Chemistry & Biology
Functional Plasticity and Allosteric Regulation
of a-Ketoglutarate Decarboxylase
in Central Mycobacterial Metabolism
Tristan Wagner,1Marco Bellinzoni,1Annemarie Wehenkel,1,3Helen M. O’Hare,2,* and Pedro M. Alzari1,*
1Institut Pasteur, Unite ´ de Biochimie Structurale (CNRS URA 2185), 25 rue du Dr. Roux, 75724 Paris, France
2Department of Infection, Immunity and Inflammation, University of Leicester, University Road, LE1 9HN Leicester, UK
3Present address: Max-Planck-Institute for Molecular Physiology, 11 Otto-Hahn Strasse, 44227 Dortmund, Germany
*Correspondence: firstname.lastname@example.org (H.M.O.), email@example.com (P.M.A.)
The a-ketoglutarate dehydrogenase (KDH) complex
is a major regulatory point of aerobic energy metab-
olism. Mycobacterium tuberculosis was reported to
lack KDH activity, and the putative KDH E1o com-
ponent, a-ketoglutarate decarboxylase (KGD), was
instead assigned as a decarboxylase or carboligase.
Here, we show that this protein does in fact sustain
KDH activity, as well as the additional two reactions,
and these multifunctional properties are shared by
the Escherichia coli homolog, SucA. We also show
that the mycobacterial enzyme is finely regulated
by an additional acyltransferase-like domain and
by the action of acetyl-CoA, a powerful allosteric
activator able to enhance the concerted protein
motions observed during catalysis. Our results un-
cover the functional plasticity of a crucial node in
bacterial metabolism, which may be important for
M. tuberculosis during host infection.
The tricarboxylic acid (TCA) cycle catalyzes the total oxidation of
acetyl units derived from other catabolic pathways and is the
major energy-generating pathway in aerobic organisms. Within
the cycle the a-ketoglutarate dehydrogenase (KDH) complex
at the branch point between energy production and nitrogen
metabolism (Bunik and Fernie, 2009). In addition to energy
metabolism, the TCA cycle is a major source of biosynthetic
precursors, which may be obtained by noncyclic flux, with
oxidative and reducing branches leading respectively to a-keto-
glutarate (KG) and succinate (Guest, 1995). Noncyclic flux is
characterized by the absence (or repression) of the KG dehydro-
genase (KDH) complex. Some microaerophilic or anaerobic
bacteria directly lack KDH genes, such as Helicobacter pylori
are linked by KG oxidase (KGO) activity (Pitson et al., 1999),
whereas others, like Escherichia coli, operate a complete TCA
cycle aerobically but can repress KDH under anaerobic condi-
tions (Keevil et al., 1979).
Mycobacterium tuberculosis (Mtb), the etiological agent of
tuberculosis, can survive for long periods in a hypoxic environ-
ment (Wayne and Hayes, 1996) but is generally considered
a strictly aerobic bacillus and has functional genes coding for
all enzymes of the standard TCA cycle (Cole et al., 1998; Mur-
thy et al., 1962). Thus, although the presence of a functional
glyoxylate shunt (McKinney et al., 2000) and an anaerobic-like
KG:ferredoxin oxidoreductase (KGO) (Baughn et al., 2009)
could bypass the requirement for KDH activity, it is, neverthe-
less, intriguing that mycobacteria were reported to lack a func-
tional KDH complex (Tian et al., 2005b), given the enormous
energetic benefits of aerobic respiration and the oxidative
environment in many sites of infection, such as the lung and
macrophages. Instead, the putative E1o component of the
KDH complex was characterized as a thiamine diphosphate
(ThDP)-dependent KG decarboxylase (KGD) to produce suc-
cinic semialdehyde (SSA) (Tian et al., 2005b) and more recently
as an efficient carboligase on glyoxylate (de Carvalho et al.,
Here, we report structural and biochemical studies of
Mycobacterium smegmatis KGD (MsKGD) and the Mtb homo-
log, Rv1248c, demonstrating that KGD is a multifunctional
enzyme capable of sustaining succinyl-transferring KDH activity,
in addition to the reported KGD and carboligase activities.
Furthermore, this multifunctionality is not a unique feature of
the mycobacterial enzyme because the E. coli E1o homolog
SucA (EcSucA) was also found to efficiently catalyze all three
reactions. However, mycobacterial KGD has low basal activities,
which might account at least in part for previous failures in
detecting KDH activity in mycobacteria, and we show here that
the enzyme requires specific activation by acetyl-CoA, the
main substrate of the TCA cycle, to achieve specific activities
comparable to those of EcSucA. The active site architecture
of MsKGD is shown to undergo a considerable structural rear-
rangement during the catalytic cycle. Acetyl-CoA binds to an
allosteric site 40 A˚away from the catalytic center and stimulates
MsKGD activity by promoting local conformational changes
that facilitate formation of the enamine-ThDP reaction interme-
diate. Taken together, our results uncover the functional versa-
tility of KGD in bacterial metabolism and highlight the presence
of three alternative, tightly regulated, pathways connecting the
differential utilization may provide an advantage for pathogen
growth within its human host.
Chemistry & Biology 18, 1011–1020, August 26, 2011 ª2011 Elsevier Ltd All rights reserved 1011
RESULTS AND DISCUSSION
Overall Structure of Mycobacterial KGD
Extensive screening for crystallization conditions of full-length
MsKGD resulted in low-resolution diffracting crystals, which
were not exploitable for structural studies. Treatment of the
enzyme with proteases led to the rapid excision of the first
115 amino acid residues from the N terminus (see Figure S1A
available online), suggesting that this region is unstructured
in the native protein and might hinder crystallization. Conse-
quently, the structure of a truncated form of the enzyme lacking
this N-terminal region (MsKGDD115) was determined at 2.74 A˚
resolution (Table 1). The protein is a homodimer, in which
each subunit is composed of several distinct domains (Fig-
ure 1A). The N-terminal domain (residues 116–360), which
was originally identified in the Corynebacterium glutamicum
homolog OdhA (Usuda et al., 1996), is only found in corynebac-
terineae enzymes but is missing in other bacterial E1o homo-
logs, such as EcSucA. This domain folds into a two-layered
open-faced sandwich motif structurally similar to the subunit
fold of trimeric acyltransferases (Knapp et al., 2000), and was
shown to encode the succinyltransferase (E2o) activity in
C. glutamicum OdhA (Hoffelder et al., 2010). The rest of the
subunit displays a similar quaternary organization to that
described for EcSucA (Figure S1B), the only other E1o compo-
nent, to our knowledge, for which a crystal structure is currently
available (Frank et al., 2007). In MsKGD the acyltransferase
domain is connected to (and interacts extensively with) a small
helical domain (MsKGD residues 361–445) involved in protein
dimerization, followed by three consecutive a/b domains (resi-
dues 446–814, 831–1091, and 1092–1227) displaying the
characteristic fold of ThDP-dependent enzymes (Figure 1A).
First described for transketolase (Lindqvist et al., 1992), this
common fold was subsequently found in other homodimeric
dehydrogenases (Muller et al., 1993) as well as in heterotetra-
meric enzymes such as eukaryotic pyruvate dehydrogenases
(PDHs), in which the basic subunit lacks an interdomain linker
connecting the first and second a/b domains and is encoded
by two different polypeptides. In the structure of the MsKGDD115
homodimer, the equivalent connecting linker (residues 810–830)
is disordered (Figure 1A) and presumably adopts a flexible
Table 1. Crystallographic Data Collection and Refinement Statistics
ESRF ID14-4 SOLEIL Proxima 1SOLEIL Proxima 1 SLS X06DASOLEIL Proxima 1
Space group P1 P21212 P1 P1P1
a, b, c (A˚)79.54, 83.24, 158.61 79.98, 151.99, 242.7279.84, 82.29, 163.4880.50, 83.32, 159.9281.06, 81.95, 161.88
a, b, g (?)
99.5, 99.1, 101.390.0, 90.0, 90.099.2, 99.0, 100.699.6, 99.0, 100.7 99.3, 97.1, 100.6
49.5–1.96 (2.07–1.96)82–2.74 (2.89–2.74)41.9–2.25 (2.37–2.25)78.1–2.2 (2.31–2.2)41.4–2.40 (2.53–2.40)
0.089 (0.445)0.074 (0.416)0.101 (0.481)0.067 (0.395)0.105 (0.562)
I/s(I)8.7 (2.2)10.3 (2.4)6.9 (2.0)8.7 (2.2)7.4 (1.9)
Completeness (%)95.2 (91.5)99.3 (97.5)95.7 (94.6)92.6 (88.5)95.8 (95.6)
Redundancy2.9 (2.4) 3.5 (3.2)2.0 (1.9)2.0 (2.0)2.5 (2.4)
Resolution (A˚)1.96 2.742.252.20 2.40
Number of reflections263,738 79,835179,896187,791 149,475
Rwork/Rfree(%)18.8/21.1 18.5/22.419.2/22.3 18.8/21.419.7/22.0
Number of atoms
Water 1,221 119 730742514
Bond lengths (A˚) 0.0100.0100.010 0.0100.008
Bond angles (?) 0.98 1.091.01 1.020.97
Highest resolution shell is shown in parenthesis.
Chemistry & Biology
Structure and Regulation of Mycobacterial KGD
1012 Chemistry & Biology 18, 1011–1020, August 26, 2011 ª2011 Elsevier Ltd All rights reserved
Pelicic, V., Jackson, M., Reyrat, J.M., Jacobs, W.R., Jr., Gicquel, B., and
Guilhot, C. (1997). Efficient allelic exchange and transposon mutagenesis in
Mycobacterium tuberculosis. Proc. Natl. Acad. Sci. USA 94, 10955–10960.
Pinsach, J., de Mas, C., and Lo ´pez-Santin, J. (2006). A simple feedback
control of Escherichia coli growth for recombinant aldolase production in
fed-batch mode. Biochem. Eng. J. 29, 235–242.
Pitson, S.M., Mendz, G.L., Srinivasan, S., and Hazell, S.L. (1999). The tricar-
boxylic acid cycle of Helicobacter pylori. Eur. J. Biochem. 260, 258–267.
Schlossberg, M.A., Bloom, R.J., Richert, D.A., and Westerfeld, W.W. (1970).
Carboligase activity of alpha-ketoglutarate dehydrogenase. Biochemistry 9,
Tian, J., Bryk, R., Itoh, M., Suematsu, M., and Nathan, C. (2005a). Variant
tricarboxylic acid cycle in Mycobacterium tuberculosis: identification of
alpha-ketoglutarate decarboxylase. Proc. Natl. Acad. Sci. USA 102, 10670–
Tian, J., Bryk, R., Shi, S., Erdjument-Bromage, H., Tempst, P., and Nathan, C.
(2005b). Mycobacterium tuberculosis appears to lack alpha-ketoglutarate
dehydrogenase and encodes pyruvate dehydrogenase in widely separated
genes. Mol. Microbiol. 57, 859–868.
Tomb, J.F., White, O., Kerlavage, A.R., Clayton, R.A., Sutton, G.G.,
Fleischmann, R.D., Ketchum, K.A., Klenk, H.P., Gill, S., Dougherty, B.A.,
et al. (1997). The complete genome sequence of the gastric pathogen
Helicobacter pylori. Nature 388, 539–547.
Trapani, S., and Navaza, J. (2008). AMoRe: classical and modern. Acta
Crystallogr. D Biol. Crystallogr. 64, 11–16.
Usuda, Y., Tujimoto, N., Abe, C., Asakura, Y., Kimura, E., Kawahara, Y.,
Kurahashi,O.,and Matsui, H.
Corynebacterium glutamicum (‘Brevibacterium lactofermentum’ AJ12036)
odhA gene encoding a novel type of 2-oxoglutarate dehydrogenase.
Microbiology 142, 3347–3354.
(1996).Molecular cloning ofthe
van den Berg, S., Lo ¨fdahl, P.A., Ha ¨rd, T., and Berglund, H. (2006). Improved
solubility of TEV protease by directed evolution. J. Biotechnol. 121, 291–298.
Waskiewicz, D.E., and Hammes, G.G. (1984). Elementary steps in the reaction
mechanism of the alpha-ketoglutarate dehydrogenase multienzyme complex
from Escherichia coli: kinetics of succinylation and desuccinylation.
Biochemistry 23, 3136–3143.
Wayne, L.G., and Hayes, L.G. (1996). An in vitro model for sequential study of
shiftdown of Mycobacterium tuberculosis through two stages of nonreplicat-
ing persistence. Infect. Immun. 64, 2062–2069.
Yamasaki, H., and Moriyama, T. (1970). Alpha-ketoglutarate: glyoxylate
carboligase activity in Escherichia coli. Biochem. Biophys. Res. Commun.
Yamasaki, H., and Moriyama, T. (1971). Purification, general properties and
two other catalytic activities of -ketoglutarate:glyoxylate carboligase of
Mycobacterium Phlei. Biochim. Biophys. Acta 242, 637–647.
Chemistry & Biology
Structure and Regulation of Mycobacterial KGD
1020 Chemistry & Biology 18, 1011–1020, August 26, 2011 ª2011 Elsevier Ltd All rights reserved