Characterization and application of the Fe(ii) and α-ketoglutarate dependent hydroxylase FrbJ

Article abstract of DOI:10.1039/c1cc13597j

The Fe(ii) and α-ketoglutarate-dependent hydroxylase FrbJ was previously demonstrated to utilize FR-900098 synthesizing a second phosphonate FR-33289. Here we assessed its ability to hydroxylate other possible substrates, generating a library of potential antimalarial compounds. Through a series of bioassays and in vitro experiments, we identified two new antimalarials. The Royal Society of Chemistry 2011.

Full text of DOI:10.1039/c1cc13597j

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
ChemComm
Cite this: Chem. Commun., 2011, 47, 10025–10027
www.rsc.org/chemcomm
COMMUNICATION
Characterization and application of the Fe(II) and a-ketoglutarate
dependent hydroxylase FrbJw
Matthew A. DeSieno,ac Wilfred A. van der Donkbc and Huimin Zhao*abcd
Received 16th June 2011, Accepted 22nd July 2011
DOI: 10.1039/c1cc13597j
The Fe(II) and a-ketoglutarate-dependent hydroxylase FrbJ was
previously demonstrated to utilize FR-900098 synthesizing a
second phosphonate FR-33289. Here we assessed its ability to
hydroxylate other possible substrates, generating a library of
potential antimalarial compounds. Through a series of bioassays
and in vitro experiments, we identified two new antimalarials.
postulated that FrbJ might be responsible for the N-hydroxy-
lation step; however, catalysis of this nature by this class
of enzyme would have been exceedingly rare. Fe(II) and
aKG-dependent enzymes can efficiently catalyze the hydro-
xylation of inactivated C–H bonds, so we investigated whether
FrbJ might be involved in the synthesis of another phosphonate
antibiotic FR-33289 ((R)-2-hydroxy-3-(N-acetyl-N-hydroxy-
amino)propylphosphonic acid).8,9 Production of this compound
was observed in both the native producer S. rubellomurinus
producer strain and a heterologous S. lividans 4G7 host. We
have recently shown that FrbJ is indeed capable of catalyzing
the hydroxylation of FR-900098 at the b-position (Fig. 1).7
This study focused on the expression, purification, and further
biochemical characterization of FrbJ and the determination
of its role in FR-900098 biosynthesis. We also demonstrated
its ability to catalyze the hydroxylation of additional substrates
as part of the creation of a new library of potential antimalarial
drugs, all of which were screened for inhibitory activity against
a malarial drug target.
The full-length frbJ gene was cloned from fosmid 4G7.7
E. coli BL21(DE3) cells were transformed with this construct
and IPTG-induced cultures were purified to homogeneity. As
previously described, the final steps in the pathway contain a
series of intermediates conjugated to the nucleotide cytidine-
50-monophosphate (CMP).7 If hydroxylation by FrbJ were to
occur before cleavage of the CMP by FrbI, CMP-50-FR-
900098 could serve as the true substrate for the enzyme. FrbJ
showed activity towards the natural substrate FR-900098 but
not the nucleotide containing substrate, indicating that nucleotide
cleavage must occur before hydroxylation in the biosynthetic
pathway (Table 1). We also checked for enzymatic activity
with another known phosphonate antibiotic, fosmidomycin
(FR-31564). This substrate contains an N-formyl group instead
of the N-acetyl group on FR-900098. FrbJ demonstrated activity
The gene frbJ was identified during the cloning and sequencing
of the biosynthetic gene cluster for the phosphonate antibiotic
FR-900098 from Streptomyces rubellomurinus.1 The enzyme is
homologous to Fe(II) and a-ketoglutarate (aKG)-dependent
hydroxylases, which catalyze substrate hydroxylation with
concomitant oxidative decomposition of aKG to succinate
and carbon dioxide.2 Enzymes within this family play a pivotal
role in a diverse set of reactions, including antibiotic synthesis
where they generate pathway precursors, intermediates, and
final products.3,4 For example, clavaminate synthase (CAS)
catalyzes three independent oxidative reactions in the bio-
synthesis of clavulanic acid.5 Interestingly, another enzyme
from this family has demonstrated epoxidase activity in the
semisynthetic pathway for the phosphonate antibiotic fosfo-
mycin. The enzyme encoded by epoA from Penicillium decumbens
converts the substrate cis-propenylphosphonic acid to the final
product fosfomycin.6 Although the natural function of this
enzyme in the native host is unknown, the epoxidase activity
signifies the broad range of functions of Fe(II) and aKG-
dependent enzymes.
Deletion studies of frbJ within the heterologous hosts
Streptomyces lividans and Escherichia coli indicated that the
enzyme was not required for FR-900098 biosynthesis, leading
to some uncertainty as to its function.1,7 It was originally
a Department of Chemical and Biomolecular Engineering, University
of Illinois at Urbana-Champaign, 600 S. Mathews Avenue, Urbana,
IL 61801, USA. E-mail: zhao5@illinois.edu;
Fax: +1-217-333-5052; Tel: +1-217-333-2631
b Departments of Chemistry and Biochemistry, University of Illinois at
Urbana-Champaign, 600 S. Mathews Avenue, Urbana, IL 61801,
USA
c Institute for Genomic Biology, University of Illinois at Urbana-
Champaign, 600 S. Mathews Avenue, Urbana, IL 61801, USA
d Department of Bioengineering, Center for Biophysics and
Computational Biology, University of Illinois at Urbana-Champaign,
600 S. Mathews Avenue, Urbana, IL 61801, USA
Fig. 1 FrbJ catalyzes the hydroxylation of FR-900098 to FR-33289
with concomitant oxidative decomposition of aKG to succinate and
carbon dioxide.
w Electronic supplementary information (ESI) available: Materials
and experimental procedures. See DOI: 10.1039/c1cc13597j
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 10025–10027 10025

View Article Online
Table 1 Kinetic parameters of FrbJ
kcat (minꢁ1
bioavailability of FR-900098 and fosmidomycin is moderate
with a resorption rate of approximately 30%.14 Recent efforts
have focused on masking the charged moiety of the phosphonate
group through esterification. Such chemically synthesized
prodrugs cross the gastrointestinal tract into the blood stream
with higher efficiency, where they are cleaved by broad speci-
ficity plasma esterases, thus releasing the free and active
phosphonate.15–17 We sought a method for the preparation
of similar esterified FR-900098 prodrugs through biosynthetic
means. The gene cluster for dehydrophos from Streptomyces
luridus contains an S-adenosyl-L-methionine (SAM)-dependent
O-methyltransferase encoded by dhpI.18 DhpI has very low
substrate specificity,19 and is able to O-methylate the phos-
phonate of a wide range of substrates, making it an excellent
candidate for further engineering.
)
KM (mM)
(meanꢀSD)
kcat/KM
(mMꢁ1 minꢁ1
)
Substrate
(meanꢀSD)
FR-900098
Fosmidomycin
CMP-50-FR-900098
147 ꢀ 31
289 ꢀ 48
509
81.9
NR
57.3 ꢀ 8.1
700 ꢀ 137
NR: no reaction detected; SD: standard deviation.
towards this structurally similar substrate, but displayed a
6-fold preference for the natural substrate. As a result of
fosmidomycin hydroxylation by FrbJ, a novel phosphonate was
able to be synthesized, (R)-2-hydroxy-3-(N-formyl-N-hydroxy-
amino)propylphosphonic acid, hereby called hydroxyfosmido-
mycin. To the best of our knowledge, this compound has never
been detected from a biological source or produced synthetically.
Heterologous production of FR-33289 in E. coli was
attempted by expressing all of the genes in the biosynthetic
cluster, excluding frbI (pETDuet-frbABCD, pRSFDuet-
frbHFG, pACYCDuet-frbEJ-dxrB). Although FR-900098
production was evident in the LC-MS analysis, no detectable
amounts of FR-33289 were detected. Cell-free lysate of E. coli
strains that only expressed FrbJ was incubated with FR-900098.
The lysate only demonstrated conversion to FR-33289 upon
addition of exogenous a-ketoglutarate, suggesting that FrbJ
may be regulated by the intracellular concentration of this
cofactor, thus preventing in vivo production in E. coli.
As shown in Table 1, FrbJ is capable of catalyzing the
hydroxylation of FR-900098 and fosmidomycin to produce
FR-33289 and hydroxyfosmidomycin, respectively. These four
phosphonates were used as the basis for the creation of a
prospective antimalarial library; each was methylated by DhpI
(Fig. 2). FR-900098 and fosmidomycin were commercially
available and the subsequent six compounds were synthesized
by in vitro reactions with purified FrbJ, DhpI, and Pfs. The
phosphonate library was then purified to remove any unreacted
substrate or cofactors. Although not identical to the di-ester
prodrugs prepared by chemical synthesis, these methylated
compounds may enhance the bioavailability of the final active
phosphonate drugs.
The biosynthesis of secondary metabolites can be effectively
regulated via the availability of precursors, cofactors, and
cellular energy originating from primary metabolism. Rate-
limiting enzymes control the flux of these critical metabolites
to ensure that secondary metabolism does not become favored
over primary metabolism. Khetan et al. described the function
of lysine-6-aminotransferase (LAT), which was determined to
be the rate-limiting enzyme in cephamycin biosynthesis from
Streptomyces clavuligerus.10 This enzyme, which utilizes aKG
as a cofactor in the conversion of lysine to a-aminoadipic acid,
is the first step in the biosynthetic pathway. The authors mea-
sured the KM for aKG to be 8.6 mM, substantially higher than
its intracellular concentration ranging from 0.3–1.3 mM.10
The KM value for LAT is also higher than that of other
characterized Fe(II) and aKG-dependent hydroxylases, which
is typically on the order of 10s of micromolar.11–13 In the same
way that LAT regulates the cephamycin biosynthetic pathway,
FrbJ may regulate the ratio of the two phosphonates FR-900098
and FR-33289 using a-ketoglutarate. The KM of aKG for the
oxidation of FR-900098 by FrbJ was determined to be 1.3 ꢀ
0.19 mM, comparable to the high KM of LAT and close to the
value of the intracellular level of aKG. One crucial difference
between the two pathways is that the entire cephamycin
biosynthesis can be regulated by the intracellular concentration
of aKG since LAT is the first enzyme in the pathway, making
the high KM of the enzyme ideal for switching the biosynthesis
of cephamycin entirely on or off. The lower KM of FrbJ is a
much weaker control valve, which can regulate the conversion
of FR-900098 to FR-33289.
The library was first screened by bioassays of an E. coli
phosphonate uptake strain.1 Bioassay plates of this FR-
900098-sensitive strain showed clear inhibition zones around
discs impregnated with FR-900098, FR-33289, fosmidomycin,
and hydroxyfosmidomycin; all of which were lost with a
resistant E. coli strain. However, there was no detectable
The dianionic charge of phosphonates at physiological pH
poses problems for their efficacy; in particular, they suffer from
poor bioavailability as a result of their high polarity. The oral
Fig. 2 Prospective antimalarial library created using FrbJ and DhpI
(with Pfs). Due to the low substrate specificity of DhpI, hydroxylation
by FrbJ was carried out before O-methylation.
c
10026 Chem. Commun., 2011, 47, 10025–10027
This journal is The Royal Society of Chemistry 2011

View Article Online
Table 2 Inhibition of purified Dxr from Plasmodium falciparum by
members of the phosphonate library
to more accurately assess these compounds, we will look to
expand the scope of the inhibition assays to include Plasmodium
species.
IC50 (nM)
(mean ꢀ SD)
KI (nM)
(mean ꢀ SD)
This work was supported by a grant from the National
Institute of General Medical Sciences GM077596 (H.Z. and
W.A.V). M.A.D. acknowledges support from the US National
Institutes of Health under Ruth L. Kirschstein National
Research Award 5 T32 GM070421 from the National Institute
of General Medical Sciences.
Substrate
FR-900098
FR-33289
Fosmidomycin
Hydroxyfosmidomycin
FR-900098-OMe
FR-33289-OMe
15 ꢀ 3
16 ꢀ 3
36 ꢀ 6
41 ꢀ 5
NI
NI
NI
NI
3.2 ꢀ 1.0
3.0 ꢀ 0.8
6.8 ꢀ 1.6
7.1 ꢀ 1.4
NI
NI
NI
NI
Fosmidomycin-OMe
Hydroxyfosmidomycin-OMe
Notes and references
1 A. C. Eliot, B. M. Griffin, P. M. Thomas, T. W. Johannes,
N. L. Kelleher, H. M. Zhao and W. W. Metcalf, Chem. Biol.,
2008, 15, 765–770.
2 R. P. Hausinger, Crit. Rev. Biochem. Mol. Biol., 2004, 39, 21–68.
3 E. G. Kovaleva and J. D. Lipscomb, Nat. Chem. Biol., 2008, 4,
186–193.
NI: no enzyme inhibition detected; SD: standard deviation.
growth inhibition for any of the corresponding methylated
compounds with the FR-900098-sensitive strain. To ensure the
lack of bioactivity on the plate was not specific for the Dxr
target, we utilized a second phosphonate antibiotic with a
completely different target. Fosfomycin, a broad-range anti-
bacterial used to treat urinary tract infections,20 and the
corresponding fosfomycin-OMe was prepared in a similar
method as described for the other phosphonates. Similar to
the prospective methylated antimalarials, there was no growth
inhibition of the sensitive E. coli strain by fosfomycin-OMe.
The methyl group is either not cleaved in these in vivo bio-
assays, preventing the release of the active phosphonate, or the
methyl ester hampers uptake.
4 C. J. Schofield and Z. Zhang, Curr. Opin. Struct. Biol., 1999, 9,
722–731.
5 M. D. Lloyd, K. D. Merritt, V. Lee, T. J. Sewell, B. Wha-Son,
J. E. Baldwin, C. J. Schofield, S. W. Elson, K. H. Baggaley and
N. H. Nicholson, Tetrahedron, 1999, 55, 10201–10220.
6 M. Watanabe, N. Sumida, S. Murakami, H. Anzai,
C. J. Thompson, Y. Tateno and T. Murakami, Appl. Environ.
Microbiol., 1999, 65, 1036–1044.
7 T. W. Johannes, M. A. DeSieno, B. M. Griffin, P. M. Thomas,
N. L. Kelleher, W. W. Metcalf and H. M. Zhao, Chem. Biol., 2010,
17, 57–64.
8 M. Okuhara, Y. Kuroda, T. Goto, M. Okamoto, H. Terano,
M. Kohsaka, H. Aoki and H. Imanaka, J. Antibiot. (Tokyo),
1980, 33, 24–28.
The library was also assessed using purified Plasmodium
falciparum Dxr (pf Dxr). The recombinant enzyme was codon-
optimized for heterologous expression in E. coli and was
designed to lack the N-terminal signal peptide.21,22 Enzymatic
activity was monitored by NADPH oxidation at 340 nm, and
subsequent inhibition values for the IC50 and KI were obtained
for each of the prospective antimalarial compounds. Similar to
the in vivo bioassays, only the free, unmethylated phosphonate
compounds demonstrated in vitro inhibition of pf Dxr. There
was no apparent decrease in the enzymatic activity upon
addition of the methylated compounds to any of the reaction
mixtures. The approximately two-fold difference between the
inhibition of FR-900098 and fosmidomycin is similar to
published data.21 Also, the inhibition constants for FR-900098
and fosmidomycin with their hydroxylated counterparts,
FR-33289 and hydroxyfosmidomycin, respectively, were
determined to be nearly identical to each other (Table 2). Even
though the natural pf Dxr substrate also contains a secondary
hydroxyl group similar to FR-33289 and hydroxyfosmidomycin,
albeit at different positions on the carbon backbone, there was
no apparent increase in inhibition compared to FR-900098
and fosmidomycin.
9 Y. Kuroda, M. Okuhara, T. Goto, M. Okamoto, H. Terano,
M. Kohsaka, H. Aoki and H. Imanaka, J. Antibiot. (Tokyo),
1980, 33, 29–35.
10 A. Khetan, L. H. Malmberg, Y. S. Kyung, D. H. Sherman and
W. S. Hu, Biotechnol. Prog., 1999, 15, 1020–1027.
11 A. K. White and W. W. Metcalf, J. Biol. Chem., 2002, 277,
38262–38271.
12 E. Eichhorn, J. R. van der Ploeg, M. A. Kertesz and T. Leisinger,
J. Biol. Chem., 1997, 272, 23031–23036.
13 F. M. Vaz, R. Ofman, K. Westinga, J. W. Back and R. J. Wanders,
J. Biol. Chem., 2001, 276, 33512–33517.
14 T. Tsuchiya, K. Ishibashi, M. Terakawa, M. Nishiyama, N. Itoh and
H. Noguchi, Eur. J. Drug Metab. Pharmacokinet., 1982, 7, 59–64.
15 A. Reichenberg, J. Wiesner, C. Weidemeyer, E. Dreiseidler,
S. Sanderbrand, B. Altincicek, E. Beck, M. Schlitzer and
H. Jomaa, Bioorg. Med. Chem. Lett., 2001, 11, 833–835.
16 R. Ortmann, J. Wiesner, A. Reichenberg, D. Henschker, E. Beck,
H. Jomaa and M. Schlitzer, Bioorg. Med. Chem. Lett., 2003, 13,
2163–2166.
17 J. Wiesner, R. Ortmann, H. Jomaa and M. Schlitzer, Arch. Pharm.,
2007, 340, 667–669.
18 B. T. Circello, A. C. Eliot, J. H. Lee, W. A. van der Donk and
W. W. Metcalf, Chem. Biol., 2010, 17, 402–411.
19 J. H. Lee, B. Bae, M. Kuemin, B. T. Circello, W. W. Metcalf,
S. K. Nair and W. A. van der Donk, Proc. Natl. Acad. Sci.
U. S. A., 2010, 107, 17557–17562.
20 D. Hendlin, E. O. Stapley, M. Jackson, H. Wallick, A. K. Miller,
F. J. Wolf, T. W. Miller, L. Chaiet, F. M. Kahan, E. L. Foltz,
H. B. Woodruff, J. M. Mata, S. Hernandez and S. Mochales,
Science, 1969, 166, 122–123.
21 H. Jomaa, J. Wiesner, S. Sanderbrand, B. Altincicek,
C. Weidemeyer, M. Hintz, I. Turbachova, M. Eberl, J. Zeidler,
H. K. Lichtenthaler, D. Soldati and E. Beck, Science, 1999, 285,
1573–1576.
22 R. E. Cobb, M.S. Thesis, University of Illinois at Urbana-
Champaign, 2011.
The combinatorial biosynthesis with FrbJ was successful in
producing two new antimalarial target compounds, FR-33289
and hydroxyfosmidomycin. The E. coli bioassays showed
some improvement for these hydroxylated compounds as the
zones of inhibition were slightly larger. Although the inhibition
of the Dxr enzyme itself does not change, the bioavailability of
these hydroxylated compounds appears to increase. In order
c
This journal is The Royal Society of Chemistry 2011
Chem. Commun., 2011, 47, 10025–10027 10027