Enzymatic logic of anthrax stealth siderophore biosynthesis: AsbA catalyzes ATP-dependent condensation of citric acid and spermidine

Article abstract of DOI:10.1021/ja072391o

Petrobactin is an iron-chelating siderophore originally isolated from Marinobacter hydrocarbonoclasticus that has been shown to play an important role in growth under iron-deficient conditions and virulence of the deadly bioterrorism agent Bacillus anthracis. It has recently been shown not to bind to siderocalin, leading it to be designated as a "stealth siderophore" that can avoid the mammalian immune system. A unique combination of nonribosomal peptide synthetase (NRPS) and NRPS-independent siderophore (NIS) synthetase enzymes is known to be required for petrobactin biosynthesis in B. anthracis. Here it is shown that AsbA from B. anthracis, the first type A NIS synthetase to be biochemically characterized, catalyzes ATP-dependent regioselective condensation of citric acid with N⁸ of spermidine, but not with N¹-(3,4-dihydroxybenzoyl)-spermidine. These results rule out a recently proposed pathway for petrobactin biosynthesis involving AsbA-catalyzed condensation of N¹-(3,4-dihydroxybenzoyl)-spermidine with citric acid and show that acylation of N¹ of spermidine with the 3,4-dihydroxybenzoyl group must occur after acylation of N⁸ of spermidine with citrate. They also provide the fundamental knowledge needed to establish a high throughput screen for inhibitors of AsbA that may provide the basis for development of new antibiotics for the treatment of deadly anthrax infections. Copyright

Full text of DOI:10.1021/ja072391o

Published on Web 06/19/2007
Enzymatic Logic of Anthrax Stealth Siderophore Biosynthesis: AsbA
Catalyzes ATP-Dependent Condensation of Citric Acid and Spermidine
†
†
‡
†
‡
Daniel Oves-Costales, Nadia Kadi, Mark J. Fogg, Lijiang Song, Keith S. Wilson, and
,
†
Gregory L. Challis*
Department of Chemistry, UniVersity of Warwick, CoVentry CV4 7AL, U.K., and Structural Biology Laboratory,
Department of Chemistry, UniVersity of York, Heslington, York YO10 5YW, U.K.
Received April 5, 2007; E-mail: g.l.challis@warwick.ac.uk
Bacillus anthracis is a Gram-positive, spore-forming bacterium
that has attracted exceptional interest because of its potential use
as a biowarfare agent. Inhalation or ingestion of B. anthracis spores
leads to systemic anthrax, which is characterized by an initial
intracellular phase, where the spores germinate and proliferate,
followed by massive bacteraemia and sepsis, which is rapidly
1
lethal. The high levels of mortality if not treated and the possibility
of making genetically modified antibiotic-resistant strains, makes
2
B. anthracis a major threat in a bioterrorism environment.
Therefore, the development and synthesis of new antibiotics active
against B. anthracis is of much current interest.
Figure 1. Structure of petrobactin 1 and organization of the asbABCDEF
gene cluster in B. anthracis that directs its biosynthesis. Genes encoding
NIS synthetases are in black, and genes encoding NRPS-like enzymes are
in gray.
As for many other pathogens, iron is an essential nutrient for B.
anthracis. Iron is tightly bound and highly regulated in eukaryotes,
and iron bioavailability for invading pathogens is therefore very
low, making its assimilation a considerable challenge for the patho-
gen and an early defensive barrier for the host. The production,
excretion, and reabsorption of small molecules with high affinity
for iron, called siderophores, is a common strategy used by patho-
Scheme 1. Revised Pathway for Petrobactin 1 Biosynthesis
Showing the Reaction Catalyzed by AsbA
3
genic bacteria to meet their iron requirements. Many siderophores
are biosynthesized by members of the well-known nonribosomal
peptide synthetase (NRPS) multienzyme family, including entero-
4
bactin, vibriobactin, yersiniabactin, and mycobactin. Biochemical
studies of these siderophore-assembling NRPS systems has led
recently to the synthesis of an effective inhibitor of mycobactin
biosynthesis that inhibits Mycobacterium tuberculosis growth. On
mammalian immune system.12 These data suggest that inhibitors
5
of the biosynthesis and uptake of 1 may be useful antianthrax agents.
Here we report the first biochemical study of AsbA, one of two
NIS synthetases involved in the assembly of 1. Our results prompt
us to propose a revised pathway for the biosynthesis of 1.
Sequence analysis of AsbA has shown that it is a “type A” NIS
the other hand, many other siderophores are biosynthesized by a
nonribosomal peptide synthetase-independent siderophore (NIS)
pathway, involving a novel family of putative synthetases about
which much less is currently known.6 Very recently, the first
biochemical study of an enzyme belonging to this family DesD,
which catalyzes the key oligomerization-macrocyclization reactions
in desferrioxamine biosynthesis, has been reported.7
Two gene clusters (asbABCDEF and bacACEBF) responsible
for siderophore biosynthesis have been identified in B. anthracis.8
While bacACEBF encodes an NRPS-dependent pathway for ba-
cillibactin biosynthesis present in all studied Bacillus species,
asbABCDEF encodes a unique combination of NRPS and NIS
synthetase enzymes responsible for the biosynthesis of petrobactin
6
synthetase, suggesting it catalyzes condensation of citric acid 2
1
with spermidine 3 (Scheme 1) or its N -3,4-dihydroxybenzoyl
derivative. To test this hypothesis, we cloned asbA into pET-
YSBLIC3C and overexpressed it in E. coli BL21star(DE3). The
resulting soluble N-terminal His -tagged derivative of AsbA was
6
purified from cell free extracts using Ni-NTA chromatography (see
Supporting Information). Because it was unclear whether 3 or its
1
N -3,4-dihydroxybenzoyl derivative was a substrate for AsbA, we
initially investigated the trapping of a presumed citryl phosphate/
8
-11
1
(Figure 1),
which is only produced by pathogenic Bacillus
13
nucleotidylate intermediate with hydroxylamine. Incubation of
9
species. Recent genetic and biochemical experiments have inves-
tigated the role of the asbABCDEF genes and the AsbC, AsbD,
and AsbE proteins in the assembly of 1, leading to a proposed
biosynthetic pathway.1 Although bacillibactin production is not
necessary for B. anthracis growth, production of 1 is required for
growth in iron-depleted media and virulence in a mouse model.8
Very recently, it has been shown that, in contrast to bacillibactin,
2+
AsbA with 2, hydroxylamine, ATP, and Mg generated a citryl
hydroxamate that could be detected by time-dependent increase in
13
absorption at 540 nm on addition of FeCl
3
.
No activity was
0,11
detected with enzyme inactivated by boiling, with nucleotide
2+
triphosphates (NTPs) other than ATP, or in the absence of Mg
(see Supporting Information).
Having established that AsbA catalyzes the ATP-dependent
1
is not bound by siderocalin, thus allowing it to evade the
6
activation of a carboxyl group of 2, as previously proposed, we
next examined whether the enzyme could catalyze ATP-dependent
condensation of 2 with 3 or its 3,4-dihydroxybenzoyl derivative.
†
University of Warwick.
University of York.
‡
8416
9
J. AM. CHEM. SOC. 2007, 129, 8416-8417
10.1021/ja072391o CCC: $37.00 © 2007 American Chemical Society

C O M M U N I C A T I O N S
1
formation of 4 (Scheme 1). It is likely that N -(3,4-dihydroxyben-
zoyl)-spermidine, which is observed to accumulate in a ∆asbA
mutant of B. anthracis,10 is a shunt metabolite resulting from the
11
biochemically demonstrated relaxed specificity of AsbE. The order
of the steps in assembly of 1 following formation of 4, that is,
1
acylation of N of 4 with 3,4-dihydoxybenzoyl-AbsD and condensa-
1
tion of 4 with 3 or N -(3,4-dihydroxybenzoyl)-spermidine, remains
to be established. Further experiments to investigate the substrate
tolerance and specificity, respectively, of AsbE and AsbB will be
required to do this. It is notable that the first step in the assembly
of 1 defined here differs from the early steps of vibriobactin
1
biosynthesis, where N of norspermidine is acylated with a 2,3-
4
7
dihydroxybenzoyl group prior to acylation of N and N with other
acyl groups.15
In conclusion, we have carried out the first biochemical studies
of AsbA, a key enzyme in petrobactin 1 biosynthesis that catalyzes
the ATP-mediated condensation of citric acid 2 with spermidine 3.
This is the first enzyme from the type A subfamily of the NIS
synthetases to be characterized and the results confirm the bioin-
formatics-derived prediction that type A enzymes catalyze NTP-
dependent condensation of citric acid with a variety of amines and
Figure 2. Extracted ion chromatograms at m/z ) 320.1 from LC-MS
2
+
analysis of incubations of 2, 3, ATP, and Mg with AsbA (top trace) and
heat-inactivated AsbA (bottom trace).
LC-MS analysis of a mixture of 2, 3, ATP, and Mg2+ incubated
with AsbA at 37 °C for 90 min identified a new compound with
m/z ) 320.1 that was absent in control incubations with inactivated
AsbA, or lacking Mg2 . The identity of this compound isolated by
semipreparative HPLC from a large-scale incubation was confirmed
+
6
alcohols. This study provides the fundamental knowledge required
to design a high-throughput screen for AsbA inhibitors that may
lead to the development of new antianthrax antibiotics targeting
petrobactin assembly.
8
as N -citryl-spermidine 4 (Figure 2) by HRMS (calculated for
+
C
13
H N O
26 3 6
, 320.1816; found, 320.1817), as well as MS/MS and
1
H NMR comparisons with a chemically synthesized authentic
standard. In contrast, no conversion of various concentrations of
chemically synthesized N -(3,4-dihydroxybenzoyl)-spermidine to
N -(3,4-dihydroxy-benzoyl)-N -citryl-spermidine could be detected
upon incubation with citric acid, ATP, Mg2 , and AsbA at 37 °C
for 90 min.
Acknowledgment. This work was supported by grants from
the U.K. BBSRC (Grant Ref. BBS/B/14450) and the EU. We thank
Mr. R. Jenkins for help with high-pressure hydrogenation and Prof.
T. D. H. Bugg for helpful discussions.
1
1
8
+
Supporting Information Available: Complete ref 2, experimental
procedures, SDS-PAGE analysis of His -AsbA expression and purifica-
6
tion, the results of the enzyme activity assays, and spectroscopic data.
This material is available free of charge via the Internet at http://
pubs.acs.org.
To determine what ATP is converted to in the AsbA-catalyzed
condensation of 2 with 3, we used continuous coupled assays for
AMP and ADP formation.14 These assays showed time-dependent
formation of the former but not the latter. No time-dependent AMP
1
formation was observed in incubations with N -(3,4-dihydroxy-
benzoyl)-spermidine in place of spermidine or R-ketoglutaric acid
in place of 2, providing further evidence that AsbA has substrate
specificity for 2 and 3 (see Supporting Information). These data
are consistent with the reaction of enzyme bound 2 and ATP to
form an acyl-adenylate intermediate that undergoes nucleophilic
References
(1) Friedlander, A. M. Curr. Clin. Top. Inf. 2000, 20, 335-349.
(
(
2) Inglesby, T. V.; et al. J. Am. Med. Assoc. 2002, 287, 2236-2252.
3) Schaible, U. E.; Kaufmann, S. H. E. Nat. ReV. Microbiol. 2004, 2, 946-
953.
8
(4) Crosa, J. H.; Walsh, C. T. Microbiol. Mol. Biol. ReV. 2002, 66, 223-
attack at its activated carboxyl group by N of 3 to form 4 (Scheme
249.
1), but do not rule out other mechanistic possibilities.
(
5) (a) Ferreras, J. A.; Ryu, J.-S.; Di Lello, F.; Tan, D. S.; Quadri, L. E. N.
Nat. Chem. Biol. 2005, 1, 29-32. (b) Somu, R. V.; Boshoff, H.; Qiao,
C.; Bennett, E. M.; Barry, C. E., III; Aldrich, C. C. J. Med. Chem. 2006,
Sherman and co-workers have recently shown that AsbC
catalyzes ATP-dependent loading of 3,4-dihydroxybenzoic acid onto
the holo-carrier protein AsbD.11 They also showed that AsbE can
49, 31-34.
(6) Challis, G. L. ChemBioChem 2005, 6, 601-611.
(
(
(
7) Kadi, N.; Oves-Costales, D.; Barona-Gomez, F.; Challis, G. L. Nat. Chem.
catalyze transfer of the 3,4-dihydroxybenzoyl group from AsbD to
Biol., accepted for publication.
1
3
. However, N -(3,4-dihydroxybenzoyl)-spermidine, the intermedi-
8) Cendrowski, S.; MacArthur, W.; Hanna, P. Mol. Microbiol. 2004, 51,
407-417.
9) Wilson, M. K.; Abergel, R. J.; Raymond, K. N.; Arceneaux, J. E. L.;
Byers, B. R. Biochem. Biophys. Res. Commun. 2006, 348, 320-325.
ate that would be required for assembly of 1, is only a minor product
of this reaction. The major product appears to be the regioisomer
8
N -(3,4-dihydroxybenzoyl)-spermidine. These data together with
(
10) Lee, J. Y.; Janes, B. K.; Passalacqua, K. D.; Pfleger, B. F.; Bergman, N.
H.; Liu, H.; Håkansson, K.; Ravindranadh, V. S.; Courtney, C. A.;
Cendrowski, S.; Hanna, P. C.; Sherman, D. H. J. Bacteriol. 2007, 189,
data on compounds accumulated in mutants lacking individual genes
from the asbABCDEF gene cluster10 led Sherman and co-workers
to propose a pathway for the biosynthesis of 1 involving AsbA-
1698-1710.
(11) Pfleger, B. F.; Lee, J. Y.; Somu, R. V.; Aldrich, C. C.; Hanna, P. C.;
1
Sherman, D. H. Biochemistry 2007, 46, 4147-4157.
catalyzed condensation of N -(3,4-dihydroxybenzoyl)-spermidine
(
12) Abergel, R. J.; Wilson, M. K.; Arceneaux, J. E. L.; Hoette, T. M.; Strong,
R. K.; Byers, B. R.; Raymond, K. N. Proc. Nat. Acad. Sci. U.S.A. 2006,
103, 18499-18503.
(
formed by the action of AsbC, AsbD, and AsbE) with 2 to give
1
8
N -(3,4-dihydroxybenzoyl)-N -citryl-spermidine, followed by AsbB-
(13) Lipmann, F.; Tuttle, L. C. J. Biol. Chem. 1945, 159, 21-28.
14) Wu, M. X.; Hill, K. A. W. Anal. Biochem. 1993, 211, 320-3.
catalyzed condensation of this intermediate with a second molecule
(
1
of N -(3,4-dihydroxybenzoyl)-spermidine. Our data show that this
(15) Keating, T. A.; Marshall, C. G.; Walsh, C. T. Biochemistry 2000, 39,
15522-15530.
pathway cannot be correct; transfer of the 3,4-dihydroxybenzoyl
1
group from AbsD to N of 3 must occur after AsbA-catalyzed
JA072391O
J. AM. CHEM. SOC.
9
VOL. 129, NO. 27, 2007 8417