On the conversion of structural analogues of (S)-2-hydroxypropylphosphonic acid to epoxides by the final enzyme of fosfomycin biosynthesis in S. fradiae

Article abstract of DOI:10.1016/j.bmcl.2007.12.012

2-Hydroxyethyl- and (S)-2-hydroxybutylphosphonic acid were prepared, starting in the latter case from (S)-2-aminobutyric acid. They were fed to cultures of Streptomyces fradiae producing fosfomycin. Only the latter (150 μg/mL of medium) was converted to the ethyl analogue of fosfomycin, isolated as 2-amino-1-hydroxybutylphosphonic acid (3%) in admixture with 2-amino-1-hydroxypropylphosphonic acid (97%) derived from fosfomycin.

Full text of DOI:10.1016/j.bmcl.2007.12.012

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3056–3059
On the conversion of structural analogues of
(S)-2-hydroxypropylphosphonic acid to epoxides by the
final enzyme of fosfomycin biosynthesis in S. fradiae
Anna Schweifer and Friedrich Hammerschmidt*
Institute of Organic Chemistry, University of Vienna, Wa¨hringerstrasse 38, A-1090 Wien, Austria
Received 15 October 2007; revised 4 December 2007; accepted 6 December 2007
Available online 26 December 2007
Abstract—2-Hydroxyethyl- and (S)-2-hydroxybutylphosphonic acid were prepared, starting in the latter case from (S)-2-aminobu-
tyric acid. They were fed to cultures of Streptomyces fradiae producing fosfomycin. Only the latter (150 lg/mL of medium) was con-
verted to the ethyl analogue of fosfomycin, isolated as 2-amino-1-hydroxybutylphosphonic acid (3%) in admixture with 2-amino-1-
hydroxypropylphosphonic acid (97%) derived from fosfomycin.
Ó 2008 Elsevier Ltd. All rights reserved.
Fosfomycin [1, (1R,2S)-epoxypropylphosphonic acid] is
a clinically utilized antibiotic of low toxicity, blocking
bacterial cell wall biosynthesis by acting as an analogue
of phosphoenolpyruvate (PEP).1–3 It is one of the rare
natural products containing a P–C bond,4 produced by
various species of Streptomyces,5 Pseudomonas syringae,
and P. viridiflava.6,7 The biosynthesis of fosfomycin was
unraveled by using feeding experiments with labeled pre-
cursors8–12 and genetic13 techniques. It comprises five
Scheme 1. Biosynthesis of fosfomycin.
steps, three of which are unique, starting from the pri-
mary metabolite PEP (2), which is rearranged reversibly
by PEP mutase to give phosphonopyruvate (3) (Scheme
Inspired by the broad substrate specificity of isopenicil-
1).14–16
lin N synthase,25 also a non-heme iron dependent oxy-
genase, we decided to prepare analogues of Hpp, in
the first place those with a hydrogen atom or an ethyl
group replacing the methyl group in 6 to study the sub-
strate specificity of the epoxidase. These homologues, 7
and 8, were probed for their conversion to the analogous
epoxides of fosfomycin by cultures of Streptomyces fra-
diae (Scheme 2). When this work was started, the epox-
idase had not yet been purified and characterized. The
cyclohexylammonium salt of 2-hydroxyethylphosphonic
acid (7) was prepared by a literature procedure.11 The
Decarboxylation17 and reduction18 produce 2-hydroxy-
ethylphosphonic acid (5). The recently elucidated meth-
ylation of 5 producing (S)-2-hydroxypropylphosphonic
acid (6, Hpp) follows a unique radical mechanism with
SAM as methyl donor.18,19 The final step is a dehy-
drogenative cyclization performed by a non-heme iron
oxygenase [(S)-2-hydroxypropylphosphonic acid epoxi-
dase, HppE].20–24 Liu et al. found that the epoxidase
converted (R)-2-hydroxypropylphosphonic and (S)-1,1-
difluoro-2-hydroxypropylphosphonic acid to 2-oxopro-
pylphosphonic acids, which has some bearing on the
mechanism of the epoxide ring closure.20
Keywords: Fosfomycin; Biosynthesis; Phosphonic acids.
*
Corresponding author. Tel.: +43 1 4277 52105; fax: +43 1 4277
9521; e-mail: friedrich.hammerschmidt@univie.ac.at
Scheme 2. Conversion of homologues of 6 to epoxides by S. fradiae.
0960-894X/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.12.012

A. Schweifer, F. Hammerschmidt / Bioorg. Med. Chem. Lett. 18 (2008) 3056–3059
3057
Scheme 4. Conversion of epoxides 9 and 10 to amino-hydroxyphos-
phonic acids by ammonia.
detection of the aminophosphonic acid 17 or 18 in the
isolated mixture.
These arguments induced us to study the behavior of
epoxides 9 and 10 toward ammonia and prepare refer-
ence compounds 17–19 (Scheme 4). Epoxide 9 will be
opened by ammonia exclusively by attack at the less hin-
dered site at C-2.31 As 2-amino-1-hydroxyethylphos-
phonic acid32 (17) was available from a previous
project and its behavior was therefore known, it was
not necessary to study the conversion of 9 to 17.
Scheme 3. Conversion of 2-aminobutyric acid to 2-hydroxyphosphon-
ic acid (S)-8 (as cyclohexylammonium salt).
However, the preparation of epoxide 10, for conve-
nience in the racemic form, and its conversion to ami-
no-hydroxybutylphosphonic acids 18 and 19 had to be
performed (Scheme 5). The used reactions are related
to the ones used for the synthesis of fosfomycin,
although some modifications had to be introduced.33
Propargyl alcohol (20) was reacted with diethyl phosp-
horochloridite to phosphite 22 as intermediate, which
underwent a smooth 2,3-sigmatropic rearrangement at
ambient temperature to allenylphosphonate 23 in 47%
yield.34 The selective hydrogenation of one double bond
was critical.35 Using Pd/CaCO3/Pb/H2 (1 atm)/chinoline
in dry ethanol furnished under optimized conditions at
room temperature the desired (Z)-1-butenylphospho-
nate 23 (42%) and a mixture of (E)-23 and possibly 2-
butenylphosphonates (ratio 2:1, 20%) by flash chroma-
synthesis of 8 is given in Scheme 3, starting from (S)-2-
aminobutyric acid.26,27 [(S)-11] as chiral precursor. The
carboxyl group of the intermediate 2-hydroxyacid was
esterified26 and the hydroxyl group was benzylated28
using Bundle’s reagent to give ester (S)-12 in an overall
yield of 18%. Reduction of the ester (81%), conversion
of the alcohol to the bromide (93%) followed by an
Arbusov reaction with triethyl phosphite (95%) furnished
2-benzyloxyphosphonate (S)-15.11,29 Removal of the
protecting groups (TMSBr for Et on phosphorus, Pd/C/
H2 for Bn) gave phosphonic acid (S)-8, which was con-
verted to the crystalline cyclohexylammonium salt for
purification, containing 1.7 mol of amine as found by
1H NMR spectroscopy. To establish the ee of the final
product,
a
sample of phosphonate (S)-15 was
hydro-genolytically deprotected and esterified with (S)-
MTPACl to yield Mosher ester 16. Similarly, a
reference sample of racemic 1530 was esterified. On the
basis of the 1H NMR spectra, the ee of (S)-15 was
found to be 98%. The salt of acid (S)-8 should have the
same ee, as the stereochemistry is not affected on depro-
tection of 15.
Isolation of fosfomycin from the broth of S. fradiae was
not possible because of its low concentration (up to
10 lg/mL) and its similar behavior to phosphoric acid
(1 mg/mL of K2HPO4 in the medium) on ion exchange
chromatography.8 Therefore, the antibiotic was con-
verted to two isomeric aminophosphonic acids by ring
opening with ammonia, of which the (1R,2R)-2-amino-
1-hydroxypropylphosphonic acid was amenable to isola-
tion by ion exchange chromatography.9 Assuming that
the analogues 9 and 10 of fosfomycin are formed in
smaller amounts than fosfomycin, if at all and then in
admixture with it, we thought that treatment of the
broth would yield amino-hydroxyphosphonic acids.
The 2-amino-1-hydroxyphosphonic acids derived from
fosfomycin and an analogue would behave similarly
upon purification. NMR spectroscopy would allow the
Scheme 5. Preparation of epoxide ( )-10 and its ring opening with
NH3.

3058
A. Schweifer, F. Hammerschmidt / Bioorg. Med. Chem. Lett. 18 (2008) 3056–3059
tography (hexanes/EtOAc 1:2). Removal of ethyl pro-
tecting groups at phosphorus effected by our standard
method (TMSBr/allyltrimethylsilane) gave the free
phosphonic acid, which was directly converted to the tri-
ethylammonium salt and epoxidized at 40 °C with H2O2
using Na2WO4Æ3H2O as catalyst.33 Anticipating difficul-
ties with its purification, the racemic epoxide was not
isolated, but heated with ammonia at 70 °C for 65 h.
The two aminophosphonic acids ( )-18 and ( )-19
were isolated by ion exchange chromatography and ob-
tained in a ratio of 2.5:1, identical with the ratio of the
analogous aminophosphonic acids obtained from fosfo-
mycin.9 2-Amino-1-hydroxypropylphosphonic acids de-
rived from fosfomycin and 17 or ( )-18 behaved
similarly upon ion exchange chromatography, thus
insuring co-isolation from the broth.
too efficiently in the methylation reaction to be used by
the epoxidase. Alternatively, epoxide 9 could be much
more prone to chemical hydrolysis than fosfomycin or
10, thus escaping transformation to aminophosphonic
acid 17 with ammonia. One really would have to com-
pare the reactions of these two substrates with the reac-
tion of (S)-2-hydroxypropylphosphonic acid with the
isolated epoxidase in vitro. The whole cell feedings ad-
dress a range of factors, not just the substrate specificity
of the epoxidase.
In summary, we have prepared the lower and higher
homologue of Hpp and fed it to S. fradiae. Evidently,
only the latter, the (S)-2-hydroxybutylphosphonic acid,
was converted to the analogue of fosfomycin, being
present in the broth as proven by conversion to amino-
phosphonic acids. The 2-aminophosphonic acids were
isolated and identified by NMR spectroscopy and found
to be present in a ratio of 3:95 derived from 10 and 1,
respectively.
The first feeding experiment was performed with 2-
hydroxyethylphosphonic acid (7) added to the growth
medium of S. fradiae (six 1 L flasks with 220 mL of med-
ium each, 500 lg/mL of 2-hydroxyethylphosphonic
acid).9 After 3 days, when the fosfomycin concentration
was 15 lg/mL of broth, the cells were harvested and the
supernatants were pooled, saturated with ammonia, and
heated at 60 °C for 3 days. The 2-amino-1-hydroxypro-
pylphosphonic acid was derived from fosfomycin possi-
bly containing 2-amino-1-hydroxyethylphosphonic acid
(17) was isolated (6 mg) by triple ion exchange chroma-
tography. The 1-amino-2-hydroxypropylphosphonic
was not retained by the first column of Dowex
50W,H+. Surprisingly, 17 could not be detected in the
product by NMR spectroscopy (1H, 31P), however 2%
of (1S,2R)-2-amino-1-hydroxypropylphosphonic acid
derived from trans-fosfomycin.36 Previous feeding
experiments with 1,1-dideuterated 2-hydroxyethylphos-
phonic acid (150 lg/mL) yielded 2-amino-1-hydroxypro-
pylphosphonic acid with 32% D. As the samples were
purified at that time by a final crystallization, the now
looked for 2-amino-1-hydroxyethylphosphonic acid
could not be found, even if it was present in the crude
product.
Note added in proof: After submitting the manuscript, a
paper by Liu et al.37 was published, reporting on the
determination of the substrate binding mode to the ac-
tive site iron of (S)-2-hydroxypropylphosphonic acid
epoxidase using17 O-enriched substrates and substrate
analogues [isobutyl-, (R)- and (S)-2-hydroxy-2-phenyl-,
and (R)- and (S)-3-hydroxybutylphosphonic acid]. (S)-
2-Hydroxy-2-phenylethylphosphonic acid was con-
verted to the epoxide by the isolated oxygenase, albeit
at much reduced rate. (S)-3-Hydroxybutylphosphonic
acid furnished (S)-3,4-dihydroxybutylphosphonic acid
and both (R)-enantiomers gave the corresponding
keto-phosphonic acids.
Acknowledgment
We thank the Austrian Fonds zur Fo¨rderung der wis-
senschaftlichen Forschung for financial support (Grant
No. P 7183-CHE).
Similarly, a second feeding experiment was set up with
(S)-2-hydroxybutylphosphonic acid (8, 150 lg/mL) as
substrate, obtained by passing down the cyclohexylam-
monium salt a column of Dowex 50W,H+. The isolated
mixture of aminophosphonic acids was very carefully
Supplementary data
All experimental procedures and spectroscopic data.
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/
j.bmcl.2007.12.012.
1
investigated by H, 31P, and 13C NMR spectroscopy.
The signals of 18 were clearly detectable in all three
spectra. The ratio of 18, 2-amino-1-hydroxypropylphos-
phonic acid derived from fosfomycin, and the 2-amino-
phosphonic acid derived from trans-fosfomycin as
determined by 31P NMR spectroscopy was 3:95:2. When
1,1-dideuterated 6 (30 lg/mL) was fed to S. fradiae, the
isolated 2-amino-1-hydroxypropylphosphonic acid con-
tained 37% of the labeled species.11
References and notes
1. Kahan, F. M.; Kahan, J. S.; Cassidy, P. J.; Kropp, H.
Ann. N.Y. Acad. Sci. 1974, 235, 364.
2. Skarzynski, T.; Mistry, A.; Wonacott, A.; Hutchinson, S.
E.; Kelly, V. A.; Duncan, K. Structure 1996, 4, 1465.
3. Ribes, S.; Taberner, F.; Domenech, A.; Cabellos, C.;
Tubau, F.; Linares, J.; Viladrich, P. F.; Gudiol, F.
J. Antimicrob. Chemother. 2006, 57, 931.
4. Seto, H.; Kuzuyama, T. Nat. Prod. Rep. 1999, 16, 589.
5. Hendlin, D.; Stapley, E. O.; Jackson, M.; Wallick, H.;
Miller, A. K.; Wolf, F. J.; Miller, T. W.; Chaiet, L.;
The missing detection of 2-amino-1-hydroxyethylphos-
phonic acid (17) as proof for the formation of epoxide
9 and the very low incorporation of 8 into epoxide 10
do not necessarily indicate that the Hpp epoxidase is a
highly substrate specific enzyme. The former compound,
the substrate for the methylase, could probably be used

A. Schweifer, F. Hammerschmidt / Bioorg. Med. Chem. Lett. 18 (2008) 3056–3059
3059
Kahan, F. M.; Foltz, E. L.; Woodruff, H. B.; Mata, J. M.;
Hernandez, S.; Mochales, S. Science 1969, 166, 122.
6. Shoji, J.; Kato, T.; Hinoo, H.; Hattori, T.; Hirooka, K.;
Matsumoto, K.; Tanimoto, T.; Kondo, E. J. Antibiot.
1986, 39, 1011.
20. Zhao, Z.; Liu, P.; Murakami, K.; Kuzuyama, T.; Seto, H.;
Liu, H. Angew. Chem., Int. Ed. 2002, 41, 4529.
21. Liu, P.; Mehn, M. P.; Yan, F.; Zhao, Z.; Que, L., Jr.; Liu,
H. J. Am. Chem. Soc. 2004, 126, 10306.
22. McLuskey, K.; Cameron, S.; Hammerschmidt, F.;
Hunter, W. N. Proc. Natl. Acad. Sci. U.S.A. 2005,
102, 14221.
7. Katayama, N.; Tsubotani, S.; Nozaki, Y.; Harada, S.;
Ono, H. J. Antibiot. 1990, 43, 238.
8. Rogers, T. O.; Birnbaum, J. Antimicrob. Agents Chemo-
ther. 1974, 5, 121.
23. Higgins, L. J.; Yan, F.; Liu, P.; Liu, H.; Drennan, C. L.
Nature 2005, 437, 834.
9. Hammerschmidt, F.; Bovermann, G.; Bayer, K. Liebigs
Ann. Chem. 1990, 1055.
24. Yan, F.; Munos, J. W.; Liu, P.; Liu, H. Biochemistry 2006,
45, 11473.
10. Hammerschmidt, F. Angew. Chem., Int. Ed. Engl. 1994,
33, 341.
11. Hammerschmidt, F.; Ka¨hlig, H. J. Org. Chem. 1991, 56,
2364.
12. Peric Simov, B.; Wuggenig, F.; La¨mmerhofer, M.; Lind-
ner, W.; Zarbl, E.; Hammerschmidt, F. Eur. J. Org. Chem.
2002, 1139.
13. Hidaka, T.; Goda, M.; Kuzuyama, T.; Takei, N.; Hidaka,
M.; Seto, H. Mol. Gen. Genet. 1995, 249, 274.
14. McQueney, M. S.; Lee, S.; Swartz, W. H.; Ammon, H. L.;
Mariano, P. S.; Dunaway-Mariano, D. J. Org. Chem.
1991, 56, 7121.
15. Seidel, H. M.; Freeman, S.; Schwalbe, C. H.; Knowles, J.
R. J. Am. Chem. Soc. 1990, 112, 8149.
16. Hidaka, T.; Iwakura, H.; Imai, S.; Seto, H. J. Antibiot.
1992, 45, 1008.
17. Nakashita, H.; Watanabe, K.; Hara, O.; Hidaka, T.; Seto,
H. J. Antibiot. 1997, 50, 212.
25. Baldwin, J. E.; Bradley, M. Chem. Rev. 1990, 90, 1079.
26. Mori, K.; Sasaki, M.; Tamada, S.; Suguro, T.; Masuda, S.
Tetrahedron 1979, 35, 1601.
27. Chenault, H. K.; Kim, M.-J.; Akiyama, A.; Miyazawa, T.;
Simon, E. S.; Whitesides, G. M. J. Org. Chem. 1987, 52,
2608.
28. Widmer, U. Synthesis 1987, 568.
29. Hammerschmidt, F. Monatsh. Chem. 1991, 122, 389.
30. Mikolajczyk, M.; Balczewski, P. Synthesis 1984, 691.
´
31. Wroblewski, A. E.; Halajewska-Wosik, A. Tetrahedron:
Asymmetry 2000, 11, 2053.
32. Hammerschmidt, F.; Lindner, W.; Wuggenig, F.; Zarbl, E.
Tetrahedron: Asymmetry 2000, 11, 2955.
33. Glamkowski, E. J.; Gal, G.; Purick, R.; Davidson, A. J.;
Sletzinger, M. J. Org. Chem. 1970, 35, 3510.
34. Mark, V. Tetrahedron Lett. 1962, 3, 281.
35. Petrov, A. A.; Ionin, B. I.; Ignatyev, V. M. Tetrahedron
Lett. 1968, 9, 15.
18. Kuzuyama, T.; Hidaka, T.; Kamigiri, K.; Imai, S.; Seto,
H. J. Antibiot. 1992, 45, 1812.
36. Woschek, A.; Wuggenig, F.; Peti, W.; Hammerschmidt, F.
ChemBioChem 2002, 3, 829.
19. Woodyer, R. D.; Li, G.; Zhao, H.; van der Donk, W. A.
J. Chem. Soc., Chem. Commun. 2007, 359.
37. Yan, F.; Moon, S.-J.; Liu, P.; Zhao, Z.; Lipscomb, J. D.;
Liu, A.; Liu, H. Biochemistry 2007, 46, 12628.