A mutant in the ADH1 gene of Chlamydomonas reinhardtii elicits metabolic restructuring during anaerobiosis.

Abstract

The green alga Chlamydomonas reinhardtii has numerous genes encoding enzymes that function in fermentative pathways. Among these, the bifunctional alcohol/acetaldehyde dehydrogenase (ADH1), highly homologous to the Escherichia coli AdhE enzyme, is proposed to be a key component of fermentative metabolism. To investigate the physiological role of ADH1 in dark anoxic metabolism, a Chlamydomonas adh1 mutant was generated. We detected no ethanol synthesis in this mutant when it was placed under anoxia; the two other ADH homologs encoded on the Chlamydomonas genome do not appear to participate in ethanol production under our experimental conditions. Pyruvate formate lyase, acetate kinase, and hydrogenase protein levels were similar in wild-type cells and the adh1 mutant, while the mutant had significantly more pyruvate:ferredoxin oxidoreductase. Furthermore, a marked change in metabolite levels (in addition to ethanol) synthesized by the mutant under anoxic conditions was observed; formate levels were reduced, acetate levels were elevated, and the production of CO(2) was significantly reduced, but fermentative H(2) production was unchanged relative to wild-type cells. Of particular interest is the finding that the mutant accumulates high levels of extracellular glycerol, which requires NADH as a substrate for its synthesis. Lactate production is also increased slightly in the mutant relative to the control strain. These findings demonstrate a restructuring of fermentative metabolism in the adh1 mutant in a way that sustains the recycling (oxidation) of NADH and the survival of the mutant (similar to wild-type cell survival) during dark anoxic growth.

Full text

Page 1

A Mutant in the ADH1 Gene of Chlamydomonas
reinhardtii Elicits Metabolic Restructuring
during Anaerobiosis1[W]
Leonardo Magneschi2*, Claudia Catalanotti, Venkataramanan Subramanian, Alexandra Dubini,
Wenqiang Yang, Florence Mus, Matthew C. Posewitz, Michael Seibert,
Pierdomenico Perata, and Arthur R. Grossman
Department of Plant Biology, Carnegie Institution for Science, Stanford, California 94305 (L.M., C.C., W.Y.,
A.R.G.); PlantLab, Institute of Life Sciences, Scuola Superiore Sant’Anna, Pisa 56124, Italy (L.M., P.P.);
Biosciences Center, National Renewable Energy Laboratory, Golden, Colorado 80401 (V.S., A.D., M.S.);
Colorado School of Mines, Department of Chemistry and Geochemistry, Golden, Colorado 80401 (V.S., M.C.P.,
M.S.); and Department of Microbiology, Department of Chemical and Biological Engineering, and Center
for Biofilm Engineering, Montana State University, Bozeman, Montana 59717 (F.M.)
The green alga Chlamydomonas reinhardtii has numerous genes encoding enzymes that function in fermentative pathways.
Among these, the bifunctional alcohol/acetaldehyde dehydrogenase (ADH1), highly homologous to the Escherichia coli AdhE
enzyme, is proposed to be a key component of fermentative metabolism. To investigate the physiological role of ADH1 in dark
anoxic metabolism, a Chlamydomonas adh1 mutant was generated. We detected no ethanol synthesis in this mutant when it was
placed under anoxia; the two other ADH homologs encoded on the Chlamydomonas genome do not appear to participate in
ethanol production under our experimental conditions. Pyruvate formate lyase, acetate kinase, and hydrogenase protein levels
were similar in wild-type cells and the adh1 mutant, while the mutant had significantly more pyruvate:ferredoxin
oxidoreductase. Furthermore, a marked change in metabolite levels (in addition to ethanol) synthesized by the mutant under
anoxic conditions was observed; formate levels were reduced, acetate levels were elevated, and the production of CO2was
significantly reduced, but fermentative H2production was unchanged relative to wild-type cells. Of particular interest is the
finding that the mutant accumulates high levels of extracellular glycerol, which requires NADH as a substrate for its synthesis.
Lactate production is also increased slightly in the mutant relative to the control strain. These findings demonstrate a
restructuring of fermentative metabolism in the adh1 mutant in a way that sustains the recycling (oxidation) of NADH and the
survival of the mutant (similar to wild-type cell survival) during dark anoxic growth.
Photosynthetic microorganisms that have evolved
in the soil, such as the unicellular green alga Chlamy-
domonas reinhardtii (Chlamydomonas throughout), are
subjected to continuous fluctuations in oxygen avail-
ability and may experience anoxic or microaerobic
conditions during the night and early morning, when
low levels of photosynthesis combined with microbial
respiration deplete the local environment of oxygen.
The anoxic environment elicits the synthesis/activation
of enzymes that ferment sugars, producing organic
acids, ethanol, CO2, and H2(Gfeller and Gibbs, 1984;
Kreuzberg, 1984; Ohta et al., 1987). We and others are
developing Chlamydomonas as a model system to eluci-
date pathways and regulatory circuits associated with
fermentation metabolism in photosynthetic, eukaryotic
microbes.
Chlamydomonas shares some metabolic features with
both vascular plants and soil microbes. It relies on
glycolytic breakdown of carbohydrate reserves and
activation of fermentation pathways for generating the
energy required for survival during periods of oxygen
depletion (Gfeller and Gibbs, 1984; Kreuzberg, 1984;
Ohta et al., 1987). A number of these fermentation
pathways are typical of those present in various pro-
karyotes and some eukaryotes (Mus et al., 2007). Some
1This work was supported by the Office of Biological and Envi-
ronmental Research, Genome to Life program, Office of Science, U.S.
Department of Energy (grants to A.R.G., M.C. P., and M.S.), by the
National Science Foundation (grant no. MCB–0235878) and the U.S.
Department of Energy (grant no. DE–FG02–07ER64427) to A.R.G., by
the Air Force Office of Scientific Research (grant no. FA9550–11–1–
0211 to M.C.P.), by the Scuola Superiore Sant’Anna (to P.P. and L.M.),
by the Regione Toscana (Programma Operativo Regionale Obiettivo 2
Fondo Sociale Europeo to L.M.), and by the National Renewable
Energy Laboratory Pension Program (to M.S.). Work at the National
Renewable Energy Laboratory was performed under U.S. Depart-
ment of Energy contract number DE–AC36–08GO28308.
2Present address: Westfa ¨lische Wilhelms-Universita ¨t Mu ¨nster,
Institut fu ¨r Biologie und Biotechnologie der Pflanzen, Hindenburg-
platz 55, Munster 48143, Germany.
* Corresponding author; e-mail magneschi@sssup.it.
The author responsible for distribution of materials integral to the
findings presented in this article in accordance with the policy
described in the Instructions for Authors (www.plantphysiol.org) is:
Leonardo Magneschi (magneschi@sssup.it).
[W]The online version of this article contains Web-only data.
www.plantphysiol.org/cgi/doi/10.1104/pp.111.191569
Plant Physiology?, March 2012, Vol. 158, pp. 1293–1305, www.plantphysiol.org ? 2012 American Society of Plant Biologists. All Rights Reserved. 1293

Page 2

enzymes that function in these pathways include
pyruvate:ferredoxin oxidoreductase (PFR), pyruvate
decarboxylase (PDC), lactate dehydrogenase (LDH),
pyruvate formate lyase (PFL), alcohol dehydrogenase
(ADH), phosphate acetyltransferase (PAT), acetate ki-
nase (ACK), and the two [FeFe] hydrogenases (HYDA1
and HYDA2) and their maturation proteins, HYDG
and HYDEF (Posewitz et al., 2004; Atteia et al., 2006;
Ghirardi et al., 2007; Mus et al., 2007; Hemschemeier
et al., 2008; Grossman et al., 2011). The anaerobic activ-
itiesof these and otherenzymes result in the secretion of
organic acids (formate, lactate, malate, acetate, and
succinate) and alcohols (ethanol and glycerol) as well
as the evolution of H2and CO2(Gfeller and Gibbs, 1984;
Kreuzberg,1984;Ohtaetal.,1987;Tsygankovetal.,2002;
Kosourov et al., 2003; Mus et al., 2007; Dubini et al.,
2009).
When Chlamydomonas experiences dark anoxic con-
ditions, the starch reserves, which are generated as a
consequence of photosynthetic activity and stored in
the chloroplast, are degraded to sugars, which may
then be metabolized to pyruvate through glycolysis,
leading to the production of ATP. Reduced pyridine
nucleotides, cogenerated during this process, are re-
oxidized through the activities of several metabolic
pathways that use glycolytic intermediates, primarily
pyruvate, as the initial substrate (Fig. 1). Interactions
among these pathways and the mechanisms by which
they are regulated are still not completely understood.
Metabolites that are synthesized as cells ferment
sugars, and the pathways responsible for their pro-
duction in enteric bacteria have been known for many
years (Harden, 1901; Clark, 1989).Fermentative metab-
olism in Escherichia coli and many other bacteria ap-
pears to have significant flexibility, and glycolytic
NADH can be recycled during anaerobic metabolism
by synthesizing and secreting various reduced metab-
olites, including ethanol, lactate, and succinate. Acetate
is also generated as a consequence of fermentation, and
whileits synthesis frompyruvategenerates ATP,itdoes
not consume NADH. Recently, the flexibility associated
with anaerobic metabolism in Chlamydomonas has been
demonstrated through the generation and analyses of
several mutant strains blocked for specific branches of
Figure 1. Chlamydomonas fermentative pathways under dark anoxic conditions. In wild-type cells (black arrows), the major
fermentative products are formate, acetate, and ethanol, with CO2and H2emitted as minor products. The metabolic pathway
that leads to the fermentative production of succinate is unveiled in the hydEF-1 mutant (Dubini et al., 2009) and is depicted in
the figure in green. An increase in the production of lactate, which is almost undetectable in fermenting wild-typecells, has been
observed in the pfl1 mutants (Philipps et al., 2011; Catalanotti et al., 2012) and is highlighted in orange. ACK1, Acetate kinase
isoform 1; ACK2, acetate kinase isoform 2; ADH, alcohol dehydrogenase (ADH1, ADH2, or ADH3 could perform this reaction;
see text); ADH1, acetaldehyde/alcohol dehydrogenase; FDX, ferredoxin; FMR, fumarate reductase; FUM, fumarase; HYDA1 and
HYDA2, two putative hydrogenases; LDH, lactate dehydrogenase; MDH, malate dehydrogenase; MME4, malic enzyme; PAT1,
phosphate acetyltransferase isoform 1; PAT2, phosphate acetyltransferase isoform 2; PDC3, pyruvate decarboxylase; PEPC,
phosphenolpyruvate carboxylase; PFL1, pyruvate formate lyase; PFR1, pyruvate:ferredoxin oxidoreductase; PYC, pyruvate
carboxylase; PYK, pyruvate kinase. This figure was modified from Grossman et al. (2011).
Magneschi et al.
1294Plant Physiol. Vol. 158, 2012

Page 3

fermentation metabolism(Dubinietal.,2009;Grossman
et al., 2011; Philipps et al., 2011; Catalanotti et al., 2012).
For example, in the Chlamydomonas hydEF-1 mutant
(Dubini et al., 2009), pyruvate metabolism is redirected
to the reverse tricarboxylic acid reactions, while in the
pfl1mutant(Philippsetal.,2011;Catalanottietal.,2012),
there is a marked increase in lactate production and a
smaller increase in ethanol synthesis; both of these
metabolites are linked to NADH reoxidation (Fig. 1).
The pfl1 mutant cells also accumulate elevated intracel-
lular levels of amino acids, which may help recycle
NADH and limit the potentially damaging conse-
quences of pyruvate accumulation (Gupta et al., 2009;
Zabalza et al., 2009). Aspects of electron rerouting
observed in the Chlamydomonas pfl1 strain are similar
to those observed for analogous E. coli mutants (Clark,
1989; Zhu and Shimizu, 2005), suggesting that the
pathways and potentially some of the compensatory
mechanisms are conserved between bacteria and eu-
karyotic green algae.
Under the fermentation conditions used, ethanol
accounts for about one-fourth of the metabolites syn-
thesized and excreted by Chlamydomonas cells during
dark anoxia (Gfeller and Gibbs, 1984; Kreuzberg, 1984;
Ohta et al., 1987). Cytoplasmic PDC and ADH activ-
ities typically ferment pyruvate to ethanol in terrestrial
plants (Dolferus et al., 1985; Mu ¨cke et al., 1995;
Ku ¨rsteiner et al., 2003; for review, see Magneschi and
Perata, 2009). PDC catalyzes the decarboxylation of
pyruvate to acetaldehyde and CO2, while ADH re-
duces acetaldehyde to ethanol (Perata and Alpi, 1991;
Perata et al., 1992), which is excreted from the orga-
nism. Chlamydomonas possesses three distinct enzymes
that are potentially important for ethanol production
when the cells become anoxic: ADH1 (a putative dual-
function alcohol/acetaldehyde dehydrogenase, previ-
ously annotated as ADHE because of its homology
with the E. coli AdhE protein [Mus et al., 2007;
Hemschemeier et al., 2008]; Joint Genome Institute
[JGI] version 4.0 protein identifier 133318, Augustus
version 5.0 protein identifier 518335) and two other
putative alcohol dehydrogenases that we designate
ADH2 (JGI version 4.0 protein identifier 121409, Au-
gustus version 5.0 protein identifier 516421) and
ADH3 (JGI version 4.0 protein identifier 82021, Au-
gustus version 5.0 protein identifier 516422). Analysis
of protein sequences by InterProScan (http://www.
ebi.ac.uk/Tools/pfa/iprscan/) suggests that both the
alcohol and the aldehyde dehydrogenase domains are
present in ADH2, whereas ADH3 possesses only the
alcohol dehydrogenase domain. In the Enterobacteri-
aceae, the AdhE enzyme represents the major route for
recycling NADH during fermentation (Clark and
Cronan, 1980; Leonardo et al., 1996), as highlighted by
the inability of an E. coli adhE mutant to grow on mini-
mal medium under anoxic conditions (Cunningham
and Clark, 1986; Gupta and Clark, 1989). Interestingly,
an AdhE/ADH1 homolog is not present in the majority
of prokaryotes, and among the eukaryotes, it has only
been identified in a few amitochondriate protists and
some green algae (Atteia et al., 2003, 2006). In the alga
Polytomella, the ADHE protein was localized to the
mitochondrion (Atteia et al., 2003), while in Chlamydo-
monas, ADH1 was found to be present in chloroplasts
(Terashima et al., 2010). This difference in subcellular
location, which could be explained by differences in
targeting sequences located at the N terminus of the
enzyme, may create differences in the intracellular traf-
ficking of metabolites. Furthermore, the Polytomella
ADHE protein lacks the conserved His residues in the
ADH-IRON2 signature of the enzyme that are respon-
sible for metal-catalyzed activation and oxygen sensitiv-
ity (Atteia et al., 2003; Supplemental Fig. S1). In contrast,
Chlamydomonas ADH1 has retained the His-containing
domain, which is also present in E. coli AdhE. These
findings suggest that although Chlamydomonas ADH1
accumulates both at the transcript and protein levels
under aerobic conditions (Whitney et al., 2011; this
work), the protein might not be enzymatically active in
an oxygen-containing atmosphere.
Based on sequence similarities with E. coli AdhE, the
dual-function Chlamydomonas ADH1 protein has been
suggested to operate downstream of PFL1, where it
would reduce acetyl-CoA to acetaldehyde and then to
ethanol, resulting in the regeneration of two molecules
of NAD+per molecule of acetyl-CoA (Mus et al., 2007;
Terashima et al., 2010; Grossman et al., 2011; Fig. 1).
ADH1 transcript levels rise when Chlamydomonas cells
are exposed to anoxic conditions (Mus et al., 2007; this
work), and anaerobiosis caused a small increase in the
level of the ADH1 protein (Terashima et al., 2010). The
PDC (designated PDC3 in Chlamydomonas) pathway is
also proposed to generate ethanol in Chlamydomonas
(Fig.1).Quantitativeproteomics-basedlocalizationdata
have shown that the metabolism of pyruvate by the
PFL1, PFR1, ADH1, and PAT2/ACK1 pathways occurs
in Chlamydomonas chloroplasts, with parallel PFL1 and
PAT1/ACK2 activities outside of the chloroplast, most
likely in mitochondria (Terashima et al., 2010).
If ADH1 is responsible for ethanol production in the
chloroplasts of Chlamydomonas, suppression of this
activity could potentially lead to an increase in reduc-
ing equivalents in the chloroplast and elicit elevated
hydrogenase activity, which would serve as an elec-
tron valve. In E. coli, adhE mutants exhibit increased
acetate production and lower lactate/formate levels
under microaerobic conditions, consistent with a
down-regulation of both PFL and LDH activities in
the mutant strain (Zhu and Shimizu, 2005). At this
point, there is no experimental evidence that defines
the participation of the three potential ADH proteins
of Chlamydomonas in fermentative ethanol production.
To address this issue, we exploited an insertional
mutagenesis- and PCR-based reverse genetic screen
(Pootakham et al., 2010; Gonzalez-Ballester et al., 2011)
and identified a Chlamydomonas adh1 mutant. Charac-
terization of this mutant has allowed us to evaluate the
impact of this enzyme on algal cells experiencing dark,
anoxic conditions and to expand our understanding of
the intricate relationships among the metabolic circuits
Anaerobic Metabolism in Chlamydomonas adh1
Plant Physiol. Vol. 158, 20121295

Page 4

associated with pyruvate fermentation as well as the
mechanisms associated with their control.
RESULTS
Identification of the Chlamydomonas adh1 Mutant
To elucidate the function of ADH1 in Chlamydomonas
metabolism, we generated an adh1 insertional mutant.
In this strain, the aminoglycoside 3’-phosphotransferase
(AphVIII) marker gene is inserted into the 15th exon
of the coding region of ADH1 (Fig. 2A) in an inverted
orientation relative to the ADH1 coding sequence. The
insertion results in the synthesis of a mutated protein
that is 43 amino acids shorter than the wild-type
protein (911 compared with 954 amino acids), with
the last 18 amino acids encoded by the fused marker
gene sequence. Most of this aberrant polypeptide is
likely rapidly degraded, as we were unable to detect
the ADH1 protein by western-blot analyses of total
proteins extracted from adh1 mutant cells grown under
either oxic or anoxic conditions (Fig. 2B). There is little
change in the level of the ADH1 transcript, based on
reverse transcription and real-time quantitative (RT-q)
PCR, in mutant relative to wild-type cells (see below),
although the transcript would be aberrant in the
mutant. The transformant analyzed here has a single
copy of the AphVIII cassette integrated into its genome,
as determined by Southern-blot hybridizations using
the AphVIII cassette as the probe (Fig. 2C) as well as by
segregation of the marker gene (always 1:1 segregation
of paromomycin-resistant and -tolerant cells, with
cosegregation of the resistant phenotype with the
insertion in ADH1, suggesting the presence of a single
marker gene insertion). The background bands on the
Southern blot, which are the same in the wild-type
(D66) and mutant strains, represent hybridizations to
the PSAD promoter and to the CYTc6 3# end, both of
which are part of the AphVIII cassette.
A wild-type copy of the ADH1 coding sequence
under the control of the PSAD promoter was trans-
formed into the adh1 mutant. The transformant was
rescued for ADH1 protein accumulation (Fig. 2B).
Furthermore, for both the rescued adh1 mutant and
wild-type cells, the level of ADH1 protein was essen-
tially the same during growth under aerobic and
anoxic conditions (Fig. 2B).
Fitness of adh1 following Exposure to Anoxic Conditions
We investigated the fitness of the adh1 mutant to
anoxic conditions. Reduction in plating efficiency for
both mutant and wild-type cells was first noted 12 h
after the shift to anoxic conditions, and by 18 h
following this shift, most of the cells did not recover.
Interestingly, the mutant shows a similar tolerance to
anoxia as wild-type cells (Fig. 3). The similar behavior
of mutant and wild-type cells suggests that the mutant
may be able to compensate for the loss of ADH1
activity, possibly as a consequence of metabolic ad-
justments, although there might still be a difference in
the fitness of these strains when they compete in their
natural environment.
Transcript and Protein Abundances for Genes Encoding
Fermentative Enzymes
To investigate how an inability to make ADH1
impacts the overall molecular responses of the cell to
anoxia, we analyzed the levels of transcripts encoding
key fermentative enzymes in mutant and wild-type
Figure 2. A Chlamydomonas mutant with an insertion in ADH1. A,
Schematic representation of the insertion of the AphVIII cassette into
ADH1. Note that the cassette is inserted in an inverted orientation
relative to the ADH1 gene. Insertion of the paromomycin cassette adds
18 amino acids to the ADH1 proteinsequencebeforethe occurrenceof
a translational stop codon; the final size of the protein is 954 amino
acids in wild-type cells and 911 amino acids in the adh1 mutant (893
amino acids from the wild-type ADH1 protein sequence and 18 amino
acids from the noncoding strand of the paromomycin cassette). UTR,
Untranslated region. B, The ADH1 protein cannot be detected in the
mutant strain (adh1 in the CC-124 background) based on immunoblot
analyses (with antibodies raised against the amino acid sequence
SGTGSEVTPFSVVTD, which is upstream of the site of insertion) but is
detected in wild-type CC-124 cells (WT) and in the strain rescued by
introduction of the ADH1 coding sequence under the control of the
PSADpromoter(adh1-ADH1). Thispromoteris able to drive expression
of the reintroduced ADH1 coding sequence under aerobic (Air) and
anoxic (Anoxia; 4 h) conditions. Moreover, accumulation of the ADH1
protein in the rescued strain is stable over a 6-h anaerobic treatment
(data not shown). The numbers below the panels indicate the relative
amount of protein, with 100% corresponding to 45 mg of total protein
loaded. C, Southern-blot analysis of genomic DNA from the original
parental wild-type strain (D66) and the adh1 mutant (in the original
D66 background) using the AphVIII cassette as a hybridization probe.
The genomic DNA was digested with PstI and PvuII, as indicated.
Arrows highlight the DNA band with the inserted AphVIII cassette;
there appears to be only one copy of the cassette inserted into the adh1
mutant genome.
Magneschi et al.
1296Plant Physiol. Vol. 158, 2012

Page 5

cells. As shown in Figure 4, the levels of nearly all of
the transcripts tested (except for PAT1 and PDC3)
increased in wild-type cells exposed to anoxic condi-
tions. In most cases, the mutant showed a similar
response to that of wild-type cells. The level of ADH1
mRNA in the adh1 mutant (analyzed with primers
designed upstream of the site of insertion) appeared to
be approximately the same as in wild-type cells ex-
posed to oxic or anoxic conditions, in spite of an
interruption within the coding region of the gene. The
levels of transcripts from PFL1, PFR1, PDC3, HYDA1,
HYDA2, PAT1, PAT2, ACK1, and ACK2 were either the
same in wild-type and adh1 mutant cells or, in some
cases (PAT2 and especially PFR1), significantly ele-
vated in the adh1 mutant relative to wild-type cells
(Fig. 4).
The change in the level of a specific transcript did
not always correlate with a change in the level of its
encoded protein. The ADH1 protein was not detected
in the adh1 mutant under oxic or anoxic conditions by
antibodies that recognize a sequence upstream of the
site of insertion (Fig. 5), even though transcript abun-
dance was high. This is not surprising, since it is not
unusual for aberrant proteins, such as the truncated
Figure 3. Anoxia tolerance of the wild type (WT) and the adh1 mutant.
The adh1 and CC-124 (wild-type) cells were grown under mixotrophic
conditions in TAP medium until they reached midlog phase (around
2 3 106cells mL21). Cells were then concentrated 10 times by centrif-
ugation and resuspension in one-tenth of the initial volume of AIB
buffer (Ghirardi et al., 1997). Culture concentrations were determined
by counting the cells in aliquots of the culture using a hemocytometer;
the cells, collected as five different aliquots from each sample, were
counted and then averaged. For all samples, cell densities were
equalized to 1.5 3 107cells mL21(the final cultures were also counted
three times to check for pipetting errors). Cultures were then subjected
to dark anoxic treatment by purging the medium with argon for 30 min
and incubating them for several hours inside an anaerobic workstation
chamber. At specific time points (0, 3, 6, 9, 12, 15, and 18 h), 5 mL of
culture was spotted onto the surface of TAP agar medium to allow
viable cells to recover under aerobic conditions for 5 d to 1 week from
the anoxic treatment. A light control (T0; cells spotted at 0 h of
treatment and left under light aerobic conditions for 1 week) was also
included to check for differences in the initial concentration of plated
cells. As diagrammed in the top left petri dish with growing colonies,
we performed two serial 1:10 dilutions of the treated cell cultures (with
AIB) prior to plating them onto the solid medium; the left side of the
petri dish always shows the growth of wild-type cells, while the right
side always shows the growth of adh1 mutant cells. Different spots
within each column are technical replicates of the same dilution.
Repetition of the experiment with two biological replicates yielded
similar results.
Figure 4. Levels of transcripts in wild-type (WT) and adh1 mutant cells
following exposureof culturesto anaerobicconditions(0, 0.5, 2, 4, and
6 h). The levels of transcripts encoding enzymes of the fermentative
pathways were analyzed by RT-qPCR, using absolute quantification of
the results that were normalized to transcript abundance at 0 h from
CC-124(wildtype[Steunouet al.,2006];transcriptlevelsattime0were
arbitrarily made 1.0 for each of the tested transcripts). Data are means of
two biological replicates, each with three technical replicates 6 SD. The
gene/protein names are as in Figure 1.
Anaerobic Metabolism in Chlamydomonas adh1
Plant Physiol. Vol. 158, 20121297

Page 6

ADH1 protein of the mutant cells, to be rapidly
degraded (Preiss et al., 2001). Western-blot analyses
were not sensitive enough to detect the 1.5-fold in-
crease in ADH1 protein after wild-type Chlamydomo-
nas cells were transferred to anoxic conditions, which
was reported previously (Terashima et al., 2010).
However, while many transcripts, including those
encoding HYDA1, HYDA2, PFL1, ACK1, and PAT2,
increased significantly as the cells became anoxic,
significant levels of the encoded proteins were already
present in oxic cultures, and shifting to anoxic condi-
tions caused little change in their abundances. How-
ever, it should be noted that the HYDA antibodies
recognize both HYDA1 and HYDA2 and that the
antibodies raised against PAT2 and ACK1 may also
detect the PAT1 and ACK2 isoenzymes (Supplemental
Fig. S2); although the predicted molecular sizes of the
ACK proteins are different enough to be separated by
SDS-PAGE (38 kD for ACK2 and 45 kD for ACK1), PAT
isoforms have almost identical predicted masses, and
even if the peptide sequence used to generate the PAT2
antibodies is highly modified in PAT1, the antibody
might not be specific for PAT2. Furthermore, the
mutant and wild-type cells exhibited similar levels of
these proteins. Interestingly, the PFR1 transcript accu-
mulated to high levels in mutant cells, with a signif-
icantly lower level of accumulation in wild-type cells
following a shift to anoxic conditions (Fig. 4). Also,
while PFR1 protein levels in wild-type cells appear to
be very low under oxic conditions and increased when
the cells were exposed to anoxia, the adh1 mutant had
relatively high levels of PFR1 prior to exposing the
cells to anoxic conditions, and the level remained high
over the entire anoxic period (Fig. 5). These results
demonstrated that PFR1 transcript levels were not
strictly coupled to protein production in the adh1
mutant strain, suggesting that posttranscriptional pro-
cesses (e.g. translation control, protein degradation)
may have a strong impact, at least in some cases, on
the final protein levels.
It has been suggested that the chloroplast-localized
PAT2/ACK1 and ADH1 pathways compete for the
substrate acetyl-CoA under conditions of oxygen dep-
rivation (Mus et al., 2007; Terashima et al., 2010;
Grossman et al., 2011). The PAT2/ACK1 pathway
leads to the synthesis of acetate and ATP, while the
ADH1 pathway would lead to the synthesis of ethanol,
regenerating two NAD+(from two NADH) for each
ethanol produced (Fig. 1). We examined the levels of
the PATand ACK proteins using antibodies generated
to PAT2 and ACK1; however, as noted above, it is not
clear that they are specific enough to distinguish
between the two isoforms (chloroplast and likely
mitochondrial) of these enzymes, especially between
PAT2 and PAT1, which have nearly identical molecu-
lar masses (Supplemental Fig. S2). In our analyses, the
PATand ACK proteins accumulated to a similar extent
in mutant and wild-type strains (Fig. 5). LDH and
PDC3 are likely not present in chloroplasts (Terashima
et al., 2010) and were not analyzed.
Extracellular Metabolite Production and H2Evolution
To directly evaluate the impact of the lack of ADH1
on changes in the activity of the various branches of
fermentation metabolism, we analyzed the accumula-
tion of metabolites excreted into the medium when
wild-type and adh1 cells transitioned from oxic to
anoxic conditions. Ethanol production was completely
abolished in the adh1 strain (Fig. 6A), suggesting that
the strain was indeed null for ADH1 activity, which is
congruent with protein accumulation data. Further-
more, while PFR1 protein levels were shown to be
significantly higher in the mutant than in wild-type
cells (especially during the early stages of anoxia), the
mutant showed little change in H2production under
dark fermentative conditions (Fig. 6B). It is plausible
that reduced ferredoxin generated by the PFR1 reac-
tion could reduce substrates other than protons (e.g.
sulfate, nitrite). The adh1-ADH1 strain (wild-type copy
of ADH1 introduced into the mutant) is rescued for the
production of ethanol (and other metabolites; Fig. 6)
and makes somewhat more H2than wild-type cells
(Fig. 6B). While acetate and formate levels increased in
both the adh1 mutant and wild-type cells following a
shift to anoxic conditions (Fig. 6A), acetate accumu-
Figure 5. Western-blot analysis of fermentative enzymes in wild-type
(WT) and adh1 mutant cells. Amido black staining of the polyvinyli-
dene difluoride membrane (bottom) is shown as a loading control. The
time following the transfer of cells to anoxic conditions is given at the
bottom. Note that the level of PFR1 is markedly up-regulated in the
adh1 mutant, while the levels of most other proteins are similar or
slightly up-regulated (e.g. PAT2 and HYDA) in adh1 relative to CC-124
(wild-type) cells. The gene/protein names are as in Figure 1. Repetition
of the experiment with two independent biological replicates yielded
essentially identical results.
Magneschi et al.
1298 Plant Physiol. Vol. 158, 2012

Page 7

lated to a significantly higher level in adh1 than in
wild-type cells, while the concentration of formate was
much higher in wild-type cells following the shift to
anoxic conditions (Fig. 6A). Furthermore, there was
essentially no CO2production in the mutant strain
(Fig. 6B). CO2production is most likely the conse-
quence of PDC3 activity, which would generate acet-
aldehyde that would have to be reduced to ethanol by
ADH or, potentially, from the PFR1 reaction if protons
(through hydrogenase) are not the ultimate electron
acceptor. The inability of the adh1 mutant to accumu-
late ethanol and CO2, with low levels of formate
production, suggests that the acetaldehyde produced
by PDC3 and the acetyl-CoA produced by PFL1 and
PFR1 are not readily reduced in mutant cells exposed
to anoxic conditions and that under such conditions
the activities of PFL1 and PDC3, the major producers
of formate and CO2, respectively, may decline. These
findings also strongly suggest that ADH1 is the only
putative alcohol dehydrogenase in Chlamydomonas
cells that is capable of reducing acetyl-CoA or acetal-
dehyde under the conditions used in this study.
Changes in fermentative metabolism in adh1 cells
would be critical to limit the generation and buildup of
substrates normally acted on by ADH1, allowing the
mutant to accommodate the block in ethanol produc-
tion. However, it would also be necessary for the
mutant to eliminate reducing equivalents that are
generated during the glycolytic production of ATP.
The adh1 strain did show some extracellular lactate
accumulation, which was not observed in wild-type
cells (Fig. 6C); however, a significantly larger increase
in lactate accumulation was observed in strains lack-
ing PFL1 (Philipps et al., 2011; Catalanotti et al., 2012).
Furthermore, transcript levels for LDH were identical
in adh1 and wild-type cells over the entire anoxic time
course (Fig. 7). Most interesting was the finding that
the medium of the adh1 mutant contained high levels
of glycerol relative to wild-type cells when the cultures
became anoxic (Fig. 6C); internal metabolite analysis
also showed higher intracellular glycerol levels in adh1
relative to the wild-type cells (Supplemental Table S2).
Extracellular lactate and glycerol do not accumulate in
the rescued strain (Fig. 6C). Glycerol is synthesized
from dihydroxyacetone phosphate (DHAP). This me-
tabolite precedes the formation of pyruvate and the 3C
oxidation (NADH formation) step in glycolysis, and
glycerol synthesis also effectively recycles one NADH.
Hence, the production of glycerol and lactate in
the adh1 mutant, as highlighted in the diagram of
fermentation metabolism presented in Figure 8, would
allow for the efficient recycling of NADH, an activity
critical for maintaining redox balance and sustaining
glycolytic production of ATP, even though the cells
are unable to reduce acetaldehyde or acetyl-CoA to
ethanol.
TheconversionofDHAPintosn-glycerol-3-phosphate,
a metabolic intermediate in glycerol synthesis, occurs
through the activity of the enzyme sn-glycerol-3-phos-
phate dehydrogenase (GPD); sn-glycerol-3-phosphate
synthesis and GPD activities were shown to be present
in isolated Chlamydomonas chloroplasts during plastidic
starch degradation in the dark (Klo ¨ck and Kreuzberg,
1989). The Chlamydomonas genome contains five genes
encodingputativeGPDenzymes;wenamedthesegenes
GPD1 (Augustus version 5.0 identifier 513084), GPD2
(Augustus version 5.0 identifier 511717), GPD3 (Augus-
tus version 5.0 identifier 511720), GPD4 (Augustus ver-
sion 5.0 identifier 509652), and GPD5 (Augustus version
5.0 identifier 343023). We also investigated whether
the observed accumulation of glycerol in adh1 mutant
Figure6. Metabolitelevelsin the wild type (WT), the adh1 mutant,and
the adh1-ADH1 rescued strain under anoxic conditions. A, External
levels of ethanol,formate, and acetate in CC-124 (wild type), adh1, and
adh1-ADH1. B, FermentativeH2productionin the wild type,adh1, and
adh1-ADH1 (top) and fermentative CO2evolution in the wild type,
adh1, and adh1-ADH1 (bottom). The CO2data were calculated as the
difference between the 30-min (anoxia) and time x (anoxia) values for
the different x data points. C, Extracellular lactate accumulation in the
wild type, adh1, and adh1-ADH1 (left) and extracellular glycerol
accumulation in the wild type, adh1, and adh1-ADH1 (right). The data
for all of the experiments presented in this figure were generated at 0.5,
2, 4, and 6 h following the imposition of anoxia and are expressed as
means of three biological replicates 6 SD. Note that wild-type cells and
the rescued strain show neither lactate nor glycerol accumulation. In
some instances, the error bars are so small that they are not seen (some
external metabolite and CO2emission data).
Anaerobic Metabolism in Chlamydomonas adh1
Plant Physiol. Vol. 158, 20121299

Page 8

cells was linked to increased levels of transcripts
encoding the GPDs (Fig. 7). While there was substan-
tial variability in the data, we observe that the levels of
some of GPD transcripts increased in wild-type and
mutant cells following the transition from oxic to
anoxic conditions (GPD3 and GPD5). Also, there do
appear to be slightly higher levels of the GPD4 tran-
script in the mutant relative to wild-type cells (al-
though in both, the transcript levels decline). More
work is required to determine if the level of GPD
activity changes as the cells become anoxic and the
roles of the different isozymes in glycerol production.
The subsequent conversion of sn-glycerol-3-phosphate
to glycerol requires either the enzyme glycerol kinase
(GK) or glycerol 3-phosphate phosphatase (GPP), both
encoded by single genes in the Chlamydomonas ge-
nome. Transcript abundances for both GK1 (Augustus
version 5.0 identifier 522152) and GPP (Augustus
version 5.0 identifier 519809) are comparable in adh1
and wild-type cells experiencing anoxia (Fig. 7).
DISCUSSION
To elucidate the physiological role of Chlamydomonas
ADH1, the E. coli AdhE homolog, we generated and
characterized an adh1 insertional mutant. Analysis of
this mutant suggests that ADH1 in this unicellular
alga catalyzes the dominant activity that reoxidizes
NADH under anoxic conditions, a key reaction that
allows glycolysis to continue to operate and to gener-
ate the energy needed for survival when cells lack
oxygen as a terminal electron acceptor. The adh1 mu-
tant cannot synthesize ethanol (below the detection
limit under the conditions that we are using), which
Figure 7. Levels of transcripts encoding LDH,
GPD (GPD1, GPD2, GPD3, GPD4, and GPD5),
GK1, and GPP in CC-124 (wild-type [WT]) and
adh1 cells following exposure of cultures to an-
aerobic conditions (0, 0.5, 2, 4, and 6 h). The
levels of transcripts were analyzed by RT-qPCR
using absolute quantification of results that were
normalized to transcript abundance at 0 h from
the wild type, as performed previously (Steunou
et al., 2006). Data are means of two biological
replicateseachwiththreetechnicalreplicates6 SD.
Figure 8. Pathways responsible for metabolite
accumulation in wild-typeand adh1mutantcells.
Red arrows indicate pathways that appear to be
up-regulated, and blue arrows indicate pathways
that appear to be down-regulated, in the adh1
mutant relative to wild-type cells. Black arrows
indicate reactions that do not appear to be sig-
nificantly altered in mutant relative to wild-type
cells. Gray arrows show pathways for which a
product was not detected in adh1 mutant cultures
but is normally detected in wild-type cultures.
This schematic is based on metabolite abun-
dances only. The gene/protein names are as in
Figure 1. This figure was modified from Grossman
et al. (2011).
Magneschi et al.
1300Plant Physiol. Vol. 158, 2012

Page 9

strongly suggests that most ethanol produced during
anoxia is a consequence of ADH1 activity and that the
other potential ADH proteins in Chlamydomonas can-
not compensate for this loss (either because they are
not able to reduce acetaldehyde and acetyl-CoA or
because of their different subcellular localizations or
expression patterns). Indeed, protein sequence analy-
sis reveals that the domain required for aldehyde
dehydrogenase activity is missing in ADH3 and that
there are modifications in ADH2 of conserved se-
quences associated with the first putative nucleotide-
binding domain and the catalytic center of the protein
(Supplemental Fig. S1). This catalytic center is con-
served in all CoA-dependent and CoA-independent
aldehyde dehydrogenases (Atteia et al., 2003). Our
results also strongly suggest a critical coupling of the
output of the PDC3 and PFL1 reactions (acetaldehyde
and acetyl-CoA, respectively) in Chlamydomonas with
ADH1 activity. A similar coupling between PFL and
AdhE in E. coli has been reported (for review, see
Clark, 1989). Although other pathways in Chlamydo-
monas may generate acetaldehyde or acetyl-CoA (sub-
strates for ADH1 and ethanol production), elevated
CO2evolution in sodium hypophosphite-treated cul-
tures (in which PFL activity is irreversibly blocked;
Knappe et al., 1984; Plaga et al., 1988) implicates
PDC3/ADH coupled reactions in the production of
ethanol (Hemschemeier et al., 2008; Catalanotti et al.,
2012). ADH1 and/or one of the other two putative
ADH proteins in Chlamydomonas could potentially
work in conjunction with PDC3 (previously referred
to as PDC1 and PDC by Mus et al. [2007] and
Terashima et al. [2010], respectively) to help manage
intracellular redox conditions during anaerobiosis,
especially when PFL1 activity is reduced or impaired.
However, the work presented here suggests that it is
predominantly ADH1 that catalyzes the reduction of
the acetaldehyde generated by PDC3 (Catalanotti
et al., 2012; this paper), at least under the conditions
used in our experiments. This coupling raises some
issues concerning the way in which ADH1 accesses
acetaldehyde, since the ADH1 protein was found
exclusively in chloroplasts of Chlamydomonas, while
PDC3 (which has no apparent transit peptide) was
putatively cytoplasmic; acetaldehyde would have to
traffic into plastids to be reduced to ethanol. The
function(s) of the other putative ADH activities in
wild-type Chlamydomonas cells is not clear, although
they may have significant roles in fermentation and
ethanol synthesis under specific environmental condi-
tions, or they may represent NADH aldehyde dehy-
drogenases with unique specificities.
It is noteworthy that E. coli mutants defective for
AdhE do not grow anaerobically on Glc, even though
the strains do have LDH activity. This growth defect
may reflect the fact that any acetyl-CoA that is gener-
ated by PFL in adhE mutants can only be metabolized
through the PAT/ACK pathway. While this would
lead to the production of an extra 1 mol ATP mol21
Glc, it would not allow for oxidative recycling of
NADH, which is critical for sustaining glycolysis
and ATP production. Similarly, vascular plants that
lack ADH activity are also more sensitive to anoxic
conditions (Jacobs et al., 1988; Saika et al., 2006). These
results demonstrate the critical nature of ADH activity
in maintaining energy/redox balance and sustained vi-
ability when organisms are exposed to anoxic conditions.
Our results show that Chlamydomonas is able to
adjust to the lack of the ADH1 protein, thus surviving
equally well as wild-type cells under anoxic condi-
tions (Fig. 3); the anoxic tolerance of the adh1 mutant
might reflect specific compensatory metabolic changes.
Indeed, in addition to eliminating the production of
ethanol, the ADH1 gene disruption elicits marked
alterations in the levels of a number of secreted me-
tabolites, including acetate, formate, lactate, and glyc-
erol. Previous work has shown that blocking various
reactions associated with fermentation metabolism
can cause changes in the flow of metabolites through
other branches of the fermentative network. For ex-
ample, Chlamydomonas strains lacking PFL1 (Philipps
et al., 2011; Catalanotti et al., 2012) compensate for the
loss of this activity by increasing the synthesis of
lactate through the activity of LDH, an enzyme that in
wild-type cells would compete with PFL1 for the
substrate pyruvate. Hence, a proportion of the pyru-
vate that accumulates in the pfl1 mutant would be
metabolized by LDH, resulting in the accumulation of
lactate and the reoxidation of 1 mol of NADH, which
would help sustain the catabolism of polysaccharides
and the synthesis of ATP in the mutant (Fig. 1).
Similarly, an E. coli pfl mutant can grow anaerobically
on acetate by maintaining energy production by me-
tabolizing pyruvate to lactate (Clark,1989). Indeed,the
LDH enzyme of E. coli has been reported to be allo-
sterically regulated, with elevated activity associated
with increased pyruvate concentrations (Tarmy and
Kaplan, 1968), suggesting that this enzyme has an
overspill function (Clark, 1989). Another example of
the metabolic flexibility of Chlamydomonas and alter-
native strategies to sustain ATP production when a
specific branch of anaerobic metabolism is blocked
comes from studies of the hydEF-1 mutant (Dubini
et al., 2009). This mutant can no longer eliminate
reducing equivalents through H2synthesis, although
it appears to sustain anaerobic ATP production by the
activation of reverse tricarboxylic acid reactions.
These reactions consume reducing equivalents and
generate succinate, which can be exported from the
cell (Fig. 1).
Analogous to the consequences of metabolic blocks
that have been reported in previous studies (Dubini
et al., 2009; Philipps et al., 2011; Catalanotti et al., 2012),
a block in the conversion of acetaldehyde and acetyl-
CoA to ethanol leads to the rerouting of metabolites to
other pathways in the fermentation network; some of
these pathways are marginally active in wild-type
cells. For the adh1 mutant placed under anoxic condi-
tions, the level of extracellular formate is decreased
while the level of acetate is increased relative to wild-
Anaerobic Metabolism in Chlamydomonas adh1
Plant Physiol. Vol. 158, 2012 1301

Page 10

type cells. This suggests a reduction in PFL1 activity
and a greater flux of acetyl-CoA through the PAT2/
ACK1 pathway, which is NADH neutral. Fermentative
H2production is not altered in adh1 cells, in spite of the
increases observed in PFR1 transcript and protein
levels. These results suggest that limitations in the
production of H2are not at the level of PFR1 and/or
that other regulatory/metabolic features of the system
limit the flow of electrons to the hydrogenase in the
adh1 mutant.
A major rerouting of metabolism in the adh1 mutant
is reflected by the accumulation of the extracellular
metabolites lactate andglycerol, which are notdetected
in cultures of anoxic wild-type cells (Fig. 6C). On the
otherhand,CO2evolution is decreased(Fig.6B),which
likely reflects metabolite redirection away from PDC3,
regardless of the level of PDC3 in the cells. The
evolutionary logic in this metabolic rewiring requires
careful consideration of the functioning of glycolysis
under anoxic conditions and the consequences of
eliminating a major reaction critical for the recycling
of reduced pyridine nucleotide. Glycolytic reactions
convert Glc to two molecules of pyruvate, and in the
process, it produces two NADH molecules that must
be reoxidized to sustain continued glycolytic ATP
formation (which is needed to maintain cell viability).
The fermentation profile observed in the adh1 mutant
indicates that approximately half of the glycolytic flux
in the mutant is being converted to glycerol and half to
pyruvate (with the pyruvate converted mostly to for-
mate/acetate and lactate). Glycerol is metabolically
derived from DHAP, a 3C glycolytic intermediate that
precedes NADH production in glycolysis. Addition-
ally, one NADH is oxidized by GPD in the conversion
of DHAP to sn-glycerol-3-phosphate, an intermediate
in the formation of glycerol. In summary, in the glyco-
lytic breakdown of a 6C sugar in the adh1 mutant, the
diversion of approximately one DHAP (3C) to glycerol
synthesis eliminates the production of an NADH, and
the NADH that is formed in the conversion of the other
DHAP to pyruvate would be oxidized in the synthesis
of glycerol, resulting in a near net zero production of
reduced pyridine nucleotide (Fig. 8), although some
NADH would also be recycled through the LDH
reaction. This rerouting of carbon flow would dramat-
ically reduce the need for the cells to reoxidize NADH
(they would not generate as much), which would
allow much of the acetyl-CoA synthesized during the
fermentative breakdown of pyruvate to be metabo-
lized to acetate in a reaction that does not require
reductant but that does synthesize an ATP in addition
to those generated by glycolysis. This change in met-
abolic flux is reflected by the appearance of glycerol
(Fig. 6C) and increased acetate accumulation (Fig. 6A)
in the medium of the adh1 mutant relative to wild-type
cells.
An increase in the levels of some GPD transcripts
(GDP3 and GDP5) does occur in both wild-type cells
and the adh1 mutant as the cells transition from aerobic
to anoxic conditions (Fig. 7). Whether these increases
in transcript levels are necessary for the synthesis of
the extracellular glycerol that accumulates in the mu-
tant strain has yet to be established (they are clearly
not sufficient for the increased glycerol production,
since they occur in both wild-type and mutant cells). In
the green halotolerant alga Dunaliella tertiolecta, it was
shown that neither de novo protein synthesis nor
covalent modification of GPD is involved in glycerol
production in response to hyperosmotic shock (Belmans
and van Laere, 1987; Sadka et al., 1989).
To our knowledge, this work represents the first
evidence for glycerol accumulation in Chlamydomonas
under anaerobic conditions when alternative NADH-
oxidizing pathways are disrupted. In order to more
thoroughly understand the fermentation circuits and
their regulation in Chlamydomonas, it is important to
generate additional mutants (single, double, and triple
mutants) to characterize how specific lesions alter both
internal and external metabolite pools, to define the
catalytic features of the various enzymes associated
with fermentation metabolism and their potential in-
teractions with each other, and to understand how
redox and metabolite levels modulate the activities of
the various fermentative pathways. It is also critical to
define the different compartments that house the
glycolytic reactions, pyruvate metabolism, and the
fate of the various pyruvate breakdown products.
Interestingly, the significant similarity of Chlamydomo-
nas fermentation to various aspects of fermentation in
the Enterobacteriaceae, such as E. coli, raises the ques-
tion of when and how these metabolic pathways were
acquired by green algae. Recently, it was suggested
that a horizontal gene transfer that occurred after the
divergenceoftheprimaryendosymbioticalgallineages
is likely responsible for the presence of eubacteria-type
Gln synthethase II in Chloroplastida such as Chlamy-
domonas (Ghoshroy et al., 2010). Therefore, some genes
encoding enzymes for key metabolic reactions in the
green algal lineage, and that function within chloro-
plasts, may have originated as a consequence of a
lateral gene transfer. In some cases, the transferred
genes are likely to have come from g-proteobacteria
(Ghoshroy et al., 2010). Chlamydomonas chloroplasts
were reported to possess all the enzymes necessary
for the conversion of Glc-6-P to CO2and water under
dark conditions (Chen and Gibbs, 1991). The initial
reactions of glycolysis (Glc-6-P to glyceraldehyde-3-
phosphate) and the oxidative pentose phosphate
pathway are present in Chlamydomonas chloroplasts
(Klein, 1986). However, it is still unclear whether
complete glycolytic breakdown of Glc to pyruvate
occurs in chloroplasts of anaerobic Chlamydomonas
cells. It is also unclear how much NADH recycling
occurs in chloroplasts of anaerobic cells. However,
based on our results, most regeneration of NAD+
from NADH in anaerobic wild-type Chlamydomonas
cells requires chloroplast ADH1 activity. Further-
more, essentially all of the work reported by us and
others suggests that there is regulated integration of
fermentation pathways in Chlamydomonas, which is
Magneschi et al.
1302Plant Physiol. Vol. 158, 2012

Page 11

probably also the case for other soil-dwelling algae.
The analysis of the adh1 mutant represents another
example of the extraordinary ability of Chlamydomo-
nas to modulate the activities of the various ferment-
ative pathways in order to maintain redox balance
and the production of ATP during anoxia, thus sur-
viving the anaerobic stress. The precise mechanisms
used to achieve physiological integration are still to
be elucidated.
MATERIALS AND METHODS
Strains and Growth Conditions
Chlamydomonas reinhardtii wild-type strains CC-124 (nit22, mt2) and D66
(CC-4425, nit22, cw15, mt+; Schnell and Lefebvre, 1993; Pollock et al., 2003) and
the adh1 mutant in both the CC-124 and D66 genetic backgrounds were
maintained on Tris-acetate-phosphate (TAP) medium, pH 7.2, solidified with
1.2% (w/v) agar at 25?C and 80 mmol photon m22s21photosynthetically
active, constant irradiance. A PCR-amplified DNA cassette (1.7 kb) encoding a
protein that confers paromomycin resistance (AphVIII) was used to transform
Chlamydomonas cells, generating a library of Chlamydomonas insertional mu-
tants (Pootakham et al., 2010; Gonzalez-Ballester et al., 2011). The adh1 mutant
allele was isolated from this library (D66 genetic background) using a PCR-
based screen (Pootakham et al., 2010; Gonzalez-Ballester et al., 2011) with the
ADH1-specific primers listed in Supplemental Table S1. The mutant was
backcrossed four times with CC-124. The CC-124 and adh1 strains (CC-124
genetic background) were grown on solid TAP medium, and once the cells
were near confluence on the plate, they were transferred to liquid TAP
medium (approximately 50 mL) and grown for 1 d at 25?C, 80 mmol photon
m22s21photosynthetically active, constant irradiance. These liquid cultures
wereused to inoculate 900 mL of TAP medium, pH 7.2, to a final concentration
of 5 3 104cells mL21(cells were counted three times for accuracy using a
hemocytometer). TAP cultures were grown in Roux bottles, stirred using
a magnetic stir bar, and vigorously bubbled with air enriched with 3% CO2.
Typically, experiments were performed with cultures that were grown sub-
sequently for 2 d to a final density of approximately 2 3 106cells mL21.
Anaerobic Induction, Sampling, and Survival
Chlamydomonas cultures grown on liquid TAP medium (approximately 2 3
106cells mL21) were concentrated 10 times by centrifugation (2,500g for 1 min)
and subsequent resuspension in 0.1 volume of anaerobic induction buffer
(AIB) containing 50 mM potassium phosphate (pH 7.0) and 3 mM MgCl2
(Ghirardi et al., 1997). The AIB suspended cultures were at 2 3 107cells
mL21(1.5 3 107cells mL21for tolerance assays). These cultures were flushed
with argon for 30 min and then incubated under anaerobic conditions (inside
an anaerobic chamber; Coy Laboratory Products) at room temperature in the
dark. Cells were collected at 4?C by centrifugation (10,000g for 1 min) at
specific times after the imposition of anoxia, and the supernatants and pellets
were separated, frozen in liquid nitrogen, and stored at 280?C for later
analyses (e.g. protein and nucleic acid). As an estimation of cell viability
following the anoxic treatment, aliquots of the cultures were sampled every 3
h over an 18-h period and spotted as serial 1:10 dilutions onto the surface of
solid TAP medium to allow for the recovery of viable cells. Each spot
represented 5 mL of a culture at an initial density of 1.5 3 107cells mL21(or
1:10 and 1:100 dilutions; see the columns on each plate in Fig. 3). A control
sample maintained under aerobic conditions in the light was included in the
survival tests.
Southern-Blot Analyses
Genomic DNA was isolated from 50-mL liquid cultures of Chlamydomonas
D66 and the original adh1 mutant (D66 background), each at 2 3 106cells
mL21, using a standard phenol-chloroform extraction protocol (Sambrook
et al., 1989). Ten micrograms of genomic DNA was digested for 2 h with 10
units of PstI or PvuII restriction endonucleases (New England Biolabs), the
DNA fragments were separated by agarose (0.8%) gel electrophoresis, the gels
were blotted overnight in 203 SSC onto nylon membranes (Bio-Rad), and the
transferred DNA was cross-linked to the membrane by UV illumination. An
alkaline phosphatase-labeled DNA probe was synthesized by chemically
cross-linking a thermostable alkaline phosphatase to the 1.7-kb AphVIII PCR
fragment (Sizova et al., 2001), which also contains the PSAD promoter and the
3# sequence from the CYTc6 gene (Fischer and Rochaix, 2001). Probe synthesis
and hybridizations were performed using the Amersham AlkPhos Direct
Labeling and Detection System according to the manufacturer’s protocol
(Amersham Biosciences). Cross-linked, membrane-bound, genomic DNAwas
hybridized overnight with the alkaline phosphatase-linked, 1.7-kb AphVIII
PCR product.
Extraction of RNA
Total RNA was isolated from frozen cell pellets using a standard phenol-
chloroform extraction protocol (Sambrook et al., 1989). The RNA was precip-
itated overnight in 4 M LiCl (final concentration) at 4?C to eliminate most of the
DNA in the preparations. To completely free the sample of genomic DNA,
approximately 40 mg of the RNAwas treatedwith 5units of RNase-freeDNase
I (Qiagen) for 1 h at room temperature (and repeated if necessary). A Qiagen
RNeasy MinElute kit was used to purify the DNase-treated total RNA and
remove degraded DNA, tRNA, 5.5S rRNA, DNase, contaminating proteins,
and potential inhibitors of the RT reaction. The A260of the eluted RNA was
determined, and its integrity was evaluated by electrophoresis on formalde-
hyde agarose gels.
RT-qPCR
The abundance of specific transcripts in the total mRNA of each sample
was quantified by RT-qPCR using the Engine Opticon system (Bio-Rad). First-
strand cDNA synthesis was primed from purified, total RNA template using
specific primers for each Chlamydomonas transcript that was examined. The RT
reaction conditions were reported previously (Mus et al., 2007; Dubini et al.,
2009); forwardand reverse primer sequences used in the RT-qPCR arelisted in
Supplemental Table S1. Amplifications were performed using the following
specific cycling parameters: an initial, single step at 95?C for 10 min (dena-
turation), then 40 cycles of 95?C for 30 s (denaturation), 60?C for 45 s
(annealing), and 72?C for 30 s (elongation), and then fluorescence measure-
ment after holding the reaction at 80?C for 10 s. This last step was incorporated
into the protocol to avoid background signals resulting from primer dimer
formation. After completing the 40 cycles, a final elongation step was
performed at 72?C for 10 min. Melt curves (65?C–100?C, heating rate of
0.2?C s21with continuous fluorescence measurements) were evaluated for all
PCRs to ensure that single DNA species were amplified. We determined both
the absolute (Steunou et al., 2006) and relative levels of each specific RNA
(normalized to the T0 sample corresponding to oxic conditions). All reactions
were performed in triplicate with at least two biological replicates.
Protein Isolation, SDS-PAGE, and Immunoblot Analysis
Frozen cell pellets were thawed and resuspended in 50 mM Tris buffer (pH
8.0) containing 10 mM EDTA, 2% SDS, 1 mM phenylmethylsulfonyl fluoride,
and 1 mM benzamidine-HCl. Protein concentrations were determined by the
bicinchoninic acid assay (Thermo Fisher Scientific) as described by the
manufacturer. For PAGE of proteins, pelleted cells were resuspended in 2%
SDS and 1 mM b-mercaptoethanol and then boiled for 5 min. Solubilized
proteins were separated by SDS-PAGE on a 10% polyacrylamide gel and then
transferred to polyvinylidene difluoride membranes by a Bio-Rad Trans-Blot
SD Semi-Dry Electrophoretic Transfer Cell following the manufacturer’s
instructions. The membranes were blocked in a 5% suspension of powdered
milk in Tris-buffered saline (pH 8.0) with 0.1% Tween 20 for 1 h prior to an
overnight incubation in the presence of primary antibodies. Dilutions of the
primary antibodies used were as follows: 1:20,000 for a-PFL1, 1:5,000 for
a-HYDA, 1:1,000 for a-PFR1, 1:2,000 for a-PAT2, 1:500 for ACK1, and 1:1,000
for ADH1. A 1:10,000 dilution of horseradish peroxidase-conjugated anti-rabbit
IgG (Promega) was used as a secondary antibody. The peroxidase activity was
detected by an enhanced chemiluminescence assay (GE Healthcare).
Antibodies
Antibodies were prepared by Agrisera (http://www.agrisera.com/en/
info/home.html) against the synthesized peptides EWLSHENRFQILERK,
Anaerobic Metabolism in Chlamydomonas adh1
Plant Physiol. Vol. 158, 20121303

Page 12

RSGRNYARDTIDRIF, and SGTGSEVTPFSVVTD of the Chlamydomonas PFR1,
PAT2, and ADH1 proteins, respectively. These peptides were conjugated to
keyhole limpet hemocyanin carrier protein via a Cys that was added to the N
terminus of each peptide. ACK1 antibodies were generated against the full-
length recombinant protein. PATand ACK antibodies might recognize the two
isoforms (chloroplast and likely mitochondrial) of these enzymes (Supple-
mental Fig. S2). Hydrogenase levels were evaluated by commercial antibodies
that recognize both HYDA1 and HYDA2 (Agrisera; no. AS09 514). PFL1
antibodies were kindly provided by Ariane Atteia at the Laboratoire de
Bioe ´nerge ´tique et Inge ´nierie des Prote ´ines in Marseille, France.
Extracellular Metabolite Analysis
Organic acids and alcohols were analyzed by liquid chromatography using
a Hewlett-Packard Series 1200 HPLC device.Dark-adapted cells were collected
atvarioustimesfollowingtheimpositionofanoxicconditionsand pelletedbya
1-min centrifugation (10,000g), and the supernatant was transferred to a new
vial and frozen in liquid N2for subsequent analysis. For organic acid analyses,
the samples were thawed, centrifuged, and filtered prior to the injection of
100 mL of the supernatant onto an Aminex HPX-87H (300 3 7.8 mm) ion-
exchange column. Metabolites in the supernatant were separated on the
column using filtered 8 mM sulfuric acid as the mobile phase; the flow rate
was 0.6 mL min21at 45?C. Organic acid retention peaks were recorded using
Agilent ChemStation software and quantified by comparisons with absorption
of known amounts of a standard for each of the organic acids. Ethanol was
detected using the refractive index detector attached to the HPLC apparatus.
CO2and H2Measurement
CO2levels were below detection limits in the serum vial head space of
anaerobically acclimated cells. Therefore, following anaerobic induction, 1 mL
of anoxic cells was transferred with a gas-tight syringe to a sealed vial into
which 1 mL of 1 M HCl was added. The acidified cell suspension was shaken
vigorously to liberate CO2, which was quantified by gas chromatography (GC;
Hewlett-Packard 5890 Series II) using a Supelco column (80/100 PORAPAK
N; 6 feet 3 1/8 inch 3 2.1 mm) coupled to a thermal conductivity detector.
Fermentative H2production was quantified from 400 mL of head space gas
withdrawn from sealed anaerobic vials and analyzed by GC (an Agilent Tech-
nologies 7890 GC system) using a Supelco column (60/80 mol sieve 5A; 6 feet 3
1/8 inch) coupled to a thermal conductivity detector.
ADH1 Complementation Construct
The 2,862-bp Chlamydomonas ADH1 coding sequence was amplified from
oligo(dT)-retrotranscribed cDNA using primers NdeIADH1Fw (5#-CATATG-
ATGTCCTCCAGCCTC-3#, introducing an NdeI site at the 5# end) and ADH1-
Rev (5#-GGAGTTCTTCTCCAAGATCAACTAA-3#), and the product was
cloned into pGEM T-Easy (Promega). A pGEM T-Easy plasmid with the
ADH1 coding sequence oriented SP6 to T7 (5# to 3#) was digested with NdeI
and EcoRI,which allowed directionalcloning intoplasmid pJM43Ble; thisvector
encodes the protein that confers resistance to the antibiotic Zeocin (Invitrogen)
and allowsconstitutive expressionofthe cloned cassette underthe control ofthe
PSAD promoter. Nuclear transformation of the Chlamydomonas adh1 mutant
(D66 background) was performed with 1.5 mg of pJM43Ble-ADH1 (linearized
with KpnI) DNA by the glass bead method (Kindle, 1990). Western-blot analysis
was used to screen for adh1 rescued strains (strains synthesizing the ADH1
protein). The rescued strain was backcrossed two times to CC-124 and D66, and
both the mutant and rescued strains were assayed for ethanol accumulation.
Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure S1. Multiple protein sequence alignment of C.
reinhardtii ADH1 with homologs from various sources.
Supplemental Figure S2. Protein sequence alignment of C. reinhardtii
PAT and ACK isoforms.
Supplemental Table S1. List of primers used in the PCR-based mutant
screening and in RT-qPCR.
Supplemental Table S2. Intracellular metabolite levels in wild-type and
adh1 cells.
Received November 29, 2011; accepted January 21,2012; published January 23,
2012.
LITERATURE CITED
Atteia A, van Lis R, Gelius-Dietrich G, Adrait A, Garin J, Joyard J,
Rolland N, Martin W (2006) Pyruvate formate-lyase and a novel route
of eukaryotic ATP synthesis in Chlamydomonas mitochondria. J Biol
Chem 281: 9909–9918
Atteia A, van Lis R, Mendoza-Herna ´ndez G, Henze K, Martin W, Riveros-
Rosas H, Gonza ´lez-Halphen D (2003) Bifunctional aldehyde/alcohol
dehydrogenase (ADHE) in chlorophyte algal mitochondria. Plant Mol
Biol 53: 175–188
Belmans D, van Laere A (1987) Glycerol cycle enzymes and intermediates
during adaptation of Dunaliella tertiolecta cells to hyperosmotic stress.
Plant Cell Environ 10: 185–190
Catalanotti C, Dubini A, Subramanian V, Yang W, Magneschi L, Mus F,
Seibert M, Posewitz MC, Grossman AR (2012) Altered metabolite flow
in Chlamydomonas reinhardtii mutants lacking pyruvate formate lyase
and both pyruvate formate lyase and alcohol dehydrogenase. Plant Cell
(in press)
Chen C, Gibbs M (1991) Glucose respiration in the intact chloroplast of
Chlamydomonas reinhardtii. Plant Physiol 95: 82–87
Clark DP (1989) The fermentation pathways of Escherichia coli. FEMS
Microbiol Rev 5: 223–234
Clark DP, Cronan JE Jr (1980) Escherichia coli mutants with altered control
of alcohol dehydrogenase and nitrate reductase. J Bacteriol 141: 177–183
Cunningham PR, Clark DP (1986) The use of suicide substrates to select
mutants of Escherichia coli lacking enzymes of alcohol fermentation. Mol
Gen Genet 205: 487–493
Dolferus R, Marbaix G, Jacobs M (1985) Alcohol dehydrogenase in
Arabidopsis: analysis of the induction phenomenon in plants and tissue
cultures. Mol Gen Genet 199: 256–264
Dubini A, Mus F, Seibert M, Grossman AR, Posewitz MC (2009) Flexi-
bility in anaerobic metabolism as revealed in a mutant of Chlamydomonas
reinhardtii lacking hydrogenase activity. J Biol Chem 284: 7201–7213
Fischer N, Rochaix JD (2001) The flanking regions of PsaD drive efficient
gene expression in the nucleus of the green alga Chlamydomonas
reinhardtii. Mol Genet Genomics 265: 888–894
Gfeller RP, Gibbs M (1984) Fermentative metabolism of Chlamydomonas
reinhardtii. I. Analysis of fermentative products from starch in dark and
light. Plant Physiol 75: 212–218
Ghirardi ML, Posewitz MC, Maness PC, Dubini A, Yu J, Seibert M (2007)
Hydrogenases and hydrogen photoproduction in oxygenic photosyn-
thetic organisms. Annu Rev Plant Biol 58: 71–91
Ghirardi ML, Togasaki RK, Seibert M (1997) Oxygen sensitivity of algal
H2-production. Appl Biochem Biotechnol 63-65: 141–151
Ghoshroy S, Binder M, Tartar A, Robertson DL (2010) Molecular evolution of
glutamine synthetase II: phylogenetic evidence of a non-endosymbiotic
gene transfer event early in plant evolution. BMC Evol Biol 10: 198–209
Gonzalez-Ballester D, Pootakham W, Mus F, Yang W, Catalanotti C,
Magneschi L, Higuera JH, de Montaigu A, Prior M, Galva ´n A, et al
(2011) Reverse genetics in Chlamydomonas: a platform for isolating
insertional mutants. Plant Methods 7: 24
Grossman AR, Catalanotti C, Yang W, Dubini A, Magneschi L, Subramanian
V, Posewitz MC, Seibert M (2011) Multiple facets of anoxic metabolism and
hydrogenproductioninthe unicellulargreenalgaChlamydomonas reinhardtii.
New Phytol 190: 279–288
Gupta KJ, Zabalza A, van Dongen JT (2009) Regulation of respiration
when the oxygen availability changes. Physiol Plant 137: 383–391
Gupta S, Clark DP (1989) Escherichia coli derivatives lacking both alcohol
dehydrogenase and phosphotransacetylase grow anaerobically by lac-
tate fermentation. J Bacteriol 171: 3650–3655
Harden A (1901) The chemical action of Bacillus coli communis and similar
organisms on carbohydrate and allied compounds. J Chem Soc 79:
610–627
Hemschemeier A, Jacobs J, Happe T (2008) Biochemical and physiological
characterization of the pyruvate formate-lyase Pfl1 of Chlamydomonas
reinhardtii, a typically bacterial enzyme in a eukaryotic alga. Eukaryot
Cell 7: 518–526
Jacobs M, Dolferus R, Van den Bossche D (1988) Isolation and biochem-
ical analysis of ethyl methanesulfonate-induced alcohol dehydrogen-
Magneschi et al.
1304 Plant Physiol. Vol. 158, 2012

Page 13

ase null mutants of Arabidopsis thaliana (L.) Heynh. Biochem Genet 26:
105–122
Kindle KL (1990) High-frequency nuclear transformation of Chlamydomo-
nas reinhardtii. Proc Natl Acad Sci USA 87: 1228–1232
Klein U (1986) Compartmentation of glycolysis and of the oxidative
pentosephosphate pathway in Chlamydomonas reinhardtii. Planta 167:
81–86
Klo ¨ck G, Kreuzberg K (1989) Kinetic properties of a sn-glycerol-3-phosphate
dehydrogenase purified from the unicellular alga Chlamydomonas reinhardtii.
Biochim Biophys Acta 991: 347–352
Knappe J, Neugebauer FA, Blaschkowski HP, Ga ¨nzler M (1984) Post-
translational activation introduces a free radical into pyruvate formate-
lyase. Proc Natl Acad Sci USA 81: 1332–1335
Kosourov S, Seibert M, Ghirardi ML (2003) Effects of extracellular pH on
the metabolic pathways in sulfur-deprived, H2-producing Chlamydomo-
nas reinhardtii cultures. Plant Cell Physiol 44: 146–155
Kreuzberg K (1984) Starch fermentation via formate producing pathway in
Chlamydomonas reinhardtii, Chlorogonium elongatum and Chlorella fusca.
Physiol Plant 61: 87–94
Ku ¨rsteiner O, Dupuis I, Kuhlemeier C (2003) The pyruvate decarboxyl-
ase1 gene of Arabidopsis is required during anoxia but not other
environmental stresses. Plant Physiol 132: 968–978
Leonardo MR, Dailly Y, Clark DP (1996) Role of NAD in regulating the
adhE gene of Escherichia coli. J Bacteriol 178: 6013–6018
Magneschi L, Perata P (2009) Rice germination and seedling growth in the
absence of oxygen. Ann Bot (Lond) 103: 181–196
Mu ¨cke U, Ko ¨nig S, Hu ¨bner G (1995) Purification and characterization of
pyruvate decarboxylase from pea seeds (Pisum sativum cv. Miko). Biol
Chem 376: 111–117
Mus F, Dubini A, Seibert M, Posewitz MC, Grossman AR (2007) Anaer-
obic acclimation in Chlamydomonas reinhardtii: anoxic gene expression,
hydrogenase induction, and metabolic pathways. J Biol Chem 282:
25475–25486
Ohta S, Miyamoto K, Miura Y (1987) Hydrogen evolution as a consump-
tion mode of reducing equivalents in green algal fermentation. Plant
Physiol 83: 1022–1026
Perata P, Alpi A (1991) Ethanol-induced injuries to carrot cells : the role of
acetaldehyde. Plant Physiol 95: 748–752
Perata P, Vernieri P, Armellini D, Bugnoli M, Tognoni F, Alpi A (1992)
Immunological detection of acetaldehyde-protein adducts in ethanol-
treated carrot cells. Plant Physiol 98: 913–918
Philipps G, Krawietz D, Hemschemeier A, Happe T (2011) A pyruvate
formate lyase-deficient Chlamydomonas reinhardtii strain provides evi-
dence for a link between fermentation and hydrogen production in
green algae. Plant J 66: 330–340
Plaga W, Frank R, Knappe J (1988) Catalytic-site mapping of pyruvate
formate lyase: hypophosphite reaction on the acetyl-enzyme interme-
diate affords carbon-phosphorus bond synthesis (1-hydroxyethyl-
phosphonate). Eur J Biochem 178: 445–450
Pollock SV, Colombo SL, Prout DL Jr, Godfrey AC, Moroney JV (2003)
Rubisco activase is required for optimal photosynthesis in the green alga
Chlamydomonas reinhardtii in a low-CO2atmosphere. Plant Physiol 133:
1854–1861
Pootakham W, Gonzalez-Ballester D, Grossman AR (2010) Identification
and regulation of plasma membrane sulfate transporters in Chlamydo-
monas. Plant Physiol 153: 1653–1668
Posewitz MC, King PW, Smolinski SL, Zhang LP, Seibert M, Ghirardi ML
(2004) Discovery of two novel radical S-adenosylmethionine proteins
required for the assembly of an active [Fe] hydrogenase. J Biol Chem
279: 25711–25720
Preiss S, Schrader S, Johanningmeier U (2001) Rapid, ATP-dependent
degradation of a truncated D1 protein in the chloroplast. Eur J Biochem
268: 4562–4569
Sadka A, Lers A, Zamir A, Avron M (1989) A critical examination of the
role of de novo protein synthesis in the osmotic adaptation of the
halotolerant alga Dunaliella. FEBS Lett 144: 93–98
Saika H, Matsumura H, Takano T, Tsutsumi N, Nakazono M (2006) A
point mutation of Adh1 gene is involved in the repression of coleoptile
elongation under submergence in rice. Breed Sci 56: 69–74
Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: A Laboratory
Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY
Schnell RA, Lefebvre PA (1993) Isolation of the Chlamydomonas regula-
tory gene NIT2 by transposon tagging. Genetics 134: 737–747
Sizova I, Fuhrmann M, Hegemann P (2001) A Streptomyces rimosus aphVIII
gene coding for a new type phosphotransferase provides stable antibi-
otic resistance to Chlamydomonas reinhardtii. Gene 277: 221–229
Steunou AS, Bhaya D, Bateson MM, Melendrez MC, Ward DM, Brecht E,
PetersJW,Ku ¨hlM,GrossmanAR(2006)Insituanalysisofnitrogenfixation
and metabolic switching in unicellular thermophilic cyanobacteria inhabit-
ing hot spring microbial mats. Proc Natl Acad Sci USA 103: 2398–2403
Tarmy EM, Kaplan NO (1968) Kinetics of Escherichia coli B D-lactate
dehydrogenase and evidence for pyruvate-controlled change in confor-
mation. J Biol Chem 243: 2587–2596
Terashima M, Specht M, Naumann B, Hippler M (2010) Characterizing the
anaerobic response of Chlamydomonas reinhardtii by quantitative pro-
teomics. Mol Cell Proteomics 9: 1514–1532
Tsygankov AA, Kosourov S, Seibert M, Ghirardi ML (2002) Hydrogen
photoproduction under continuous illumination by sulfur-deprived,
synchronous Chlamydomonas reinhardtii cultures. Int J Hydrogen Energy
27: 1239–1244
Whitney LA, Loreti E, Alpi A, Perata P (2011) Alcohol dehydrogenase and
hydrogenase transcript fluctuations during a day-night cycle in Chla-
mydomonas reinhardtii: the role of anoxia. New Phytol 190: 488–498
Zabalza A, van Dongen JT, Froehlich A, Oliver SN, Faix B, Gupta KJ,
Schma ¨lzlin E, Igal M, Orcaray L, Royuela M, et al (2009) Regulation of
respiration and fermentation to control the plant internal oxygen con-
centration. Plant Physiol 149: 1087–1098
Zhu J, Shimizu K (2005) Effect of a single-gene knockout on the metabolic
regulation in Escherichia coli for D-lactate production under micro-
aerobic condition. Metab Eng 7: 104–115
Anaerobic Metabolism in Chlamydomonas adh1
Plant Physiol. Vol. 158, 20121305