Literature Article Nature, 2010 Explain the assigned literature article (Nature, 2010, 464, 1077), as if another student in CHEM 130B was your audience. An

Literature Article Nature, 2010 Explain the assigned literature article (Nature, 2010, 464, 1077), as if another student in CHEM 130B was your audience. An important component of scientific writing is to be concise, so you will have a 1-page limit to give a complete and accurate explanation. Remember that you may need to read sources that are cited or look at the Supporting Material for a full understanding of the paper. Please focus on the biochemical aspects of the paper rather than the phylogenetic studies, making sure to include key experiments and results. Vol 464 | 15 April 2010 | doi:10.1038/nature08884
LETTERS
Fructose 1,6-bisphosphate aldolase/phosphatase
may be an ancestral gluconeogenic enzyme
Rafael F. Say1 & Georg Fuchs1
Most archaeal groups and deeply branching bacterial lineages
harbour thermophilic organisms with a chemolithoautotrophic
metabolism. They live at high temperatures in volcanic habitats at
the expense of inorganic substances, often under anoxic conditions1.
These autotrophic organisms use diverse carbon dioxide fixation
mechanisms generating acetyl-coenzyme A, from which gluconeogenesis must start2–4. Here we show that virtually all archaeal groups
as well as the deeply branching bacterial lineages contain a bifunctional fructose 1,6-bisphosphate (FBP) aldolase/phosphatase with
both FBP aldolase and FBP phosphatase activity. This enzyme is
missing in most other Bacteria and in Eukaryota, and is heat-stabile
even in mesophilic marine Crenarchaeota. Its bifunctionality
ensures that heat-labile triosephosphates are quickly removed and
trapped in stabile fructose 6-phosphate, rendering gluconeogenesis
unidirectional. We propose that this highly conserved, heat-stabile
and bifunctional FBP aldolase/phosphatase represents the pacemaking ancestral gluconeogenic enzyme, and that in evolution gluconeogenesis preceded glycolysis5.
The theory of a chemoautotrophic origin of life by transitionmetal-catalysed, autocatalytic carbon fixation assumes that chemoevolution took place in a hot volcanic flow setting6,7. It has become
evident that in the phylogenetic tree of life the Archaea and deeply
branching lineages of the Bacteria harbour thermophiles that thrive
in volcanic environments on volcanic gases and inorganic substrates
under anoxic or microaerobic conditions1 (Fig. 1). From carbon
dioxide (CO2) or carbon monoxide (CO) they synthesize activated
acetic acid, acetyl-coenzyme A (acetyl-CoA), as biosynthetic starting
material. Their energy metabolism often makes use of hydrogen or
CO as electron donors, and CO2 or sulphur compounds serve as
electron acceptors in anaerobic respiration. Therefore, these chemolithoautotrophic organisms may serve as models for the study of
primordial metabolism requiring the synthesis of organic building
blocks from inorganic carbon. In none of these microorganisms does
the Calvin–Benson–Bassham cycle seem to operate in CO2 fixation.
Instead, other autotrophic pathways are functioning that have in
common the formation of acetyl-CoA from inorganic carbon2–4.
If so, the biosynthesis of sugars requires gluconeogenesis to start
from acetyl-CoA as a precursor. Pyruvate and phosphoenolpyruvate
(PEP) formation from acetyl-CoA and CO2 may differ2–4. In contrast,
gluconeogenesis starting from PEP seems to be uniform. All enzyme
activities and genes of a trunk Embden–Meyerhof–Parnas gluconeogenic pathway, which are necessary for the formation of FBP from
PEP, are assumed to be present in Archaea. This is in contrast to the
great diversity of glycolytic pathways and enzymes present in
Archaea8. However, it has been generally difficult or impossible to
detect FBP aldolase activity9. In many cases this enzyme activity could
only be measured in the direction of FBP formation, whereas detection of the reverse reaction, FBP cleavage, has failed9,10. This discrepancy is puzzling as the FBP aldolase reaction is freely reversible.
1
Furthermore, tracer studies with several autotrophic Archaea
revealed a labelling pattern of hexoses that was consistent with the
classical gluconeogenic route involving FBP aldolase10,11.
None of the archaeal genomes sequenced so far contains a classical
FBP aldolase of class I (Schiff base intermediate, mainly found in
Eukaryota) or class II (metal based catalysis, mainly in Bacteria and
Fungi) (Supplementary Table 1). However, one small group harbours the gene encoding a different archaeal class IA aldolase12,13; this
enzyme has a common evolutionary origin with class I and II aldolases14. Its function as FBP aldolase has been shown experimentally
only in three cases generally by measuring the glycolytic direction.
The other aldolase genes are only distantly related to this FBP aldolase
and may have a different function (for example, in the archaeal
aromatic biosynthesis pathway15; Supplementary Table 1). The
majority of archaeal genomes sequenced so far lack any proven
FBP aldolase gene, whereas generally the gene encoding an archaeal
type V FBP phosphatase is present16 (Supplementary Table 1). This
phosphatase also catalyses the sought-after FBP aldolase reaction, as
the following experiments show.
We searched for FBP aldolase plus FBP phosphatase activity in
extracts from autotrophically grown cells of the thermophilic Archaea
Ignicoccus hospitalis, Metallosphaera sedula and Thermoproteus neutrophilus, the central carbon metabolism of which has been studied
recently2–4,10. The assay, at 65–85 uC, is based on the measurement of
phosphate release, when triosephosphates were incubated with cell
extracts. However, it was difficult to measure FBP aldolase activity
owing to the heat instability of triosephosphates producing toxic
methylglyoxal and forming phosphate, which interfered with the assay.
The half-life of triosephosphates at 80 uC, pH 7, was confirmed to be
4 min12. Furthermore, we could not detect the reverse reaction; that is,
FBP cleavage.
However, FBP aldolase and phosphatase activities were readily
detected using a discontinuous spectrophotometric assay for fructose
6-phosphate formation (for assay see Supplementary Information).
The specific FBP aldolase activities were 0.02 mmol min21 (per mg of
protein; 85 uC) in I. hospitalis, 0.006 mmol min21 (per mg of protein;
75 uC) in M. sedula, and 0.02 mmol min21 (per mg of protein; 85 uC) in
T. neutrophilus. The specific FBP phosphatase activities were 1.5-fold
as high. To convert efficiently the aldol reaction product FBP into
fructose 6-phosphate, we added purified recombinant archaeal type
V FBP phosphatase from I. hospitalis (Igni_0363) as auxiliary enzyme
to the assay to pull forward the reaction. Only FBP phosphatase
activity was observed when the intermediate FBP was supplied as
substrate. Surprisingly, addition of triosephosphates resulted in a
burst of FBP aldolase activity, even when cell extract was omitted.
The archaeal type V FBP phosphatase has been studied in detail16,17
and its crystal structure was solved18; however, the main physiological
function of the enzyme seems to have gone unnoticed. We meticulously purified the enzyme so that it showed no contaminating protein
Mikrobiologie, Fakulta?t Biologie, Universita?t Freiburg, Scha?nzlestraße 1, D-79104 Freiburg, Germany.
1077
©2010 Macmillan Publishers Limited. All rights reserved
LETTERS
NATURE | Vol 464 | 15 April 2010
a
Euryarchaeota
Archaeoglobus
fulgidus
Thermoplasmatales
*
Halobacteriales
Methanosarcinales
Thermococcales
Methanomicrobiales
Methanopyrus
kandleri
*
68
Methanobacteriales
54 60
*
Sulfolobales
Ko
Na
no
ar
c
eq hae
ra
u
u
cr rch ita m
yp a ns
to eu
filu m
m
Methanococcales
Desulfurococcales
Thermoproteales
*
*
*
Mesophillic group 1
Crenarchaeota
Crenarchaeota
b
Bacteria
Cyanobacteria
*
Actinobacteria
Firmicutes
Fu
so
*
ba
ct
ia 58 ?-Proteobacteria
?-Proteobacteria
29
62
Aquificae
?-Proteobacteria
46
Thermotogae
er
hl
C
Bacteroidetes
or
ob
i
Deinococcus–
Thermus
Planctomycetes
Spirochaetes Chlamydiae
?-Proteobacteria
?-Proteobacteria
Figure 1 | Phylogenetic unrooted trees of Archaea and Bacteria. Boxes
indicate phyla containing FBP aldolase/phosphatase. Asterisks indicate
phyla from which enzymes were studied. Only bootstrap values ,70% are
indicated. a, Archaebacterial tree based on analyses of 64 conserved proteins
(59 genomes). Red lines represent (hyper)thermophilic Archaea (.65 uC);
blue lines mesophilic Archaea. Autotrophic lineages are marked with a dot.
b, Eubacterial tree based on concatenated 37 ribosomal proteins (120
genomes30). Note that the systematic positions of the Thermotogae,
Aquificae, Chloroflexi (not shown on the tree), the Deinococcus–Thermus
lineage and in some respects also the deeply branching, thermophilic,
acetogenic Clostridia/Firmicutes23 are considered as early branching22.
(Supplementary Fig. 1 and Supplementary Table 2), but it still catalysed both aldolase and phosphatase reactions, demonstrating bifunctionality. The kinetic constants at 48 uC for I. hospitalis FBP aldolase
activity were maximal velocity (vmax) of 0.54 mmol min21 (per mg of
protein) and Michaelis constant (Km) of 0.23 mM for triosephosphates. The FBP phosphatase activity (vmax of 0.88 mmol min21
(per mg of protein); Km of 0.02 mM for FBP) was nearly twice as high
as the aldolase activity, as observed in cell extract (for optimal pH see
Supplementary Fig. 2).
As further proof of bifunctionality, recombinant putative FBP
phosphatases were overproduced and partly purified from five other
Archaea covering different lineages (Fig. 2). We also expressed the
synthetic FBP phosphatase gene of Cenarchaeum symbiosum19, a
member of the marine group I Crenarchaeota (Fig. 2); similar genes
allocated to this marine group occur in high numbers in the GOS
databases. All six recombinant archaeal enzymes exhibited both FBP
aldolase and phosphatase activities. This enzyme, now referred to as
FBP aldolase/phosphatase, was previously considered as a potential
determinant of hyperthermophily, besides reverse gyrase, a type I
DNA topoisomerase able to stabilize DNA at high temperature by
introducing positive supercoils20. Notably, even the Cenarchaeum
enzyme, adapted to ocean temperature, was heat-stabile at 70 uC,
with a half-life of 20 min at 82 uC, and the Ignicoccus enzyme even
survived boiling for 1 h (Supplementary Fig. 3). The catalytic properties of the Cenarchaeum enzyme were analysed at 40 uC and are
shown in Fig. 3. AMP, ADP or glucose (2 mM), known allosteric
regulators of FBP phosphatases, had no effect. The lower activation
energy of the Cenarchaeum enzyme compared to the Ignicoccus
enzyme (Supplementary Fig. 4) indicates that these enzymes are
adapted to the respective cold or hot optimal temperature for growth.
The specific activity of the enzyme in cell extract is generally low,
reflecting the minor need for carbohydrates in Archaea. Even in
Escherichia coli, with its high content of sugars in lipopolysaccharides,
the biosynthetic fluxes leading away from hexosephosphates add up
to only ,13% of all biosynthetic fluxes.
A data base search of sequenced genomes revealed that almost all
Archaea contain the corresponding FBP aldolase/phosphatase gene
(e-values ,10275), except for halophilic and a very few methanogenic
Archaea that harbour, in most cases, the genes for other types of FBP
aldolases and phosphatases (Supplementary Table 1). In most
Archaea that do not grow on sugars the FBP aldolase/phosphatase
gene is the only candidate gene for both FBP aldolase and FBP phosphatase. A look at the regulation of the FBP aldolase/phosphatase gene
corroborates our expectation. Thermococcus kodakarensis forms this
enzyme solely under gluconeogenic, but not under glycolytic, conditions16,17 (using the Embden–Meyerhof–Parnas pathway)8. A deletion
mutant could grow under glycolytic, but not under gluconeogenic,
conditions. Yet, complementation of this mutant by a different monofunctional FBP phosphatase did not restore growth under gluconeogenic conditions17. This is consistent with the notion that the deleted
enzyme has, in addition, FBP aldolase activity. In contrast, Sulfolobus
solfataricus uses a branched Entner–Doudoroff pathway for glycolysis8, in which FBP is not an intermediate, and therefore may
tolerate constitutive expression of FBP aldolase/phosphatase21.
A gene highly similar to the archaeal FBP aldolase/phosphatase gene
(e-values ,10280) is also present in members of the deeply branching
bacterial phyla22 (Fig. 2). They include genera of Aquificae (Aquifex,
Hydrogenobaculum, Hydrogenivirga), Thermotogae (Petrotoga),
Chloroflexi (Roseiflexus, Dehalococcoides), the Deinococcus–Thermus
group (Thermus), as well as the mostly homoacetogenic, thermophilic
Clostridia/Firmicutes (Carboxydibrachium, Thermoanaerobacter,
Moorella, Pelotomaculum, Carboxydothermus, Natranaerobius) that
also exhibit traits of a deeply branching phylum23,24 (for phylogenetic
positions of the phyla see Fig. 1b). As a proof of concept, we overproduced the enzymes from Thermus thermophilus and Moorella thermoacetica and showed that they were an FBP aldolase/phosphatase,
like the archaeal enzyme. There are only rare exceptions to the rule that
this bifunctional enzyme is restricted to the Archaea and deeply
branching, mostly thermophilic and autotrophic Bacteria (e-value
,10260). Probably all those bacteria have to perform a unidirectional
gluconeogenesis from C2 or C3 compounds under some conditions,
which may have favoured the acquisition of the gene by lateral transfer
(for example, from mesophilic group 1 Crenarchaeota). Examples
are the pathogenic Coxiella burnetii (c-Proteobacteria), the denitrifying Nitrococcus mobilis (c-Proteobacteria) and the symbiotic
Bradyrhizobium japonicum (a-Proteobacteria) (Fig. 2). Syntrophs
like Syntrophus aciditrophicus (d-Proteobacteria) depend on a close
spatial contact with other Bacteria or (methanogenic) Archaea, which
favours not only interspecies hydrogen transfer but also lateral gene
transfer. Saccharopolyspora erythraea (Actinobacteria) contains many
1078
©2010 Macmillan Publishers Limited. All rights reserved
LETTERS
mu
um
xie
ja
po
lla
ni
c
b
he
rm urn um
e
o
aq phil tii
u
u
ati
s
the
cu
Nitr
rm
s
op
oco
h
ccu
Syntr
s m ilus
ophu
obil
s ac
is
iditro
Hydro
phic
genob
us
aculu
m sp.
Aquifex ae
olicus
Hydrogenivirga sp.
Na
tra
Th
nae
Aquificae
st
erm
rob
ius
us
ha
r
cc
‘M
tus
ida
nd
Ca
is
lud
nnie
lii
ripa
ma
Sa
us
i
ex
M
Methanopyrus kandleri
Me
tha
no
cal
dococcus jan
naschii
Metha
no
co
ccus ae
Meth
olicus
ano
c
o
Meth
c
cus v
ano
oltae
coc
Me
cus
tha
va
n
oco
cc
ofl
lor
Ch
De
ha
lo
op
co
oly
cc
sp
eth
o
or
Me
an
a e ide
tha
or
ryt s s
eg
no
p
hr
ula
cu
ae .
lleu
a
s m boo
aris nei
nig ’
ri
Carboxydothermus hydrogenoformans
m
opropionicu
ulum therm
Pelotomac
xviator’
dis auda
ru
a
fo
c
ul
ti
oace
tus ‘Des
us
therm
Candida
olic
rella
Moo
than nsis
e
d
u
ge m
pse
con
cter
g
cu
a
n
b
te
cifi sp.
ero
ter
pa
ana
m
us lzii
bac
rmo
ex
hiu
The
ero
o
fl
a
c
i
n
a
a
nh nes
se
ibr
rmo
Ro
ste ge
xyd
a
The
o
c
no
rb
s
a
e
u
C
h
x
et
ifle
es
se
id
Ro
co
c
co
lo
ha
De
eth
an o
gen
m
ka
s
s
cu
en
oc
nd
sis
c
e
en
ro m p
ng
u
i
lf
u
uil
su ofil
aq
ax
e
n
m
m
D er
e
ga
st
um
eu
Th ldivir
rot
ndic
isla
Ca rmop
ilus
oph
lum
e
u
h
u
c
e tr
T
oba
us n
Pyr
rote
s
p
ti
o
n
o
rm
alidif
The
lum c
bacu
Pyro
rophilum
culum ae
ba
ro
Py
arsenaticum
Pyrobaculum
Thermoplasma volcanium 1
Thermoplasma volc
anium 2
Thermop
lasma ac
idophilu
Picro
m
philu
s torr
Meth
idus
anos
Arc
aeta
hae
therm
oglo
Pe
oph
tro
bus
ila
tog
Ac
fulg
am
idu
ob
Ac iduli
s
il
p
is
rof
Th idu
u
l
er
nd
m ipro
um
fu
oc
bo
oc ndu
on
cu
m
ei
sg
b
1
am oon
ei
m
2
at
ol
er
an
s
s
.
eu
in
sp
s
i
ur
s
ns
nn
cu
are
oc us o
ak
c
oc
od
us
m coc us k
hil
er
c
o
rop
Th m
ba
oc
er
oc
us
us
cc
Th erm
ios
co
fur
Th rmo
us
e
i
occ
Th
yss
roc
s ab
Py
shii
ccu
oco
oriko
Pyr
us h
r sp.
cocc
obacte
Pyro
rm
nothe
Metha
dmanae
phaera sta
Methanos
r smithii
Methanobrevibacte
icutes
Firm
bi
Co
er
is
ns
ke
at
ch
ta
Th
zo
rc
ha
eo
Bac
ter
ia
hi
ur
ya
yr
Candidatus ‘Korarchaeum
cryptofilum’
Sulfolobus
tokodaii
Sulfolo
bus ac
idocald
Meta
llosp
arius
haera
Sulf
sedu
olob
la
Su
u
s is
lfolo
la
n
Ign
dicu
bu s
i
s
solf
Hy cocc
ata
us
ricu
Ae pert
ho
s
sp
St rop herm
ita
y
ap
lis
us
hy rum
bu
lo
pe
tyl
th
icu
rn
er
i
x
s
m
us
m
ar
in
us
De
in
Th o c o
er cc
m u
us s
ad
ta
Br
Cr
en
ar
o
ae
ch
um
bios e
sym
ot
aeum
hae
arc
arch
us
n
Cen
cre
ritim 2
red
ma
te
ultu
eo
ilus
Unc
ha
e3
um
ot 1
arc
op
ae te
ro s
ren
ch eo
Nit
ec
ar
a
rin
en rch
ma
cr
a
red
ine cren
ltu
ar
e
m
cu
in
d
Un
ar
re
m
ltu
d
cu
re
Un
lt u
cu
Un
C
–
ne ota
ari ae
M arch
n
re
Korarchaeota
NATURE | Vol 464 | 15 April 2010
E
ic Euryarchaeota
Figure 2 | Phylogenetic tree of FBP aldolases/phosphatases (compare with
the ribosomal proteins trees in Fig. 1). Species in red are members of earlybranching lineages of the Bacteria; the Clostridia cluster is shown in black
characters. Species for which their enzymes were overproduced and studied
here are underlined. The tree was constructed using the neighbour-joining
algorithm. When maximum-parsimony or maximum-likelihood algorithms
were used, the same groups were obtained. Bootstrap values higher than 75%
are marked with dots. (Hyper)thermophiles are marked by red lineages;
autotrophs by blue boxes.
integrative and conjugative elements25, which may have facilitated the
acquisition of the FBP aldolase/phosphatase gene.
The phylogenetic tree of the enzyme (Fig. 2) largely corresponds to
the phylogenetic archaeal (Fig. 1a) and bacterial (Fig. 1b) ribosomal
proteins trees. Note that the phylogenetic position of Nanoarchaeota,
Korarchaeota and ‘marine group I’ Crenarchaeota is currently under
discussion. Both ribosomal proteins and enzyme trees clearly separate ‘marine Crenarchaeota’ from Crenarchaeota, which were suggested recently to form a new archaeal phylum26. The apparent
association of members of Methanomicrobiales (Methanoculleus
marisnigri and Candidatus ‘Methanoregula boonei’) with Bacteria in
Fig. 2 is probably not significant. The absence of the gene in the
heterotrophic Halobacteriales and in several Methanosarcinales
and Methanomicrobiales (Euryarchaeota) is interpreted as loss of
the gene in these derived phyla. Loss of function is expected for
Nanoarchaeum equitans, which has lost the genes for all biosynthetic
pathways27. The presence of the gene in some but not all major
bacterial phyla can reflect either loss (in the late-branching phyla)
or gain (in the early-branching lineages) since they diverged from a
common ancestor. It is impossible to conclude this case here with
certainty because of the low bootstrap values in the deep branches.
Yet, the marked coincidence of the presence of the highly conserved
gene in the deeply branching (mostly autotrophic and thermophilic)
bacterial lineages as well as the distinct lineages in the phylogenetic
enzyme tree support loss in the late-branching lineages rather than
gain in the early bacterial lineages. Why of all phyla should these
deep branches have acquired the gene? The data may, however, be
interpreted differently, as indicating that the gene is archaeal specific
and that very early it was transferred twice laterally from Archaea to
Bacteria, followed by independent vertical transfer. Such an ambiguous
situation is similar to the chimaeric nature of Thermotogales28.
Whereas ribosomal protein genes strongly place Thermotogales as a
sister group to Aquificales, the majority of genes with sufficient phylogenetic signal show affinities to Archaea and Clostridia/Firmicutes.
Many of the bacterial species harbouring the gene (Fig. 2) are members
of the deeply branching Clostridia/Firmicutes23. The huge impact of
lateral gene transfer, often unrecognized, on prokaryote genome evolution has been impressively documented29.
What makes FBP aldolase/phosphatase a peculiar aldolase? The
enzyme contains four Mg21 ions required to bind the phosphate
residue (C1-phosphate in FBP and C3-phosphate in dihydroxyacetone phosphate (DHAP))18 and is consequently inactivated by EDTA
(80%; a typical feature of class II enzymes). The crystal structure of
the Sulfolobus tokodaii enzyme18 shows that substrate binding
requires the interaction of two subunits (Fig. 4a). A conserved lysine
(Lys 133) and tyrosine (Tyr 348) approximate to the C2 and C4
hydroxyl groups of FBP may be essential for catalysis (Fig. 4b and
Supplementar…
Purchase answer to see full
attachment