Petrobactin biosynthesis: AsbB catalyzes condensation of spermidine with N8-citryl-spermidine and its N1-(3,4-dihydroxybenzoyl) derivative

Article abstract of DOI:10.1039/b809353a

The AsbB enzyme, which is involved in the biosynthesis of the virulence-conferring siderophore petrobactin in Bacillus anthracis, is shown to catalyze efficient ATP-dependent condensation of spermidine, but not N ¹-(3,4-dihydroxbenzoyl)-spermidine, with N⁸-citryl- spermidine or N¹-(3,4-dihydroxbenzoyl)-N⁸-citryl- spermidine, suggesting that N¹-(3,4-dihydroxbenzoyl)-spermidine is very unlikely to be a significant intermediate in petrobactin biosynthesis, contrary to previous suggestions. The Royal Society of Chemistry.

Full text of DOI:10.1039/b809353a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Petrobactin biosynthesis: AsbB catalyzes condensation of spermidine with
N8-citryl-spermidine and its N1-(3,4-dihydroxybenzoyl) derivativew
Daniel Oves-Costales,a Nadia Kadi,a Mark J. Fogg,b Lijiang Song,a
Keith S. Wilsonb and Gregory L. Challis*a
Received (in Cambridge, UK) 3rd June 2008, Accepted 23rd July 2008
First published as an Advance Article on the web 31st July 2008
DOI: 10.1039/b809353a
The AsbB enzyme, which is involved in the biosynthesis of the
virulence-conferring siderophore petrobactin in Bacillus anthra-
cis, is shown to catalyze efficient ATP-dependent condensation
of spermidine, but not N1-(3,4-dihydroxbenzoyl)-spermidine,
with N8-citryl-spermidine or N1-(3,4-dihydroxbenzoyl)-N8-
citryl-spermidine, suggesting that N1-(3,4-dihydroxbenzoyl)-
spermidine is very unlikely to be a significant intermediate in
petrobactin biosynthesis, contrary to previous suggestions.
same type of condensation reaction, whereas others are pre-
dicted to catalyse only a single condensation. We recently
reported biochemical investigations of AsbA, a type A NIS
synthetase, and DesD, an ‘‘iterative’’ type C NIS synthetase,
that catalyze key reactions in the biosynthesis of the Bacillus
anthracis siderophore petrobactin 1 (Scheme 1) and the
Streptomyces siderophores desferrioxamines B, G1 and E,
respectively.5,6 AsbA and DesD were the first members of the
NIS synthetase superfamily to be biochemically-characterized.
Because many pathogens depend on siderophore-mediated
iron uptake from their hosts for growth and survival, the
molecular machinery used by bacteria for the assembly,
excretion and uptake of siderophores offers new targets for
the development of novel antibiotics.7 For example, effective
inhibitors of mycobactin biosynthesis in Mycobacterium
tuberculosis that inhibit growth under iron-limiting conditions
have recently been reported.8 Among the microorganisms that
could be used for bioterrorism, B. anthracis is arguably one of
the most dangerous.9 The development of new antibiotics
active against B. anthracis is therefore of much importance.
Two gene clusters (asbABCDEF and bacACBEF) have been
reported to direct siderophore biosynthesis in B. anthracis.10
bacACBEF encodes an NRPS-dependent pathway for the
biosynthesis of bacillibactin, a siderophore present in all
studied Bacillus species. On the other hand, asbABCDEF
Iron is an essential trace element for almost all living organ-
isms. Despite being the fourth most abundant element on
Earth’s crust, its predominant presence as insoluble polymeric
oxo-hydroxy complexes makes iron bioavailability very low in
the environment. Therefore, in order to meet their iron
requirements, many microorganisms have evolved the ability
to produce and excrete small molecules with a high affinity for
iron called siderophores.1 Once a siderophore has scavenged
iron from its environment, the ferric–siderophore complex is
actively taken up by microbial cells.1
Many siderophores are biosynthesized by members of the
extensively-studied non-ribosomal peptide synthetase (NRPS)
multienzyme superfamily, e.g. mycobactin, enterobactin,
vibriobactin and yersiniabactin.2 On the other hand, a rapidly
growing group of structurally-diverse siderophores is bio-
synthesized by an emerging family of new synthetases with
no sequence or structural similarity to NRPSs via the so-called
NRPS-independent siderophore (NIS) pathway.3 NIS synthe-
tases are classified as type A, type B or type C according to
sequence similarity criteria.3 A predictive model hypothesises
that type A enzymes catalyse the condensation of one of the
prochiral carboxyl groups in citric acid with amines or alco-
hols; type B enzymes catalyse the condensation of the
g-carboxyl group of a-ketoglutarate with amines; and type C
enzymes catalyse either the condensation of a monoamide
derivative of citric acid with an amine or alcohol, or the
oligomerisation and macrocyclisation of o-amino-carboxylic
acids that contain a hydroxamate group.3,4 Thus some type C
enzymes are predicted to catalyse multiple iterations of the
a Department of Chemistry, University of Warwick, Coventry,
UK CV4 7AL. E-mail: g.l.challis@warwick.ac.uk;
Fax: +44 (0)24 7652 4112; Tel: +44 (0)24 7657 4024
b Department of Chemistry, University of York, Heslington, York,
UK YO10 5YW
w Electronic supplementary information (ESI) available: Experimental
procedures, SDS-PAGE analysis of His6-AsbB expression and puri-
fication, spectroscopic and kinetic data. See DOI: 10.1039/b809353a
Scheme 1 Possibile reactions catalysed by AsbB in petrobactin 1
biosynthesis.
ꢀc
This journal is The Royal Society of Chemistry 2008
4034 | Chem. Commun., 2008, 4034–4036

View Article Online
Fig. 2 Key correlations observed in the COSY (bold lines) and
HMBC (arrows) spectra of 6 and 7.
Fig. 1 Extracted ion chromatograms from LC-MS analyses, in
positive ion mode, of incubations of AsbB, ATP, and Mg2+ with 4
and 2 (top trace, m/z 447.3), 4 and 3 (second trace, m/z 583.3), 5 and 2
(third trace, m/z 583.3), and 5 and 3 (bottom trace, m/z 719.3).
incubated equal quantities of each separately with equal
quantities of 2 or synthetic 3,5 AsbB, a nucleotide triphosphate
(NTP) and Mg2+ at 37 1C for 90 min. LC-ESI-MS analysis of
the mixtures showed new compounds with m/z 447.3 for
incubation of 4 with 2, m/z 583.3 for incubations of 4 with 3
and 5 with 2, and m/z 719.3 for incubation of 5 with 3 (Fig. 1).
These new compounds were absent in incubations using heat-
inactivated AsbB, lacking Mg2+, or using NTPs other than
ATP or GTP, suggesting that all four possible roles we
envisioned for AsbB are, a priori, possible. Although LC-
ESI-MS does not allow quantitation of the relative amounts of
1, 6 and 7 produced in the reactions, qualitatively, 6 was
formed in the largest amounts in these reactions, followed by 7
and then 1. Only trace amounts of 1 could be detected in the
reaction of 5 with 3, suggesting that this is the least efficient of
the reactions catalysed by AsbB. The ability of AsbB to use
either ATP or GTP is not surprising, given the structural
similarity of the purine bases in these NTPs. Large scale
incubations of 2 with either 4 or 5, AsbB, Mg2+ and ATP
were carried out. 6 and 7 were isolated from these reactions by
semi-preparative HPLC and their identities were confirmed by
HRMS, MS-MS and 1-D and 2-D NMR experiments (Fig. 2).
To gain more insight into both the mechanism and the rate
of the different AsbB-catalyzed reactions, we used continuous
coupled assays for AMP and ADP formation.14 These assays
showed time-dependent formation of AMP but not ADP,
consistent with the reaction of the enzyme-bound citrate
derivatives 4 and 5 with ATP to form an acyl-adenylate
intermediate that undergoes nucleophilic attack by 2 or 3 to
form a new amide bond. To establish which of the four
encodes a unique hybrid NRPS-NIS system for the assembly
of 1, which is present only in pathogenic Bacillus species.11
While the production of bacillibactin is not necessary for B.
anthracis growth, 1 has been shown to be required for growth
in iron-depleted media and virulence in a mouse model,10
probably due to its ability to evade the mammalian immune
system.12 These data suggest that inhibitors of the biosynthesis
and uptake of 1 may be useful anti-anthrax agents.
Recent biochemical studies of the biosynthesis of 1 have
shown that AsbC, a NRPS-like adenylating enzyme, catalyzes
the ATP-dependent acylation of the phosphopantetheine thiol
of the carrier protein AsbD with 3,4-dihydroxybenzoic acid.13
They also showed that AsbE can catalyze the transfer of the
3,4-dihydroxybenzoyl group from AsbD to N1 and N8 of
spermidine 2 to afford N1-(3,4-dihydroxybenzoyl)-spermidine
3 and N8-(3,4-dihydroxybenzoyl)-spermidine, respectively,13
demonstrating that AsbE has relaxed substrate specificity.
On the other hand, the NIS synthetase AsbA has been shown
to catalyze the ATP-dependent condensation of citric acid
with N8 of spermidine 2 to afford N8-citryl-spermidine 4.5
AsbA did not catalyze condensation of N1-(3,4-dihydroxy-
benzoyl)-spermidine 3 with citrate, suggesting that the transfer
of the 3,4-dihydroxybenzoyl group from AsbD to N1 of
spermidine occurs after AsbA-catalyzed formation of 4. Here
we report the first biochemical study of AsbB, the other NIS
synthetase involved in the biosynthesis of 1, and show that it
preferentially catalyzes ATP-dependent condensation of N8-
citryl-spermidine 4 or N1-(3,4-dihydroxybenzoyl)-N8-citryl-
spermidine 5 with spermidine 2.
Table 1 Relative rates of AMP formation catalysed by AsbB with
different substrate combinations
Sequence analysis of AsbB indicates that it is a type C NIS
synthetase.3 We therefore envisioned different possible roles
for AsbB in petrobactin biosynthesis; it could catalyze con-
densation of either 4 or 5 with 2 or 3 (Scheme 1). To
investigate these hypotheses, we cloned asbB into pET-YS-
BLIC3C and overexpressed it in E. coli BL21star(DE3). The
resulting soluble N-terminal His6-tagged derivative of AsbB
was purified from cell-free extracts using Ni-NTA and gel
filtration chromatography. We synthesised 4 and 5 and
Entry
Substrates
Relative ratea
1
2
3
4
5
6
a
2 + 4
2 + 5
4
3 + 4
3 + 5
5
57.10 ꢁ 0.15
38.63 ꢁ 5.76
1.12 ꢁ 0.02
1.20 ꢁ 0.15
1.00 ꢁ 0.12
1.03 ꢁ 0.13
Rate values are the mean of three independent measurements and
errors are quoted as ꢁ one standard deviation.
ꢀc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4034–4036 | 4035

View Article Online
Our data suggest a revised pathway for petrobactin biosynth-
esis in which 4, 5, 6 and 7 are all viable intermediates and AsbE
functions as an acyltransferase that could transfer the 3,4-
dihydroxybenzoyl group from the phosphopantetheine thiol
of AsbD to the primary amino groups in 4, 6 and 7, rather
than 2 as previously suggested13 (Scheme 2).
AsbB is only the second type C NIS synthetase to be
biochemically characterized. Our results confirm the bioinfor-
matics-derived prediction that some type C enzymes can
catalyze NTP-dependent condensation of a derivative of citric
acid with amines.3 DesD, another type C enzyme, has been
shown experimentally to catalyse the ATP-dependent iterative
condensation of o-amino-carboxylic acids to form oligomeric
products, some of which undergo subsequent ATP-dependent
macrocylisation.6 This raises an intriguing question for future
research: how does AsbB avoid catalysing oligomerization
(and subsequent macrocyclisation) of 4? Clearly, the type C
subfamily of NIS synthetases should be divided into two
groups: The AsbB-like group, which catalyses a single con-
densation reaction and can therefore be thought of as ‘‘mod-
ular’’, and the DesD-like group, which catalyses multiple
condensation reactions and can therefore be referred to as
‘‘iterative’’.
In conclusion, we have carried out the first biochemical
study of AsbB, a key yet promiscuous enzyme in petrobactin
biosynthesis. Our results provide the basic biochemical knowl-
edge required to screen for inhibitors of AsbB, which may
ultimately lead to the development of new antibiotics active
against B. anthracis that abrogate petrobactin biosynthesis.
This work was supported by a grant from the BBSRC
(grant ref. BB/FO13760/1).
Notes and references
1 U. E. Schaible and S. H. E. Kaufmann, Nat. Rev. Microbiol., 2004,
2, 946–953.
2 J. H. Crosa and C. T. Walsh, Microbiol. Mol. Biol. Rev., 2002, 66,
223–249.
3 G. L. Challis, ChemBioChem, 2005, 6, 601–611.
4 F. Barona-Gomez, U. Wong, A. Giannakopulos, P. J. Derrick and
G. L. Challis, J. Am. Chem. Soc., 2004, 126, 16282–16283.
5 D. Oves-Costales, N. Kadi, M. J. Fogg, L. Song, K. S. Wilson and
G. L. Challis, J. Am. Chem. Soc., 2007, 129, 8416–8417.
6 N. Kadi, D. Oves-Costales, F. Barona-Gomez and G. L. Challis,
Nat. Chem. Biol., 2007, 3, 652–656.
Scheme 2 Revised pathway for petrobactin biosynthesis.
7 (a) M. Marcus and M. A. Marahiel, Microbiol. Mol. Biol. Rev.,
2007, 71, 413–451; (b) R. R. Monfeli and C. Beeson, Infect. Disord.
Drug Targets, 2007, 7, 213–220.
8 (a) J. A. Ferreras, J.-S. Ryu, F. Di Lello, D. S. Tan and L. E. N.
Quadri, Nat. Chem. Biol., 2005, 1, 29–32; (b) R. V. Somu, H.
Boshoff, C. Qiao, E. M. Bennett, C. E. Barry III and C. C. Aldrich,
J. Med. Chem., 2006, 49, 31–34.
9 T. V. Inglesby, T. O0Toole, D. A. Henderson, J. G. Bartlett, M. S.
Ascher, E. Eitzen, A. M. Friedlander, J. Gerberding, J. Hauer, J.
Hughes, J. McDade, M. T. Osterholm, G. Parker, T. M. Perl, P. K.
Russell and K. Tonat, J. Am. Med. Assoc., 2002, 287, 2236–2252.
10 S. Cendrowski, W. MacArthur and P. Hanna, Mol. Microbiol.,
2004, 51, 407–417.
11 M. K. Wilson, R. J. Abergel, K. N. Raymond, J. E. L. Arceneaux and
B. R. Byers, Biochem. Biophys. Res. Commun., 2006, 348, 320–325.
12 R. J. Abergel, M. K. Wilson, J. E. L. Arceneaux, T. M. Hoette, R.
K. Strong, B. R. Byers and K. N. Raymond, Proc. Natl. Acad. Sci.
U. S. A., 2006, 103, 18499–18503.
reactions catalysed by AsbB is the fastest, we measured the
relative initial rates of AMP formation using saturating con-
centrations of ATP, 2 or 3, and 4 or 5. The rate of AMP
formation from condensation of 2 with 4 or 5 was high relative
to the background rate of AMP formation resulting from ATP
hydrolysis (Table 1, entries 1–3 and 6). On the other hand, the
rate of AMP formation from condensation of 3 with 4 or 5 was
not significantly higher than the background rate of AMP
formation (Table 1, entries 3–6). These data show that AsbB
has a strong preference for 2 over 3 as a substrate, but a much
less marked preference for 4 over 5.
The inefficiency with which AsbB catalyzes condensation of
3 with 4 or 5, together with our previously reported observa-
tion that AsbA catalyzes condensation of 2 but not 3 with
citric acid to form 4,5 strongly suggests that 3 is unlikely to be
13 B. F. Pfleger, J. Y. Lee, R. V. Somu, C. C. Aldrich, P. C. Hanna
and D. H. Sherman, Biochemistry, 2007, 46, 4147–4157.
14 M. X. Wu and K. A. W. Hill, Anal. Biochem., 1993, 211, 320–323.
a
significant intermediate in petrobactin biosynthesis.
ꢀc
This journal is The Royal Society of Chemistry 2008
4036 | Chem. Commun., 2008, 4034–4036