For bacterial growth on substrates with only one or two carbon
atoms, special assimilation pathways are required. In 1957, the
glyoxylate cycle of Kornberg and Krebs was described for bacterial
growth on [C.sub.2] compounds such as ethanol and acetate. However this
pathway did not operate in some photosynthetic bacteria and in some
methylotrophs when they were growing on Ce compounds, so an alternative
pathway must exist. By 1973 Quayle's serine cycle had been
described for methylotrophs growing on [C.sub.1] compounds such as
methanol, but the pathway was incomplete, the unknown part also
functioning during growth on [C.sub.2] compounds. After more than 35
further years of research, the ethylmalonyl-CoA (EMC) pathway for growth
on [C.sub.2] compounds, of photosynthetic bacteria has recently been
elucidated. This pathway also operates in methylotrophs during growth on
[C.sub.2] compounds, and on [C.sub.1] compounds by way of the serine
cycle. This review is a celebration of half a century of research and of
the fascinating result of that research.
Keywords: bacterial metabolism, serine cycle, serine/EMC cycle,
ethylmalonyl-CoA (EMC) pathway, EMC pathway, methylotrophs, Rhodobacter,
Methylobacterium, methylaspartate cycle, glyoxylate cycle
This review was originally planned to provide a brief summary of
the final elucidation of the serine cycle in methylotrophic bacteria
which grow on C x compounds such as methane, methanol or methylamine.
Our understanding of this cycle was thought to be almost complete when I
wrote a review for Science Progress in 1975 (1) but it has taken a
further 35 years for its completion. It has depended upon the
elucidation of the ethylmalonyl-CoA (EMC) pathway for growth on
[C.sub.2] compounds in some purple photosynthetic bacteria, almost half
a century after their 'problem' was first appreciated. This
review has thus expanded to include historical accounts of both of these
It is usual to give a list of acknowledgements at the end of a
publication but I want to emphasise the contributions of many people to
the work described here and also to acknowledge their generous
encouragement and help in the preparation of this manuscript: these
friends include Birgit Alber, Ivan Berg, John Bolbot, Mila
Chistoserdova, Tobi Erb, Georg Fuchs, Pat Goodwin, Peter Large, Mary
Lidstrom, Hans Kornberg, David Peel and Julia Vorholt.
Bacterial growth on [C.sub.1] and [C.sub.2] compounds
This review concerns the carbon assimilation pathways of aerobic
methylotrophs growing on [C.sub.1] compounds and of bacteria able to
grow on [C.sub.2] compounds such as ethanol and acetate as their sole
source of carbon. Energy for growth of methylotrophs is obtained by
oxidation of their [C.sub.1] substrates by specific dehydrogenases, and
a complete Krebs TCA cycle is not needed (2,3). By contrast, during
growth on [C.sub.2] compounds, energy is obtained by oxidation of
acetyl-CoA by the TCA cycle or, in photoheterotrophs, from light.
Assimilation of all carbon substrates when provided as the sole
source of carbon requires their conversion to the intermediates of
central metabolism containing 3 or 4 carbon atoms, which then provide
precursors for biosynthesis. During growth of methylotrophs on [C.sub.1]
compounds every carbon carbon bond must be created. In this they
resemble autotrophic bacteria and photosynthetic organisms which achieve
this by the Calvin-Benson-Bassham ribulose bisphosphate pathway for
fixation of carbon dioxide. Some methylotrophs do use this pathway but
the majority use specific methylotroph carbon assimilation pathways (3).
Similarly, a special pathway for growth on [C.sub.2] compounds is
needed because there is no route for direct formation of [C.sub.3] or
[C.sub.4] compounds from [C.sub.2] compounds such as ethanol or acetate,
or during growth on other substrates metabolised exclusively by way of
acetyl-CoA such as long chain fatty acids and 3-hydroxybutyrate. Of
course the intermediates of the TCA cycle can provide many of the
central metabolites required for biosynthesis but they cannot be
withdrawn without 'stopping' this oxidative cycle. In the
absence of other enzymes the only function of the TCA cycle is to
oxidise acetyl-CoA to two molecules of carbon dioxide. During metabolism
of substrates with three or more carbon atoms supplementary enzymes are
required that replenish the cycle by carboxylation of pyruvate or
phophoenolpyruvate [PEP] to oxaloacetate.
During growth of bacteria on [C.sub.2] compounds the pathway that
replenishes the intermediates of the TCA cycle was described in 1957 by
Kornberg and Krebs; this pathway is the glyoxylate cycle which results
in production of one molecule of malate from two molecules of
acetyl-CoA; oxaloacetate, produced from the malate can then be withdrawn
for biosynthesis (Figure 1) (4). Because, in effect, it bypasses the two
decarboxylating enzymes of the TCA cycle, Kornberg and Madsen suggested
that the cycle can also appropriately be called the glyoxylate bypass
(5). It requires two key enzymes, isocitrate lyase and malate synthase,
in addition to some of the TCA cycle enzymes. Kornberg later suggested
that these enzymes, and others involved in 'replenishing' the
TCA cycle during extraction of intermediates for biosynthesis, should be
named anaplerotic enzymes (6).
The operation of the proposed glyoxylate cycle during growth on
acetate was confirmed by Kornberg (7) using short-term incubation
experiments with radioactive [sup.14]C acetate, based on those used by
Calvin's group to elucidate the path of carbon from carbon dioxide
in photosynthesis. His conclusions were further confirmed in experiments
with Quayle in which they showed that the labelling pattern within cell
constituents was consistent with the pathway but not with alternatives
(8). Such experiments became the key type of experiment that led to the
elucidation in methylotrophs of the first part of the serine cycle by
Quayle and his colleagues.
[FIGURE 1 OMITTED]
The growth of methylotrophs using the serine cycle
There are at least four pathways for assimilation of reduced
[C.sub.1] compounds in methylotrophs, most of which were proposed, and
mainly elucidated, by J.R. (Rod) Quayle and his colleagues (3). The
serine cycle is different from the other pathways in having carboxylic
and amino acids as intermediates instead of the usual carbohydrates.
This was the first pathway described for growth on methanol but the last
to be completed.
The initial proposal by Quayle, with David Peel and Peter Large
(9,10), was based on short term labelling experiments using [sup.14]C
bicarbonate, [sup.14]C formate and [sup.14]C methanol in
Methylobacterium extorquens AM1 (then called Pseudomonas AM1). In these
experiments whole cells are incubated with the radioactive substrate and
samples taken every few seconds up to 2 minutes, inactivated in boiling
ethanol and non-volatile ethanol-soluble components separated by 2-way
paper chromatography. All radioactive components are eluted, counted,
purified and identified. In these experiments it is not the total label
in each compound at each time but the percentage in each compound of the
total label incorporated that is important (for example see Figure 2).
When the percentage labels are plotted against time, those showing a
negative slope are likely to be very early intermediates in metabolism
of the added substrate. For example, during photosynthesis [sup.14]C
bicarbonate is first incorporated into phosphoglycerate and the label
appears later in phosphorylated sugars. In contrast, the label from
methanol in M. extorquens was first seen in serine whereas [sup.14]C
bicarbonate first appeared in malate and aspartate which would be in
equilibrium with oxaloacetate which cannot be seen in this sort of
experiment; the distribution of [sup.14]C label within each compound was
subsequently determined. Two possible pathways (one linear, the other
cyclic) were proposed; in both the key fixation steps were addition of
formaldehyde to glycine to give serine, and carboxylation of a [C.sub.3]
compound derived from the serine to give oxaloacetate. The cyclic
pathway was confirmed by Alan Salem's discovery, of a novel
malyl-CoA lyase which cleaves malyl-CoA to acetyl-CoA plus glyoxylate
(11). Figure 3 shows the serine cycle as it stood in 1973. In summary,
the two carbon atoms of acetyl-CoA are produced from one molecule of
formaldehyde plus one carbon dioxide, both derived from methanol
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
The pathway was strongly supported by characterisation of the
proposed enzymes, many of them novel, and most of which are induced
during methylotrophic growth. Further confirmation of their importance
in the pathway was the isolation of mutants lacking them which had lost
the ability to grow on [C.sub.1] compounds but retained the ability to
grow on multicarbon compounds such as succinate. Pat Goodwin showed that
some of the serine cycle enzymes are co-ordinately regulated (12),
suggesting that, as expected for genes of a well-defined metabolic
pathway, the serine cycle genes are encoded on an operon, which has
subsequently been amply confirmed by her group (13) and by Mary
Lidstrom, Mila Chistoserdova and their colleagues (14,15).
[FIGURE 4 OMITTED]
The fate of the acetyl-CoA produced in the serine cycle
The serine cycle as described in Figure 3 produces acetyl-CoA from
formaldehyde and carbon dioxide, raising the obvious question of the
subsequent route for assimilation of the acetyl-CoA into [C.sub.3] and
[C.sub.4] compounds, and thence into cell material. The first obvious
answer is that the glyoxylate cycle must be operating here as it does
during growth of other bacteria on acetate (Figure 1). Indeed, some
methylotrophs do have the key enzyme (isocitrate lyase) of that pathway
and they assimilate [C.sub.1] compounds by what is known as the
[icl.sup.+] serine cycle (3) (Figure 4). The glyoxylate produced by the
malyl-CoA lyase reaction is the precursor of one molecule of glycine.
The acetyl-CoA is oxidised to a second glyoxylate by a route involving
isocitrate lyase (ICL) plus some (non-decarboxylating) TCA cycle
enzymes. However, it had been shown previously that there is no
isocitrate lyase present during methylotrophic growth of M. extorquens,
as was later confirmed for the majority of methylotrophs (3).
So, the question remaining is this: how is acetyl-CoA metabolised
to [C.sub.3] and [C.sub.4] compounds in M. extorquens during
methylotrophic growth? This question has taken more than 30 years to
answer and the search for this answer is the subject of the rest of this
review. It has depended on studies of [C.sub.2] metabolism by the
methylotrophs and on the elucidation of a novel pathway, the
ethylmalonyl-CoA (EMC) pathway for photoheterotrophic growth of bacteria
on [C.sub.2] compounds.
The pathway for acetyl-CoA assimilation during growth of
Methylobacterium extorquens on ethanol, 3-hydroxybutyrate, malonate,
pyruvate, lactate and propane 1,2-diol (1972-1980)
Most of the work I will describe in this slightly self-indulgent
section was done by Pat Dunstan (now Goodwin), Ian Taylor and John
Bolbot in my lab. Our interest in this topic started, when Pat set out
to isolate, using the penicillin enrichment technique, mutants of M.
extorquens lacking its quinoprotein methanol dehydrogenase [MDH] which
is also responsible for oxidation of ethanol during growth on that
substrate. We assumed that the pathway for assimilation of ethanol
[[C.sub.2]] must be different from that for methanol [[C.sub.1]], and
that the only common feature of their metabolism would be the initial
oxidation step. Therefore, Pat set out to characterise mutants that were
able to grow on succinate but not on methanol or ethanol. Remarkably,
three different types of mutant were isolated, two of which were unable
to oxidise the alcohols. These included MDH mutants as expected, and
also mutants lacking cytochrome c which had important implications in
our study of electron transport during methanol oxidation; we now know
that methanol is oxidised by a novel periplasmic electron transport
chain in which MDH is coupled directly to a specific c-type cytochrome
(16). However, a third type of mutant (e.g. mutant PCT48) was able to
oxidise both alcohols but was unable to grow on them, suggesting that
there might be some step that is common to assimilation of [C.sub.1] and
[C.sub.2] substrates; growth of mutant PCT48 was restored by inclusion
of glyoxylate together with methanol or ethanol in the growth medium. We
showed that the glyoxylate cycle is unlikely to be involved because M.
extorquens lacks the key enzyme isocitrate lyase during growth on
methanol or ethanol. We therefore set out to elucidate the common part
of the pathway for assimilation of [C.sub.1] and [C.sub.2] compounds
that appeared to involve oxidation of acetyl-CoA to glyoxylate
[FIGURE 5 OMITTED]
In short-term labelling experiments, we studied incorporation of
[sup.14]C acetate in cells grown on either methanol or ethanol using
wild type bacteria or mutant PCT48. If the glyoxylate cycle and TCA
cycle were operating then early label should be seen in citrate, but it
was not. Early label was found in malate and aspartate, and in
glycollate (with ethanol-grown cells) or glycine (in methanol grown
cells). No glycollate was detected when [sup.14]C acetate was used with
mutant PCT48 grown on succinate and induced with ethanol plus acetate
[note that glycollate reflects the presence of glyoxylate which is not
visible in these experiments]. We therefore concluded that there is a
novel pathway involving the oxidation of acetyl-CoA to glyoxylate in M.
extorquens grown on methanol or ethanol; it would appear to be like the
serine cycle as shown in Figure 4 but with an alternative route for
oxidation of acetyl-CoA to glyoxylate (12,17,18) (Figure 5).
The same unknown pathway was later shown to be involved, as
expected, during growth on 3-hydroxybutyrate (18,19) and malonate (19);
and unexpectedly, on pyruvate (20,21), lactate (20,21) and
1,2-propanediol (22). These [C.sub.3] compounds are in effect
metabolised as [C.sub.2] compounds; they cannot be converted to
[C.sub.4] compounds by carboxylation because M. extorquens lacks
pyruvate carboxylase or PEP synthase. Figure 5 summarises the pathways
as understood in 1980.
The glyoxylate regeneration cycle (GRC) for oxidation of acetyl-CoA
to glyoxylate in the serine cycle
In 1983 at one of the regular International Symposia on Microbial
Growth on [C.sub.1] Compounds, a paper was presented by Shimizu, Ueda
and Sato on the physiological role of vitamin [B.sub.12] in
Protaminobacter ruber (very similar to M. extorquens), which produced
relatively large amounts of vitamin [B.sub.12]. Their demonstration that
crude extracts could catalyse production of glyoxylate and propionyl-CoA
from mesaconyl-CoA (with methylmalyl-CoA as intermediate) led them to
propose a pathway for oxidation of acetyl-CoA to glyoxylate that
involves mesaconyl-CoA, methylmalyl-CoA, methylmalonyl-CoA,
propionyl-CoA and succinyl-CoA (Figure 6) (23). Part of this sequence of
intermediates was subsequently confirmed to be involved in oxidation of
acetyl-CoA to glyoxylate in the serine cycle (from mesaconyl-CoA to
glyoxylate plus succinate). Their experiments provided the first direct
evidence that the key unknown intermediates would be derivatives of
Coenzyme A; it should be noted that although such intermediates were
often the subject of speculation, they would never have been seen in
earlier studies because they are destroyed during preparation and
separation of metabolic intermediates.
Remarkably, although the cycle as shown in Figure 6 is not the
route for assimilation of acetyl-CoA in M. extorquens or P. ruber, this
cycle has now been shown by Ivan Berg and colleagues to be the route for
assimilation of [C.sub.2] compounds in haloarchea; this third strategy
for assimilation of [C.sub.2] compounds is now named the methylaspartate
The involvement of derivatives of Coenzyme A in oxidation of
acetyl-CoA to glyoxylate was confirmed 12 years later by Pat Goodwin who
described the isolation and sequencing of a gene that complemented the
PCT48 mutation which had led to inability to oxidise acetyl-CoA to
glyoxylate; this gene (meaA) encoded an unknown type of
[B.sub.12]-dependent mutas (25), now known to be ethylmalonyl-CoA mutase
(see below). At the same time, and using a similar approach,
Chistoserdova and Lidstrom described this gene and also concluded that
propionyl-CoA carboxylase (encoded by pccA) and an unknown dehydrogenase
(encoded by adhA) are also involved (26); the adhA designation was later
changed to ccr, encoding crotonyl-CoA reductase (27).
[FIGURE 6 OMITTED]
With the exception of the contribution of Shimizu and colleagues,
advances in our understanding of the serine cycle in the two decades
after 1980 depended on the development of methylotroph genetics and
mapping of M. extorquens genes (see above). In particular, from about
1991 Mary Lidstrom and Mila Chistoserdova started to publish a lengthy
series of papers on the enzymes and genes of the serine pathway, and of
methanol oxidation, in M. extorquens, establishing an important centre
for future genetic studies of methylotrophs (15,28-29), leading to
publication in 2003 of a preliminary version (29) and, in 2009, the
complete genome of M. extorquens (30). Their experimental approach to
solving the 'missing' part of the serine cycle was to use
genome analysis to select candidate genes for enzymes that might be
involved. Candidate genes were tested by mutation, and potential
intermediates identified by monitoring the fates of labelled acetate,
butyrate and bicarbonate in these mutants and wild type bacteria
(27,31-33). They showed that the most likely start of the route for
oxidation of acetyl-coA to glyoxylate (the glyoxylate regeneration
cycle; GRC; Figure 7), would be conversion of acetyl-CoA to crotonyl-CoA
by enzymes similar (or identical) to those involved in biosynthesis of
poly-3-hydroxybutyrate or oxidation of fatty acids, and that a later
part of the GRC would involve carboxylation of propionyl-coA to
methylmalonyl-CoA followed by conversion to succinyl-CoA. Their
(unavoidably) speculative GRC had 18 intermediates between acetyl-CoA
and glyoxylate (27). Although many of these were later shown not to be
involved, one key intermediate, ethylmalonyl-CoA, was later shown to be
essential in what has become known as the ethylmalonyl-CoA (EMC)
[FIGURE 7 OMITTED]
The eventual elucidation by Fuchs' group in Freiburg of this
pathway for glyoxylate formation from acetyl-CoA in the serine cycle
came from definitive enzymological work on the pathway for growth on
[C.sub.2] compounds of photosynthetic bacteria lacking the glyoxylate
[FIGURE 8 OMITTED]
The pathway for assimilation of [C.sub.2] compounds in some purple
photosynthetic bacteria lacking the glyoxylate cycle
After the glyoxylate cycle was established for bacterial growth on
[C.sub.2] compounds (Figure 1) it was very soon demonstrated that this
pathway is not universal; it does not operate in many methylotrophs (3),
or in Streptomyces (34), or in some photosynthetic bacteria such as
Rhodospirillum rubrum and Rhodobacter sphaeroides which are unable to
synthesise the key enzyme of the glyoxylate cycle, isocitrate lyase
(35). These photoheterotrophic bacteria use acetate as a carbon and
energy source in the dark or as a carbon source when grown anaerobically
with light as their source of energy. As early as 1951 it was shown by
Cutinelli and colleagues (36), using [sup.13]C[H.sub.3]. [sup.14]COONa,
that acetate was incorporated into protein amino acids without its prior
oxidation to C[O.sub.2], indicating the presence of a metabolic pathway
for direct assimilation of acetate; there was also significant
incorporation into protein amino acids from labelled bicarbonate. The
existence of such a novel pathway in Rhodospirillum rubrum was confirmed
by Benedict (37) and Hoare (38) who showed that [sup.14]C acetate was
accumulated in citramalate ([alpha]-methylmalate) and this led Ivanovsky
and colleagues (39,40) to propose, in 1997, a citramalate cycle for
acetate assimilation (Figure 8), based in part on work by Osumi and
Katsuki who had demonstrated 20 years previously that these bacteria
contain an enzyme catalysing the reversible condensation of glyoxylate
and propionyl-CoA to [C.sub.5] compounds (41).
Key steps of the proposed citramalate cycle (Figure 8) are the
condensation of acetyl-CoA and pyruvate to citramalate
([alpha]-methylmalate); this is converted to [beta]-methylmalyl-CoA,
which is cleaved to glyoxylate plus propionyl-CoA. Pyruvate is then
regenerated from propionyl-CoA by way of methylmalonyl-CoA and
phosphoenolpyruvate. The glyoxylate condenses with a second molecule of
acetyl-CoA to form malate. Work on these photosynthetic bacteria
eventually led to the elucidation of the ethylmalonyl-CoA (EMC) pathway,
operating during growth in bacteria growing on [C.sub.2] compounds and
also during methylotrophic growth.
The elucidation of the ethylmalonyl-CoA (EMC) pathway for bacterial
growth on [C.sub.2] compounds
Recent work on this pathway was initiated by Georg Fuchs, with
Birgit Alber, Tobi Erb, Ivan Berg and their colleagues in Freiburg; the
success of this work depended on a multitude of experimental approaches
including genetic analysis, proteomics, analysis of metabolites (linked
to Coenzyme A) and excellent conventional enzymology. They showed in
2005 that malyl-CoA lyase was induced during growth on acetate of
Rhodobacter capsulatus; its gene (mcll) was also present in
Rhodospirillum rubrum, Rhodobacter sphaeroides and Streptomyces
coelicolor, and was 51% identical to the malyl-CoA lyase gene in the
methylotroph M. extorquens (42). A key observation was that the purified
enzyme was also able to catalyse cleavage of [beta]-methylmalyl-CoA to
propionyl-CoA plus glyoxylate. This led to their conclusion that this
enzyme functions twice in the pathway (Figure 9). It catalyses cleavage
of methylmalyl-CoA to propionyl-CoA plus glyoxylate; it subsequently
catalyses the condensation of glyoxylate with acetylCoA to give
malyl-CoA and thence malate. The propionyl-CoA from the cleavage
reaction is carboxylated to methylmalonyl-CoA, eventually to give
succinate. This left the question of the origin of the
[beta]-methylmalyl-CoA; this was investigated using mutagenesis, genome
analysis, proteomics, and HPLC and mass spectrometry of metabolites
produced by cell free extracts of wild type and mutant bacteria (43). In
their proposed pathway (Figure 9) two molecules of acetyl-CoA are
condensed to acetoacetyl-CoA which is then reduced to
3-hydroxybutyryl-CoA. In an unknown reaction a [C.sup.4] derivative of
this is carboxylated to an unknown [C.sub.5] compound. This gives rise
to mesaconyl-CoA, which is hydrated to give the [beta]-methylmalyl-CoA.
They showed that a gene cluster present in the genomes of all bacteria
proposed to use this pathway contained ccr-and meaA-like genes,
suggesting that a crotonyl-CoA reductase and a [B.sub.12]-dependent
mutase (unknown specificity) may be involved in the 'missing'
part of the pathway.
[FIGURE 9 OMITTED]
The discovery of three novel enzymes in Rhodobacter sphaeroides to
complete the EMC pathway
The solution of the 'missing' part of the pathway
depended on the discovery of three novel enzymes in R. sphaeroides.
These are crotonyl-CoA carboxylase/reductase, ethylmalonyl-CoA mutase
and methylsuccinyl-CoA dehydrogenase (Figure 10).
Crotonyl-CoA carboxylase / reductase (44)
This novel enzyme was described in Freiburg in the year marking the
50th anniversary of the publication in 1957 of the glyoxylate cycle by
Kornberg and Krebs (1957). Because a key part of the 'missing'
part of the pathway from acetyl-CoA involved carboxylation and reduction
(Figure 9), this work was initiated by investigating products of cell
free extracts incubated with NADPH, [sup.14]C bicarbonate and various
thioesters. The best substrate was crotonyl-CoA and the product was
shown by HPLS-MS to be ethylmalonylCoA; this was then confirmed by 2D
NMR spectroscopy. It had been shown previously that crotonyl-CoA
reductase (from the ccr
[FIGURE 10 OMITTED]
gene) is essential for growth on [C.sub.2] compounds by
Streptomyces and by M. extorquens, and it was thought that this
reductase might also catalyse carboxylation.
The ccr gene from R. sphaeroides was therefore cloned into E. coli,
and extracts shown to catalyse the carboxylation of crotonyl-CoA in the
presence of NADPH. The enzyme catalysing this reaction was purified,
thoroughly characterised and named crotonyl-CoA carboxylase/reductase:
crotonyl-CoA + NADPH + H+ + C[O.sub.2] [right arrow]
ethylmalonyl-CoA + [NADP.sup.+]
The NADPH donates a hydride to the [beta]-carbon of crotonyl-CoA
whose [alpha]-carbon is then carboxylated to yield the product
ethylmalonyl-CoA (44-45). This inducible enzyme, catalysing the
reductive carboxylation of an enoyl-CoA ester, is a reaction
unprecedented in biology and is one of the key reactions of the new
proposed ethylmalonyl-CoA pathway; its presence can therefore be used as
a good indicator of the operation of the pathway in other organisms. For
example it is also present in Streptomyces growing on butyrate and also
on M. extorquens during growth on methanol (44).
Ethylmalonyl-CoA mutase; a new category of coenzyme
[B.sub.12]-dependent acyl-CoA mutases (46)
In Fuchs' lab, using NMR spectroscopy it was shown that, in
the presence of coenzyme [B.sub.12],
[3-carboxy-[sup.14]C]ethylmalony-COA was rapidly converted by crude cell
extracts of R. sphaeroides to [4-carboxy-[sup.14]C]methylsuccinyl-CoA,
demonstrating that a rearrangement of the carbon skeleton had taken
place. The mutase is encoded by the gene ecm which is equivalent to meaA
in M. extorquens, encoding the unknown mutase involved in the proposed
glyoxylate regeneration cycle (GRC) (25,26). The ethylmalonyl-CoA mutase
was expressed from ecm in E. coli and the novel enzyme purified and
characterised. It is specific for ethylmalonyl-CoA and completely
distinct from the methylmalonyl-CoA mutase that catalyses the conversion
of methylmalonyl-CoA to succinyl-CoA later in the pathway.
The substrate for the mutase is (2R)-ethylmalonyl-CoA whereas the
product of the carboxylase/reductase reaction (above) is
(2S)ethylmalonyl-CoA; an epimerase to catalyse their interconversion is
therefore required. This 'promiscuous' enzyme, also
responsible for the epimerisation of methylmalonyl-CoA later in the
pathway, is called ethylmalonyl-CoA/methylmalonyl-CoA epimerase and is
encoded by the epi gene.
Methylsuccinyl-CoA dehydrogenase (47)
This third enzyme, recently described by Erb, Fuchs and Alber,
'closes' the ethylmalonyl-CoA pathway for acetate assimilation
(47). This FAD-linked dehydrogenase (encoded by mcd) is highly specific
for (2S)-methylsuccinyl-CoA, oxidising it to mesaconyl-CoA.
The discovery of these three novel enzymes enabled the description
of the ethylmalonyl-CoA (EMC) pathway for assimilation of acetate in
these photosynthetic bacteria.
Summary of the EMC pathway for growth on acetate in Rhodobacter
Figures 11 and 12 summarise this pathway for growth on [C.sub.2]
compounds as described by Fuchs' group as a result of their work,
culminating in the description of the three key enzymes that convert
crotonyl-CoA to methylsuccinyl-CoA (Figure 10). The genes encoding these
novel enzymes, ccr, ecm (meaA) and (sometimes) mcd, are clustered
together on the same genetic locus on the chromosome of those bacteria
that grow on acetate (or other [C.sub.2] compounds) but not having the
classic glyoxylate cycle, including Rhodobacter sphaeroides,
Rhodospirillum rubrum, Streptomyces coelicolor and M. extorquens (44).
It is remarkable that two of the enzymes function twice in the pathway.
A single epimerase is responsible for conversion of
(2S)-ethylmalonyl-CoA and (2S)-methylmalonyl-CoA. Similarly a single
lyase (encoded by mcl1) is responsible for cleavage of methylmalyl-CoA
to propionyl CoA plus glyoxylate, and also for condensation of
glyoxylate and acetyl-CoA to malyl-CoA. A homologue of this lyase
(encoded by mcl2) was shown to be upregulated during growth on acetate.
Gene activation and biochemical studies of the two enzymes confirmed the
dual role of the Mcl1 lyase and showed that mcl2 encodes a malyl-CoA
thioesterase (48). The two enzymes together thus catalyse the
condensation of glyoxylate and acetyl-CoA to malate, a reaction which is
usually catalysed by the single enzyme malate synthase in bacteria
having the glyoxylate pathway. The overall carbon balance (47) of the
EMC pathway is:
3acetyl-CoA + 2C[O.sub.2] [right] arrow] malate + succinate
[FIGURE 11 OMITTED]
It is worth noting that the elucidation of the EMC pathway for
oxidation of acetyl-CoA to glyoxylate has not only solved long standing
problems in the photosynthetic bacteria, and the methylotrophs [see
below] but, as pointed out in a valuable recent summary of the pathway
by Birgit Alber, it also has biotechnological implications, providing
the opportunity of tapping into the potential of using the new
intermediates and enzymes to produce value-added products (49).
[FIGURE 12 OMITTED]
The ethylmalonyl-CoA (EMC) pathway in methylotrophs during their
growth on [C.sub.2] compounds
All the previously accumulated evidence relating to growth on
[C.sub.2] compounds of methylotrophs like M. extorquens is consistent
with the operation of the EMC pathway, but one problem requires some
discussion. Malate synthase catalyses the condensation of glyoxylate and
acetyl-CoA to malate and is essential in all bacteria that use the
glyoxylate cycle; the proposed alternative EMC pathway in M. extorquens
also requires this enzyme, and we showed in 1976 that mutant ICT51
(Figure 5) which had lost malate synthase activity had also lost the
ability to grow on [C.sub.2] compounds, although able to grow on
[C.sub.1] compounds or succinate (19). It was then shown that this
'malate synthase' activity is due to a combination of the
reversible malyl-CoA lyase plus a malyl-CoA hydrolase (50). The
'malate synthase' activity was also lost from mutant PCT57
which lacked malyl-CoA lyase (Figure 5). About 30 years later similar
conclusions have been drawn for R. sphaeroides during growth on acetate
by the EMC pathway (42,48); the enzyme that catalyses the cleavage of
methylmalyl-CoA in the first part of the pathway also catalyses
condensation of glyoxylate and acetyl-CoA in the later part of the
pathway [hence its name malylCoA/methylmalyl-CoA lyase].
In M. extorquens there remain unanswered questions relating to the
properties of some mutants lacking the ability to grow on [C.sub.2]
compounds. The more recent experiments on this topic are relatively
complex; furthermore it is, of course, extraordinarily difficult to
attempt to reconcile some growth properties with levels of enzymes
required in wild type and mutant in experiments spread over almost 40
years. For further discussion of this the reader is referred to the
original literature (3,11,12,18-20,50-53).
The role of the ethylmalonyl-CoA (EMC) pathway in the serine cycle
for growth on [C.sub.1] compounds by methylotrophs
After elucidation of the EMC pathway for oxidation of acetyl-CoA to
glyoxylate in Rhodobacter sphaeroides, it was immediately apparent that
this pathway could also provide the answer to the related problem in
methylotrophs growing by the serine cycle, and this solution was
published in the same paper as the first proposal of the EMC pathway by
Fuchs and his colleagues (44). Figure 13 summarises their EMC pathway as
it links to the serine cycle during growth of M. extorquens on [C.sub.1]
compounds. It is the same as when it operates during growth on [C.sub.2]
compounds, the only difference being the fate of the glyoxylate produced
from acetyl-CoA. During growth on [C.sub.1] compounds the glyoxylate is
transaminated to glycine which condenses with formaldehyde to give
serine; whereas during growth on [C.sub.2] compounds the glyoxylate
condenses with acetyl-CoA, eventually producing malate. In the depiction
of the complete serine cycle as shown in Figure 13 the carbon balance
3[C.sub.1] + 4C[O.sub.2] [right arrow] 1[C.sub.3] + 1[C.sub.4]
It should be noted that this pathway is consistent with all major
previous studies (described above) of metabolism in wild type and
mutants. Such studies have been extended by Fuchs' group who have
demonstrated all the reactions of the ethylmalonyl-CoA pathway in
methanol-grown M. extorquens (52). They showed regulation of all the
enzymes involved in response to [C.sub.1], [C.sub.2] and [C.sub.4]
substrates, with (as would be expected) significantly higher activities
of EMC pathway enzymes in methanol- and acetate-grown cells compared
with succinate-grown cells. Similar conclusions were drawn in an
extensive study by Lidstrom and colleagues in which they added methanol
to a succinate-limited chemostat culture of M. extorquens and followed
the transition of the cells from succinate to methanol growth, measuring
transcription, enzyme activity, metabolites and fluxes (53). This
integrated analysis provided information about the transient imbalance
and the response to carbon starvation when they are shifted out of the
multi-carbon metabolic mode, the flux redistribution that occurs while
the cells are adapting to the new [C.sub.1] metabolic mode, and the
relationships between transcript, enzyme activity, metabolite and flux
during this transition.
[FIGURE 13 OMITTED]
Further confirmation of the operation of the EMC pathway during
methanol assimilation by M. extorquens has been provided by an
investigation by Julia Vorholt and colleagues in Zurich, using a
remarkably powerful set of experimental approaches (54). The development
of an original method of liquid chromatography high-resolution mass
spectrometry (LC-HRMS) allowed for the identification of almost all
intermediates of the ethylmalonyl-CoA pathway. Detailed and conclusive
information on its role as the predominant process for glyoxylate
formation was obtained from experiments in which kinetic isotopomer
profiles collected by LCHRMS during short-term experiments with
[1-[sup.13]C] acetate were combined with steady-state isotopomer
distributions measured by NMR ([sup.13]C metabolomics!). The
steady-state distribution of [[sup.13]C] isotopomers in glycine was
subsequently used to determine the metabolic origin of glyoxylate in M.
extorquens under purely methylotrophic conditions. Their results
suggested that the propionyl-CoA produced in the EMC pathway
'recycles' by way of methylmalonyl-CoA, and succinyl-CoA to
malyl-CoA which is then cleaved to give a second molecule of glyoxylate.
This is incorporated into their proposed depiction of the serine cycle
shown in Figure 14. The overall carbon balance of this cycle is then:
3MeOH + 3C[O.sub.2] [right arrow] 2[C.sub.3]
Exactly the same enzymes and intermediates are involved in the
cycle operating to produce two molecules of PEP as are involved in the
depiction in Figure 13 in which the balance is:
3HCHO + 4C[O.sub.2] [right arrow] 1[C.sub.3] + 1[C.sub.4]
The difference in the two representations is the fate of some of
the propionyl-CoA. It is obvious that oxaloacetate, PEP, succinyl-CoA
and acetyl-CoA must always be produced for biosynthesis of amino acids,
carbohydrates, haem, and fatty acids and poly-[beta]-hydroxybutyrate. It
is equally obvious therefore that no single representation can summarise
the operation of the serine/EMC cycle in all growth conditions. For
convenience in comparing the flow of carbon from substrate to cell
material in all the pathways discussed in this review, balances taken
from those pathways, but normalised to produce oxaloacetate, have been
put together in Table 1.
[FIGURE 14 OMITTED]
I should like to record that the ethylmalonyl-CoA (EMC) pathway
was, appropriately, first presented in 2006 by Tobi Erb, in Magdalen
College, Oxford during the same Gordon Conference on Microbial
Metabolism of [C.sub.1] Compounds at which I gave a lecture
commemorating the work of the late Rod Quayle who would have been so
excited to have seen the work he had started more than half a century
ago on the serine cycle in methylotrophs to be so magnificently
(1.) Anthony, C. (1975) The biochemistry of methylotrophic
micro-organisms. Sci. Prog., 62, 167-206.
(2.) Taylor, I.J. and Anthony, C. (1976) A biochemical basis for
obligate methylotrophy: properties of a mutant of Pseudomonas AM1
lacking 2-oxoglutarate dehydrogenase. J. Gen. Microbiol., 93, 259 265.
(3.) Anthony, C. (1982) The biochemistry of methylotrophs. Academic
(4.) Kornberg, H.L. and Krebs, H.A. (1957) Synthesis of cell
constituents from [C.sub.2]-units by a modified tricarboxylic acid
cycle. Nature, 179, 988-991.
(5.) Kornberg, H.L. and Madsen, N.B. (1957) Synthesis of
C4-dicarboxylic acids from acetate by a 'glyoxylate bypass' of
the tricarboxylic acid cycle. Biochim. Biophys. Acta, 24, 651.
(6.) Kornberg, H.L. (1965) Anaplerotic sequences in microbial
metabolism. Angew. Chem., 4, 558-565.
(7.) Kornberg, H.L. (1958) The metabolism of [C.sub.2] compounds in
micro-organisms. 1. The incorporation of [2-[sup.14]C] acetate by
Pseudomonas fluorescens, and by a Corynebacterium, grown on ammonium
acetate. Biochem. J., 68, 535-542.
(8.) Kornberg, H.L. and Quayle, J.R. (1958) The metabolism of
[C.sub.2] compounds in micro-organisms. 2. The effect of carbon dioxide
on the incorporation of [[sup.14]C] acetate by acetate-grown Pseudomonas
KBI. Biochem. J., 68, 542-549.
(9.) Large, P.J., Peel, D. and Quayle, J.R. (1961) Microbial growth
on [C.sub.1] compounds: Synthesis of cell constituents by methanol- and
formate-grown Pseudomonas AM1 and methanol-grown Hyphomierobium vulgare.
Biochem. J., 81, 470-480.
(10.) Large, P.J., Peel, D. and Quayle, J.R. (1962) Microbial
growth on [C.sub.1] compounds: Distribution of radioactivity in
metabolites of methanol-grown Pseudomonas AM1 after incubation with
[[sup.14]C] methanol and [[sup.14]C] bicarbonate. Biochem. J., 82,
(11.) Salem, A.R., Hacking, A.J. and Quayle, J.R. (1973a) Cleavage
of malyl-Coenzyme A into acetyl Coenzyme A and glyoxylate by Pseudomonas
AM1 and other [C.sub.1]-unit-utilizing bacteria. Biochem. J., 136,
(12.) Dunstan, P.M., Anthony, C. and Drabble, WT. (1972) The role
of glyoxylate, glycollate and acetate in the growth of Pseudomonas AM1
on ethanol and [C.sub.1] compounds. J. Gen. Microbiol., 128, 107-115.
(13.) Stone, S. and Goodwin, P.M. (1989) Characterization and
complementation of mutants of Methylobacterium extorquens AM1 which are
defective in [C.sub.1] assimilation. J. Gen. Microbiol., 135, 227-235.
(14.) O'Connor, M.E. (1981) Extension of the model concerning
linkage of genes coding for [C.sub.1] related functions in
Methylobacterium organophilum. Appl. Environ. Microbiol., 41, 437-441.
(15.) Chistoserdova, L.V., Kalyuzhnaya, M.G. and Lidstrom, M.E.
(2009) The expanding world of methylotrophic metabolism. Annu. Rev.
Microbiol., 63, 477-99.
(16.) Anthony, C. and Williams, P.W. (2003) The structure and
function of methanol dehydrogenase. Biochim. Biophys. Acta, 1467, 18-23.
(17.) Dunstan, EM., Anthony, C. and Drabble, W.T. (1972). The
involvement of glycollate in the metabolism of ethanol and of acetate by
Pseudomonas AM1. Biochem. J., 128, 99-106.
(18.) Dunstan, EM. and Anthony, C. (1973). The role of acetate
during growth of Pseudomonas AM1 [C.sub.1] compounds, ethanol and
3-hydroxybutyrate. Biochem. J., 132, 797-801.
(19.) Taylor, I.J. and Anthony, C. (1976) Acetyl-CoA production and
utilization during growth of the facultative methylotroph, Pseudomonas
AM1, on ethanol, malonate and 3-hydroxybutyrate. J. Gen. Microbiol., 95,
(20.) Salem, A.R., Wagner, C., Hacking, A.J. and Quayle, J.R.
(1973) The metabolism of lactate and pyruvate by Pseudomonas AM1. J.
Gen. Microbiol., 76, 375-388.
(21.) Bolbot, J.A. and Anthony, C. (1980) The metabolism of
pyruvate by the facultative methylotroph, Pseudomonas AM1. J. Gen.
Microbiol., 120, 233-244.
(22.) Bolbot, J.A. and Anthony, C. (1980) The metabolism of
1,2-propanediol by the facultative methylotroph, Pseudomonas AM1. J.
Gen. Microbiol., 120, 245-254.
(23.) Shimizu, S., Ueda, S. and Sato, K. (1984). Physiological role
of vitamin [B.sub.12] in a methanol utilising bacterium, Protaminobacter
ruber. In: Crawford, R.L. and Hanson, R.S. (eds.), Microbial growth on
[C.sub.1] compounds, pp. 113-117. American Society for Microbiology,
(24.) Maria Khomyakova, M., Bukmez, O., Thomas, L.K., Erb, T.J. and
Berg, I.A. (2011) A Methylaspartate Cycle in Haloarchaea. Science, 21,
(25.) Smith, L.M., Meijer, WG., Dijkhuizen, L. and Goodwin, P.M.
(1996) A protein having similarity with methymalonyl-CoA mutase is
required for the assimilation of methanol and ethanol by
Methylobacterium extorquens. Microbiology, 142, 675-684.
(26.) Chistoserdova, L.V. and Lidstrom, M.E. (1996) Molecular
characterization of a chromosomal region involved in the oxidation of
acetyl-CoA to glyoxylate in the isocitrate-lyase-negative methylotroph
Methylobacterium extorquens AM1. Microbiology, 142, 1459-1468.
(27.) Korotkova, N., Chistoserdova, L., Kuksa, V. and Lidstrom,
M.E. (2002) Glyoxylate regeneration pathway in the methylotroph
Methylobacterium extorquens AM1. J. Bacteriol., 184, 1750-1758.
(28.) Lidstrom, M.E., Anthony, C., Biville, E, Gasser, E, Goodwin,
E et al. (1994) New unified nomenclature for genes involved in the
oxidation of methanol in gram-negative bacteria. FEMS Microbiol. Lett.,
(29.) Chistoserdova, L., Chen, S.W., Lapidus, A. and Lidstrom, M.E.
(2003). Methylotrophy in Methylobacterium extorquens AM1 from a genomic
point of view. J. Bacteriol., 185, 2980-2987.
(30.) Vuilleumier, S., Chistoserdova, L., Lee, M.-C., Bringel, E,
Lajus, A. et al. (2009). Methylobacterium genome sequences: a reference
blueprint to investigate microbial metabolism of [C.sub.1] compounds
from natural and industrial sources. PLoS ONE, 4:e5584.
(31.) Korotkova, N. and Lidstrom, M.E. (2001). Connection between
poly-[beta]-hydroxybutyrate biosynthesis and growth on [C.sub.1] and
[C.sub.2] compounds in the methylotroph Methylobacterium extorquens AM1.
J Bacteriol., 183, 1038-1046.
(32.) Korotkova, N. and Lidstrom, M.E. (2004). MeaB is a component
of the methylmalonyl-CoA mutase complex required for protection of the
enzyme from inactivation. J Biol. Chem., 279, 13652-13658.
(33.) Korotkova, N., Lidstrom, M.E. and, Chistoserdova, L. (2005).
Identification of genes involved in the glyoxylate regeneration cycle in
Methylobacterium extorquens AM1, including two new genes, meaC and meaD.
J. Bacteriol., 187, 1523-1526.
(34.) Han, L. and Reynolds, K.A. (1997) A novel alternate
anaplerotic pathway to the glyoxylate cycle in streptomycetes. J.
Bacteriol., 179, 5157-5164.
(35.) Kornberg, H.L., and Lascelles, J. (1960) The formation of
isocitratase by the thiorhodaceae. J Gen. Microbiol., 23, 511-517.
(36.) Cutinelli, C., Ehrensvard, G., Reio, L., Saluste, E. and
Stjernholm, R. (1951) Acetic acid metabolism in Rhodospirillum rubrum
under anaerobic condition. Arkiv Kemi, 3, 315-322.
(37.) Benedict, C.R. (1962) Early products of [[sup.14]C]-acetate
incorporation in resting cells of Rhodospirillum rubrum. Biochim.
Biophys. Acta, 56, 620-622.
(38.) Hoare, D.S. (1963) The photo-assimilation of acetate by
Rhodospirillum rubrum. Biohem. J., 87, 284-301.
(39.) Ivanovsky, R.N., Krasilnikova, E.N. and Berg, I.A. (1997) A
proposed citramalate cycle for acetate assimilation in the purple
non-sulfur bacterium Rhodospirillum rubrum. FEMS Microbiol. Lett., 153,
(40.) Berg, I.A., Filatova, V. and Ivanovsky, R.N. (2002)
Inhibition of acetate and propionate assimilation by itaconate via
propionyl-CoA carboxylase in isocitrate lyase-negative purple bacterium
Rhodospirillum rubrum. FEMS Microbiol. Lett., 216, 49-54.
(41.) Osumi, T. and Katsuki, H. (1977) A novel pathway for
L-citramalate synthesis in Rhodospirillum rubrum. Biochem., 81, 771-778.
(42.) Meister, M., Saum, S., Alber, B.E. and Fuchs, G. (2005)
L-Malyl-Coenzyme A/[beta]-methylmalyl-coenzyme A lyase is involved in
acetate assimilation of the isocitrate lyasenegative bacterium
Rhodobacter capsulatus. J. Bacteriol., 187, 1415-1425.
(43.) Alber, B.E., Spanheimer, R., Ebenau-Jehle, C. and Fuchs, G.
(2006) Study of an alternate glyoxylate cycle for acetate assimilation
by Rhodobacter sphaeroides. Mol. Microbiol., 61, 297-309.
(44.) Erb, T.J., Berg, I.A., Brecht, V., Muller, M., Fuchs, G. and
Alber, B.E. (2007). Synthesis of [C.sub.5]-dicarboxylic acids from
[C.sub.2]-units involving crotonyl-CoA carboxylase/reductase: the
ethylmalonyl-CoA pathway. Proc. Natl. Acad. Sci. USA, 104, 10631-10636.
(45.) Erb, T.J., Brecht, V., Muller, M., Fuchs, G. and Alber, B.E.
(2009) Carboxylation mechanism and stereochemistry of crotonyl-CoA
carboxylase/reductase. Proc. Natl. Acad. Sci. USA, 106, 8871-8876.
(46.) Erb, T.J., Retey, J., Fuchs, G. and Alber, B.E. (2008).
Ethylmalonyl-CoA mutase from Rhodobacter sphaeroides defines a new
subclade of coenzyme [B.sub.12]-dependent acyl-CoA mutases. 1 Biol.
Chem., 283, 32283-32293.
(47.) Erb, T.J., Fuchs, G. and Alber, B.E. (2009)
[2S]-Methylsuccinyl-CoA dehydrogenase closes the ethylmalonyl-CoA
pathway for acetyl-CoA assimilation.
Mol. Microbiol., 73, 992-1008.
(48.) Erb, T.T., Frerichs-Revermann, L., Fuchs, G. and Alber, B.E.
(2010) The apparent malate synthase activity of Rhodobacter sphaeroides
is due to two paralogous enzymes, (3S)-Malyl-Coenzyme
A(CoA)/[beta]-Methylmalyl-CoA Lyase and (3S)-Malyl-CoA Thioesterase. J.
Bacteriol., 192, 1249-1258.
(49.) Alber, B.E. (2011) Biotechnological potential of the
ethylmalonyl-CoA pathway. Appl. Microbial. Biotechnol., 89, 17-25.
(50.) Cox, R.B. and Quayle, J.R. (1976). Synthesis and hydrolysis
of malylcoenzyme A by Pseudomonas AM1: an apparent malate synthase
activity. J. Gen. Microbiol., 95, 121-133.
(51.) Okubo, Y., Yang, S., Chistoserdova, L. and Lidstrom, M.E.
(2010) Alternative route for glyoxylate consumption during growth on
two-carbon compounds by Methylobacterium extorquens AM 1. J. Bacteriol.,
(52.) Smejkalova, H., Erb, T.J. and Fuchs, G (2010) Methanol
assimilation in Methylobacterium extorquens AM1: demonstration of all
enzymes and their regulation. PLoS ONE, 5(10), e13001.
(53.) Skovran, E., Crowther, G.J., Guo, X., Yang, S. and Lidstrom,
M.E. (2010). A systems biology approach uncovers cellular strategies
used by Methylobacterium extorquens AM1 during the switch from multi- to
single-carbon growth. PLoS One, 5(11), e14091.
(54.) Peyraud, R., Kiefer, P., Christen, R, Massou, S., Portais,
J.C. and Vorholt, J.A. (2009). Demonstration of the ethylmalonyl-CoA
pathway by using [sup.13]C metabolomics. Proc. Natl. Acad. Sci. USA,
(55.) Anthony, C. (1978). The prediction of growth yields in
methylotrophs. J. Gen. Microbiol., 104, 91-104.
Christopher Anthony is Emeritus Professor of Biochemistry,
University of Southampton, UK. His main interests are the biochemistry
of methylotrophs during growth on [C.sub.1] and [C.sub.2] compounds. He
first discovered their unusual enzyme for oxidation of methanol
(methanol dehydrogenase) and its novel prosthetic group (PQQ). He
subsequently concentrated on this quinoprotein, and related
quinoproteins with their associated electron transport chains, using the
techniques of continuous culture, spectrophotometry, X-ray
crystallography, and molecular genetics. Email: C.Anthony@soton.ac.uk;
Dedication: I should like to dedicate this historical review to
Hans L. Kornberg and J. Rod Quayle.
Table 1 Summary of the carbon and energy balances of the
assimilation pathways. Growth substrates must provide acetyl-CoA,
for synthesis of fatty acids and 3-hydroxybutyrate (for carbon
storage as poly 3-hydroxybutyrate, and C3 and C4 compounds for
synthesis of carbohydrates, amino acids and nucleic acids. The
relative requirements for these metabolic precursors will depend on
the growth conditions. The pathways and carbon balances in this
review have been presented in the Figures as originally described
but the balances in this Table have been normalised for production
of oxaloacetate (OAA) so that they may be compared more readily. It
has been assumed that production of thioesters requires ATP and the
production of AMP plus pyrophosphate, and that conversion of
succinyl-CoA to succinate produces ATP (although the energetically
equivalent GTP may be produced). Similarly [NAD.sup.+] and
[NADP.sup.+] are treated as being energetically equivalent and
written as [NAD(P).sup.+]. Water, and protons involved in
[NAD(P).sup.+]-linked reactions, are omitted. It should be noted
that during growth on [C.sub.2] compounds by way of acetyl-CoA
growth yields (g of cells/mole substrate consumed) will be limited,
as is usual with bacteria, by ATP. By contrast, during growth by
way of the serine cycles growth yields will be limited by NAD(P)H
(reductant). This is because reductant is required for [in effect]
reduction of C[O.sub.2], and because oxidation of the [C.sub.1]
substrates to HCHO provides no NAD(P)H (3,53,55)
The glyoxylate cycle for assimilation from acetyl-CoA (Fig. 1)
2acetyl-CoA + 2[NAD.sup.+] + FAD OAA + 2NADH + [FADH.sub.2] + 2CoA
The methylaspartate pathway for assimilation of acetyl-CoA (Fig. 6)
2acetyl-CoA + 2[NAD(P).sup.+] OAA + 2NAD(P)H + [FADH.sub.2] +
+ FAD + ATP AMP + [PP.sub.i] + 2CoA
The ethylmalonyl-CoA pathway for assimilation of acetyl-CoA
(Fig. 11 and 12)
3acetyl-CoA + 2C[O.sub.2] + HAD 20AA + 2[FADH.sub.2] + 3CoA
The serine cycle [ICL variant] for assimilation of formaldehyde
2HCHO + 2 C[O.sub.2] + 2NAD(P)H OAA + 2[NAD(P).sup.+] +
+ FAD + 3ATP [FADH.sub.2] + 2ADP + 2[P.sub.i]
+ AMP + [PP.sub.i]
The serine cycle [EMC variant] for assimilation of formaldehyde
(Figs. 13 and 14)
3HCHO + 5C[O.sub.2] + 6NAD(P)H + 20AA + 6[NAD(P).sup.+] +
2 FAD + 5ATP 2[FADH.sub.2] + 3ADP + 2AMP +
2 [PP.sub.i] + 3[P.sub.i]