tricornutum (diatom) arerelated to the FAD2 group. The desaturases of nematodes(cluster III) diverge from both cluster II and cluster IV. Inter-estingly, �3 desaturases (except the nematode enzyme) divergefrom cluster IV of �12 desaturases. This is supported by a high

bootstrap confidence level (89%). The cyanobacterial DesBprotein was identified as the origin of the plant �3 desaturases,including both chloroplast and ER isozymes.

These results show close relationships of cyanobacterialDesA and DesB with FAD6 and FAD7 in chloroplasts ofplants and green algae (green lineage), respectively. Amongcyanobacteria, marine species, such as Prochlorococcus mari-nus, have a cluster IV enzyme but no DesA or DesB. Theseenzymes were probably acquired by horizontal gene transfer,and this result should not be considered evidence that the C.merolae �12 desaturase originated from marine cyanobacteria.This is in clear contrast with �9 desaturases, for which bothmarine and freshwater species of cyanobacteria have orthologousDesC proteins.

These results suggest that the desaturases of C. merolae areunrelated to the enzymes of the cyanobacteria and the greenlineage. Assuming the monophyletic origin of the red and

FIG. 2. Phylogenetic analysis of putative �12 desaturase in C. merolae (Cme). (A) Compressed phylogenetic tree of �9, �12, and �3desaturases. This tree was obtained by the neighbor-joining method, using MEGA 2 software. A full tree and sequence information are availablefrom the authors upon request. Each number at the branch points indicates a bootstrap confidence level. (B) Detailed analysis of phylogeneticrelationships of desaturases in cluster IV by neighbor-joining and maximum parsimony methods. (C) Best topology of cluster IV by the maximumlikelihood method. (D) Alignment of the N-terminal regions of desaturases in cluster IV. Putative transit peptides for targeting to the ER andplastids are shown. The candidate initiation codons that were tested for the experiment shown in Fig. 3 are shown.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

green lineages (27, 28, 42), this indicates that the red alga doesnot keep cyanobacterial desaturases and retains the desatu-rases that existed before the cyanobacterial endosymbiosis. Inaddition, the �12 desaturases of the diatom P. tricornutum,PtFAD2 and PtFAD6, also clustered with the �12 desaturasesof C. merolae and marine cyanobacteria (Fig. 2B and C).Neighbor joining (Fig. 2B) and maximum likelihood (Fig. 2C)as well as maximum parsimony (Fig. 2B) did not give a con-sistent relationship of these desaturases within this subcluster.PtFAD2 and PtFAD6 are known to be targeted to the ER andplastids, respectively (14).

Localization of �12 desaturase. The N terminus of PtFAD2is similar to the N terminus of other ER-localized FAD2 pro-teins of plants and fungi (Fig. 2D), suggesting the presence ofsimilar signal sequences. However, the �12 desaturase of C.merolae has an N-terminal extension, which is partially similarto the N-terminal extension of PtFAD6. Here we show theentire N-terminal sequence of the putative �12 desaturase(CMK291C). However, in the current version of annotationgiven in the Cyanidioschyzon website (http://merolae.biol.s.u-tokyo.ac.jp/), the CMK291C sequence begins from the secondmethionine. There is an in-frame upstream methionine codon,which could act as the initiation codon, and we used the entiresequence for the analysis in the present study. Intracellularlocalization of the putative �12 desaturase of C. merolae(CMK291C) was examined using GFP fusion constructs (Fig.3). The polypeptide starting from Met1 was targeted to mito-chondria, the polypeptide starting from Met2 was targeted tothe ER, and the polypeptide beginning from Met3 showed noclear localization. It is interesting that the fluorescence of theMet2 construct is localized to the ER as well as to the mem-branes surrounding the nucleus (ER and nuclear envelope). Itis therefore reasonable that the desaturase is translated from

the second methionine, as described in the current database,and targeted to the ER.

Labeling of lipids. The computational analysis suggestedthat there is no desaturase in the plastid, but our lipid analysisshowed that typical plastid lipids containing 18:2, such asMGDG, exist in this red alga. To solve this problem, we ana-lyzed the flow of lipid synthesis within the cell by pulse labelingwith radioactive acetate, followed by a chase period (Fig. 4).About 40% of added radiocarbon was incorporated into thetotal lipid fraction. The results of two-dimensional TLC indi-cated that PC and MGDG were the major labeled lipids. In-terestingly, DGDG was not labeled efficiently after the 1-hlabeling period, but it was densely labeled after the 20-h chaseperiod. The radioactivity in PC decreased steadily during thechase period, while the radioactivity in MGDG increased.These results are consistent with the flow of carbon from PC toMGDG and then to DGDG, as documented for higher plants(10).

Figure 5 shows the results of fatty acid analysis using arecently developed reversed-phase argentation TLC method.In the total lipid fraction, desaturation of fatty acids was clearlydetected. In MGDG, however, only 16:0 and 18:2 were de-tected, with a decrease in 16:0 and an increase in 18:2. Nopossible intermediate, such as 18:0, 18:2, or 16:1, was detectedas a labeled acid. For DGDG, labeling in the fatty acids wasnoted after the 6-h chase period. 16:0 was labeled first, andafter the 20-h chase period, 18:2 was also labeled. In PC,various fatty acids were labeled, and the changes were ex-plained by desaturation, except for the initial high level oflabeling of 18:2, which might suggest rapid turnover of a smallpool of PC-bound 18:2.

Lipid molecular species were also analyzed (Fig. 6). InMGDG, 18:2/16:0 was the major labeled molecular species

FIG. 3. Targeting of putative �12 desaturase. There are three methionine codons that could act as the initiation codon in the putative �12desaturase gene (CMK291C), as shown in Fig. 2D. The �12 sequences starting from the first, second, and third methionine codons were fused withthe GFP gene and introduced into the onion epidermis by particle bombardment. Fluorescence of GFP and Nomarski differential interferenceimages are shown. The control (GFP alone) is shown in Fig. S12 in the supplemental material. Tungsten particles are visible within the cells assmall black patches.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

from the beginning. At later times, 18:2/18:2 was also labeled.The rapid labeling of the 18:2/16:0 species of MGDG was notfound in similar labeling experiments with plants (18), algae(17, 38, 50), or cyanobacteria (47) and seemed quite strange.This molecular species was recovered from the plate, and thefatty acids were analyzed (Fig. 5E). It was clear that the initiallylabeled 18:2/16:0-MGDG was labeled only in its 16:0 part, notin its 18:2 part. After the chase, both fatty acids were labeled.The delay in 18:2 labeling can be explained by the large sizeof the pool that provides 18:2. This suggests that MGDG issynthesized from 16:0 and 18:2 but that the two acids aresupplied from different compartments, namely, the rapidly la-beled 16:0 is supplied within the plastids, whereas the slowlylabeled 18:2 is supplied from outside the plastids, possibly fromthe ER. Labeling of DGDG molecular species was similar tothat for MGDG. In PC, various molecular species were la-

beled, and the radioactivity shifted from less unsaturated tomore unsaturated species. This suggests that the desaturationof fatty acids occurs mainly on PC, as in other eukaryotes (10, 17).

The initial molecular species of galactolipid synthesis was

FIG. 4. Labeling of polar lipids with radioactive acetate. C. merolaecells were incubated with [14C]acetate for 1 h (A) and then chased for20 h (B). Lipids were separated by two-dimensional TLC, and then theradioactivity of each lipid spot was counted (C). The radioactivityexperiments were repeated three times, but representative results areshown throughout the paper for consistency of data.

FIG. 5. Analysis of radioactivity in fatty acids in the total lipids (A),MGDG (B), DGDG (C), and PC (D). Each lipid class was isolated bytwo-dimensional TLC and recovered from the gel. Each isolated lipidwas then subjected to methanolysis. Fatty acid methyl esters wereextracted, separated by reversed-phase argentation TLC, and thenquantified. (E) Autoradiogram of reversed-phase argentation TLCanalysis of fatty acids in the 18:2/16:0 molecular species of MGDGbefore (lane 0) and after (lane 20) the chase. The original autoradio-grams for panels A to D are shown in Fig. S10 in the supplementalmaterial.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

also confirmed by incorporation of UDP-[U-14C]galactose, us-ing isolated plastids (Fig. 7). 18:2/16:0 and 18:2/18:2 were themajor labeled molecular species in both classes of lipids, al-though their exact proportions were not similar.

DISCUSSION

Lack of trienoic acids. The results presented above indicatethat C. merolae does not contain polyunsaturated acids withmore than three double bonds. This is unusual for a photo-synthetic eukaryote. Among the eukaryotes, Saccharomycescerevisiae and some species of yeasts are the only organisms thatpossess only saturated and monounsaturated acids (15). Pho-tosynthetic organisms in general have polyunsaturated acidssuch as 18:3, 20:4, or 22:6. The only known photosyntheticorganisms that lack polyunsaturated acids are Synechococcus

sp. strain PCC 6301 (formerly called Anacystis nidulans [49])and its close relative PCC 7942 as well as some unicellularspecies of cyanobacteria (22). Polyunsaturated acids such as18:3 are known to be important for two major reasons. First,18:3 and 20:4 are precursors to physiologically active signalingcompounds, such as jasmonate and prostaglandins in plantsand animals, respectively. The lack of 18:3 implies that jas-monate might not act as a signal in this microalga. Second, 18:3is required for maintaining photosynthetic activities at low andhigh temperatures. At low temperatures, this acid keeps thefluidity of membrane lipids to increase the tolerance to chillingand freezing (37). In addition, 18:3 is known to increase toler-ance to high temperatures in plants (32). This does not meanthat C. merolae does not have a tolerance to both cold and hightemperatures. In fact, this alga can grow in a wide range oftemperatures, from 25 to 50°C.

Pathway of lipid and fatty acid biosynthesis in C. merolae. C.merolae contains common glycerolipids that are usually foundin plants and algae. Genomic analysis supports the hypothesisthat this alga possesses a standard pathway of biosynthesis ofthese glycerolipids, except for DGDG (see Fig. S9 in the sup-plemental material). The biosynthesis of DGDG in plants isknown to be catalyzed by the products of the DGD1/DGD2genes. However, C. merolae does not have a homolog of theseplant-type galactosyltransferases. A survey of the Gclust data-base (data set CZ20x0, cluster 2825) identified a putative gly-cosyltransferase (Ycf82) shared by cyanobacteria and the C.merolae plastid genome. The disruption of the Synechocystishomolog (slr1508) resulted in a lack of DGDG (I. Sakurai, N.Mizusawa, H. Wada, and N. Sato, unpublished data). Thisgene is also being analyzed in different laboratories, and wehope that this is the structural gene for the cyanobacterial andred algal DGDG synthesis enzyme. We also note that furtherstudies are needed to obtain a conclusion about the presenceof cardiolipin in C. merolae.

Biochemical analysis clearly indicated that C. merolae cansynthesize saturated fatty acids and mono- and diunsaturatedfatty acids. The gene involved in the synthesis of �3-trans-16:1,which is a typical fatty acid present at the C-2 position of PG

FIG. 6. Analysis of radioactivity in various molecular species ofMGDG (A), DGDG (B), and PC (C). Each lipid class was isolated bytwo-dimensional TLC. The molecular species were separated byargentation TLC and then quantified. The original autoradiogramsfor panels A to D are shown in Fig. S11 in the supplementalmaterial.

FIG. 7. Incorporation of radioactive galactose into various molec-ular species of MGDG (A) and DGDG (B) in isolated chloroplasts ofC. merolae.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

in photosynthetic eukaryotes, is still unknown and is not asubject of the current discussion. The genomic analysis indi-cated that the pathway of fatty acid desaturation in this algamust be very different from that in flowering plants. Althoughfatty acid biosynthesis certainly occurs only in the plastids in C.merolae, as evidenced by the localization experiments usingGFP (Table 3; see Fig. S12 in the supplemental material), thesynthesis of oleate occurs in the ER and in the form of bothacyl-CoA and acyl lipid (Table 3). This is fundamentally dif-ferent from the �9 desaturation in flowering plants, in whicholeate is produced only by the stearoyl ACP desaturase in thechloroplast, an enzyme that catalyzes desaturation by a differ-ent mechanism from that of acyl lipid desaturases (53). Thegenes encoding �7 acyl lipid desaturases have also been re-

ported for plants, with some of their products acting as MGDGdesaturases (19).

The second desaturation, namely, �12 desaturation, is cat-alyzed by the only �12 acyl lipid desaturase, which is likely tobe localized in the ER (Fig. 3). However, MGDG, DGDG, andPG consist mainly of 1-(18:2)-2-(16:0) species, with smallamounts of 1-(18:2)-2-(18:2) molecular species (Table 2). Thisindicates that the molecular species of these plastid lipids aresynthesized by the coupled supply of 16:0 within the plastid and18:2 from the outside. The role of the ER in acyl group de-saturation was suggested by the results of tracer experimentswith the cryptophyte Cryptomonas sp. (39), a descendant ofsecondary red algal endosymbiosis. In Cryptomonas, the radio-label was initially incorporated into PC and then moved to

FIG. 8. Comparison of pathways of galactolipid biosynthesis in C. merolae and flowering plants. In flowering plants, each of the prokaryotic(within the plastid) and eukaryotic (via the ER) pathways can produce unsaturated galactolipids, and the proportion of each pathway is differentin different plants. In C. merolae, simultaneous functioning of both pathways is absolutely required for the synthesis of MGDG. The synthesis ofMGDG is catalyzed by the plant-type enzyme. Although a homolog of cyanobacterial glucosyltransferase (Sll1377) was detected in the genome,there is no evidence for the production of GlcDG. The plant-type enzyme for the synthesis of DGDG does not exist in C. merolae. We suspectthat another enzyme is used in cyanobacteria and C. merolae for the synthesis of DGDG.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

MGDG. This and other results indicate that PC is the majorsubstrate of desaturation in Cryptomonas. An essentially simi-lar flow of carbon from PC to MGDG was found in C. merolae(Fig. 4).

Origins of desaturases. The desaturases in plants and algaehave two different origins, namely, either the eukaryotic hostor the cyanobacterial endosymbiont. The diversification of �9and �12 desaturases had already occurred before the creationof eukaryotes, but some cyanobacterial enzymes were alsotransferred to plants and algae during endosymbiosis. It is clearthat the �3 desaturases were generated from a �12 desaturase,but the exact organism in which �3 desaturase was created wasnot identified. Fig. 2A clearly shows that the cyanobacterial �3desaturase was not created within cyanobacteria. This is aquestion to be answered in the future.

The phylogenetic analysis (Fig. 2) suggested a cyanobacterialorigin of �3 desaturases of plants. Both plastidic (FAD7 andFAD8) and microsomal (FAD3) �3 desaturases are monophy-letic and originated from cyanobacterial DesB. Likewise, theplastidic �12 desaturase FAD6 in plants originated from cya-nobacterial DesA. The �12 desaturase of C. merolae is a sisterto eukaryotic �12 desaturases (FAD2) localized in the ER.Therefore, C. merolae completely lacks the desaturase genes ofcyanobacterial origin, namely, desA, desB, FAD3, FAD6,FAD7, and FAD8. It is not yet clear if this is due to the selectiveloss of cyanobacterial enzymes in the lineage of primitiverhodophytes after endosymbiosis.

In cyanobacteria, the conversion of 18:0 to 18:1 occurs onacyl lipids by the action of DesC (Fig. 2). Neither DesC norstearoyl ACP desaturase is present in C. merolae. Desaturationof 18:0 to 18:1 therefore proceeds by the two �9 desaturases,which have no direct relationship with �9 desaturases of plantsor cyanobacteria (Fig. 2).

The similarity of the green lineage and the cyanobacteriawith respect to desaturases is in clear contrast with the situa-tion for the genomic machinery in the plastid (41, 42, 52).Various DNA-binding proteins of cyanobacterial origin areretained in the red lineage, whereas the genomic machineryacquired eukaryotic components in the green lineage. Thegenealogy of the enzymes of MGDG and DGDG synthesissupports the association of red algae and cyanobacteria. Theseobservations shed light on the multiple or mosaic origins oflipid-related enzymes in the red algae.

Coupled pathway of MGDG synthesis. All of the results oftracer experiments clearly indicate that MGDG is synthesizedfrom 16:0 and 18:2 within plastids. The DG moiety of MGDGis either 18:2/16:0 or 18:2/18:2 from the beginning (Fig. 7). 16:0can be supplied within the plastids, whereas 18:2 cannot besupplied within the plastids because of the lack of desaturaseswithin the plastids. It is most likely that PC in the ER is the siteof desaturation, because all intermediate fatty acids were de-tected in the PC. A summary of galactolipid biosynthesis in C.merolae is illustrated in Fig. 8. Biosynthesis of MGDG there-fore requires a supply of extraplastidic 18:2. This is the mostinteresting characteristic of C. merolae. We call this the cou-pled pathway, as opposed to the prokaryotic and eukaryoticpathways that were explained in the introduction. The coupledpathway of MGDG synthesis is a result of the total lack ofdesaturases of cyanobacterial origin and of stearoyl ACP de-saturase.

In red algae other than those in the Cyanidiales (17, 35), C18

unsaturated fatty acids are not abundant. In Porphyra yezoensis,for example, C18 polyunsaturated acids amount to 1% of thegalactolipids (4). This can be explained if, as in C. merolae, thesynthesis of unsaturated fatty acids takes place mostly in the ERrather than in the plastids. The elongation of unsaturated fattyacids is also active in the ER, which results in vast productionof C20 or C22 unsaturated fatty acids that characterize non-Cyanidiales red algal galactolipids. Although we still do notknow if red algae in general lack stearoyl ACP desaturase, thisenzyme could be an interesting switch that changes the meta-bolic flow of fatty acids in red algae.

ACKNOWLEDGMENTS

This work was supported in part by grants-in-aid from JSPS (no.13440234, 15370017, and 18017005), a grant-in-aid for priority areasfrom MEXT (Genome Biology), a grant-in-aid for creative scientificresearch (16GS0304), and a grant from the Takeda Science Founda-tion.

REFERENCES

1. Adachi, J., and M. Hasegawa. 1992. Amino acid substitution of proteinscoded for in mitochondrial DNA during mammalian evolution. Jpn.J. Genet. 67:187–197.

2. Allen, M. B. 1959. Studies with Cyanidium caldarium, an anomalously pig-mented chlorophyte. Arch. Microbiol. 32:270–277.

3. Altschul, S. F., T. L. Madden, A. A. Schafer, J. Zhang, Z. Zhang, W. Miller,and D. J. Lipman. 1997. Gapped BLAST and PSI-BLAST: a new generationof protein database search programs. Nucleic Acids Res. 25:3389–3402.

4. Araki, S., T. Sakurai, T. Omata, A. Kawaguchi, and N. Murata. 1986. Lipidand fatty acid composition in the red alga Porphyra yezoensis. Jpn. J. Phycol.34:94–100.

5. Armbrust, E. V., J. A. Berges, C. Bowler, B. R. Green, D. Martinez, N. H.Putnam, S. Zhou, A. E. Allen, K. E. Apt, M. Bechner, M. A. Brzezinski, B. K.Chaal, A. Chiovitti, A. K. Davis, M. S. Demarest, J. C. Detter, T. Glavina, D.Goodstein, M. Z. Hadi, U. Hellsten, M. Hildebrand, B. D. Jenkins, J. Jurka,V. V. Kapitonov, N. Kroger, W. W. Lau, T. W. Lane, F. W. Larimer, J. C.Lippmeier, S. Lucas, M. Medina, A. Montsant, M. Obornik, M. S. Parker, B.Palenik, G. J. Pazour, P. M. Richardson, T. A. Rynearson, M. A. Saito, D. C.Schwartz, K. Thamatrakoln, K. Valentin, A. Vardi, F. P. Wilkerson, andD. S. Rokhsar. 2004. The genome of the diatom Thalassiosira pseudonana:ecology, evolution, and metabolism. Science 306:79–86.

6. Awai, K., T. Kakimoto, C. Awai, T. Kaneko, Y. Nakamura, K. Takamiya, H.Wada, and H. Ohta. 2006. Comparative genomic analysis revealed a gene formonoglucosyldiacylglycerol synthase, an enzyme for photosynthetic mem-brane lipid synthesis in cyanobacteria. Plant Physiol. 141:1120–1127.

7. Beisson, F., A. J. K. Koo, S. Ruuska, J. Schwender, M. Pollard, J. J. Thelen,T. Paddock, J. J. Salas, L. Savage, A. Milcamps, V. B. Mhaske, Y. Cho, andJ. B. Ohlrogge. 2003. Arabidopsis genes involved in acyl lipid metabolism. A2003 census of the candidates, a study of the distribution of expressedsequence tags in organs, and a web-based database. Plant Physiol. 132:681–697.

8. Bligh, E. G., and W. J. Dyer. 1959. A rapid method of total lipid extractionand purification. Can. J. Biochem. Physiol. 37:911–917.

9. Botte, C., C. Jeanneau, L. Snajdrova, O. Bastien, A. Imberty, C. Breton, andE. Marechal. 2005. Molecular modeling and site-directed mutagenesis ofplant chloroplast monogalactosyldiacylglycerol synthase reveal critical resi-dues for activity. J. Biol. Chem. 280:34691–34701.

10. Browse, J., and C. Somerville. 1991. Glycerolipid synthesis: biochemistry andregulation. Annu. Rev. Plant Physiol. Plant Mol. Biol. 42:467–506.

11. Cavalier-Smith, T. 2003. Genomic reduction and evolution of novel geneticmembranes and protein-targeting machinery in eukaryote-eukaryote chi-maeras (meta-algae). Philos. Trans. R. Soc. Lond. B 358:109–134.

12. Chenna, R., H. Sugawara, T. Koik, R. Lopez, T. J. Gibson, D. G. Higgins,and J. D. Thompson. 2003. Multiple sequence alignment with the Clustalseries of programs. Nucleic Acids Res. 31:3497–3500.

13. Chiu, W.-I., Y. Niwa, W. Zeng, T. Hirose, H. Kobayashi, and J. Sheen. 1996.Engineered GFP as vital reporter in plants. Curr. Biol. 6:325–330.

14. Domergue, F., P. Spiekermann, J. Lerchl, C. Beckmann, O. Kilian, P. G.Kroth, W. Boland, U. Zahringer, and E. Heinz. 2003. New insight intoPhaeodactylum tricornutum fatty acid metabolism. Cloning and functionalcharacterization of plastidial and microsomal delta12-fatty acid desaturases.Plant Physiol. 131:1648–1660.

15. Erwin, J. A. 1973. Comparative biochemistry of fatty acids in eukaryoticmicroorganisms, p. 41–143. In J. A. Erwin (ed.), Lipids and biomembranes ofeukaryotic microorganisms. Academic Press, New York, NY.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

16. Fischer, W., E. Heinz, and M. Zeus. 1973. The suitability of lipase fromRhizopus arrhizus delemar for analysis of fatty acid distribution in dihexosyldiglycerides, phospholipids and plant sulfolipids. Hoppe-Seyler’s Z. Physiol.Chem. 354:1115–1123.

17. Harwood, J. L., and A. L. Jones. 1989. Lipid metabolism in algae. Adv. Bot.Res. 16:1–53.

18. Heemskerk, J. W. M., H. Schmidt, U. Hammer, and E. Heinz. 1991. Biosyn-thesis and desaturation of prokaryotic galactolipids in leaves and isolatedchloroplasts from spinach. Plant Physiol. 96:144–152.

19. Heilmann, I., S. Mekhedov, B. King, J. Browse, and J. Shanklin. 2004.Identification of the Arabidopsis palmitoyl-monogalactosyldiacylglycerol 7-desaturase gene FAD5, and effects of plastidial retargeting of Arabidopsisdesaturases on the fad5 mutant phenotype. Plant Physiol. 136:4237–4245.

20. Itoh, R., K. Toda, H. Takahashi, H. Takano, and T. Kuroiwa. 1998. Delta-9fatty acid desaturase gene containing a carboxyl-terminal cytochrome b5domain from the red alga Cyanidioschyzon merolae. Curr. Genet. 33:165–170.

21. Katayama, K., I. Sakurai, and H. Wada. 2004. Identification of anArabidopsis thaliana gene for cardiolipin synthase located in mitochon-dria. FEBS Lett. 577:193–198.

22. Kenyon, C. N., R. Rippka, and R. Y. Stanier. 1972. Fatty acid compositionand physiological properties of some filamentous blue-green algae. Arch.Mikrobiol. 83:216–236.

23. Kleinschmidt, M. G., and V. A. McMahon. 1970. Effect of growth temper-ature on the lipid composition of Cyanidium caldarium. II. Glycolipid andphospholipid components. Plant Physiol. 46:290–293.

24. Kumar, S., K. Tamura, I. B. Jakobsen, and M. Nei. 2001. MEGA2: molec-ular evolutionary genetics analysis software. Bioinformatics 17:1244–1245.

25. Kuroiwa, T. 1998. The primitive red algae Cyanidium caldarium andCyanidioschyzon merolae as model system for investigating the dividingapparatus of mitochondria and plastids. BioEssays 20:344–354.

26. Marquardt, T. C., and R. F. Wilson. 1998. An improved reversed-phasethin-layer chromatography method for separation of fatty acid methyl esters.J. Am. Oil Chem. Soc. 75:1889–1892.

27. Martin, W., T. Rujan, E. Richly, A. Hansen, S. Cornelsen, T. Lins, D. Leister,B. Stoebe, M. Hasegawa, and D. Penny. 2002. Evolutionary analysis ofArabidopsis, cyanobacterial, and chloroplast genomes reveals plastid phy-logeny and thousands of cyanobacterial genes in the nucleus. Proc. Natl.Acad. Sci. USA 99:12246–12251.

28. Matsuzaki, M., O. Misumi, T. Shin-i, S. Maruyama, M. Takahara, S.Miyagishima, T. Mori, K. Nishida, F. Yagisawa, K. Nishida, Y. Yoshida,Y. Nishimura, S. Nakao, T. Kobayashi, Y. Momoyama, T. Higashiyama,A. Minoda, M. Sano, H. Nomoto, K. Oishi, H. Hayashi, F. Ohta, S.Nishizaka, S. Haga, S. Miura, T. Morishita, Y. Kabeya, K. Terasawa, Y.Suzuki, Y. Ishii, S. Asakawa, H. Takano, N. Ohta, H. Kuroiwa, K.Tanaka, N. Shimizu, S. Sugano, N. Sato, H. Nozaki, N. Ogasawara, Y.Kohara, and T. Kuroiwa. 2004. Genome sequence of the ultrasmall uni-cellular red alga Cyanidioschyzon merolae 10D. Nature 428:653–657.

29. Mitchell, A. G., and C. E. Martin. 1995. A novel cytochrome b5-like domainis linked to the carboxyl terminus of the Saccharomyces cerevisiae �9 fattyacid desaturase. J. Biol. Chem. 270:29766–29772.

30. Moreira, D., H. Le Guyader, and H. Philippe. 2000. The origin of red algaeand the evolution of chloroplasts. Nature 405:69–72.

31. Moretti, A., and R. Nazzaro. 1980. Fatty acids in thermoacidophilic algae.Delpinoa 21:3–11.

32. Murakami, Y., M. Tsuyama, Y. Kobayashi, H. Kodama, and K. Iba. 2000.Trienoic fatty acids and plant tolerance of high temperature. Science 287:476–479.

33. Napier, J. A., O. Sayanova, P. Sperling, and E. Heinz. 1999. A growing familyof cytochrome b5-domain fusion proteins. Trends Plant Sci. 4:2–4.

34. Nozaki, H., M. Matsuzaki, M. Takahara, O. Misumi, H. Kuroiwa, M.Hasegawa, T. Shin-i, Y. Kohara, N. Ogasawara, and T. Kuroiwa. 2003.

Phylogeny of plastids based on cladistic analysis of gene loss inferredfrom complete plastid genome sequences. J. Mol. Evol. 56:485–497.

35. Nyberg, H., and K. Koskimies-Soininen. 1984. The glycolipid fatty acids ofPorphyridium purpureum cultured in the presence of detergents. Phyto-chemistry 23:751–757.

36. Riekhof, W. R., B. B. Sears, and C. Benning. 2005. Annotation of genesinvolved in glycerolipid biosynthesis in Chlamydomonas reinhardtii: discoveryof the betaine lipid synthase BTA1Cr. Eukaryot. Cell 4:242–252.

37. Routaboul, J.-M., S. F. Fischer, and J. Browse. 2000. Trienoic fatty acids arerequired to maintain chloroplast function at low temperatures. Plant Physiol.124:1697–1705.

38. Sato, N. 1991. Lipids in Cryptomonas CR-1. I. Occurrence of betaine lipids.Plant Cell Physiol. 32:819–825.

39. Sato, N. 1991. Lipids in Cryptomonas CR-1. II. Biosynthesis of betaine lipidsand galactolipids. Plant Cell Physiol. 32:845–851.

40. Sato, N. 2000. SISEQ: manipulation of multiple sequence and large databasefiles for common platforms. Bioinformatics 16:180–181.

41. Sato, N. 2001. Was the evolution of plastid genetic machinery discontinuous?Trends Plant Sci. 6:151–156.

42. Sato, N. 2006. Origin and evolution of plastids: genomic view on the unifi-cation and diversity of plastids, p. 75–102. In R. R. Wise and J. K. Hoober(ed.), The structure and function of plastids. Springer, Berlin, Germany.

43. Sato, N. 2006. Gclust server: phylogenetic profiling with pre-definedorganism sets. Genome Inform. 17:P156.

44. Sato, N., and M. Furuya. 1984. The composition of lipids and fatty acidsdetermined at various stages of haploid and diploid generations in the fernAdiantum capillus-veneris. Physiol. Plant 62:139–147.

45. Sato, N., M. Ishikawa, M. Fujiwara, and K. Sonoike. 2005. Mass identifica-tion of chloroplast proteins of endosymbiont origin by phylogenetic profilingbased on organism-optimized homologous protein groups. Genome Inform.16:56–68.

46. Sato, N., and N. Murata. 1980. Temperature shift-induced responses in lipidsin the blue-green alga, Anabaena variabilis. The central role of diacylmono-galactosylglycerol in thermo-adaptation. Biochim. Biophys. Acta 619:353–366.

47. Sato, N., and N. Murata. 1982. Lipid biosynthesis in the blue-green alga,Anabaena variabilis. II. Fatty acids and lipid molecular species. Biochim.Biophys. Acta 710:279–289.

48. Sato, N., and N. Murata. 1988. Membrane lipids. Methods Enzymol. 167:251–259.

49. Sato, N., N. Murata, Y. Miura, and N. Ueta. 1979. Effect of growth temper-ature on lipid and fatty acid compositions in the blue-green algae, Anabaenavariabilis and Anacystis nidulans. Biochim. Biophys. Acta 572:19–28.

50. Sato, N., Y. Nemoto, and M. Furuya. 1988. Lipids of Chattonella antiqua(Raphidophyceae). Plant Physiol. Biochem. 26:93–98.

51. Sato, N., K. Sekine, Y. Kabeya, S. Ehira, M. Onuma, and N. Ohta. 2004.Discontinuous evolution of plastid genomic machinery: radical replacementof major DNA-binding proteins. Endocytobiosis Cell Res. 15:286–293.

52. Sato, N., K. Terasawa, K. Miyajima, and Y. Kabeya. 2003. Organization,developmental dynamics and evolution of the plastid nucleoid. Int. Rev.Cytol. 232:217–262.

53. Shanklin, J., and E. B. Cahoon. 1998. Desaturation and related modificationof fatty acids. Annu. Rev. Plant Physiol. Plant Mol. Biol. 49:611–641.

54. Sperling, P., P. Ternes, T. K. Zank, and E. Heinz. 2003. The evolution ofdesaturases. Prostaglandins Leukot. Essent. Fatty Acids 68:73–95.

55. Toda, K., H. Takano, S. Miyagishima, H. Kuroiwa, and T. Kuroiwa. 1998.Characterization of a chloroplast isoform of serine acetyltransferase fromthe thermo-acidiphilic red alga Cyanidioschyzon merolae. Biochim. Biophys.Acta 1403:72–84.

56. Yoon, H. S., J. D. Hackett, G. Pinto, and D. Bhattacharya. 2002. A singleorigin of the peridinin- and fucoxanthin-containing plastids in dinoflagellatesthrough tertiary endosymbiosis. Proc. Natl. Acad. Sci. USA 99:15507–15512.



http://ec.asm.org/

Dow

nloaded from

http://ec.asm.org/

genomic and biochemical analysis of lipid biosynthesis in ... · istry to reveal similarities and...

Documents