Evidence for radical-mediated catalysis by HppE: A study using cyclopropyl and methylenecyclopropyl substrate analogues

Article abstract of DOI:10.1021/ja3078126

(S)-2-Hydroxypropylphosphonic acid epoxidase (HppE) is an unusual mononuclear iron enzyme that catalyzes the oxidative epoxidation of (S)-2-hydroxypropylphosphonic acid ((S)-HPP) in the biosynthesis of the antibiotic fosfomycin. HppE also recognizes (R)-2-hydroxypropylphosphonic acid ((R)-HPP) as a substrate and converts it to 2-oxo-propylphosphonic acid. To probe the mechanisms of these HppE-catalyzed oxidations, cyclopropyl- and methylenecyclopropyl-containing compounds were synthesized and studied as radical clock substrate analogues. Enzymatic assays indicated that the (S)- and (R)-isomers of the cyclopropyl-containing analogues were efficiently converted to epoxide and ketone products by HppE, respectively. In contrast, the ultrafast methylenecyclopropyl-containing probe inactivated HppE, consistent with a rapid radical-triggered ring-opening process that leads to enzyme inactivation. Taken together, these findings provide, for the first time, experimental evidence for the involvement of a C2-centered radical intermediate with a lifetime on the order of nanoseconds in the HppE-catalyzed oxidation of (R)-HPP.

Full text of DOI:10.1021/ja3078126

Communication
pubs.acs.org/JACS
Evidence for Radical-Mediated Catalysis by HppE: A Study Using
Cyclopropyl and Methylenecyclopropyl Substrate Analogues
Hui Huang,†,‡ Wei-chen Chang,†,‡ Pei-Jing Pai,§ Anthony Romo,† Steven O. Mansoorabadi,†
David H. Russell,§ and Hung-wen Liu*,†
†Division of Medicinal Chemistry, College of Pharmacy, and Department of Chemistry and Biochemistry, University of Texas at
Austin, Austin, Texas 78712, United States
§Department of Chemistry, Texas A&M University, College Station, Texas 77843, United States
S
* Supporting Information
Scheme 1. Proposed Mechanism for the Oxidation of (S)-
and (R)-HPP by HppE
a
ABSTRACT: (S)-2-Hydroxypropylphosphonic acid epox-
idase (HppE) is an unusual mononuclear iron enzyme that
catalyzes the oxidative epoxidation of (S)-2-hydroxypro-
pylphosphonic acid ((S)-HPP) in the biosynthesis of the
antibiotic fosfomycin. HppE also recognizes (R)-2-
hydroxypropylphosphonic acid ((R)-HPP) as a substrate
and converts it to 2-oxo-propylphosphonic acid. To probe
the mechanisms of these HppE-catalyzed oxidations,
cyclopropyl- and methylenecyclopropyl-containing com-
pounds were synthesized and studied as radical clock
substrate analogues. Enzymatic assays indicated that the
(S)- and (R)-isomers of the cyclopropyl-containing
analogues were efficiently converted to epoxide and
ketone products by HppE, respectively. In contrast, the
ultrafast methylenecyclopropyl-containing probe inacti-
vated HppE, consistent with a rapid radical-triggered
ring-opening process that leads to enzyme inactivation.
Taken together, these findings provide, for the first time,
experimental evidence for the involvement of a C2-
centered radical intermediate with a lifetime on the order
of nanoseconds in the HppE-catalyzed oxidation of (R)-
HPP.
a
(A) Conversion of (S)-HPP (2) to fosfomycin (1). (B) Conversion
of (R)-HPP (3) to 2-oxopropylphosphonic acid (4).
HppE revealed the formation of an FeIII−OOH species in the
(partially) rate-limiting step of O2 activation.9
Interestingly, HppE can also recognize and efficiently oxidize
(R)-HPP (3), the enantiomer of the natural substrate, to 2-
oxopropylphosphonic acid (4) (Scheme 1B).10 The catalytic
versatility and substrate flexibility of HppE suggest that both
the regiochemistry of the hydrogen atom abstraction steps and
the nature of the resulting radical intermediates vary with the
stereochemical properties of the substrate. Crystallographic and
spectroscopic experiments showed that both enantiomers of
HPP (2 and 3) act as bidentate ligands to the mononuclear
iron, such that only a single hydrogen atom (1-HR for 2 and 2-
H for 3) is poised for abstraction by a reactive iron/oxygen
species (likely FeIII−OO·).11,12 These results led to the current
working hypothesis for the mechanisms of the HppE-catalyzed
reactions shown in Scheme 1.
Although non-heme iron enzymes are thought to initiate
substrate oxidation via hydrogen atom abstraction, leading to
the formation of substrate-derived radical intermediates,13
experimental evidence for the involvement of discrete radical
intermediates in these enzymatic systems is often lacking.14
Specifically for the HppE-catalyzed reactions, the intermediacy
of the proposed substrate radical species (5 and 6) has never
been verified experimentally. To address this issue and gain
support for the proposed reaction mechanisms, substrate
analogues bearing strategically incorporated cyclopropyl or
methylenecyclopropyl moieties were prepared and analyzed as
osfomycin ((1R, 2S)-1,2-epoxypropylphosphonic acid, 1)
Fis a widely used broad-spectrum antibiotic with activity
against both Gram-negative and Gram-positive bacterial
infections.1 Fosfomycin blocks bacterial cell wall biosythesis
by inhibiting UDP-N-acetyl-glucosamine-3-O-enolpyruvyltrans-
ferase (MurA) through the covalent modification of an active
site cysteine residue.2 A proposed biosynthetic pathway for
fosfomycin was first reported in 1995.3 One of the key steps in
this pathway is the formation of the epoxide ring, which is
catalyzed by (S)-2-hydroxypropylphosphonic acid epoxidase
(HppE), a mononuclear non-heme iron enzyme.4,5 Unlike most
epoxidases that catalyze direct oxygen atom insertion into an
alkene precursor,6 the oxirane moiety of fosfomycin is
generated by forming a C−O bond between C1 and the 2-
OH group of the substrate (S)-2-hydroxypropylphosphonic
acid (2, (S)-HPP) in a dehydrogenation reaction (Scheme
1A).3,7 HppE catalysis employs molecular oxygen as the oxidant
and consumes a stoichiometric amount of NADH as the
electron donor.5,8 18O kinetic isotope effects (KIEs) studies of
Received: August 6, 2012
Published: September 24, 2012
© 2012 American Chemical Society
16171
dx.doi.org/10.1021/ja3078126 | J. Am. Chem. Soc. 2012, 134, 16171−16174

Journal of the American Chemical Society
Communication
mechanistic probes of HppE catalysis. The radical-induced ring-
opening of cyclopropylcarbinyl or methylenecyclopropylcarbin-
yl radicals are well-known.15,16 Thus, the detection of ring-
opened turnover products or radical induced enzyme
inactivation would provide evidence for the occurrence of
radical intermediates during catalysis and their lifetimes. In this
paper, we report the syntheses of a set of cyclopropyl- and
methylenecyclopropyl-containing substrate analogues ((S)-7,
(R)-7 and 16), the evaluation of these compounds as radical
clock probes for the HppE reaction, and the mechanistic
implications of the incubation results.
The syntheses of (S)-7 and (R)-7 are depicted in Scheme 2.
Treatment of the bromo compound 817 with P(OEt)3 under
Scheme 2. Synthetic Schemes for the Preparation of (S)- and
(R)-7
Figure 1. 1H NMR end-point assay of HppE with (R)- and (S)-7. (A)
Initial time point recorded 3 min after incubation of substrate with
enzyme; (B) 36 min time point showing >90% conversion of (S)-7 to
13; (C) 36 min time point showing >90% conversion of (R)-7 to 15
(see Figure S4−1, S4−2 for details).
Scheme 3. Conversion of (S)-7 to 13 and (R)-7 to 15 by
HppE
Michaelis-Arbuzov reaction conditions followed by reduction of
the resulting phosphonate 9 with NaBH4 gave a racemic
mixture of alcohol 10. Resolution of this pair of alcohols was
attempted using various lipases. This method attempts to utilize
the asymmetric environment in the lipase active site to
stereoselectively acetylate one of the enantiomers, rendering
the resulting acetate separable from the unreacted alcohol.
Unfortunately, all lipase systems tested showed poor stereo-
selectivity toward 10.18 However, it was later found that
efficient resolution of the racemic acetate 11 could be achieved
through asymmetric hydrolysis with pig liver esterase. Indeed,
after three rounds of acetylation and esterase-catalyzed
hydrolysis, both (R)- and (S)-10 were obtained in >94%
enantiomeric excess (ee) as determined by 31P NMR analysis
using quinine as a chiral shifting reagent (Figure S2.1−1).18c
Subsequent deprotection of the phosphonate ethyl ester by
TMSBr gave the desired substrate analogues (S)-7 and (R)-7.
The purified (S)-7 and (R)-7 were each incubated with
HppE and the reactions were monitored by 1H NMR
spectroscopy (Figure 1). Both compounds were found to be
substrates of HppE and each was converted to a single product
(Scheme 3). A 1,2-epoxide (13) was produced when (S)-7 was
employed as the substrate, whereas a keto-product (15) was
obtained from the incubation with (R)-7. The chemical nature
of these products was further verified by high-resolution mass
spectrometry (Figure S5−1), and the structure of 15 was also
confirmed by comparison with a synthetic standard. The regio-
and stereoselectivity of the turnover of (S)-7 and (R)-7 are
consistent with those established for 2 and 3. Notably, the
cyclopropyl group is retained in both products (13 and 15).
The conversion of (S)-7 to 13 is not surprising, since the
putative C1 radical intermediate (12) is not expected to affect
the cyclopropyl group at the C3 position. However, the
conversion of (R)-7 to 15 as the sole product is in contrast to
the anticipation that the C2-centered carbinyl radical (i.e., 14),
once formed, would trigger ring-opening of the adjacent
cyclopropyl group. At first glance, this observation appears to
contradict the proposed radical-mediated oxidation mechanism.
However, the retention of the cyclopropyl group in the product
may be ascribed to the subsequent one-electron oxidation step
(14 → 15) being more facile than the competing ring-opening
reaction.
Considering that the ring-opening rate constant of
(methylenecyclopropyl)carbinyl radical (k = 6 × 109 s−1 at
298 K)19 is nearly 2 orders of magnitude greater than that of
the cyclopropylcarbinyl radical (k = 8.6 × 107 s−1 at 298 K),20 a
substrate analogue carrying a methylenecyclopropyl group (16)
was designed and synthesized as a more sensitive probe to
detect the transient formation of radical intermediate by HppE.
The synthetic route to 16 started with nucleophilic addition of
methyldiethylphosphonate to ester 17,21 followed by Luche
reduction of the resulting ketone (18) to give 19 (Scheme 4).
Scheme 4. Synthetic Scheme for the Preparation of 16
16172
dx.doi.org/10.1021/ja3078126 | J. Am. Chem. Soc. 2012, 134, 16171−16174

Journal of the American Chemical Society
Communication
Since attempts to resolve this racemic alcohol using various
lipases and esterases failed, 19 was converted to 16 as a racemic
mixture. While enantiomerically pure substrates are preferred,
experiments performed with the racemic sample can still
provide evidence to discern the possible formation of radical
intermediates in the HppE-catalyzed reactions.
(2R)-16. Two mechanistic scenarios can be envisioned for the
inactivation of HppE by the 2R-epimer of 16 (Scheme 6). In
Scheme 6. Proposed Mechanisms for the Inactivation of
HppE by (2R)-16
When compound 16 was incubated with HppE, formation of
a new product was detected by NMR spectroscopy (Figure 2).
route a, 16 is first converted to the corresponding 2-oxo
phosphonate 24 in a manner analogous to the oxidation of (R)-
3 and (R)-7 by HppE. Since the cyclopropyl group of 24 is
primed by the newly formed electron-withdrawing 2-keto
functionality, it may be susceptible to the attack by an active
site nucleophile in a Michael-addition reaction.22 An alternative
explanation for the observed inactivation (route b) lies in the
ultrafast radical-triggered ring-opening of the methylenecyclo-
propyl group, whereby the C2 radical intermediate (25)
rearranges to an allylic radical that inactivates the enzyme.
To distinguish between these two possibilities, the 2-keto
product (24) was chemically synthesized and tested as an
inhibitor of HppE activity. (S)-HPP (2) was added to the
HppE reaction system that had been preincubated with an
excess amount of 24 (20-fold excess than enzyme), and the
enzyme activity was determined following fosfomycin for-
mation by 1H NMR (Figure S4−3). No difference in the rate or
extent of fosfomycin formation was detected in the presence or
absence of 24, clearly indicating that this putative turnover
product is not an inactivator of HppE. Therefore, a mechanism
initiated by a radical-triggered ring-opening is a more likely
mode of inactivation of HppE by (R)-17. Consistent with a
radical-induced enzyme modification, HppE activity could not
be restored after extensive dialysis of the inactivated enzyme.
Moreover, given the differences in ring-opening rates of the
cyclopropylcarbinyl and methylenecyclopropylcarbinyl radicals
and the fact that inactivation was not observed with (R)-7, the
lifetime of radical 22 can be estimated to be on the order of
nanoseconds, placing it within range of a true intermediate.
In this work, cyclopropyl- and methylenecyclopropyl-based
radical probes were employed to study the mechanism of
HppE-catalysis. Specifically, the cyclopropyl-containing ana-
logues (S)-7 and (R)-7 were found to be processed by HppE to
epoxide 13 and ketone 15 products, respectively. Despite the
presence of a cyclopropyl radical reporting group in these
structures, the reactions of (S)-7 and (R)-7 are free from
apparent radical-triggered ring-opening events. In contrast,
incubation with the ultrafast radical probe 16 led to the
formation of 24 and irreversible enzyme inactivation. While
product 20 is thought to be generated from (2S)-16 via C1-
centered radical intermediate 21, the observed inactivation is
likely a result of radical induced ring-opening of the C2-
centered radical intermediate (22) derived from (2R)-16.
These observations are significant because they provide the first
experimental evidence supporting the involvement of substrate
radical intermediate(s) in an HppE-catalyzed reactions, and
also allow estimation of the rate of the subsequent electron
transfer step to be between 8.6 × 107 and 6 × 109 s−1 based on
1
Figure 2. H NMR assay of HppE with 16. Spectra were recorded at
different time points. Only the region (5.1−5.4 ppm) of the olefinic
protons is shown. No further consumption of 16 was observed after
∼10 min.
However, the conversion was incomplete with no further
consumption of 16 observed after ∼10 min. No fosfomycin
formation could be detected when the natural substrate, (S)-
HPP (2), was incubated with HppE that had been pretreated
with 16, suggesting that this compound (16) also inactivates
the enzyme. The fact that 16 can act as both a substrate and an
inactivator for HppE may simply be a consequence of the
racemic nature of 16 used for these assays. Namely, the
outcome of the incubation is governed by the C2 stereo-
chemistry of 16, with one epimer serves as the substrate and the
other acts as an inactivator.
The structure of the turnover product of 16 is a molecule
featuring both an epoxide ring and an intact methylenecyclo-
propyl group. This is supported by the appearance of new
peaks at 2.7−2.9 and 5.3 ppm, which are characteristic
resonances for the oxiranyl protons and the olefinic protons
of a terminal alkene, respectively. By analogy to the reaction
with (S)-7, this product may be assigned to the corresponding
1,2-epoxyl compound 20 derived from the (2S)-epimer of 16
(Scheme 5). The presence of 20 in the reaction system was also
confirmed using high resolution mass spectrometry ([M − H]−
calcd for C6H8O4P−, m/z 175.0166; found, m/z 175.0160)
(Figure S5−2).
As (2S)-16 is likely to be a substrate of HppE, the
inactivation of HppE may thus be attributed to the action of
Scheme 5. Proposed Reactions of (2S)- and (2R)-16 with
HppE
16173
dx.doi.org/10.1021/ja3078126 | J. Am. Chem. Soc. 2012, 134, 16171−16174

Journal of the American Chemical Society
Communication
Nat. Chem. Biol. 2008, 4, 186. (i) Bollinger, J. M., Jr.; Krebs, C. Curr.
Opin. Chem Biol. 2007, 11, 151. (j) van der Donk, W. A.; Krebs, C.;
Bollinger, J. M., Jr. Curr. Opin. Struct. Biol. 2010, 20, 673.
(14) For examples see: (a) Ryle, M. J.; Liu, A.; Muthukumaran, R. B.;
Ho, R. Y. N.; Koehntop, K. D.; McCracken, J., Jr.; Que, L.; Hausinger,
R. P. Biochemistry 2003, 42, 1854. (b) Howard-Jones, A. R.; Elkins, J.
M.; Clifton, I. J.; Roach, P. L.; Adlington, R. M.; Baldwin, J. E.;
Rutledge, P. J. Biochemistry 2007, 46, 4755.
(15) (a) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1980, 13, 317.
(b) Newcomb, M. Tetrahedron 1993, 49, 1151. (c) Silverman, R. B.
Acc. Chem. Res. 1995, 28, 335. (d) Newcomb, M.; Toy, P. H. Acc.
Chem. Res. 2000, 33, 449. (e) Salaun, J. In Topics in Current Chemistry;
Springer-Verlag: Berlin, 2000; Vol. 207, pp 1−67. (f) Ortiz de
Montellano, P. R. Chem. Rev. 2010, 110, 932.
the ring-opening rate constants of the cyclopropylcarbinyl and
methylenecyclopropylcarbinyl radicals.
ASSOCIATED CONTENT
* Supporting Information
■
S
Details of synthetic methods for compounds 7, 15, 16, and 24,
and experimental procedures of the enzymatic reactions of
HppE with each substrate. This material is available free of
charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
h.w.liu@mail.utexas.edu
■
(16) (a) Lenn, N. D.; Shih, Y.; Stankovuch, M. T.; Liu, H.-w. J. Am.
Chem. Soc. 1989, 111, 3065. (b) Lai, M. T.; Li, D.; Oh, E.; Liu, H.-w. J.
Am. Chem. Soc. 1993, 115, 1619.
Author Contributions
‡H.H. and W.-c.C. contributed equally.
(17) Laurent, M.; Ceresiat, M.; Marchand-Brynaert, J. Eur. J. Org.
Chem. 2006, 3755.
Notes
The authors declare no competing financial interest.
(18) (a) Zhang, Y. H.; Yuan, C. Y.; Li, Z. Y. Tetrahedron 2002, 58,
2973. (b) Zurawinski, R.; Nakamura, K.; Drabowicz, J.; Kielbasinski,
P.; Mikolajczyk, M. Tetrahedron: Asymmetry 2001, 12, 3139.
ACKNOWLEDGMENTS
■
This work is supported in part by grants from the National
Institutes of Health (GM040541 to H.-w.L) and the Robert A.
Welch Foundation (F-1511 to H.-w.L. and A-1176 to D.H.R.).
We thank Steve Sorey for his help on acquiring the NMR
spectra.
̇
́ ́
(c) Zymanczyk-Duda, E.; Skwarczynski, M.; Lejczak, B.; Kafarski, P.
Tetrahedron: Asymmetry 1996, 7, 1277.
(19) Horner, J. H.; Johnson, C. C.; Lai, M. T.; Lin, H.-w.;
Martinesker, A. A.; Newcomb, M.; Oh, E. Bioorg. Med. Chem. Lett.
1994, 4, 2693.
(20) Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1991,
113, 5687.
(21) Lai, M. T.; Liu, L. D.; Liu, H.-w. J. Am. Chem. Soc. 1991, 113,
7388.
(22) Agnihotri, G.; He, S. M.; Hong, L.; Dakoji, S.; Withers, S. G.;
Liu, H.-w. Biochemistry 2002, 41, 1843.
REFERENCES
(1) Raz, R. Clin. Microbiol. Infect. 2012, 18, 4.
(2) (a) Marquardt, J. L.; Brown, E. D.; Lane, W. S.; Haley, T. M.;
Ichikawa, Y.; Wong, C. H.; Walsh, C. T. Biochemistry 1994, 47, 10646.
■
(b) Eschenburg, S.; Priestman, M.; Schonbrunn, E. J. Biol. Chem. 2004,
280, 3757.
̈
(3) Hidaka, T.; Goda, M.; Kuzuyama, T.; Takei, N.; Hidaka, M.; Seto,
H. Mol. Gen. Genet. 1995, 249, 274.
(4) Liu, P.; Murakami, K.; Seki, T.; He, X.; Yeung, S. M.; Kuzuyama,
T.; Seto, H.; Liu, H.-w. J. Am. Chem. Soc. 2001, 123, 4619.
(5) Liu, P.; Liu, A.; Yan, F.; Wolfe, M. D.; Lipscomb, J. D.; Liu, H.-w.
Biochemistry 2003, 42, 11577.
(6) Thibodeaux, C. J.; Chang, W.-c.; Liu, H.-w. Chem. Rev. 2012, 112,
1681.
(7) (a) Hammerschmidt, F.; Bovermann, G.; Bayer, K. Liebigs Ann.
Chem. 1990, 1055. (b) Hammerschmidt, F. J. Chem. Soc., Perkin Trans.
1 1991, 1993.
(8) (a) Yan, F.; Munos, J. W.; Liu, P.; Liu, H.-w. Biochemistry 2006,
45, 11473. (b) Munos, J. W.; Moon, S.-J.; Mansoorabadi, S. O.; Chang,
W.-c.; Hong, L.; Yan, F.; Liu, A.; Liu, H.-w. Biochemistry 2008, 47,
8726.
(9) Mirica, L. M.; McCusker, K. P.; Munos, J. W.; Liu, H.-w.;
Klinman, J. P. J. Am. Chem. Soc. 2008, 130, 8122.
(10) Zhao, Z.; Liu, P.; Murakami, K.; Kuzuyama, T.; Seto, H.; Liu,
H.-w. Angew. Chem., Int. Ed. 2002, 41, 4529.
(11) (a) Woschek, A.; Wuggenig, F.; Peti, W.; Hammerschmidt, F.
ChemBioChem 2002, 3, 829. (b) Higgins, L. J.; Yan, F.; Liu, P.; Liu, H.-
w.; Drennan, C. L. Nature 2005, 437, 838.
(12) (a) Yan, F.; Moon, S.-J.; Liu, P.; Zhao, Z.; Lipscomb, J. D.; Liu,
A.; Liu, H.-w. Biochemistry 2007, 46, 12628. (b) Yun, D.; Dey, M.;
Higgins, L. J.; Yan, F.; Liu, H.-w.; Drennan, C. L. J. Am. Chem. Soc.
2011, 133, 11262.
(13) (a) Kappock, T. J.; Caradonna, J. P. Chem. Rev. 1996, 96, 2659.
(b) Hegg, E. L.; Que, L. Eur. J. Biochem. 1997, 250, 625. (c) Rocklin,
A. M.; Tierney, D. L.; Kofman, V.; Brunhuber, N. M. W.; Hoffman, B.
M.; Christoffersen, R. E.; Reich, N. O.; Lipscomb, J. D.; Que, L. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 7905. (d) Fitzpatrick, P. F. Annu. Rev.
Biochem. 1999, 68, 355. (e) Gibson, D. T.; Parales, R. E. Curr. Opin.
Biotechnol. 2000, 11, 236. (f) Prescott, A. G.; Lloyd, M. D. Nat. Prod.
Rep. 2000, 17, 367. (g) Costas, M.; Mehn, M. P.; Jensen, M. P.; Que,
L., Jr. Chem. Rev. 2004, 104, 939. (h) Kovaleva, E. G.; Lipscomb, J. D.
16174
dx.doi.org/10.1021/ja3078126 | J. Am. Chem. Soc. 2012, 134, 16171−16174