Content uploaded by Gerard Charles Dismukes
Author content
All content in this area was uploaded by Gerard Charles Dismukes on Dec 21, 2017
Content may be subject to copyright.
Regulatory branch points affecting protein and lipid
biosynthesis in the diatom Phaeodactylum
tricornutum
L. Tiago Guerra
a,b
, Orly Levitan
c
, Miguel J. Frada
c,1
, Jennifer S. Sun
a
,
Paul G. Falkowski
c
, G. Charles Dismukes
a,
*
a
Waksman Institute of Microbiology, Rutgers, The State University of New Jersey, 190 Frelinghuysen rd.,
Piscataway, NJ 08854, USA
b
Department of Chemistry, Princeton University, Princeton, NJ 08544, USA
c
Environmental Biophysics and Molecular Ecology Program, Institute of Marine and Coastal Sciences, Rutgers,
The State University of New Jersey, New Brunswick, NJ 08901, USA
article info
Article history:
Received 18 March 2013
Received in revised form
14 September 2013
Accepted 2 October 2013
Available online xxx
Keywords:
Microalgae
Biofuels
Nitrogen metabolism
GS/GOGAT
Lipid
Carbon partition
abstract
It is widely established that nutritional nitrogen deprivation increases lipid accumulation
but severely decreases growth rate in microalgae. To understand the regulatory branch
points that determine the partitioning of carbon among its potential sinks, we analyzed
metabolite and transcript levels of central carbon metabolic pathways and determined the
average fluxes and quantum requirements for the synthesis of protein, carbohydrates and
fatty acid in the diatom Phaeodactylum tricornutum. Under nitrate-starved conditions, the
carbon fluxes into all major sinks decrease sharply; the largest decrease was into proteins
and smallest was into lipids. This reduction of carbon flux into lipids together with a
significantly lower growth rate is responsible for lower overall FA productivities implying
that nitrogen starvation is not a bioenergetically feasible strategy for increasing biodiesel
production. The reduction in these fluxes was accompanied by an 18-fold increase in
a-ketoglutarate (AKG), 3-fold increase in NADPH/NADP
þ
, and sharp decreases in glutamate
(GLU) and glutamine (GLN) levels. Additionally, the mRNA level of acetyl-CoA carboxylase
and two type II diacylglycerol-acyltransferases were increased. Partial suppression of ni-
trate reductase by tungstate resulted in similar trends at lower levels as for nitrate star-
vation. These results reveal that the GS/GOGAT pathway is the main regulation site for
nitrate dependent control of carbon partitioning between protein and lipid biosynthesis,
while the AKG/GL(N/U) metabolite ratio is a transcriptional signal, possibly related to redox
poise of intermediates in the photosynthetic electron transport system.
ª2013 Elsevier Ltd. All rights reserved.
Abbreviations: GS, glutamine synthetase; GOGAT, glutamine oxoglutarate aminotransferase; GDH, glutamate dehydrogenase; GLN,
glutamine; GLU, glutamate; AKG, a-ketoglutarate.
*Corresponding author. Tel.: þ1 848 445 6786; fax: þ848 445 5735.
E-mail address: dismukes@rci.rutgers.edu (G.C. Dismukes).
1
Current address: Department of Plant Sciences, Weizmann Institute of Science, P.O. B. 26, Rehovot 76100, Israel.
Available online at www.sciencedirect.com
http://www.elsevier.com/locate/biombioe
biomass and bioenergy xxx (2013) 1e10
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
0961-9534/$ esee front matter ª2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.biombioe.2013.10.007
1. Introduction
The appropriation of photosynthetically fixed carbon in
microalgae is strongly affected by environmental conditions.
Under optimal growth conditions, most fixed carbon is allo-
cated to proteins [1]. In contrast, under suboptimal conditions,
especially nitrogen limitation, cells divert most of the fixed
carbon into nitrogen deficient compounds such as lipids and/
or carbohydrates, at the expense of growth [2]. The tradeoff
between growth rate and lipid content has been one of the
biggest factors preventing the industrial production of
algal biodiesel at prices competitive with fossil sources [3].
Hence, considerable effort has gone into understanding the
molecular mechanisms that determine the fate of photosyn-
thetically fixed carbon under stress conditions [2]. In this
paper we use a metabolic analysis to identify the regulatory
branch points in the ‘carbon decision tree’ under nitrogen
limitation in a marine diatom.
Inspection of the intermediate metabolism of eukaryotic
algae (Fig. 1) reveals several potential regulatory branch points
that could control the fate of carbon. Pyruvate (Pyr) is the first
key intermediate in this process. This 3-carbon molecule can
take several alternative paths in the metabolic network which
will determine if the carbon is deposited into amino or fatty
acids (Fig. 1). Pyruvate can be directly aminated to generate
alanine and thus proteins (denoted as step (1) in Fig. 1). More
commonly however, pyruvate undergoes decarboxylation to
Fig. 1 eModel of metabolic pathways and branching points affecting lipid biosynthesis in P. tricornutum under nitrogen
starvation. Relevant branching points are signaled in parenthesis and described in the introduction. Metabolites and
enzymatic steps are color coded according to the fold change of metabolite abundance or mRNA in the eNOL
3condition in
comparison to the control according to the legend. The names of the enzymes responsible for relevant reactions are given
next to the corresponding step. Multistep biosynthetic pathways are noted by broken arrows. For TAG biosynthesis the
arrow color refers to the DGAT step only. The chrysolaminarin synthesis pathway may happen in the cytosol and/or in the
chloroplast. Here it is represented only the cytosol pathway for simplicity. The GS/GOGAT cycle proteins are predicted to be
localized in the mitochondrion, cytosol or plastid. Here they were represented only in the mitochondrion and cytosol for
simplicity. Gene abbreviations are given in Table S1 and metabolite abbreviations are given in the legend of Fig. 2.PSe
Photosystem, FD eferredoxin.
biomass and bioenergy xxx (2013) 1e102
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
form acetyl-CoA (AcCoA), (denoted as step (2) in Fig. 1). AcCoA,
the second key intermediate, can either undergo successive
oxidation in the TCA cycle or be used by acetyl-CoA carbox-
ylase (ACCase) to form malonyl-CoA (MaCoA), which is the
first committed intermediate in the synthesis of lipids [4].Ifit
enters the TCA cycle it produces a-ketoglutarate (AKG), which
can be used for assimilation of ammonium via the glutamine
synthetase/glutamine oxoglutarate aminotransferase (GS/
GOGAT) cycle, yielding glutamine (GLN) and glutamate (GLU),
(step (3) in Fig. 1)[5]. GLU, the third intermediate, can also be
produced by the NAD(P)H-dependent amination of AKG via
the reversible glutamate dehydrogenase (GDH), step (4) in
Fig. 1. However the contribution of this pathway for assimi-
lation of ammonium was shown to be minor in relation to the
GS/GOGAT cycle as its Michaelis constant K
m
for ammonium
is 1000-fold higher than GS’s [5]. Thus, AKG is an essential
carbon source for the synthesis of glutamine (GLN) and
glutamate (GLU), from which many other amino acids and
cofactors are produced. Finally, two anapleurotic pathways
exist to replenish AKG in the TCA cycle: pyruvate may un-
dergo carboxylation to form oxaloacetate via step (5) in Fig. 1,
or AKG can be derived from the deamination of glutamic acid
by the catabolism of proteins via GDH in step (4). In conclu-
sion, there are at least five key carbon branch points that can
influence the supply of intermediates, and hence the fate of
carbon, all of which intersect in a small part of the metabolic
processes centered on pyruvate and the TCA cycle.
In eukaryotic microalgae, nitrate reductase (NR) is
responsible for the NAD(P)H dependent reduction of nitrate to
nitrite, and is sequentially coupled to ferredoxin-dependent
nitrite reductase (NiR) for production of ammonium which is
used in amino acid and protein synthesis (Fig. 1)[6]. Hence,
coordinated regulation of NR by several factors including the
cellular redox poise provides the initial mechanism for con-
trolling the conversion of intracellular nitrate for downstream
biosynthesis reactions [7]. The cellular redox poise has also
been implicated in controlling pathways of carbon meta-
bolism such as lipid biosynthesis [8].
Although phylogenetically distant from diatoms, cyano-
bacteria use AKG and GLN or GLU to report the C:N status of
the cell in real time [9,10]. The levels of these metabolites are
sensed by NtcA, a global nitrogen regulator that can function
as both a transcriptional activator and a repressor. For
instance, NtcA binds and activates the promoters of glnA
(encoding glutamine synthetase), the nir operon (nitrate up-
take) and icd (isocitrate dehydrogenase) while it also represses
transcription of gifA and gifB (encoding inactivating factors of
glutamine synthetase) [10].
Based on the logic of intermediate metabolism of acetyl-
CoA, Dunahay and coworkers (1996) tried to increase carbon
flow towards lipid biosynthesis by increasing the activity of
ACCase. In the first genetic transformation of a diatom, these
researchers overexpressed ACCase in Cyclotella cryptica [11].
The transformant strain had increased ACCase activity, but
the phenotype did not show any significant change in lipid
content [2]. Similar results were achieved when ACCase of
Arabidopsis thaliana was expressed in Brassica napus and Sym-
phytum tuberosum [12,13]. These results indicate that ACCase
does not control the flux of carbon into lipids, thus, leaving the
regulatory branch point for lipid biosynthesis poorly defined.
In the present study we sought to understand the mecha-
nism of carbon partitioning under nitrate depletion in the ma-
rine diatom Phaeodactylum tricornutum. We probed changes in
intracellular metabolite pools and transcripts of genes involved
in central carbon metabolism. For this, we compared cells
grown under nitrate replete (control), nitrate starved (NO
3),
and nitrate replete with tungstate (W). In the absence or limi-
tation for molybdenum, tungsten can substitute in the active
site of nitrate reductase, rendering the enzyme catalytically
inactive [14]. Our results on the influence of nitrate assimila-
tion/deprivation on terminal products in P. tricornutum reveal
that regulation occurs at the nexus of the TCA/(GS/GOGAT)/
GDH pathways, similar to cyanobacteria, and is consistent with
the multi-functional role of AKG in microbial phototrophs.
2. Materials and methods
2.1. Culturing system and sampling
P. tricornutum was obtained from the National Center for Ma-
rine Algae and Microbiota (NCMA, formerly CCMP) accession
Pt1 8.6 (clone CCMP 632) was maintained in an artificial sea
water medium [15] supplemented with F/2 nutrients [16] and
buffered with 2 mmol L
1
Tris to pH 8 (hereafter called F/2).
Cultures were grown at 18 C, with air bubbling, under
continuous white LED light at a photon flux density of
200 mmol m
2
s
1
. Pre-inocula were grown with NaNO
3
(0.88 mmol L
1
) as the sole nitrogen source. Exponentially
growing pre-inocula were centrifuged (5500 g, 10 min) and
washed twice with nitrate-free F/2 medium and inoculated in
triplicate at a concentration of 2.5 10
5
cells mL
1
into fresh F/
2 medium with nitrate (0.88 mmol L
1
, control), F/2 medium
without any nitrogen source (NO
3) and F/2 medium with
nitrate (0.88 mmol L
1
) and sodium tungstate (0.88 mmol L
1
,
W) replacing molybdate in the trace metal mixture. All cul-
tures were grown for 3 days in the same conditions as above.
Known numbers of cells were then filtered, flash frozen in
liquid nitrogen, and kept at 80 C until analysis. The only
exception was the metabolite extraction which was per-
formed on fresh samples.
2.2. Cell number, cell volume and chlorophyll
determination
Cell numbers were measured with a Coulter counter multi-
sizer 3 (Beckman Coulter Inc, Fullerton, CA, USA). After 3 days
of growth, cell volume (approximated to a prolate spheroid)
was determined by measuring the length and width of 100
representative cells of each condition with the image analysis
software imageJ (http://imagej.nih.gov/ij/)[17]. Chlorophyll a
was measured from a known amount of cells after filtration
and mechanic homogenization in a cold mixture of acetone at
90% volume fraction [18].
2.3. Fatty acid methyl esters and triacylglycerols
determination
For measuring total fatty acids content we transesterified the
cells’ fatty acids to get fatty acids methyl esters (FAMEs).5 10
7
biomass and bioenergy xxx (2013) 1e10 3
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
cells were filtered onto 25 mm GF/F filters (Whatman) and
subsequently extracted and transesterified in a single step as
reported previously [19]. Analysis of the hexane fraction con-
taining the FAMEs was performed by gas chromatography as in
Ref. [14]. Helium was used as the carrier gas at 179 kPa.
Triacylglycerols (TAGs) were extracted with methyl tert-
butyl ether from 5 10
7
cells collected on GF/F filters [20].TAG
estimation was performed by thin layer chromatography (TLC).
Briefly, afterresuspension of dried samples in a known volume
of chloroform:methanol:water mix (60 mL:30 mL:4.5 mL),
samples were spotted on silica gel 60 10 cm 20 cm TLC
plates (Millipore) with hexane:diethylether:acetic acid mix
(80 mL:20 mL:2 mL) as running solvent. The plates were
sprayed with 50 mg L
1
primuline in an 80% volume fraction
acetone solution, dried for several hours and the TAGs were
visualized under a UV lamp. ImageJ [17] was used to quantify
the intensity of the spots which were compared to spots of
known amounts of a commercial mono-, di-, and triglyceride
standard mix (1787, SigmaeAldrich).
2.4. Carbohydrate content determination
5∙10
7
cells were collected in 1.2 mm pore size polycarbonate
filters (Millipore, Billerica, MA, USA). Total carbohydrate levels
(intracellular reducing pentoses and hexoses) were measured
by reaction with anthrone [21] with a procedure adapted to
diatoms [22].
2.5. Total protein determination
For total protein determination, 2 10
7
cells were filtered onto
1.2 mm pore size polycarbonate filters and re-suspended in 1
denaturing extraction buffer (300 mL), containing 140 mmol L
1
Tris base, 105 mmol L
1
TriseHCl, 0.5 mmol L
1
EDTA, 20 g L
1
lithium dodecyl sulfate, 10% volume fraction glycerol, and
protease inhibitor (1 mL:200 mL, P2714, SigmaeAldrich). After
re-suspension, the samples underwent two cycles of freezing
in liquid nitrogen and thawing with a Microson ultrasonic cell
disruptor XL (Qsonica, Newtown, CT, USA). After pelleting
insoluble cell debris total protein concentration was
measured using a modified Lowry assay (Bio-Rad DC 500-0111,
Hercules, CA, USA). Bovine gamma globulin was used as the
protein standard.
2.6. Metabolite extraction and LC-MS/MS
Intracellular metabolite extraction and analysis was made via
liquid chromatography mass spectrometry (LC-MS/MS).
210
7
cells were filtered onto a 0.45 mm pore size nylon filter
(Pall Corporation, Port Washington, NY, USA) and immedi-
ately cold extracted [23]. Samples were then analyzed on a
1200-series Agilent LC system coupled to an Agilent 6410 Tri-
ple Quadrupole (QQQ) mass spectrometer (Agilent Technolo-
gies, Santa Clara, CA, USA), using previously optimized
Selective Reaction Monitoring parameters [23]. Ion-pairing
chromatography was employed using a Pursuit XRs3 C18
(50 mm 2 mm, Varian, Palo Alto, CA, USA). Mobile phase A
consisted of 10 mmol L
1
tributylamine (pairing agent),
11 mmol L
1
CH
3
OOH while mobile phase B was pure meth-
anol. The chromatography program of mixture of mobile
phase B into mobile phase A was as follows: 0 min (Volume
fraction of B: 0%), 8 min (Volume fraction of B: 40%), 10 min
(Volume fraction of B: 40%), 12.5 min (Volume fraction of B:
90%), 18 min (Volume fraction of B: 90%), 18.1 min (Volume
fraction of B: 0%). A 6 min post-run equilibration period at 0% B
was used. Intracellular metabolite pools were quantified by
comparison to a calibration curve established by spiking
a standard mixture of the target compounds at
10 nmol L
1
e10,000 nmol L
1
into a sample background and
determining the per-metabolite linear response of the in-
strument signals to the expected concentrations.
2.7. Metabolic analysis and real time qPCR
Genes that are involved in lipid biosynthesis and central car-
bon pathways, were selected from the literature [24] and from
both the genome (http://genome.jgi-psf.org/Phatr2/Phatr2.
home.html)[25] and the expressed sequence tag database
(http://www.diatomics.biologie.ens.fr/EST3/seq.php)[26,27].
The prediction of localization of target genes was made using
Mitoprot [28], TargetP1.1 [29] and signalP4.0 [30] online tools.
Total RNA was extracted from 1 10
8
cells collected on 1.2 mm
pore size polycarbonate filters with an RNAeasy plant mini kit
(Qiagen, Venlo, Netherlands). Ambion
TurboDNase (Life
Technologies, Carlsbad, CA, USA) was used to remove DNA
contamination. PCR was used before cDNA generation to
confirm that there was no DNA contamination. Total RNA
quantification and quality assessment was made spectro-
photometrically with a Nanodrop 1000 (Thermo Scientific).
cDNA generation was done using oligodT as primers and the
SuperScript
III reverse transcriptase kit (Life Technologies).
The resulting cDNA-generation reaction mixture (1 mL) was
directly used as the template for qPCR. Primers for target
genes were constructed with NCBI Primer-BLAST (http://
www.ncbi.nlm.nih.gov/tools/primer-blast/), designed to
anneal near to the 30end of the mRNA (Table S1). Standard
molecules of known molecular weight, for copy number
calculation, were generated by cloning each amplicon onto a
TOPO
TA (Life Technologies) cloning kit. qPCR was made
using the Applied Biosystems SYBR
Green PCR Master mix
(Life Technologies) on a Mx3000P QPCR system (Agilent
Technologies). All standard curves had at least 5 points,
covered 5 orders of magnitude and had an R
2
>0.94. Each qPCR
amplification on biological samples was then compared to its
specific calibration curve to calculate the gene copy number
present. Calculated copy numbers were normalized to total
RNA extracted.
2.8. Quantum requirement and fluxes calculations
Carbon fluxes (pmol cell
1
d
1
) into proteins, carbohydrates
and fatty acids were calculated according to the equation
Nm¼(dN/dt), where Nis moles of carbon in protein, car-
bohydrate or fatty acids measured after 3 days, mthe cellular
specific growth rate and dN/dtthe average flux. The average
carbon mass fraction in proteins (44%) was calculated using
the mass of carbon per mass of amino acid normalized by the
relative abundance of each amino acid according to the ge-
netic code. The mass fraction of carbon in carbohydrates (40%)
was calculated using glucose as a standard. The mass fraction
biomass and bioenergy xxx (2013) 1e104
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
of carbon in fatty acids (76%) was calculated using the mass of
carbon per mass of fatty acid normalized by the abundance of
each fatty acid measured in this study. The quantum
requirement for each pool (4
1
) was calculated according to
[31] using the values for a* and E
k
values reported in Ref. [14].
3. Results
3.1. Physiological responses to nitrate limitation
The control condition presented a growth rate of
0.86 d
1
0.04 d
1
, while the growth rate of the eNO
3and the
W cultures were 0.36 and 0.76-fold of the control cultures,
respectively (Table 1). Chlorophyll a(Chl a) content, which
generally reflects total biosynthetic nitrogen accumulation at a
constant irradiance, was lower for both eNO
3and W cultures
by 0.83 and 0.50-fold, respectively, vs. the control condition
(0.12 pg 0.02 pg cell
1
)(Table 1). The volume of the cells in the
eNO
3cultures did not change significantly in relation to the
control, while the ones in the W condition increased their
volume by approximately 1.18-fold (p-value <0.01).
Regarding the main carbon sinks, the eNO
3cultures, had a
1.21-fold increase in levels of FA per cell accompanied by a
0.74-fold decrease in proteins and a 0.29-fold decrease in the
carbohydrates relative to the control. In the W condition, the
FA per cell increased by 1.14-fold, while the protein levels
decreased by 0.40-fold and the total carbohydrate levels did
not change relative to the control. TAGs constituted a slightly
larger mass fraction of the total measured FA in the eNO
3
(93%) and W (82%) compared to the control (76%) (Table 1).
3.2. Changes in intracellular metabolite pools in
response to nitrate limitation
In order to obtain further insight into the metabolic state of
cells, the intracellular metabolite pools were measured by
LC-MS/MS (Fig. 2). The eNO
3condition produced an overall
decrease of several metabolites in comparison to the control.
Most significantly, the nitrogen carriers for biosynthesis, GLU
and GLN were reduced 7 and 55-fold, respectively, in the eNO
3
case compared to the control condition. As these metabolites
are formed from AKG and ammonium, we looked at in-
termediates of the TCA cycle. The AKG level increased by
18-fold compared to the control condition, while other TCA
metabolites downstream such as succinate increased nearly
3-fold, while fumarate and malate remained unchanged.
These data provide definitive evidence for control occurring at
(GS/GOGAT)/GDH (steps (3) and (4) in Fig. 1).
The large increase in AKG suggests either increased flux
into AKG from the upstream portion of the TCA cycle, greater
input through amino acid catabolism via GDH, or decreased
output of AKG into the GS/GOGAT pathway. To further
constrain the source of AKG, we examined the levels of the
TCA metabolites upstream prior to the typically irreversible
step at isocitrate dehydrogenase. This analysis revealed a
3-fold decrease in levels of citrate, acetyl-CoA (AcCoA) and
pyruvate under eNO
3condition, suggestive of either faster
depletion or reduced input into these pools. AcCoA is the
immediate precursor to malonyl-CoA (MaCoA) for FA biosyn-
thesis. MaCoA was significantly decreased (10-fold), in the
eNO
3condition.
As expected, the total pool sizes of adenosine phosphates
and both pyridine nucleotide pairs were also highly decreased
in the eNO
3case (Fig. 2). However, the adenylate cell energy
charge (CEC) decreased, from 0.74 0.03 in the control to
0.63 0.06 in the eNO
3condition (p-value <0.05) (Table 2).
Regarding the cellular pyridine nucleotide redox poise, the
NADH/NAD
þ
ratio was unchanged, while the NADPH/NADP
þ
ratio increased 3-fold in the eNO
3condition vs. the control (p-
value <0.01) (Table 2). This significant increase in the ratio of
NADPH/NADP
þ
is consistent with lower nitrate reduction to
ammonium and lower activity of other sinks such as the GS/
GOGAT cycle and carbon fixation.
In general, the W condition presented fewer changes in the
metabolite pools than the eNO
3condition in comparison to
the control (Fig. 2). The levels of AKG, AcCoA, MaCoA, aden-
osine phosphates, NADH and NAD
þ
were unchanged, while
the level of NADP
þ
decreased by about 60% relative to the
control. The CEC and the NADH/NAD
þ
redox poise, were also
unchanged for the W condition in comparison to the control.
The GLU and GLN levels decreased approximately 1.7 and
12-fold, respectively, indicating some degree of nitrogen lim-
itation. The NADPH/NADP
þ
redox poise showed a statistically
significant increase, albeit much smaller in magnitude than
that of the eNO
3condition (Table 2).
3.3. mRNA abundance of genes of the central carbon
metabolism in response to nitrate limitation
In order to relate the metabolite data to the genetic response
of cells, the mRNA abundance of key genes involved in the key
branch points were measured by real-time RT-qPCR (Fig. 3).
The total RNA (ribosomal þmessenger þtransfer) measured
in the eNO
3condition, was only 44% of that of the control
(p-value <0.01, n¼6). In contrast, the total RNA levels
extracted from the W condition were 2.3-fold higher than
Table 1 ePhysiological parameters of P. tricornutum after
3 days of growth in nitrate replete (control), nitrate
limited (LNOL
3) and tungstate (W) conditions. In
parentheses are the values as a fraction of the control set
to 100. Here and elsewhere, the mean and one standard
deviation of biological triplicates are shown.
Control NO
3W
m(d
1
) 0.86 0.04 (100) 0.31 0.01 (36) 0.66 0.03
(76)
Cell volume (mm
3
) 74.0 22.9 (100) 70.8 23.4 (96) 87.6 33.3
(118)
Chl a(pg cell
1
) 0.12 0.02 (100) 0.02 0.01 (17) 0.06 0.01
(50)
Protein (pg cell
1
) 13.6 0.1 (100) 3.6 0.6 (26) 8.2 0.6 (60)
Carbohydrate
(pg cell
1
)
7.7 0.3 (100) 5.5 0.9 (71) 8.2 0.7
(106)
FAMEs (pg cell
1
) 2.9 0.1 (100) 3.5 0.09 (121) 3.3 0.04
(114)
Mass fraction of
FAMEs as TAG
76 4936822
RNA (pg$cell
1
) 0.06 0.01 (100) 0.02 0.01 (33) 0.14 0.02
(233)
biomass and bioenergy xxx (2013) 1e10 5
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
those of the control condition (p-value <0.01, n¼6) (Table 1).
For all 18 genes tested, there was a general increase in the
abundance of their mRNA levels in both conditions compared
to the control. In the eNO
3condition, genes involved in TCA
cycle were amongst the ones that presented highest in-
creases. Specifically, the highest increase (65-fold) was
observed in isocitrate dehydrogenase (Isocitrate DH) mRNA. In
respect to lipid biosynthesis genes, a 35-fold increase in the
ACCase mRNA abundance was observed in the eNO
3condi-
tion, while the mRNA levels of diacylglycerol-acyltransferase
(DGAT)2B and DGAT2D were increased 3- and 8-fold, respec-
tively, in comparison to the control. The W condition gener-
ally had higher fold changes than those of the nitrogen-
starved condition. The mRNA levels of isocitrate dehydroge-
nase increased 90-fold, ACCase increased 28-fold, while the
DGAT2B and DGAT2D increased 8 and 20-fold respectively. A
very large increase in mRNA abundance of nitrate reductase
(NR) was also observed in response to nitrogen starvation (26-
fold) and W conditions (248-fold).
4. Discussion
4.1. GS/GOGAT/GDH are important regulatory branch
points
The results of this study strongly suggest that the key branch
point in lipid biosynthesis in P. tricornutum centers on the
metabolism of AKG (Fig. 1). In the eNO
3condition, the general
decrease in metabolite pools is consistent with lower overall
metabolic activity due to the global translational limitation
this nitrogen deprivation inevitably imposes. Nonetheless, the
changes in the levels of glutamate (GLU), glutamine (GLN), and
a-ketoglutarate (AKG) indicate that the GS/GOGAT/GDH
branch points, steps (3) and (4) of Fig. 1, the shift the flux of
carbon from protein into lipid biosynthesis. The observed
large decrease in the pool sizes of GLU and GLN, in parallel
with the 18-fold increase in AKG and 3-fold increase in
NADPH/NADP
þ
ratio, are consistent with an arrest of the
GS/GOGAT cycle due to ammonium depletion at step (3). The
18-fold increase in AKG is likely also due to recycling of amino
acids from protein catabolism, including chlorophyll associ-
ated proteins [a large decrease in chlorophyll aand protein
levels were observed in this condition, (Table 1)] through the
GDH pathway. This pathway is readily reversible and, because
Fig. 2 eRatio of abundance of key cellular metabolites at
steady state growth determined by LC-MS/MS. The
abundance of each metabolite in the control condition was
used as reference and set to 1. **Student ttest p-
value <0.01, *Student ttest p-value <0.05. AA eamino-
acids; FA efatty acids; CEC eadenylate cell energy charge;
Pyr Nuc ePyridine nucleotides; UDP-Gluc eUDP-glucose;
G1P eGlucose-1-phosphate; R5P eRibulose-5-phosphate;
X5P eXilulose-5-Phosphate; F6P eFructose-6-Phsphate;
FBP eFructose bis-phosphate; DHAP eDihydroxyacetone;
GAP eGlyceraldehyde-3-phosphate; 3 PG e3-
phosphoglycerate, PEP ePhosphoenolpyruvate; AcCoA e
Acetyl-CoA; AKG ea-ketoglutarate; GLN eGlutamine; GLU
eGlutamate; MaCoA eMalonyl-Co; AMP, ADP, ATP e
adenosyl-(mono,di,tri)phosphate; NAD
D
, NADH e
Nicotinamide adenine dinucleotide; NADP
D
, NADPH -
Nicotinamide adenine dinucleotide phosphate.
Table 2 eAdenylate Cell Energy Charge (CEC) and
Pyridine nucleotide ratios in the 3 conditions tested.
Adenylate CEC was calculated according to the formula
([ATP] D½[ADP])/([AMP] D[ADP] D[ATP]). *Student t-test
p-value <0.05.**Student t-test p-value <0.01.
Control NO
3W
Adenylate CEC 0.74 0.03 0.63 0.06* 0.75 0.06
NADH/NAD
þ
1.22 0.21 1.16 0.10 1.27 0.06
NADPH/NADP
þ
1.2 0.06 3.7 0.9** 1.6 0.2*
biomass and bioenergy xxx (2013) 1e106
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
of the high K
m
for ammonium, it operates primarily to convert
GLU to AKG. Analogously, this pathway was suggested to have
a mainly catabolic role in Chlamydomonas reinhardii [32].
Interestingly, the mRNA level of the icd gene, which codes
for isocitrate DH (responsible for synthesizing AKG) was
increased 65-fold under nitrate deprivation. Yet, it remains to
be clarified if the higher mRNA levels observed in our experi-
ments are translated into higher protein levels. Our results
strongly suggest that blockage of the GS/GOGAT branch point
due to low ammonium availability is responsible for the
decrease in protein content observed. The system attempts to
compensate for this by elevating the icd gene abundance,
attempting to make more AKG, which is not in shortage. As a
consequence, the system redirects newly fixed carbon into
lipids (via AcCoA and MaCoA) which requires no nitrogen for
its biosynthesis and, in addition, is a more effective way to
relieve the stronger reducing intracellular environment [3].
The observed redirection of carbon metabolism into FA and
TAGs under eNO
3conditions is consistent with the increases
in the mRNA levels of ACCase, DGAT2B and DGAT2D, which
are key steps for FA and TAG biosynthesis pathways [33,34].
Increased mRNA levels of the chloroplast ACCase were also
recently reported in nitrogen depleted P. tricornutum cultures,
albeit with a lower magnitude [35].
The replacement of Mo by W generated an intermediary
response between the control and the eNO
3condition, likely
due to the incomplete inactivation of NR activity. Nitrogen
limitation in this condition was apparent by the decrease in
GLN and GLU levels which are in agreement with the 40%
lower proteins per cell, relative to the control. The AKG level
was, however similar to that of the control. This result is
consistent with an incomplete inactivation of NR activity by
W, which would still allow some nitrogen assimilation by the
GS/GOGAT cycle, preventing the accumulation of AKG.
The FA and TAGs per cell were also increased in the W
conditions, but not as much as in nitrogen-starved. This is
consistent with the slight but significant increase in the
NADPH/NADP
þ
ratio observed and with an incomplete NR
inactivation by W. Lastly, the changes in mRNA levels of iso-
citrate DH, NR, ACCase, DGAT2B and DGAT2D were, in gen-
eral, larger than what was observed in nitrogen starved
conditions. This result is somewhat unexpected, as W causes
only a mild nitrogen limitation. Together with the fact that we
observed a 2.3-fold increase in total RNA in the W condition
and a 1.18-fold increase in cell volume, relative to the control,
this indicates that there are unspecific effects of tungstate in
cells. Furthermore, a specific deregulation of nitrate reductase
mRNA by tungstate was previously reported in the higher
plant Nicotiana tabacum [36]. RNAi NR knock-downs could
further clarify the importance of this enzyme on the fate of
carbon, circumventing the unspecific effects of tungstate in
cellular physiology.
4.2. Putative nitrogen sensing mechanisms in diatoms
Taking the eNO
3and the W conditions together, we suggest
that the cell senses the ratio of AKG to GLN or GLU (GL(N/U)) in
addition to its overall redox state. FA and TAG synthesis have
a higher NADPH requirement in comparison with carbohy-
drates and proteins. Consistent with this requirement, the
ratio of NADPH/NADP
þ
, which is generally tightly regulated,
was increased 3-fold relative to the control, implying a
significantly more reducing intracellular environment,
including a more reduced plastoquinone (PQ) pool. The con-
trol of microalgal nuclear genes involved in photosynthesis,
carbon fixation, carbohydrate metabolism and nitrogen
assimilation (including nitrate reductase) has been previously
proposed to be controlled, at least in part, by the redox state of
the PQ pool [7,37,38]. The increase in the NADPH/NADP
þ
ratio
observed could then, not only facilitate lipid biosynthesis as a
substrate, but also be a sign of a reduced PQ pool that can
regulate nuclear genes for lipid biosynthesis.
Under mild nitrate limitation the decrease in the pools of
GLN and GLU are responsible for the AKG/GL(N/U) ratio in-
crease. However, under nitrate starvation both the increase of
AKG and the decrease of GL(N/U) are responsible for the large
ratio increase. In N. tabacum leaves, it was suggested that the
transcription level of NR was antagonistically controlled by
AKG (activator) and GLN (repressor). The ratio of these
Fig. 3 emRNA abundances of target genes as measured by
real-time RT-qPCR. Logarithm of the ratio of change of
mRNA abundance of target genes in the nitrate depleted
and tungstate conditions relative to the control condition
whose copy number was set to 1. The calculated copy
number for each gene was normalized to total RNA. Gene
name abbreviations are given in Table S1.
biomass and bioenergy xxx (2013) 1e10 7
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
metabolites was proposed to be more important for nitrogen
sensing than their individual concentrations [39]. In cyano-
bacteria, similar changes in AKG, GLU and GLN concentration
were reported to take place upon nitrogen limitation [9]. These
metabolites signal the nitrogen availability and start a cascade
of responses, through the global nitrogen regulator NtcA, that
leads to the increase in expression of several genes including
isocitrate DH, NR, GS/GOGAT and several nitrate transporters
[10]. Interestingly, both isocitrate DH and NR mRNA were also
over-represented in our experimental conditions, and there is
expressed sequence tag indication that the mRNA levels for
nitrate transporters and GS/GOGAT genes are also increased
under nitrogen limitation (Table S2). Our results suggest that a
similar nitrogen sensing mechanism that responds to the
AKG/GL(N/U) ratio may exist in P. tricornutum. A similar
conclusion was reached by analyzing the protein expression
levels in the diatom Thalassiosira pseudonana in response to
nitrogen starvation [40]. However, no ortolog of NtcA could be
found in the genome of P. tricornutum, implying that the signal
transduction between AKG/GL(N/U) and gene expression
must use a different protein or mechanism.
The general increase in mRNA abundance of nitrogen
assimilation genes seems contradictory to the arrest of the
GS/GOGAT cycle observed here under eNO
3conditions,
although it is unclear if the increased mRNA levels are
translated into higher protein levels or increased enzymatic
activity (untested). In the diatom T. pseudonana these proteins
were observed to be increased in response to nitrogen star-
vation and it was suggested that they could serve, together
with the urea cycle, to more efficiently recycle nitrogen rich
compounds from catabolic processes [40]. This is similar to
what is observed in cyanobacteria which also have an urea
cycle [41].The physiological result of this putative sensing
mechanism is however, different enitrate starvation leads to
accumulation of high levels of carbohydrates in cyanobacteria
[42], and to high lipid accumulation in diatoms.
4.3. Fluxes and quantum requirements of the main
carbon pools
The average flux into each major carbon pool as well as the
quantum requirement for placing carbon into each pool was
calculated as described in the methods section. These values
allow the quantification of the carbon that is placed into each
sink per unit time and the energetic cost of that flux. The flux
into each pool decreases exponentially with the increase of its
quantum requirements (Table 3,Fig. 4). This robust relation-
ship does not differentiate between carbon pools (protein,
carbohydrate or FA) or growth conditions (control, NO
3or
W). In the control condition, protein has the largest influx of
carbon and the lowest quantum requirement, while FA has
the highest quantum requirement and the lowest carbon
influx. In the eNO
3condition, the quantum requirement for
all three carbon pools increased significantly relative to the
control: 7-fold for proteins, 2.6-fold for carbohydrates and 1.5-
fold for FA. This translates into a quantitatively related
reduction of total carbon flux into these pools: 10-fold for
proteins, 3.7-fold for carbohydrates and 2-fold for FA. This
overall lower total carbon flux is in agreement with the lower
growth rate and general lower levels of metabolite in-
termediates measured by LC-MS/MS. Despite the measured
increase in the mass fraction of carbon deposited into FA in
the eNO
3condition relative to the control (41% vs. 20%,
respectively) the total carbon flux into FA is still 2-fold lower
than in the control. This lower total flux can also explain the
2.7 and 10-fold lower levels of AcCoA and MaCoA relative to
the control. Nonetheless, it is possible that the decrease of
AcCoA and MaCoA is being overestimated as the local con-
centration of AcCoA and MaCoA inside the chloroplast may be
considerably higher than the average cellular concentration.
This means that the global productivity of FA in a eNO
3cul-
ture is severely diminished relative to the control, since both
the flux of carbon to FA per cell and the number of cells is
decreased. Thus, this treatment is not a good option for
biotechnological production of biodiesel.
The addition of W to the media affects mainly protein
biosynthesis and consequently cell growth. The quantum
requirement for protein synthesis increased 1.8-fold and
leads to a 2-fold decrease of carbon flux into proteins. The
quantum requirements for FA and carbohydrates were
similar to the control and thus the carbon fluxes into those
pools were also indistinguishable from those of the control
condition. This is consistent with the levels of AcCoA and
MaCoA that remained stable in this treatment, while GLU and
GLN levels decreased sharply as described above. Similarly to
the eNO
3case, the lower levels of proteins and the lower
Table 3 eQuantum requirements and flux of carbon into for protein, carbohydrates and fatty acids in the control, eNOL
3
and W conditions. The fluxes and 4
L1
was calculated as described in the methods section using the m, protein,
carbohydrate or lipid data from Table 1
Control NO
3W
Protein 4
1
(mol$mol
1
)7517 532 109 137 40
Flux C
protein
(pmol cell
1
d
1
) 0.43 0.02 0.04 0.01 0.2 0.02
Mass fraction flux
Proteina
53 25 38
Carbohydrate 4
1
(mol mol
1
) 146 37 383 77 151 46
Flux C
carbohydrate
(pmol cell
1
d
1
) 0.22 0.02 0.06 0.01 0.18 0.02
Mass fraction flux C
carbohydratea
27 34 35
Fatty acid 4
1
(mol mol
1
) 204 50 317 19 197 45
Flux C
fatty acid
(pmol cell
1
d
1
) 0.16 0.01 0.07 0.01 0.14 0.01
Mass fraction flux
fatty acida
20 41 28
a
Assumes that the sum of fluxes of carbon for protein, carbohydrate and fatty acid is equal to 100.
biomass and bioenergy xxx (2013) 1e108
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
growth rates will limit the productivity of cultures treated
with W in the presence of nitrate, even if the flux of carbon to
FA remains unchanged. The trends observed here for FAs
fluxes and accumulation are consistent with the ones
calculated previously under similar conditions [14]. However,
in that study these parameters were not calculated for pro-
teins or carbohydrates.
5. Conclusions
In conclusion, here we demonstrated that under nitrogen
starvation all fluxes of carbon into the major carbon sinks are
considerably decreased. The addition of tungstate mimicked
mild nitrogen limitation. Our metabolite data strongly sug-
gests that the GS/GOGAT/GDH pathways are key branch
points controlling the allocation of carbon and the recycling of
proteins, in P. tricornutum. The ratio of AKG/GL(N/U) likely re-
ports the nitrogen availability in the cell triggering the gene
expression responses observed, by an unknown signal
pathway. The high NADPH/NADP
þ
ratio may facilitate lipid
biosynthesis and implies a more reduced cellular environ-
ment, which could trigger previously proposed nuclear gene
expression mechanisms. These putative signal transduction
cascades (both AKG/GL(N/U) sensitive and the redox sensitive)
deserve further investigation in diatoms, as they may provide
new targets for inducing lipid biosynthesis in actively growing
cells. Thus, future studies may focus on controlling the AKG/
GL(N/U) ratio by genetically knocking down nitrate reductase
or proteins in the GDH/GS/GOGAT pathways, while simulta-
neously promoting higher fluxes of carbon towards lipid
biosynthesis. This potentially could be achieved by over
expressing ACCase in conjunction with DGATs, while
increasing the NADPH/NADP
þ
ratio by over expressing the
ferredoxin:NADPH oxidoreductase.
Acknowledgments
This work was supported by the DOE-EERE, grant number DE-
EE0003373. The authors acknowledge Dr. Jorge Dinamarca
Cerda for assistance with TAG analysis, as well as Dr. Sang
Hoon Lee for initial experimental planning. LTG was sup-
ported by a doctoral fellowship FCT-MCTES (reference code
SFRH/BD/61387/2009). OL and MJF were funded by a gift by
James Gibson to PGF. We would like to acknowledge assis-
tance from Rutgers University, the Aresty Research Center for
Undergraduates, and the Funding Unit for JSS support. The
authors acknowledge Agilent Technologies, Inc. for their
partnership and support in LC/MS method development and
instrumentation support.
Appendix A. Supplementary data
Supplementary data related to this article can be found online
at http://dx.doi.org/10.1016/j.biombioe.2013.10.007.
references
[1] Myers J. On the algae: thoughts about physiology and
measurements of efficiency. In: Falkowski PG, editor.
Primary productivity in the sea. New York: Plenum Press;
1980. p. 1e16.
[2] Sheehan J, Dunahay T, Benemann J, Roessler P. A look back
at the US department of energy’s aquatic species program:
biodiesel from algae. In: Energy Do, editor. Golden, CO:
National Renewable Energy Laboratory; 1998.
[3] Hu Q, Sommerfeld M, Jarvis E, Ghirardi M, Posewitz M,
Seibert M, et al. Microalgal triacylglycerols as feedstocks for
biofuel production: perspectives and advances. Plant J
2008;54:621e39.
[4] Schwender J, Ohlrogge JB. Probing in vivo metabolism by
stable isotope labeling of storage lipids and proteins in
developing Brassica napus embryos. Plant Physiol
2002;130:347e61.
[5] Zehr JP, Falkowski PG. Pathway of ammonium assimilation
in a marine diatom determined with the radiotracer
13
N.
J Phycol 1988;24:588e91.
[6] Campbell WH. Nitrate reductase structure, function and
regulation: bridging the gap between biochemistry and
physiology. Annu Rev Plant Physiol Plant Mol Biol
1999;50:277e303.
[7] Giordano M, Chen Y-B, Koblizek M, Falkowski PG. Regulation
of nitrate reductase in Chlamydomonas reinhardtii by the
redox state of the plastoquinone pool. Eur J Phycol
2005;40:345e52.
[8] Sasaki Y, Nagano Y. Plant acetyl-CoA carboxylase: structure,
biosynthesis, regulation, and gene manipulation for plant
breeding. Biosci Biotechnol Biochem 2004;68:1175e84.
[9] Muro-Pastor MI, Reyes JC, Florencio FJ. Cyanobacteria
perceive nitrogen status by sensing intracellular 2-
oxoglutarate levels. J Biol Chem 2001;276:38320e8.
[10] Herrero A, Muro-Pastor AM, Flores E. Nitrogen control in
cyanobacteria. J Bacteriol 2001;183:411e25.
[11] Dunahay TG, Jarvis EE, Dais SS, Roessler PG. Manipulation of
microalgal lipid production using genetic engineering. Appl
Biochem Biotech 1996;57-58:223e31.
Fig. 4 eRelationship between carbon fluxes and quantum
requirements. Values of quantum requirements and fluxes
are indicated in Table 3. An exponential decay equation
was fitted to the points with a correlation coefficient of
0.97. The markers used for each condition are given in the
inset table.
biomass and bioenergy xxx (2013) 1e10 9
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007
[12] Klaus D, Ohlrogge J, Neuhaus HE, Do
¨rmann P. Increased fatty
acid production in potato by engineering of acetyl-CoA
carboxylase. Planta 2004;219:389e96.
[13] Roesler K, Shintani D, Savage L, Boddupalli S, Ohlrogge J.
Targeting of the Arabidopsis homomeric acetyl-coenzyme A
carboxylase to plastids of rapeseeds. Plant Physiol
1997;113:75e81.
[14] Frada MJ, Burrows EH, Wyman KD, Falkowski PG. Quantum
requirements for growth and fatty acid biosynthesis in the
marine diatomPhaeodactylum tricornutum (Bacilloriophyceae) in
nitrogen replete and limitedconditions. J Phycol 2013;49:381e8.
[15] Goldman JC, McCarthy JJ. Steady state growth and
ammonium uptake of a fast-growing marine diatom. Limnol
Oceanogr 1978;23:695e703.
[16] Guillard RR, Ryther JH. Studies of marine plankton diatoms. I.
Cyclotella nana (Husted) and Detonula confervacea (Cleve).
Can J Microbiol 1962;8:229e39.
[17] Abramoff MD, Magalha
˜es PJ, Ram SJ. Image processing with
ImageJ. Biophotonics Int 2004;11:36e42.
[18] Jeffrey S, Humphrey G. New spectrophotometry equations
for determining chlorophyll a, chlorophyll b, chlorophyll c-1
and chlorophyll c-2 in higher plants, algae and natural
phytoplankton. Biochemie und Physiologie der Pflanzen
1975;167:191e4.
[19] Rodrı´guez-Ruiz J, Belarbi E-H, Sa
´nchez JLG, Alonso DL. Rapid
simultaneous lipid extraction and transesterification for
fatty acid analyses. Biotechnol Tech 1998;12:689e91.
[20] Matyash V, Liebisch G, Kurzchalia TV, Shevchenko A,
Schwudke D. Lipid extraction by methyl-tert-butyl ether for
high-throughput lipidomics. J Lipid Res 2008;49:1137e46.
[21] Trevelyan WE, Harrison JS. Studies on yeast metabolism. I.
Fractionation and microdetermination of cell carbohydrates.
Biochem J 1952;50:298e303.
[22] Post AF, Dubinsky Z, Wyman K, Falkowski PG. Physiological
responses of a marine planktonic diatom to transitions in
growth irradiance. Mar Ecol Prog Ser 1985;25:141e9.
[23] Bennette NB, Eng JF, Dismukes GC. An LCeMS-based
chemical and analytical method for targeted metabolite
quantification in the model cyanobacterium Synechococcus
sp. PCC 7002. Anal Chem 2011;83:3808e16.
[24] Kroth PG, Chiovitti A, Gruber A, Martin-Jezequel V, Mock T,
Parker MS, et al. A model for carbohydrate metabolism in the
diatom Phaeodactylum tricornutum deduced from comparative
whole genome analysis. PLoS ONE 2008;3:e1426.
[25] Bowler C, Allen AE, Badger JH, Grimwood J, Jabbari K, Kuo A,
et al. The Phaeodactylum genome reveals the evolutionary
history of diatom genomes. Nature 2008;456:239e44.
[26] Maheswari U, Montsant A, Goll J, Krishnasamy S,
Rajyashri KR, Patell VM, et al. The diatom EST database. Nucl
Acids Res 2005;33:D344e7.
[27] Maheswari U, Mock T, Armbrust EV, Bowler C. Update of the
diatom EST database: a new tool for digital transcriptomics.
Nucl Acids Res 2009;37:D1001e5.
[28] Claros MG, Vincens P. Computational method to predict
mitochondrially imported proteins and their targeting
sequences. Eur J Biochem 1996;241:779e86.
[29] Emanuelsson O, Nielsen H, Brunak S, von Heijne G.
Predicting subcellular localization of proteins based on their
N-terminal amino acid sequence. J Mol Biol
2000;300:1005e16.
[30] Petersen TN, Brunak S, von Heijne G, Nielsen H. SignalP 4.0:
discriminating signal peptides from transmembrane regions.
Nat Methods 2011;8:785e6.
[31] Falkowski PG, Dubinsky Z, Wyman K. Growth-irradiance
relationships in phytoplankton. Limnol Oceanogr
1985;30:311e21.
[32] Cullimore JV, Sims AP. Pathway of ammonia assimilation in
illuminated and darkened Chlamydomonas reinhardii.
Phytochemistry 1981;20:933e40.
[33] Chen JE, Smith AG. A look at diacylglycerol acyltransferases
(DGATs) in algae. J Biotechnol 2012;162:28e39.
[34] Gong Y, Zhang J, Guo X, Wan X, Liang Z, Hu CJ, et al.
Identification and characterization of PtDGAT2B, an
acyltransferase of the DGAT2 acyl-Coenzyme A:
diacylglycerol acyltransferase family in the diatom
Phaeodactylum tricornutum. FEBS Lett 2013;587:481e7.
[35] Valenzuela J, Mazurie A, Carlson RP, Gerlach R, Cooksey KE,
Peyton BM, et al. Potential role of multiple carbon fixation
pathways during lipid accumulation in Phaeodactylum
tricornutum. Biotechnol Biofuels 2012;5:40.
[36] Deng M, Moureaux T, Caboche M. Tungstate, a molybdate
analog inactivating nitrate reductase, deregulates the
expression of the nitrate reductase structural gene. Plant
Physiol 1989;91:304e9.
[37] Durnford DG, Falkowski PG. Chloroplast redox regulation of
nuclear gene transcription during photoacclimation.
Photosynth Res 1997;53:229e41.
[38] Nott A, Jung HS, Koussevitzky S, Chory J. Plastid-to-nucleus
retrograde signaling. Annu Rev Plant Biol 2006;57:739e59.
[39] Ferrario-Me
´ry S, Masclaux C, Suzuki A, Valadier M-H, Hirel B,
Foyer CH. Glutamine and a-ketoglutarate are metabolite
signals involved in nitrate reductase gene transcription in
untransformed and transformed tobacco plants deficient in
ferredoxin-glutamine-a-ketoglutarate aminotransferase.
Planta 2001;213:265e71.
[40] Hockin NL, Mock T, Mulholland F, Kopriva S, Malin G. The
response of diatom central carbon metabolism to nitrogen
starvation is different from that of green algae and higher
plants. Plant Physiol 2012;158:299e312.
[41] Tolonen AC, Aach J, Lindell D, Johnson ZI, Rector T, Steen R,
et al. Global gene expression of Prochlorococcus ecotypes in
response to changes in nitrogen availability. Mol Syst Biol
2006;2:53.
[42] Stevens SE, Balkwill DL, Paone DAM. The effects of nitrogen
limitation on the ultrastructure of the cyanobacterium
Agmenellum quadruplicatum. Arch Microbiol 1981;130:204e12.
biomass and bioenergy xxx (2013) 1e1010
Please cite this article in press as: Guerra LT, et al., Regulatory branch points affecting protein and lipid biosynthesis in the
diatom Phaeodactylum tricornutum, Biomass and Bioenergy (2013), http://dx.doi.org/10.1016/j.biombioe.2013.10.007