characterization of type 2 diacylglycerol acyltransferases ... · chlamydomonas reinhardtii harbors...

Characterization of type 2 diacylglycerol acyltransferases inChlamydomonas reinhardtii reveals their distinct substratespecificities and functions in triacylglycerol biosynthesis

Jin Liu1,2, Danxiang Han3, Kangsup Yoon3, Qiang Hu3,* and Yantao Li1,*1Institute of Marine and Environmental Technology, University of Maryland Center for Environmental Science and University

of Maryland Baltimore County, Baltimore, MD 21202, USA,2Department of Energy and Resources Engineering, College of Engineering, Peking University, Beijing 100871, China, and3Center for Microalgal Biotechnology and Biofuels, Institute of Hydrobiology, Chinese Academy of Sciences, Wuhan 430072,

China

Received 3 December 2015; revised 4 February 2016; accepted 8 February 2016; published online 26 February 2016.

*For correspondence (e-mails [email protected] or [email protected]).

SUMMARY

Diacylglycerol acyltransferases (DGATs) catalyze a rate-limiting step of triacylglycerol (TAG) biosynthesis in

higher plants and yeast. The genome of the green alga Chlamydomonas reinhardtii has multiple genes

encoding type 2 DGATs (DGTTs). Here we present detailed functional and biochemical analyses of Chlamy-

domonas DGTTs. In vitro enzyme analysis using a radiolabel-free assay revealed distinct substrate specifici-

ties of three DGTTs: CrDGTT1 preferred polyunsaturated acyl CoAs, CrDGTT2 preferred monounsaturated

acyl CoAs, and CrDGTT3 preferred C16 CoAs. When diacylglycerol was used as the substrate, CrDGTT1 pre-

ferred C16 over C18 in the sn-2 position of the glycerol backbone, but CrDGTT2 and CrDGTT3 preferred C18

over C16. In vivo knockdown of CrDGTT1, CrDGTT2 or CrDGTT3 resulted in 20–35% decreases in TAG con-

tent and a reduction of specific TAG fatty acids, in agreement with the findings of the in vitro assay and

fatty acid feeding test. These results demonstrate that CrDGTT1, CrDGTT2 and CrDGTT3 possess distinct

specificities toward acyl CoAs and diacylglycerols, and may work in concert spatially and temporally to syn-

thesize diverse TAG species in C. reinhardtii. CrDGTT1 was shown to prefer prokaryotic lipid substrates and

probably resides in both the endoplasmic reticulum and chloroplast envelope, indicating its role in prokary-

otic and eukaryotic TAG biosynthesis. Based on these findings, we propose a working model for the role of

CrDGTT1 in TAG biosynthesis. This work provides insight into TAG biosynthesis in C. reinhardtii, and paves

the way for engineering microalgae for production of biofuels and high-value bioproducts.

Keywords: Chlamydomonas reinhardtii, diacylglycerol acyltransferase, lipid metabolism, triacylglycerol,

algae, biofuels.

INTRODUCTION

Triacylglycerols (TAGs) are energy-rich reduced carbon

reserves that are commonly found in algae, higher plants,

fungi and animals. Under stress conditions such as nutri-

ent deprivation or high light irradiance, many microalgae

accumulate high levels of TAGs as storage compounds (Hu

et al., 2008). Recently, growing demand for sustainable

fuels has revived interest in exploring microalgae as a

renewable feedstock for biofuel production. However, TAG

productivity from the naturally occurring microalgae cur-

rently used for oil production remains far below the theo-

retical maximum (Sheehan et al., 1998; Hu et al., 2008; Li

et al., 2011).

Triacylglycerol biosynthesis has been intensively studied

in yeast and higher plants, and is thought to occur mainly

via two routes: an acyl CoA-independent route and an acyl

CoA-dependent route. The acyl CoA-independent route

involves transfer of a fatty acyl moiety from a phospholipid

to diacylglycerol (DAG) to form TAG; this transfer is medi-

ated by a phospholipid:DAG acyltransferase (PDAT)

(Dahlqvist et al., 2000; Stahl et al., 2004; Yoon et al., 2012;

Fan et al., 2013). By contrast, the acyl CoA-dependent route

uses acyl CoA as the acyl donor and involves three

sequential acylations of glycerol-3-phosphate, with the

final acylation catalyzed by DAG acyltransferase (DGAT)

© 2016 The AuthorsThe Plant Journal © 2016 John Wiley & Sons Ltd

3

The Plant Journal (2016) 86, 3–19 doi: 10.1111/tpj.13143

(Ohlrogge and Jaworski, 1997). Several types of DGATs

have been identified in higher plants, including DGAT1

and DGAT2, two structurally distinct groups of membrane-

bound acyltransferases (Ohlrogge and Jaworski, 1997;

Lung and Weselake, 2006). Additional DGATs include a

type 3 soluble DGAT that is found in the cytosol of Arachis

hypogaea (peanut) (Saha et al., 2006) and a bi-functional

DGAT/wax ester synthase found in the bacterium Acineto-

bacter calcoaceticus (Kalscheuer and Steinbuchel, 2003).

DGATs have been proposed to play a regulatory role in

plant TAG biosynthesis, because DGAT catalyzes the only

committed step in TAG biosynthesis. Moreover, modula-

tion of DGAT activity in Brassica napus altered oil produc-

tion (Lung and Weselake, 2006). In Arabidopsis thaliana,

mutation of DGAT1 affected seed oil biosynthesis, but

mutation of DGAT2 did not (Zou et al., 1999). By contrast,

work in Vernicia fordii (tung tree) indicated that DGAT2 is

essential for TAG biosynthesis (Shockey et al., 2006), sug-

gesting that DGAT enzymes have distinct roles in TAG

biosynthesis in different organisms. Sequence analysis of

algal genomes has identified DGAT1 and DGAT2 homo-

logs (Merchant et al., 2007; Wagner et al., 2010; Guiheneuf

et al., 2011; Radakovits et al., 2012; Vieler et al., 2012), but

the regulation and contribution of different DGAT genes to

TAG biosynthesis remain poorly understood.

Chlamydomonas reinhardtii has been widely used as a

model organism to study TAG metabolism, and many

genes involved in TAG biosynthesis have been identified in

the C. reinhardtii genome, including one PDAT gene, one

DGAT1 gene, five DGAT2 genes and one DGAT3-like gene

(Merchant et al., 2007; Boyle et al., 2012; Blaby et al., 2013).

DGATs, rather than PDAT, are thought to be the major con-

tributors to TAG formation in C. reinhardtii under nitrogen

deprivation (Hu et al., 2008; Yoon et al., 2012).

Chlamydomonas reinhardtii harbors five type 2 DGAT

genes (named DGTT genes in this organism), in contrast to

Arabidopsis thaliana, which possesses only one type 2

DGAT. The presence of multiple DGTT genes in C. rein-

hardtii raises the possibilities that they may be expressed in

distinct subcellular compartments, possess differential sub-

strate specificities, and respond differently to abiotic stress

conditions, thus contributing synergistically to formation of

diverse TAG species. Previously, the C. reinhardtii DGAT

genes, DGTTs in particular, were partially characterized by

functional complementation in TAG-deficient yeast strains

(Boyle et al., 2012; Hung et al., 2013; Sanjaya et al., 2013) or

by heterologous expression in Arabidopsis (Sanjaya et al.,

2013). An in vitro assay to dissect the substrate specificity

of CrDGTT2 and CrDGTT3 was performed using radiola-

beled C16:0 and C18:1 CoAs, but specificity with respect to

the other substrate, DAG, was not explored (Sanjaya et al.,

2013). While these studies confirmed the acyltransferase

activity of CrDGTT1, CrDGTT2 and CrDGTT3, their substrate

specificity beyond C16:0 and C18:1 CoAs remains to be

studied. Over-expression and knockdown of CrDGTT1,

CrDGTT2 or CrDGTT3 in C. reinhardtii have also been

attempted (Deng et al., 2012; La Russa et al., 2012; Iwai

et al., 2014); however, changes in TAG content and fatty

acid composition were not characterized.

In higher plants, glycerolipids with a C18 acyl group in

the sn-2 position are assembled in the endoplasmic reticu-

lum (ER) via the eukaryotic pathway, whereas those with a

C16 acyl group in sn-2 are assembled in plastids via the

prokaryotic pathway (Ohlrogge and Browse, 1995). The

prokaryotic pathway has been suggested to mediate TAG

synthesis and lipid droplet (LD) formation in the chloro-

plast of C. reinhardtii (Fan et al., 2011; Goodson et al.,

2011; Tsai et al., 2015). In this study, we developed a radio-

label-free in vitro assay that revealed the different sub-

strate specificities of CrDGTT1, CrDGTT2 and CrDGTT3

with respect to both acyl CoAs and DAGs. We also gener-

ated individual CrDGTT knockdown lines of these three

genes using an artificial microRNA-mediated approach,

and investigated the functions of these enzymes in vivo via

lipidomic analysis. CrDGTT1 was found to prefer prokary-

otic lipid substrates and to probably reside in the ER and

chloroplast envelope, indicating its role in prokaryotic TAG

biosynthesis. We discuss the implications of these results

for TAG metabolism in microalgae, and describe potential

biotechnological applications.

RESULTS

Cloning of CrDGAT genes and functional analysis in yeast

Using quantitative real-time PCR, we confirmed that nitro-

gen (N) deprivation triggered the up-regulation of

CrDGAT1 (Cre01.g045900), CrDGTT1 (Cre12.g557750),

CrDGTT2 (Cre02.g121200), CrDGTT3 (Cre06.g299050),

CrDGTT4 (Cre03.g205050), CrDGTT5 (Cre02.g079050) and

CrDGAT3-like (Cre06.g310200) at the mRNA level (Fig-

ure S1), in agreement with previous RNA-seq data (Miller

et al., 2010; Blaby et al., 2013; Goodenough et al., 2014).

Up-regulation of CrDGAT genes under N deprivation coin-

cided with a drastic increase in TAG content, from a basal

level of 3 nmol mg�1 dry weight on day 0 to 146 nmol g�1

on day 4 (Figure S2a).

We next introduced these six CrDGAT genes into the

TAG-deficient Saccharomyces cerevisiae strain H1246

(Ddga1Dlro1Dare1Dare2). Expression of CrDGTT1, CrDGTT2and CrDGTT3 restored TAG biosynthesis, as indicated by

prominent TAG spots on a TLC plate of the lipid extracts

from the transformed yeast cells, but a TAG spot was not

observed for the control expressing the empty vector (Fig-

ure S3a). The CrDGTT1-expressing strain produced the

highest level of TAG, followed by strains expressing

CrDGTT2 and CrDGTT3. Yeast strains expressing CrDGAT1,

CrDGTT4 or CrDGAT3-like did not produce any detectable

TAG.

© 2016 The AuthorsThe Plant Journal © 2016 John Wiley & Sons Ltd, The Plant Journal, (2016), 86, 3–19

4 Jin Liu et al.

To test whether CrDGTTs are active on polyunsaturated

fatty acids (PUFAs) present in C. reinhardtii but not in

yeast, we assessed the growth response of CrDGTT-

expressing S. cerevisiae H1246 to feeding with free fatty

acids (FFAs). H1246 cells lacking DGAT activity did not effi-

ciently metabolize FFA, and FFA feeding severely impaired

the growth of these cells; by contrast, heterologous expres-

sion of DGATs in H1246 cells enabled them to import FFA

for sequestration into TAG, thus leading to FFA detoxifica-

tion and restoration of cell growth (Siloto et al., 2009). We

examined the response of CrDGTT-expressing H1246 cells

to feeding with four FFAs: C16:1n7, C18:1n9, C18:2 and

C18:3n3 (Figure S3b). Similar to the control cells express-

ing the empty vector, the H1246 cells expressing CrDGAT1,

CrDGTT4 or CrDGAT3-like showed no growth in the pres-

ence of any of these FFAs, further supporting the likelihood

that the encoded enzymes have no acyltransferase activity.

By contrast, H1246 cells expressing CrDGTT1, CrDGTT2 or

CrDGTT3 exhibited various growth responses to FFA feed-

ing. When fed with C16:1n7, CrDGTT1-expressing yeast

cells exhibited the highest growth rate, indicating that

CrDGTT1 probably has greater activity than CrDGTT2 and

CrDGTT3 toward C16:1n7 CoA. When fed with C18:1n9,

CrDGTT2-expressing yeast cells accumulated the most bio-

mass, suggesting that CrDGTT2 had the highest activity

toward C18:1n9 CoA. As CrDGTT1-expressing H1246 cells

produced more biomass when supplemented with either

C18:2 or C18:3n3 than did CrDGTT2-expressing cells,

CrDGTT1 may have higher activity than CrDGTT2 toward

C18:2 and C18:3n3 CoAs. Feeding with C18:2 or C18:3n3

completely blocked cell growth of CrDGTT3-expressing

H1246 cells, suggesting that CrDGTT3 cannot use these

two acyl CoAs.

Acyl CoA substrate specificity of CrDGTT1, CrDGTT2 and

CrDGTT3

Over 30 major TAG species are present in C. reinhardtii

(Yoon et al., 2012), suggesting that the DGAT isoforms

have different substrate preferences. To test this possibil-

ity, we developed an in vitro radiolabel-free DGAT assay to

quantitatively measure the activity and substrate specificity

of CrDGTTs toward a wide range of acyl CoAs and DAGs

with minimum background signals (Figure S4). The

amount of TAG product showed a good linear relationship

with the acyl CoA concentrations (62.5–250 lM) under ourtested conditions (Figure S5). Compared to the radiola-

beled in vitro DGAT assay, our radiolabel-free assay was

less sensitive and required a higher level of TAG products

for DGAT activity determination. Thus, we used a higher

concentration of acyl CoA (250 lM) in our subsequentexperiments. As the sn-2 position of TAG contained pre-

dominantly C16:0, while the sn-1/sn-3 positions mainly

contained C18:1n9 (Figure S2b,c), we first used the

prokaryotic C18:1n9/C16:0 DAG as an acyl acceptor for the

in vitro assay. We examined ten acyl CoAs, including most

of the fatty acids present in the sn-1/sn-3 positions of TAG

(Figure S2b) and four fatty acids that are not present in

C. reinhardtii, i.e. c-linolenoyl CoA (C18:3n6(d6) CoA),eicosapentaenoyl CoA (C20:5n3 CoA), docosahexaenoyl

CoA (C22:6n3 CoA) and lignoceryl CoA (C24:0 CoA). The

microsomal protein fraction from the empty vector-expres-

sing H1246 cells, which produced no TAG, was used as a

negative control (Figure S6a).

Unsaturated acyl CoAs, particularly polyunsaturated acyl

CoAs, were the preferred substrates for CrDGTT1 (Fig-

ure 1a). CrDGTT1 had a considerably higher activity toward

C16:1 than C16:0. For C18 CoAs, CrDGTT1 had the greatest

activity toward C18:3n, followed by C18:2, C18:1n9 and

C18:0. CrDGTT1 preferred shorter-chain acyl CoAs when

the number of double bonds was the same; for instance,

C16:0 and C16:1 resulted in a greater level of TAG accumu-

lation than C18:0 and C18:1, respectively. CrDGTT1 was

also shown to have strong activity toward C20:5n3 and

C22:6n3 CoAs. By contrast, CrDGTT2 exhibited a preference

for monounsaturated acyl CoAs over saturated ones (Fig-

ure 1b). However, when supplied with acyl CoAs of the

same acyl chain length, CrDGTT2, unlike CrDGTT1, showed

only a slight preference for polyunsaturated acyl CoAs over

monounsaturated ones. CrDGTT2 showed comparable

activities toward C16:0 and C18:0 CoAs; for unsaturated

acyl CoAs, it had similar activity toward C16:1 and C18:1

CoAs. Interestingly, CrDGTT2 also showed strong activity

toward C20:5n3 and C22:6n3 CoAs. In contrast to CrDGTT1

and CrDGTT2, CrDGTT3 showed low activity toward

polyunsaturated acyl CoAs and preferred C16 CoAs, partic-

ularly C16:1 CoA, over C18 and other longer-chain fatty acyl

CoAs (Figure 1c). We also evaluated the preference of

CrDGTT1, CrDGTT2 and CrDGTT3 for acyl CoAs using

another eukaryotic DAG (C18:1n9/C18:1n9 DAG) as the acyl

acceptor, and observed similar results (Figure S7).

Quantitatively, CrDGTT1, CrDGTT2 and CrDGTT3 had

comparable enzymatic activities on C16:0 CoA, while

CrDGTT1 had the greatest activity toward C16:1 CoA com-

pared with other DGTTs (Figure 1d). CrDGTT2 showed

greater activity for C18:0 and C18:1n9 CoAs than did the

other two DGTTs. On the other hand, CrDGTT1 showed the

highest activity toward C18 polyunsaturated CoAs. These

data indicate that CrDGTT1 contributes predominantly to

incorporation of C18 polyunsaturated fatty acyl CoAs,

while CrDGTT2 contributes more to incorporation of C18:0

and C18:1n9 into TAG.

DAG substrate specificity of CrDGTT1, CrDGTT2 and

CrDGTT3

We further investigated the activity and preference of

CrDGTT1, CrDGTT2 and CrDGTT3 for various DAGs. A

microsomal fraction prepared from control H1246 cells

expressing the empty vector did not catalyze formation of


Diacylglycerol acyltransferases in Chlamydomonas 5

TAG from any of the DAGs tested (Figure S6b). When

using C18:3n6(d6) CoA as the acyl donor, CrDGTT1 showed

strong and comparable activity toward C16:1/C16:1,

C18:1n9/C16:0, C16:0/C18:1n9 and C18:1n9/C18:1n9 DAGs,

but weak or no activity toward C16:0/C16:0, 1,3-C18:1n9/

C18:1n9, C18:2/C18:2 and C18:3n3/C18:3n3 DAGs (Fig-

ure 2a). Interestingly, when the acyl donor was C16:0 or

C18:1n9 CoA, which represent the most abundant fatty

acids in the sn-1/sn-3 positions of Chlamydomonas TAGs

(Figure S2b), CrDGTT1 demonstrated considerably higher

activity toward C18:1n9/C16:0 DAG than toward eukaryotic

DAGs with C18 in the sn-2 position. These data indicate

that CrDGTT1 prefers prokaryotic DAGs as the acyl accep-

tor for TAG biosynthesis. CrDGTT2, unlike CrDGTT1, pre-

ferred the eukaryotic C16:0/C18:1n9 and C18:1n9/C18:1n9

DAGs over prokaryotic DAGs for TAG formation (Fig-

ure 2b). CrDGTT3 had slightly higher activity toward the

eukaryotic C16:0/C18:1n9 and C18:1n9/C18:1n9 DAGs than

toward C18:1n9/C16:0 DAG (Figure 2c).

Feeding with free fatty acids enhances TAG biosynthesis

and affects the fatty acid profile of Chlamydomonas

Incorporation of FFAs into TAG under N deprivation is cat-

alyzed primarily through the Kennedy pathway in C. rein-

hardtii. To determine how fatty acids were incorporated

into TAG in Chlamydomonas, and the enzymatic activity of

DGATs in vivo, we coupled feeding of exogenous FFAs to

C. reinhardtii to treatment with cerulenin, a b-ketoacyl ACPsynthase I inhibitor that is known to inhibit de novo fatty

acid biosynthesis in algae (Liu et al., 2012). In the presence

of 10 lM cerulenin, the TAG content decreased by morethan 80% in C. reinhardtii under N deprivation conditions

(Figure S8a). Cerulenin caused a drastic decrease in the rel-

ative content of C18:1n9 and a concomitant enrichment of

C16:4, C18:0, C18:1n11, C18:3n6, C18:3n3 and C18:4 (Fig-

ure S8b). This decrease was partially restored by FFA feed-

ing, with C18:1n9 being the most effective (Figure S8a).

Although the FFA feeding experiments did not rule out the

contribution of other acyltransferases, such as PDAT, to

TAG biosynthesis under N deprivation, previous studies

indicate that the primary route for TAG biosynthesis under

N deprivation is through the DGAT-catalyzed pathway

(Yoon et al., 2012). Thus, the restoration of TAG biosynthe-

sis following FFA feeding may primarily result from action

of the CrDGATs. FFA feeding led to an increase in not only

the fed FFA, but also of the desaturated forms of the

respective fed FFA. For example, C16:1n7 feeding led to a

drastic increase in the relative content of not only C16:1n7,

accounting for 35.6% of the fatty acids in TAG (TAG FAs),

but also an increase in C16:2 and C16:4 content (Fig-

ure S8b). As no desaturase is known or predicted to desat-

urate fatty acids linked to TAG (Li-Beisson et al., 2015),

some of the exogenous FFAs were probably first incorpo-

rated into polar lipids for further desaturation and then

released by lipases for subsequent TAG assembly (Fig-

ure S8b). We also observed incorporation of exogenous

FFAs into DAG (Figure S9).

Figure 1. The various substrate specificities of CrDGTT1, CrDGTT2 and

CrDGTT3 for acyl CoAs.

(a–c) TLC analysis of lipids resulting from in vitro enzymatic reactions ofCrDGTT1 (a), CrDGGT2 (b) and CrDGTT3 (c) with various acyl CoAs.

C18:1n9/C16:0 DAG was used as the acyl acceptor. C18:3n6(d6) represents

C18:3n6(D6,9,12). C1, microsome; C2, microsome plus 250 lM DAG.(d) Quantification of the specific activity of CrDGTT1, CrDGTT2 and

CrDGTT3 by GC-MS. Values are means � SD (n = 3).


6 Jin Liu et al.

It is worth noting that the increased accumulation of

fatty acids in TAG as a result of FFA feeding under N depri-

vation occurred not only in the sn-1/sn-3 positions (Fig-

ure S8c), but also in the sn-2 position of TAG (Figure S8d).

For example, C16:1n7 feeding led to a 45-fold increase in

the relative content of C16:1n7 in the sn-1/sn-3 positions of

TAG, and a sixfold increase in the sn-2 position of TAG,

indicating that the acyl CoAs from exogenous FFAs were

incorporated into TAG by acyltransferases, probably

DGATs, under N deprivation, in both the sn-1/sn-3 and sn-

2 positions. These data suggest that Chlamydomonas

DGATs use both eukaryotic and prokaryotic substrates for

TAG biosynthesis.

Cerulenin also caused a significant decrease in the con-

tent of polar membrane lipids, including monogalactosyl-

diacylglycerol (MGDG), digalactosyldiacylglycerol (DGDG),

diacylglycerol trimethylhomoserine (DGTS), phosophatidyl-

glycerol (PG) and phosphatidyl inositol (PI); by contrast, the

level of sulfurquinovosyl diacylglycerol (SQDG) was barely

altered (Figure S8e). FFA feeding restored polar lipid

biosynthesis in cerulenin-treated cells to various degrees

(Figure S8e). Among the four tested FFAs (i.e., C16:1n7,

Figure 2. The various specificities of CrDGTT1, CrDGTT2, and CrDGTT3 for DAGs.

TLC analysis of lipids resulting from in vitro enzymatic reactions of CrDGTT1 (a), CrDGTT2 (b) and CrDGTT3 (c) with various DAGs (1, C16:0/C16:0; 2, C16:1/

C16:1; 3, C18:1n9/C16:0; 4, C16:0/C18:1n9; 5, C18:1n9/C18:1n9; 6, 1,3-C18:1n9/C18:1n9; 7, C18:2/C18:2; 8, C18:3n3/C18:3n3). The acyl CoA donor used in each reac-

tion is indicated on the panel. The numbers above each panel indicate the relative TAG values, normalized to those obtained when C18:1n9/C16:0 DAG was used

as the acyl acceptor (set as 100).



C18:1n9, C18:2 and C18:3n3), only C18:3n3 restored MGDG

levels (Figure S8e). By contrast, all four FFAs resulted in an

increase in DGDG (Figure S8e), and altered its fatty acid

profile in an FFA-specific manner (Figure S10). For example,

we observed a considerable increase in the relative con-

tent of C16:1n7 (32-fold increase) in DGDG upon C16:1n7

feeding.

CrDGTT1 knockdown suppresses TAG accumulation and

modifies polyunsaturated fatty acid composition

To investigate the function and possible biological role of

individual DGATs in C. reinhardtii, we used an artificial

microRNA-mediated gene silencing approach (Molnar

et al., 2009) to generate CrDGTT knockdown lines. To

enhance knockdown vector expression and increase knock-

down efficiency under N deprivation conditions, we

replaced the original psaD promoter with a hybrid constitu-

tive HSP70/RBCS2 promoter to drive expression of the arti-

ficial microRNA sequence (Figure S11). This increased the

knockdown efficiency, in some cases from approximately

50 to 85%.

Two CrDGTT1 knockdown lines, CrDGTT1-11i and

CrDGTT1-13i, exhibited a reduction in CrDGTT1 transcripts

of over 70% under both N-replete and N-deprived condi-

tions (Figure 3a). CrDGTT1 knockdown was also confirmed

at the protein level by immunoblotting (Figure 3b). No dif-

ference in TAG content was observed between the control

and CrDGTT1 knockdown lines under N-replete conditions

(Figure 3c, day 0). Upon stress induction, the TAG content

in CrDGTT1-11i and CrDGTT1-13i decreased by 27.0 and

34.3%, respectively, compared to the control (day 1). As N

deprivation stress persisted (days 2–4), the significant TAGreduction in the knockdown lines remained prominent.

Notably, CrDGTT1-11i and CrDGTT1-13i exhibited an

increase in the membrane lipids MGDG and PG but a

decrease in SQDG. By contrast, DGDG, DGTS and PI

showed no significant difference between the control and

knockdown lines (Figure 3d).

We further studied the effect of CrDGTT1 knockdown on

TAG fatty acid composition. When compared to the con-

trol, TAG showed a significant decrease in the relative con-

tent of C16:1, C18:1n9, C18:2, C18:3n6 and C18:3n3 and an

increase in C16:4, C18:0 and C18:1n7 in the CrDGTT1

knockdown lines under N deprivation conditions (Figure 3e

and Figure S12). A similar decrease in the relative content

of these fatty acids in the sn-1/sn-3 positions of TAG was

observed (Figure 3f). In the sn-2 position of TAG, the rela-

tive content of C16:0, C16:1 and C18:1n9 was significantly

decreased in the CrDGTT1 knockdown line (Figure 3g),

suggesting that CrDGTT1 was actively involved in both

prokaryotic and eukaryotic TAG biosynthesis. The knock-

down lines showed a decrease in nearly all fatty acids per

cell dry weight, but most significantly in C16:1, C18:1n9,

C18:2, C18:3n6 and C18:3n3 (Figure S13). These in vivo

results are in good agreement with the in vitro CrDGTT1

assay data (Figure 1a,d), further supporting the preference

of CrDGTT1 for unsaturated acyl CoAs, particularly PUFAs.

No significant difference in the fatty acid composition of

the six classes of polar lipids (i.e. MGDG, PG, SQDG,

DGDG, DGTS and PI) was observed between the control

and knockdown line (Figure S14).


modifies unsaturated fatty acid composition


CrDGTT2-45i, exhibited a >65% decrease in CrDGTT2 tran-scripts under both N-replete and N-deprived conditions

(Figure 4a). CrDGTT2 knockdown lines showed little differ-

ence in TAG content compared with the control under N-

replete conditions (Figure 4a, day 0). However, under N-

deprived conditions, we observed 23% and 20% decreases

in TAG content in CrDGTT2-42i and CrDGTT2-45i, respec-

tively, on day 1, and the reduction remained significant,

but less pronounced, on days 2–4 (Figure 4b). CrDGTT2knockdown increased the content of the major thylakoid

membrane lipids MGDG, DGDG and PG, but decreased the

SQDG content (Figure 4c), similar to our findings for the

CrDGTT1 knockdown lines.

The CrDGTT2 knockdown lines had less C16:1, C18:1n9,

C18:3n6 and C18:3n3 in TAG under N-deprived conditions

than did the control (Figures 4d and S15). These fatty acids

were also less abundant in the sn-1/sn-3 positions of TAG

in the CrDGTT2 knockdown lines under N deprivation (Fig-

ure 4e). The relative content of C16:0, C16:1 and C18:1n9 in

the sn-2 position of TAG was significantly attenuated in

the CrDGTT2 knockdown line (Figure 4f). The contents of

all types of TAG fatty acids per dry cell weight decreased

significantly, with C16:1 and C18:1n9 decreasing to the

greatest extent (Figure S16). Together, these results sug-

gest that CrDGTT2 shows a substrate preference for unsat-

urated acyl CoAs during TAG biosynthesis in vivo

(Figure 1b,d).


modifies saturated and monounsaturated fatty acid

composition


CrDGTT3-15i, exhibited a decrease in CrDGTT3 transcripts

by over 80% compared to the control under both

N-replete and N-deprived conditions (Figure 5a). Under

N-deprived conditions, the TAG content of CrDGTT3-3i

and CrDGTT3-15i cells decreased significantly compared

to the control (Figure 5b). CrDGTT3 knockdown caused

an increase in MGDG and PG content and a decrease in

SQDG (Figure 5c). Under both N-replete and N-deprived

conditions, the CrDGTT3 knockdown mutants showed a

significant decrease in the relative content of C16:0,

C16:1 and C18:1n9 in TAG compared to the control


8 Jin Liu et al.

Figure 3. CrDGTT1 knockdown suppressed TAG biosynthesis and altered TAG fatty acid profiles.

(a) DGTT1 mRNA levels in two CrDGTT1 knockdown lines compared with a control transformed with the empty vector (control, Ctr), as determined by quantita-

tive real-time PCR. The relative expression level of DGTT was normalized to that of the gene encoding b-actin.(b) CrDGTT1 protein expression level in the knockdown lines, as determined by immunoblotting. Protein concentration (represented by numbers above the

blots) was determined using ImageJ software (http://rsb.info.nih.gov/ij/index.html), and normalized to the concentration obtained for the control, which was set

as 1.

(c) TAG content under N-replete and N-deprived conditions in the control and CrDGTT1 knockdown lines. Cells grown under N-replete conditions (day 0) were

transferred to TAP-N medium for 4 days.

(d) Polar lipid contents in the control and CrDGTT1 knockdown lines under N-deprived conditions (day 2).

(e) Relative abundance of TAG fatty acid groups in the control and CrDGTT1 knockdown lines under N-deprived conditions (day 2).

(f,g) Relative abundance of fatty acid groups in the sn-1/sn-3 (f) and sn-2 (g) positions of TAG in the control and CrDGTT1-13i knockdown line under N-deprived

conditions (day 2).

Values are means � SD (n = 3). Asterisks indicate statistically significant differences compared with the control (t test, P < 0.05).



http://rsb.info.nih.gov/ij/index.html

(Figures 5d and S17). The decrease in these fatty acids

was also evident in the sn-1/sn-3 positions of TAG (Fig-

ure 5e). The relative content of C16:0, C16:1 and C18:1n9

in the sn-2 position of TAG was significantly attenuated

in the CrDGTT3 knockdown line (Figure 5f). Furthermore,

the contents of C16:0, C16:1 and C18:1n9 per cell dry


(a) CrDGTT2 mRNA levels in two CrDGTT2 knockdown lines compared with the control (Ctr), as determined by quantitative real-time PCR.

(b) TAG content under N-replete and N-deprived conditions in the control and CrDGTT2 knockdown lines.

(c) Polar lipid contents in the control and CrDGTT2 knockdown lines under N-deprived conditions (day 2).

(d) Relative abundance of TAG fatty acid groups in the control and CrDGTT2 knockdown lines under N-deprived conditions (day 2).

(e,f) Relative abundance of fatty acid groups in the sn-1/sn-3 (e) and sn-2 (f) positions of TAG in the control and CrDGTT2-42i knockdown line under N-deprived

conditions (day 2).



10 Jin Liu et al.

weight in TAG of the CrDGTT3 knockdown line were con-

siderably lower than in the control, whereas CrDGTT3

knockdown had little effect on the abundance of PUFAs

such as C18:2, C18:3n3 and C18:3n6 (Figure S18). These

in vivo data are in good agreement with the findings

of our in vitro assay (Figure 1c,d), indicating that

CrDGTT3 has a substrate preference for C16 saturated

and monounsaturated fatty acids.


(a) CrDGTT3 mRNA levels in two CrDGTT3 knockdown lines compared with the control (Ctr), as determined by quantitative real-time PCR.

(b) TAG content under N-replete and N-deprived conditions in the control and CrDGTT3 knockdown lines.

(c) Polar lipid contents in the control and CrDGTT3 knockdown lines under N-deprived conditions (day 2).

(d) Relative abundance of TAG fatty acid groups in the control and CrDGTT3 knockdown lines under N-deprived conditions (day 2).

(e,f) Relative abundance of fatty acid groups in the sn-1/sn-3 (e) and sn-2 (f) positions of TAG in the control and CrDGTT3-15i knockdown line under N-deprived

conditions (day 2).




CrDGTT1 is considerably up-regulated under nitrogen

deprivation

CC-4348 (sta6), a starch-null mutant that is defective in the

gene encoding an ADP-glucose pyrophosphorylase sub-

unit, produces more TAG than the wild-type strain (Li

et al., 2010; Fan et al., 2012; Blaby et al., 2013; Goode-

nough et al., 2014). We examined CrDGAT expression and

TAG levels in CC-4348 and the reference strain CC-503 (Fig-

ure 6). Our quantitative real-time PCR results showed that

CC-4348 had higher transcript levels of CrDGTT1, CrDGTT2

and CrDGTT3 than did CC-503 (Figures 6a and S19), coinci-

dent with a 100% increase in TAG content in CC-4348 rela-

tive to CC-503 (Figure 6b). Upon N deprivation, CrDGTT1

transcript levels increased by up to 54-fold, while those for

CrDGTT2 and CrDGTT3 increased by only three- to fivefold.

At the translational level, CrDGTT1 protein abundance

increased upon N deprivation, with a 2.9-fold increase at

0 h and a 14.9-fold increase at 48 h for CC-4348 compared

with CC-503 at 0 h, and a 3.8-fold increase for CC-503 at

48 h compared with CC-503 at 0 h (Figure 6c). Interest-

ingly, when the fatty acid profile of TAG was assessed in

N-deprived CC-503 and CC-4348 cells, the latter were found

to have a significantly higher relative abundance of C16:1,

C18:1n9, C18:2 and C18:3n3 (Figure S20), consistent with

the acyl CoA specificity of CrDGTT1 (Figure 1a) and the

higher expression level of CrDGTT1 in CC-4348 (Figure 6a).

CrDGTT1 is probably located in the ER and chloroplast

envelope

Type 2 DGATs are thought to catalyze TAG assembly in

the ER in yeast and higher plants (Sorger and Daum, 2003;

Chapman and Ohlrogge, 2012). In contrast to TAGs in

higher plants, Chlamydomonas TAGs contained up to 85%

C16 fatty acids in the sn-2 position (Figure S2c). Similarly,

DAGs also contained predominantly C16:0 in the sn-2 posi-

tion (over 50%, Figure S21), suggesting that TAGs are

probably synthesized through both the prokaryotic and

eukaryotic pathways in C. reinhardtii, with the former con-

tributing more to TAG biosynthesis. Consistent with this

possibility, CrDGTT1 has a preference for C16 acyl groups

over C18 acyl groups in the sn-2 position of DAGs (Fig-

ure 2a). These results suggest that CrDGTT1 is involved in

prokaryotic TAG biosynthesis, and call into question its

proposed exclusive localization to the ER. Bioinformatic

analysis using TargetP 1.1 (http://www.cbs.dtu.dk/services/

TargetP/) and SignalP 4.1 (http://www.cbs.dtu.dk/services/

SignalP/) failed to predict the subcellular localization of

CrDGTT1. To examine the subcellular localization of

CrDGTT1 experimentally, we prepared subcellular fractions

from CC–503 enriched with the ER, mitochondria (MTC),chloroplast envelope (CpE), thylakoid membrane (TM) and

LDs. To check for possible cross-contamination among

these fractions, we used four markers (binding immuno-

Figure 6. CrDGTT1 was up-regulated upon N deprivation, and probably localized to the ER and chloroplast envelope (CpE).

(a) CrDGTT1 mRNA levels in CC-503 and CC-4348 cells under N-deprived conditions.

(b) TAG content in CC-503 and CC-4348 cells under N-deprived conditions.

(c) CrDGTT1 protein levels in CC-503 and CC-4348 cells, as determined by immunoblot analysis. The CrDGTT1 control was derived from the crude protein frac-

tion of CrDGTT1-expressing H1246 yeast cells. Protein concentration was normalized to the CrDGTT1 expression level in CC-503 at 0 h, which was set as 1.

(d) Subcellular localization of CrDGTT1. CC-503 cells grown under N-deprived conditions for 1 day were used for organelle preparation. MTC, mitochondria; ER,

endoplasmic reticulum; LD, lipid droplet; CpE, chloroplast envelope; TM, thylakoid membrane. LHCb1, AOX, BIP and Toc34 were used as markers for the TM,

MTC, ER and CpE, respectively.


12 Jin Liu et al.

http://www.cbs.dtu.dk/services/TargetP/http://www.cbs.dtu.dk/services/TargetP/http://www.cbs.dtu.dk/services/SignalP/http://www.cbs.dtu.dk/services/SignalP/

globulin protein (BIP), light-harvesting complex b1 subunit

(LHCb1), alternative oxidase (AOX), and translocase of

chloroplast 34 (Toc34)), which specifically target the ER,

TM, MTC and CpE fractions, respectively. Only the MTC

fraction showed signs of contamination with trace

amounts of TM and ER (Figure 6d). Immunoblot analysis

using the CrDGTT1 antibody revealed that CrDGTT1 was

enriched in the ER and CpE fractions, indicating that

CrDGTT1 resides in both the ER and CpE (Figure 6d). As

no visible contamination of the CpE fraction by ER or TM

was observed, the CrDGTT1 signal in the CpE fraction was

unlikely to be due to contamination with ER or TM. The

intracellular localization of CrDGTT1 requires further confir-

mation using GFP and in situ immunolocalization

approaches. Unfortunately, our analysis of CrDGTT1–GFPfusions (including mCherry and Clover vectors) failed to

reveal any significant florescent signal in wild-type

Chlamydomonas or the UVM4 strain (Neupert et al., 2009)

that shows improved transgene expression efficiency,

probably due to the low expression level of membrane-

bound proteins such as DGTT1, as also found for the galac-

toglycerolipid lipase PLASTID GALACTOGLYCEROLIPID

DEGRADATION1 (PGD1) (Li et al., 2012). We also

attempted to determine the subcellular localization of

CrDGTT2 and CrDGTT3, but did not detect these proteins,

as the antibodies raised against CrDGTT2 and CrDGTT3

failed to work.

DISCUSSION

CrDGTT1, CrDGTT2 and CrDGTT3 have distinct

acyltransferase activities and substrate specificities

Diacylglycerol acyltransferases catalyze the only committed

step of TAG biosynthesis, and have been suggested to be

rate-limiting enzymes in TAG biosynthesis in higher plants

(Ohlrogge and Jaworski, 1997; Lung and Weselake, 2006).

Unlike higher plants, many microalgae possess a relatively

large number of DGAT genes encoding type 2 DGATs.

Whereas the Arabidopsis genome harbors only one DGAT2

gene, C. reinhardtii has five DGAT genes (Miller et al.,

2010; Boyle et al., 2012), Phaeodactylum tricornutum has

four (Gong et al., 2013), Chlorella pyrenoidosa has five (Fan

et al., 2014), and Nannochloropsis oceanica has eleven

(Vieler et al., 2012; Wang et al., 2014). The presence of mul-

tiple DGAT2 genes may underlie the biosynthesis of large

numbers of TAG species in microalgae.

Four independent studies including the present study

have been performed to characterize C. reinhardtii DGAT

genes by functional complementation in various TAG-defi-

cient yeast strains (Boyle et al., 2012; Hung et al., 2013;

Sanjaya et al., 2013). While the previous studies confirmed

that CrDGTT1, CrDGTT2 and CrDGTT3 had acyltransferase

activity and contributed to TAG biosynthesis in yeast, we

provided further evidence that CrDGTT1, CrDGTT2 and

CrDGTT3 possess distinct specificities toward various acyl

CoAs and DAGs.

The availability of limited numbers of radiolabeled sub-

strates and the complexity of performing radiolabeling

experiments are the main limitations for performing

in vitro DGAT assays. Greer et al. (2014) developed a radio-

label-free method for DGAT activity determination using a

short-chain DAG (dioctanoyl DAG) as the acyl acceptor.

Based on high-temperature GC-MS, this method eliminates

the use of labeled substrates and TLC, and is thus time-

saving and cost-effective, but is limited to use of dioc-

tanoyl DAG as the acyl acceptor. In our in vitro assay, we

used non-radiolabeled acyl CoAs and DAGs to evaluate

DGAT activity for a wide range of substrates present in

C. reinhardtii, including PUFAs and medium-chain DAGs

that had not previously been evaluated. Compared to the

radiolabeled in vitro DGAT assay, our TLC-based radiola-

bel-free assay may be used to test a wider range of sub-

strates, but is less sensitive and requires a higher level of

TAGs for DGAT activity determination. To increase the

amount of TAG product, we used a relatively high acyl

CoA concentration (250 lM, Figure S5), which, althoughhigher than that used in the radiolabeled assays

(15–50 lM), was still within the linear range under ourexperimental conditions (Figure S5) and did not introduce

artifacts.

In some previous DGAT in vitro assays, detergent was

used to dissolve DAG (Xu et al., 2008; Wagner et al., 2010;

Greer et al., 2015). In other studies, no detergent was used

and DAG was added to the assay mixture from an alcohol

stock (Milcamps et al., 2005; Shockey et al., 2006). Simi-

larly, in our assays, after addition of microsomal protein to

the assay mixture, we delivered DAG from a 50 mM stock

in ethanol by vigorous shaking. We also tested the inclu-

sion of Tween-20 (0.05–0.2% v/v) in the assay mixture, andour results suggest that this treatment did not significantly

improve DGAT activity. Using our optimized assay

method, we demonstrated that CrDGTT1, CrDGTT2 and

CrDGTT3 had distinct substrate preferences for acyl CoAs

with different chain lengths and numbers of double bonds.

We also tested eight DAGs, and were able to distinguish

DGAT preference with respect to prokaryotic DAGs (where

sn-2 contains C16) and eukaryotic DAGs (where sn-2 con-

tains C18): CrDGTT1 preferred prokaryotic DAGs, whereas

CrDGTT2 and CrDGTT3 preferred eukaryotic DAGs (Fig-

ure 2). The distinct substrate specificities of CrDGTTs on

acyl CoAs and DAGs suggest that the multiple DGAT genes

in Chlamydomonas may work in concert to use a wide

range of substrates for biosynthesis of diverse TAG species

in this microalga.

Intriguingly, CrDGTT1 and CrDGTT2 were highly active

toward C20:5 and C22:6 CoAs (Figure 1). CrDGTT2 also

exhibited activity toward C22:1 CoA, as indicated by a

competition assay (Sanjaya et al., 2013). Our work quanti-



tatively determined the specific activity of DGAT toward

C20:5 CoA. CrDGTT1 and CrDGTT2 catalyzed the produc-

tion of approximately 200 nmol TAG per mg protein per

hour, indicating a relatively high activity of these two

DGTTs toward 20:5 CoA, given that reported DGAT activi-

ties range from 1 to 300 nmol TAG formed per mg protein

per hour (Oelkers et al., 2002; Shockey et al., 2006; Stone

et al., 2006; Xu et al., 2008; Liu et al., 2011; Zhang et al.,

2013; Greer et al., 2015).

The differences in substrate specificity of CrDGTT1,

CrDGTT2 and CrDGTT3 may be attributable to amino acid

diversification in the functional domains of these proteins.

Bioinformatics analysis revealed the presence of multiple

conserved motifs, including the YFP and HPHG motifs

(Stone et al., 2006; Liu et al., 2011), in the DGAT2 family

(Cao, 2011; Chen and Smith, 2012). Among CrDGTT1,

CrDGTT2 and CrDGTT3, only CrDGTT1 possesses the

HPHG motif; CrDGTT2 and CrDGTT3 contain a variant of

this motif of the form FPHG (Hung et al., 2013). The varia-

tion in the first amino acid residue of this motif may con-

tribute to the different catalytic activity of CrDGTT1,

CrDGTT2 and CrDGTT3, as observed for S. cerevisiae Dga1

and mouse DGAT2 (Stone et al., 2006; Liu et al., 2011).

CrDGTT1, CrDGTT2 and CrDGTT3 are critical for stress-

associated TAG biosynthesis in Chlamydomonas

Individual knockdown of CrDGTT1, CrDGTT2 or CrDGTT3

led to statistically significant 20–35% decreases in TAGcontent under N deprivation conditions (Figures 3c, 4b and

5b), indicating the critical and non-redundant roles of these

CrDGTT genes in stress-induced TAG biosynthesis in

Chlamydomonas. By contrast, knockdown of CrDGTT1,

CrDGTT2 or CrDGTT3 did not significantly change the TAG

content under N-replete conditions. Considering that

CrPDAT knockdown with an efficiency of 65% resulted in a

58% decrease in TAG content under N-replete conditions

(Yoon et al., 2012), CrPDAT but not CrDGTT is probably the

major contributor to TAG biosynthesis under favorable

conditions.

Consistent with the in vitro assay results (Figures 1 and

2), knockdown of CrDGTT1, CrDGTT2 and CrDGTT3 indi-

vidually led to different changes in TAG content and fatty

acid profiles (Figures 3–5). Taking the in vitro and in vivodata together, we infer that CrDGTT1 is the main enzyme

catalyzing the incorporation of C18 PUFAs into TAG,

CrDGTT2 is active toward unsaturated C16 and C18 acyl

CoAs, and CrDGTT3 prefers C16 saturated and monounsat-

urated fatty acids as substrates. The distinct substrate

specificities of CrDGTTs may lead to accumulation of vari-

ous TAG species in C. reinhardtii under N deprivation.

Over-expression of CrDGTT genes in C. reinhardtii has

been used to dissect their function in vivo. However, previ-

ous studies over-expressing CrDGTT1, CrDGTT2 or

CrDGTT3 individually failed to show any appreciable differ-

ence in TAG content, either when expression of the trans-

genes was driven by a PsaD promoter under favorable

growth conditions (La Russa et al., 2012) or by a SQD2 pro-

moter under phosphorus starvation conditions (Iwai et al.,

2014). It is possible that over-expression of an endogenous

gene for TAG biosynthesis is subjected to certain regula-

tory loops or negative feedback inhibition in C. reinhardtii.

This possibility is partially supported by the fact that

heterologous expression of a Brassica napus DGAT2 gene

in C. reinhardtii promoted TAG accumulation and altered

its fatty acid composition (Ahmad et al., 2014). Interest-

ingly, over-expression of CrDGTT4 led to an increase in

TAG accumulation in C. reinhardtii (Iwai et al., 2014), sug-

gesting that the failure of CrDGTT4 to rescue TAG biosyn-

thesis in H1246 is probably because either its activity is

below our detection limit or certain enzyme co-factors are

not present in yeast. As CrDGTT4 transcripts were much

less abundant than CrDGTT1, CrDGTT2 and CrDGTT3 tran-

scripts (Figure S1), over-expression of CrDGTT4 may be

subject to less feedback regulation, thus resulting in

enhanced TAG biosynthesis. It is worth noting that intro-

ducing CrDGTT2 into Arabidopsis increased TAG accumu-

lation in leaves and the relative abundance of unsaturated

C18 fatty acids in TAG species (Sanjaya et al., 2013), in

agreement with our results that CrDGTT2 is active toward

unsaturated C18 CoAs (Figure 1).

C16 fatty acids are the main acyl groups found in the sn-

2 position of Chlamydomonas TAGs (Figure S2c), suggest-

ing that most DAG backbones for TAG biosynthesis are

derived from the prokaryotic pathway in the chloroplast.

Recent studies revealed the presence of plastidic LDs in

addition to cytosolic LDs in the sta6 mutant (Fan et al.,

2011; Goodson et al., 2011), and showed that LDs con-

tained certain plastid membrane proteins and lipids (Tsai

et al., 2015). These data suggest that a prokaryotic pathway

for TAG biosynthesis exists in the C. reinhardtii chloro-

plast. This was further supported by our in vitro and

in vivo data showing that CrDGTT1, CrDGTT2 and

CrDGTT3 accepted both plastid- and ER-derived DAGs for

TAG assembly. In particular, CrDGTT1 prefers prokaryotic

DAGs over eukaryotic ones, supporting its role in prokary-

otic TAG biosynthesis. Our immunoblot analysis with sub-

cellular fractions revealed that CrDGTT1 resided in both

the ER and CpE (Figure 6d). LDs, the sink for TAG storage,

had no detectable CrDGTT1 protein, consistent with previ-

ous proteomics studies in which no DGAT was found in

the LD fraction (Moellering and Benning, 2010; Nguyen

et al., 2011). Taken together, we propose a working model

for the role of CrDGTT1 in TAG biosynthesis under N depri-

vation conditions (Figure 7). Fatty acids from de novo

biosynthesis or membrane lipid turnover pathways consti-

tute the acyl CoA pool, and enter the Kennedy pathway for

biosynthesis of prokaryotic and eukaryotic DAGs. CrDGTT1

probably resides in both the CpE and ER, and accesses


14 Jin Liu et al.

these DAGs and acyl CoAs to synthesize prokaryotic and

eukaryotic TAG species packed into ER-derived and possi-

bly plastid-derived LDs. In the current model, we cannot

exclude the possibility that acyl CoAs and DAGs of

different origins may be translocated between the ER and

chloroplast by diffusion along the LD-delimiting mono-

layer, and used by CrDGTT1 for TAG assembly, as hypoth-

esized by Goodson et al. (2011). The exact mechanisms for

dual targeting of CrDGTT1 remain to be explored, as cur-

rent prediction programs such as TargetP 1.1 and SignalP

4.1 that are trained on sets of plant genes fail to identify

any signal peptide in this protein. Interestingly, a recent

study by Manandhar-Shrestha and Hildebrand (2015) sug-

gested that Thalassiosira pseudonana DGAT2 was also

present in two subcellular organelles, shifting from the

chloroplast in the exponential growth phase to the ER in

the stationary growth phase.

TAG biosynthesis interacts with membrane lipid turnover

in Chlamydomonas

Knockdown of CrDGTTs led to an increase in MGDG con-

tent with a concomitant decrease in TAG content (Fig-

ures 3d, 4c and 5c), suggesting a role for MGDG turnover

in TAG biosynthesis (Li et al., 2012; L�egeret et al., 2016).

MGDG is probably hydrolyzed by galactolipases that are

up-regulated during MGDG degradation under N depriva-

tion conditions (Li et al., 2012; Yoon et al., 2012; Gargouri

et al., 2015; L�egeret et al., 2016). PGD1, a galactolipase of

C. reinhardtii, hydrolyzed MGDG with a preference for the

sn-1 position, and its disruption led to attenuated TAG

biosynthesis (Li et al., 2012), supporting a role for MGDG

turnover in TAG biosynthesis. It is worth noting that PGD1

acts on de novo-synthesized MGDG (which predominantly

contains molecular species 18:1n9/16:0 (sn-1/sn-2)), but not

on the mature species which predominantly contains

molecular species 18:3n3/16:4 (sn-1/sn-2), and releases

C18:1n9 as the main product (Li et al., 2012). Consistent

with this, in our FFA feeding experiment, C18:1n9 feeding

did not affect the MGDG content or its fatty acid profile

(Figures S8e and S10), presumably because PGD1

specifically degrades MGDG species synthesized from

C18:1n9-enriched DAG, leading to MGDG homeostasis and

accumulation of TAG enriched in C18:1n9. Feeding of

C18:3n3, on the other hand, increased the MGDG content

in Chlamydomonas, possibly because PGD1 lacks function

toward highly unsaturated MGDG species.

As Chlamydomonas TAGs contain the signature fatty

acids of MGDG (C16:4 and C18:3n3), certain unknown

lipases may mediate turnover of mature MGDG (C18:3n3/

C16:4) and release these highly unsaturated acyl groups

for TAG assembly. CrPDAT is a multi-functional enzyme

with proven galactolipase function (Yoon et al., 2012). Sim-

ilar to PGD1, CrPDAT prefers MGDG to DGDG as a sub-

strate for release of acyl groups. However, whether PDAT

hydrolyzes mature MGDG species remains unknown. A

pathway mediated by a galactolipid:galactolipid galactosyl-

transferase may also play a role in MGDG remodeling

(Benning and Ohta, 2005; Moellering and Benning, 2011).

The galactolipid:galactolipid galactosyltransferase transfers

a galactosyl moiety from an MGDG molecule to a second

galactolipid molecule (MGDG or DGDG), resulting in for-

mation of highly unsaturated DAG for TAG biosynthesis in

Arabidopsis. It would be interesting to characterize galac-

tolipid:galactolipid galactosyltransferase and confirm its

activity toward MGDG in microalgae.

Biotechnological implications

In this study, we demonstrated that CrDGTT1, CrDGTT2

and CrDGTT3 catalyzed formation of various TAG species

in vitro, and contributed to the biosynthesis and accumula-

tion of diverse TAG species in C. reinhardtii in response to

N deprivation. As C. reinhardtii contains predominantly

prokaryotic TAGs, and CrDGTT1, but not CrDGTT2 or

CrDGTT3, preferred prokaryotic substrates, CrDGTT1 is

probably the key enzyme responsible for TAG production

in C. reinhardtii. Thus, CrDGTT1 represents a promising

gene target for metabolic engineering of this and other

related microalgal species for improved lipid production,

for example through enzyme engineering or over-expres-

sion of CrDGTT1 homologs. CrDGTT1 and CrDGTT2 had

high activities on x-3 long-chain PUFAs, particularly eico-sapentaenoic acid (20:5), thus providing opportunities to

engineer microalgae or other microbes to over-produce

eicosapentaenoic acid-rich TAGs.

Figure 7. Proposed working model illustrating the localization and possible

roles of CrDGTT1 in Chlamydomonas TAG biosynthesis.

For simplicity, a single LD is shown bridging the ER and CpE. Not all inter-

mediates or reactions are displayed. CrDGTT1 (red font) resides in both the

chloroplast envelope (CpE) and ER. Arrows indicate catalytic steps in the

glycerolipid biosynthesis pathway.



EXPERIMENTAL PROCEDURES

Strains and growth conditions

The Chlamydomonas strains used, including the wild-type strainCC-1690 (mt+), the cell wall-deficient strain CC-503 (mt+ cw92) andthe sta6 mutant (CC-4348), were obtained from the Chlamy-domonas Resource Center (http://www.chlamycollection.org/). Allstrains were grown in Tris-acetate-phosphate (TAP) medium con-taining 7.5 mM NH4Cl at 23°C under continuous illumination of40 lmol photons m�2 sec�1. To impose N deprivation, cells werecollected in the stationary growth phase, washed with N-free TAPmedium (TAP-N) at room temperature for 5 min, and resuspendedin TAP-N. For feeding with external fatty acids, C16:1n7, C18:1n9,C18:2 and C18:3n3 were supplemented at concentrations of 62.5–250 lM as described by Siloto et al. (2009), under nitrogen depri-vation in the presence of 10 lM cerulenin, a specific inhibitoragainst fatty acid biosynthesis.

RNA isolation and quantitative real-time PCR

RNA was isolated from algal samples using TRI reagent (Invitro-gen, https://www.thermofisher.com/us/en/home/brands/invitro-gen.html) according to the manufacturer’s instructions. cDNAsynthesis and quantitative real-time PCR were performed asdescribed by Liu et al. (2012) using an Applied Biosystems 7500fast real-time PCR system (http://www.thermofisher.com/us/en/home/brands/applied-biosystems.html) with SYBR Green PCRMaster Mix (Invitrogen). Primer sequences used for real-time PCRare listed in Table S1. The mRNA expression level was normalizedusing the actin gene as the internal control.

Gene cloning, expression in yeast, and yeast lipid analysis

Chlamydomonas DGAT genes were cloned into the yeastexpression vector pYES2-CT (Invitrogen). The PCR primers forcloning are listed in Table S1. The recombinant CrDGAT con-structs were then transformed into the TAG-deficient S. cere-visiae strain H1246 (Ddga1Dlro1Dare1Dare2) (Sandager et al.,2002) using an S.c. EasyCompTM transformation kit (Invitrogen).Fatty acids were fed to yeast cultures as described by Silotoet al. (2009), with supplementation of C16:1n7, C18:1n9, C18:2and C18:3n3 at a concentration of 250 lM under galactoseinduction.

Lipid extraction and analysis of yeast cells were performed asdescribed by Yoon et al. (2012). For lipid quantification, lipids on TLCplates were visualized using iodine vapor, recovered, transesterified,and analyzed by GC-MS, as previously described (Li et al., 2010).

Yeast microsome preparation and in vitro CrDGAT activity

assay

Yeast transformants were collected after growth in liquid galactosemedium lacking uracil for 15 h at 30°C. The cell pellets were resus-pended in cell lysis buffer containing 5% glycerol, 20 mM Tris/HClpH 8.0, 0.3 M ammonium sulfate, 10 mM MgCl2, 1 mM EDTA, 1 mMdithiothreitol, 19 EDTA-free protease inhibitor cocktail set X (Cal-biochem, http://www.calbiochem.com), 1 mM phenylmethanesul-fonyl fluoride to an OD600 of approximately 100, and lysed bypassing twice through a French press (Spectronics Instruments,www.spectronic.com) at an internal pressure of 103 421 Kpa. Celldebris was removed from the suspension by centrifugation at10 000 g for 10 min at 4°C, and the supernatant was centrifugedfurther at 100 000 g for 1 h at 4°C. The resulting microsomal mem-brane pellets were resuspended in microsomal storage buffer

(50 mM Tris/HCl, pH 7.5, 10% glycerol) to a protein concentration of10 lg ll�1 for immediate use or storage at �80°C.

The DGAT in vitro assay was performed in a 200 ll volume ofassay mixture containing 50 mM potassium phosphate (pH 7.5),10 mM MgCl2, 40 lg microsomal membrane protein, 250 lM acylCoA and 250 lM DAG (from a 50 mM stock in ethanol). Reactionswere incubated at 30°C for 1 h, and the lipids were extracted anddetected as described above. The microsome fraction alone, themicrosome fraction with acyl CoA, or the microsome fraction withDAG were used as controls, and the background levels of TAGwere subtracted from the data for DGAT activity analysis. The acylCoAs included palmitoyl CoA (C16:0 CoA), hexadecenoyl CoA(C16:1 CoA), stearoyl CoA (C18:0 CoA), oleoyl CoA (C18:1n9 CoA),linoleoyl CoA (C18:2 CoA), a-linolenoyl CoA (C18:3n3 CoA), c-lino-lenoyl CoA (C18:3n6(d6) CoA, which is different from C18:3n6(D5,9,12) present in Chlamydomonas), eicosapentaenoyl CoA(C20:5n3 CoA), docosahexaenoyl CoA (C22:6n3 CoA), andlignoceryl CoA (C24:0 CoA). The DAGs included C16:0/C16:0DAG, C16:1/C16:1 DAG, C18:1n9/C16:0 DAG, C16:0/C18:1n9 DAG,C18:1n9/C18:1n9 DAG, 1,3-C18:1n9/C18:1n9 DAG, C18:2/C18:2 DAGand C18:3n3/C18:3n3 DAG. C16:1/C16:1, C18:2/C18:2 and C18:3n3/C18:3n3 DAGs were purchased from Larodan Fine Chemicals(http://www.larodan.com/), and C18:1n9/C16:0 DAG was preparedby partial digestion of C18:1n9/C16:0/C18:1n9 TAG using Rhizopusarrhizus lipase (Sigma-Aldrich, http://www.sigmaaldrich.com/) andrecovery of DAG. The other lipid standards were purchased fromAvanti Polar Lipids (http://avantilipids.com/).

Artificial microRNA-mediated gene silencing

The amiRNA vectors targeting CrDGTT genes were designed asdescribed by Molnar et al. (2009). The resulting forward andreverse oligonucleotides (Table S1) for CrDGTT1, CrDGTT2 andCrDGTT3 were annealed and cloned into the SpeI site of thepChlamyRNA4 vector (Figure S11), which was derived from pChla-myRNA2 (Chlamydomonas Resource Center, http://www.chlamy-collection.org/) by substituting ARG7 with a b-tubulin-aphVIIIcassette from the plasmid pKS-aphVIII-lox (ChlamydomonasResource Center). These vectors were linearized using ScaI, trans-formed into Chlamydomonas, and screened as previouslydescribed (Yoon et al., 2012).

Organelle preparation and immunoblotting

To isolate cellular organelles, 2 liters of CC-503 cell culture werecollected and washed in ice-cold PBS. Thylakoid membranes andchloroplast envelope membranes were isolated as previouslydescribed (Chua and Bennoun, 1975; Mendiola-Morgenthaleret al., 1985) with some modifications. Algal cells were disruptedusing BioNeb (VWR, https://vwr.com/) at 138 Kpa, and the cellhomogenate was layered onto a sucrose gradient containing0.9 M/0.6 M sucrose and centrifuged at 10 000 g, 4°C for 1 h. Thy-lakoid membranes floating on the 1.3 M sucrose fraction were col-lected, and the membranes were pelleted by centrifugation(100 000 g, 4°C for 10 min).

To isolate LDs, cells were disrupted as described above. Thesupernatant was mixed with 10 ml ice-cold isolation buffer con-taining 60% w/v sucrose. The density-adjusted cell lysate (12 ml)was loaded into an ultracentrifuge tube and topped with 12 mlisolation buffer containing 5% w/v sucrose. The samples werecentrifuged at 100 000 g for 1 h at 4°C. LDs floating on the isola-tion buffer were collected and resuspended in isolation buffer con-taining 5% w/v sucrose and 100 mM sodium carbonate. Thesucrose gradient containing crude LDs was layered on the isola-tion buffer, and centrifuged at 100 000 g, 4°C for 30 min to purify


16 Jin Liu et al.

http://www.chlamycollection.org/https://www.thermofisher.com/us/en/home/brands/invitrogen.htmlhttps://www.thermofisher.com/us/en/home/brands/invitrogen.htmlhttp://www.thermofisher.com/us/en/home/brands/applied-biosystems.htmlhttp://www.thermofisher.com/us/en/home/brands/applied-biosystems.htmlhttp://www.calbiochem.comwww.spectronic.comhttp://www.larodan.com/http://www.sigmaaldrich.com/http://avantilipids.com/http://www.chlamycollection.org/http://www.chlamycollection.org/https://vwr.com/

LDs. ER membranes and mitochondria were isolated as describedby Bessoule et al. (1995) and Eriksson et al. (1995), respectively.

The protein samples from organelle fractions and whole cellswere used for immunoblotting as previously described (Yoonet al., 2012). Recombinant anti-CrDGTT1 antibody was obtainedfrom Agrisera AB (http://www.agrisera.com/).

Chlamydomonas lipid extraction and analysis

Algal lipid extraction from Chlamydomonas and TLC analysis ofneutral lipids were performed as previously described (Yoonet al., 2012). For TLC analysis of polar lipids, a mixture of chloro-form/methanol/acetic acid/water (25/4/0.7/0.3 by volume) was usedas the mobile phase. Lipid detection by charring and lipid quantifi-cation by GC-MS were performed as previously described (Liet al., 2010). TAG positional analysis was performed using R. ar-rhizus lipase as described by Li et al. (2012).

Statistical analysis

All experiments were performed using biological triplicates toensure reproducibility. Values presented are means � SD.Statistical analyses were performed using the SPSS statisticalpackage (SPSS Inc., http://www-01.ibm.com/software/analytics/spss/). The paired-samples t test was applied. Differences wereconsidered statistically significant at P values

Cao, H. (2011) Structure-function analysis of diacylglycerol acyltransferase

sequences from 70 organisms. BMC Res Notes, 4, 249.

Chapman, K.D. and Ohlrogge, J.B. (2012) Compartmentation of triacylglyc-

erol accumulation in plants. J. Biol. Chem. 287, 2288–2294.Chen, J.E. and Smith, A.G. (2012) A look at diacylglycerol acyltransferases

(DGATs) in algae. J. Biotechnol. 162, 28–39.Chua, N.H. and Bennoun, P. (1975) Thylakoid membrane polypeptides of

Chlamydomonas reinhardtii: wild-type and mutant strains deficient in

photosystem II reaction center. Proc. Natl Acad. Sci. USA, 72, 2175–2179.

Dahlqvist, A., Stahl, U., Lenman, M., Banas, A., Lee, M., Sandager, L.,

Ronne, H. and Stymne, H. (2000) Phospholipid:diacylglycerol acyltrans-

ferase: an enzyme that catalyzes the acyl-CoA-independent formation of

triacylglycerol in yeast and plants. Proc. Natl Acad. Sci. USA, 97, 6487–6492.

Deng, X.-D., Gu, B., Li, Y.-J., Hu, X.-W., Guo, J.-C. and Fei, X.-W. (2012) The

roles of acyl-CoA:diacylglycerol acyltransferase 2 genes in the biosynthe-

sis of triacylglycerols by the green algae Chlamydomonas reinhardtii.

Mol. Plant, 5, 945–947.Eriksson, M., Gardestrom, P. and Samuelsson, G. (1995) Isolation, purifica-

tion, and characterization of mitochondria from Chlamydomonas rein-

hardtii. Plant Physiol. 107, 479–483.Fan, J., Andre, C. and Xu, C. (2011) A chloroplast pathway for the de novo

biosynthesis of triacylglycerol in Chlamydomonas reinhardtii. FEBS Lett.

585, 1985–1991.Fan, J., Yan, C., Andre, C., Shanklin, J., Schwender, J. and Xu, C. (2012) Oil

accumulation is controlled by carbon precursor supply for fatty acid syn-

thesis in Chlamydomonas reinhardtii. Plant Cell Physiol. 53, 1380–1390.Fan, J., Yan, C., Zhang, X. and Xu, C. (2013) Dual role for phospholipid:dia-

cylglycerol acyltransferase: enhancing fatty acid synthesis and diverting

fatty acids from membrane lipids to triacylglycerol in Arabidopsis leaves.

Plant Cell, 25, 3506–3518.Fan, J., Cui, Y., Wan, M., Wang, W. and Li, Y. (2014) Lipid accumulation and

biosynthesis genes response of the oleaginous Chlorella pyrenoidosa

under three nutrition stressors. Biotechnol. Biofuels, 7, 17.

Gargouri, M., Park, J.-J., Holguin, F.O., Kim, M.-J., Wang, H., Deshpande,

R.R., Shachar-Hill, Y., Hicks, L.M. and Gang, D.R. (2015) Identification of

regulatory network hubs that control lipid metabolism in Chlamy-

domonas reinhardtii. J. Exp. Bot. 66, 4551–4566.Gong, Y., Zhang, J., Guo, X., Wan, X., Liang, Z., Hu, C.J. and Jiang, M.

(2013) Identification and characterization of PtDGAT2B, an acyltrans-

ferase of the DGAT2 acyl-coenzyme A:diacylglycerol acyltransferase fam-

ily in the diatom Phaeodactylum tricornutum. FEBS Lett. 587, 481–487.Goodenough, U., Blaby, I., Casero, D. et al. (2014) The path to triacylglyc-

eride obesity in the sta6 strain of Chlamydomonas reinhardtii. Eukaryot.

Cell, 13, 591–613.Goodson, C., Roth, R., Wang, Z.T. and Goodenough, U. (2011) Structural

correlates of cytoplasmic and chloroplast lipid body synthesis in Chlamy-

domonas reinhardtii and stimulation of lipid body production with acet-

ate boost. Eukaryot. Cell, 10, 1592–1606.Greer, M., Zhou, T. and Weselake, R. (2014) A novel assay of DGAT activity

based on high temperature GC/MS of triacylglycerol. Lipids, 49, 831–838.Greer, M.S., Truksa, M., Deng, W., Lung, S.C., Chen, G. and Weselake, R.J.

(2015) Engineering increased triacylglycerol accumulation in Saccha-

romyces cerevisiae using a modified type 1 plant diacylglycerol acyl-

transferase. Appl. Microbiol. Biotechnol. 99, 2243–2253.Guiheneuf, F., Leu, S., Zarka, A., Khozin-Goldberg, I., Khalilov, I. and Bous-

siba, S. (2011) Cloning and molecular characterization of a novel acyl-

CoA:diacylglycerol acyltransferase 1-like gene (PtDGAT1) from the dia-

tom Phaeodactylum tricornutum. FEBS J. 278, 3651–3666.Hu, Q., Sommerfeld, M., Jarvis, E., Ghirardi, M., Posewitz, M., Seibert, M.

and Darzins, A. (2008) Microalgal triacylglycerols as feedstocks for bio-

fuel production: perspectives and advances. Plant J. 54, 621–639.Hung, C.-H., Ho, M.-Y., Kanehara, K. and Nakamura, Y. (2013) Functional

study of diacylglycerol acyltransferase type 2 family in Chlamydomonas

reinhardtii. FEBS Lett. 587, 2364–2370.Iwai, M., Ikeda, K., Shimojima, M. and Ohta, H. (2014) Enhancement of

extraplastidic oil synthesis in Chlamydomonas reinhardtii using a type-2

diacylglycerol acyltransferase with a phosphorus starvation-inducible

promoter. Plant Biotechnol. J. 12, 808–819.

Kalscheuer, R. and Steinbuchel, A. (2003) A novel bifunctional wax ester

synthase/acyl-CoA:diacylglycerol acyltransferase mediates wax ester and

triacylglycerol biosynthesis in Acinetobacter calcoaceticus ADP1. J. Biol.

Chem. 278, 8075–8082.La Russa, M., Bogen, C., Uhmeyer, A., Doebbe, A., Filippone, E., Kruse, O.

and Mussgnug, J.H. (2012) Functional analysis of three type-2 DGAT

homologue genes for triacylglycerol production in the green microalga

Chlamydomonas reinhardtii. J. Biotechnol. 162, 13–20.L�egeret, B., Schulz-Raffelt, M., Nguyen, H.M., Auroy, P., Beisson, F., Peltier,

G., Blanc, G. and Li-Beisson, Y. (2016) Lipidomic and transcriptomic anal-

yses of Chlamydomonas reinhardtii under heat stress unveil a direct

route for the conversion of membrane lipids into storage lipids. Plant

Cell Environ. doi: 10.1111/pce.12656.

Li, Y., Han, D., Hu, G., Dauvillee, D., Sommerfeld, M., Ball, S. and Hu, Q.

(2010) Chlamydomonas starchless mutant defective in ADP-glucose

pyrophosphorylase hyper-accumulates triacylglycerol. Metab. Eng. 12,

387–391.Li, Y., Han, D., Sommerfeld, M. and Hu, Q. (2011) Photosynthetic carbon

partitioning and lipid production in the oleaginous microalga Pseu-

dochlorococcum sp. (Chlorophyceae) under nitrogen-limited conditions.

Bioresour. Technol. 102, 123–129.Li, X., Moellering, E.R., Liu, B., Johnny, C., Fedewa, M., Sears, B.B., Kuo,

M.-H. and Benning, C. (2012) A galactoglycerolipid lipase is required for

triacylglycerol accumulation and survival following nitrogen deprivation

in Chlamydomonas reinhardtii. Plant Cell, 24, 4670–4686.Li-Beisson, Y., Beisson, F. and Riekhof, W. (2015) Metabolism of acyl-lipids

in Chlamydomonas reinhardtii. Plant J. 82, 504–522.Liu, Q., Siloto, R.M.P., Snyder, C.L. and Weselake, R.J. (2011) Functional

and topological analysis of yeast acyl-CoA:diacylglycerol acyltrans-

ferase 2, an endoplasmic reticulum enzyme essential for triacylglycerol

biosynthesis. J. Biol. Chem. 286, 13115–13126.Liu, J., Sun, Z., Zhong, Y., Huang, J., Hu, Q. and Chen, F. (2012) Stearoyl-

acyl carrier protein desaturase gene from the oleaginous microalga

Chlorella zofingiensis: cloning, characterization and transcriptional analy-

sis. Planta, 236, 1665–1676.Lung, S.-C. and Weselake, R. (2006) Diacylglycerol acyltransferase: a key

mediator of plant triacylglycerol synthesis. Lipids, 41, 1073–1088.Manandhar-Shrestha, K. and Hildebrand, M. (2015) Characterization and

manipulation of a DGAT2 from the diatom Thalassiosira pseudonana:

improved TAG accumulation without detriment to growth, and implica-

tions for chloroplast TAG accumulation. Algal Res. 12, 239–248.Mendiola-Morgenthaler, L., Eichenberger, W. and Boschetti, A. (1985) Isola-

tion of chloroplast envelopes from Chlamydomonas. Lipid and polypep-

tide composition. Plant Sci. 41, 97–104.Merchant, S.S., Prochnik, S.E., Vallon, O. et al. (2007) The Chlamydomonas

genome reveals the evolution of key animal and plant functions. Science,

318, 245–250.Milcamps, A., Tumaney, A.W., Paddock, T., Pan, D.A., Ohlrogge, J. and Pol-

lard, M. (2005) Isolation of a gene encoding a 1,2-diacylglycerol-sn-

acetyl-CoA acetyltransferase from developing seeds of Euonymus alatus.

J. Biol. Chem. 280, 5370–5377.Miller, R., Wu, G., Deshpande, R.R. et al. (2010) Changes in transcript abun-

dance in Chlamydomonas reinhardtii following nitrogen deprivation pre-

dict diversion of metabolism. Plant Physiol. 154, 1737–1752.Moellering, E.R. and Benning, C. (2010) RNA interference silencing of a

major lipid droplet protein affects lipid droplet size in Chlamydomonas

reinhardtii. Eukaryot. Cell, 9, 97–106.Moellering, E.R. and Benning, C. (2011) Galactoglycerolipid metabolism

under stress: a time for remodeling. Trends Plant Sci. 16, 98–107.Molnar, A., Bassett, A., Thuenemann, E., Schwach, F., Karkare, S.,

Ossowski, S., Weigel, D. and Baulcombe, D. (2009) Highly specific gene

silencing by artificial microRNAs in the unicellular alga Chlamydomonas

reinhardtii. Plant J. 58, 165–174.Neupert, J., Karcher, D. and Bock, R. (2009) Generation of Chlamydomonas

strains that efficiently express nuclear transgenes. Plant J. 57, 1140–1150.

Nguyen, H.M., Baudet, M., Cuin�e, S. et al. (2011) Proteomic profiling of oil

bodies isolated from the unicellular green microalga Chlamydomonas

reinhardtii: with focus on proteins involved in lipid metabolism. Pro-

teomics, 11, 4266–4273.


18 Jin Liu et al.

info:doi/10.1111/pce.12656

Oelkers, P., Cromley, D., Padamsee, M., Billheimer, J.T. and Sturley, S.L.

(2002) The DGA1 gene determines a second triglyceride synthetic path-

way in yeast. J. Biol. Chem. 277, 8877–8881.Ohlrogge, J. and Browse, J. (1995) Lipid biosynthesis. Plant Cell, 7, 957–

970.

Ohlrogge, J.B. and Jaworski, J.G. (1997) Regulation of fatty acid synthesis.

Annu. Rev. Plant Physiol. Plant Mol. Biol. 48, 109–136.Radakovits, R., Jinkerson, R.E., Fuerstenberg, S.I., Tae, H., Settlage, R.E.,

Boore, J.L. and Posewitz, M.C. (2012) Draft genome sequence and

genetic transformation of the oleaginous alga Nannochloropis gaditana.

Nat. Commun. 3, 686.

Saha, S., Enugutti, B., Rajakumari, S. and Rajasekharan, R. (2006) Cytosolic

triacylglycerol biosynthetic pathway in oilseeds. Molecular cloning and

expression of peanut cytosolic diacylglycerol acyltransferase. Plant Phys-

iol. 141, 1533–1543.Sandager, L., Gustavsson, M.H., Stahl, U., Dahlqvist, A., Wiberg, E., Banas,

A., Lenman, M., Ronne, H. and Stymne, S. (2002) Storage lipid synthesis

is non-essential in yeast. J. Biol. Chem. 277, 6478–6482.Sanjaya , Miller, R., Durrett, T.P. et al. (2013) Altered lipid composition and

enhanced nutritional value of Arabidopsis leaves following introduction

of an algal diacylglycerol acyltransferase 2. Plant Cell, 25, 677–693.Sheehan, J., Dunahay, T., Benemann, J. and Roessler, P. (1998) A Look Back

at the US Department of Energy’s Aquatic Species Program: Biodiesel

from Algae. Report number NREL/TP-580-24190. Golden, CO: National

Renewable Energy Laboratory.

Shockey, J.M., Gidda, S.K., Chapital, D.C., Kuan, J.C., Dhanoa, P.K., Bland,

J.M., Rothstein, S.J., Mullen, R.T. and Dyer, J.M. (2006) Tung tree

DGAT1 and DGAT2 have nonredundant functions in triacylglycerol

biosynthesis and are localized to different subdomains of the endoplas-

mic reticulum. Plant Cell, 18, 2294–2313.Siloto, R.P., Truksa, M., He, X., McKeon, T. and Weselake, R. (2009) Simple

methods to detect triacylglycerol biosynthesis in a yeast-based recombi-

nant system. Lipids, 44, 963–973.Sorger, D. and Daum, G. (2003) Triacylglycerol biosynthesis in yeast. Appl.

Microbiol. Biotechnol. 61, 289–299.Stahl, U., Carlsson, A.S., Lenman, M., Dahlqvist, A., Huang, B., Banas, W.,

Banas, A. and Stymne, S. (2004) Cloning and functional characterization

of a phospholipid:diacylglycerol acyltransferase from Arabidopsis. Plant

Physiol. 135, 1324–1335.Stone, S.J., Levin, M.C. and Farese, R.V. (2006) Membrane topology and

identification of key functional amino acid residues of murine acyl-CoA:

diacylglycerol acyltransferase-2. J. Biol. Chem. 281, 40273–40282.Tsai, C.-H., Zienkiewicz, K., Amstutz, C.L., Brink, B.G., Warakanont, J., Ros-

ton, R. and Benning, C. (2015) Dynamics of protein and polar lipid

recruitment during lipid droplet assembly in Chlamydomonas reinhardtii.

Plant J. 83, 650–660.Vieler, A., Wu, G., Tsai, C.-H. et al. (2012) Genome, functional gene annota-

tion, and nuclear transformation of the heterokont oleaginous alga Nan-

nochloropsis oceanica CCMP1779. PLoS Genet. 8, e1003064.

Wagner, M., Hoppe, K., Czabany, T., Heilmann, M., Daum, G., Feussner, I.

and Fulda, M. (2010) Identification and characterization of an acyl-CoA:

diacylglycerol acyltransferase 2 (DGAT2) gene from the microalga O.

tauri. Plant Physiol. Biochem. 48, 407–416.Wang, D., Ning, K., Li, J. et al. (2014) Nannochloropsis genomes reveal evo-

lution of microalgal oleaginous traits. PLoS Genet. 10, e1004094.

Xu, J., Francis, T., Mietkiewska, E., Giblin, E.M., Barton, D.L., Zhang, Y.,

Zhang, M. and Taylor, D.C. (2008) Cloning and characterization of an

acyl-CoA-dependent diacylglycerol acyltransferase 1 (DGAT1) gene from

Tropaeolum majus, and a study of the functional motifs of the DGAT

protein using site-directed mutagenesis to modify enzyme activity and

oil content. Plant Biotechnol. J. 6, 799–818.Yoon, K., Han, D., Li, Y., Sommerfeld, M. and Hu, Q. (2012) Phospholipid:di-

acylglycerol acyltransferase is a multifunctional enzyme involved in

membrane lipid turnover and degradation while synthesizing triacylglyc-

erol in the unicellular green microalga Chlamydomonas reinhardtii. Plant

Cell, 24, 3708–3724.Zhang, C., Iskandarov, U., Klotz, E.T., Stevens, R.L., Cahoon, R.E., Nazare-

nus, T.J., Pereira, S.L. and Cahoon, E.B. (2013) A thraustochytrid diacyl-

glycerol acyltransferase 2 with broad substrate specificity strongly

increases oleic acid content in engineered Arabidopsis thaliana seeds. J.

Exp. Bot. 64, 3189–3200.Zou, J., Wei, Y., Jako, C., Kumar, A., Selvaraj, G. and Taylor, D.C. (1999) The

Arabidopsis thaliana TAG1 mutant has a mutation in a diacylglycerol

acyltransferase gene. Plant J. 19, 645–653.



characterization of type 2 diacylglycerol acyltransferases ... · chlamydomonas reinhardtii harbors...

Documents