Metabolism of pentose sugars in the hyperthermophilic archaea Sulfolobus solfataricus and ...

20 Pages·2010·0.31 MB·English

JBC Papers in Press. Published on August 24, 2010 as Manuscript M110.146332
The latest version is at http://www.jbc.org/cgi/doi/10.1074/jbc.M110.146332
Metabolism of pentose sugars in the hyperthermophilic archaea Sulfolobus solfataricus
and Sulfolobus acidocaldarius

Charlotte E.M. Nunn1, Ulrike Johnsen2, Peter Schönheit2, Tobias Fuhrer3, Uwe Sauer3, David W.
Hough1 and Michael J. Danson1

1 Centre for Extremophile Research, Department of Biology and Biochemistry, University of Bath,
Bath, BA2 7AY, U.K.

2 Institut für Allgemeine Mikrobiologie, Christian-Albrechts-Universität Kiel, Am Botanischen
Garten 1-9, D-24116 Kiel, Germany

3 Institute for Molecular Systems Biology, ETH Zürich, CH-8093 Zürich, Switzerland

Running title: Pentose metabolism in Sulfolobus

Communicating author:
D
o
w
Professor Michael J. Danson n
lo
Centre for Extremophile Research ad
e
d
Department of Biology and Biochemistry fro
University of Bath m
h
Bath, BA2 7AY, U.K. ttp
://w
w
Tel: [phone] w
FAX: [phone] .jbc
.o
Email: [email protected] rg
b/
y
g
u
e
s
t o
n
J
a
n
u
a
ry
3
, 2
0
1
9
Copyright 2010 by The American Society for Biochemistry and Molecular Biology, Inc.
ABSTRACT production of ATP (Fig. 1) (2-5). Glucose
We have previously shown that the dehydrogenase catalyses the conversion of
hyperthermophilic archaeon, Sulfolobus glucose to gluconate, which is then dehydrated
solfataricus, catabolises D-glucose and D- to 2-keto-3-deoxygluconate (KD-gluconate) by
galactose to pyruvate and glyceraldehyde via a gluconate dehydratase. KD-gluconate in turn is
non-phosphorylative version of the Entner- cleaved to pyruvate and glyceraldehyde via 2-
Doudoroff pathway. At each step, one enzyme is keto-3-deoxygluconate aldolase (KDG-
active with both C6 epimers, leading to a aldolase), the glyceraldehyde generating a
metabolically-promiscuous pathway. On further second molecule of pyruvate via the actions of
investigation, the catalytic promiscuity of the glyceraldehyde oxidoreductase, glycerate
first enzyme in this pathway, glucose kinase, enolase and pyruvate kinase.
dehydrogenase, has been shown to extend to the In vitro kinetic analyses of glucose
C5 sugars, D-xylose and L-arabinose. In the dehydrogenase, gluconate dehydratase and
current paper, we establish that this promiscuity KDG-aldolase from S. solfataricus showed
for C6 and C5 metabolites is also exhibited by these enzymes are also capable of catalysing
the third enzyme in the pathway, 2-keto-3- the catabolism of galactose, the C4 epimer of
deoxygluconate aldolase, but that the second step glucose, to pyruvate and glyceraldehyde,
requires a specific C5-dehydratase, the gluconate leading to the suggestion that the pathway
dehydratase being active only with C6 exhibits a metabolic promiscuity towards these
metabolites. The products of this pathway for two hexose sugars (5-7). Since recombinantly- D
o
w
the catabolism of D-xylose and L-arabinose are produced glucose dehydrogenase has good n
lo
pyruvate and glycolaldehyde, pyruvate entering activity with the pentose sugars D-xylose and a
d
e
tdheec arbocixtyrilca tiona ctoid a cetcyyl-ccloe enzyamfteer A. oWxied ahtaivvee Lth-ea rabcoinnodseen s(a6t)i,o na ndo fK DthGe- aldCo-2la sec ocmatpaolyusneds d from
h
identified and characterised the enzymes, both glycolaldehyde and pyruvate to form a 2-keto- ttp
native and recombinant, that catalyse the 3-deoxypentanoate (H.J. Lamble, D.W. Hough, ://w
w
conversion of glycolaldehyde to glycolate and S.D. Bull and M.J. Danson, unpublished data), w
then to glyoxylate, which can enter the citric acid both enzymes are likely to have a role in the .jb
c
.o
cycle via the action of malate synthase. catabolism of the naturally occurring pentose rg
Evidence is also presented that similar enzymes sugars, D-xylose and L-arabinose, as well as of by/
g
for this pentose sugar pathway are present in the hexose sugars D-glucose and D-galactose. u
e
s
Sulfolobus acidocaldarius, and metabolic tracer However, the purified gluconate dehydratase t o
n
studies in this archaeon demonstrate its in vivo was found to be specific for gluconate and Ja
n
operation, in parallel with a route involving no galactonate (7). ua
ry
aldol cleavage of the 2-keto-3-deoxy-pentanoates A number of questions arise from these 3
, 2
but direct conversion to the citric acid cycle C5- observations. Does Sulfolobus possess a 0
1
9
metabolite, 2-oxoglutarate. dehydratase specific for D-xylonate and L-
arabinonate, thereby establishing a pathway for
C5-sugar catabolism that uses the same glucose
INTRODUCTION dehydrogenase and KDG-aldolase involved in
Sulfolobus solfaraticus and Sulfolobus the utilization of C6 sugars? If so, what is the
acidocaldarius are hyperthermophilic archaea metabolic fate of glycolaldehyde? Finally,
that grow optimally at 78-85°C, pH 2-4, and are what is the relationship of this possible
able to utilise a variety of carbon sources, metabolic route to that proposed by Brouns et
including the four most-commonly occurring al (8) for the catabolism of the non-natural
sugars in nature, D-glucose, D-galactose, D- isomer, D-arabinose, by S. solfataricus,
xylose and L-arabinose (1). whereby the 2-keto-3-deoxy-arabinonate is not
Metabolism of glucose in S. solfataricus aldol-cleaved but is converted to 2-oxoglutarate
and S. acidocaldarius proceeds via a non- through the action of a KD-arabinonate
phosphorylative variant of the Entner-Doudoroff dehydratase and 2,5-dioxopentanoate
pathway, which generates pyruvate with no net dehydrogenase?
3
In the current paper, we demonstrate that Growth of S. acidocaldarius and preparation
Sulfolobus possesses a C5-specific dehydratase of cell extracts
that catalyses the dehydration of D-xylonate and S. acidocaldarius (DSM 639) was grown
L-arabinonate to 2-keto-3-deoxy-D-xylonate aerobically in Erlenmeyer flasks at 72°C in the
(KD-xylonate), and 2-keto-3-deoxy-L- presence of 20 mM D-xylose. The medium
arabinonate (KD-arabinonate), respectively, and contained (per litre): 0.25 g MgSO 7H O, 1.7 g
4 2
the enzymes glycolaldehyde oxidoreductase and K SO , 0.028 g FeSO .7H O, 1 g (NH ) SO ,
2 4 4 2 4 2 4
glyoxylate reductase to catalyse the conversion 0.28 g KH PO , 0.07 g CaCl .2H O and 1 ml
2 4 2 2
of glycolaldehyde to glycolate and then to trace element solution (per litre: 15 g MgSO 7
4
glyoxylate, which enters the citric acid cycle via H O, 1 g MnSO .H O, 5 g NaCl, 0.9 g CoCl .6
2 4 2 2
the action of malate synthase. The genes for two H O, 0.05 g H BO , 0.9 g ZnSO .7 H O, 0.05 g
2 3 3 4 2
of these three enzymes have been cloned and Na MoO .2H O, 0.13 g NiCl .6H O, 0.0015 g
2 4 2 2 2
expressed, and the recombinant enzymes Na SeO .5H O). The medium was adjusted to
2 3 2
characterised. Furthermore, through in vivo pH 2.5 with H SO . Cell extracts were
2 4
metabolite tracer studies, we have shown that D- prepared as described for S. solfataricus.
xylose is partitioned between this newly-
discovered pathway, involving aldol cleavage Enzyme assays
and subsequent glyoxylate formation, and the Unless stated otherwise, the conditions stated
pathway identified for the metabolism of D- below are those used for the assays of enzymes
arabinose involving direct conversion to 2- from S. solfataricus; minor variations were D
o
oxoglutarate (8). used for the assays at 50-60°C of enzymes from wn
lo
S. acidocaldarius. ad
e
EXPERIMENTAL PROCEDURES Xylose dehydrogenase activity was determined d from
h
as previously described (6). Assays were ttp
Growth of S. solfataricus and preparation of performed at 70°C in 1 ml of 100 mM HEPES, ://w
w
cell extracts pH 8.0. The assay mixture contained 0.5 mM w
Freeze-dried S. solfataricus P2 (DSM 1617) was NADP+ and 5 mM D-xylose, and the reaction .jbc
.o
obtained from DSMZ (Germany) and grown was started by the addition of enzyme. The rg
aerobically in Erlenmeyer flasks at 78°C in basal production of NADPH was followed by the by/
salt medium (9) containing (per litre) 0.28 g increase in absorbance at 340nm (ε = 6,220 M-1 gu
e
s
KH PO , 1.30 g (NH ) SO , 0.25 g MgSO .7H 0, cm-1). One unit of enzyme activity is defined t o
2 4 4 2 4 4 2 n
0.25 g CaCl2.7H20, adjusted to pH 3.5. The as the amount of enzyme required to produce 1 Jan
u
medium also contained a 1% (v/v) trace element µmol of NADPH per minute. a
ry
solution (20.0 mg FeCl .6H 0, 4.5 mg 3
3 2 , 2
Na B O .10H 0, 1.8 mg MnCl .4H 0, 0.22 mg Xylonate dehydratase activity was determined 0
2 4 7 2 2 2 1
9
ZnSO .7H 0, 0.05 mg CuCl .2H 0, 0.03 mg using the method of Lamble et al (7). Assays
4 2 2 2
Na MoO .4H 0, 0.03 mg VOSO .2H 0, 0.01 mg were performed in 400µl of 100 mM Bis-Tris,
2 4 2 4 2
CoSO .7H 0), 0.05% (w/v) yeast extract and 2% pH 7.0, containing 10 mM MgCl and 5 mM D-
4 2 2
(w/v) D-glucose, D-xylose, or L-arabinose. Cells xylonate. The reaction was started by the
were harvested by centrifugation at 5,000 g for addition of enzyme and was incubated for 10
10 min and resuspended at 0.2 g/ml in 50 mM min at 70°C before being stopped by
Tris-HCl, pH 8.0, and incubated for 30 min. transferring 100 µl of the reaction mixture to 10
Cells were lysed by four 30-second bursts of µl of 12% (w/v) trichloroacetic acid. After
sonication using a 150-watt Ultrasonic clarification by centrifugation at 16,000 g for 5
Disintegrator (MSE Scientific Instruments, min, the formation of KD-xylonate was
Crawley, UK) and the soluble fraction of the cell quantified spectrophotometrically at 549 nm (ε
extract obtained by centrifugation at 12,500 g for = 67,800 M-1 cm-1) by the reaction with
30 min. thiobarbituric acid (10).
4
2-Keto-3-deoxyxylonate aldolase activity was
assayed in the direction of aldol synthesis using a Isocitrate lyase activity was determined as
thiobarbituric acid assay as described in (10). described in (12). The assay was performed at
Assays were carried out at 70°C in 50 mM 70°C in 1 ml of 100 mM Bis-Tris, pH 7.5,
sodium phosphate buffer, pH 6.0, with 50 mM containing 100 mM NaCl and 20 mM MgCl .
2
pyruvate and 10 mM glycolaldehyde. The assay mixture contained 2.5 mM
phenylhydrazine, 15 mM isocitrate, and the
Two enzymatic activities were measured for the reaction was started by the addition of enzyme.
conversion of glycolaldehyde to glycolate. The reaction was followed by the increase in
Aldehyde oxidoreductase was assayed at 70°C in absorbance at 324nm (ε = 17,000 M-1 cm-1)
I ml of 100mM Bis-Tris buffer, pH 6.8, 2mM corresponding to the production of 2-
2,6-dichlorophenolindophenol (DCPIP) and (phenylhydrazono)-acetate. One unit of
5mM glycolaldehyde; the assay was started by enzyme activity is defined as the amount of
the addition of substrate, and the reduction of the enzyme required to produce 1 µmol of
DCPIP was followed at 595nm (ε = 22,000 M-1 glyoxylate per minute.
cm-1). Glycolaldehyde dehydrogenase activity
was determined at 70°C in 1ml of 50mM 2,5-Dioxopentanoate dehydrogenase (2-
potassium phosphate buffer, pH 6.0. The assay oxoglutarate semialdehyde dehydrogenase)
mixture contained 1 mM NADP+ and 50 mM assays were performed at 70°C in 1 ml of
glycolaldehyde, and the reaction was started by 100mM HEPES, pH7.5, as described in (8). D
o
the addition of enzyme. The reduction of The assay mixture contained 1 mM NADP+ and wn
lo
NADP+ was followed by the increase in 10 mM pentanedial, and the reaction was ad
e
aubnsito robfa ennczey amt e3 4a0ctnimvi t(yε i=s d6e,2fi2n0e dM a-1s cthme- 1a).m oOunnet sretadrutecdti onb y oft heN AaDddPi+t iown aso f foellnozwyemde . by Tthhee d from
h
of enzyme required to produce 1 µmol of increase in absorbance at 340nm. One unit of ttp
reduced DCPIP or NADPH per minute. enzyme activity is defined as the amount of ://w
w
enzyme required to produce 1µmol of NADPH w
Glycolate dehydrogenase (glyoxylate reductase) per minute. .jbc
.o
activity was determined as described in (11). rg
Assays were performed at 70°C in 1 ml of 100 Gene cloning and expression of recombinant by/
g
mM HEPES, pH 7.5. The assay mixture enzymes ue
s
contained 0.2 mM NADH and 10 mM All genes in this study were PCR-amplified t o
n
glyoxylate, and the reaction was started by the from S. solfataricus genomic DNA. PCR- Ja
n
addition of enzyme. The oxidation of NADH amplification was carried out using Phusion ua
ry
was followed by the decrease in absorbance at followed by A-tailing with Taq polymerase. 3
, 2
340nm. One unit of enzyme activity is defined Amplified DNA fragments were purified by 0
1
9
as the amount of enzyme required to produce 1 electrophoresis in a 1% (w/v) agarose gel,
µmol of NAD+ per minute. extracted using the Qiaex II Gel Extraction kit
(Qiagen Ltd, Crawley, UK) and, unless
Malate synthase activity was determined as otherwise stated, were ligated into the pGEMT-
described in (12). The assay was performed at Easy intermediate cloning vector. After
70°C in 1 ml of 20 mM EPPS, pH 8.0, 100 mM sequence determination to ensure PCR fidelity,
KCl. The assay mixture contained 0.1 mM fragments were ligated into pET vectors as
DTNB, 0.14 mM acetyl-coenzyme A (acetyl- detailed below; unless otherwise stated,
CoA) and 10 mM glyoxylate, and the reaction restriction sites were used that removed the
was started by the addition of enzyme. The His-tag from the vector.
increase in absorbance at 412 nm (ε = 13,600 M-1
cm-1) was followed with time, one unit of D-Xylonate dehydratase
enzyme activity being defined as the amount of A putative second dehydratase gene (SS02665)
enzyme required to produce 1 µmol of CoASH was identified through a BLAST search of the
per minute. S. solfataricus genome using the gluconate
5
dehydratase gene sequence (SSO3198). The resulting cell extract was then incubated at
gene was cloned into pET3a and pET19b vectors 80°C for 15 min to remove the majority of
(Novagen, Nottingham, UK) using the Nde1 and E.coli proteins, before being subjected to gel
BamH1 sites to produce both untagged and His- filtration chromatography on a GE Healthcare
tagged constructs, respectively. The plasmids Superdex 200 column (1 x 30 cm) using 50
were transformed into a series of host E. coli mM Tris-HCl buffer, pH 8.5, containing 100
strains including BL21(DE3), KRX, Rosetta, mM NaCl.
C41 and C43. Strains containing either vector
were grown in LB or Overnight Express media Malate synthase
(Novagen, Nottingham, UK) containing The putative malate synthase gene (SSO1334)
appropriate antibiotics. LB cultures were grown annotated in the genome of S. solfataricus was
with and without IPTG induction. cloned into pET19b using the NcoI and XhoI
sites. The gene was expressed in E. coli
2,5-dioxopentanoate (2-oxoglutarate Rosetta cells grown at 37°C in Overnight
semialdehyde) dehydrogenase Express medium to produce ~50% soluble
The 2,5-dioxopentanoate dehydrogenase gene recombinant enzyme. Cells were harvested and
(SSO3117) annotated in the genome of S. a cell extract was prepared and heated treated at
solfataricus was cloned into pET28a (Novagen) 80°C for 10 min as described above. The
using the Nde1 and NcoI sites. The gene was recombinant malate synthase was then purified
expressed in E. coli Rosetta cells grown at 37°C by anion exchange chromatography on a D
o
w
in Overnight Express medium. A cell extract HiTrap Q Sepharose column (GE Healthcare) n
lo
was prepared as described previously and then using 50 mM Tris-HCl, pH 8.5, and an elution a
d
e
cinocliu bparotetedi nast. 70°C for 10 min to precipitate E. gfirlatrdaiteinotn oonf a0 -S1uMpe rdNeaxC l2,0 0f oclololuwmedn (b1y x g3e0l d from
h
The gene was also cloned from S. cm) in 50 mM Tris-HCl, pH 8.5. ttp
acidocaldarius into pET17b using NdeI and ://w
w
BamHI sites. The gene was expressed in E. coli Isocitrate lyase w
Rosetta-pLysS cells grown at 37°C for 4 h. After The putative isocitrate lyase gene (SSO1333) .jb
c
.o
preparation of a cell extract, the recombinant annotated in the genome of S. solfataricus was rg
protein was purified by heat precipitation at 72°C cloned into pET19b using the NcoI and XhoI by/
g
for 30 min, followed by gel filtration sites. The gene was expressed in Rosetta™ u
e
s
chromatography on a Superdex 200 column (1 x grown at 37°C overnight in Overnight Express t o
n
60 cm) in 100 mM Tris, pH 7.4, containing 150 medium, both singularly and co-expressed with Ja
n
mM NaCl. the malate synthase gene. A cell extract was ua
ry
prepared as described previously and was then 3
, 2
Glyoxylate reductase incubated at 70°C for 10 min. 0
1
9
A putative glyoxylate reductase gene (SSO3187)
was identified in the genome of S. solfataricus Protein concentration and xylose concentration
using a BLAST search with the glyoxylate Protein concentrations were determined by the
reductase sequence from Pyrococcus horikoshii method of Bradford (13) using a calibration
(11). The gene was cloned into pET3a using the curve constructed with bovine serum albumin.
NcoI and XhoI sites. Expression of the Xylose concentrations in growth supernatants
recombinant enzyme was carried out in E. coli were determined according to (14) using a
Rosetta cells grown at 37°C overnight in calibration curve constructed with D-xylose.
Overnight Express medium. Cells were
harvested by centrifugation at 5,000 g for 10 min 13C-labelling experiments
and resuspended at 0.2 g/ml in 50 mM Tris-HCl, S. acidocaldarius was grown aerobically at
pH 8.0. After 30 min at room temperature, the 72°C, with shaking at 150 rpm, in 100 ml
cells were lysed by three 30-second bursts of Erlenmeyer flasks containing 20 ml medium,
sonication and cell debris was removed by pH 2.5, containing 25 mM D-[1-13C]-xylose or
centrifugation at 12,500 g for 30 min. The D-[2-13C]-xylose. Cell aliquots were harvested
6
during mid-exponential growth by centrifuging 2 Fig 2. This scheme predicts that Sulfolobus
ml of culture broth at 8,000 g for 10 min at 10°C. will possess a C5-specific dehydratase, the
The dried biomass pellet was hydrolysed in 1.5 enzymes glycolaldehyde oxidoreductase or
ml of 6 M HCl for 24 h at 110°C in sealed 2-ml glycolaldehyde dehydrogenase, glycolate
Eppendorf tubes and desiccated overnight in a dehydrogenase (glyoxylate reductase) and
heating block at 85°C under a constant air- malate synthase, and that the aldolase will
stream. The hydrolysate was dissolved in 50 µl catalyse the cleavage of both KD-xylonate and
of 99.8% (v/v) dimethyl formamide and KD-arabinonate. The malate formed would
transferred to a new Eppendorf cup within a few then enter the citric acid cycle for complete
seconds. For derivatization, 30 µl N-methyl-N- oxidation to CO and reducing equivalents.
2
(tert-butyldimethylsilyl)-trifluoroacetamide were Also shown in Fig. 2 is a possible parallel
added, which readily silylates hydroxyl groups, pathway whereby KD-xylonate and KD-
thiols, primary amines, amides and carboxyl arabinonate are converted to 2-oxoglutarate.
groups (15), and the mixture was shaken at 550 This pathway is identical to that proposed by
rpm and 85°C for 60 min. 1 µl of the derivatized Brouns et al. (8) for the catabolism of D-
sample was then injected into a 6890N Network arabinose in S. solfataricus, the 2-keto-3-
GC system, combined with a 5975 Inert XL deoxy-D-arabinonate of that pathway being
Mass Selective Detector (Agilent Technologies, identical to 2-keto-3-deoxy-D-xylonate
Böblingen, Germany) and analysed as described produced in the above scheme.
in (16,17). The GC temperature profile was As detailed below, we present several D
o
w
160°C for 1 min, increased to 320°C at 20°C per lines of evidence for the pathways shown in Fig n
lo
min, and held at 320°C for 1 min. The injector 2. First, with respect to the newly-proposed a
d
e
t1e:m10p,e arantdu rteh ew falos ws erta tteo t2o3 10.°5C m, lt/hme insp; ltiht er actairor iteor gplaythcowlaayld efhory dteh,e tmheet abenozliysmme ofa cCti5v-istuiegsa rsh avviae d from
h
gas was helium in an HP-5MS column (30 m x been assayed and found in cell extracts of S. ttp
0.25 mm, 0.25 µm coated) (Agilent solfataricus and S. acidocaldarius grown on ://w
w
Technologies). either C6 or C5 sugars as a sole carbon source, w
Mass spectra of the derivatized amino and nearly all the respective S. solfataricus .jb
c
.o
acids alanine, aspartate, glutamate, proline, genes have been cloned and expressed and the rg
serine and threonine were corrected for the recombinant enzymes characterised. Secondly, by/
g
natural abundance of all stable isotopes and for with respect to the in vivo partitioning of u
e
s
unlabeled biomass from the inoculum. Glycine, metabolites between the two pathways, we t o
n
histidine, isoleucine, leucine, lysine, methionine, report 13C-metabolic labelling studies. Ja
n
phenylalanine, tyrosine and valine were not used ua
ry
in this study, and arginine, asparagine, cysteine, Enzyme assays of cell extracts 3
, 2
glutamine and tryptophan were not detectable. The first critical test for the proposed pathway 0
1
9
of C5-sugar catabolism in S. solfataricus is to
assay for the native enzyme activities in cell
RESULTS AND DISCUSSION extracts of this organism. S. solfataricus was
grown on either D-glucose or D-xylose as sole
A proposed catabolic pathway for D-xylose and carbon source, and cell extracts prepared as
L-arabinose described in Experimental Procedures. Enzyme
Our previous findings that S. solfataricus glucose activities were measured in these
dehydrogenase can catalyse the oxidation of both unfractionated extracts and are shown in Fig 3.
D-xylose and L-arabinose (6), and that the KDG- It should be noted that these activities are those
aldolase can catalyse the aldol cleavage of 2- measured in the forward direction as defined by
keto-3-deoxypentanoate (until now, assayed only the scheme shown in Fig 2, except for KDG-
in the condensation direction), led us to consider aldolase and glycolate dehydrogenase. In the
what might be the catabolic fate of these two C5 case of glycolate dehydrogenase (also known as
sugars. Based on known enzymatic activities in glyoxylate reductase), the equilibrium is far
other organisms, a proposed scheme is shown in towards glycolate and therefore the enzyme
7
was assayed using glyoxylate and NAD+ as carbon source, the glyoxylate cycle being
substrates. required for the biosynthesis of C4 compounds
Activities of xylose dehydrogenase and from the C2 nutrient (19). Significantly, in
KDG-aldolase did not differ between glucose- Sulfolobus growing on xylose, malate synthase
and xylose-grown cells. This is not unexpected, is produced in the absence of an isocitrate
in that glucose dehydrogenase (encoded by gene lyase, consistent with the metabolic scheme
SSO3003) has comparable activity with D- proposed (Fig 2), although the latter enzyme
glucose, D-galactose, D-xylose and L-arabinose was induced in cells grown on glucose. As
(6). Furthermore, activities of glucose pointed out by Uhrigshardt et al (12),
dehydrogenase and xylose dehydrogenase were significant levels of acetate have been found in
found not to be additive in cell extracts of either S. acidocaldarius growing on glucose, and this
S. solfataricus or S. acidocaldarius (data not may explain the induction of both glyoxylate
shown), which provides in vivo evidence that cycle enzymes in the presence of this C6
indeed both sugars are oxidized by the same substrate.
dehydrogenase, i.e glucose/xylose All the enzyme activities for the
dehydrogenase. The KDG-aldolase (SSO3197) pathway are also found in cell extracts of S.
is also active with the 2-keto-3-deoxy-derivatives acidocaldarius grown on xylose (Table 1), and
of these four sugars (see below). Glycolaldehyde the same potential catabolic routes of C5-sugars
oxidation to glycolate was found with DCPIP or are therefore likely for this species.
NADP+ as electron acceptors. The activity with D
o
w
DCPIP is likely to be the glyceraldehyde Characterisation of xylonate-arabinonate n
lo
oxidoreductase reported by Kardinahl et al (18) dehydratase a
d
e
iwna sS . sahcoiwdonc atlod arbiues , awn hialsctt ivthitayt wofi th thNe AD2,P5+- Gdeihvyednr otgheen asper oamndis KcuDoGus- alndaotluarsee foorf bogtlhu cCo6se- d from
h
dioxopentanoate dehydrogenase identified by and C5-metabolites, it was important to ttp
Brouns et al (8) as part of the D-arabinose determine whether there was also a single ://w
w
degradation pathway (see ‘Recombinant enzyme dehydratase or two separate enzymes catalysing w
characterisation’). The oxidoreductase activity the dehydration of gluconate/galactonate and .jb
c
.o
in both S. solfataricus and S. acidocaldarius is 7 xylonate/arabinonate. When cell extracts of S. rg
to 10-fold higher than that of the dehydrogenase, solfataricus were subjected to either gel by/
g
and so would be the more significant enzyme filtration or ion-exchange chromatography, the u
e
s
candidate for the oxidation of glycolaldehyde in xylonate dehydratase was clearly separated t o
n
the proposed pathway. from the gluconate dehydratase (Fig. 4). The Ja
n
It was not surprising to find that the former enzyme was also active with L- ua
ry
activities of glyoxylate reductase and malate arabinonate, and the latter with D-galactonate 3
, 2
synthase are higher in cells grown on xylose than (7), suggesting that the two enzymes are 0
1
9
on glucose, as these enzymes are proposed to be promiscuous for C5 and C6 sugar acids,
involved in C5 sugar catabolism, but not C6. respectively. As shown in Fig. 4, the two
On the other hand, the levels of glycolaldehyde enzymes are produced in approximately equal
oxidoreductase would not be expected to change catalytic amounts.
between growth on glucose and xylose if this Using the semi-purified xylonate
enzyme activity is indeed from the dehydratase after ion-exchange
glyceraldehyde oxidoreductase that functions in chromatography, the dependence of measured
C6 sugar metabolism. However, unexpectedly, velocity on substrate concentration showed that
the levels of xylonate dehydratase were similar in the enzyme exhibited what appeared to be
glucose- and xylose-grown S. solfataricus, even pronounced substrate inhibition with both D-
though this enzyme has no detectable activity xylonate and L-arabinonate. However, the data
with gluconate or galactonate. do not give a good fit to the classical substrate
Malate synthase and isocitrate lyase are inhibition equation, indicating that other factors
usually coincidentally produced in aerobic may be influencing the dependence of enzyme
micro-organisms growing on acetate as sole velocity on substrate concentration.
8
Interestingly, the purified D-arabinonate laboratory with respect to its aldol-cleavage
dehydratase identified in S. solfataricus is also activity with KD-gluconate and KD-galactonate
subject to what appears to be substrate inhibition (6), and with its ability to catalyse the aldol-
(8). condensation of pyruvate and glyceraldehyde.
Gel filtration chromatography showed KD-xylonate and KD-arabinonate are not
that the gluconate dehydratase appeared to be commercially available, and therefore the
active as a monomer of M ≈ 48,000. On the kinetics of only the aldol-condensation activity
r
other hand, xylonate dehydratase activity eluted of this enzyme with pyruvate and the C2
with a M ≈ 165,000, indicating that the active compound glycolaldehyde have been
r
enzyme is a homo-tetramer. investigated (Table 1). The V of 31 µmol
max
min-1 mg protein-1 is 2-fold greater than that
Recombinant enzyme characterisation with glyceraldehyde. NMR analysis of the
To confirm the activities and substrate product generated confirmed it to be a 2-keto-
specificities of the enzymes in the proposed 3-deoxy-pentanoate; however, it clearly could
pathway for C5 sugar catabolism (Fig. 2), and to not reveal the relative amounts of the two
determine their kinetic parameters, the genes possible enantiomers.
encoding the enzymes have been cloned from S. To demonstrate the aldol cleavage
solfataricus and expressed in E. coli. In each activity of KDG-aldolase with both KD-D-
case, heat-treatment of the cell extract gave xylonate and KD-L-arabinonate, coupled
recombinant enzyme that was of >90% purity, assays were carried out. D-xylonate and L- D
o
w
although in some cases a chromatography step arabinonate were incubated separately with the n
lo
was then carried out to achieve a homogeneous semi-purified xylonate dehydratase described a
d
e
apnredp aarrea tcioonm. p aTrehde wdaittha tahroes er efcoourndde dw iitnh Tenabzylem 1e aKbDov-aer,a abnindo tnhaet ep roductiown aosf KD-xylomneaates uarnedd d from
h
assays from cell extracts. spectrophotometrically by the reaction with ttp
thiobarbituric acid (10). Incubation of the ://w
w
Xylose dehydrogenase products from the two dehydratase reactions w
The recombinant glucose dehydrogenase from S. with pure recombinant KDG-aldolase verified .jb
c
.o
solfataricus has been previously reported from that the latter enzyme catalysed the aldol rg
our group, and shown to have comparable cleavage of both enantiomers to pyruvate and by/
g
activity with D-glucose, D-galactose, D-xylose, glycolaldehyde. The data are therefore u
e
s
and L-arabinose (6). In the context of the consistent with the one aldolase enzyme being t o
n
pathway for C5 sugar catabolism, the kinetic involved in the catabolism of both the C6 and Ja
n
parameters with D-xylose and L-arabinose were C5 sugars (Fig. 2). ua
ry
determined (Table 1). 3
, 2
Glycolaldehyde oxidoreductase and 0
1
9
Xylonate dehydratase glycolaldehyde dehydrogenase
As described in experimental procedures, a Glycolaldehyde DCPIP oxidoreductase most
putative second dehydratase gene (SS02665) was likely reflects a side activity of the
identified in the genome sequence, with its glyceraldehyde DCPIP oxidoreductase reported
translated protein product having 62% amino by Kardinahl et al (18). Since the enzyme is
acid sequence identity to the gluconate extremely oxygen sensitive and found to
dehydratase. The gene was cloned but attempts contain a molybdenum cofactor, no attempts
to generate a soluble and active recombinant were made to clone and express heterologously
enzyme using a variety of E. coli, Baculovirus the genes (SS02636, SS02637 and SS02639)
and fungal expression systems have been encoding the 3 enzyme components.
unsuccessful. Brouns et al (8) reported that S.
solfataricus gene SSO3117 encodes an
KDG-aldolase. aldehyde dehydrogenase that, when
Recombinant KDG-aldolase from S. solfataricus recombinantly expressed, possessed catalytic
has been previously characterised in our activity with 2,5-dioxopentanoate,
9
glyceraldehyde and glycolaldehyde. We have acetate production under these growth
confirmed this observation, found that the conditions.
enzyme is specific for NADP+, and report the At this point, the data presented from
kinetic parameters for the substrates pentanedial assays in cell extracts and with the purified
(as an analogue of 2,5-dioxopentanoate) and recombinant enzymes are consistent with the
glycolaldehyde (Table 1). metabolism of D-xylose and L-arabinose via
the two possible pathways outlined in Fig 2. It
Glycolate dehydrogenase (Glyoxylate reductase) was therefore important to investigate their
The gene annotated as glyoxylate reductase operation in vivo, and the extent to which
(SS03187) in the S. solfataricus genome was metabolites are partitioned between them. This
cloned and expressed in E.coli to give a soluble, was performed using metabolic labelling
active recombinant protein, that was purified by studies with 13C-sugars.
anion exchange chromatography and gel
filtration (Fig. 5). The enzyme was found to be Metabolic labelling studies
NAD-dependent, and the kinetic parameters were Fig 2 shows the two proposed pathways for the
determined (Table 1). catabolism of D-xylose and L-arabinose. It is
important to note that Brouns et al (8) studied
Malate synthase and isocitrate lyase the catabolism of D-arabinose in S. solfataricus
Purified malate synthase and semi-purified and found that the resulting 2-keto-3-deoxy-D-
isocitrate lyase have been characterised arabinonate is converted to 2-oxoglutarate D
o
w
previously in S. acidocaldarius, and evidence for through the action of KD-arabinonate n
lo
a functional glyoxylate cycle in this organism has dehydratase and 2,5-dioxopentanoate a
d
e
bsoelefna tparreicsuens tegde n(e1s2 )a.n nIno tathteed c uasrr emnat lpaatep esry, ntthhea sSe. dareahbyidnroongaetne asise . id eHnotiwcaelv etro, 22--kkeettoo--33--ddeeooxxyy--DD-- d from
h
(SS01334) and isocitrate lyase (SS01333) were xylonate, and therefore can also undergo aldol- ttp
cloned and expressed in E. coli, and the proteins cleavage to pyruvate and glycolaldehyde as we ://w
w
purified as described in Experimental have proposed for the catabolism of D-xylose w
Procedures. The kinetic parameters (Table 1) and L-arabinose. Of the two Sulfolobus species .jb
c
.o
were similar to those reported for the considered here, the ability of S. acidocaldarius rg
corresponding S. acidocaldarius enzymes, as to grow on minimal media in the absence of by/
g
were the dependence of isocitrate lyase activity yeast extract (whereas S. solfataricus cannot) u
e
s
on the presence of Mg2+ and the oligomeric makes it ideal for metabolic labeling studies; t o
n
nature of the two enzymes as judged by SDS- that is, S. acidocaldarius will grow on 13C- Ja
n
PAGE and gel filtration (malate synthase: homo- labelled sugars as a sole carbon source, thereby ua
ry
dimer of M ≈ 189,000; isocitrate lyase: homo- permitting the flow of C5 sugars via these two 3
r , 2
tetramer of M ≈ 190,000). possible pathways to be determined. 0
r 1
9
In both S. acidocaldarius and S. To determine the relative in vivo
solfataricus, the malate synthase and isocitrate activities of xylose degradation pathways in S.
lyase genes are found 484 and 461 base pairs acidocaldarius, we performed labelling
apart, respectively, and TATA boxes (20) are experiments using [1-13C]- and [2-13C]-D-
found upstream of both genes, indicating that xylose. Label in the proteinogenic amino acids
they do not exist in an operon. Furthermore, no was determined by GC-MS, which gives direct
Shine-Dalgarno sequence could be identified information about the labelling patterns of the
upstream of either gene; such sequences are not precursors such as 3P-glycerate, pyruvate,
found upstream of isolated genes in S. oxaloacetate or 2-oxoglutarate. Absence of the
solfataricus (21), with most transcripts pentose phosphate pathway was confirmed by
completely lacking 5’untranslated regions (22). the absence of labeling in serine (Table 2), as
As stated above, isocitrate lyase activity is found catabolism of labelled xylose via xylulose-5P
in cells grown on glucose (12), which may be and 3P-glycerate would lead to label in this
linked with the finding in S. acidocaldarius of amino acid.
10
The first of the two alternative pathways labelling information within the citric acid
postulated for xylose catabolism (via KDG- cycle intermediates, we conducted tracer
aldolase) would lead to incorporation of C1 to experiments using [2-13C]-xylose. Glyoxylate
C3 of xylose into pyruvate, the direct precursor condenses with acetyl-CoA to form malate,
of alanine. The second pathway via 2,5- which now carries the label at position 2. If
dioxopentanoate leaves the carbon backbone malate is formed from glutamate labelled at
unchanged. Our data from tracer experiments position 2, the label ends up in position 1 or 4
using [1-13C]-xylose show that the first pathway of malate and leads to alanine labeled at
contributes to about half of the pyruvate position 1, as observed to be about 10% (Table
formation, yielding 50% unlabeled alanine and 2). The threonine and aspartate are labelled at
50% alanine labeled at position one (Table 2). both positions 1 and 4; however, within the
We also observed labelled glutamate and proline, resolution of this analysis, we can only estimate
but this cannot be explained by anaplerotic the possible label at position 2, which would
reactions from pyruvate to oxaloacetate and then give a 15-30% range for the formation of
to glutamate because the amino acids threonine malate from glyoxylate (Table 2).
and aspartate, derived from oxaloacetate, were
completely unlabeled (Table 2). The activity of
pyruvate ferredoxin oxidoreductase can be CONCLUDING REMARKS
excluded because the labelled C1 of pyruvate The in vivo labelling studies with S.
would be lost as CO . Therefore, we conclude acidocaldarius indicate that D-xylose (and D
2 o
w
that the second pathway degrading xylose via probably, therefore, L- arabinose) is degraded n
lo
2,5-dioxopentanoate directly to glutamate takes simultaneously via two routes, each at about a
d
e
KplDacGe -asilmdoullatsaen eaonuds layc cwoiuthn tsth feo rf irastt lpeaastht w5a0y% v oiaf 5p0yr%uv. a teO nane d rogulytec oilnalvdoelhvyeds ea, lwdohli cchl eiasv faugret hetor d from
h
the glutamate. Due to the loss of label in both converted to glyoxylate, glycolate and finally ttp
pyruvate and 2-oxoglutarate ferredoxin malate. In the second route, D-xylose is ://w
w
oxidoreductase reactions, we cannot exclude converted to 2-oxoglutarate, the KD-xylonate w
reformation of 2-oxoglutarate through the citric being dehydrated to 2-oxoglutarate- .jb
c
.o
acid cycle or the so-called pyruvate shunt, and semialdehyde, which is subsequently oxidized rg
therefore the 50% labels are a lower limit. to 2-oxoglutarate. The enzymes of the first by/
g
The glyoxylate formed in the degradation route, which constitutes a novel pathway of u
e
s
of xylose to pyruvate via the KDG-aldolase route pentose degradation, have been described and t o
n
is proposed to end up in malate. The C1-label of characterised from S. solfataricus and S. Ja
n
pyruvate is lost as CO2 in the formation of acidocaldarius in the current paper. uary
acetyl-CoA required to condense with It has also been proposed that the 3
, 2
glyoxylate, and therefore the malate formed Entner-Doudoroff pathway in S. solfataricus is 0
1
9
through this pathway is unlabelled, and may branched (3), whereby KD-gluconate is
account for the unlabelled alanine fraction phosphorylated to 2-keto-3-deoxy-6-
obtained by gluconeogenic reactions from phosphogluconate via a KDG-kinase; KDG-
malate/oxaloacetate to pyruvate (Table 2). In aldolase then catalyses the cleavage to pyruvate
addition, unlabelled malate can also originate and glyceraldehyde-3P, generating a pathway
from glutamate since the C1-label of glutamate is that operates in parallel with the non-
lost as CO in the 2-oxoglutarate oxidoreductase phosphorylative route. Through kinetic and
2
reaction. structural studies of KDG-kinase and KDG-
In addition to the main conclusion from aldolase, we have shown that this part-
the [1-13C]-xylose tracer studies, that both phosphorylative pathway in S. solfataricus is
pathways account for about half of the formation also promiscuous for the metabolism of both
of pyruvate or 2-oxoglutarate respectively, we glucose and galactose (23, 24); however, we
tried to differentiate between formation of currently have no evidence for a similar semi-
unlabelled malate from glyoxylate or from phosphorylative pathway being involved in the
glutamate. To obtain additional positional catabolism of the C5 sugars.