CHARACTERIZATION OF LIPASES INVOLVED IN TRIACYLGLYCEROL
METABOLISM IN CHLAMYDOMONAS REINHARDTII
By
Xiaobo Li

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Plant Biology
2012

ABSTRACT
CHARACTERIZATION OF LIPASES INVOLVED IN TRIACYLGLYCEROL
METABOLISM IN CHLAMYDOMONAS REINHARDTII
By
Xiaobo Li
Triacylglycerol (TAG) is the major storage lipid for eukaryotic organisms, including
animals, fungi, land plants and algae. With a higher energy density than carbohydrates and
proteins, TAG from plant sources serves as an ideal feedstock for biofuel production. By
transesterification, TAGs can be converted to fatty acid methylesters (FAMEs), which are similar
to components of diesel derived from petroleum in carbon chain length and viscosity. Microalgae
grow faster and have a larger potential for industrial biofuel production than land plants. In
microalgae TAG production can be induced by nutrient stresses such as nitrogen (N) deprivation.
Insights into the mechanism of oil production are required for further optimization of the biofuel
production

process.

As

an

experimental

organism

for

studying

lipid

metabolism,

Chlamydomonas reinhardtii represents the most studied alga with multiple genetic and
molecular biological tools available. In a transcript-profiling study to compare transcript
abundance of genes in Chlamydomonas under normal growth conditions and N deprived
conditions, a large number of predicted lipase-encoding genes appeared to be differentially
regulated under the two conditions. This observation suggests that lipases play an important role
in TAG metabolism of Chlamydomonas.
Conceivably, TAG production can be enhanced by suppressing TAG degradation
enzymes, such as TAG lipases. One avenue of research pursued a reverse genetic screen of
several lipase candidates by heterologous expression in a yeast mutant deficient in TAG lipases.
Over-expression of one of the candidate genes, named CrLIP1, rescued the yeast mutant.

Recombinant CrLIP1 protein exhibited lipolytic activity on diacylglycerol (DAG) and polar
lipids, but not TAG. RNA suppression of the CrLIP1 gene led to delayed TAG degradation in
Chlamydomonas, possibly through reduced capacity to degrade DAG, which is an intermediate
in TAG turnover. It is also possible that CrLIP1 acts on TAG but requires an unknown cofactor
which was missing in the lipase assays.
To complement the transcript profiling and reverse genetic approaches, an unbiased
forward genetic screen was performed to obtain mutants with abnormal TAG amounts. Focus
was given to a low TAG mutant disrupted in a gene designated Plastid Galactolipid Degradation
1 (PGD1), which was identified as a lipase-encoding gene. In the above-mentioned
transcriptomic study, PGD1 was up-regulated following N deprivation. The pgd1 mutant is
decreased in acyl fluxes from galactolipids to TAG. Recombinant PGD1 protein exhibited lipase
activity on the substrate monogalactosyldiacylglycerol (MGDG). Multiple lines of evidence
suggest PGD1 takes part in an acylation/deacylation cycle to export de novo synthesized fatty
acids from the plastid.
The low TAG mutant pgd1 provides a tool for understanding the consequences of
reduced TAG production. The pgd1 mutant exhibited chlorosis, accumulation of oxidative
damage and loss of viability, which was reversed by the exposure to 3-(3,4-dichlorophenyl)-1,1dimethylurea (DCMU), a specific inhibitor for photosynthetic electron transport. This indicates
that cells synthesize fatty acids and TAG to relieve the strain of excess electron transport and
prevent the accompanying oxidative stress. These studies demonstrate the physiological
significance of the accumulation of TAG under stress conditions and will impact how scientists
approach the improvement of the production of biofuels from algae.

ACKNOWLEDGEMENTS

In August 2007 I left the Department of Bioengineering, Xi’an Jiaotong University and
started my Ph.D. in the DOE-Plant Research Lab at Michigan State University. The degreegranting program for me is the Department of Plant Biology. After three lab rotations, in April
2008, I decided to work with Dr. Christoph Benning on algal lipid metabolism. Christoph asked
his colleague Min-Hao Kuo to join the project because he knew I need to use yeast as a tool for
certain experiments. This co-advising strategy proved successful in providing guidance from
different perspectives and training me a broad set of skills.
Christoph is a great advisor, not just because of his ideas in science, also because of his
ways of nurturing a graduate student. He accepted me to his lab because he saw my passion,
efforts and creativity during my rotation despite my lack of experience in molecular biology.
One year after I joined the lab, I lost my projects because the mutants were not well maintained.
His encouragement helped me calm down and start another project. After I got slightly mature in
research, he allowed me a lot of freedom in the directions to go. Min-Hao is also a great advisor.
From frequent discussions with him, he has brilliant, neat and sometimes crazy ideas on many
questions. Min-Hao was initially not an expert in lipid metabolism and algae. But every time
when I talk with him he can point out some connections between my research and other
biological questions. That’s how we found the longevity phenotype of the yeast lipase mutant.
Min-Hao cares about the training of his students, too. There were several times that he helped me
on oral presentations by giving suggestions slide by slide.
My committee members Jianping Hu and Barbara B. Sears helped by providing me
suggestions on experiments and giving me right papers to read. Barb also did some experiments

iv

on my project. Several other professors generously allowed me to use their equipment. I was a
frequent user of the microplate reader in Thomas D. Sharkey lab and the Real-time PCR machine
located in Gregg A. Howe lab. My colleagues are great complements to my advisors and
committee. Each of them has his/her own expertise and everyone is glad to provide help on
techniques. Eric R. Moellering in the Benning lab and JianjunLuo in Kuo lab are the two
colleagues that trained me the most. Here, I have to mention that, the algal mutant that I worked
on was originally found by Eric R. Moellering.
More broadly, I want to thank MSU, the College of Natural Sciences and the departments
for their support. My programs are DOE-Plant Research Lab and Plant Biology. But I also want
to acknowledge the Department of Biochemistry and Molecular Biology as I do experiments in
BMB labs.
My friends in East Lansing spent a lot of time together with me. We played hard after
working hard. Last but not least, I am grateful to my family (parents and two sisters) and Shasha
Zhang for their love and support regardless what I choose to do during and after my PhD.

v

TABLE OF CONTENTS

LIST OF TABLES ....................................................................................................................... viii
LIST OF FIGURES ....................................................................................................................... ix
CHAPTER 1
Algal triacylglycerol metabolism and lipases-a literature review....................................................1
Introduction ..........................................................................................................................2
Chlamydomonas reinhardtii as a model system ….................................................................... 3
Comparison between Arabidopsis and Chlamydomonas lipid metabolism pathways….. ... 5
Phosphatidylcholine (PtdCho) is absent in Chlamydomonas ..............................................7
A general overview of triacylglycerol (TAG) biosynthesis in Chlamydomonas ................8
Recent progress in the genomics, genetics and cell biology of Chlamydomonas TAG
metabolism .........................................................................................................................11
Potential roles of lipases in triacylglycerol metabolism ...................................................13
Aims of this research .........................................................................................................16
CHAPTER 2
Rapid triacylglycerol turnover in Chlamydomonas reinhardtii requires a lipase with broad
substrate specificity........................................................................................................................23
Abstract ..............................................................................................................................24
Introduction ........................................................................................................................25
Materials and Methods .......................................................................................................27
Results ................................................................................................................................35
Discussion ..........................................................................................................................45
CHAPTER 3
A galactoglycerolipid lipase is required for triacylglycerol accumulation following nitrogen
deprivation in Chlamydomonas reinhardtii ...................................................................................72
Abstract ..............................................................................................................................73
Introduction ........................................................................................................................74
Materials and Methods .......................................................................................................77
Results ................................................................................................................................86
Discussion ..........................................................................................................................99
Conclusions ......................................................................................................................105
CHAPTER 4
Chlamydomonas shunts electrons into triacylglycerols to avoid photooxidative stress during
nutrient deprivation ......................................................................................................................130
Abstract ............................................................................................................................131
Introduction ......................................................................................................................132
vi

Materials and Methods .....................................................................................................133
Results and Discussion ....................................................................................................136
CHAPTER 5
Conclusions and perspectives ......................................................................................................149
CrLIP1 is involved in the regulation of TAG turnover in Chlamydomonas. ..................150
Oleate can be exported from the plastid through an acylation/deacylation cycle mediated
by PGD1...........................................................................................................................153
Triacylglycerol accumulation in Chlamydomonas reinhardtii prevents oxidative stress
under N deprivation .........................................................................................................156
Reference List ..............................................................................................................................159

vii

LIST OF TABLES

Table 2.1.Yeast strains used in this study ......................................................................................68
Table 2.2.Plasmid constructs used in this study ............................................................................69
Table 2.3.Oligonucleotide primers used in this study....................................................................70
Table 2.4.Putative lipase-encoding genes selected for expression in yeast ...................................71
Table 3.1.Oligonucleotide primers used in this study .................................................................129

viii

LIST OF FIGURES

Figure 1.1. Structure of triacylglycerol and transesterification of triacylglycerol to produce fatty
acid methyl esters ..........................................................................................................................17
Figure 1.2.Comparison of major membrane glycerolipid biosynthesis pathways between
Arabidopsis and Chlamydomonas .................................................................................................18
Figure 1.3.Structures of PtdCho and DGTS. .................................................................................20
Figure 1.4.TAG synthesis catalyzed by DGAT and PDAT activities ...........................................21
Figure 1.5.Lipases affect TAG content by degrading TAG or providing acyl groups for TAG
synthesis .........................................................................................................................................22
Figure 2.1.Expression of CrLIP1 rescues a yeast mutant deficient in TAG lipases ......................49
Figure 2.2.Metabolic labeling and chase studies of TAG and phospholipid synthesis in yeast
strains over-expressing CrLIP1 .....................................................................................................51
Figure 2.3.Recombinant CrLIP1 fails to hydrolyze triacylglycerol in vitro ..................................53
Figure 2.4.Western blotting for hexahistidine-tagged CrLIP1 protein purified from E. coli cells
........................................................................................................................................................54
Figure 2.5.Recombinant CrLIP1 fails to degrade steryl esters in vitro .........................................55
Figure 2.6.CrLIP1 shares sequence similarity to two human DAG lipases DAGLα and DAGLβ
........................................................................................................................................................56
Figure 2.7.Recombinant CrLIP1 degrades diacylglycerol in vitro ................................................57
Figure 2.8.Recombinant CrLIP1 hydrolyzes phosphatidylethanolamine, the major membrane
lipid of E. coli ................................................................................................................................59
Figure 2.9.Recombinant CrLIP1 degrades major polar lipids of Chlamydomonas in vitro… ......61
Figure 2.10.Recombinant CrLIP1 cleaves both acyl groups of phosphatidylcholine with a
preference for sn-1 position ..........................................................................................................62
Figure 2.11.Inverse correlation between CrLIP1 transcription and TAG abundance in
Chlamydomonas ...........................................................................................................................63
Figure 2.12.Suppression of CrLIP1 expression delays TAG turnover in Chlamydomonas. .........64
ix

Figure 2.13.A model for how CrLIP1 might affect TAG metabolism .........................................66
Figure 2.14.RT-PCR quantitative comparison of RACK1, CrLIP1 and DGTT1 transcripts in
vegetatively growing cells ............................................................................................................67
Figure 3.1.Phenotypes of the pgd1 mutant compared to the wild-type parental strain (WT)......106
Figure 3.2.Growth curves of wild-type parental strain and pgd1 in standard TAP medium .......107
Figure 3.3.Molecular characterization the pgd1 mutant…. .........................................................108
Figure 3.4.Genetic complementation of pgd1 phenotypes with wild-type genomic DNA .........110
Figure 3.5.Detailed lipid analysis of the wild-type parental strain (WT) and pgd1 mutant in Nreplete medium (+N) and 48 h after transfer to N-depleted medium (-N)...................................112
Figure 3.6.Fatty acid compositions of DGTS, PtdEtn, MGDG, DGDG, PtdGro of the wild-type
parental strain (WT) and pgd1 mutant in N-replete medium (+N) and 48 h after transfer to Ndepleted medium (-N) ..................................................................................................................114
Figure 3.7.Positional analysis of TAG of the wild-type parental strain (WT) and pgd1 mutant 48
h after transfer to N-depleted medium (-N) .................................................................................116
Figure 3.8.In vivo pulse-chase acetate labeling of lipids in the wild-type parental strain (WT) and
the pgd1 mutant before N deprivation .........................................................................................117
Figure 3.9.In vivo pulse-chase acetate labeling of lipids in the wild-type parental strain (WT) and
the pgd1 mutant............................................................................................................................118
Figure 3.10.In vivo pulse-chase acetate labeling of PtdEtn, PtdGro, SQDG and PtdIns in the
wild-type parental strain (WT) and the pgd1 mutant ...................................................................119
Figure 3.11.Activity of the recombinant PGD1 protein on MGDG and DGDG. ........................120

Figure 3.12.Hydrolysis of Chlamydomonas derived MGDG by PGD1 ......................................122
Figure 3.13.Quantitative hydrolysis of E. coli derived MGDG by PGD1 ...................................124
Figure 3.14.Activity of the recombinant PGD1 protein on DGTS, PtdEtn, PtdGro and SQDG .126
Figure 3.15.Abbreviated model proposing a function for PGD1 in TAG biosynthesis ..............128

x

Figure 4.1.Viability of wild-type parental strain (WT) and the pgd1 mutant following N
deprivation ...................................................................................................................................142
Figure 4.2.A scheme on the role of TAG production in redox control ........................................144
Figure 4.3.Examples of primary and secondary screen for pgd1 suppressors .............................145
Figure 4.4.Analysis of TAG amount and fatty acid compositions in FB 12 and FB16 ...............146
Figure 4.5.Growth curves of FB16, pgd1 mutant and its wild-type (WT) parental strain in TAP
(A) and TMP media (B) ...............................................................................................................147
Figure 4.6.Molecular characterization FB12 ...............................................................................148

xi

Chapter 1
Algal triacylglycerol metabolism and lipases-a literature review.

1

Introduction.
Triacylglycerol (TAG) is composed of a glycerol backbone with three esterified fatty acyl
chains. The acyl chains esterified to glycerol are usually stereospecifically numbered as sn-1, 2
and 3 (Figure 1.1). In contrast to lipids contained in cell membranes such as phospholipids, TAG
lacks a polar headgroup. Therefore, TAG is classified as a “neutral lipids” while membrane
lipids are commonly called “polar lipids”. TAG is synthesized universally by animals, fungi,
land plants and algae to store carbon and energy. Through transesterification with methanol, a
TAG molecule can be converted into three fatty acid methylesters (FAMEs, see Figure 1.1)
which can be blended into petroleum diesel and used by current diesel engines (Durrett et al.,
2008). Microalgae produce a substantial amount of TAGs which provide a potential feedstock
for biofuel production to relieve the widely-recognized energy crisis (Hu et al., 2008). To further
improve the yield and lower the cost of algal biofuel production, a more detailed understanding
of algal lipid metabolism pathways is urgently required. Algal lipid metabolism is relatively
understudied compared to that of land plants. Most of the knowledge of algal lipid metabolism is
based on in silico annotation with information derived from seed plants, especially the model
plant Arabidopsis thaliana. In the model alga Chlamydomonas reinhardtii, orthologs for most
genes involved in glycerolipid biosynthesis can be identified in the genome (Riekhof et al.,
2005). However, plants produce TAGs in the seeds while the unicellular algae such as
Chlamydomonas accumulate TAGs under nutrient deprivation (Weers and Gulati, 1997). Hence
regulation of related enzymes is expected to be different between seed plants and algae.
Transcript profiling and reverse genetics have been performed recently in Chlamydomonas to
reveal genes central to TAG biosynthesis (Miller et al., 2010; Boyle et al., 2012). This chapter

2

will summarize current understanding and remaining questions for algal lipid metabolism with a
focus on the metabolism of TAG.

Chlamydomonas reinhardtii as a model system
Chlamydomonas reinhardtii is a unicellular green alga with a cell diameter of about 10
µm (Fang et al., 2006). Chlamydomonas strains have been isolated from the soil and fresh water.
th

Studies on Chlamydomonas have been recorded since the 19

century, initially on sexual

reproduction by J.N. Goroschankin in 1875, and then on multiple processes such as cell motility,
phototaxis and photosynthesis (Harris, 1989).
Chlamydomonas haploid cells grow vegetatively through mitotic divisions but can be
induced to enter a sexual cycle through nitrogen starvation and subsequent mixing of cells of two
+

-

mating types (mt and mt ). The diploid zygote forms and undergoes meiosis to produce four
progeny spores (Goodenough et al., 1995). The predominance of the haploid stage has important
implications for genetic research on Chlamydomonas. A mutation in a haploid strain will have an
immediate effect while in diploid organisms, it can be compensated by the wild-type allele. With
the sexual cycle, strains can be crossed to combine different traits. Additionally, a fraction of
diploid cells obtained from mating can be maintained vegetatively (Ebersold, 1967), facilitating a
complementation test for different mutations or maintenance of a mutation that is lethal in a
haploid background.
The flagella of a Chlamydomonas cell share the 9+2 microtubule doublet structure with
human cilia (Silflow and Lefebvre, 2001). Thus, research on the structure and function of
Chlamydomonas flagella have yielded principles that have been applied to human health. The
primary cause of the mouse autosomal-recessive polycystic kidney disease proved to be a defect

3

in a protein which has been discovered in the intraflagellar transport particles in Chlamydomonas
(Rosenbaum and Witman, 2002). Studies of mutants with abnormal flagellar size or function
(McVittie, 1972) and characterization of the flagellar proteome (Pazour et al., 2005) revealed
numerous genes involved in flagellar biogenesis and movement.
Chlamydomonas has a single chloroplast that is conserved between green algae and land
plants but different from the rhodoplasts in red algae in ultrastructure and composition of
pigments (Keeling, 2004; Wise and Hoober, 2006). Chlamydomonas can grow in the dark on
organic carbon sources such as acetate as well as photoautotrophically with CO2 as the sole
carbon source (Grossman et al., 2003). As a result, mutants that are completely blocked in
photosynthesis can be maintained by supplementing acetate in the growth medium. Studies of
Chlamydomonas photosynthesis have yielded knowledge in key processes such as dynamics of
state transitions and cyclic photosynthetic electron transport (Finazzi, 2005). In the recent years,
the mechanism of photoprotection has been elucidated to a large extent through characterizing
mutants deficient in non-photochemical quenching (Niyogi, 1999).
Studies of carbon storage in Chlamydomonas have focused on starch biosynthesis and
turnover. Several mutants deficient in starch biosynthesis have been isolated and the mutated loci
identified (Zabawinski et al., 2001). Seminal research on Chlamydomonas lipid metabolism has
dealt

with

the

biosynthesis

and

function

of

membrane

lipids

such

as

sulfoquinovosyldiacylglycerol (Riekhof et al., 2003), but studies on algal triacylglycerol (TAG)
metabolism are preliminary. Chlamydomonas is unrivaled among algae in the molecular genetic
toolset developed, including the random insertional mutagenesis (Kindle, 1990), RNA
suppression technique (Molnar et al., 2009) and the use of GFP expression for protein
localization (Neupert et al., 2009). Genetic characterization of TAG metabolism pathways in

4

Chlamydomonas will reveal genes that can be engineered for optimization of TAG production as
biofuel feedstock.

Comparison between Arabidopsis and Chlamydomonas lipid metabolism pathways
Figure 1.2 outlines the major pathways for membrane glycerolipid biosynthesis in
Arabidopsis thaliana and Chlamydomonas reinhardtii based on biochemical studies and genome
annotation (Riekhof et al., 2005). The pathways of TAG metabolism will be discussed later. In
seed plants such as Arabidopsis, fatty acids are synthesized in the stroma of the chloroplast
(Ohlrogge et al., 1979). The assembly of glycerolipids occurs through two pathways in the
plastid and at the endoplasmic reticulum (ER) respectively (Roughan and Slack, 1982). In the
chloroplast, two fatty acids attached to the acyl-carrier protein (ACP) are incorporated into the
backbone of glycerol-3-phosphate sequentially, by the glycerol-3-phosphate acyltransferase
(GPAT, Figure 1.2, arrow 1) and lyso-phosphatidic acid acyltransferase (LPAT, Figure 1.2,
arrow 2) enzymes respectively, to produce phosphatidic acid (PtdOH). PtdOH can be converted
into diacylglycerol (DAG) through the removal of the phosphate group by the PtdOH
phosphatase (PAP, Figure 1.2, arrow 3). Alternatively, fatty acids can be exported to the ER and
activated into acyl-coenzyme A (acyl-CoA, Figure 1.2 arrows 10 and 11), the substrate for the
GPAT (Figure 1.2, arrow 12) and LPAT (Figure 1.2, arrow 13) enzymes associated with the ER
(Ohlrogge and Browse, 1995). Biochemical evidence indicates that in the plastid, PtdOH is the
precursor for phosphatidylglycerol (PtdGro, Figure 1.2, arrows 7, 8, 9) (Xu et al., 2006) while
monogalactosyldiacylglycerol (MGDG, Figure 1.2, arrow 5) and sulfoquinovosyldiacylglycerol
(SQDG, Figure 1.2, arrow 4) are derived from DAG with UDP-galactose and UDPsulfoquinovose as the headgroup donors (Benning and Somerville, 1992a, b; Yu et al., 2002).

5

Digalactosyldiacylglycerol (DGDG) is formed upon the transfer of another galactose to MGDG
(Figure 1.2, arrow 6) (Dormann et al., 1999). In plants, phosphatidylethanolamine (PtdEtn) and
phosphatidylcholine (PtdCho) are the major glycerolipids that are synthesized in the ER.
Previous radiolabeling experiments in plants showed that DAG backbones of glycerolipid
assembled at the ER can cycle back to the plastid for plastidic lipid synthesis (Roughan et al.,
1980). The two sources of plastid lipids are referred as the “eukaryotic” (ER) and “prokaryotic”
(plastid) pathways (Roughan and Slack, 1982). DAG backbones from these two pathways could
be distinguished based on the length of fatty acids at the sn-2 position; the LPAT in the
chloroplast prefers C16 fatty acids as opposed to the LPAT of the ER, which uses C18 acylCoAs (Frentzen, 1998). In Chlamydomonas, positional analysis showed that DAG backbones of
the plastidic lipids were almost exclusively from the prokaryotic pathway (Giroud et al., 1988).
This suggests that Chlamydomonas lacks the lipid trafficking from the ER back to the
chloroplast, which has been extensively studied in Arabidopsis (Figure 1.2, arrow 18) (Xu et al.,
2003; Awai et al., 2006; Lu et al., 2007; Xu et al., 2008).
For pathways that are common to Arabidopsis and Chlamydomonas, the latter is usually
less redundant. One prominent example is the galactolipid biosynthesis pathway. In Arabidopsis
MGD1 and DGD1 catalyze the synthesis of MGDG and DGDG respectively under normal
growth conditions, while MGD2, MGD3 and DGD2 are induced under phosphate deprivation to
produce more galactolipids as functional replacements to phosphatidylcholine (Benning and
Ohta, 2005). Additionally, a processive enzyme, galactolipid:galactolipid galactosyltransferase
(GGGT, a.k.a. SFR2) which uses MGDG instead of UDP-galactose as the galactose donor is
induced under freezing stress, generating trigalactosyldiacylglycerol (TGDG) (Moellering et al.,
2010). Among the above-mentioned glycosyltransferases, a BLAST (Altschul et al., 1997)

6

search of the Chlamydomonas genome (Riekhof et al., 2005) yielded orthologs for MGD1 and
DGD1 only. The lack of alternative galactolipid synthesis pathways may arise from the fact that
Chlamydomonas does not have phosphatidylcholine (Giroud et al., 1988) as discussed below and
phosphate deprivation would be a relatively mild stress.

Phosphatidylcholine (PtdCho) is absent in Chlamydomonas
In terms of lipid composition, the most striking difference between Chlamydomonas and
Arabidopsis is the lack of PtdCho but presence of a betaine lipid diacylglycerol-N,N,N,trimethylhomoserine (DGTS) in Chlamydomonas (Giroud et al., 1988). In plants, PtdCho plays a
central role in lipid metabolism. The fatty acids exported from the plastid were proposed to be
incorporated into PtdCho first (Bates et al., 2007), possibly by the action of an acyl-CoA:lysophosphatidylcholine acyltransferase (Stahl et al., 2008), then liberated by an unidentified
phospholipase (arrows 19 and 20 in Figure 1.2). PtdCho is also a major substrate for multiple
desaturases that produce polyunsaturated fatty acids. Through PtdCho synthesis and degradation,
the DAG backbone containing polyunsaturated fatty acids can be incorporated into TAG, a
pathway important for the industrial value of plant oil (Vogel and Browse, 1996; Bates et al.,
2012). Thus, the absence of PtdCho will result in substantial differences in Chlamydomonas
TAG metabolism from that of plants.
DGTS is a lipid which has been found in some algal species (Eichenberger, 1982), a
photosynthetic bacterium Rhodabacter sphaeroides (Benning et al., 1995), and non-seed plants
such as ferns (Eichenberger, 1993). Despite the lack of phosphorus, the headgroup of DGTS
resembles that of PtdCho (Figure 1.3). Furthermore, DGTS and PtdCho have similar biophysical
properties (Sato and Murata, 1991). It has been hypothesized that DGTS replaces PtdCho in its

7

structural role of as the major extraplastidic membrane lipid. The pathway of DGTS biosynthesis
was initially elucidated in Rhodabacter sphaeroides. BtaA encodes the enzyme that transfers the
3-amino 3-carboxylpropyl residue from S-adenosylmethionine to DAG while the BtaB protein
conducts the successive methylation of the N atom in the amino group (Klug and Benning,
2001). Interestingly, these two activities are harbored by a single protein named BTA1 in
Chlamydomonas. The protein sequence at the N-terminus of BTA1 is similar to that of the
bacterial BtaB while the C-terminus is similar to that of BtaA (Riekhof et al., 2005).

A general overview of triacylglycerol (TAG) biosynthesis in Chlamydomonas
Triacylglycerol is composed of a glycerol backbone and three fatty acyl chains which
usually contain 16-20 carbon atoms (Figure 1.1). TAG can be derived from DAG by one of the
two alternative routes: diacylglycerol acyltransferase (DGAT) or phospholipid:diacylglycerol
acyltransferase (PDAT). DGAT was initially discovered in human and mammals, with two
isoforms-DGAT1 and DGAT2. The enzymes transfer an acyl group from acyl-Coenzyme A to a
DAG backbone (Figure 1.4A), These two isoforms both catalyze the acylation of DAG but do
not share sequence similarity (Oelkers et al., 1998). In the seed plant Arabidopsis, DGAT1 but
not DGAT2 was found to contribute to seed oil biosynthesis (Zhang et al., 2009). More recently,
a novel protein conferring DGAT activity was found in developing peanut cotyledons; it has
sequence similarity to proteins of neither the DGAT1 nor the DGAT2 family. This protein was
thus designated DGAT3. An important difference is that DGAT3 is soluble while DGAT1 and
DGAT2 are membrane-bound (Saha et al., 2006; Rani et al., 2010).

8

PDAT activity was demonstrated for yeast and plants not longer after the discovery of the
first DGAT (Dahlqvist et al., 2000). The major difference between PDAT and DGAT is that
PDAT utilizes a phospholipid molecule instead of acyl-CoA as the acyl donor (Figure 1.4B).
For algae in various lineages, nutrient deprivation, especially nitrogen (N) deprivation, is
a common trigger for TAG accumulation. Examples include green algae such as Neochloris
oleobundans (Pruvost et al., 2009) or the eustigmatophyte Nannochloropsis sp. (Suen et al.,
1987). For diatoms, which contain silicates as their cell wall components, limitation of silicon
also induces TAG accumulation (Roessler, 1988). In addition to the nutrient availabilities, light
intensity and other factors such as temperature also play a role in regulating algal TAG
biosynthesis. For example, in Nannochloropsis salina, high light stress has been reported to
increase TAG content (Van Wagenen et al., 2012). Thus, besides the obvious role of storing
energy, TAG has been proposed to serve as an electron sink to prevent the formation of reactive
oxygen species from photosynthesis (Hu et al., 2008).
As a model green alga, Chlamydomonas has been repeatedly observed to accumulate
TAG under N-deprivation (Weers and Gulati, 1997; Wang et al., 2009; Moellering and Benning,
2010). Many studies on Chlamydomonas TAG metabolism have been performed on cells grown
photoheterotrophically, with acetate in the medium (Tris-acetate-phosphate, TAP) and constant
light. Under this condition, both acetate and light seem to be required for TAG production (Fan
et al., 2011; Goodson et al., 2011). Increasing the concentration of acetate from 17 mM to 60
mM greatly enhances TAG production (Fan et al., 2012). However, a parallel study was not
performed in which CO2 was passed through the medium and it remains unclear whether it is the
carbon source or the energy contained in acetate that contributes to the increase of TAG content.

9

As mentioned above, starch is another major energy reservoir in Chlamydomonas that is
stored upon N-deprivation and importantly, the starchless mutants were found to overaccumulate TAG (Work et al., 2010). This finding has significant value for industrial biofuel
production if it holds true for promising algal production species. Additionally, characterizations
of the starchless mutants will help elucidate the mechanism through which TAG synthesis is
induced.
Inhibition of fatty acid synthesis by cerulenin substantially decreased the amount of TAG
that accumulates under N-deprivation (Fan et al., 2011), indicating that de novo synthesized fatty
acids are important building blocks of TAG molecules. Indeed, the most abundant fatty acids
sequestered into TAG include the de novo synthesized fatty acids-16:0 and 18:1 (Fan et al.,
2011).

However, polyunsaturated fatty acids such as 18:3 and 16:4 were also observed

(Moellering and Benning, 2010). Since most desaturases act on the membrane glycerolipids
instead of free fatty acids or acyl-ACPs (Napier et al., 1999), the polyunsaturated fatty acids are
most likely obtained from the degradation of membrane lipids. This deduction is consistent with
the observation that the chloroplast of Chlamydomonas degenerates during N-deprivation
(Moellering and Benning, 2010).
The fact that approximately equal amounts of C16 and C18 fatty acids were contained in
TAG of Chlamydomonas (Fan et al., 2011) was initially surprising. In seed plants, TAG is
exclusively from extraplastidic origins. As a result, TAG contains only C18 fatty acids at the sn2 position and more than half C18 fatty acids in total (Lu et al., 2009). In contrast, positional
analysis showed that TAG in Chlamydomonas has mostly C16 fatty acids at sn-2, which
indicates a chloroplast origin for the DAG backbones used in TAG biosynthesis and explains the
abundance of C16 fatty acids in TAG (Fan et al., 2011).

10

Recent progress in understanding TAG metabolism through the genomics, genetics and cell
biology of Chlamydomonas
TAG is stored in globular organelles commonly referred to as lipid droplets (Farese and
Walther, 2009). In order to exist in the aqueous cytoplasm, the surface of lipid droplets is
covered with a monolayer of polar lipids, containing mostly phosphatidylcholine (Krahmer et al.,
2011). Since membranes are usually defined as bilayers of polar lipids, the monolayer of polar
lipids has been called a “hemi-membrane” (Tauchi-Sato et al., 2002). As discussed above,
PtdCho is lacking in Chlamydomonas and probably some other polar lipids function in its place
to coat the surface of lipid droplets.
Chlamydomonas also deviates from seed plants in the locations and protein composition
of lipid droplets. The major lipid droplet proteins differ between mammals (“PAT” family of
proteins) and plants (oleosins for seed lipid droplets) while for baker’s yeast, a major structural
protein has not yet been identified (Athenstaedt et al., 1999; Jolivet et al., 2004; Walther and
Farese, 2009). In Chlamydomonas, the most abundant protein in the lipid droplet proteome was
designated MLDP (Major Lipid Droplet Protein). Bioinformatic investigations showed that
orthologs of MLDP can only be found in green algal species (Moellering and Benning, 2010).
Despite the lack of sequence similarity between MLDP and Arabidopsis oleosin, RNA
suppression lines of MLDP gene in Chlamydomonas resembled the Arabidopsis oleosin mutants
in that lipid droplets are enlarged (Moellering and Benning, 2010). This suggests a similar
function between Chlamydomonas MLDP and plant oleosin in maintaining the surface structure
of lipid droplets.
In plant seeds, lipid droplets are cytosolic (Huang, 1996) but in senescent leaves, small
lipid droplets called plastoglubuli have been observed within the chloroplasts (Martin and

11

Wilson, 1984). In the unicellular alga Chlamydomonas, when a starchless mutant was used for
ultrastructural analysis of lipid droplets, surprisingly, two independent groups observed a
substantial number of lipid droplets which are similar to the cytosolic lipid droplets in size but
located in the plastid or connected to the plastidic envelopes (Fan et al., 2011; Goodson et al.,
2011). This might have to do with the above-mentioned finding that TAG in Chlamydomonas is
mostly derived from the DAG backbones of chloroplast origin (Fan et al., 2011). Localization
studies of the enzymes catalyzing TAG biosynthesis (discussed below) will be required to clarify
the process of lipid droplet biogenesis.
The Next-Generation Sequencing technique is revolutionizing molecular genetic research
for all organisms. The frequency of detection of a transcript by sequencing can serve as an
indicator of its abundance. Solexa sequencing (Illumina Inc.) combined with 454 sequencing
(454 Life Sciences) were carried out by Miller et al. (2010) to reveal genes differentially
regulated by N-deprivation. The transcript abundance of multiple genes encoding metabolic
enzymes or transcription factors increased or decreased. This study was followed by a time
course analysis of transcript abundance with time points ranging from minutes to days of Ndeprivation (Boyle et al., 2012). Both investigations obtained similar results regarding the
enzymes predicted to be involved in TAG metabolism. One example is that the DGTT1 gene
predicted to encode a type two diacylglycerol acyltransferase is increased more than 50 fold in
its transcript abundance after 48 h of N-deprivation. Specific genes such as the PDAT1 have
been selected from the databases for reverse genetic studies (Boyle et al., 2012).
One important discovery from both of the transcriptomic investigations is that a large
number of predicted lipase-encoding genes were observed to be differentially regulated by Ndeprivation. Among the 130 candidate lipase genes in Chlamydomonas, 35 (27%) are up-

12

regulated and 11 (8.5%) are down-regulated while the numbers for all of the genes in the genome
average are 13% and 11% respectively. Possible functions of these putative lipases are discussed
below.

Potential roles of lipases in triacylglycerol metabolism
The term “lipase” was originally used for the enzymes that degrade TAG (Feinberg,
1991). The enzymes that hydrolyze phospholipids are referred to as phospholipases, which can
be further divided into phospholipase A1 (PLA1), phospholipase A2 (PLA2), phospholipase B
(PLB), phospholipase C (PLC) and phospholipase D (PLD). PLC and PLD digest the
phosphodiester bonds on the glycerol backbone side and the polar headgroup side respectively
while PLA1 and PLA2 remove the acyl chains at the sn-1 and sn-2 positions respectively
(Vandeene and Dehaas, 1966). If a phospholipase sequentially cleaves off the two acyl groups, it
is designated as PLB (Wang, 2001). Both TAG lipases and PLA1/PLA2 catalyze the hydrolysis
of the carboxyl ester bonds between the acyl chains and their glycerol backbone. In fact, multiple
proteins were found to possess lipolytic activity against both TAG and phospholipids (Kuusi et
al., 1982). In the recent years, the term “lipase” has been used to denote many proteins with acyl
hydrolase activity against polar lipids (phospholipids, galactolipids, etc.) instead of
triacylglycerol (Ellinger et al., 2010; Bonaventure et al., 2011) . In this dissertation, “putative
lipase” will also be used to describe a protein predicted to have an acyl hydrolase motif but
lacking information about its substrate specificity.
TAG metabolism can be affected by lipases in different ways. Intuitively, a TAG
molecule can be hydrolyzed by a TAG lipase (reaction 1 in Figure 1.5), DAG lipase and

13

monoacylglycerol (MAG) lipase consecutively. In Baker’s yeast, deletion of genes encoding
TAG lipases from the genome has been shown to double the amount of TAG accumulated
(Athenstaedt and Daum, 2005; Kurat et al., 2006). The system in which TAG degradation has
been most intensively studied is the mammalian adipose tissue. In human beings, the hormone
sensitive lipase (HSL) was first found to liberate fatty acids from fat in the adipose tissue
(Vaughan et al., 1964) and was thought to be the major TAG lipase until the recent finding of the
adipose triglyceride lipase (ATGL) (Zimmermann et al., 2004). Characterizations of mutants
deficient in these two enzymes and in vitro kinetic studies indicate that ATGL is the major TAG
lipase while HSL mostly degrades the DAG generated by ATGL activity to form MAG, a
substrate for the third lipase-monoacyglycerol lipase (MAG) (Zechner et al., 2009).
Activities of the above-mentioned lipases are regulated at multiple levels. Insulin inhibits
phosphorylation of HSL, which facilitates the attachment of cytosolically localized HSL to the
surface of lipid droplets (Slavin et al., 1994). An adipogenic transcription factor PPARgamma
was reported to transcriptionally activate mouse ATGL (Kim et al., 2006). In the recent years,
CGI-58, a protein which shares sequence similarities to lipases but with the serine catalytic
center replaced during evolution, emerged as an important regulator for fat mobilization. Binding
to CGI-58 is required for ATGL to exhibit full TAG hydrolysis activity (Zimmermann et al.,
2009).
During the past decade, major TAG lipases have been identified in Arabidopsis
(Eastmond, 2006), fruit fly (Gronke et al., 2005) and the budding yeast (Athenstaedt and Daum,
2003, 2005; Kurat et al., 2006). These lipases as well as the mammalian ATGLs belong to the
family of patatin-like lipases (Zimmermann et al., 2004) while many members of this family
exhibit lipolytic activity against phospholipids instead of TAG (Banerji and Flieger, 2004).

14

Lipases that attack polar lipids can also affect the metabolism of TAG. The acyl group
generated by polar lipid hydrolysis can be used a substrate for the DGAT reaction (reaction 2,
Figure 1.5). In plants, it has been proposed that the fatty acid exported from the chloroplast to
cytosol, after being activated into acyl-CoA, is first incorporated into PtdCho through an
acyltransferase. A phospholipase then removes this acyl group to the cytosolic acyl-CoA pool,
which can be used for the biosynthesis of TAG and other extraplastidic lipids (Bates et al.,
2007). In Chlamydomonas, PtdCho is absent but it remains possible that another lipid serves this
role.
Besides the trafficking of fatty acids through the acyl-editing mechanism, breakdown of
membrane components in Chlamydomonas may also require lipases. Under N-deprivation, the
fatty acids contained in TAG include polyunsaturated species such as 16:4 and 18:3, which could
only be obtained by membrane lipid degradation (Moellering and Benning, 2010). The lipases
involved in this process are likely encoded by those genes up-regulated by N-deprivation (Miller
et al., 2010).
Last but not least, lipases may affect cell metabolism by producing or removing lipid
signaling molecules. In plants, upon pathogen infection, two galactolipases, DGL and DAD,
remove the 18:3 fatty acids from plastidic lipids, which are then utilized by the oxylipin pathway
to produce the phytohormone jasmonic acid (JA) (Hyun et al., 2008; Ellinger et al., 2010). While
the presence of JA in Chlamydomonas is not reported, there are probably other lipid molecules
playing signaling roles.

15

Aims of the thesis research
This thesis aims to identify novel genes involved in TAG metabolism in
Chlamydomonas. Three investigations have been performed: Chapter 2 describes reverse
genetics on several candidate TAG lipases. One lipase designated CrLIP1 was discovered and
characterized. Chapter 3 reports a forward genetic screen for mutants with abnormal amounts of
TAG. A mutant deficient in TAG accumulation was characterized extensively. Chapter 4
describes further the low-TAG mutant, which was used as a tool to explore the physiological
significance of TAG accumulation.

16

Figure 1.1. Structure of triacylglycerol and transesterification of triacylglycerol to produce
fatty acid methyl esters. R1, R2 and R3 represent hydrocarbon chains.

17

Figure 1.2 Comparison of major membrane glycerolipid biosynthesis pathways between
Arabidopsis and Chlamydomonas. Regular arrows indicate processes that can be found in both
organisms. Dashed arrows represent processes unique to Arabidopsis while the bold arrow is
unique to Chlamydomonas. Only major substrates or products are shown for each reaction.
Phosphatidylserine and phosphatidylinositol are less relevant for discussions in this dissertation
and are skipped. Phosphatidylglycerol is also partly synthesized in extraplastidic membranes, but
in trace amounts compared to the plastic pathway and is thus skipped. Abbreviations: ACP, acyl
carrier protein; AdoMet, S-adenosyl methionine; CDP-Cho, cytidine diphosphate-choline; CDPDAG, cytidine diphosphate-diacylglycerol; CDP-Etn, cytidine diphosphate-ethanolamine; CTP,

18

Figure 1.2 (cont’d)
cytidine triphosphate; DAG, diacylglycerol; DGDG, digalactosyldiacylglycerol; DGTS,
diacylglycerol-N,N,N,-trimethylhomoserine; FAS, fatty acid synthesis; G-3-P, glycerol-3phosphate; lyso-PtdOH, lyso-phosphatidic acid; MGDG, monogalactosyldiacylglycerol; PGP,
phosphatidyllgycerol-phosphate;

PtdCho,

phosphatidylcholine;

PtdEtn,

phosphatidylethanolamine; PtdGro, phosphatidylglycerol; PtdOH, phosphatidic acid; SQDG
sulfoquinovosyldiacylglycerol; UDP-Gal; UDP-SQ, uridine diphosphate-sulfoquinovose. Names
of enzymes or processes for each arrow: 1, plastid glycerol-3-phosphate acyltransferase; 2,
plastid lyso-phosphatidic acid acyltransferase (chloroplast); 3, plastid phosphatidic acid
phosphatase; 4, SQDG synthase; 5, MGDG synthase; 6, DGDG synthase; 7, CDP-DAG
synthase; 8, PGP synthase; 9, PGP phosphatase; 10, thioesterase; 11, long-chain acyl-CoA
synthetase; 12, ER associated glycerol-3-phosphate acyltransferase; 13, ER-associated lysophosphatidic acid acyltransferase; 14, ER associated phosphatidic acid phosphatase; 15, CDPethanolamine:diacylglycerol ethanolaminephosphotransferase; 16, betaine lipid synthase; 17,
CDP-choline:diacylglycerol cholinephosphotransferase; 18, lipid trafficking from ER to plastid,
catalyzed by multiple protein components; 19, acyl-CoA:lyso-PtdCho acyltransferase 20,
phospholipase.

19

Figure 1.3. Structures of PtdCho and DGTS. R represents a hydrocarbon chain.

20

Figure 1.4. TAG synthesis catalyzed by DGAT and PDAT activities. X represents an organic
group that varies between different phospholipids.

21

Figure 1.5. Lipases affect TAG content by degrading TAG or providing acyl groups for
TAG synthesis. X represents a polar headgroup that varies between different polar lipids.

22

Chapter 2
Rapid triacylglycerol turnover in Chlamydomonas reinhardtii requires a lipase with broad
substrate specificity

1

1

This research was published in: Li, X., C Benning and M Kuo. (2012) Rapid triacylglycerol
turnover in Chlamydomonas reinhardtii requires a lipase with broad substrate specificity.
Eukaryot.Cell. Epub ahead of print. I contributed to the data shown in all of the figures.

23

ABSTRACT
When deprived of nitrogen (N), the photosynthetic microalga Chlamydomonas
reinhardtii accumulates large quantities of triacylglycerols (TAG), affording a promising source
of biofuel. Prominent transcriptional changes have been found associated with conditions leading
to TAG accumulation, suggesting that key enzymes for TAG metabolism might be among those
fluctuating in their expression during TAG synthesis and breakdown. Using a Saccharomyces
cerevisiae lipase null mutant strain for functional complementation, I identified the CrLIP1gene
from Chlamydomonas based on its ability to suppress the lipase deficiency-related phenotypes of
the yeast mutant. In Chlamydomonas, an inverse correlation was found between the
CrLIP1relative transcript level and TAG abundance when Chlamydomonas cultures were
reversibly N-deprived. CrLIP1 protein expressed and purified from E. coli exhibited lipolytic
activity against diacylglycerol (DAG) and polar lipids. The lipase domain of CrLIP1 is most
similar to two human DAG lipases, DAGLα and DAGLβ. The involvement of CrLIP1 in
Chlamydomonas TAG hydrolysis was corroborated by reducing the abundance of the CrLIP1
transcript with an artificial microRNA, which resulted in an apparent delay in TAG lipolysis
when N was re-supplied. Together, these data suggest that CrLIP1 facilitates TAG turnover in
Chlamydomonas primarily by degrading DAG presumably generated from TAG hydrolysis.

24

INTRODUCTION
Plants, animals and fungi accumulate triacylglycerols (TAGs) to store excessive carbon
and energy. For industrial biofuel production, triacylglycerol can be converted into fatty acid
methyl esters (FAMEs)through transesterification, the major components of biodiesel (Durrett et
al., 2008). TAG harvested from cells can thus be a source of biofuel. Thanks to the faster growth
rate and higher oil content, microalgae can in theory produce TAG at yields surpassing those of
plants (Luo et al., 2010b). However, details of lipid metabolism in algae are poorly understood.
Chlamydomonas reinhardtii has a long history of serving as a model system to study cell
motility and photosynthesis (Rochaix, 1995; Merchant et al., 2007; Harris et al., 2009). The
recent development of genetic and molecular biological tools such as nuclear genome
transformation (Kindle, 1990) and gene silencing with artificial microRNA (Molnar et al., 2009)
has also made Chlamydomonas an excellent model alga for a wider spectrum of research topics,
including lipid metabolism and biofuel production. In Chlamydomonas the glycerolipid
composition was explored (Giroud et al., 1988) and the lipid metabolism pathways have been
annotated in silico(Riekhof et al., 2005). Certain enzymes for TAG metabolism have also been
characterized experimentally(Boyle et al., 2012). Similar to many other algae, Chlamydomonas
accumulates TAG under nitrogen deprivation(Wang et al., 2009; Moellering and Benning, 2010).
In plants, TAG biosynthesis starts with plastid production of fatty acids, which are then
esterified to glycerol-3-phosphate, forming phosphatidic acid (PtdOH). The endoplasmic
reticulum (ER) is another site for lipid assembly and fatty acid modification (Roughan and Slack,
1982). Diacylglycerol is produced after the removal of the phosphate by a PtdOH phosphatase.
Diacylglycerol

acyltransferases

(DGAT)

and

phospholipid:diacylglycerolacyltransferases(PDAT) then transfer an acyl chain from acyl

25

coenzyme A or a phospholipid to DAG to synthesize TAG (Zhang et al., 2009), which can be
considered to be a vehicle for carbon storage. TAG turnover (lipolysis) is initiated by the action
of TAG lipases, which generate DAG and free fatty acids (Eastmond, 2006). DAG can be further
hydrolyzed into monoacylglycerol and glycerol. Certain lipases hydrolyze all acyl chains from
the glycerol backbone, while others may act specifically on one kind of glycerolipid(Utsugi et
al., 2009). To engineer TAG content, both TAG biosynthetic and hydrolytic enzymes have to be
considered. For example, overexpressing an Arabidopsis DGAT increases seed oil (Jako et al.,
2001), whereas deleting the major TAG lipases in the budding yeast Saccharomyces cerevisiae
results in enhanced accumulation of TAG in cells (Athenstaedt and Daum, 2003, 2005; Kurat et
al., 2006). In Chlamydomonas, functions of two DGATs and a PDAT have been described
(Boyle et al., 2012; La Russa et al., 2012a, b). Lipases for the TAG turnover chain reactions are
yet to be identified.
Like many other microalgae, Chlamydomonas reinhardtii cells do not maintain a large TAG pool
during vegetative growth, but do so following nutrient deprivation, most notably nitrogen (N).
Seventy-two hours after N removal, TAG accounts for 50% of all cellular fatty acids, which
results from both increased fatty acid biosynthesis and membrane remodeling (Miller et al.,
2010). These metabolic flux changes are accompanied by genome-wide adaptation of
transcription, including a large number of putative lipase-encoding genes that are up- or downregulated (Miller et al., 2010). I reasoned that the repressed genes during TAG accumulation
may encode lipases that degrade TAG or its initial breakdown products (i.e., diacylglycerol and
monoacylglycerol). The N-deprivation-induced lipase genes, on the other hand, might be
involved in shuffling fatty acids from membrane lipids to TAG. Here, I describe the
identification of a lipase-encoding gene required for rapid TAG turnover in Chlamydomonas.

26

MATERIALS AND METHODS
Yeast methods
Yeast cells were grown according to standard procedures (Sherman, 1991). Yeast
transformation was performed using the lithium acetate method as previously described (Gietz et
al., 1992). Diameters of yeast cells were determined using the MicroSuite Basic software
(Olympus) after conducting microscopy. A uniform lower limit of 4.1 µm long-axis cell length
was arbitrarily set to exclude young daughter cells. All yeast strains (Table 1)were derived from
theyMK839 (Kuo et al., 1998) background, a derivative of S288C.To delete TGL4, PCR was
done with primers 4DF and 4DR (all primer sequences are shown in Table 2.3) to amplify
Kluyveromyceslactis TRP1 from pBS1479 (Rigaut et al., 1999). The PCR product was then
transformed into yMK839 for tryptophan prototroph selection to obtain yXL001. The deletion
was verified by genomic PCR. To delete TGL3from yXL001, PCR was performed to amplify
KanMX6 from pFA6a-KanMX6 (Wach et al., 1994). The PCR product was then gel purified and
used as the template for another PCR with primers 3DF and 3DR. The new PCR product was
transformed to yXL001 and cells were plated onto YPD with 200μg/ml G418 for selection,
creating yXL005 after genomic PCR verification of the tgl3∆::KanMX allele. In some cases,
yeast cells were grown in casamino acids (CAA) medium minus uracil (CAA –ura) that was
essentially identical to the synthetic complete medium (Sherman, 1991) except the supplement of
0.5% casamino acids and uracil dropout.
To create a cDNA clone for CrLIP1 expression in yeast, Chlamydomonas RNA was
prepared with the RNeasy Plant Mini Kit (Qiagen). Reverse transcription was conducted with
Superscript II reverse transcriptase (Invitrogen) to obtain cDNA. CrLIP1 coding sequence was

27

amplified using Phusion polymerase (NEB) and primers 595F and 595R. The PCR product was
co-transformed into yeast cells together with a NotI linearized yeast expression vector pMK595
(Luo et al., 2010a) by homologous recombination (Ma et al., 1987). A single clone of pMK595
with CrLIP1 integrated (pMK595CrLIP1) was obtained through transformation of the yeast
crude DNA into E. coli and sequenced to be correct (Table 2.2). pMK595 and pMK595CrLIP1
were transformed into yMK839 to obtain yXL023 (referred to as WT-vectin Figures) and
yXL026 (WT-CrLIP1), and into yXL005 to obtain yXL077 (tgl3Δtgl4Δ-vect) and yXL080
(tgl3Δtgl4Δ-CrLIP1) respectively.

Chlamydomonas strains and growth conditions
+

The cell wall-less strain dw15-1 (cw15, nit1, mt ) obtained from Arthur Grossman
(Carnegie Institution, Stanford) was used for all experiments performed on Chlamydomonas.
Liquid and solid cultures were grown in Tris-acetate-phosphate (TAP) medium as described
before(Moellering and Benning, 2010). For N-deprivation or re-supply, cells were collected by
centrifugation at 3000 x g (4°C, 3 min), washed twice and re-suspended in TAP-N (TAP medium
with NH4Cl omitted) or TAP respectively.

PCR
Quantitative real-time PCR on CrLIP1 was performed using primers qRT-F and qRT-R
with the Applied Biosystems 7500 Fast Real-time PCR system. Data were normalized to the
commonly used RACK1gene using primers RACK1-F and RACK1-R (Chang et al., 2005). The
2(-∆∆C(T)) method(Livak and Schmittgen, 2001)was employed for data analysis. RT-PCR for

28

the agarose gel electrophoresis was performed using the same primers with the real-time PCR for
RACK1 and CrLIP1. For DGTT1, DGTT1-F and DGTT1-R were used.

Western blotting
To examine the expression of the CrLIP1cDNA in yeast, log-phase yeast cells were
collected by centrifugation (3000 x g, 5 min) and then re-suspended in 2X SDS-PAGE loading
dye (0.12 M Tris-HCl, pH 6.8, 0.04%bromophenol blue,4%SDS,20%glycerol, 5.6% βmercaptoethanol) with the same volume of acid-washed glass beads (0.45 mm, Sigma-Aldrich).
The mixture was boiled in a water bath for 5 min followed by 5 min of vortexing. After an
additional cycle of boiling and vortexing, the lysates were centrifuged at 21,000 x g for 1 min
and the supernatant was analyzed by SDS-PAGE. To examine the subcellular localization of
recombinant CrLIP1, yeast cell pellets were suspended in 0.1 M PBS (pH 7.4) containing the
Complete protease inhibitor cocktail (Roche, one tablet for every 10 ml of buffer). Glass beads
(0.45 mm) were added, followed by vigorous agitation in a Mini-Bead Beater (BioSpec) for 45
sec and 1-minute ice incubation. The beating was repeated 3 times. After the last bead-beating,
lysates were collected by spinning through a pinhole punched at the bottom of the tube (1,000 x
g, 5 min, 4°C). The supernatant of the low-speed spinning was fractionated by 100,000 x g
centrifugation (90 min, 4°C). The secondary supernatant was defined as the soluble fraction in
Figure 2.1. The membrane-enriched pellet was suspended in PBS containing the protease
inhibitor tablet. For immunoblotting, the 12CA5 monoclonal antibody (Roche) was used to probe
for proteins with an HA epitope tag and the His monoclonal antibody (H1029, Sigma-Aldrich)
was used for the hexahistidine-tagged protein in E. coli.

29

Lipid analysis
To extract lipids, Chlamydomonas and E. coli cells were harvested and re-suspended in
the lipid extraction solvent containing methanol, chloroform, andformic acid (88%) (2:1:0.1 by
volume). In addition, the bacterial cells expressing CrLIP1 were shaken at 26°C for 6 hours after
overnight IPTG (isopropyl-β-D-thiogalactopyranoside) induction before cells were collected
(described below). I found that this extra step allowed better lipase action in E. coli. For lipid
extraction, Chlamydomonas and E. coli cells were vortexed for 30 sec in the extraction solvent
for cell lysis. For yeast, cells were re-suspended in the solvent; and glass beads (0.45 mm) were
added. Yeast lipids were extracted by using a Mini-Bead Beater (BioSpec) as described above
for protein preparation, except that 0.5 volume of lipid extraction buffer (1M KCl, 0.2M H3PO4)
was added to the organic phase after its collection through spinning through a pinhole at the
bottom of the tube. For Thin-layer chromatography (TLC), lipids were loaded onto Silica Gel 60
plates (EMD chemicals, #5721-7) and developed in petroleum ether-diethyl ether-acetic acid
(80:20:1 by volume, to separate neutral lipids) or chloroform-methanol-acetic acid-H2O
(75:13:9:3 by volume, for separation of polar lipids). Lipids were visualized by exposure to
iodine vapor. For quantification of fatty acid methylesters derived from a certain lipid species on
the TLC plate, corresponding spots were scraped and subjected to transesterification and GLC as
previously described(Rossak et al., 1997).

Yeast metabolic labeling
To measure the rate of lipid synthesis, yeast cells were grownin CAA –ura medium for
14

48 hat 26°C to early stationary phase. To each 5-ml culture 20μCi [ C] acetate (specific activity

30

45-60 mCi/mmol; Perkin Elmer)was added. Cultures were then continually shaken (app. 150 rpm
in a Lab-Line floor incubator) at 26°C and were harvested 25 min, 40 min, 65 min, 100 min and
140 min after acetate supplementation. Total lipids were extracted (see above) and resolved by
TLC resolution (petroleum ether:diethylether:acetic acid = 80:20:1 by volume)that separated
neutral lipids. TAG, DAG, and TLC origins (containing a mixture of polar lipids) were isolated
for radioactivity measurement by liquid scintillation counting. For yeast cellularTAG hydrolysis
analysis, 50μCi (0.83-1.11 μmol) [14C]-acetate was added to 5 ml cultures grown for 48 hat
26°C. After 4 h of labeling, cells were collected by centrifugation, washed with fresh medium
and diluted 28 fold with fresh CAA –ura medium to an optical density (at 600 nm) of 0.4-0.6.
The cells were shaken for another 5 h with aliquots frozen at each time point. The growth curves
were determined with cells grown under exactly the same conditions except that 1 μmol nonlabeled acetate was used. Lipid analysis was as above for lipid synthesis rate analysis.

Recombinant protein production
The CrLIP1 coding sequence was amplified using primers LICF and LICR and the
Phusion polymerase. The PCR product was then integrated into an expression vector pMK1006
through ligation-independent cloning(Aslanidis and de Jong, 1990). BL21 CodonPlus
(Stratagene) transformed with empty pMK1006 vector or pMK1006CrLIP1 were grown to the
log phase at 37°C and protein production was induced by addition of IPTG to a final
concentration of 0.5 mM. After 16 h of shaking at 220 rpm at 16°C, cells were harvested by
centrifugation (6,000 x g, 5 min, 4°C). For protein preparation, cells were collected and
suspended in the lysis buffer (20 mM Tris-HCl, pH 7.9, 10% glycerol, 150 mMNaCl, 1 mM
dithiothreitol) and subjected to three freeze-thaw cycles, followed by sonication on ice (90 times

31

of alternating 1-sec pulse and 0.3-sec rest, with 1 minute ice-chilling after every 15 cycles;
Branson digital sonicator 250 with 25% energy output was used). Cell-free extracts were
obtained through centrifugation at 21,000 x g for 15 min and applied to a Ni-NTA affinity
purification column (Qiagen). CrLIP1 was eluted with the lysis buffer supplemented with 200
mM imidazole. A spectrophotometric assay with the Bio-Rad Protein Assay Dye Reagent
Concentrate (BioRad #500-006 EDU)(Bradford, 1976) was used to determine protein
concentrations.

Lipase assay
For the lipase assay with various substrates, lipids dissolved in organic solvents were
dried under a stream of nitrogen gas. The dried lipid was then dissolved in 350 µL of 0.1 M
phosphate saline buffer (PBS, pH 7.4) containing varying concentrations of Triton X-100 (see
below) by sonication (Sonicator 3000 with a Misonix microprobe)for 6 x 10 sec (power setting
1.5). Purified CrLIP1 protein or the same volume of protein storage buffer was added.
Dithiothreitol (DTT) was added fresh to the final concentration of 2 mM. The mixture was then
vortexed and incubated at room temperature.
For the TAG lipase assay, 10 µCitriolein (specific activity 30-120 Ci/mmol; Perkin
Elmer) and 5.6nmol olive oil was dried and re-suspended in 350 µL PBS containing 0.11 mM
Triton X-100. The mixture was split in halves and 18µg of purified CrLIP1 in 20 µL protein
storage buffer was added to one of the aliquots. The mixture was then vortexed vigorously for 5
sec and incubated at room temperature for 12 hr. I tried multiple other conditions, with varied
amounts of olive oil (0-250 nmol), DTT or supplementation of CaCl2. Emulsifiers including
phosphatidylcholine/phosphatidylinositol mix and BSA were also used to replace Triton X-100

32

without an observable effect on the result. For the DAG lipase assay, 150 nmoldiolein (Avanti
Polar Lipids) was dried and sonicated as described for TAG, except that 0.53 mM instead of 0.11
mM Triton X-100 was used. Purified CrLIP1, 18µg in 50 µL protein storage buffer, was added.
The mixture was then vortexed briefly and incubated at room temperature for 16 hr.
For assays containing radioactive polar lipids, Chlamydomonas cells were grown to log
14

phase in TAP medium, harvested and suspended in medium with 30 µCi of [ C] acetate
mentioned above (final concentration 20 - 23.3 µM). Cells were harvested after another 4 h of
standard growth (see above). Total lipids were extracted and resolved by TLC developed in
chloroform-methanol-acetic acid-H2O (75:13:9:3 by volume). Individual lipids were visualized
by radiography and then scraped off the plate before extraction with chloroform-methanol (1:1
by volume) from silica gel. Lipids (35 nmol with total radioactivity of 10,000-40,000 dpm) were
dried and dissolved as described for TAG. For lipase assays, the emulsified lipid substrates were
divided into halves, with 18 µg purified CrLIP1 added to one set and the same volume of protein
storage buffer added to the other set. The mixture was incubated for 12 h at room temperature.
To prepare radioactive steryl esters for in vitro lipolysis reactions, 1 ml of early stationary-phase
tgl3Δtgl4Δ-vect yeast cells were diluted to 25 ml fresh CAA –ura medium with additional 30 µCi
of [14C]-acetate and grown for 48 h at 26°C. Total lipids were extracted and resolved by TLC as
described above. Steryl esters were then isoaltedfrom the TLC plate. Assay conditions were the
same as lipids derived from Chlamydomonas cells, with 35 nmol labeled steryl esters (10,
9

000dpm) as the substrates. For assays with PtdCho, 60 nmol 18:1∆ /16:0 PtdCho (Sigma9

Aldrich) or 18:1∆ /16:0 PtdCho (Avanti Polar Lipids) was dried and dissolved as described for
TAG. CrLIP1 (36 µg) was added and aliquots were flash frozen at 0, 6 and 16 h. The 0 h control

33

was used to verify the intactness of PtdCho. The mixture was incubated for 12 h at room
temperature. Background levels of fatty acids carried over with purified CrLIP1 protein were
estimated in a control reaction without substrate lipid supplied, and subtracted from the free fatty
acids data obtained with substrate.
All the lipase reactions were quenched by adding two volumes of lipid extraction solvent.
Lipid products were then analyzed by TLC as described above. For Rhizopusarrihizus lipase
(Sigma-Aldrich) digestion, 200 µg of the enzyme preparation was used and the incubation time
was shortened to 20 min.

AmiRNA construct
The artificial microRNA construct to silence CrLIP1 expression was generated according
to Molnar et al. (2009). Briefly, four primers were designed using the MicroRNA Designer
protocols (http://wmd3.weigelworld.org/cgi-bin/webapp.cgi?page=Help). The primers were
annealed by 5 min of boiling and overnight cooling to form a double-stranded fragment with
CrLIP1 targeting sequences and overhangs compatible with the SpeI digestion site. This
fragment was then integrated into SpeI linearized pChlamiRNA3int vector (Molnar et al., 2009)
through ligation and transformation. To generate CrLIP1 knock-down lines, the artificial
microRNA construct (pChlamiRNA3intCrLIP1) or the empty vector was digested with KpnI
(NEB)and transformed

into the dw15-1Chlamydomonasstrain

using the glass bead

transformation method (Kindle, 1990). TAP agar containing 10 µg/ml paromomycin was used
for selection. Real-time PCR described above was used to screen for lines with reduced mRNA
abundance of CrLIP1.

34

Bioinformatics
A BLAST search(Altschul et al., 1997) of CrLIP1 protein sequence against human
genome was performed on National Center for Biotechnology Information website
(http://blast.ncbi.nlm.nih.gov/). Protein sequences were aligned with the ClustalW2 program
(Larkin et al., 2007).

RESULTS
Heterologous expression of CrLIP1 complemented a yeast mutant deficient in TAG lipase
activity
To identify the Chlamydomonas lipases involved in TAG turnover, Iuseda yeast mutant
deficient in TAG lipase for functional complementation. In yeast, the major TAG lipases are
encoded by TGL3 and TGL4(Athenstaedt and Daum, 2003, 2005; Kurat et al., 2006). Their
deletion results in hyper-accumulation of TAG in stationary-phase cells. This phenotype can be
rescued by ectopic expression of a mouse adipose triglyceride lipase (Kurat et al., 2006). To
choose Chlamydomonas candidate genes for a lipase that suppresses the yeast lipid overstocking
phenotype, putative lipases based on Gene Ontology of the JGI Chlamydomonas 3.0 database
were cross-referenced toour transcriptomic studies of N-replete and depleted cells (Miller et al.,
2010), and to the data from a lipid droplet proteomics study(Moellering and Benning,
2010).Eight genes were selected for initial studies. The respective cDNAs were placed under the
control of the constitutive ADH1 promoter from a yeast high-copy plasmid. An N-terminal triple
HA tag was added to the coding sequence. These genes were tentatively named CrLIP1 through
CrLIP8 (summarized in Table 2.4). CandidatesCrLIP1, CrLIP3, CrLIP5, CrLIP6 and
CrLIP8were successfully cloned and expressed in the tgl3∆ tgl4∆ double knockout mutant. Of

35

the five genes tested, CrLIP1 (protein ID 184308 and 519543, JGI Chlamydomonas 4.0; a.k.a.
FAP12 (Pazour et al., 2005), see Discussion)consistently showed reduction of TAG and changes
in cell growth (see below). For the remainder of this work, I focused on CrLIP1 for its role in
lipolysis.
I first examined the expression of CrLIP1 in yeast. As expected for a lipase, the majority
of recombinant CrLIP1 was associated with the membrane-enriched fraction that also included
lipid droplets and TAG (Figures 2.1A -C). Importantly, the TAG over-accumulation phenotype
of thetgl3Δtgl4Δ mutant (Athenstaedt and Daum, 2005) was partially suppressed by CrLIP1
overexpression (Figure 2.1D). There was no discernible change in the steady-state levels of DAG
or major phospholipids, initially suggesting that CrLIP1 primarily affected the TAG
accumulation or hydrolysis in yeast. Although CrLIP1 did not decrease the TAG content in wildtype background to a statistically significant level (Figure 2.1D), it caused consistent differences
in the fatty acid compositions of TAG in both wild-type and thetgl3Δtgl4Δ mutant, with an
increase in the relative amount of 16:1 and a decrease in 16:0 (Figure 2.1E). This might be due to
substrate specificity of CrLIP1 if it is a bona fide lipase. Besides these lipid phenotypes, I also
noticed that the tgl3∆ tgl4∆ mutant grew to a higher optical density (OD) in the stationary phase
(Figure 2.1F). Overexpressing CrLIP1 prevented this OD increase. One of the underlying
reasons for this phenomenon can be ascribed to the changes in cell size, as revealed by the
measurement of cellular diameter under a light microscope (Figure 2.1G). Overall, the double
lipase knockout strain had a higher number of large cells than did the wild-type strain. Larger
particles should absorb more light, resulting in higher OD. Over-expressing CrLIP1 in the
tgl3Δtgl4Δ background caused the overall cell size distribution to approach that of the wild-type

36

cells. Together, these results demonstrated that the Chlamydomonas CrLIP1 protein was able to
functionally complement the yeast TAG lipases Tgl3 and Tgl4.

Heterologous expression of CrLIP1 lowers the rate of TAG biosynthesis in yeast
To explore the role of CrLIP1 in yeast TAG metabolism in a more direct way, metabolic
labeling experiments were performed. Stationary phase cells were chosen as TAG accumulates
when cells enter stationary phase (Larkin et al., 2007). Additionally, from 48 h to 60 h, the
increase in cell number was within 4 % for all strains tested (Figure 2.1F), thus the cell number
during the 2-h labeling period was regarded as constant, hence permitting us to normalize
changes in lipid labeling to cell numbers. The radioactivity incorporated into TAG, DAG, and
polar lipids (mainly phospholipids in yeast) was measured (Figure 2.2A-C). Consistent with the
steady state lipid analysis results (Figure 2.1D), CrLIP1 lowered the rate of TAG accumulation
(Figure 2.2A) while the effect on polar lipid was relatively minor (Figure 2.2B). Interestingly,
labeling of DAG increased in the first 90 min after the addition of radioactive acetate, and
dropped afterwards. Figure 2.2C showed that the incorporation of label into DAG was faster than
that of TAG (Figure 2.2A), consistent with DAG being the precursor for TAG biosynthesis. The
expression of CrLIP1 did not affect the initial DAG labeling, but rather accelerated its reduction.
As the immediate precursor of TAG, the reduction in DAG labeling likely reflected its
conversion to TAG. The faster appearance of label in DAG than in TAG and the decrease of
label in DAG after 90 min are consistent with a substrate product relationship of DAG and TAG,
respectively. The faster decline of DAG label in the presence of CrLIP1 would be consistent with
CrLIP1 being a DAG lipase, which could affect the synthesis of TAGs as observed. Indeed, in
vitro experiments (below) confirmed that CrLIP1 has DAG lipase activity.

37

When growing out of the stationary phase, yeast degrades TAG to produce precursors
supporting multiple cellular functions, including membrane polar lipid biosynthesis and energy
consumption (Larkin et al., 2007). To shed light on the function of CrLIP1 in yeast, I employed
metabolic pulse-chase labeling. Toward this end, stationary phase yeast cells were labeled with
radioactive acetate and then transferred to fresh medium without the label. Radioactivity in TAG
and total polar lipids was quantified. During the course of the chasing period, all strains showed
varying degrees (up to about 30%) of increase in the optical density (Figure 2.2F), a result of
cells undergoing mitosis. To accommodate the differences in cell density, I measured the
radioactivity of TAG and polar lipids over a unit volume (per ml) of all strains, and normalized it
to the starting counts. During the outgrowth, the label in cellular polar lipids increased whereas
TAG labeling remained constant in wild-type cells (open squares, Figures 2.2D and 2.2E).
Labeling of DAG was very low at the times tested, consistent with very rapid turnover of the
DAG pool following the onset of the chase phase of the experiment. The increase of total
radioactivity contained in polar lipids likely resulted from the incorporation of persisting
radiolabeled acetate and fatty acids. On the other hand, the total amount of TAG remained steady,
which I suggest to be a result of balanced TAG synthesis and hydrolysis. Indeed, in the tgl3∆
tgl4∆ double lipase knockout strain (open triangles, Figures 2.2D and 2.2E), the total TAG
amount increased drastically. Importantly, consistent with a role in lipolysis, expression of
CrLIP1 in both wild-type and yeast lipase null strains caused a reduction in TAG labeling levels
when cells emerged out of the stationary phase (close triangles and squares, Figures 2.2D). In
addition, labeling of polar lipids also showed a clear decrease in cells expressing the algal
CrLIP1 protein (Figure 2.2E). Together, these results strongly suggest a lipolytic function of
CrLIP1.

38

Recombinant CrLIP1 hydrolyzes diacylglycerol but not triacylglycerol in vitro
To examine the lipase activity of CrLIP1 biochemically, the cDNA of CrLIP1 was cloned
into an E. coli expression vector. A hexahistidine-tagged CrLIP1 protein was produced, purified
(Figure 2.3A), and subjected to TAG hydrolysis assays (Figure 2.3B). Purified CrLIP1 ran as a
doublet on SDS-PAGE (Figure 2.3A). Although the underlying reason for this phenomenon
remains unclear, the fact that both bands reacted with anti-His antibody (Figure 2.4) indicated
that these two bands were isoforms of CrLIP1, not co-purified contaminants. Radioactive triolein
supplemented with olive oil was incubated with the purified CrLIP1 protein before TLC analysis
of the reaction products. Neither iodine vapor stain (not shown) nor autoradiography detected
production of free fatty acids by CrLIP1 (Figure 2.3B). Multiple attempts with varying reaction
conditions all failed to detect TAG lipase activity of CrLIP1. I also tried another neutral lipidsteryl esters extracted from yeast cells as the substrate and detected no lipolytic activity either
(Figure 2.5).
I had noted that BLAST searches against non-redundant databases revealed considerable
similarity between CrLIP1 and the human sn-1-specific diacylglycerol lipases DAGLα and
DAGLβ (Bisogno et al., 2003) (identity scores of 22 with DAGLα and 26 with DAGLβ) (see
Figure 2.6). This prompted us to test the activity of CrLIP1 against DAG. Commercial radiolabeled DAG was not available. To choose a DAG substrate capable of differentiating the
reaction products from bacterial fatty acids co-purified with the recombinant CrLIP1 protein, I
9

took advantage of the fact that E. coli does not synthesize oleic acids (18:1∆ ) (Kito et al., 1972),
and thus used diolein [1,2-Di(cis-9-octadecenoyl)glycerol] as the substrate for in vitro lipolysis

39

assays. Reaction mixtures were resolved by TLC, followed by gas-liquid chromatography (GLC)
of FAMEs to identify the fatty acid products. At the position of free fatty acids, there was no
visible staining in the absence of CrLIP1 (Figure 2.7A), demonstrating that diolein by itself
remained intact through the entire reaction. In the presence of CrLIP1, on the other hand, iodine
staining clearly revealed the production of free fatty acids. Indeed, GLC analysis of the free fatty
acids isolated from the TLC plate determined unequivocally the presence of oleic acid when
CrLIP1 was present in the reaction (Figure 2.7C). I thus concluded that CrLIP1 possesses DAG
lipase activity.

Recombinant CrLIP1 degrades membrane lipids of E. coli
Using crude lysates ofE. colifor the initial testing of the CrLIP1-dependent lipase activity,
I noticed a prominent signal of free fatty acids prior to the addition of an exogenous substrate,
suggesting that CrLIP1 may have had already acted on an abundant bacterial lipid species. To
explore this possibility further, I isolated bacterial lipids for TLC analysis, and observed the
presence of substantial amounts of free fatty acids in the CrLIP1-containing extracts (Figure
2.8A).In E. coli, phosphatidylethanolamine (PtdEtn) is the most abundant glycerolipid (Rietveld
et al., 1993) and I suspected that the free fatty acids were at least partly derived from PtdEtn
hydrolysis. If so, one of the PtdEtn hydrolysis intermediates, lyso-PtdEtn, might be detectable in
the bacterial lipid extracts when CrLIP1 was overproduced. Indeed, when the sn-1-specific
Rhizopusarrihizus lipase (Fischer et al., 1973)was used to generate a standard for lyso-PtdEtn,
lyso-PtdEtn increased along with free fatty acids (Figure 2.8B). GLC analysis provided the
quantitative support for both free fatty acids and lyso-PtdEtn increase and PtdEtn reduction in the
CrLIP1-overexpressing sample (Figure 2.8C). These results strongly suggest that CrLIP1 is

40

capable of hydrolyzing glycerophospholipids such as PtdEtn that are the major components of
biological membranes. Intriguingly, in the neutral TLC (Figure 2.8A) on lipids extracted from E.
coli cells, a new compound (indicated by the arrow) was detected upon CrLIP1 over-expression.
I do not yet know the identity and origin of this species. While this phenomenon is likely related
to the biochemical activity of CrLIP1, definitive identification of this bacterial lipid species will
require additional efforts beyond the scope of the current work.

Recombinant CrLIP1 hydrolyzes polar lipids
Many lipases such as the human ATGL (Jenkins et al., 2004) and yeast Tgl4 (Rajakumari
and Daum, 2010) act on a broad spectrum of glycerolipid substrates. The apparent CrLIP1dependent lipolysis of PtdEtn in E. coliprompted us to explore what other membrane lipids might
be substrates for CrLIP1. I focused on abundant Chlamydomonaspolar lipids that each
constituted greater than 10% of the membrane lipids: MGDG (monogalactosyldiacylglycerol),
DGTS (diacylglycerol-N,N,N-trimethylhomoserine), PtdEtn, PtdGro (phosphatidylglycerol) and
DGDG (digalactosyldiacylglycerol). Radio-labeled polar lipids were prepared by growing cells
in [14C]-acetate containing medium since acetate predominantly labels acyl chains rather than the
glycerol backbone (Schneider and Roessler, 1994). These substrates would therefore allow us to
follow the lipase activity by tracking the production of free fatty acids. Lipolysis reaction
products were resolved by TLC and detected by autoradiography. For the five substrates used,
lipolysis was documented by the apparent production of radioactive free fatty acids when CrLIP1
was present (Figures 2.9A and B). While CrLIP1 appeared to be more active on phospholipids
(i.e., PtdEtn and PtdGro) under the assay conditions, it was difficult to quantitatively determine
the substrate preference for these polar lipids, because they were mixtures of molecular species

41

with varying specific activities due to the preparation of the substrates by pulse-labeling of
Chlamydomonas cells. Nonetheless, as shown in Figures 2.7 to 2.9, I was able to demonstrate the
lipase activity of CrLIP1 against DAG and polar lipids.

Recombinant CrLIP1 hydrolyzes both acyl chains of phosphatidylcholine (PtdCho) with a
preference for the sn-1 position
Diacylglycerol and polar lipids harbor two acyl chains at the sn-1 and sn-2 positionsof the
glycerol backbone. Many lipases prefer one position over the other. For example, the lipase from
Rhizopusarrhizus strongly prefers the sn-1 ester bond (Fischer et al., 1973). The human DAGLs
also prefer sn-1 over sn-2 although both activities were detectable (Bisogno et al., 2003). To
explore the position preference of CrLIP1, I chose phosphatidylcholine (PtdCho) as the substrate.
PtdCho is not present in Chlamydomonas, but gave us an opportunity to test a number of
9

commercially available molecular species. Specifically, I used 18:1∆ /16:0 (sn-1/sn-2) and
9

16:0/18:1∆ PtdCho (Figure 2.10). By following the time-dependent liberation of fatty acids from
each substrate, the preference (or the lack of which) for position and/or saturation was
observable. Lipase reactions were sampled at 0, 6, and 16 h, and PtdCho, lyso-PtdCho, and free
fatty acids were extracted from the corresponding spots on TLC plates for GLC analysis. Only
negligible amounts of free fatty acids and lyso-PtdCho were present at 0 h (Figures 2.10C and D).
After 6 h of incubation, both free fatty acids and lyso-PtdCho emerged along with a decrease in
PtdCho (Figures 2.10E and F). For both PtdCho species, fatty acids were mainly, but not
exclusively released from the sn-1 position, regardless of the acyl chain. From 6 to 16 h (Figures
2.10G and H), the amount of PtdCho remained largely unchanged, whereas the level of lyso-

42

PtdCho decreased, suggesting CrLIP1 was also able to hydrolyze lyso-PtdCho. Based on these
resultsI can conclude that CrLIP1 prefers the sn-1 position, but can also hydrolyze the acyl chain
at the sn-2 position. Moreover, the sn-1 oleate and palmitate were mobilized equally well by
CrLIP1, suggesting that both oleate and palmitate are comparable substrates for CrLIP1.

Involvement of CrLIP1 in Chlamydomonas TAG metabolism
Although I have yet to directly observe TAG lipolysis by using a purified recombinant
CrLIP1 protein, it is possible that CrLIP1 is indirectly involved in cellular TAG metabolism.
This is because remodeling of lipids, membrane and storage, is one of the responses to Ndeprivation that leads to TAG accumulation (Moellering and Benning 2010). In addition, DAG is
both a precursor and a product (intermediate) for TAG and membrane lipid metabolism (Miege
and Marechal, 1999). The lipase action of CrLIP1 as shown in Figures 7 to 10 may impact the
flux of DAG and polar lipids, which could eventually affect TAG biosynthesis and breakdown. I
took several approaches to test this hypothesis. First, I examined the relationship between
changes in CrLIP1 transcript and TAG abundance following N-deprivation and N resupply. If
CrLIP1 plays a role in storage lipid metabolism in Chlamydomonas cells, it was expected that
the expression of theCrLIP1gene would also respond to N fluctuation. Chlamydomonas cells
started to accumulate TAG shortly after N removal and continued to increase their TAG content
(Figure 2.11A)(Wang et al., 2009; Moellering and Benning, 2010; Siaut et al., 2011). N resupply
triggered fast TAG mobilization. Consistent with a role in lipolysis, the transcript of CrLIP1
declined steadily when TAG was accumulating (Figure 2.11B). Notably, there was a surge of
CrLIP1 transcript that coincided with the N resupply-triggered TAG mobilization. The obvious

43

inverse correlation between CrLIP1 transcript level and TAG abundance strongly suggests that
CrLIP1 exerts a catabolic function in Chlamydomonas TAG metabolism.
To further link CrLIP1 expression to TAG metabolism, I used an artificial microRNA
(AmiRNA) (Molnar et al., 2009) to reduce the expression of CrLIP1. Sixty independent
transformants were screened for a reduction of CrLIP1 mRNA level, and two lines (1-44 and 156) exhibited 70-80% reduction (Figure 2.12A). These two lines and a vector-only control were
subjected to N-deprivation and TAG quantification. I only observed a slight increase in the TAG
levels for the 1-56 line 48 and 72 hours into N-deprivation (Figure 2.12B). The lack of a
significant change in the TAG pool size in CrLIP1 knockdown lines seemed to be in agreement
with the N-deprivation-dependent decline of CrLIP1 mRNA, that is, the CrLIP1 lipase activity
was likely to be down-regulated when TAG synthesis prevailed in the overall energy flux.
Artificial knockdown of CrLIP1 transcript level thus caused a minimal effect on the total TAG
level. On the other hand, when ammonium was added to N-deprived cultures to trigger TAG
breakdown, knocking down CrLIP1 expression resulted in a delay of TAG mobilization. The
amount of TAG normalized to that of total lipids was significantly higher in the CrLIP1
knockdown lines than that in the empty vector control (Figure 2.12C). This phenomenon held
true when the TAG content was measured as an absolute amount per cell at 24 h after the resupply of N (Figure 2.12D). This observation conforms to a lipolytic role of CrLIP1in vivo, and
is reminiscent of the slower TAG turnover during post-germination growth of the sdp1lipase
mutant of Arabidopsis thaliana (Eastmond, 2006). Intriguingly, the maximal TAG amount in the
sdp1 mutant seeds is also indistinguishable from the wild-type seeds (Eastmond, 2006). I thus
conclude that the Chlamydomonas CrLIP1 plays a critical role in ensuring fast TAG turnover
when cells emerge from N-deprivation.

44

DISCUSSION
CrLIP1 affects TAG content possibly by degrading DAG
From the data presented above, I suggest that CrLIP1 affects Chlamydomonas and yeast
TAG metabolism in different ways. In vitro, the recombinant CrLIP1 protein clearly can
hydrolyze DAG and phospholipids, but not TAG under the standard test conditions. On the other
hand, metabolism of TAG in yeastand in Chlamydomonas is affected by changes in the amount
of CrLIP1. Given that DAG is an immediate precursor for TAG synthesis and an intermediate of
TAG degradation, changes in DAG pool sizes will presumably impact TAG abundance in
different directions. Accordingly, as shown in Figure 2.13, I proposethat CrLIP1 participates in
TAG turnover in Chlamydomonas by facilitating DAG removal during TAG mobilization
(“reaction 1”, Figure 2.13), hence contributing to expeditious conversion of TAG into cell
building blocks and metabolic energy following the re-introduction of N. Repression of CrLIP1
expression by artificial microRNA may transiently increase the DAG pool, which then delays
TAG turnover by product inhibition through feedback control. The overall rate of TAG
hydrolysis is thus reduced. However, possibly due to functional redundancy, or as a result of
metabolism of DAG through other routes such as conversion to phosphatidic acid (Arisz et al.,
2009), TAG that accumulates following N-deprivation eventually will be turned over
quantitatively (Figure 2.12B). In yeast, the tgl3∆ tgl4∆ double mutant has no detectable TAG
lipase activity (“reaction 2”, Figure 2.13)(Kurat et al., 2006). This argues against the notion that
CrLIP1 suppresses the TAG accumulation phenotype solely by expediting DAG removal
following the first TAG hydrolysis step, as the lipolytic route of DAG production is blocked by

45

the deletion of TAG lipases. Rather, Isuggest two hypotheses. First, yeast gradually accumulates
TAG when cells enter the stationary phase (Clausen et al., 1974; Leber et al., 1994). The
introduction of a heterologous DAG lipase CrLIP1 may result in a shortage of DAG supply for
TAG biosynthesis (“reaction 3”, Figure 2.13), hence partially preventing the TAG accumulating
phenotype seen in the lipase knockout strains. Secondly, due to the pleiotropic metabolic as well
as signaling functions of DAG, the steady-state DAG amount has to be tightly controlled (Shen
et al., 2011). The action of CrLIP1 in yeast may cause alterations in TAG metabolism by an
indirect route involving DAG signaling. Nevertheless, I cannot exclude the possibility that
CrLIP1 is indeed a TAG lipase in vivo but requires an unknown co-factor that was absent in our
assay mixtures or a post-translational modification of the protein which could not be performed
in E. coli cells.
In addition to phenotypes of TAG metabolism, overexpressing CrLIP1 in yeast also
affected the rate of polar lipid accumulation in cells emerging from stationary phase. The
observed phospholipid lipase activity (Figures 2.8-2.10) can account for the above phenotype,
although the stationary-phase cells appear to be relatively insensitive to CrLIP1 overexpression
(Figures 2.1D and 2.2B). Alternatively, as DAG can be converted to phosphatidic acid and other
phospholipids including PtdEtn and PtdCho (Weyer et al., 2010), CrLIP1 may simply hydrolyze
and remove DAG from the pool for polar lipid biosynthesis, thus creating a futile cycle
hampering the rate of polar lipid accumulation.

Is there another function of CrLIP1?
In Chlamydomonas, CrLIP1 appears to be actively expressed during vegetative growth
when cells maintain a negligible amount of steady-state TAG (see time 0, Figure 2.11). The

46

qRT-PCR results in Figure 2.11 are consistent with our earlier work, in which global RNAseq
analysis was employed to examine transcriptomic changes duringN-deprivation. The
CrLIP1transcript hits were 2921 ± 687 under N-replete conditions, and were down to 1335 ± 149
following 48 hours of N-deprivation (Miller et al., 2010). Comparatively, the average transcript
hits of all genes were only 570 during vegetative growth.qRT-PCR verified the abundant CrLIP1
mRNA when compared with a constitutively expressed RACK1 gene (Figure 2.14). These data
demonstrate the active expression of CrLIP1 during vegetative growth. Assuming that transcript
abundance correlates with the protein level, these expression data suggest that CrLIP1 may be a
relatively abundant protein during vegetative growth. Given that many Chlamydomonas major
membrane lipids such as MGDG and DGDG are derived from DAG (Miller et al., 2010), and
that CrLIP1 is capable of hydrolyzing DAG and these major Chlamydomonas polar lipids, it
seems possible thatCrLIP1 may be involved in DAG and membrane lipid turnover during mitotic
growth when active membrane synthesis and remodeling take place. In contrast, following Ndeprivation, DAG is synthesized but then channeled into TAG production by DGATs(Boyle et
al., 2012). A DAG lipase activity of CrLIP1 would be at odds with the need for cells to
accumulate TAG, hence the observed decline of CrLIP1 transcript following N-deprivation. To
delineate the potential additional function of CrLIP1 in vegetative cells, the localization of
CrLIP1 may have provided some useful information. Unfortunately, our attempts to express a
CrLIP1-GFP fusion protein have not been successful. Intriguingly, CrLIP1 was detected in a
flagellum proteomic survey, and was hence designated FAP12 (Flagella Associated Protein 12)
(Pazour et al., 2005). Thus far, I have not observed a discernible flagellar phenotype associated
with either CrLIP1 knock-down lines characterized above (data now shown). One possibility is
that the CrLIP1repression in theAmiRNA lines was not sufficient to bring about an apparent

47

flagellar phenotype. On the other hand, given the similarity between CrLIP1 and the human
DAG lipases DAGLα and DAGLβ that are localized in the plasma membrane (Bisogno et al.,
2003), it is possible that part of the flagella-associated CrLIP1 population may have derived from
the plasma membrane as the flagellar membrane is confluent with the plasma
membrane(Rosenbaum and Witman, 2002). Thus, the presence of CrLIP1 in flagella does not
contradict our hypothesis of CrLIP1 involvement in TAG metabolism.
CrLIP1 did not exhibit a TAG lipase activity under our test conditions, and repressing
CrLIP1 expression delayed but did not prevent TAG turnover. Therefore, other TAG lipases in
Chlamydomonas reinhardtii remain to be identified.

48

Figure 2.1. Expression of CrLIP1 rescues a yeast mutant deficient in TAG lipases. (A)
Western blotting for HA-tagged CrLIP1 protein expressed in yeast. Whole cell extracts from
tgl3Δtgl4Δyeast cells bearing either the empty pMK595 vector (vect) or pMK595CrLIP1 were
centrifuged to obtain a membrane-enriched pellet (P). 10 µg of proteins from both the soluble (S)
and pellet fractions were loaded for SDS-PAGE before western blotting (panel A) or Coomassie
brilliant blue staining (panel B). (C) Thin-layer chromatogram of lipids extracted from

49

Figure 2.1. (cont’d)
membrane-enriched fractions containing 10 µg of proteins. Lipids were visualized by exposing
the plate to iodine vapor. (D) Cellular contents of fatty acid methylesters derived from major
glycerolipids. Lipids were extracted from 2-day old stationary-phase yeast cells. Lipid
abbreviations: TAG, triacylglycerol; DAG, diacylglycerol; PtdEtn, phosphatidylethanolamine;
PtdCho, phosphatidylcholesterol; PtdSer, phosphatidylserine; PtdIns, phosphatidylinositol. (E)
Fatty acid compositions of TAG. Fatty acids are designated as chain length:number of double
bonds. For D and E, averages of three replicates are shown with errors bars indicating SD (n=3).
(F) Growth curves of yeast strains in uracil dropout medium supplemented with 0.5% casamino
acids (CAA-ura). Stationary phase cells were diluted to an optical density (at 600 nm) of 0.1 to
start the time course. Representative results from more than three independent experiments are
shown here. Data are presented as averages from three measurements of the same culture. Error
bars indicate standard deviation. G, Distribution of yeast cell diameters in µm (long axis). Cells
were grown in medium without uracil for 48 hr. For each strain, two populations from two single
colonies were analyzed. More than 200 cells were measured for each population. Percentages of
cells within each size range are shown.

50

Figure 2.2. Metabolic labeling and chase studies of TAG and phospholipid synthesis in
yeast strains over-expressing CrLIP1.

14

C acetate was added to stationary phase yeast cultures.

Samples were taken at time points indicated, with TAG (panel A), polar lipids (panel B), and
DAG (panel C) extracted and quantified. The radioactivity was normalized to the cell number.

51

Figure 2.2. (cont’d)
Data are averages ± SD (n=3). (D and E) Chase experiments. Stationary phase yeast cells were
labeled, briefly washed and then diluted to fresh, non-labeled medium. Cells were collected at
four time points, with the radioactivity TAG (panel D) and polar lipids (panel E) quantified, and
normalize to the starting counts. (F) Quantification of cell numbers over the course of chase
experiments. The optical density at 600 nm was measured. For panels D, E and F, data are
averages of two technical replicates with the error bars indicating standard deviations.

52

Figure 2.3. Recombinant CrLIP1 fails to hydrolyze triacylglycerol in vitro. (A) SDS-PAGE
of bacterial whole cell lysates (WCL, 6 µg) and purified CrLIP1 protein. Arrows point to two
potential isoforms of recombinant CrLIP1. The unlabeled lane shows molecular mass markers.
(B) Thin-layer chromatogram of lipids extracted from triacylglycerol lipase assay mixtures with
radioactive triolein and olive oil as the substrate. An autoradiogram of TLC is shown. For the
buffer-only control lane, same volume of CrLIP1 storage buffer was used instead of the purified
CrLIP1 protein. Rhizopus lipase (RL) digested triolein was loaded as the marker for free fatty
acids.

53

Figure 2.4. Western blotting for hexahistidine-tagged CrLIP1 protein purified from E. coli
cells. Two µg purified protein was loaded for immunodetection. Note the doublets of CrLIP1
isoforms.

54

Figure 2.5. Recombinant CrLIP1 fails to degrade steryl esters in vitro. Thin-layer
chromatographic resolution of lipids extracted from lipase assay mixtures with radioactive steryl
esters derived from yeast cells as the substrate. TLC was developed in petroleum ether-diethyl
ether-H2O (80:20:1 in volume). An autoradiograph is shown. The position of free fatty acids was
marked by running oleic acids in the TLC and aligning the film to the TLC plate. Appearance of
radioactive free fatty acids is considered as evidence for esterase activity on steryl esters.

55

Figure 2.6. CrLIP1 shares sequence similarity with two human DAG lipases DAGLα and
DAGLβ. Shown are Clustal W2 results using the lipase domains of each of the three proteins.
The “GXSXG” hydrolase motif is boxed. Part of the sequence for each protein was used as
indicated by the residue numbers on the right side. Asterisk (*), colon (:) and period (.) denote
conserved residues, conservative substitutions and semiconservative substitutions respectively.

56

Figure 2.7. Recombinant CrLIP1 degrades diacylglycerol in vitro. (A) Thin-layer

57

Figure 2.7. (cont’d)
chromatogram of lipids extracted from diacylglycerol lipase assay mixtures with diolein as the
substrate for CrLIP1. Lane 1, control reaction with diolein and CrLIP1 storage buffer; lane 2,
reaction with diolein and purified CrLIP1; lane 3, control reaction with purified CrLIP1without
9

diolein to show fatty acids contained in the protein prep. Diolein and oleic acid (18:1Δ ) were
loaded as markers. The deep staining at the origins might be from ingredients of the reaction
buffer such as dithiothreitol or glycerol. (B-D) Gas-liquid chromatograms of fatty acid methyl
esters (FAMEs) derived from free fatty acids scraped from the TLC plate shown in panel A.
Arrows point to the location or expected location of oleic FAME.

58

Figure 2.8 Recombinant CrLIP1 hydrolyzes phosphatidylethanolamine, the major
membrane lipid of E. coli. (A) Thin-layer chromatograms of total lipids extracted from E. coli
cells. Note the CrLIP1-dependent production of a significant increase of free fatty acids in the E.
coli lysate. The arrow points to an unidentified lipid. (B) TLC of in vitro PtdEtn hydrolysis

59

Figure 2.8 (cont’d)
reactions. CrLIP1 causes the reduction of the PtdEtn substrrate, and produces free fatty acids and
lyso-PtdEtn. Markers include PtdEtn and oleic acid from a commercial source, and lyso-PtdEtn
generated by the Rhizopus lipase (RL). (C) Quantitative analysis of CrLIP1-dependent PtdEtn
hydrolysis in E. coli. Data are presented as the ratio of fatty acids in each lipid to fatty acids in
the total lipid extract. Averages and standard deviations are shown (n=3).

60

Figure 2.9. Recombinant CrLIP1 degrades major polar lipids of Chlamydomonas in vitro. A, a
thin-layer chromatogram of CrLIP1-catalyzed polar lipid hydrolysis. Each polar lipid was
isolated from Chlamydomonas cells metabolically labeled with

14

C acetate. Positions of the

polar lipid substrates are labeled on the left with arrows near the corresponding substrates. Lipid
abbreviations:

MGDG,

monogalactosyldiacylglycerol;

DGTS,

diacylglycerol-N,N,N-

trimethylhomoserine; PtdEtn, phosphatidylethanolamine; PtdGro, phosphatidylglycerol; DGDG,
digalactosyldiacylglycerol. Chloroform-methanol-acetic acid-H2O (75:13:9:3 in volume) was
used to develop the TLC. B, for a better resolution between MGDG and free fatty acids, MGDG
assay mixtures were loaded to another TLC developed in petroleum ether-diethyl ether-H2O
(80:20:1 in volume). MGDG and DGDG digested with Rhizopus lipase (RL) were loaded to
provide markers for free fatty acids. X-ray films exposed to TLC plates are shown.

61

Figure 2.10. Recombinant CrLIP1 cleaves both acyl groups of phosphatidylcholine with a
preference for sn-1 position. (A and B) Schematic representations of the two
phosphatidylcholine (PtdCho) substrate. Circled letter “P” denotes phosphate. (C-H) Quantitative
analyses of PtdCho lipolysis products. Each lipid species was scraped off the TLC plate,
transesterified to FAMEs, and quantified by GLC. Panels C, E, and G represent products derived
9

9

from 18:1Δ /16:0 PtdCho, whereas panels D, F, and H are from the 16:0/18:1Δ PtdCho
reactions. Averages and standard deviations are shown (n=3).

62

Figure 2.11. Inverse correlation between CrLIP1 transcription and TAG abundance in
Chlamydomonas. (A) TAG abundance through N deprivation and resupply. Data are presented
as the molar ratio of fatty acids contained in TAG to total fatty acids in the lipid extract. Logphase cells were N-deprived for 72 hr. After N re-supply, cells were grown for another 36 hr.
Portions of cultures at each time point were harvested for lipid and mRNA extraction. (B)
CrLIP1 mRNA abundance from the same cultures examined in panel A. The mRNA was
quantified by qRT-qPCR. CrLIP1 transcripts were normalized to the constitutively expressed
RACK1, and the CrLIP1/RACK1 ratio at the time of N-deprivation initiation was set as 1. In
both A and B, data are shown as averages ± SD (n=3).

63

Figure 2.12. Suppression of CrLIP1 expression delays TAG turnover in Chlamydomonas.
(A) Expression of CrLIP1 in two AmiRNA lines compared to an empty-vector control. mRNA
quantification was expressed as CrLIP1/RACK1 ratio, and then normalized to the empty-vector
control. (B) TAG accumulation at 48 and 72 hr of nitrogen deprivation. (C) TAG degradation

64

Figure 2.12. (cont’d)
after nitrogen re-supply described in Figure 7. (D) TAG amount normalized by cell number 24 h
after N re-supply to N deprived cells as described for panel C. Data are shown as averages ± SD
(A, n=3; B, n=4; D, n=3).*per t-test statistically different at p< 0.05; **per t-test statistically
different at p< 0.01.

65

Figure 2.13. A model for how CrLIP1 might affect TAG metabolism. The DAG lipase
activity of CrLIP1 may affect TAG biogenesis and hydrolysis in yeast and Chlamydomonas,
respectively. Enzymes are italicized. Note that the model stresses components directly related to
CrLIP1 over-expression or knock-down. Not all enzymes or metabolites are shown.
Abbreviations for enzymes and lipids: ACP, acyl carrier protein; DAG, diacylglycerol; DGAT,
diacylglycerol acyltransferase; DGDG, digalactosyldiacylglycerol; DGK, diacylglycerol kinase;
DGTS, diacylglycerol-N,N,N-trimethylhomoserine; FA, fatty acid; G-3-P, glycerol-3-phosphate;
MGDG, monogalactosyldiacylglycerol; PLC, phospholipase C; PLD, phospholipase D; PtdOH,
phosphatidic acid; TAG, triacylglycerol; TAGL, triacylglycerol lipase.

66

Figure 2.14. RT-PCR quantitative comparison of RACK1, CrLIP1 and DGTT1 transcripts
in vegetatively growing cells. No template control was performed for each gene. Calculated
sizes for PCR products are: 139 bp for RACK1, 138 bp for CrLIP1 and 98 bp for DGTT1.

67

Table 2.1. Yeast strains used in this study.
Strain

Relevant genotype

Source or Reference

yMK839

MATa leu2-3 trp1 ura3-52

Kuo et al., 1998

yXL001

MATa leu2-3 trp1 ura3-52tgl4Δ::K.l. TRP1

This study

yXL005

MATa leu2-3 trp1 ura3-52tgl4Δ::K.l. TRP1tgl3Δ::KanMX This study

yXL023

MATa leu2-3 trp1 ura3-52/pMK595 [2µ URA3]

This study

yXL026

MATa leu2-3 trp1 ura3-52/ pMK595CrLIP1 [2µ URA3]

This study

yXL077

MATa

leu2-3

trp1

ura3-52tgl4Δ::K.l. This study

TRP1tgl3Δ::KanMX/ pMK595 [2µ URA3]
yXL080

MATa

leu2-3

trp1

ura3-52tgl4Δ::K.l. This study

TRP1tgl3Δ::KanMX/ pMK595CrLIP1 [2µ URA3]

68

Table 2.2. Plasmid constructs used in this study

Plasmid

Main features

Source

or

Reference
pMK595

2µ URA3/PADH1-3xHA

Luo et al., 2010

pMK595CrLIP1

2µURA3/PADH1-3xHA-CrLIP1

This study

pMK1006

ori amp pT7-His6

pMK1006CrLIP1

ori amp pT7-His6-CrLIP1

pChlamiRNA3int

ori amp aphVIII pPSAD

r

unpublished

r

This study

r

Molnar et al., 2009

pChlamiRNA3intCrLIP1 ori ampraphVIII pPSAD-CrLIP1 targeting This study
oligo

69

Table 2.3. Oligonucleotide primers used in this study. All primer sequences are written in 5’
to 3’ direction. N represents a random nucleotide.
Name

Sequence

qRT-F

GAGCCTGAGTACGCCAAGTC

qRT-R

AATATCGTGCTTGGGAGTGG

RACK1-F

GTCATCCACTGCCTGTGCTT

RACK1-R

CCTTCTTGCTGGTGATGTTG

DGTT1-F

AGTCCCACGGAGGCTGAG

DGTT1-R

ACACCCTTGCCGTACTTGTC

4DF

ATGAGCAGCAAAATATCAGATCTTACATCTACACAAAATAAGTGAA
GCTTGATATCGAAT

4DR

TTATTGAGTAAAACTGGTGGTGCTTTTGTGCTGCGTTGAAGATACGA
CTCACTATAGGGC

KMF

TGAAGCTTGATATCGAATCGGATCCCCGGGTTAATTAA

KMR

TACGACTCACTATAGGGCGAATTCGAGCTCGTTTAAAC

3DF

ATGAAGGAAACGGCGCAGGAATACAAGGTGTCTGCTGTAATATGAA
GCTTGATATCGAAT

3DR

CTACCTACTCCGTCTTGCTCTTATTATGTCGTCTAATTTAAATACGAC
TCACTATAGGGC

595F

TATCCATATGACGTTCCAGATTACGCTGCTCAGTGCGGCCGCATGGG
TTGCGGAGCGAGC

595R

GAATTTCGACGGTATCGGGGGGATCCACTAGTTCTAGCTAGACTAAC
CGACCTCCTGCTT

LICF

AAAGAGGCGCGTNATGGGTTGCGGAGCGAGC

LICR

CGGAAGGCGCGTTACGCTGCCGGCGGCGG

P1

CTAGTAAGAGCCTCCTCATTGGGATATCTCGCTGATCGGCAC

P2

CATGGGGGTGGTGGTGATCAGCGCTATATCTTAATGAGGAGGCTCTT
G

M1

CTAGCAAGAGCCTCCTCATTAAGATATAGCGCTGATCACCACCACC

M2

CCCATGGTGCCGATCAGCGAGATATCCCAATGAGGAGGCTCTTA

70

Table 2.4. Putative lipase-encoding genes selected for expression in yeast
Name

Protein ID in JGI Reason for selection Result
Chlamydomonas 4.0 and reference
database

CrLIP1

184308/519543

Down-regulation under 519543
nitrogen

was

the

right

model

(N) according to sequencing of the

deprivation (Miller et coding sequence amplified by PCR.
al., 2010)

This gene complemented the lipase
null mutant of yeast

CrLIP2

191373

Up-regulation under N Not expressed (data not shown)
deprivation (Miller et
al., 2010)

CrLIP3

160378

Found in lipid droplet No major effect on growth and TAG
proteomics (Moellering content
and Benning, 2010)

CrLIP4

319684

Down-regulation under Not cloned (data not shown)
N deprivation (Miller et
al., 2010)

CrLIP5

141065

Down-regulation under No major effect on growth and TAG
N deprivation (Miller et content
al., 2010)

CrLIP6

148864

Up-regulation under N No major effect on growth and TAG
deprivation (Miller et content
al., 2010)

CrLIP7

157360

Up-regulation under N Not expressed
deprivation (Miller et
al., 2010)

CrLIP8

344422

Up-regulation under N No major effect on growth and TAG
deprivation (Miller et content
al., 2010)

71

Chapter 3
A galactoglycerolipid lipase is required for triacylglycerol accumulation following nitrogen
deprivation in Chlamydomonas reinhardtii

1

1

This project was done in collaboration with several colleaguesand is part of a paper in press at
The Plant Cell inXiaobo Li, Eric R. Moellering, Bensheng Liu, Cassandra Johnny, Marie
Fedewa, Barbara B. Sears, Min-Hao Kuo, Christoph Benning. (2012) A galactoglycerolipid
lipase is required for triacylglycerol accumulation and survival following nitrogen deprivation in
Chlamydomonas reinhardtii. Eric R. Moellering initiated the mutant screen. Marie Fedewa
participated in the screen. Bensheng Liu contributed Figure 3.3B. Barbara. B. Sears performed
crossing between the mutant and another strain, providing progeny that were analyzed by me for
lipid content (Figure 3.1B).

72

ABSTRACT
Following nitrogen (N) deprivation microalgae accumulate triacylglycerols. To gain
mechanistic insights into this phenomenon, I identified mutants with reduced TAG content
following N deprivation in the model alga Chlamydomonas reinhardtii. In one of the mutants the
disruption of a galactoglycerolipid lipase-encoding gene, tentatively designated Plastid
Galactoglycerolipid Degradation 1 (PGD1), was responsible for the primary phenotype: reduced
TAG content, altered TAG composition, and reduced galactoglycerolipid turnover. The
recombinant PGD1 protein, which was purified from E. coli extracts, hydrolyzed
monogalactosyldiacylglycerol into its lyso-lipid derivative. In vivo pulse chase labeling
identified galactoglycerolipid pools as a major source of fatty acids esterified in triacylglycerols
following N deprivation. Moreover the fatty acid flux from plastid lipids to triacylglycerol was
decreased in the pgd1 mutant. Apparently, de novo synthesized fatty acids in Chlamydomonas
are, at least partially, first incorporated into plastid lipids before they enter triacylglycerol
synthesis.

73

INTRODUCTION
Triacylglycerol (TAG) is a universal storage lipid in plants, algae, fungi, and animals.
TAG is composed of a glycerol backbone to which three fatty acyl chains are esterified. By
transesterification with methanol TAG can be converted into fatty acid methyl esters (FAMEs)
commonly referred to as biodiesel (Durrett et al. 2008). Microalgae have been considered as
sustainable feedstock for the production of biofuels because they accumulate substantial amounts
of TAG following nutrient deprivation. Theoretical calculations suggest that microalgae can
surpass crop plants in their TAG yield per land area used (Weyer et al. 2010). Despite the recent
interest in microalgae, this phylogenetically diverse group of photosynthetic organisms is not
well understood at the molecular and biochemical levels, and the mechanistic basis of algal lipid
metabolism and of TAG accumulation still needs to be explored in detail. Much of the current
molecular understanding of photosynthetic lipid biosynthesis is based on work with Arabidopsis
thaliana and other land plant models, providing paradigms that may not be directly transferable
given their evolutionary divergence from microalgae. Indeed, our current information on lipid
metabolism in the green algal model Chlamydomonas reinhardtii, which is mostly based on
genome annotation (Riekhof et al. 2005) or early labeling and lipid profiling experiments
(Giroud et al. 1988, Giroud and Eichenberger 1989), suggests that lipid metabolism in this
organism is distinct in crucial aspects from that of land plants. Most strikingly, Chlamydomonas
lacks phosphatidylcholine (PtdCho), but instead contains the betaine lipid diacylglycerol-N,N,N,trimethylhomoserine (DGTS).
Seed plants typically have two assembly pathways for glycerolipids (Roughan and Slack
1982). Fatty acids are synthesized de novo in the plastid while attached to acyl carrier proteins
(ACPs) (Ohlrogge et al. 1979). Acyltransferases at the inner chloroplast envelope membrane

74

transfer acyl groups from acyl-ACPs to glycerol 3-phosphate leading to the formation of
phosphatidic acid (PtdOH), the precursor of glycerolipids of the thylakoid membrane.
Alternatively, fatty acids are exported from the plastid for assembly of extraplastidic
glycerolipids including TAGs at the endoplasmic reticulum (ER). Because the acyltransferases
associated with the inner plastid envelope membrane and the ER have different acyl group
preferences, glycerolipids assembled by the two pathways can be distinguished based on their
acyl group composition (Heinz and Roughan 1983). In Chlamydomonas, the analysis of the acyl
groups in the glyceryl backbone of the galactoglycerolipids, which are the predominant lipids in
the thylakoid membranes, suggests that their assembly is completely dependent on the plastid
pathway (Giroud et al. 1988). In contrast, in seed plants such as Arabidopsis the
galactoglycerolipid molecular species are nearly equally derived from the ER and the plastid
assembly pathway (Browse et al. 1986), thus requiring an elaborate system of lipid transfer
between the ER and the plastid envelopes (Benning 2009).
In particular, the lack of phosphatidylcholine (PtdCho) in Chlamydomonas is expected to
affect other aspects of glycerolipid metabolism. For example, isotope labeling of cytosolic lipids
in pea leaves indicated that most of the acyl groups synthesized de novo in the plastid are first
incorporated into PtdCho instead of PtdOH (Bates et al. 2007). Thus, it was proposed that acyl
editing of PtdCho is an important aspect of fatty acid export from the plastid, cycling acyl groups
through PtdCho before they enter the cytosolic acyl-CoA pool, which ultimately provides acyl
groups for glycerolipid assembly at the ER. The lack of PtdCho in Chlamydomonas raises
several questions, particularly whether an alternative mechanism of acyl editing, possibly
involving DGTS or another lipid, or a mechanism completely independent of acyl editing exists,
which is involved in the export of fatty acids from the plastid. These apparent differences in the

75

cellular organization of basic membrane glycerolipid metabolism in Chlamydomonas may also
extend to differences in the formation of TAGs and their storage in lipid droplets (Liu and
Benning 2012). Typically, lipid droplets are formed at the ER in all eukaryotic cells. However,
recent reports on TAG a starch-less mutant of Chlamydomonas with increased accumulation in
Chlamydomonas of oil suggest that TAG-containing lipid droplets are present in plastids(Fan et
al. 2011, Goodson et al. 2011), raising the possibility that TAG is either directly assembled in
plastids, or imported into them.
Aside from the basic mechanisms of glycerolipid assembly in Chlamydomonas, the
details of the regulation of TAG synthesis are unclear as well. Like other microalgae,
Chlamydomonas produces lipid droplets filled with TAGs following nutrient deprivation
(Moellering and Benning 2010, Wang et al. 2009), conditions that involve genome-wide
transcriptional changes (Castruita et al. 2011, Miller et al. 2010). Intriguingly, among the genes
up-regulated or down-regulated by N deprivation was a large number of genes annotated to
encode lipases (Miller et al. 2010). One of the down-regulated genes was recently shown to
encode a lipase involved in TAG turnover in Chlamydomonas (Chapter 2). As a complement to
transcript profiling in revealing genes involved in TAG metabolism or its regulation, Eric
Moellering and I developed a genetic screen for mutants with abnormal TAG levels following N
deprivation. Here I describe the identification of a low-TAG mutant with a lesion in a
galactoglycerolipid lipase-encoding gene. This gene was among the up-regulated lipase-encoding
genes following N deprivation(Miller et al. 2010), consistent with a role for acyl editing or
turnover of galactoglycerolipids during TAG formation in Chlamydomonas.

76

MATERIALS AND METHODS
Strains and growth conditions
+

The cell wall-less dw15-1 (cw15, nit1, mt ) strain of Chlamydomonas reinhardtii was
obtained from A. Grossman and is referred to as the wild-type (with regard to PGD1) parental
strain throughout. This strain was crossed to CC-198 (er-u-37, str-u-2-60, mt-; Chlamydomonas
Resource Center; http://www.chlamycollection.org) for genetic analysis. Cells were grown in
Tris-acetate-phosphate (TAP) medium (20 mM Tris, 0.1 g/L MgSO4 7H2O (0.4 mM), 0.05 g/L
CaCl2 2H2O (0.34 mM), 10 mL/L glacial acetic acid, 10 mM NH4Cl, 1 mM phosphate and trace
elements (Harris 1989)) under continuous light (70-80 μmol m

-2 -1

s ) at 22°C or ambient room

temperature (~22°C) for solid media, which contained 1.5 % agar. Ammonium chloride was
omitted from N-depleted (TAP-N) medium. To induce TAG biosynthesis, cells were collected by
centrifugation (3000 x g, 4°C, 3 min), washed twice with TAP-N and finally resuspended in
TAP-N of the same volume. For spotting on TAP-N agar, approximately 106 cells from a logphase culture were concentrated in 5 μL.

Primary mutant screen
Plasmid disruption was used to generate mutants of the wild-type parental strain dw15-1.
Transformation using glass beads was performed as previously described (Kindle 1990) using
the pHyg3 plasmid conferring resistance to Hygromycin B (Berthold et al. 2002). The plasmid
was linearized with NdeI (all the restriction endonucleases were purchased from New England
Biolabs, http://www.neb.com). After 8 hrs of recovery, cells were spread onto agar-solidified

77

TAP medium containing 10 μg/mL Hygromycin B. Colonies were picked into 96-well plates
with 1.1 mL TAP medium and grown for three days. For N deprivation and induction of TAG
biosynthesis, small culture droplets (~3 μL) were transferred with a 48-pin replicator to inoculate
a new 96-well plate containing TAP medium containing 0.5 mM ammonium chloride and grown
for 6-7 days under continuous light (70-80 μmol m

-2 -1

s ) at ambient room temperature (~22°C).

For normalization within a 96-well plate, chlorophyll fluorescence was used. For this purpose
100 μL of N-deprived cells were transferred to a black 96-well plate (Black Flat Bottom
Polystyrene NBS™ Microplate 3991, Corning, http://www.corning.com) and read at 455 nm
excitation with an emission filter cut off >685 nm using a FLUOstar Optima 96-well plate reader
(BMG Labtech, http://www.bmglabtech.com). In the same plate to visualize neutral lipids, 100
μL Nile-Red (Sigma-Aldrich; http://www.sigmaaldrich.com) stock solution (5 μg/mL in 10%
methanol containing 0.04 % Triton X-100) was added. The wavelength settings for Nile-Red
fluorescence were 455 nm for excitation and 550-560 nm for emission. A background reading
for this filter set was obtained prior to the addition of Nile-Red (cells only) and subtracted. The
neutral lipid specific signal was calculated as [(Nile-Red fluorescence – background
fluorescence)/chlorophyll fluorescence]. To identify outliers in individual 96 plate sets, the
Median Absolute Deviation (MAD) was determined as [1.482 X median X |individual valuemedian|] according to (Rousseeuw and Croux 1993) and the z-score was calculated as
[(individual value-median)/MAD] as previously described for another mutant screen (Lu et al.
2008). The threshold for the z-score was set at +/- 3.

Genetic analysis

78

In preparation for crossing, the pgd1 mutant in the dw15-1 background and CC-198 were
separately grown for five days on TAP with 4 g/L yeast extract, transferred with a sterile loop to
TAP agar with 10 % the normal concentration of N for two days, and then suspended at high
density in test tubes with sterile water and placed on a shaker overnight. On the following day,
the two cell types were combined using 0.5 mL aliquots, removed after 2-3 h of mating and
plated onto TAP medium completely lacking N and solidified with 4% agar. After one day in the
light, the zygote plates were moved to the dark. Five days after mating, zygospores were sampled
as described by Harris (Harris 1989). On the day following transfer of the zygospores to normal
TAP solidified with 2% agar, meiotic progeny were identified under a dissecting microscope and
separated with a glass needle. After 7 days, the colonies were sufficiently large to transfer to
non-selective media for subsequent replica-plating.

DNA and RNA techniques
Genomic DNA of Chlamydomonas was prepared according to (Newman et al. 1990). For
Southern blotting, genomic DNA was digested with BamHI and resolved by agarose gel
electrophoresis (10 μg DNA per lane). DNA was transferred to a nylon membrane (Amersham
Hybond N+, GE Healthcare, http://www.gelifesciences.com) and fixed under ultraviolet light.
Digoxigenin (DIG) labeling of the probe, DNA transfer, and signal detection were performed
using a kit from Roche (http://www.roche.com) following the manufacturer’s instructions. The
probe was generated through PCR amplification of a 234 bp region within the hygromycin B
resistance cassette with primers SF and SR (all primer sequences can be found in Supplemental
Table 1).

79

For genotyping and “SiteFinding” PCR (Tan, G. H. et al. 2005), Taq polymerase
(Invitrogen, http://www.invitrogen.com) was used. For genotyping, the PCR conditions were
according to the protocol provided by Invitrogen with primers F1, S2-1, and R. SiteFinding PCR
was conducted according to (Tan, G. H. et al. 2005) with minor modifications and with primers
optimized for the pHyg3 plasmid. The primers used for finding the insertion in PGD1 were:
SiteFinder6 in combination with S1-1 and S1-2, SiteFinder8 in combination with S2-1 and S2-2.
In addition, nested primers SFP1 and SFP2 were used for both combinations.
RNA was isolated using the RNeasy Plant Mini Kit (Qiagen, http://www.qiagen.com)
according to the manufacturer’s instructions. To obtain cDNA as the template for RT-PCR, RNA
was subjected to reverse transcription with Superscript II reverse transcriptase (Invitrogen). For
real-time PCR, the commonly used reference gene RACK1 was employed for normalization
using previously reported primers (Chang et al. 2005). Primers used for PGD1 were qF and qR.
Data were analyzed with the 2(- ΔΔC(T)) method (Livak and Schmittgen 2001).
For expression of PGD1 in E. coli, cDNA was originally amplified with primers CF1 and
CR1 using the Failsafe PCR Kit from Epicentre (http://www.epibio.com). The PCR product was
then integrated into the NotI-linearized yeast vector pMK595 (Luo et al. 2010) by homologous
recombination in Saccharomyces cerevisiae(Ma et al. 1987). The fusion plasmid was recovered
by transforming E. coli with yeast DNA extract and designated pXL1238. This plasmid was then
used as a template for PCR using primers CF2 and CR2 and Phusion polymerase (New England
Biolabs) to generate a fragment with BamHI and SalI sites. The PCR product was ligated into
pCR-Blunt (Invitrogen) and cut out with BamHI and SalI. This fragment was ligated into
pLW01-DsRed (Lu and Benning 2009) to generate plasmid pXL1256. pXL1256 was sequenced
and mutations were found. The mutated region was removed by restriction digestion and the

80

remaining backbone was ligated with digested RT-PCR product of that region to obtain plasmid
pXL1262. Another PCR was performed to amplify PGD1 cDNA from pXL1262 using primers
CF3 and CR3. PCR product and an expression vector pMK1006 (provided by M.-H. Kuo) were
combined using a ligation independent cloning procedure (Aslanidis and de Jong 1990) and
sequenced for confirmation. In this plasmid, PGD1 expression was under the control of a T7
promoter and the resulting fusion protein was N-terminally tagged with poly-histidine.

Mutant complementation
A co-transformation protocol was used to introduce wild-type sequences into the pgd1
mutant in the dw15-1 (nit-) background. Plasmid pMN24 (Fernandez et al. 1989) containing the
Chlamydomonas nitrate reductase gene NIT1 as selection marker was digested with BamHI and
used for glass bead transformation of the pgd1 mutant. A bacterial artificial chromosome (BAC)
5E6 containing wild-type genomic DNA was obtained from the Clemson University Genomics
Institute (http://www.genome.clemson.edu), and was digested with KpnI and AseI to excise a 9.5
kb fragment containing the PGD1 genomic DNA. In each transformation, 0.25 μg linearized
pMN24 and 0.3 μg gel purified BAC fragment were used. TAP plates containing 0.5 mM nitrate
instead of 10 mM ammonium were used for selection. The nitrate served initially as the nitrogen
source and the low concentration led eventually to conditions of N deprivation and chlorosis of
the pgd1 mutant but not the complemented lines nor the wild-type parental control. After
transformation of pgd1 with pMN24, colonies from non-complemented lines formed and
bleached within approximately three weeks when grown under continuous light (70-80 μmol m
-1

-2

s ) at ambient room temperature (~22°C). Complemented lines forming green colonies were re-

81

streaked and maintained on agar solidified TAP medium with 10 mM nitrate as the sole N source
to avoid growth of potentially contaminating non-transformed cells.

Lipid analysis and pulse-chase labeling
Lipid extraction, thin-layer chromatography (TLC) of neutral lipids, transesterification
and gas-liquid chromatography were done according to Moellering and Benning (2010). Briefly,
lipids were extracted from cell pellets with methanol, chloroform, 88% formic acid (2:1:0.1 by
volume). To the extract 0.5 volume of 1 M KCl, 0.2 M H3PO4 was added, mixed and phases
were separated by low speed centrifugation. For TAG quantification, lipids were resolved by
TLC on Silica G60 plates (EMD chemicals, #5721-7, http://www.emdchemicals.com) developed
in petroleum ether-diethyl ether-acetic acid (80:20:1 by volume). Polar lipids were separated on
the same plate using chloroform-methanol-acetic acid-H2O (75:13:9:3 by volume) as solvent. To
analyze lyso-glycolipids during for the PGD1 assay, acetone-toluene-H2O (91:30:7.5 by volume)
was used, instead. Brief exposure to iodine vapor was employed for visualization of lipids.
Transesterification of each lipid and separation of fatty acid methyl esters by GLC were as
previously described (Rossak et al. 1997). Transesterification was conducted on pellets with a
known number of cells to determine the cellular total fatty acid content. Staining with α-naphtol
(Benning et al. 1995) was used for the PGD1 assay to detect galactoglycerolipids.
For pulse-chase labeling experiments, cells were grown to log phase in TAP medium and
either used directly (Figure 3.8), or transferred to TAP-N medium and grown for 12 h to induce
8

N deprivation. Cells were harvested and resuspended at a concentration of 3-8 X 10 per mL
either in modified TAP (Figure 3.8) or TAP-N medium (Figure 3.9 and Figure 3.10) containing 6

82

mM acetate (normal TAP contains 17.5 mM). To these cultures [14C-U]-acetate (specific activity
45-60 mCi/mmol; Perkin Elmer, http://www.perkinelmer.com) was added to provide 1 μCi/mL.
In a typical experiment after 1-4 h of incubation in the light 20-40% of the labeled acetate was
incorporated as determined by liquid scintillation counting. At the end of the pulse labeling
phase, cells were centrifuged and washed to remove the labeled acetate, and cells were
resuspended in TAP-N containing the normal amount of acetate to initiate the chase phase. Lipid
extracts were prepared as described above, split in half, and analyzed for polar lipids DGTS,
PtdEtn, MGDG, DGDG, PtdGro, and a mixture of SQDG and PtdIns, which could not be
individually analyzed due to their low total amount in this experiment. Material at the origin of
the TLC was also analyzed and included. The other half of the sample was analyzed for nonpolar lipids DAG and TAG. Lipids were isolated from the TLC plates and incorporation of label
into each lipid was quantified by scintillation counting. These lipid fractions were summed up
and percentages for each lipid fraction were calculated.

Recombinant protein production and PGD1 assay
BL21 (codon+) E. coli strains harboring the empty pMK1006 vector or the pMK1006PGD1 construct were grown to log phase at 37°C. Isopropyl-β-D-thiogalactopyranoside was
added to the final concentration of 0.5 mM to induce protein production. Cells were harvested
after 3 h of induction. To extract proteins, cells were resuspended in lysis buffer (20 mM TrisHCl, pH 7.9, 10% glycerol, 150 mM NaCl, 1 mM dithiothreitol). The mixture was then frozen in
liquid nitrogen and thawed at room temperature for three cycles and sonicated to lyse cells.
Lysates were centrifuged at 21,000xg for 15 min to obtain inclusion bodies. Inclusion bodies
were washed with 5 mL/g wash buffer (4 M urea, 0.5 M NaCl, 1 mM EDTA, 1 mg/ml sodium

83

deoxycholate, 50 mM Tris-Cl pH 8.0) twice and denatured with solubilization buffer (8 M urea,
50 mM Tris-Cl pH 8.0, 10 mM dithiothreitol) by incubation at 50°C for 20 min. The supernatant
was collected after centrifugation at 21, 000Xg for 30 min and subjected to Ni-NTA affinity
purification as described before(Lu and Benning 2009). His-tagged PGD1 protein was eluted
with solubilization buffer containing 200 mM imidazole. Aliquots of purified proteins were
diluted in 15 different buffers of the QuickFold Kit (AthenaES http://www.athenaes.com/),
assayed for lipase activity, and 40 X dilution into protein refolding buffer (50 mM Tris-Cl pH
8.5, 9.6 mM NaCl, 0.4 mM KCl, 1 mM EDTA, 0.5 M arginine, 0.75 M Guanidine-HCl, 0.05%
polyethylene glycol 3350, 1 mM dithiothreitol) was found to be optimal for PGD1. After 1 h
incubation at 4 °C, proteins were aliquoted and kept frozen at -80 °C. Protein concentration was
determined with Bio-Rad Protein Assay Dye Reagent Concentrate (http://www.bio-rad.com)
according to the manufacturer’s instructions.
To prepare lipid substrates from Chlamydomonas or E. coli cells, lipids were extracted
from 48 h N-deprived Chlamydomonas cells or IPTG induced E. coli cells expressing cucumber
MGDG synthase (Shimojima et al. 1997) and resolved by polar TLC. Corresponding bands were
isolated and lipids were recovered from silica gel by extraction with chloroform-methanol (1:1
by volume). For each PGD1 reaction, 75 nmol lipid substrates extracted from Chlamydomonas,
or E. coli cells expressing the cucumber MGDG synthase were used. The organic solvent was
removed under an N2 stream and the lipids were resuspended in 350 μL 0.1 M phosphate saline
buffer (PBS; pH 7.4) with 4.28 mM Triton X-100 and dispersed by sonication (Sonicator 3000
with microprobe, Misonix, http://www.misonix.com) for 6 X 10 s (power setting 1.5) on ice.
Then 100 μL additional PBS was added. Per assay 10 μg refolded PGD1 protein (quantified as
stated above) in 50 μL protein refolding buffer were added. As a negative control, 50 μL protein

84

refolding buffer only were added. The PGD1 refolding buffer inhibited Rhizopus lipase (SigmaAldrich). Therefore, 10 μg lipase dissolved in 50 μL PBS instead of protein refolding buffer was
used unless otherwise noticed. Dithiothreitol was added to a final concentration of 1 mM from a
freshly prepared stock solution. The mixture was sonicated again for 5 s and incubated at
ambient temperature (~22°C). After 6 h incubation (3 h for Rhizopus lipase to prevent potential
loss of lyso-lipid standards), reactions were stopped by the addition of 1 mL solvent used for
lipid extraction, and lipid extracts were analyzed by TLC described above. For gas
chromatograms on free fatty acids and lyso-MGDG generated by PGD1, 9 h was used to obtain
more prominent signals. To measure the velocity of MGDG hydrolysis, reactions were quenched
after 3 h of incubation.

Positional analysis of TAG
Positional analysis of TAG was performed with Rhizopus lipase using a similar
procedure as described above. Briefly, lipids were extracted from 48 h N-deprived
Chlamydomonas cells and resolved by neutral TLC. TAG was extracted from silica gel with
chloroform-methanol (1:1 by volume) as above. Approximately 10 μg was dried under an N2
stream and resuspended in PBS containing Triton X-100 as above. Rhizopus lipase was
dissolved in PBS and 20 μg was added to the emulsified TAG preparation. Lipids were extracted
from the reaction mixtures and resolved by neutral lipid TLC (described above). Free fatty acids
and monoacylglycerol spots were scraped for transesterification as above. Background levels of
fatty acids carried over with Rhizopus lipase were estimated in a control reaction without
substrate lipid supplied, and subtracted from the free fatty acids data obtained with substrate.

85

RESULTS
Isolation of TAG mutants
To generate Chlamydomonas mutants with altered TAG content, random, insertional
gene-disruption was conducted by introducing a linearized pHyg3 plasmid (Berthold et al. 2002)
into the cell wall-less Chlamydomonas strain dw15-1.1, which is referred to as the parental wildtype strain (because it is wild-type (with regard to its lipid content and synthesis). Hygromycin
B-resistant transgenic lines were picked into a 96-well plate and induced for TAG accumulation
by transfer to low-N medium. During the primary screen, Nile-Red fluorescence-staining of
neutral lipids (Chen et al. 2011, Kimura et al. 2004) was used to monitor neutral lipids in a highthroughput mode using a 96-well plate reader. Putative mutants differing in Nile-Red
fluorescence from the wild-type parental strain based on statistical criteria as defined under
Material and Methods were reanalyzed by extracting lipids, separating them by thin-layer
chromatography (TLC), followed by quantification of TAG-derived fatty acid methylesters by
gas liquid chromatography (GLC). Of 34,000 independent transgenic lines generated, 80 were
initially found to exhibit an altered Nile-Red fluorescence intensity, of which six mutants with
robust and reproducible changes in TAG levels were eventually isolated. The focus here is on the
characterization of one of the low-TAG mutants, initially designated line E12. After in-depth
analysis it was renamed plastid galactoglycerolipid degradation 1 (pgd1), the designation used
from here on.

The pgd1 mutant has reduced TAG and becomes chlorotic following N deprivation

86

Over the course of three days following N deprivation the pgd1 mutant showed an
approximately 50% reduction in the ratio of fatty acids in TAG over total fatty acids in the lipid
extract, a parameter that allows a robust comparison of relative TAG content between different
lines, in this case pgd1 and the wild-type parental strain dw15-1 (Figure 3.1A). Because nonhomologous integration of linearized plasmids into the Chlamydomonas genome can potentially
occur multiple times in a single line, genetic linkage of Hygromycin B resistance and the lipid
phenotype were examined to confirm insertional tagging of the gene responsible for the lipid
phenotype, a prerequisite for subsequent gene identification. Towards this end, the pgd1 mutant
was crossed with CC-198, a cell-walled strain (mating type-) and close relative of dw15-1, which
is mating type+ and the wild-type parental strain of pgd1. Strains CC-198 and dw15-1 were
compared for their lipid composition, but did not show major differences in TAG content. After
48 h of N deprivation, the ratio of fatty acids in TAG over total fatty acids in extracts was 0.46 ±
0.04 for dw15-1and 0.51 ± 0.04 for CC-198. A total of 83 meiotic progeny lines were analyzed,
of which 40 were resistant and 43 sensitive to Hygromycin B. The observed ratio approached the
hypothetical 2:2 segregation ratio suggesting a single plasmid insertion in the genome, although
the statistical limitations of the experiment would allow for multiple, but very tightly linked
plasmid insertions. Lipid analysis was performed on 14 progeny lines (Figure 3.1B) and the
results were compared to the wild-type parental strain and pgd1. The TAG fatty acid to total fatty
acid ratio of the eight Hygromycin B sensitive lines was similar to that of the parental strain,
while the six resistant lines showed a ratio similar to that of the original pgd1 mutant. Thus, the
Hygromycin B resistance marker appeared to be closely
linked to the mutation causing the lipid phenotype.

87

In addition to TAG deficiency, pgd1 cells gradually developed chlorosis and fully
bleached over the course of 12 days following N-depletion (Figure 3.1C), which was
accompanied by reduced cell viability (see below). However, in N-replete medium there was no
discernible difference in growth between the wild-type parental strain and pgd1 (Figure 3.2).
Thus, the ability to produce or maintain TAG seems to be required for the long-term viability of
the cells following N deprivation, which provides a clue towards a physiological role of TAG
accumulation under nutrient stress that will be further explored below.

The pgd1 lipid phenotype is caused by disruption of a putative lipase gene
To identify the plasmid insertion site, “SiteFinding” PCR (polymerase chain reaction)
(Tan et al. 2005) was employed. Random primers combined with primers annealing to the
positive strand of the Hygromycin B resistance gene (aph7”)on the pHyg3 plasmid (Figure 3.3A,
primer S1) generated two partial pHyg3 plasmid sequences present in opposite orientations as
depicted in Figure 3.3A, but no bona fide genomic flanking DNA. Probing a Southern blot of
BamHI-digested pgd1 genomic DNA with a pHyg3 fragment as indicated in Figure 3.3A, two
signals were observed (Figure 3.3B), while only one would be expected for a single insertion due
to the presence of only a single BamHI site in pHyg3. However, probing pgd1 genomic DNA
double-digested with PstI (no site in pHyg3) and XhoI (single site in pHyg3) with the same
probe, a single band was present (Figure 3.3B). Together these data suggested that two pHyg3
fragments were present in opposite orientations at the pgd1 locus. No true signal was obtained
from genomic DNA of the wild-type parental strain (Figure 3.3B).
Through “SiteFinding” PCR with plasmid-specific, nested primers S2-1 and S2-2
complementary to the other end of pHyg3 (Figure 3.3A, S2), a flanking genomic DNA (to the

88

right side of the insertion as shown in Figure 3.3A) was amplified. Sequencing indicated that one
end of the insertion bordered sequences within the predicted untranslated region of a gene
previously

annotated

as

CGLD15

(Conserved

in

Green

Lineage

and

Diatoms15;

Chlamydomonas v5.3 genome in the Phytozome database, http://www.phytozome.net/) on
chromosome 3 (position 6320421 - 6327099; gene locus Cre03.g193500) (Merchant et al. 2007),
which I designated PGD1 based on functional analysis presented below. A conserved catalytic
triad of Ser-Asp-His was predicted for the translated protein sequence of this gene, which is a
typical motif for hydrolases such as lipases (acyl hydrolases). The flanking genomic sequence on
the left side of the insertion (refer to Figure 3.3A) was obtained by PCR with primers F1 and S3
(Figure 3.3A). Sequence analysis of this fragment showed that the insertion was accompanied by
the deletion of 14 bp of genomic sequence that is unlikely to affect the neighboring gene 5’ to
PGD1. Based on these analyses, PGD1 was considered to be the most likely affected gene in the
pgd1 mutant and responsible for the observed lipid phenotype.
Insertions into the promoter or untranslated region of PGD1 were expected to affect gene
expression. Quantifying PGD1 transcript levels by real-time PCR (Figure 3.3C) showed greatly
reduced expression of this gene in the pgd1 mutant. The real time PCR results also confirmed the
increased expression of the PGD1 gene in the wild-type parental strain following N deprivation
previously observed during global transcript analysis (Miller et al. 2010). The up-regulation of
PGD1 expression following N deprivation in parallel with TAG accumulation suggested that the
gene product might play a role in TAG biosynthesis.
To independently confirm that the phenotypes of the pgd1 mutant were indeed caused by
the insertion into PGD1 described above, complementation analysis with a PGD1-containing
fragment from the bacterial artificial chromosome (BAC) clone, 5E6 (Grossman et al. 2003), was

89

conducted. The fragment used for transformation contained 2 kb 5’ and 1 kb 3’ of the predicted
PGD1 gene and was devoid of other predicted open reading frames. The pMN24 plasmid
(Fernandez et al. 1989) containing the NIT1 gene encoding nitrate reductase was used in a cotransformation experiment for selection on agar plates with nitrate as the N source. (Note, the
parental wild-type strain dw15-1 as well as pgd1 carry a mutation in the genomic NIT1 gene). To
screen for DNA fragments rescuing the observed chlorosis phenotype of pgd1 on N-limited
medium, I developed a “Single Step N Deprivation-Colony Color Screen” method. Agar plates
containing 0.5 mM instead of 10 mM nitrate were used for selection allowing colonies to form,
which then became N-deprived as nitrate was depleted. Under these conditions, pgd1 mutant
colonies turned from green to white within three weeks while colonies of the wild-type parental
strain or pgd1 colonies harboring an introduced wild-type copy of the PGD1 gene were expected
to remain green (Figure 3.4A). When the PGD1 genomic fragment was co-transformed with the
NIT1 marker, approximately 5-10 % colonies remained green. This frequency is at the lower end
of the range for previously reported co-transformation efficiencies (Kindle 1990). Eight colonies
scored as green and another eight colonies scored as white were chosen and the phenotype was
confirmed by spotting cells onto -N agar plates (Figure 3.4B). Genotyping was performed on the
junction of the plasmid insertion to confirm the presence of the gene disruption typical for the
pgd1 mutant (Figure 3.4C). Primers F1 and R were expected to give a signal specific for PGD1
either in the genome or introduced through the fragment, while primers S2-1 and R were
expected to give a signal specific for pgd1. According to this reasoning, seven of the eight green
lines (G1-G7) and one of the white lines (W4) contained the wild-type PGD1 gene (Figure
3.4C). The presence of a signal from a combination of S2-1 and R indicative of the presence of
the pgd1 background ruled out contamination by the parental strain. It seems likely that in outlier

90

G8 a secondary mutation caused the observed suppressor phenotype and in outlier line W4 the
introduced PGD1 gene was either mutated, or not adequately expressed, or silenced. Quantitative
lipid analysis of three green colony-forming lines (G1-3) showed that they regained their ability
to accumulate TAGs to similar levels as the parental strain (Figure 3.4D).

Extraplastidic lipids of pgd1 are affected in a consistent way
The fact that disrupting PGD1 led to lower TAG content argues against its gene product’s
role as a TAG lipase, because decreased TAG lipase activity in the mutant would be expected to
increase TAG content. An alternative hypothesis was that PGD1 releases acyl groups from
membrane lipids. The activation of the released fatty acids by formation of acyl-CoAs would
then make them available for TAG synthesis. To identify the lipid substrates for such a presumed
lipase,

the

abundant

membrane

monogalactosyldiacylglycerol

lipids

DGTS,

(MGDG),

phosphatidylethanolamine

digalactosyldiacylglycerol

(PtdEtn),
(DGDG),

phosphatidylglycerol (PtdGro), sulfoquinovosyldiacylglycerol (SQDG), and phosphatidylinositol
(PtdIns) were analyzed in the wild-type parental strain and pgd1 grown on N-replete and Ndepleted medium for 48 h (Figure 3.5A). The relative fraction of DGDG increased following N
deprivation as recently reported (Fan et al. 2011). However, no statistically significant difference
between the relative amounts of the respective membrane lipid classes for the wild-type parental
strain and the pgd1 mutant was observed.
Although relative amounts of membrane lipid classes were not altered under the growth
conditions used, it seemed possible that specific molecular species within each lipid class
represented by differences in the respective acyl group substituents were altered in the mutant.
For example, changes in fatty acid profiles of glycerolipids have been diagnostic in determining

91

whether the ER or plastid pathways of lipid assembly were affected in the respective mutants of
Arabidopsis (Kunst et al. 1988, Xu et al. 2003). Overall the decreased TAG content in pgd1 was
reflected by a reduced total amount of fatty acids per cell (Figure 3.5C) raising the question
whether specific TAG molecular species were missing in pgd1 consistent with the disruption of
one of several hypothetical TAG assembly pathways. Indeed, the total fatty acid profile of pdg1
∆9

was altered. Most prominently the relative fraction of oleate (18:1 ; number of carbons :
number of double bonds and position of double bonds from the carboxyl end) was reduced
(Figure 3.5B). Following N deprivation the wild-type parental strain showed an increase in the
relative amount of oleate that was not observed for pgd1, while the acyl composition of pgd1
was indistinguishable from that of the parental strain under N-replete growth conditions (Figure
3.5B). When the fatty acyl group profile of individual lipids following N deprivation was
examined, a decrease in oleate was observed for pgd1 not only in TAG (Figure 3.5D), but also in
DGTS (Figure 3.6A) and PtdEtn (Figure 3.6B). The latter two are presumed to be extraplastidic
membrane lipids (Giroud et al. 1988, Giroud and Eichenberger 1989), while the exclusive
location of TAG in cytosolic lipid droplets has recently been questioned (Fan et al. 2011,
Goodson et al. 2011). Oleate accounts for approximately 25% of the acyl groups in TAG, but
only up to 10% in DGTS or PtdEtn explaining why a loss of a specific molecular species
containing this fatty acid has more drastic effects on overall TAG content than on that of DGTS
and PtdEtn. The plastid lipids MGDG (Figure 3.6C), DGDG (Figure 3.6D), and PtdGro (Figure
3.6E), were not altered in their acyl composition. As apparently only extraplastidic lipids are
affected in thepgd1mutant, it seems possible that PGD1 activity affects the export of acyl groups

92

from the plastid or the assembly of extraplastidic lipids, assuming that the fraction of TAG
missing in pgd1 is extraplastidic.

Oleate is decreased in the sn-1 and sn-3 position of TAGs in pgd1
To gain more information on the origin of the diacylglycerol moiety for TAG
∆9

biosynthesis and possible role of oleate (18:1 ) in limiting TAG biosynthesis in pgd1,
positional analysis of TAG acyl groups was conducted with Rhizopus arrhizus lipase. Rhizopus
lipase specifically hydrolyzes the sn-1 position of membrane glycerolipids or the sn-1/sn-3
positions of TAG and is frequently used for the positional analysis of acyl groups in
glycerolipids (Fischer et al. 1973, Siebertz and Heinz 1977). Consistent with previous
observations (Fan et al. 2011), the sn-2 position of TAG is mostly composed of C16 acyl groups
while sn-1/sn-3 positions contain both C16 and C18 acyl; groups (Figure 3.7). While a decrease
in oleate in the sn-1 or sn-3 position was obvious, the method did not allow us to distinguish
between the two positions. For sn-1/sn-3, the relative content of 18:4, 18:3

ω6

and 18:1

∆11

were 2-

fold higher in the pgd1 mutant than in the wild-type parental strain (Figure 3.7). This was also
seen in the total composition of all TAG acyl groups (Figure 3.5D). Interestingly, 18:4 and
18:3

ω6

are mostly found in the extraplastidic lipids DGTS (Figure 3.6A) and PtdEtn (Figure
∆11

3.6B). Vaccenic acid (18:1

∆9

) is produced through elongation of 16:1 , at least in plants

(Nguyen et al. 2010), and is presumed to be extraplastidic. In view of an approximate 50%
reduction of TAG in the pgd1 mutant, the 2-fold relative increase in these three fatty acids
suggests that the supply of TAG precursors from extraplastidic lipid turnover was not affected.

93

Precursor fluxes from plastid lipids to TAG are reduced in pgd1
The analysis described above can only provide a static picture of lipid composition.
However, the defect in a putative lipase encoding gene in the pgd1 mutant suggested that lipid
remodeling or turnover might play a role in TAG accumulation in Chlamydomonas following N
deprivation. To observe possible changes in the dynamics of lipid metabolism in the pgd1
mutant, I employed pulse-chase labeling of membrane lipids using [14C]-acetate which can be
converted easily to precursors of fatty acid biosynthesis in plastids. The labeling pulse was
provided either before (150 min pulse duration, Figure 3.8) or during N deprivation (200 min
pulse duration initiated 12 h after the start of N deprivation, Figure 3.9 and Figure 3.10). Lipids
were extracted as indicated and fractions of label incorporation into major lipids during the chase
stage were calculated. The difference in the incorporation of label into TAG between the wildtype parental strain and pgd1 was more prominent when the labeling pulse was applied during N
deprivation (Figure 3.9) compared to its application prior to N deprivation (Figure 3.8). This
observation suggested that exchange of de novo synthesized acyl groups through the membrane
lipid pool into TAGs during N deprivation might involve PGD1, rather than the conversion of
preexisting membrane lipids formed under N-replete conditions to TAGs. When the pulse was
applied following N deprivation, plastid lipids, especially galactoglycerolipids, were rapidly
labeled (Figure 3.9). Conditions were chosen such that the total label in the lipid extract during
the chase phase remained approximately the same. In the wild-type parental strain (Figure 3.9A)
an increase of label in TAG was observed in parallel with a decrease of label in membrane lipids
suggesting the incorporation of acyl or diacylglyceryl groups derived from membrane lipids into
TAGs. After 25 h of chase the label remaining in membrane lipids was lower than that in TAGs.

94

This situation was reversed in the pgd1 mutant (Figure 3.9B). In particular the fraction of label in
the two galactoglycerolipids DGDG and MGDG remained much higher in pgd1. Because
MGDG is its precursor, DGDG was labeled with some delay. In fact, it was the most highly
labeled lipid in the pgd1 mutant extracts presumably because the transfer of labeled acyl or
diacylglyceryl groups from MGDG into TAGs was disrupted in the mutant. When the pulse was
applied prior to N deprivation (Figure 3.8), MGDG was the most highly labeled lipid reflecting
the fact that it is also the most abundant lipid under N-replete conditions, when TAG
biosynthesis is repressed. Apparently, following N deprivation de novo synthesized acyl groups
in the plastid are incorporated first into plastid membrane lipids, in particular MGDG, prior to
becoming incorporated into TAGs, and the incorporation of acyl groups into TAG seems to
require PGD1. Thus MGDG serves as precursor for a fraction of acyl or diacylglycerol groups,
those containing oleic acid (see Figure 3.5), incorporated into TAGs following N deprivation.
This process is disrupted in the mutant and more MGDG is converted to DGDG instead of TAG.
While PtdGro was rapidly turned over, the rates of turnover remained approximately the same in
the pgd1 mutant and the wild type (Figure 3.10). Based on these data I concluded thatPGD1 is
likely to be a galactoglycerolipid lipase prompting me to tentatively name the gene Plastid
Galactoglycerolipid Degradation 1 (PGD1).

PGD1 hydrolyzes acyl groups of MGDG with a preference for the sn-1position
To more directly determine the biochemical activity of PGD1, I produced the
recombinant protein in E. coli. The recombinant protein was affinity-purified from denatured
inclusion bodies (Figure 3.11A), renatured, and offered various substrate lipids in a lipase
activity assay. To control for spontaneous lipid hydrolysis or lyso-lipid contamination I assayed

95

the protein refolding buffer without proteins. As a positive control I assayed Rhizopus lipase. I
employed MGDG of different molecular composition as substrates to determine the enzyme’s
preference. “Mature” MGDG was isolated from Chlamydomonas cells, which predominantly
contain molecular species 18:3/16:4 (sn-1/sn-2). Using this substrate, the lyso-MGDG product
band that was generated was rather faint (Figure 3.11B). By analyzing the different fractions, I
Δ9

found that in the remaining MGDG after PGD1 hydrolysis, the 16:0 and 18:1

acyl groups

selectively disappeared while the major acyl groups 16:4 and 18:3 remained (Figure 3.12A and
5B). This suggested that PGD1 prefers to hydrolyze de novo synthesized MGDG (18:1

Δ9

/16:0)

Δ9

and the reaction stops when 18:1

/16:0 is depleted in the assay mixture. Hypothesizing that

PGD1 only hydrolyzes de novo-formed MGDG to release 18:1
Δ9

reduction of 18:1

Δ9

for further TAG biosynthesis,

in TAG of the pgd1 mutant can be explained. The lyso-MGDG obtained

using MGDG purified from Chlamydomonas exclusively contained 16:0 (Figure 3.12C),
suggesting that PGD1 prefers the sn-1 position. However, the low amount of free fatty acids
generated (Figure 3.12D) made it difficult to draw a firm conclusion, considering that there may
be lipids or fatty acids co-purified with the PGD1 protein.
To confirm that PGD1 prefers less desaturated molecular species, MGDG synthesized in
E. coli by the activity of recombinant cucumber MGDG synthase (Shimojima et al. 1997) was
used. In the buffer control lane, presumably no hydrolysis occurred as no generation of lysoMGDG was seen (Figure 3.11 B). MGDG synthesized in E. coli contains a combination of
Δ11

18:1

Δ9

, 16:0 or 16:1 acyl groups and is similar to the newly assembled MGDG in

Chlamydomonas. As indicated in Figure 3.11B by the intensity of the lyso-MGDG band, PGD1

96

was more active on the MGDG species produced in E. coli using the cucumber MGDG synthase
than the mature, mostly unsaturated MGDG from Chlamydomonas.
To obtain more information on the substrate preference of PGD1, I compared the acyl
composition of the remaining MGDG substrate, lyso-MGDG and free fatty acid products after
incubation with the corresponding fractions obtained from Rhizopus lipase hydrolysis. Rhizopus
lipase was inhibited by the buffer used for PGD1 refolding (Figure 3.11B). To generate lysolipid (including lyso-MGDG) standard, PBS instead of protein refolding buffer was used to
dissolve Rhizopus lipase for all the reactions mentioned below. The E. coli derived MGDG
Δ9

contains 16:0, 16:1

Δ11

and 18:1

(Figure 3.11D, untreated sample) with 16:0 and 18:1
Δ9

present in the sn-1 (Figure 3.11F), and 16:1

Δ11

and 18:1

Δ11

in the sn-2 position (Figure 3.11E), as

determined by Rhizopus lipase digestion. After PGD1 hydrolysis, the substrate MGDG was
partly remaining (Figure 3.11B). However, all the three major fatty acids were decreased to a
similar extent (Figure 3.11G). Thus, the remaining MGDG was due to limited enzyme activity
instead of the preference between different molecular species within E. coli-derived MGDG.
Lyso-MGDG (Figure 3.11H) and free fatty acids (Figure 3.11I) generated by PGD1 resembled
the corresponding fractions following Rhizopus lipase digestion in fatty acid compositions,
indicating that PGD1 prefers acyl groups at the sn-1 position similar to Rhizopus lipase.
I also explored the kinetics of PGD1 mediated hydrolysis of E. coli-derived MGDG.
Lyso-MGDG was detectable in 30 min and continuously increased within 4.5 h of incubation
(Figure 3.13A). At 4.5 h, the bulk of MGDG substrate still remained and I chose a 3 h incubation
time to test the relationship between reaction velocity and substrate MGDG availability. It should
be noted that MGDG is not soluble and, therefore, classic enzyme kinetics is not directly

97

applicable in this case. It should also be cautioned that the purified PGD1 enzyme went through
a denaturation process, and the lipase activity may not be completely regained during refolding
for all molecules present. Lyso-MGDG instead of free fatty acids was quantified as free fatty
acids can be either derived from MGDG or lyso-MGDG. The hydrolysis of MGDG exhibited
saturation as expected for an enzyme catalyzed reaction, with a linear increase in reaction
velocity up to an apparent MGDG concentration of 300 μM (Figure 3.13B).

PGD1 does not act on DGDG
During the labeling experiment shown in Figure 3.9, labeling of DGDG increased to a
greater extent in the pdg1 mutant than did MGDG, suggesting it might be a possible substrate of
PGD1. To test this possibility, DGDG extracted from Chlamydomonas was used as a substrate in
the PGD1 assay. However, no formation of lyso-DGDG was detected by sugar-specific staining
(Figure 3.11C). When E. coli-derived MGDG and Chlamydomonas derived-DGDG were offered
in equal amounts, only lyso-MGDG was formed, showing that PGD1 used MGDG but not
DGDG in this competition experiment, which might reflect the in vivo situation in which both
lipids are present in the same membrane. While a single MGDG molecular species 18:3/16:4
predominates in Chlamydomonas (Figure 3.6C), DGDG is represented by a greater variety of
molecular species with mostly 16:0, at the sn-2 position, and considerable amounts of 18:1

Δ9

,

18:2 and 18:3 acyl groups present at the sn-1 position of the glycerol moiety (Giroud et al. 1988),
(Figure 3.6D). Apparently, none of these DGDG molecular species is hydrolyzed by PGD1
under the conditions used. In addition, I tested PGD1 activity on other major membrane lipids
prepared from Chlamydomonas extracts. PGD1-dependent generation of lyso-DGTS and lyso-

98

PtdEtn were detectable by iodine staining, but much lower than those generated by Rhizopus
lipase (Figure 3.14A and 7B). Lyso-SQDG hydrolysis was not detectable by sugar-specific
staining (Figure 3.14D). Repeated trials of PtdGro hydrolysis by Rhizopus lipase failed to yield
an obvious lyso-PtdGro band (Figure 3.14), which is expected to run slower than PtdGro on TLC
plates. This might be due to the fact that the sn-2 position of PtdGro is composed of 16:0 and
Δ3

16:1

only (Giroud et al. 1988) while iodine stains lipids by binding to double bonds.

Nevertheless, a major decrease in the lipid substrate after PGD1 hydrolysis was visible for E.
coli MGDG (Figure 3.11B) but not PtdGro nor DGTS, PtdEtn and SQDG. At this time, synthetic
molecular species are not available for lipids such as MGDG and DGTS so I could not compare
lipids with exactly the same acyl compositions but different head groups.

DISCUSSION
The primary phenotype through which the pgd1 mutant was identified, is a reduction in
the accumulation of TAGs following N deprivation (Figure 3.1). Because the affected gene
(Figure 3.3), PGD1, encodes a galactoglycerolipid lipase with preference for the sn-1 position of
the respective glyceryl backbone (Figure 3.11), I had to reconsider how TAGs are synthesized in
Chlamydomonas. A mechanistic model that accommodates my current findings on the pgd1
mutant and the PGD1 protein, and that places PGD1 into the context of overall cellular lipid
metabolism is shown in Figure 3.15. Because the pgd1 mutant still accumulates considerable
amounts of TAGs, additional pathways not involving PGD1 are contributing to TAG
accumulation in N-deprived cells. For simplification purposes, a single lipid droplet is shown
receiving TAG assembled at the ER or the plastid envelopes.

99

An MGDG deacylation/acylation cycle involved in TAG biosynthesis of Chlamydomonas
Figure 3.15 summarizes my current working hypothesis of PGD1 function in TAG
metabolism. I have placed the galactoglycerolipids, in particular MGDG, at the center of the
plastid envelope pathway as my pulse-chase labeling experiment (Figure 3.9) suggested that acyl
groups or entire DAG moieties move through the membrane lipid fraction of the chloroplast
prior to incorporation into TAGs. While the labeling of galactoglycerolipids in the pgd1 mutant
was increased, the relative amounts of the membrane lipid classes did not change in the mutant.
Apparently, the pool size of MGDG and other membrane lipids is strictly controlled to maintain
functional photosynthetic membranes.
The importance of galactoglycerolipids in TAG metabolism in Chlamydomonas may
arise from the fact that this alga lacks PtdCho (Giroud et al. 1988). PtdCho is one of the most
rapidly labeled and metabolized membrane lipids in seed plants and acyl exchange involving
PtdCho has been suggested to play a role in the export of fatty acids relevant for extraplastidic
membrane lipid biosynthesis including that of TAGs (Bates et al. 2007, Bates et al. 2009).
Chlamydomonas membranes contain the betaine lipid DGTS which has been widely assumed to
play at least some of the roles of PtdCho in Chlamydomonas due to similarities in structure and
biophysical properties of the two lipids (Sato and Murata 1991). However, my labeling and lipid
analysis data showed no differences for DGTS between the parental strain and the pgd1 mutant
and suggested it is not involved, at least in the aspect of TAG biosynthesis that is affected in the
pgd1 mutant. Although DGTS and PtdEtn molecular species isolated from Chlamydomonas were
hydrolyzed by recombinant PGD1 to a limited extent (Figure 3.14A and 7B), DGTS and PtdEtn
showed the same change in molecular species composition in the mutant, i.e., the reduction of

100

oleic acid containing species (Figure 3.5 and Figure 3.6), as was seen also for TAGs. It should be
noted that DGTS and PtdEtn are extraplastidic membrane lipids. Oleate (18:1

Δ9

) and palmitate

(16:0) typically are the de novo synthesized acyl groups incorporated into the glyceryl backbone.
Thus the reduction of oleate in TAG in pgd1 suggests that the TAG affected in this mutant is
derived from glyceryl moieties containing these de novo synthesized acyl groups. Pulse-chase
labeling data obtained when labeled acetate was added prior to N deprivation showed few
differences between the mutant and the parental strain (Figure 3.8). However, stark differences
were observed, when the pulse was given following N deprivation (Figure 3.9) suggesting that
the fraction of TAG requiring PGD1 activity indeed involves de novo fatty acid biosynthesis.
As DGTS and PtdEtn in the pgd1 mutant are likely down-stream products of PGD1
activity just like TAG as discussed above, it seems probable that a plastid lipid serves as the true
substrate for PGD1 and that PGD1 may be involved in cycling newly synthesized fatty acids
through the plastid polar lipid pool. DGDG is not likely a major substrate for PGD1 in cells as it
is also not a substrate for PGD1 in vitro, even though it is highly labeled during pulse-chase
experiments in the pgd1 mutant (Figure 3.9). The delay in labeling of DGDG as compared to
MGDG is consistent with biosynthesis of DGDG by galactosylation of MGDG. Thus if cycling
of acyl groups into TAG through the MGDG pool is reduced in pgd1, the reduced flux from
MGDG to TAG allows for greater availability of MGDG for DGDG biosynthesis explaining
increased labeling of this lipid in the mutant.
In vitro lipase assays suggested that PGD1 prefers MGDG over DGDG, with a preference
for MGDG molecular species with fewer double bonds over 18:3/16:4 species, and a preference
for acyl groups at the sn-1 position over sn-2. To explain these observations, I propose an acyl-

101

editing cycle (Figure 3.15, process 1) involving MGDG assembled from de novo synthesized
fatty acids (18:1Δ9/16:0). One function of such a cycle involving a transient MGDG pool might
be the transfer of fatty acids synthesized in the plastid through the plastid envelope membranes
Δ9

effectively accomplishing a net export of 18:1

Δ9

acyl groups. As 18:1

is a major fatty acid in

TAG (Figure 3.5D), but not in DGTS or PtdEtn (Figure 3.6A and 2B), TAG is the most affected
of the extraplastidic lipids in the pgd1 mutant. Interestingly, MGDG in the pgd1 mutant did not
Δ9

accumulate 18:1

(Figure 3.6C), but apparently continued to become desaturated to its mature

18:3/16:4 molecular species. Alternatively, it seems likely that MGDG assembly from
newlysynthesizedacyl groups is feedback inhibited, adjusting the flow through the pathway to
the demands of the cell. Thus, the total amount of fatty acids in the pgd1 mutant is lower than in
the wild-type parental strain (Figure 3.5C) and the relative amount of 18:1

Δ9

in the total lipid

extracts was reduced (Figure 3.5D).
Mature MGDG found in thylakoid membranes is predominantly composed of the
18:3/16:4 species (Figure 3.6C). Perhaps the presence of the unusual 16:4 acyl group, or other
highly unsaturated acyl groups, protects MGDG degradation by PGD1, because it is necessary
building block of the photosynthetic membrane, while allowing cycling of de novo synthesized
acyl groups through the MGDG pool. This process is not perfect as 18:3 and 16:4 acyl groups
were present in TAGs of the wild-type parental strain and the pgd1 mutant indicating some
turnover of mature MGDG (Figure 3.15, process 2). It is likely that following severe N
deprivation photosynthetic membranes containing mature MGDG species are degraded to some
extent, perhaps involving lipases different from PGD1, and that these acyl or diacylglycerol
moieties find their way into the TAG fraction. But the bulk of the membrane has to be

102

maintained for a rapid recovery when nitrogen is re-supplied. Thus, the resistance of mature
Chlamydomonas MGDG to PGD1-catalyzed hydrolysis supports the hypothesis of de novo
synthesized acyl group cycling through a specific MGDG subpool. For this hypothesis to be
valid, an acyl-ACP:lyso-MGDG acyltransferase activity is required for acylation of lyso-MGDG.
Such an enzyme activity with a preference for the sn-1 position has been described for a
cyanobacterium (Chen et al. 1988). I assume that Chlamydomonas has an ortholog, although the
identity of the gene is not yet known.
One possible shortcoming of the proposed hypothesis is that no chloroplast transit peptide
was predicted for PGD1 suggesting it to be extraplastidic. However, a cytosolic location would
give PGD1 access to MGDG in the outer leaflet of the outer envelope membrane, where this
lipid has been shown to be present (Joyard et al. 1991). Analogously, the outer envelope protein,
SFR2, of Arabidopsis (Moellering et al. 2010) acts on MGDG. While I attempted to resolve this
issue, subcellular localization of PGD1 using GFP fusion constructs was unsuccessful.
Alternative localization approaches such as immunolocalization of PGD1 will have to await the
availability of antibodies.

How does oleate availability affect TAG biosynthesis?
The fatty acid composition of mutant TAG (Figure 3.5D and Figure 3.7) showed besides
∆9

∆11

a lack of 18:1 a higher relative abundance of 18:1

ω6

, 18:3

and 18:4 acyl groups. To explain

these observations, I considered the different sources for acyl groups that might be present in
∆9

TAGs: 1. de novo synthesis (Figure 3.15, process1), for which 18:1

is the diagnostic fatty acid

most increased following N deprivation (Figure 3.5C); 2. plastid membrane lipid degradation

103

ω3

(Figure 3.15, process 2) indicated by the presence of 16:4 and 18:3

in TAGs derived from

mature MGDG; and 3. extraplastidic membrane lipid modification and turnover (Figure 3.15;
∆11

process 3) characterized by the presence of 18:1

ω6

, 18:3

and 18:4 acyl groups in TAG. Fatty

acids such as 16:0 can be derived from multiple processes and are less diagnostic. It has been
demonstrated that the DAG backbone for TAG species mostly originates from the chloroplast
pathway since the sn-2 position of TAGs of Chlamydomonas contains up to 80% 16-carbon fatty
acids (Figure 3.7) and (Fan et al. 2011). I suggest that the plastid DAG pool primarily contributes
to TAG biosynthesis. Plastid DAG can derive from both, de novo assembly and plastid
membrane lipid degradation (Figure 3.15, process 2). Turnover of de novo synthesized MGDG
(Figure 3.15, process 1) will contribute to the cytosolic acyl-CoA pool which provides acyl
groups for the sn-3 position in TAGs. Similarly, lipid modification and turnover at the ER
(Figure 3.15, process 3) likely provide acyl-CoA substrate for the diacylglycerol acyltransferase
∆9

(DGAT). The absence of PGD1 impairs the export of 18:1

with two consequences: 1. decrease

∆9

of 18:1 -CoA as substrate for TAG biosynthesis (Figure 3.5D, arrow 1); and 2. decrease of
∆9

18:1 in the DAG backbones of DGTS and PtdEtn as shown in Figure 3.6A and 2B (Arrows 2
and 3).
∆11

The relative amounts of fatty acids from extraplastidic membrane lipid turnover (18:1
ω6

18:3

,

and 18:4) were doubled in TAG of the pgd1 mutant (Figure 3.5D and Figure 3.7).

Considering the approximately 2-fold decrease in TAG content in the pgd1 mutant, this can be

104

explained by the fact that extraplastidic membrane lipid turnover (Figure 3.15, process 3) was
not affected by the pgd1 mutation, while the total TAG amount was decreased 2-fold.
In contrast, the relative amounts of fatty acids provided by mature plastid membrane lipid
ω3

turnover (16:4 and 18:3

) remained the same or only slightly increased (Figure 3.5D and Figure

3.7). As a consequence, the absolute amounts of these two fatty acids in TAGs were decreased.

CONCLUSIONS
The isolation of the pgd1 mutant led to the discovery of a galactolipid lipase that plays a
role in TAG accumulation following N deprivation in Chlamydomonas. This finding was not
predicted based on our current knowledge of lipid metabolism in seed plants. However,
Chlamydomonas lacks PtdCho, which is the polar lipid in plants on which the modification of
acyl groups followed by acyl exchange happens. Thus, one might wonder whether the TAG
assembly pathway presented here is specific to Chlamydomonas. A cursory check suggests that
there are possible orthologs of PGD1 in plants and other algae and that the hypothesis outlined in
Figure 3.15 may, therefore, also have some relevance to TAG biosynthesis in plants and other
algae, at least under certain growth conditions and perhaps with modifications.

105

Figure 3.1. Phenotypes of the pgd1 mutant compared to the wild-type parental strain (WT).
(A) Time course of triacylglycerol (TAG) accumulation following N deprivation and (B)
phenotypic analysis of progenies from a cross between pgd1 and CC-198. Hs and Hr indicate
Hygromycin B sensitive and resistant lines respectively. (A, B) The ratio of fatty acids (FA) in
TAGs over total fatty acids in the lipid extracts is shown. Averages of three independent
measurements are provided. Error bars indicate standard deviation. (C) Appearance of the same
patches of N-deprived cells palced on agar-solidified TAP-N medium, 0-12 days after plating.
(For interpretation of the references to color in this and all other figures, the reader is referred to
the electronic version of this dissertation.)

106

Figure 3.2. Growth curves of wild-type parental strain (WT) and pgd1 mutant in regular
TAP medium. Cells were grown to stationary phase and inoculated into fresh TAP medium to
an optical density (at 750 nm) of 0.04. This experiment was repeated more than three times with
two biological replicates each time. A representative result is shown here. Each data point is the
average from three technical replicates with relative standard deviations smaller than 3 %.

107

Figure 3.3. Molecular characterization the pgd1 mutant. (A) A schematic representation
of the pHyg3 insertion into the genome of the pgd1 mutant. The triangle represents the
insert. Thick arrows indicate the orientation of the positive strand of the aph7’’ gene
conferring resistance to Hygromycin B. The PGD1 gene model is shown in a 5’ to 3’
direction from left to right, with exons and introns represented by black boxes and
connecting lines, respectively. 5’ and 3’ untranslated regions are shown as white boxes.
Crosshatched boxes indicate the position of the Southern probe used in (B). PCR primer

108

Figure 3.3. (cont’d)
sites are indicated by small arrows, which are not drawn to scale. The binding sites for the
“SiteFinding” primers and nested primers for the two ends of the insertion are shown as single
arrows S1 and S2, respectively. Sequences for all primers can be found in Supplemental Table 1.
(B) Southern blot of the pHyg3 insertion and surrounding genomic DNA. Genomic DNA was
digested with BamHI or PstI and XhoI and probed with the fragment (cross-hatched box) as
shown in (A) PstI cuts outside the insert and sites are not shown in (A). (C) Reverse
transcription-quantitative PCR of the PGD1 transcript in the wild-type parental strain (WT) and
the pgd1 mutant grown for 48 h in TAP (+N) or TAP-N (-N) medium. The abundance of PGD1
mRNA was normalized to RACK1. Data are presented as average ± SD (n=3).

109

Figure 3.4. Genetic complementation of pgd1 phenotypes with wild-type genomic DNA.

110

Figure 3.4. (cont’d)
(A) Section of an agar plate with pgd1 mutant colonies 20 days after transformation with a wildtype PGD1 containing fragment. Green colonies (arrows) are presumed to be complemented
lines; white colonies show the pgd1 characteristic chlorosis phenotype. (B) Confirmation of the
phenotypes of lines which form green (G1-8) and white colonies (W1-8) following restreaking
and 10 d of growth on TAP-N. The pgd1 mutant as well as the wild-type parental strain (WT) are
included. (C) Genotyping of the different lines. A scheme depicting the insertion site shows the
primer locations with arrows; sections of DNA gels with PCR products obtained with PCR
primers as indicated are shown below. Primer sizes are not to scale. (D) Quantitative analysis of
TAG of three lines rescued with PGD1 genomic DNA after 48 h of growth in TAP-N medium.
The ratio of fatty acids (FA) in TAGs over total fatty acids in the lipid extracts is shown.
Averages of three independent measurements are provided. Error bars indicate standard
deviation.

111

Figure 3.5. Detailed lipid analysis of the wild-type parental strain (WT) and pgd1 mutant in
N-replete medium (+N) and 48 h after transfer to N-depleted medium (-N). (A) Relative

112

Figure 3.5. (cont’d)
abundance of major polar lipid classes. (B) Relative fatty acid (FA) composition and (C) cellular
contents of total cellular fatty acids. (D) Composition of fatty acids esterified to TAG. Averages
of three replicates are shown with errors bars indicating SD. Lipid abbreviations: DGTS,
diacylglycerol-N,N,N-trimethylhomoserine;

DGDG,

digalactosyldiacylglycerol;

MGDG,

monogalactosyldiacylglycerol; PtdEtn, phosphatidylethanolamine; PtdGro, phosphatidylglycerol;
PtdIns, phosphatidylinositol; SQDG, sulfoquinovosyldiacylglycerol. Fatty acids are designated
as chain length: number of double bonds. Positions of double bonds are indicated with Δ
(counting from carboxyl group) or ω (counting from the methyl group). In (B) 16:2 is composed
of 16:2

Δ7,10

and 16:2

Δ10,13

.

113

Figure 3.6. Fatty acid compositions of DGTS, PtdEtn, MGDG, DGDG, PtdGro of the wildtype parental strain (WT) and pgd1 mutant in N-replete medium (+N) and 48 h after
transfer to N-depleted medium (-N). Lipid abbreviations are as defined for Figure 3.5.
Averages of three replicates are shown with error bars indicating SD. Fatty acids are designated
as chain length: number of double bonds. Positions of double bonds are indicated with Δ

114

Figure 3.6. (cont’d)
(counting from carboxyl group) or ω (counting from the methyl group).

115

Figure 3.7. Positional analysis of TAG acyl groups of the wild-type parental strain (WT)
and pgd1 mutant 48 h after transfer to N-depleted medium (-N). Purified TAG from
Chlamydomonas cells were hydrolyzed by Rhizopus lipase. Free fatty acids were presumed to be
derived from sn-1/sn-3 position of the glyceryl backbone of TAG (A) while monoacylglycerol
contains the sn-2 position acyl groups (B). Averages of three replicates are provided with error
bars indicating standard deviation.

116

Figure 3.8. In vivo pulse-chase acetate labeling of lipids in the wild-type parental strain
(WT) and the pgd1 mutant before N deprivation. Labeled acetate was added prior to the
transfer to TAP-N medium. The length of the

14

C acetate labeling pulse was 150 min after which

the cells were transferred to TAP-N medium lacking labeled acetate. Cells were collected at the
times indicated and lipid extracts were prepared and analyzed. The fraction of label in all
analyzed lipids is given; only lipids containing the bulk of the label or those most relevant for the
discussion are shown. Lipid abbreviations are as defined for Figure 3.5. The data are from one
representative experiment of a series of independent experiments.

117

Figure 3.9. In vivo pulse-chase acetate labeling of lipids in the wild-type parental strain
(WT) and the pgd1 mutant. Labeled acetate was added 12 h following transfer of cells to TAPN medium. The length of the

14

C acetate labeling pulse was 200 min after which the cells were

transferred to TAP-N medium lacking labeled acetate. Cells were collected at the times indicated
and lipid extracts were prepared and analyzed. The fraction of label in all analyzed lipids is
given; lipids containing the bulk of the label or those most relevant for the discussion are shown
in this figure. Fractions of label in other lipids are shown in Figure 3.10. Lipid abbreviations are
as defined for Figure 4. The data are from one representative experiment of a series of
independent experiments.

118

Figure 3.10. In vivo pulse-chase acetate labeling of PtdEtn, PtdGro, SQDG and PtdIns in
the wild-type parental strain (WT) and the pgd1 mutant. The data are from the same set of
experiments as shown in Figure 3.9. Lipid abbreviations are as described in Figure 3.5. SQDG,
PtdIns and TLC origin were scraped together as one fraction (SPO).

119

Figure 3.11. Activity of the recombinant PGD1 protein on MGDG and DGDG. (A) SDSPAGE of purified PGD1 protein and whole cell lysates (WCL) from E. coli cells expressing the
PGD1 open reading frame and the empty vector control. Protein loading was 6 μg per lane for
whole cell lysates. Purified PGD1 protein loaded was 1 μg (Quantified as in Methods; possibly
biased by components in refolding buffer). Proteins were stained by Coomassie Brilliant Blue.
The arrow points towards the PGD1 protein. (B) A thin-layer chromatogram of polar lipids from
the lipase assay mixtures to which either mature MGDG extracted from Chlamydomonas, or
MGDG extracted from the E. coli (E. coli) strain over-expressing cucumber MGDG synthase
were added as substrates. (C) A thin-layer chromatogram of polar lipids from the lipase assay
mixtures to which either DGDG extracted from Chlamydomonas alone or mixed in with E. coli
derived MGDG at a 1:1 molar ratio were added as substrates. Glycolipids were visualized with
α-naphtol reagent. Reaction products obtained with refolded PGD1 protein, blank refolding
buffer and Rhizopus arrhizus lipase dissolved in protein refolding buffer (RaLip-R) or PBS

120

Figure 3.11. (cont’d)

(RaLip-P) were analyzed. (D)-(I) Gas liquid chromatograms of methyl esters derived from a
buffer control reaction containing E. coli-derived MGDG (D) or different fractions after lipase
reaction with E. coli-derived MGDG as discussed in details in the text. As an internal standard,
15:0 was used.

121

Figure 3.12. Hydrolysis of Chlamydomonas-derived MGDG by PGD1. Gas liquid

122

Figure 3.12. (cont’d)
chromatograms of methyl esters derived from different fractions after lipase reaction are shown
as discussed in detail in the text. As an internal standard 15:0 was used. (A) input
Chlamydomonas MGDG; (B) remaining MGDG after PGD1 hydrolysis; (C) and (D) lysoMGDG and free fatty acids generated from PGD1 hydrolysis, respectively.

123

Figure 3.13. Quantitative hydrolysis of E. coli derived MGDG by PGD1. (A) A thin-layer

124

Figure 3.13. (cont’d)
chromatogram of polar lipids from the PGD1 assay mixtures to which MGDG extracted from the
E. coli strain over-expressing cucumber MGDG synthase was added as substrate. Glycolipids
were visualized with α-naphtol reagent. One aliquot of the same volume was extracted and
loaded onto the TLC for each time point. (B) Dependence of MGDG hydrolysis on MGDG
concentration. Reaction velocity was quantified as the amount of lyso-MGDG generated per min
per mg purified PGD1 protein. A representative result is shown for each panel.

125

Figure 3.14. Activity of the recombinant PGD1 protein on DGTS, PtdEtn, PtdGro and
SQDG. Thin-layer chromatograms of polar lipids from the PGD1 assay mixtures to which

126

Figure 3.14. (cont’d)
Chlamydomonas-derived membrane lipids were added as substrates. (A), diacylglycerol-N,N,Ntrimethylhomoserine (DGTS); (B) PhosphatidylEthanolamine (PtdEtn); (C) Phosphatidylglycerol
(PtdGro); (D) Sulfoquinovosyldiacylglycerol (SQDG). Exposure to iodine vapor was used to
visualize lipids for reactions on DGTS, PtdEtn and PtdGro. SQDG and lyso-SQDG were stained
by α-naphtol reagent. Substrates treated with Rhizopus lipase was used to generate the lyso-lipid
standards. A representative result is shown for each panel.

127

Figure 3.15. Hypothesis placing PGD1 into the context of overall cellular lipid metabolism
explaining its role in TAG biosynthesis. For simplicity a single lipid droplet (LD) is shown
forming at the endoplasmic reticulum (ER) or the chloroplast envelope (Cp Env). Thylakoid
membranes harboring the two photosystems have been omitted. Three lipid turnover processes
discussed in the text are indicated by numbers: 1. Turnover of newly synthesized MGDG; 2. acyl
group modification and lipid turnover at the ER; 3 turnover of mature MGDG and other
thylakoid lipids at the plastid envelopes. Enzymes, protein complexes and processes are
italicized: DGAT, diacylglycerol acyltransferase; FAS, fatty acid synthase complex; Fd,
ferredoxin; FNR, ferredoxin: NADP+ reductase; PSI and II, Photosystem I and II. Substrates and
products: DAG, diacylglycerol; e-, electron; FA, fatty acid (carbons: double bonds); MGDG,
monogalactosyldiacylglycerol; PL, polar lipids; TAG, triacylglycerol; ROS, reactive oxygen
species. Not all intermediates or reactions involved are shown.

128

Table 3.1. Oligonucleotide primers used in this study. All primer sequences are written in 5’
to 3’ direction. N indicates a random nucleotide.
Name

Sequence

SF

ACCAACATCTTCGTGGACCT

SR

CTCCTCGAACACCTCGAAGT

SiteFinder6

SiteFinder8

CACGACACGCTACTCAACACACCACCTCGCACAGCGTCCTCAAGCG
GCCGCNNNNNNGCAT
CACGACACGCTACTCAACACACCACCTCGCACAGCGTCCTCAAGCG
GCCGCNNNNNNGCAG

SFP1

CACGACACGCTACTCAACAC

SFP2

ACTCAACACACCACCTCGCACAGC

S1-1

ACTGCTCGCCTTCACCTTCC

S1-2

CTGGATCTCTCCGGCTTCAC

S2-1

ATAGGGGTTCCGCGCACAT

S2-2

CCGAAAAGTGCCACCTGAC

S3

GTCATCCCATGGAAGCTTGG

F1

ACATCGTGAATGGCAAAACA

R

ATTGCGCGGGTTTAGAACTT

qF

AGCCAGCTATTGTCGCACTT

qR

CAAGAAATCCGCTGACATCC

CF1

CR1

TATCCATATGACGTTCCAGATTACGCTGCTCAGTGCGGCCGCATGAG
CCAGCTATTGTCG
GAATTTCGACGGTATCGGGGGGATCCACTAGTTCTAGCTAGATCACC
GGCAGGCGTGTGG

CF2

GGATCCGATGAGCCAGCTATTGTCG

CR2

GTCGACCCGGCAGGCGTGTGGGTC

CF3

AAAGAGGCGCGTCATGAGCCAGCTATTGTCG

CR3

CGGAAGGCGCGTCACCGGCAGGCGTGTGG

129

Chapter 4
Chlamydomonas shunts electrons into triacylglycerols to avoid photooxidative stress
during nutrient deprivation

1

1

Two undergraduates worked on this project under my supervision. Fabian Brandenburg
contributed data for Figure 4.3. Cassandra Johnny contributed part of data for Figures 4.1, 4.4,
4.5 and 4.6. I provided supervision, summarized the data and wrote this chapter. Figures 4.1 and
4.2 are part of a paper in press at The Plant Cell in Xiaobo Li, Eric R. Moellering, Bensheng Liu,
Cassandra Johnny, Marie Fedewa, Barbara B. Sears, Min-Hao Kuo, Christoph Benning. (2012)
A galactoglycerolipid lipase is required for triacylglycerol accumulation and survival following
nitrogen deprivation in Chlamydomonas reinhardtii.

130

ABSTRACT
A mutant in Chlamydomonas reinhardtii was previously found to be deficient in
triacylglycerol (TAG) biosynthesis under nitrogen (N) deprivation and designated pgd1 (Chapter
3). In this study, the pgd1 mutant is used to explore the physiological significance for TAG
production in microalgae. The pgd1 mutant exhibited chlorosis, accumulated reactive oxygen
species (ROS) and lost viability under N deprivation. Blocking photosynthetic electron transport
by 3-(3,4-dichlorophenyl)-1,1-dimethylurea reversed these phenotypes. This suggests that TAG
accumulation serves as an electron sink under N deprivation to prevent the generation of reactive
oxygen species through reduction of molecular oxygen (Mehler reaction). A suppressor screen
was conducted in the pgd1 mutant background for suppression of the chlorosis phenotype and
several suppressor mutants were obtained successfully. Biochemical, physiological and genetic
characterizations of these suppressors are in progress.

131

INTRODUCTION
Under nitrogen (N) deprivation, algae such as Chlamydomonas reinhardtii produce
substantial amount of triacylglycerols (TAGs) (Wang et al., 2009; Moellering and Benning,
2010). The physiological significance of this process has not been extensively explored.
Understanding of this process will have implications on how algae can be modified to produce
more TAG for large-scale biofuel production. In microalgae, besides N deprivation, other
stresses such as high salinity also increase TAG content substantially (Weers and Gulati, 1997;
Hu et al., 2008). Therefore, instead of being just carbon and energy storage, TAG accumulation
is likely a mechanism conferring stress resistance. Hu et al. hypothesized that under stress
conditions, electrons accumulate in the photosynthetic electron transport chain. Over-reduction
of the photosynthetic electron transport chain can lead to a burst of reactive oxygen species
(ROS) (Mehler, 1951; Asada, 1994), which may damage membrane lipids, proteins and nucleic
acids. By channeling electrons into the production of NADPH, which can be efficiently
consumed by fatty acid synthesis, the oxidative stress can be relieved (Hu et al., 2008). This
hypothesis is supported by the observation that TAG is increased in Nannochloropsis salina cells
receiving light above the saturating light intensity (Van Wagenen et al., 2012). However, it
could not be directly tested without a mutant strain that is impaired in TAG accumulation.
In a previous screen for mutants with altered TAG metabolism, a mutant disrupted in the
Plastid Galactolipid Degradation 1 (PGD1) gene was found to be 50 % decreased in the TAG
content. PGD1 encodes a lipase that facilitates an acylation/deacylation cycle on de novo
synthesized monogalactosyldiacylglycerol (MGDG) and provides 18:1

∆9

for extraplastidic lipid

biosynthesis (Chapter 3). Importantly, the pgd1 mutant was specifically affected in the amount of

132

TAG but not the major membrane lipids (Figure 3.5) and thus provides an excellent tool to
explore the consequences that occur when TAG is less abundant.
In this research, the pgd1 mutant is characterized in its survival rate, production of
reactive oxygen species, and response to the inhibition of the photosynthetic electron transport
under N deprivation. A suppressor screen in the pgd1 mutant background was also performed to
test the relationship between TAG production and redox control.

MATERIALS AND METHODS
Strains and growth conditions
The pgd1 mutant strain was derived from dw15-1 (cw15, nit1, mt+) strain of
Chlamydomonas reinhardtii obtained from the Arthur Grossman Lab as described in Chapter 3.
The dw15.1 strain is thus termed the wild-type parental strain (WT). Conditions for
photoheterotrophic growth in Tris-acetate-phoshate (TAP) medium (Harris, 1989) and N
deprivation were as described before (Moellering and Benning, 2010). For photoautotrophic
growth, HCl was used instead of glacial acetic acid for neutralization and the medium is termed
Tris-minimal-phoshate (TMP). For spotting on TAP-N agar, log-phase cells were condensed to 2
7

X10 cells/ml and 5 μL was loaded for each strain.

Chlorophyll, viability and TBARS analyses
Chlorophylls were extracted from fresh cell pellets with 80% acetone and concentrations
were calculated from the absorbance values at 647 nm and 664 nm according to (Zieger and
Egle, 1965). To assess cell viability, cells were grown in liquid cultures of TAP or TAP-N. On
days 0, 2, 4, and 7 a set volume of culture was diluted and spread onto agar-solidified TAP

133

medium supplemented with 0.4 % yeast extract. To assess cell viability, colony forming units
were counted one week later. Cells from a second aliquot were fixed in 3.7% formaldehyde (in
water) and counted using a hemocytometer under a microscope. Viability percentages (colonies
formed per total cells counted each day) were normalized to the values on day 0. TBARS were
prepared by extraction with thiobarbituric acid/trichloroacetic acid solution (0.3 % and 3.9 %
respectively) and determined by measuring absorbance at 532 nm as previously described (Baroli
-1

-1

et al., 2003). The extinction coefficient used was 155 mM cm .

Primary and secondary suppressor screen
Mutants were generated by plasmid insertion in the pgd1 mutant background through
glass bead transformation (Kindle, 1990). The selectable marker used was the NIT1 gene which
encodes the nitrate reductase and confers the ability to grow on medium with nitrate instead of
ammonium. The NIT1 DNA was obtained through BamHI and SalI digestion of the pMN24
(Fernandez et al., 1989), followed by gel purification (all the restriction endonucleases were
purchased from New England Biolabs, http://www.neb.com). For each transformation 250 ng
DNA was used. The transformants were then inoculated in TAP and grown in the dark on a
shaker for 8h, and spread on low nitrate plates (TAP with 0.5 mM nitrate) as described in
Chapter 3. Typically colonies were visible in a week and after another week, most colonies
turned white. The colonies that remained green were restreaked and maintained on solid TAP
medium with 10 mM nitrate.
For a second screen, the putative suppressors were grown in regular TAP together with
dw15-1 and the pgd1 mutant to log-phase. Cells were then spotted to a TAP-N agar plate and

134

2 -1

incubated under continuous light (70-80 μmol m- s ). After 10 days, the appearance of the
putative suppressor was compared to that of the pgd1 mutant and dw15-1 by the naked eyes.
DNA and RNA techniques
Genomic DNA of Chlamydomonas strains was prepared according to (Newman et al.,
1990). SiteFinding PCR was conducted according to (Tan et al., 2005) to identify the insertion
locus of the NIT1 gene, with minor modifications in the PCR conditions and with primers
optimized for the NIT1 marker. The primers used for insertion identification in pgd1fb12 were:
SiteFinder5

(CACGACACGCTACTCAACACACCACCTCGCACAGCGTCCTCAAGCGG-

CCGCNNNNNNGCTA, 5’-3’, N represents a random nucleotide) in combination with L2
(ATGTTC-ACGGAGCACCTCTT) and L1 (GACCGCCAGATCCAGTTCTA). In addition,
nested primers SFP1 (CACGACACGCTACTCAACAC) and SFP2 (ACTCAACACACCACCTCGCACAGC) were used. For reverse transcriptase-PCR (RT-PCR), pgd1fb12qRTF (TGCTTCAAGGAGGAGACGTT) and pgd1fb12qRTR (CCCAGGTCACGTGTGTAGAA) were used to
amplify a region of cDNA in gene 525683 (see Results and Discussion). RACK1-F
(GTCATCCACTGCCTGTGCTT) and RACK1-R (CCTT-CTTGCTGGTGATGTTG) were used
to amplify the RACK1 cDNA.

Lipid analysis
Lipid extraction, thin-layer chromatography (TLC) for lipid separation and gas-liquid
chromatography (GLC) were performed as previously described (Moellering and Benning,
2010). Briefly, cell pellets was resuspended and vortexed in methanol, chloroform, 88% formic
acid (2:1:0.1 by volume). To the mixture, 0.5 volume of lipid extraction buffer (1 M KCl, 0.2 M
H3PO4) was added and vortexed, followed by low speed centrifugation for phase separation. For

135

TAG quantification and fatty acid profile analysis, lipids were loaded to TLC on Silica G60
plates (EMD chemicals, #5721-7, http://www.emdchemicals.com) and developed in petroleum
ether-diethyl ether-acetic acid (80:20:1 by volume). Spots were scraped for transesterification
and GLC after brief exposure to iodine vapor.

RESULTS AND DISCUSSION
The pgd1 mutant loses viability following N deprivation
In contrast to the wild-type parental strain, the pgd1 mutant liquid cultures and colonies
on agar solidified medium became chlorotic 5-9 days following N deprivation (Figure 4.1A). I
took advantage of this observation during the complementation analysis described in Chapter 3
(Figure 3.4). The increasing chlorosis correlated well with the decrease in the chlorophyll
content (Figure 4.1B), as well as the decline in the viability of the pgd1 strain following N
deprivation (Figure 4.1C). Following N deprivation Chlamydomonas typically ceases cell growth
after approximately one cell cycle (James et al., 2011). However, these cells continue to capture
light with their photosynthetic light harvesting complexes. If electron acceptors become limiting
due to the cessation of growth under these conditions, photosynthetic electron transport chain
components may become overreduced. Indeed, it has been hypothesized that enhanced fatty acid
synthesis and sequestration of acyl groups in TAG provide an electron sink, because acyl groups
are among the most reduced carbon compounds that algae can produce (Hu et al., 2008). A
+

potential consequence of TAG deficiency is the increase in the the NADPH/NADP ratio. This
is because NADPH is a major reductant in fatty acid synthesis. With decreasing availability of
+

NADP , molecular oxygen may become an alternative electron acceptor for photosystem I. Thus
when photosynthetic electron transport exceeds the capacity of the NADP+ pool to accept

136

electrons, in the pgd1mutant due to decreased TAG synthesis, superoxide may be generated and
further converted to H2O2 and hydroxyl radicals (Mehler, 1951). Overproduction of these highly
cytotoxic reactive oxygen species (ROS) may lead to cell death. To begin to test this hypothesis,
I took advantage of the herbicide DCMU, which specifically inhibits photosynthetic electron
transfer at the acceptor side of PS II (Trebst et al., 1970). DCMU treatment is well known to
decrease the generation of superoxide and other reactive oxygen species from PS I (Davies et al.,
1996; Wen et al., 2008; Robert et al., 2009). When DCMU was added to N-deprived cultures,
chlorosis and loss of viability were suppressed (Figure 4.1A, B, and C). To further verify this
hypothesis, I analyzed thiobarbituric acid reactive substances (TBARS), a product of reactive
oxygen species (Baroli et al. 2003), and observed a burst of TBARS in the pgd1 mutant, which
was also reversed by DCMU (Figure 4.1D). As expected, DCMU did not rescue the pgd1 TAG
phenotype, but did decrease TAG levels of the wild-type parental strain (Figure 4.1E) because of
the decrease in electrons provided for NADPH generation as recently observed by others (Fan et
al., 2012). A model explaining how cells shunt electrons to TAG biosynthesis to prevent the
generation of ROS is shown in Figure 4.2.
The assay employed to detect thiobarbituric acid reactive substances (TBARS) is
commonly used to measure the consequences of oxidative stress in Chlamydomonas (Baroli et
al., 2003; Fischer et al., 2007). As products of lipid peroxidation, TBARS are easier to detect
than ROS themselves which are short-lived (Shulaev and Oliver, 2006). I observed a strong
accumulation of TBARS in the pgd1 mutant on day 7 of N deprivation (Figure 4.1D). However,
this effect was preceded by the loss of ability to form colonies indicating a loss in cell viability
(Figure 4.1C). Similarly, Baroli et al. tested the ability to form colonies and TBARS
accumulation in a Chlamydomonas mutant sensitive to high light and also observed a similar lag

137

in TBARS formation (Baroli et al., 2004). It is likely that lower amounts of ROS can cause loss
of colony forming ability, while the formation and accumulation of detectable levels of TBARS
requires more time. However, we cannot exclude the alternate explanation that cell death itself is
the cause of TBARS accumulation in the pgd1 mutant.
If the proposed hypothesis that TAG biosynthesis mitigates ROS formation at PSI
following N deprivation is correct, I expect to identify more mutants deficient in TAG
accumulation that lose viability following N deprivation. In fact, the essentiality of TAG
accumulation opens new opportunities for additional forward genetic screens of mutants
compromised in genes required for TAG biosynthesis and its regulation, or even photosynthetic
electron transport.

Suppressor mutants were successfully obtained
To further understand the physiological significance of TAG accumulation, a suppressor
screen was conducted in the background of the pgd1 mutant, for suppression of the chlorosis
phenotype. The pgd1 mutant was derived from the dw15-1 strain which was disrupted in the
nitrate reductase gene. Taking advantage of that, insertion mutagenesis was performed with the
NIT1 marker and colonies were selected on agar solidified medium containing nitrate instead of
ammonium as the nitrogen source. The concentration of nitrate used was 0.5 mM, which is lower
than the regular concentration of 10 mM. As described in Chapter 3, the low concentration
allowed colonies to form and after two weeks, most colonies show the phenotype of the pgd1 as
the nitrate is depleted.
Among 10, 000 colonies obtained from transformation, approximately 25 colonies
remained green (Figure 4.3A). These 25 strains were then grown up to log-phase and spotted to

138

an agar plate without nitrogen source for a confirmation of phenotype. Figure 4.3B shows
examples of a confirmed suppressor and a false positive strain. Two confirmed suppressors,
pgd1fb12 and pgd1fb16, have been further characterized as described below.

No suppressor mutant is rescued in the amount or fatty acid composition of TAG
If the pgd1 mutant loses viability due to the overreduced photosynthetic electron transport
chain caused by decreased fatty acid synthesis as hypothesized in Figure 4.2, increased fatty acid
synthesis will rescue the mutant. TAG amount and fatty acid composition was analyzed in all the
∆9

suppressor mutants. The pgd1 mutant showed low amount of TAG and low 18:1

abundance in

TAGs. However, in most of the suppressor mutants the amount of TAG or the percentage of
∆9

18:1

in TAG were the same as in the pgd1 mutant. The pgd1fb12 and pgd1fb16 are shown in
∆9

Figure 4.4. Interestingly, the fraction of 18:1 in the TAG was even lower in mutant pgd1fb12
than in the original pgd1mutant (Figure 4.4).
Why wasn’t a mutant isolated with a rescued TAG content? It is possible that more genes
are involved in photosynthetic electron transport or redox control than those involved in TAG
biosynthesis and that the suppressor screen is not yet saturated. Another possibility is that the
pgd1 mutation serves as a bottleneck for fatty acid export and mutations upstream or downstream
cannot compensate for this defect.

Mutant pgd1fb16 exhibited slower growth
To investigate whether any of the suppressor mutants is blocked in photosynthetic
electron transport, three mutants (pgd1fb3, pgd1fb12 and pgd1fb16) were grown in parallel with

139

the pgd1 mutant and its wild-type parental strain (WT) in TMP or TAP under constant light.
Hypothetically, a mutation that decreases photosynthetic electron transport would lower the
phototrophic growth. While pgd1fb3 and pgd1fb12 did not exhibit a difference from the pgd1
mutant and WT, pgd1fb16 had a slower growth rate (Figure 4.5) in both TMP and TAP.
Photoheterotrophic growth in TAP under constant light is dependent on both photosynthesis and
the organic carbon source-acetate for energy. At this time, it remains elusive whether the slower
growth phenotype of pgd1fb16 is due to a defect in photosynthesis or other processes such as cell
cycle progression. To differentiate between these two possibilities, growth will be tested in TAP
medium in the dark. If the growth of pgd1fb16 is similar to that of the pgd1 mutant, pgd1fb16 is
probably affected in photosynthesis. Otherwise, the low growth rate in TMP and in TAP under
light should be attributed to other reasons.

pgd1fb12 is disrupted in a putative aldehyde dehydrogenase gene and possibly several other
genes
SiteFinding PCR (Tan et al., 2005) was performed on several mutants to identify the
insertion loci of NIT1 marker with success on pgd1fb12 and pgd1fb16 (Figure 4.6A). Band 1
was amplified from every strain, including the pgd1 mutant. Since the mutation in the
endogenous NIT1 gene in the dw15-1 strain was not a deletion, this band is likely to be from
amplification of that locus. Bands 2 and 3 were specific for pgd1fb12 and pgd1fb16 respectively
and have been sequenced and BLASTed (Altschul et al., 1997) against the Chlamydomonas
genome (http://genome.jgi-psf.org/Chlre4/Chlre4.home.html). For pgd1fb12, the insertion is
mapped to 5 bp above the 5’ UTR region of a gene predicted to encode an aldehyde
dehydrogenase (protein ID: 525683, chromosome_8:3499179-3508643). A diagram of the

140

insertion is shown in Figure 4.6B. For pgd1fb16, BLAST (Altschul et al., 1997) yielded multiple
hits in the genome and further experiments are required to elucidate the insertion locus.
RT-PCR was done to confirm that the insertion affected the expression of 525683.Under
both N replete (+N) and N-deprived (-N) conditions, the transcript of 525683 is missing while
that of a positive control gene-RACK1 can be detected (Figure 4.6C). We concluded that in
mutant pgd1fb12, 525683 is abolished in the transcription. However, whether the mutation in
525683 is responsible for the phenotype will have to await the identification of the other junction
of the plasmid and complementation experiments.

141

Figure 4.1. Biochemical and physiological characterization of wild-type parental strain
(WT) and the pgd1 mutant following N deprivation. (A) Appearance of cultures grown in
TAP-N for the number of days indicated. The electron transport chain inhibitor DCMU dissolved
in dimethyl sulfoxide was present at a final concentration of 2 μM as indicated. Two

142

Figure 4.1. (cont’d)
representative cultures per line are shown. (B) Time course of total cellular chlorophyll (Chl)
content. (C) Time course of cell viability relative to day 0, the start of N deprivation following
transfer to TAP-N medium. (D) Time course of cellular thiobarbituric acid reactive substances
(TBARS) content. (E) TAG accumulation presented as ratio of fatty acids (FA) in TAGs over
total fatty acids in the lipid extracts after 2 d of N deprivation. For all quantitative data, three
replicates were averaged with SD indicated by the error bars.

143

Figure 4.2. A scheme on the role of TAG production in redox control. Enzymes, protein
complexes and processes are italicized. Abbreviations: FAS, fatty acid synthase complex; Fd,
ferredoxin; FNR, ferredoxin: NADP+ reductase; PSI and II, Photosystem I and II; TAG,
triacylglycerol; ROS, reactive oxygen species. Not all intermediates or reactions involved are
shown. The dashed arrow indicates multiple steps.

144

Figure 4.3. Examples of primary and secondary screen for pgd1 suppressors. WT represents
dw 15.1, the wild-type parental strain for pgd1. A, primary screen. Transformants were spread to
TAP media with 0.5 mM nitrate as the nitrogen source. Typically it takes one week for colonies
to form and after another week most colonies will turn white. The colonies that remain green (as
pointed to by the arrow) will be picked and maintained as putative suppressors. These strains will
be confirmed in the phenotype by spotting same number of log-phase cells to TAP-N which
completely lacks ammonium or nitrate. After 10 days, pgd1 mutant can be told from its wildtype parental strain in appearance and the candidate strains will be visually scored to be
suppressed or not. One example of positive and negative result is shown.

145

Figure 4.4. Analysis of TAG amount and fatty acid compositions in pgd1fb12 and pgd1fb16.
WT indicates dw 15.1, the wild-type parental strain for pgd1. A, amount of TAG accumulated
after 48 h of growth in N depleted medium. Data are presented as the ratios of fatty acids
esterified to TAG versus the total fatty acids in the lipid extract. B, percentages of each fatty acid
contained in TAG. Fatty acids are shown as number of carbon: number of double bonds with Δ
(counting from carboxyl group) or ω (the first double bond counting from the methyl group)
indicating the position of double bonds. Trace amount of 16:1 and 16:2 constitute around 1% of
TAG fatty acids and are skipped in the graph. Data shown are averages from two biological
replicates with error bars representing standard deviation.

146

Figure 4.5. Growth curves of pgd1fb16, pgd1 mutant and dw15.1 (WT) in TAP (A) and
TMP media (B). Cells were grown to saturation and inoculated into fresh medium to an optical
density (at 750 nm) of 0.04. Each data point is the average from three technical replicates with
relative standard deviations smaller than 3 %.

147

Figure 4.6. Molecular characterization pgd1fb12. A, Products of SiteFinding PCR with
genomic DNA of several mutants resolved on agarose gel. B, a gene model (5’-3’) of obtained
from

the

Chlamydomonas

genome

database

(http://genome.jgi-

psf.org/Chlre4/Chlre4.home.html). Red boxes and blue boxes represent exons and the
untranslated regions (UTRs) respectively. C, Products of RT-PCR with RNA obtained from cells
grown with nitrogen or cells nitrogen deprived for 48 h as the templates.

148

Chapter 5.
Conclusions and perspectives.

149

The field of algal lipid metabolism has been growing rapidly during the past several
years. While algae of the Nannochloropsis genus produce triacylglycerol up to 50 % of their dry
weight (Radakovits et al., 2012), Chlamydomonas has long been used as a model system for cell
motility and photosynthesis studies (Grossman et al., 2003). Genetic techniques such as nuclear
transformation (Kindle, 1990) and RNA suppression (Molnar et al., 2009) are well developed.
TAG accumulation under nitrogen (N) deprivation was documented in Chlamydomonas
more than a decade ago (Weers and Gulati, 1997) and was recently confirmed by several
independent groups (Wang et al., 2009; Moellering and Benning, 2010). N-deprivation was more
effective than high salinity and high light for induction of TAG production (Fan et al., 2011). To
understand the molecular mechanisms involved in TAG accumulation, several research groups
performed transcript profiling to reveal genes up- or down-regulated by N-deprivation. Many
genes predicted to encode lipid metabolism-related enzymes have been shown to be differentially
regulated (Miller et al., 2010; Boyle et al., 2012). Among these genes were a large number that
encode lipases (Miller et al., 2010). During my Ph.D research, two projects were conducted on
lipase candidates which affect the turnover and biosynthesis of TAG respectively.

CrLIP1 is involved in the regulation of TAG turnover in Chlamydomonas.
The research described in Chapter 2 aimed to discover a lipase involved in TAG
degradation. When N is supplied to N-deprived Chlamydomonas cells, the stored TAG is
degraded, providing precursors for membrane lipid biosynthesis or energy for cell growth
(Nguyen et al., 2011). In yeast, inhibition of TAG turnover by deleting lipases has been
demonstrated to increase TAG content (Athenstaedt and Daum, 2003, 2005) and I was curious
whether this approach could be applied to algae to increase their TAG production.

150

Several putative lipase-encoding genes were heterologously expressed in the yeast lipase
deficient mutant tgl3Δtgl4Δ(Athenstaedt and Daum, 2005; Kurat et al., 2006) for functional
complementation. The candidates were chosen based on the transcriptional regulation by Ndeprivation. Down-regulated candidate genes were thought to be more likely to encode TAG
lipases as lipolysis is expected to be inhibited under N-deprivation. Several up-regulated
candidates were also considered, as it is possible that a lipase protein is produced along with
TAG so it can be rapidly activated through post-translational modifications once N is resupplied. Among five candidate genes screened, over-expression of a down-regulated candidate,
CrLIP1, reversed the hyper-accumulation of TAG in the yeast mutant. RNA suppression of the
CrLIP1 gene in Chlamydomonas delayed the TAG turnover after N re-supply to N-deprived
cultures. However, the TAG accumulation was largely unaffected. This suggests a difference in
the physiology of TAG metabolism between yeast and Chlamydomonas. In yeast, TAG increases
when cells leave log phase and enter stationary phase (Athenstaedt and Daum, 2003). The fact
that tgl3Δtgl4Δ accumulates more TAG than wild-type strains (Athenstaedt and Daum, 2005;
Kurat et al., 2006) indicates that TAG synthesis and degradation probably occur together with
synthesis surpassing degradation. In Chlamydomonas, during N-deprivation, TAG degradation
might be negligible compared with its synthesis. As a result, CrLIP1 knock down lines did not
over-accumulate TAG during N-deprivation but exhibited delayed TAG turnover.
Recombinant CrLIP1 protein was produced and purified from E. coli cells for in vitro
lipase assays. CrLIP1 degraded diacylglycerol (DAG) and polar lipids. However, no TAG
hydrolysis was observed upon CrLIP1 treatment under various reaction conditions. That
observation led to a major unsolved question: is CrLIP1 a DAG lipase or a TAG lipase? As overexpression of CrLIP1 in yeast decreased cellular content of TAG but not DAG, it is possible that

151

CrLIP1 uses TAG as the major substrate but the activity requires an unknown cofactor which
was missing in the in vitro system. Alternatively, the CrLIP1 protein produced in yeast may have
undergone a post-translational modification (e.g. phosphorylation) which did not occur in E. coli
due to the absence of the corresponding mechanism. Future assays should include common
cofactors such as ADP or NAD or should be performed with CrLIP1 protein produced in yeast
cells. To overcome the low abundance of the HA (hemagglutinin) tagged CrLIP1 protein in yeast
strains (not detectible on SDS-PAGE by Coomassie Brilliant Blue staining), it can be enriched
by immunoprecipitation with an anti-HA antibody.
Some other evidence seems to suggest that CrLIP1 might be mainly a DAG lipase as it
was detected in the flagellar proteome of Chlamydomonas (Pazour et al., 2005). The flagellar
membrane is physically connected to the plasma membrane, and is the location for human DAG
lipases (Bisogno et al., 2003). In contrast, TAG lipases in yeast and Arabidopsis were found to
be localized on lipid droplets (Athenstaedt and Daum, 2003, 2005; Eastmond, 2006; Kurat et al.,
2006). For the yeast strain over-expressing CrLIP1, localization can be accomplished by
immunofluorescence by taking advantage of the HA tag. In Chlamydomonas, production of a
GFP fusion protein failed so production of a protein with an epitope tag or generation of an
antibody against CrLIP1 with the recombinant protein from E. coli will be favored.
Even if CrLIP1 is shown to have in vitro TAG lipase activity, it is not likely the major
TAG lipase in Chlamydomonas because in the AmiRNA lines the TAG turnover was slightly
decreased but not abolished. After 36 h of growth in N-containing medium, the TAG contents in
both the knock down lines and the empty vector control were below 3 % of total lipids. If
CrLIP1 is not responsible, then which gene in Chlamydomonas is encoding the major TAG
lipase? In the model seed plant Arabidopsis thaliana, mutation in the lipase-encoding SDP1 gene

152

blocked seed oil liberation and consequently, resulted in slower post-germinative growth
(Eastmond, 2006). A BLAST (Altschul et al., 1997) search of the Chlamydomonas genome
(http://genome.jgi-psf.org/pages/blast.jsf?db=Chlre4) with SDP1 protein sequence showed a
candidate with an E-Value of 2.97E-98. The cDNA of this candidate gene was recently cloned
(Jaruswan Warakanont, unpublished) and future work should focus on biochemical
characterization of this protein and phenotypic analysis of its RNA suppression lines.

Oleate can be exported from the plastid through an acylation/deacylation cycle mediated
by PGD1.
To complement the reverse genetic approach, a large-scale forward genetic screen was
performed to obtain mutants with abnormal TAG amounts. This led to the identification of the
PGD1 gene (described in Chapter 3) which was actually one of the putative lipase-encoding
genes up-regulated under N-deprivation (Miller et al., 2010).
∆9

Disruption of the PGD1 gene lowered the relative amount of oleate (18:1 ) in several
glycerolipids, including the storage lipid TAG and the membrane lipids diacylglycerol-N,N,N,trimethylhomoserine (DGTS) and phosphatidylethanolamine (PtdEtn). All three of these are
likely to be located outside of the plastid. In seed plants PtdEtn has been shown to be absent
from the plastid (Jouhet et al., 2007) and DGTS is presumed to extraplastidic (Giroud and
Eichenberger, 1988). TAG is at least partly extraplastidic in Chlamydomonas as some lipid
droplets clearly exist in the cytosol despite the finding of plastid-located lipid droplets in
starchless mutants of Chlamydomonas (Fan et al., 2011; Goodson et al., 2011). The fact that
these three lipids lack the same fatty acid strongly suggest that export of 18:1

153

∆9

from the plastid

is blocked by the pgd1 mutation. Pulse-chase labeling showed that upon the switch to Ndeprivation, de novo synthesized acyl groups go through the most abundant plastid lipid
monogalactosyldiacylglycerol (MGDG) before being incorporated into TAG. In vitro assays
proved PGD1 to be a lipase which digests the sn-1 position of MGDG, with a preference for de
novo

synthesized

species

∆9

(18:1 /16:0).

These

data

support

the

presence

of

an

∆9

acylation/deacylation cycle acting on lyso-MGDG/MGDG to export 18:1 .
This discovery is an important advance in understanding how the fatty acids are exported
from the plastid for the green lineage organisms. However, due to the complexity of the fatty
acid export process, several major questions remain unsolved. The first question is: is there
another mechanism for exporting fatty acids? There are two major reasons that I suspect other
mechanisms exist to export fatty acids from the plastid. First of all, the acylation/deacylation
cycle proposed applies only to 18:1

∆9

while 16:0 is another major fatty acid exported to the

cytosol (Bates et al., 2007). The activity of PGD1 on the sn-1 position of de novo synthesized
MGDG does not directly facilitate the trafficking of 16:0. Secondly, the PGD1 gene is upregulated by approximately 10 fold under N-deprivation while the vegetative growth under N
∆9

replete conditions should also require the export of 18:1 . Probably there is a basal mechanism
for the export while the acylation/deacylation cycle of MGDG is induced upon stress conditions
that require a higher capacity of fatty acid export. This interpretation also explains the unaffected
vegetative growth of pgd1 mutant.
Details of the acylation/deacylation mechanism also require further study. One immediate
question is where the PGD1 protein is localized. No transit peptide was identified for PGD1 with

154

the TargetP program (http://www.cbs.dtu.dk/services/TargetP/). However, even if PGD1 is
cytosolically localized, it would have access to the MGDG pool as MGDG is present on the
outer envelope, inner envelope as well the thylakoids of a chloroplast (Jouhet et al., 2007).
Similar to the case of CrLIP1, I was unable to create a successful PGD1 cDNA-GFP fusion
protein was not successfully produced. This might be due to a transcriptional silencing
mechanism which is hypothesized to be present in Chlamydomonas to repress exogenous genes
(Neupert et al., 2009). Antibody against PGD1 is being raised with a synthetic PGD1 fragment
and will be tested by my co-workers in the future.
If the proposed model is true, an acyltransferase which uses lyso-MGDG as the acyl
acceptor, acyl-ACP (acyl carier protein) as the acyl donor, and prefers to acylate at the sn-1
position is required to complete the cycle. Such an acyltransferease has been discovered in a
cyanobacterium (Chen et al., 1988). Given the widely accepted endosymbiosis hypothesis that
plastids evolved from an engulfed cyanobacterium (Howe et al., 2008), I expect this
acyltransferase exists in Chlamydomonas as well. However, this enzyme was not purified to
homogeneity and the protein sequence is not currently available. Identification of this
acyltransferase can be accomplished if more mutants with phenotypes similar to those of pgd1
are obtained. Alternatively, this acyltransferase can be identified through biochemical
characterizations of multiple candidate acyltransferases.
∆9

Last but not least, it remains unclear where the exported 18:1 is incorporated into the
DAG backbone for TAG biosynthesis. Different from seed plants, the sn-2 position of TAGs in
Chlamydomonas is almost exclusively C16 fatty acids, which indicates a chloroplast origin of
DAG backbones for TAG synthesis by the diacylglycerol acyltransferase (DGAT) activity (Fan

155

et al., 2011). However, in a starchless mutant, lipid droplets were observed both inside and
outside of the plastid (Fan et al., 2011; Goodson et al., 2011) while in its parental strain only
cytosolic lipid droplets were observed despite that they are also associated with the plastid
envelopes as well as the ER membrane (Goodson et al., 2011). The exact location of TAG
synthesis will await the identifications of the major diacylglycerol acyltransferases.
This work has important implications for the lipid metabolism of land plants. In the
model seed plant Arabidopsis thaliana, people have proposed a similar acylation/deacylation
cycle, with phosphatidylcholine (PtdCho) as the intermediate (Bates et al., 2007). Recombinant
PGD1 protein exhibited activity on both MGDG and PtdCho (unpublished). BLAST (Altschul et
al., 1997) of the PGD1 protein sequence against the Arabidopsis genome (http://arabidopsis.org/)
indicated

three

genes

with

high

similarity

which

are

possibly

involved

in

the

acylation/deacylation cycle on PtdCho. While PtdCho is absent in Chlamydomonas (Riekhof et
al., 2005) both lipids are present in bulk in Arabidopsis (Jouhet et al., 2007). Expressing the
proteins from these three genes in vitro and testing their substrate preference is necessary to
establish their functions. Phenotypic analysis of mutants disrupted in one or more of these three
genes would be a complementary approach to reveal whether they are involved in fatty acid
export in Arabidopsis. Algae accumulate TAG under certain stresses while plants produce oil in
certain tissues such as seeds and pollen (Zhang et al., 2009). Tissue-specific expression databases
will provide important clues about the roles of these genes.

Triacylglycerol accumulation in Chlamydomonas reinhardtii prevents oxidative stress under
N-deprivation

156

TAG accumulation under N-deprivation has been repeatedly reported without extensive
studies of the physiological significance of this phenomenon. The low TAG mutant pgd1 was
used as a tool to understand the consequences of lower TAG accumulation as described in
Chapter 4.
The pgd1 mutant exhibited chlorosis, accumulated thiobarbituric acid reactive substances
(TBARS) and lost viability after 7 days of N-deprivation. These phenotypes could not be
ascribed to altered membrane lipid metabolism as the pgd1 mutant is specifically altered in the
amount of TAG without a major difference from the wild-type parental strain in membrane lipid
composition. Inhibition of photosynthetic electron transport by exposure to 3-(3,4dichlorophenyl)-1,1-dimethylurea (DCMU) (Trebst et al., 1970) in the medium reversed the
phenotypes of the pgd1 mutant. The following model was proposed to explain the observations
above: under N-deprivation, cells stop dividing, and consequently the NADPH generated from
photosynthetic electron transport might be in excess compared to N-replete conditions. This can
lead to the reduction of alternative electron acceptors including molecular oxygen (Asada, 1994).
By synthesizing fatty acids which are highly reduced compared to carbohydrates and proteins,
NADPH can be sequestered to prevent the overreduction of the photosynthetic electron transport
chain.
While the above model is supported by the avoidance of chlorosis, TBARS accumulation
and loss of viability following treatment with DCMU, it can be further corroborated if multiple
independent mutants deficient in TAG production exhibit similar phenotypes. A suppressor
screen has been performed in the pgd1 mutant background for suppression of the chlorosis
phenotype (Chapter 4). Several suppressors that result from second site disruption have been
isolated. Further molecular and physiological characterization of two suppressors is underway. I

157

expect some of the mutants to be affected in aspects of the generation/detoxification of reactive
oxygen species or photosynthesis.
Under N deprivation, Chlamydomonas cells also accumulate a considerable amount of
starch (Work et al., 2010), which might function as another sink for the photosynthetic electron
flux. In mutants that were disabled in starch biosynthesis, sta6 and sta7, TAG was overaccumulated (Fan et al., 2012; Work et al., 2010). Apparently TAG biosynthesis served to
consume the NADPH which could not be allocated to starch biosynthesis. With a low-TAG
mutant pgd1, it is possible that starch accumulation will be enhanced to partly relieve the
overreduction of the photosynthetic electron transport chain, and pgd1 still loses viability
because starch synthesis consumes NADPH less efficiently than fatty acid synthesis on a carbon
number base (Hu et al., 2008). If this hypothesis is correct, a double mutant with both the pgd1
mutation and the sta6 or sta7 mutation will lose viability earlier than the pgd1 mutant.
This project also provides clues about how TAG synthesis can be induced. Conceivably,
instead of being triggered by the nutrient deprivation, the overreduction of the photosynthetic
electron transport chain may be the signal to activate the lipid biosynthesis machinery. These
possibilities can be tested by further pharmacological manipulations of photosynthesis or
forward genetic screens for abnormal regulation on important marker genes such as the DGAT
genes. Answers to these questions will contribute to the development of algae as important
sources of bioenergy.

158

REFERENCE LIST

159

REFERENCE LIST

Altschul, S.F., Madden, T.L., Schaffer, A.A., Zhang, J., Zhang, Z., Miller, W., and Lipman,
D.J. (1997). Gapped BLAST and PSI-BLAST: a new generation of protein database
search programs. Nucleic Acids Res. 25:3389-3402.
Arisz, S.A., Testerink, C., and Munnik, T. (2009). Plant PA signaling via diacylglycerol
kinase. Biochim. Biophys. Acta 1791:869-875.
Asada, K. (1994). Production and action of active oxygen species in photosynthetic tissues. In
Causes of photooxidative stress and amelioration of defence systems in plants, C.F.F.
P.M.Mullineaux, ed (Boca Raton: CRC Press), pp. 77-104.
Aslanidis, C., and de Jong, P.J. (1990). Ligation-independent cloning of PCR products (LICPCR). Nucleic Acids Res. 18:6069-6074.
Athenstaedt, K., and Daum, G. (2003). YMR313c/TGL3 encodes a novel triacylglycerol lipase
located in lipid particles of Saccharomyces cerevisiae. J. Biol. Chem. 278:23317-23323.
Athenstaedt, K., and Daum, G. (2005). Tgl4p and Tgl5p, two triacylglycerol lipases of the
yeast Saccharomyces cerevisiae are localized to lipid particles. J. Biol. Chem.
280:37301-37309.
Athenstaedt, K., Zweytick, D., Jandrositz, A., Kohlwein, S.D., and Daum, G. (1999).
Identification and characterization of major lipid particle proteins of the yeast
Saccharomyces cerevisiae. J Bacteriol 181:6441-6448.
Awai, K., Xu, C., Tamot, B., and Benning, C. (2006). A phosphatidic acid-binding protein of
the chloroplast inner envelope membrane involved in lipid trafficking.
Proc.Natl.Acad.Sci.U.S.A. 103:10817-10822.
Banerji, S., and Flieger, A. (2004). Patatin-like proteins: a new family of lipolytic enzymes
present in bacteria? Microbiology 150:522-525.
Baroli, I., Do, A.D., Yamane, T., and Niyogi, K.K. (2003). Zeaxanthin accumulation in the
absence of a functional xanthophyll cycle protects Chlamydomonas reinhardtii from
photooxidative stress. Plant Cell 15:992-1008.
Baroli, I., Gutman, B.L., Ledford, H.K., Shin, J.W., Chin, B.L., Havaux, M., and Niyogi,
K.K. (2004). Photo-oxidative stress in a xanthophyll-deficient mutant of
Chlamydomonas. J. Biol. Chem. 279:6337-6344.

160

Bates, P. D., Durrett, T. P., Ohlrogge, J. B., and Pollard, M. (2009) Analysis of acyl fluxes
through multiple pathways of triacylglycerol synthesis in developing soybean embryos.
Plant Physiol. 150:55-72.
Bates, P.D., Ohlrogge, J.B., and Pollard, M. (2007). Incorporation of newly synthesized fatty
acids into cytosolic glycerolipids in pea leaves occurs via acyl editing. J.Biol.Chem.
282:31206-31216.
Bates, P.D., Fatihi, A., Snapp, A.R., Carlsson, A.S., Browse, J.A., and Lu, C. (2012). Acyl
Editing and Headgroup Exchange are the Major Mechanisms that Direct Polyunsaturated
Fatty Acid Flux into Triacylglycerols. Plant Physiol. doi: 10.1104/pp.112.204438
Benning, C. (2009) Mechanisms of lipid transport involved in organelle biogenesis in plant
cells. Annu. Rev. Cell Dev. Biol. 25:71-91.
Benning, C., and Somerville, C.R. (1992a). Isolation and genetic complementation of a
sulfolipid-deficient mutant of Rhodobacter sphaeroides. J Bacteriol 174:2352-2360.
Benning, C., and Somerville, C.R. (1992b). Identification of an operon involved in sulfolipid
biosynthesis in Rhodobacter sphaeroides. J Bacteriol 174:6479-6487.
Benning, C., and Ohta, H. (2005). Three enzyme systems for galactoglycerolipid biosynthesis
are coordinately regulated in plants. J. Biol. Chem. 280:2397-2400.
Benning, C., Huang, Z.H., and Gage, D.A. (1995). Accumulation of a novel glycolipid and a
betaine lipid in cells of Rhodobacter sphaeroides grown under phosphate limitation.
Arch. Biochem. Biophys. 317:103-111.
Berthold, P., Schmitt, R., and Mages, W. (2002) An engineered Streptomyces hygroscopicus
aph 7" gene mediates dominant resistance against hygromycin B in Chlamydomonas
reinhardtii. Protist 153:401-412.
Bisogno, T., Howell, F., Williams, G., Minassi, A., Cascio, M.G., Ligresti, A., Matias, I.,
Schiano-Moriello, A., Paul, P., Williams, E.J., Gangadharan, U., Hobbs, C., Di
Marzo, V., and Doherty, P. (2003). Cloning of the first sn1-DAG lipases points to the
spatial and temporal regulation of endocannabinoid signaling in the brain. J Cell Biol
163:463-468.
Bonaventure, G., Schuck, S., and Baldwin, I.T. (2011). Revealing complexity and specificity
in the activation of lipase-mediated oxylipin biosynthesis: a specific role of the Nicotiana
attenuata GLA1 lipase in the activation of jasmonic acid biosynthesis in leaves and roots.
Plant Cell Environ. 34:1507-1520.
Boyle, N.R., Page, M.D., Liu, B., Blaby, I.K., Casero, D., Kropat, J., Cokus, S., HongHermesdorf, A., Shaw, J., Karpowicz, S.J., Gallaher, S., Johnson, S., Benning, C.,
Pellegrini, M., Grossman, A., and Merchant, S.S. (2012). Three acyltransferases and a

161

nitrogen responsive regulator are implicated in nitrogen starvation-induced
triacylglycerol accumulation in Chlamydomonas. J. Biol. Chem. 287: 15811-15825
Bradford, M.M. (1976). A rapid and sensitive method for the quantitation of microgram
quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem.
72:248-254.
Browse, J., Warwick, N., Somerville, C. R., and Slack, C. R. (1986) Fluxes through the
prokaryotic and eukaryotic pathways of lipid synthesis in the '16:3' plant Arabidopsis
thaliana. Biochem. J. 235:25-31.
Castruita, M., Casero, D., Karpowicz, S. J., Kropat, J., Vieler, A., Hsieh, S. I., Yan, W. H.,
Cokus, S., Loo, J. A., Benning, C., Pellegrini, M., and Merchant, S. S. (2011) Systems
biology approach in Chlamydomonas reveals connections between copper nutrition and
multiple metabolic steps. Plant Cell 23:1273-1292.
Chang, C.W., Moseley, J.L., Wykoff, D., and Grossman, A.R. (2005). The LPB1 gene is
important for acclimation of Chlamydomonas reinhardtii to phosphorus and sulfur
deprivation. Plant Physiol. 138:319-329.
Chen, H.H., Wickrema, A., and Jaworski, J.G. (1988). Acyl-acyl-carrier protein:
lysomonogalactosyldiacylglycerol acyltransferase from the cyanobacterium Anabaena
variabilis. Biochim. Biophys. Acta 963, 493-500.
Chen, W., Sommerfeld, M., and Hu, Q. A. (2011) Microwave-assisted Nile red method for in
vivo quantification of neutral lipids in microalgae. Biores. Technol. 102:135-141.
Clausen, M.K., Christiansen, K., Jensen, P.K., and Behnke, O. (1974). Isolation of lipid
particles from baker's yeast. FEBS Lett.43:176-179.
Dahlqvist, A., Stahl, U., Lenman, M., Banas, A., Lee, M., Sandager, L., Ronne, H., and
Stymne, S. (2000). Phospholipid:diacylglycerol acyltransferase: an enzyme that catalyzes
the acyl-CoA-independent formation of triacylglycerol in yeast and plants. Proc. Natl.
Acad. Sci. U. S. A. 97:6487-6492.
Davies, J.P., Yildiz, F.H., and Grossman, A. (1996). Sac1, a putative regulator that is critical
for survival of Chlamydomonas reinhardtii during sulfur deprivation. EMBO J. 15:21502159.
Dormann, P., Balbo, I., and Benning, C. (1999). Arabidopsis galactolipid biosynthesis and
lipid trafficking mediated by DGD1. Science 284:2181-2184.
Durrett, T.P., Benning, C., and Ohlrogge, J. (2008). Plant triacylglycerols as feedstocks for
the production of biofuels. Plant J. 54:593-607.

162

Eastmond, P.J. (2006). SUGAR-DEPENDENT1 encodes a patatin domain triacylglycerol
lipase that initiates storage oil breakdown in germinating Arabidopsis seeds. Plant Cell
18:665-675.
Ebersold, W.T. (1967). Chlamydomonas reinhardi: heterozygous diploid strains. Science
157:447-449.
Eichenberger, W. (1982). Distribution of diacylglyceryl-O-4'-(N,N,N,-trimethyl)homoserine in
different algae. Plant Sci. Lett. 24:91-95.
Eichenberger, W. (1993). Betaine lipids in lower plants. Distribution of DGTS, DGTA and
phospholipids, and the intracellular localization and site of biosynthesis of DGTS. Plant
Physiol. Biochem. 31:213-221.
Ellinger, D., Stingl, N., Kubigsteltig, II, Bals, T., Juenger, M., Pollmann, S., Berger, S.,
Schuenemann, D., and Mueller, M.J. (2010). DONGLE and DEFECTIVE IN
ANTHER DEHISCENCE1 lipases are not essential for wound- and pathogen-induced
jasmonate biosynthesis: redundant lipases contribute to jasmonate formation. Plant
Physiol. 153:114-127.
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 synthesis in Chlamydomonas
reinhardtii. Plant Cell Physiol. 53:1380-1390.
Fang, S.C., de los Reyes, C., and Umen, J.G. (2006). Cell size checkpoint control by the
retinoblastoma tumor suppressor pathway. PLoS Genet 2:e167.
Farese, R.V., Jr., and Walther, T.C. (2009). Lipid droplets finally get a little R-E-S-P-E-C-T.
Cell 139:855-860.
Feinberg, D. (1991). Glycerol Ethers-Synthesis, configuration Analysis, and a brief review of
their lipase-catalyzed reactions. In Synthesis and Chemistry of Agrochemicals II, pp.
269-277.
Fernandez, E., Schnell, R., Ranum, L.P.W., Hussey, S.C., Silflow, C.D., and Lefebvre, P.A.
(1989). Isolation and Characterization of the Nitrate Reductase Structural Gene of
Chlamydomonas reinhardtii. Proc.Natl.Acad.Sci.U.S.A. 86:6449-6453.
Finazzi, G. (2005). The central role of the green alga Chlamydomonas reinhardtii in revealing
the mechanism of state transitions. J Exp Bot 56:383-388.

163

Fischer, B.B., Krieger-Liszkay, A., Hideg, E., Snyrychova, I., Wiesendanger, M., and
Eggen, R.I. (2007). Role of singlet oxygen in chloroplast to nucleus retrograde signaling
in Chlamydomonas reinhardtii. FEBS Lett. 581:5555-5560.
Fischer, W., Heinz, E., and Zeus, M. (1973). The suitability of lipase from Rhizopus arrhizus
delemar for analysis of fatty acid distribution in dihexosyl diglycerides, phospholipids
and plant sulfolipids. Hoppe Seylers Z Physiol. Chem. 354:1115-1123.
Frentzen, M. (1998). Acyltransferases from basic science to modified seed oils. Fett. Lipid
100:161-166.
Gietz, D., St Jean, A., Woods, R.A., and Schiestl, R.H. (1992). Improved method for high
efficiency transformation of intact yeast cells. Nucleic Acids Res. 20:1425.
Giroud, C., and Eichenberger, W. (1988). Fatty-Acids of Chlamydomonas reinhardtii Structure, Positional Distribution and Biosynthesis. Biological Chemistry Hoppe-Seyler
369:18-19.
Giroud, C. and Eichenberger, W. (1989) Lipids of Chlamydomonas reinhardtii - Incorporation
of [C-14] acetate, [C-14] palmitate and [C-14] oleate into different lipids and evidence
for lipid-linked desaturation of fatty acids. Plant Cell Physiol. 30:121-128.
Giroud, C., Gerber, A., and Eichenberger, W. (1988). Lipids of Chlamydomonas reinhardtii Analysis of Molecular-Species and Intracellular Site(S) of Biosynthesis. Plant Cell
Physiol. 29:587-595.
Goodenough, U.W., Armbrust, E.V., Campbell, A.M., and Ferris, P.J. (1995). Annual
Review of Plant Physiology and Plant Molecular Biology. Plant Cell 46:21-44.
Goodson, C., Roth, R., Wang, Z.T., and Goodenough, U. (2011). Structural correlates of
cytoplasmic and chloroplast lipid body synthesis in Chlamydomonas reinhardtii and
stimulation of lipid body production with acetate boost. Eukaryot. Cell 10:1592-1606.
Gronke, S., Mildner, A., Fellert, S., Tennagels, N., Petry, S., Muller, G., Jackle, H., and
Kuhnlein, R.P. (2005). Brummer lipase is an evolutionary conserved fat storage
regulator in Drosophila. Cell Metab 1:323-330.
Grossman, A.R., Harris, E.E., Hauser, C., Lefebvre, P.A., Martinez, D., Rokhsar, D.,
Shrager, J., Silflow, C.D., Stern, D., Vallon, O., and Zhang, Z. (2003).
Chlamydomonas reinhardtii at the crossroads of genomics. Eukaryot. Cell 2:1137-1150.
Harris, E.H. (1989). Chlamydomonas Sourcebook. (San Diego: Academic Press).
Harris, E.H., Stern, D.B., and Witman, G. (2009). The Chlamydomonas Sourcebook:
Introduction to Chlamydomonas and Its Laboratory Use. (San Diego: Academic Press).

164

Heinz, E. and Roughan, G. (1983) Similarities and differences in lipid metabolism of
chloroplasts isolated from 18:3 and 16:3 plants. Plant Physiol. 72:273-279.
Howe, C.J., Barbrook, A.C., Nisbet, R.E., Lockhart, P.J., and Larkum, A.W. (2008). The
origin of plastids. Philos Trans R Soc Lond B Biol Sci 363:2675-2685.
Hu, Q., Sommerfeld, M., Jarvis, E., Ghirardi, M., Posewitz, M., Seibert, M., and Darzins,
A. (2008). Microalgal triacylglycerols as feedstocks for biofuel production: perspectives
and advances. Plant J. 54:621-639.
Huang, A.H. (1996). Oleosins and oil bodies in seeds and other organs. Plant Physiol 110:10551061.
Hyun, Y., Choi, S., Hwang, H.J., Yu, J., Nam, S.J., Ko, J., Park, J.Y., Seo, Y.S., Kim, E.Y.,
Ryu, S.B., Kim, W.T., Lee, Y.H., Kang, H., and Lee, I. (2008). Cooperation and
functional diversification of two closely related galactolipase genes for jasmonate
biosynthesis. Dev. Cell. 14:183-192.
Jako, C., Kumar, A., Wei, Y., Zou, J., Barton, D.L., Giblin, E.M., Covello, P.S., and Taylor,
D.C. (2001). Seed-specific over-expression of an Arabidopsis cDNA encoding a
diacylglycerol acyltransferase enhances seed oil content and seed weight. Plant Physiol.
126:861-874.
James, G.O., Hocart, C.H., Hillier, W., Chen, H., Kordbacheh, F., Price, G.D., and
Djordjevic, M.A. (2011). Fatty acid profiling of Chlamydomonas reinhardtii under
nitrogen deprivation. Bioresour Technol 102:3343-3351.
Jenkins, C.M., Mancuso, D.J., Yan, W., Sims, H.F., Gibson, B., and Gross, R.W. (2004).
Identification, cloning, expression, and purification of three novel human calciumindependent phospholipase A2 family members possessing triacylglycerol lipase and
acylglycerol transacylase activities. J. Biol. Chem. 279:48968-48975.
Jolivet, P., Roux, E., D'Andrea, S., Davanture, M., Negroni, L., Zivy, M., and Chardot, T.
(2004). Protein composition of oil bodies in Arabidopsis thaliana ecotype WS. Plant
Physiol Biochem 42:501-509.
Jouhet, J., Marechal, E., and Block, M.A. (2007). Glycerolipid transfer for the building of
membranes in plant cells. Prog. Lipid Res. 46:37-55.
Joyard, J., Block, M. A., and Douce, R. (1991) Molecular aspects of plastid envelope
biochemistry. Eur. J. Biochem. 199:489-509.
Keeling, P.J. (2004). Diversity and evolutionary history of plastids and their hosts. American J.
Bot. 91:1481-1493.

165

Kim, J.Y., Tillison, K., Lee, J.H., Rearick, D.A., and Smas, C.M. (2006). The adipose tissue
triglyceride lipase ATGL/PNPLA2 is downregulated by insulin and TNF-alpha in 3T3L1 adipocytes and is a target for transactivation by PPARgamma. Am J Physiol
Endocrinol Metab 291:E115-127.
Kimura, K., Yamaoka, M., and Kamisaka, Y. (2004) Rapid estimation of lipids in oleaginous
fungi and yeasts using Nile red fluorescence. J. Microbiol. Meth. 56:331-338.
Kindle, K.L. (1990). High-frequency nuclear transformation of Chlamydomonas reinhardtii.
Proc. Natl. Acad. Sci. U. S. A. 87:1228-1232.
Kito, M., Aibara, S., Kato, M., and Hata, T. (1972). Differences in fatty acid composition
among phosphatidylethanolamine, phosphatidylglycerol and cardiolipin of Escherichia
coli. Biochim. Biophys. Acta 260:475-478.
Klug, R.M., and Benning, C. (2001). Two enzymes of diacylglyceryl-0-4 '-(N,N,N,trimethyl)homoserine biosynthesis are encoded by btaA and btaB in the purple bacterium
Rhodobacter sphaeroides. Proc. Natl. Acad. Sci. U. S. A. 98:5910-5915.
Krahmer, N., Guo, Y., Wilfling, F., Hilger, M., Lingrell, S., Heger, K., Newman, H.W.,
Schmidt-Supprian, M., Vance, D.E., Mann, M., Farese, R.V., Jr., and Walther, T.C.
(2011). Phosphatidylcholine synthesis for lipid droplet expansion is mediated by
localized activation of CTP:phosphocholine cytidylyltransferase. Cell Metab 14:504-515.
Kuo, M.H., Zhou, J., Jambeck, P., Churchill, M.E., and Allis, C.D. (1998). Histone
acetyltransferase activity of yeast Gcn5p is required for the activation of target genes in
vivo. Genes Dev 12:627-639.
Kunst, L., Browse, J., and Somerville, C. (1988) Altered regulation of lipid biosynthesis in a
mutant of Arabidopsis deficient in chloroplast glycerol-3-phosphate acyltransferase
activity. Proc. Natl. Acad. Sci. U. S. A. 85:4143-4147.
Kurat, C.F., Natter, K., Petschnigg, J., Wolinski, H., Scheuringer, K., Scholz, H.,
Zimmermann, R., Leber, R., Zechner, R., and Kohlwein, S.D. (2006). Obese yeast:
triglyceride lipolysis is functionally conserved from mammals to yeast. J. Biol. Chem.
281:491-500.
Kuusi, T., Nikkila, E.A., Taskinen, M.R., Somerharju, P., and Ehnholm, C. (1982). Human
postheparin plasma hepatic lipase activity against triacylglycerol and phospholipid
substrates. Clin Chim Acta 122:39-45.
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 epub ahead of print.

166

Larkin, M.A., Blackshields, G., Brown, N.P., Chenna, R., McGettigan, P.A., McWilliam,
H., Valentin, F., Wallace, I.M., Wilm, A., Lopez, R., Thompson, J.D., Gibson, T.J.,
and Higgins, D.G. (2007). Clustal W and Clustal X version 2.0. Bioinformatics 23:29472948.
Leber, R., Zinser, E., Zellnig, G., Paltauf, F., and Daum, G. (1994). Characterization of lipid
particles of the yeast, Saccharomyces cerevisiae. Yeast 10:1421-1428.
Liu, B., and Benning, C. (2012). Lipid metabolism in microalgae distinguishes itself. Curr.
Opin. Biotechnol. Epub ahead of print.
Livak, K.J., and Schmittgen, T.D. (2001). Analysis of relative gene expression data using realtime quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25:402-408.
Lu, B. and Benning, C. (2009) A 25-amino acid sequence of the Arabidopsis TGD2 protein is
sufficient for specific binding of phosphatidic acid. J. Biol. Chem. 284:17420-17427.
Lu, B., Xu, C., Awai, K., Jones, A.D., and Benning, C. (2007). A small ATPase protein of
Arabidopsis, TGD3, involved in chloroplast lipid import. J.Biol.Chem. 282:3594535953.
Lu, C., Xin, Z., Ren, Z., Miquel, M., and Browse, J. (2009). An enzyme regulating
triacylglycerol composition is encoded by the ROD1 gene of Arabidopsis.
Proc.Natl.Acad.Sci.U.S.A. 106:18837-18842.
Lu, Y., Savage, L. J., Ajjawi, I., Imre, K. M., Yoder, D. W., Benning, C., Dellapenna, D.,
Ohlrogge, J. B., Osteryoung, K. W., Weber, A. P., Wilkerson, C. G., and Last, R. L.
(2008) New connections across pathways and cellular processes: industrialized mutant
screening reveals novel associations between diverse phenotypes in Arabidopsis.Plant
Physiol. 146:1482-1500.
Luo, J., Xu, X., Hall, H., Hyland, E.M., Boeke, J.D., Hazbun, T., and Kuo, M.H. (2010a).
Histone h3 exerts a key function in mitotic checkpoint control. Mol Cell Biol 30:537549.
Luo, J., Xu, X., Hall, H., Hyland, E.M., Boeke, J.D., Hazbun, T., and Kuo, M.H. (2010b).
Histone h3 exerts a key function in mitotic checkpoint control. Mol Cell Biol 30:537549.
Ma, H., Kunes, S., Schatz, P.J., and Botstein, D. (1987). Plasmid construction by homologous
recombination in yeast. Gene 58:201-216.
Martin, B.A., and Wilson, R.F. (1984). Subcellular localization of TAG synthesis in spinach
leaves. Lipids 19:117-121.

167

McVittie, A. (1972). Flagellum mutants of Chlamydomonas reinhardii. J Gen Microbiol 71:525540.
Mehler, A.H. (1951). Studies on Reactions of Illuminated Chloroplasts .2. Stimulation and
Inhibition of the Reaction with Molecular Oxygen. Arch. Biochem. Biophys. 34:339-351.
Merchant, S.S., Prochnik, S.E., Vallon, O., Harris, E.H., Karpowicz, S.J., Witman, G.B.,
Terry, A., Salamov, A., Fritz-Laylin, L.K., Marechal-Drouard, L., Marshall, W.F.,
Qu, L.H., Nelson, D.R., Sanderfoot, A.A., Spalding, M.H., Kapitonov, V.V., Ren, Q.,
Ferris, P., Lindquist, E., Shapiro, H., Lucas, S.M., Grimwood, J., Schmutz, J.,
Cardol, P., Cerutti, H., Chanfreau, G., Chen, C.L., Cognat, V., Croft, M.T., Dent,
R., Dutcher, S., Fernandez, E., Fukuzawa, H., Gonzalez-Ballester, D., GonzalezHalphen, D., Hallmann, A., Hanikenne, M., Hippler, M., Inwood, W., Jabbari, K.,
Kalanon, M., Kuras, R., Lefebvre, P.A., Lemaire, S.D., Lobanov, A.V., Lohr, M.,
Manuell, A., Meier, I., Mets, L., Mittag, M., Mittelmeier, T., Moroney, J.V.,
Moseley, J., Napoli, C., Nedelcu, A.M., Niyogi, K., Novoselov, S.V., Paulsen, I.T.,
Pazour, G., Purton, S., Ral, J.P., Riano-Pachon, D.M., Riekhof, W., Rymarquis, L.,
Schroda, M., Stern, D., Umen, J., Willows, R., Wilson, N., Zimmer, S.L., Allmer, J.,
Balk, J., Bisova, K., Chen, C.J., Elias, M., Gendler, K., Hauser, C., Lamb, M.R.,
Ledford, H., Long, J.C., Minagawa, J., Page, M.D., Pan, J., Pootakham, W., Roje,
S., Rose, A., Stahlberg, E., Terauchi, A.M., Yang, P., Ball, S., Bowler, C.,
Dieckmann, C.L., Gladyshev, V.N., Green, P., Jorgensen, R., Mayfield, S., MuellerRoeber, B., Rajamani, S., Sayre, R.T., Brokstein, P., Dubchak, I., Goodstein, D.,
Hornick, L., Huang, Y.W., Jhaveri, J., Luo, Y., Martinez, D., Ngau, W.C., Otillar,
B., Poliakov, A., Porter, A., Szajkowski, L., Werner, G., Zhou, K., Grigoriev, I.V.,
Rokhsar, D.S., and Grossman, A.R. (2007). The Chlamydomonas genome reveals the
evolution of key animal and plant functions. Science 318:245-250.
Miege, C., and Marechal, E. (1999). 1,2-sn-Diacylglycerol in plant cells: Product, substrate and
regulator. Plant. Physiol. Biochem. 37:795-808.
Miller, R., Wu, G., Deshpande, R.R., Vieler, A., Gartner, K., Li, X., Moellering, E.R.,
Zauner, S., Cornish, A.J., Liu, B., Bullard, B., Sears, B.B., Kuo, M.H., Hegg, E.L.,
Shachar-Hill, Y., Shiu, S.H., and Benning, C. (2010). Changes in transcript abundance
in Chlamydomonas reinhardtii following nitrogen deprivation predict 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., Muthan, B., and Benning, C. (2010). Freezing tolerance in plants requires
lipid remodeling at the outer chloroplast membrane. Science 330:226-228.
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.

168

Napier, J.A., Michaelson, L.V., and Stobart, A.K. (1999). Plant desaturases: harvesting the fat
of the land. Curr. Opin. Plant Biol. 2:123-127.
Neupert, J., Karcher, D., and Bock, R. (2009). Generation of Chlamydomonas strains that
efficiently express nuclear transgenes. Plant J. 57:1140-1150.
Newman, S.M., Boynton, J.E., Gillham, N.W., Randolphanderson, B.L., Johnson, A.M.,
and Harris, E.H. (1990). Transformation of Chloroplast Ribosomal-Rna Genes in
Chlamydomonas - Molecular and Genetic-Characterization of Integration Events.
Genetics 126:875-888.
Nguyen, H.M., Baudet, M., Cuine, S., Adriano, J.M., Barthe, D., Billon, E., Bruley, C.,
Beisson, F., Peltier, G., Ferro, M., and Li-Beisson, Y. (2011). Proteomic profiling of
oil bodies isolated from the unicellular green microalga Chlamydomonas reinhardtii:
with focus on proteins involved in lipid metabolism. Proteomics 11:4266-4273.
Nguyen, H. T., Mishra, G., Whittle, E., Pidkowich, M. S., Bevan, S. A., Merlo, A. O.,
Walsh, T. A., and Shanklin, J. (2010) Metabolic engineering of seeds can achieve
levels of omega-7 fatty acids comparable with the highest levels found in natural plant
sources. Plant Physiol.154:1897-1904.
Niyogi, K.K. (1999). PHOTOPROTECTION REVISITED: Genetic and Molecular Approaches.
Annu Rev Plant Physiol Plant Mol Biol 50:333-359.
Oelkers, P., Behari, A., Cromley, D., Billheimer, J.T., and Sturley, S.L. (1998).
Characterization of two human genes encoding acyl coenzyme A:cholesterol
acyltransferase-related enzymes. J. Biol. Chem.273:26765-26771.
Ohlrogge, J., and Browse, J. (1995). Lipid biosynthesis. Plant Cell 7:957-970.
Ohlrogge, J.B., Kuhn, D.N., and Stumpf, P.K. (1979). Subcellular localization of acyl carrier
protein in leaf protoplasts of Spinacia oleracea. Proc. Natl. Acad. Sci. U. S. A. 76:11941198.
Pazour, G.J., Agrin, N., Leszyk, J., and Witman, G.B. (2005). Proteomic analysis of a
eukaryotic cilium. J Cell Biol.170:103-113.
Pruvost, J., Van Vooren, G., Cogne, G., and Legrand, J. (2009). Investigation of biomass and
lipids production with Neochloris oleoabundans in photobioreactor. Bioresour Technol
100:5988-5995.
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.

169

Rajakumari, S., and Daum, G. (2010). Multiple functions as lipase, steryl ester hydrolase,
phospholipase, and acyltransferase of Tgl4p from the yeast Saccharomyces cerevisiae. J.
Biol. Chem. 285:15769-15776.
Rani, S.H., Krishna, T.H., Saha, S., Negi, A.S., and Rajasekharan, R. (2010). Defective in
cuticular ridges (DCR) of Arabidopsis thaliana, a gene associated with surface cutin
formation, encodes a soluble diacylglycerol acyltransferase. J. Biol. Chem. 285:3833738347.
Riekhof, W.R., Sears, B.B., and Benning, C. (2005). Annotation of genes involved in
glycerolipid biosynthesis in Chlamydomonas reinhardtii: discovery of the betaine lipid
synthase BTA1Cr. Eukaryot. Cell 4:242-252.
Riekhof, W.R., Ruckle, M.E., Lydic, T.A., Sears, B.B., and Benning, C. (2003). The
sulfolipids 2 '-O-acyl-sulfoquinovosyldiacylglycerol and sulfoquinovosyldiacylglycerol
are absent from a Chlamydomonas reinhardtii mutant deleted in SQD1. Plant Physiol.
133:864-874.
Rietveld, A.G., Killian, J.A., Dowhan, W., and de Kruijff, B. (1993). Polymorphic regulation
of membrane phospholipid composition in Escherichia coli. J. Biol. Chem. 268:1242712433.
Rigaut, G., Shevchenko, A., Rutz, B., Wilm, M., Mann, M., and Seraphin, B. (1999). A
generic protein purification method for protein complex characterization and proteome
exploration. Nat Biotechnol 17:1030-1032.
Robert, G., Melchiorre, M., Racca, R., Trippi, V., and Lascano, H.R. (2009). Apoplastic
superoxide level in wheat protoplast under photooxidative stress is regulated by
chloroplast redox signals: Effects on the antioxidant system. Plant Science 177:168-174.
Rochaix, J.D. (1995). Chlamydomonas reinhardtii as the photosynthetic yeast. Annu Rev Genet
29:209-230.
Roessler, P.G. (1988). Effects of Silicon Deficiency on Lipid-Composition and Metabolism in
the Diatom Cyclotella-Cryptica. Journal of Phycology 24:394-400.
Rosenbaum, J.L., and Witman, G.B. (2002). Intraflagellar transport. Nat Rev Mol Cell Biol
3:813-825.
Rossak, M., Schafer, A., Xu, N., Gage, D.A., and Benning, C. (1997). Accumulation of
sulfoquinovosyl-1-O-dihydroxyacetone in a sulfolipid-deficient mutant of Rhodobacter
sphaeroides inactivated in sqdC. Annu. Rev. Plant Physiol. Plant Mol. Biol. 340:219230.
Roughan, P.G., and Slack, C.R. (1982). Cellular-Organization of Glycerolipid Metabolism.
Annual Review of Plant Physiology and Plant Molecular Biology 33:97-132.

170

Roughan, P.G., Holland, R., and Slack, C.R. (1980). The role of chloroplasts and microsomal
fractions in polar-lipid synthesis from [1-14C]acetate by cell-free preparations from
spinach (Spinacia oleracea) leaves. Biochem J 188:17-24.
Rousseeuw, P. J. and Croux, C. (1993) Alternatives to the median absolute deviation. Am. Stat.
Assoc. 88:1273-1283.
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 Physiol. 141:1533-1543.
Sato, N., and Murata, N. (1991). Transition of lipid phase in aqueous dispersions of
diacylglyceryltrimethylhomoserine. Biochim. Biophys. Acta 1082:108-111.
Schneider, J.C., and Roessler, P. (1994). Radiolabeling studies of lipids and fatty acids in
nannochloropsis (eustigmatophyceae), an oleaginous marine alga. J. Phycol. 30:594-598.
Shen, S., Wu, J., Srivatsan, S., Gorentla, B.K., Shin, J., Xu, L., and Zhong, X.P. (2011).
Tight regulation of diacylglycerol-mediated signaling is critical for proper invariant NKT
cell development. J Immunol 187:2122-2129.
Sherman, F. (1991). Getting started with yeast. Methods Enzymol 194:3-21.
Shimojima, M., Ohta, H., Iwamatsu, A., Masuda, T., Shioi, Y., and Takamiya, K. (1997)
Cloning of the gene for monogalactosyldiacylglycerol synthase and its evolutionary
origin. Proc. Natl. Acad. Sci. U. S. A. 94:333-337.
Shulaev, V., and Oliver, D.J. (2006). Metabolic and proteomic markers for oxidative stress.
New tools for reactive oxygen species research. Plant Physiol. 141:367-372.
Siaut, M., Cuine, S., Cagnon, C., Fessler, B., Nguyen, M., Carrier, P., Beyly, A., Beisson, F.,
Triantaphylides, C., Li-Beisson, Y., and Peltier, G. (2011). Oil accumulation in the
model green alga Chlamydomonas reinhardtii: characterization, variability between
common laboratory strains and relationship with starch reserves. BMC Biotechnol 11:7.
Siebertz, H. P. and Heinz, E. (1977) Labelling experiments on the origin of hexa- and
octadectrienoic acids in galactolipids from leaves. Z. Naturforsch. 32c:193-205.
Silflow, C.D., and Lefebvre, P.A. (2001). Assembly and motility of eukaryotic cilia and
flagella. Lessons from Chlamydomonas reinhardtii. Plant Physiol. 127:1500-1507.
Slavin, B.G., Ong, J.M., and Kern, P.A. (1994). Hormonal regulation of hormone-sensitive
lipase activity and mRNA levels in isolated rat adipocytes. J Lipid Res 35:1535-1541.

171

Stahl, U., Stalberg, K., Stymne, S., and Ronne, H. (2008). A family of eukaryotic
lysophospholipid acyltransferases with broad specificity. FEBS Lett. 582:305-309.
Suen, Y., Hubbard, J.S., Holzer, G., and Tornabene, T.G. (1987). Total Lipid Production of
the Green-Alga Nannochloropsis Sp Qii under Different Nitrogen Regimes. J. Phycol.
23:289-296.
Tan, G.H., Gao, Y., Shi, M., Zhang, X.Y., He, S.P., Cheng, Z.L., and An, C.C. (2005).
SiteFinding-PCR: a simple and efficient PCR method for chromosome walking. Nucleic
Acids Res. 33, e122.
Tauchi-Sato, K., Ozeki, S., Houjou, T., Taguchi, R., and Fujimoto, T. (2002). The surface of
lipid droplets is a phospholipid monolayer with a unique Fatty Acid composition. J. Biol.
Chem. 277:44507-44512.
Trebst, A., Harth, E., and Draber, W. (1970). On A New Inhibitor of Photosynthetic ElectronTransport in Isolated Chloroplasts. Zeitschrift fur Naturforschung Part B-Chemie
Biochemie Biophysik Biologie und Verwandten Gebiete B 25:1157-1159.
Utsugi, A., Kanda, A., and Hara, S. (2009). Lipase specificity in the transacylation of
triacylglycerin. Journal of Oleo Science 58:123-132.
Van Wagenen, J., Miller, T.W., Hobbs, S., Hook, P., Crowe, B., and Huesemann, M. (2012).
Effects of Light and Temperature on Fatty Acid Production in Nannochloropsis Salina.
Energies 5:731-740.
Vandeene, L.L., and Dehaas, G.H. (1966). Phosphoglycerides and Phospholipases. Annu. Rev.
Biochem. 35:157-194.
Vaughan, M., Berger, J.E., and Steinberg, D. (1964). Hormone-Sensitive Lipase and
Monoglyceride Lipase Activities in Adipose Tissue. J. Biol. Chem. 239:401-409.
Vogel,

G., and Browse, J. (1996). Cholinephosphotransferase and Diacylglycerol
Acyltransferase (Substrate Specificities at a Key Branch Point in Seed Lipid
Metabolism). Plant Physiol. 110:923-931.

Wach, A., Brachat, A., Pohlmann, R., and Philippsen, P. (1994). New heterologous modules
for classical or PCR-based gene disruptions in Saccharomyces cerevisiae. Yeast 10:17931808.
Walther, T.C., and Farese, R.V. (2009). The life of lipid droplets. Biochimica. Biophysica.
Acta-Mol. Cell Biol. Lipids 1791:459-466.
Wang, X. (2001). Plant Phospholipases. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52:211-231.

172

Wang, Z.T., Ullrich, N., Joo, S., Waffenschmidt, S., and Goodenough, U. (2009a). Algal lipid
bodies: stress induction, purification, and biochemical characterization in wild-type and
starchless Chlamydomonas reinhardtii. Eukaryot. Cell 8:1856-1868.
Wang, Z.T., Ullrich, N., Joo, S., Waffenschmidt, S., and Goodenough, U. (2009b). Algal
lipid bodies: stress induction, purification, and biochemical characterization in wild-type
and starchless Chlamydomonas reinhardtii. Eukaryot. Cell 8:1856-1868.
Weers, P.M.M., and Gulati, R.D. (1997). Growth and reproduction of Daphnia galeata in
response to changes in fatty acids, phosphorus, and nitrogen in Chlamydomonas
reinhardtii. Limnology and Oceanography 42:1584-1589.
Wen, F., Xing, D., and Zhang, L.R. (2008). Hydrogen peroxide is involved in high blue lightinduced chloroplast avoidance movements in Arabidopsis. J Exp Bot 59:2891-2901.
Weyer, K.M., Bush, D.R., Darzins, A., and Willson, B.D. (2010). Theoretical maximum algal
oil production. Bioenergy Res 3, 204-213.
Wise, R.R., and Hoober, J.K. (2006). The Diversity of Plastid Form and Function.
(Netherlands: Springer).
Work, V.H., Radakovits, R., Jinkerson, R.E., Meuser, J.E., Elliott, L.G., Vinyard, D.J.,
Laurens, L.M., Dismukes, G.C., and Posewitz, M.C. (2010). Increased lipid
accumulation in the Chlamydomonas reinhardtii sta7-10 starchless isoamylase mutant
and increased carbohydrate synthesis in complemented strains. Eukaryot. Cell 9:12511261.
Xu, C., Fan, J., Cornish, A.J., and Benning, C. (2008). Lipid trafficking between the
endoplasmic reticulum and the plastid in Arabidopsis requires the extraplastidic TGD4
protein. Plant Cell 20:2190-2204.
Xu, C., Fan, J., Riekhof, W., Froehlich, J.E., and Benning, C. (2003). A permease-like
protein involved in ER to thylakoid lipid transfer in Arabidopsis. EMBO J. 22:23702379.
Xu, C., Yu, B., Cornish, A.J., Froehlich, J.E., and Benning, C. (2006). Phosphatidylglycerol
biosynthesis in chloroplasts of Arabidopsis mutants deficient in acyl-ACP glycerol-3phosphate acyltransferase. Plant J. 47:296-309.
Yu, B., Xu, C., and Benning, C. (2002). Arabidopsis disrupted in SQD2 encoding sulfolipid
synthase is impaired in phosphate-limited growth. Proc. Natl. Acad. Sci. U. S. A.
99:5732-5737.
Zabawinski, C., Van Den Koornhuyse, N., D'Hulst, C., Schlichting, R., Giersch, C., Delrue,
B., Lacroix, J.M., Preiss, J., and Ball, S. (2001). Starchless mutants of Chlamydomonas

173

reinhardtii lack the small subunit of a heterotetrameric ADP-glucose pyrophosphorylase.
J Bacteriol 183:1069-1077.
Zechner, R., Kienesberger, P.C., Haemmerle, G., Zimmermann, R., and Lass, A. (2009).
Adipose triglyceride lipase and the lipolytic catabolism of cellular fat stores. J Lipid Res
50:3-21.
Zhang, M., Fan, J., Taylor, D.C., and Ohlrogge, J.B. (2009). DGAT1 and PDAT1
acyltransferases have overlapping functions in Arabidopsis triacylglycerol biosynthesis
and are essential for normal pollen and seed development. Plant Cell 21:3885-3901.
Zieger, R., and Egle, K. (1965). Zur quantitativen Analyse der Chloroplasten pigmente. I.
Kritische
Überprüfung
der
spektralphotometrischen
Chlorophyllbestimmung.
Beitr.Biol.Pflanz 41:11-37.
Zimmermann, R., Lass, A., Haemmerle, G., and Zechner, R. (2009). Fate of fat: the role of
adipose triglyceride lipase in lipolysis. Biochim. Biophys. Acta 1791:494-500.
Zimmermann, R., Strauss, J.G., Haemmerle, G., Schoiswohl, G., Birner-Gruenberger, R.,
Riederer, M., Lass, A., Neuberger, G., Eisenhaber, F., Hermetter, A., and Zechner,
R. (2004). Fat mobilization in adipose tissue is promoted by adipose triglyceride lipase.
Science 306:1383-1386.

174