Studies on the biodegradation of fosfomycin: Synthesis of 13C-labeled intermediates, feeding experiments with rhizobium huakuii PMY1, and Isolation of Labeled Amino Acids from Cell Mass by HPLC

Article abstract of DOI:10.1002/chem.201100725

Racemic (1R*,2R*)-1,2-dihydroxy-[1-¹³C ₁]propylphosphonic acid and 1-hydroxy-[1-¹³C ₁]acetone were synthesized and fed to R. huakuii PMY1. Alanine and a mixture of valine and methionine were isolated as their N-acetyl derivatives from the cell hydrolysate by reversed-phase HPLC and analyzed by NMR spectroscopy. It was found that the carbon atoms of the respective carboxyl groups were highly ¹³C-labeled (up to 65 %). Hydroxyacetone is therefore considered an obligatory intermediate of the biodegradation of fosfomycin by R. huakuii PMY1.

Full text of DOI:10.1002/chem.201100725

FULL PAPER
DOI: 10.1002/chem.201100725
13
Studies on the Biodegradation of Fosfomycin: Synthesis of C-Labeled
Intermediates, Feeding Experiments with Rhizobium huakuii PMY1, and
Isolation of Labeled Amino Acids from Cell Mass by HPLC
John W. McGrath,[a] Friedrich Hammerschmidt,*[b] Hanspeter Kꢀhlig,[b]
Frank Wuggenig,[b] Gꢁnther Lamprecht,[c] and John P. Quinn*[a]
Abstract: Racemic (1R*,2R*)-1,2-dihydroxy-[1-13C1]propylphosphonic acid and 1-
hydroxy-[1-13C1]acetone were synthesized and fed to R. huakuii PMY1. Alanine
and a mixture of valine and methionine were isolated as their N-acetyl derivatives
from the cell hydrolysate by reversed-phase HPLC and analyzed by NMR spec-
troscopy. It was found that the carbon atoms of the respective carboxyl groups
were highly 13C-labeled (up to 65%). Hydroxyacetone is therefore considered an
obligatory intermediate of the biodegradation of fosfomycin by R. huakuii PMY1.
Keywords: amino acids · biotrans-
formations · fosfomycin · hydroxy-
AHCTUNGTERGaNNUN cetone · isotopic labeling
Introduction
phate (Pi), as the preferred phosphorus source, is vital for
any living cell. Under conditions of phosphorus starvation,
gene systems are induced, the products of which are in-
volved in the acquisition and assimilation of phosphorus
from a range of organic phosphorus substrates, including
phosphonates. For the most part, microorganisms use phos-
phonates only as a phosphorus source, with this breakdown
being strictly regulated by the pho regulon via the Pi re-
leased from the phosphonate substrate. Some bacteria, how-
Fosfomycin (1) is one of the few natural products to contain
[1]
À
a chemically very stable P C bond. It is produced by
strains of Streptomyces and Pseudomonas syringae and is
utilized as a clinical antibiotic that blocks[2] the formation of
bacterial cell walls.
À
À
ever, produce P C bond-cleaving enzymes (P C hydro-
AHCTUNGTREGlNNUN ases), which are not subject to control by Pi, thereby facili-
tating microbial utilization of the phosphonate as a phos-
phorus, carbon (and nitrogen), and energy source.[3,8] Some
years ago, we reported the isolation of the bacterium Rhi-
zobium huakuii PMY1 using fosfomycin at 10 mm as a
carbon or carbon and phosphorus source with concomitant
Pi release.[9]
Elucidation of the fosfomycin biosynthetic pathway, which
comprises five metabolic steps (starting from phosphoenol-
pyruvate), has proved challenging, since three of these steps
involve extraordinary chemical transformations.[3–7] Mean-
while, its mineralization remains largely unexplored, as is
the case for the majority of man-made and biogenic phos-
phonates, with the exception of 2-aminoethylphosphonic
acid (2). Acquisition of adequate supplies of inorganic phos-
More recently, we demonstrated that of the four stereo-
isomeric 1,2-dihydroxypropylphosphonic acids only the
(1R,2R)-isomer supported growth, as evidenced by the re-
lease of Pi.[10] None of the salts of (Æ)-1,2-epoxybutyl-, (Æ)-
1,2-dihydroxyethyl-, 2-oxopropyl-, (S)-2-hydroxypropyl-,
(Æ)-1-hydroxypropyl-, or (Æ)-1-hydroxy-2-oxopropylphos-
[a] Dr. J. W. McGrath, Dr. J. P. Quinn
School of Biological Sciences, The Queenꢀs University of Belfast
Medical Biology Centre
À
phonic acid supported growth. In vitro P C bond cleavage
activity was detected only for the latter phosphonic acid. R.
huakuii PMY1 grew well on (R)- and (S)-lactic acid and hy-
droxyacetone, but less well on propionic acid, and not on
acetone or (R)- and (Æ)-1,2-propanediol.[10] Furthermore,
phosphite is not involved in fosfomycin biodegradation since
it cannot be used as a phosphorus source by R. huakuii
PMY1, and no phosphite oxidase activity could be detected
in PMY1 cell-free extracts.[10] These results indicate that the
enzymes involved in the biodegradation of fosfomycin are
very specific and that the first step is a formal hydrolytic
ring opening. Most probably, a glycol monoester[11] is
97 Lisburn Road, Belfast BT9 7BL (Northern Ireland)
Fax : (+44)28-90975877
E-mail: j.mcgrath@qub.ac.uk
j.quinn@qub.ac.uk
[b] Prof. Dr. F. Hammerschmidt, Prof. Dr. H. Kꢁhlig, Dr. F. Wuggenig
Institute of Organic Chemistry, University of Vienna
Wꢁhringerstrasse 38, 1090 Vienna (Austria)
Fax : (+43)1-4277-9521
E-mail: friedrich.hammerschmidt@univie.ac.at
[c] Prof. Dr. G. Lamprecht
Institute of Analytical Chemistry, University of Vienna
Wꢁhringerstrasse 38, 1090 Vienna (Austria)
Chem. Eur. J. 2011, 17, 13341 – 13348
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
13341

Scheme 1. Putative pathway for the biodegradation of fosfomycin.
formed as an enzyme-bound intermediate, which is hydro-
lyzed to (1R,2R)-3 (see Scheme 1), the stereochemical out-
come of which is net inversion of the configuration at C-2.
Interestingly, FosX proteins from M. loti and Listeria cata-
lyze the metal-ion-assisted ring opening of fosfomycin by
water at C-1, giving the (1S,2S)-diol, as a genomic encoded
mechanism of resistance.[12]
Results and Discussion
To date, our studies on the mineralization of fosfomycin by
cells of R. huakuii PMY1 have focused on the P moiety by
quantifying Pi release into the culture medium; fosfomycin
consumption during microbial growth has also been moni-
tored by GC/MS analysis.[9] However, as yet we have no in-
formation as to what happens to the three-carbon unit of
fosfomycin during degradation, and whether hydroxyace-
tone and lactic acid are obligatory intermediates in this
pathway (Scheme 1). We reasoned that it would be possible
to address these points through two complementary feeding
Scheme 2. Synthesis of unlabeled and 13C-labeled (1R*,2R*)-1,2-(Æ)-di-
AHCTUNGTRENGNhUN ydroxypropylphosphonic acids 3.
diethyl trimethylsilyl phosphite to protected lactaldehyde
and was optimized in the unlabeled series.[10]
experiments
starting
with
(1R,2R)-1,2-dihydroxy-
The acetaldehyde cyanohydrin formed as an intermediate
in a mixture of DMF, acetaldehyde, NaCN, imidazole, and
TIPSCl (TIPS=triisopropylsilyl) was silylated immediately
and the product was isolated in 81% yield after 60 h. Re-
duction of the nitrile (Æ)-8 in dry toluene, followed by
acidic work-up to hydrolyze the imine, furnished the crude
TIPS-protected lactaldehyde, which was used immediately
without further purification.[13] Diethyl trimethylsilyl phos-
phite was added to a solution of the aldehyde in dry toluene
at À788C and then the mixture was allowed to warm to
108C.[10] Removal of the volatile components under reduced
pressure and selective monodesilylation with HCl in dry eth-
anol yielded a mixture of racemic diastereomers, which
were separated by flash chromatography. The predominating
and less polar isomer (Æ)-9, with the relative (1R*,2R*) con-
figuration as deduced from the configuration of (Æ)-3, was
obtained in 53% yield, whereas the more polar (Æ)-10 with
the (1R*,2S*) configuration was obtained in 7% yield. The
desired isomer (Æ)-9 was deprotected in two steps, first with
HF in CH3CN/H2O in 98% yield, and then with TMSBr/al-
lylTMS. The free phosphonic acid was converted into the
ammonium salt, which was identical by NMR spectroscopic
analysis to an authentic sample of the ammonium salt of
[1-13C1]propylphosphonic acid (3). As fosfomycin and all ob-
ligatory intermediates of the biodegradation are used as
carbon and energy sources, part of them will be incorporat-
ed into cellular components. We assume that the phosphorus
is cleaved from a three-carbon-atom portion and that the
functionalized propane is further transformed to pyruvic
acid, which is incorporated into amino acids. We planned to
hydrolyze the harvested cells, after feeding, with 6m HCl
and to isolate the labeled amino acids, possibly after deriva-
tization, by an HPLC method. NMR spectroscopy of the
fractions obtained would allow determination of the location
and quantification of the incorporated 13C atoms from the
precursors.
Synthesis of 13C-labeled 1,2-dihydroxypropylphosphonic
acid: The synthesis of homochiral (1R,2R)-1,2-dihydroxy-
[1-13C1]propylphosphonic acid is quite difficult. We opted for
the much more easily accessible racemic mixture, as we had
established for the unlabeled series that the (1S,2S)-enantio-
mer does not interfere with the biodegradation of the
(1R,2R)-enantiomer by R. huakuii PMY1.[10] The synthesis
as outlined in Scheme 2 was built on the known addition of
13342
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13341 – 13348

Biodegradation of Fosfomycin in R. huakuii PMY1 Cells
FULL PAPER
(1R,2R)-1,2-dihydroxypropylphosphonic acid derived from
fosfomycin.[10] The 13C-labeled species (Æ)-[1-13C1]3 was ac-
cessed analogously in similar yield, except that NaCN was
replaced by commercially available Na13CN (99% 13C).
was the strongest signal in the 13C NMR spectrum. The C-1
position of N-acetylvaline was labeled (23% 13C), as evi-
denced by the doublet (J=5.8 Hz) at d=4.14 ppm flanked
by doublets (J=5.8 Hz) in the 1H NMR spectrum, which
collapsed to a doublet upon 13C decoupling. Furthermore,
the carboxyl signal was strongly enhanced in the 13C NMR
spectrum. The methionine was essentially unlabeled at any
position. These findings show that the three-carbon-atom
fragment of fosfomycin stays intact as long as it is attached
Feeding experiments and isolation of amino acids: Feeding
experiments with labeled precursors are very useful and a
standard methodology in biosynthetic studies. The natural
products formed are isolated and investigated for the incor-
poration of the label. However, elucidation of the minerali-
zation pathway of a natural product or a xenobiotic is more
complicated, as the degradative intermediates, on their way
to water and CO2 as the final metabolic products, are rarely
detected. Knowing that fosfomycin is used as a carbon and
energy source by cells of R. huakuii PMY1, we reasoned
that isolation of labeled amino acids biosynthesized from its
degradation product(s) could help to deduce the biodegra-
dation pathway.
Cells of R. huakuii PMY1 were grown in minimal salts
medium with phosphate-free yeast extract, but omitting the
essential amino acids used previously, supplemented with
undiluted 1,2-dihydroxypropylphosphonic acid (Æ)-[1-13C1]3
(7.5 mm) as the sole phosphorus, carbon, and energy
source.[9] After 10 days of growth at 308C on an orbital
shaker at 100 rpm, 2.2 mm Pi had been released into the cul-
ture medium. No spontaneous Pi release was observed
either in uninoculated control experiments or from cultures
incubated in the absence of substrate. Cells were harvested
at 10000ꢃg for 15 min at 48C and washed once in 10 mm 2-
(N-morpholino)ethanesulfonic acid (MES)/NaOH buffer
(pH 7.0).
À
to the phosphorus atom. The P C bond cleavage releases a
propane derivative of unknown substitution pattern, possi-
bly hydroxyacetone. Propane-1,2-diol has been excluded as
an intermediate of the bio-
AHCTUNGTERGdNNUN egradation pathway of fosfo-
mycin.[10] If 2-oxophosphonic
acid (R)-4 (Scheme 1) is an ob-
ligatory intermediate and not
just a shunt metabolite, an al-
ternative pathway for its degra-
dation could be envisaged
(Scheme 3). A thiamine diphos-
phate-dependent enzyme could
degrade it to acetaldehyde and
formylphosphonic acid (12),
which would be further metab-
olized. The mechanism for this
Scheme 3. Alternative pathway
for the biodegradation of 2-ox-
opropylphosphonic acid (R)-4.
À
C C bond breaking is some-
what reminiscent of the trans-
ketolase-catalyzed reaction of the Calvin cycle.[14] However,
this cleavage is not compatible with the above results.
The sterilized cell suspension was hydrolyzed in 6m HCl
and the liberated amino acids were N-acetylated by reaction
with acetic anhydride. This derivatization and the use of re-
versed-phase HPLC for the isolation of N-acetyl amino
acids were necessary, since preliminary experiments had
shown that alanine isolated by HPLC as its hydrochloride
Synthesis and feeding of 1-hydroxy-[1-13C1]acetone: If hy-
droxyacetone is a degradation product of fosfomycin, its C-
1-labeled species should give the same labeling pattern for
alanine and valine as 1,2-dihydroxy-[1-13C1]propyl-
phosphonic acid. To test this hypothesis, 13C-labeled hydroxy-
acetone was prepared after optimizing the reaction sequence
in the unlabeled series (Scheme 4).
1
was too impure by H NMR spectroscopy (600.13 MHz). It
had to be purified a second time, but there were still impuri-
ties present. Furthermore, reversed-phase HPLC additional-
ly allowed the isolation of a mixture of N-acetylvaline and
N-acetylmethionine. Finally, the cells grown on the C-1 13C-
labeled 1,2-dihydroxypropylphosphonic acid were worked-
up and gave N-acetylalanine (0.4 mg) and a mixture of N-
acetylvaline and N-acetylmethionine (0.3 mg, ratio 1:0.26 by
1H NMR) in sufficient purity for investigation by NMR
spectroscopy (1H and 13C NMR at 600.13 and 150.90 MHz,
respectively) in D2O. For comparison, commercially avail-
able reference samples were used for HPLC and NMR spec-
troscopy. The alanine was labeled at C-1 (52%, from the
height of the signals due to 3-H at d=1.32 ppm, doublet
with J=7.3 Hz, flanked by satellite doublets with JACTHNUTRGNEUNG
(13C,H) =
4.3 Hz). The 2-H signal (d=4.18 ppm) was composed of a
quartet (J=7.3 Hz) due to the unlabeled species, overlap-
ping with a doublet of quartets (J=7.3 and 5.0 Hz) due to
the 13C-labeled species. On 13C decoupling, the 2-H signal
collapsed to a quartet with J=7.3 Hz. The C-1 resonance
Scheme 4. Synthesis of labeled and unlabeled hydroxyacetone.
Chem. Eur. J. 2011, 17, 13341 – 13348
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13343

F. Hammerschmidt, J. P. Quinn et al.
Isopropenylmagnesium bromide was treated with
paraformaldehyde to give 2-methyl-2-propenol (16), which
ACHTUNGTRENNUNG
was not purified because of its low boiling point, but directly
esterified with 4-nitrobenzoyl chloride/pyridine to give p-ni-
trobenzoate 17 in 68% yield. The 3,5-dinitrobenzoyl group
was also tested as a protective group to facilitate the han-
dling of this small molecule, but was found to be less well
suited for the requisite steps than the 4-nitrobenzoyl group.
The ester 17 was ozonolyzed and then treated with Ph3P as
a reductant to yield the protected hydroxyacetone 18 as a
crystalline product in 87% yield. It was transesterified with
NaOMe in MeOH, and the hydroxyacetone (5) formed was
obtained in 53% yield after purification by flash chromato-
ACHTUNGTRENNUNG
graphy and distillation. The 13C-labeled isotopomer [1-13C1]5
was prepared by the same steps in similar yield.
Cells of R. huakuii PMY1 were grown on minimal salts
medium supplemented with labeled hydroxyacetone (5 mm)
as the sole source of carbon and energy; 1 mmolLÀ1 of
KH2PO4 was added as a phosphorus source. The cells were
worked-up as before and N-acetylalanine and a mixture of
N-acetylvaline and N-acetylmethionine (ratio 1:0.33) were
isolated by reversed-phase HPLC. Alanine and valine were
found to be 13C-labeled at C-1, 52 and 65%, respectively.
Methionine was not labeled by [1-13C1]5. These data are
taken as proof that hydroxyacetone is an obligatory inter-
mediate in the biodegradation of fosfomycin.
À
Scheme 5. Mechanisms for cleavage of P C bonds.
phite and an alcohol 21, which would also be compatible
with hydrolytic cleavage of a P C bond. From a mechanistic
À
À
point of view, cleavage of the P C bond in (R)-1-hydroxy-2-
oxopropylphosphonate could have some precedence in the
biodegradation of 2-aminoethylphosphonic acid (AEP)
(Scheme 5, route B). It is first converted by a transaminase
to phosphonoacetaldehyde (22) and then by phosphonatase
to acetaldehyde and orthophosphate.[17] The latter enzyme
was found to have an essential lysine in the active site. It
forms a protonated Schiff base (imine) 23 with the substrate,
having an electron sink at the b-carbon atom (relative to
Two pathways have been proposed for the degradation of
hydroxyacetone. The first involves oxidation via methyl-
glyoxal to pyruvate, and is catalyzed by acetol dehydrogen-
ase [EC 1.1.1.-] and methylglyoxal dehydrogenase [EC
1.2.1.23].[15] In the second pathway, C C cleavage of
À
hydroxy
G
acetate and a C1-compound, possibly formaldehyde.[16] The
results with the 13C-labeled compound showed that the first
pathway must be operative in R. huakuii PMY1. Further-
more, acetol dehydrogenase and methylglyoxal dehydrogen-
ase activities of 6.2 U per mg of cell-extract protein and
6.7 U per mg of cell-extract protein, respectively, were de-
tected in extracts of R. huakuii PMY1 cells grown on fosfo-
mycin (5 mm) as a phosphorus, carbon, and energy source.
À
phosphorus), and accommodates the electron pair of the P
C bond when an enzyme nucleophile attacks phosphorus.
Subsequent hydrolysis of the resultant acetaldehyde en-
AHCTUNGTREGaNNUN mine 24 and the phosphoenzyme 25 gives acetaldehyde
(26) and Pi. A similar mechanism could be operative in the
cleavage of the P C bond in (R)-1-hydroxy-2-oxopropyl-
phosphonic acid, except that a hydroxyacetone enamine in-
termediate is involved, which is finally hydrolyzed to
À
No acetol monoACHTUNGTRENNUNGoxygenase activity could be detected.
As the growth medium does not contain sufficient
amounts of alanine and valine, but evidently sufficient me-
thionine, the former two have to be biosynthesized from
pyruvate formed in two steps from hydroxyacetone.[14] It is
transaminated from glutamate to give alanine. Two mole-
cules of pyruvate are condensed with loss of CO2 to form a-
acetolactic acid, which is converted in three steps to a-keto-
hydroxyACTHNGUTERaNNUG cetone. Alternatively, a metal-ion-complexed eno-
late could be an equally good leaving group to facilitate fis-
À
sion of the P C bond. More experiments are necessary to
clarify this point. These two mechanisms are reminiscent of
both the lysine and the Zn2+-dependent fructose-1,6-bi-
sphosphate aldolases of the glycolytic cycle. They cleave a
À
C C bond, with a protonated enamine and a metal-com-
A
plexed enolate of dihydroxyacetone phosphate, respectively,
acid 13C-labeled at C-1 is used, valine 13C-labeled at C-1 will
serving as the leaving group.[18] Cleavage of the P C bond in
À
result, in agreement with our results.
phosphonates without an electron sink at the b-carbon atom
Enzymatic cleavage of P C bonds in phosphonates 19
most likely involves radical intermediates.[19]
À
studied so far furnished orthophosphate and organic com-
À
À
pounds 20, with an H C bond in place of the P C bond
(Scheme 5, route A).[8] This formal hydrolytic process re-
duces the carbon atom and oxidizes the phosphorus atom.
No enzyme has been found that effects formation of phos-
13344
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13341 – 13348

Biodegradation of Fosfomycin in R. huakuii PMY1 Cells
FULL PAPER
Merck L-3000 diode-array detector and fractions were collected with an
Isco model Foxy fraction collector. Fractions containing N-acetylalanine
were pooled and purified by reinjection. Isocratic elution was performed
with a mixture of 0.02m HCO2H/MeOH, 4:1 (v/v). Pooled fractions con-
taining N-acetylmethionine and N-acetylvaline were further purified on a
Superspher C18 column of dimensions 250ꢃ4 mm. Isocratic elution was
performed with a mixture of 0.01m HCO2H/MeOH, 9:1 (v/v) at a flow
Conclusion
We have shown that the isolation of alanine and a mixture
of valine and methionine as N-acetyl derivatives by re-
versed-phase HPLC helped to unravel the biodegradation of
fosfomycin in R. huakuii PMY1. Feeding with racemic
rate of 0.8 mLminÀ1
.
(1R*,2R*)-1,2-dihydroxypropylphosphonic
acid
and
(Æ)-(2-Triisopropylsilyloxy)propanenitrile and (Æ)-2-triisopropylsilyloxy-
[1-13C1]propanenitrile ((Æ)-8 and (Æ)-[1-13C1]8): Acetaldehyde (0.159 g,
0.20 mL, 3.60 mmol) was added to a stirred mixture of NaCN (0.147 g,
3.0 mmol, powdered), imidazole (0.450 g, 6.60 mmol), TIPSCl (0.617 g,
0.69 mL, 3.20 mmol), and dry DMF (5.0 mL) at 48C under argon. After
stirring for 60 h at room temperature, water (20 mL) and Et2O (40 mL)
were added. The organic phase was removed and the aqueous phase was
extracted with further Et2O (20 mL). The combined organic layers were
washed with HCl (1m) and a saturated aqueous solution of NaHCO3,
dried (MgSO4), and concentrated under reduced pressure. The residue
was flash chromatographed (hexanes/CH2Cl2, 5:1, Rf =0.27) to give nitrile
(Æ)-8 (0.550 g, 81%) as a colorless oil. 1H NMR (400.13 MHz, CDCl3):
d=4.65 (q, J=6.7 Hz, 1H; CHCN), 1.58 (d, J=6.7 Hz, 3H; CH3), 1.22–
1.04 ppm (m, 21H; iPr); 13C NMR (100.61 MHz, CDCl3): d=120.7, 58.1,
23.3, 17.8 and 17.7 (6C), 11.9 ppm (3C); IR (NaCl): n˜ =2946, 2868, 2230,
1464, 1375, 1132, 1065 cmÀ1; elemental analysis calcd (%) for C12H25NOSi
(227.4): C 63.71, H 11.10; found: C 63.91, H 11.39.
Similarly, (Æ)-2-triisopropylsilyloxy-[1-13C1]propanenitrile ((Æ)-[1-13C1]8)
(5.2 g, 85%) was prepared from acetaldehyde (1.54 g, 1.96 mL,
34.84 mmol, 1.3 equiv) and Na13CN (1.34 g, 26.8 mmol, 99% 13C).
1H NMR (400.13 MHz, CDCl3): d=4.63 (dq, J=6.5, 5.0 Hz, 1H;
CHCN), 1.56 (dd, J=6.5, 6.1 Hz; CH3), 1.19–1.02 ppm (m, 21H; iPr);
13C NMR (100.61 MHz, CDCl3): d=120.7, 58.1 (d, J=61.2 Hz; CH13CN),
23.3, 17.8 and 17.7 (6C), 11.9 ppm (3C).
hydroxy
acetone, both 13C-labeled at C-1, resulted in the for-
mation of alanine and valine that were highly labeled at C-
1. These findings are taken as evidence that (1R,2R)-1,2-di-
hydroxypropylphosphonic acid is converted into (R)-1-hy-
droxy-2-ACHTUNGTRENNUNGoxopropylphosphonic acid, which is the substrate
À
for the enzyme cleaving the P C bond. The hydroxyacetone
formed is dehydrogenated to pyruvic acid via methylglyoxal.
The former is used in part for the biosynthesis of amino
acids. Therefore, our feeding experiments have also elucidat-
ed the biodegradation of hydroxyacetone by cells of R. hua-
kuii PMY1.
Experimental Section
General: 1H, 13C (in part J-modulated, not for 13C-labeled compounds),
and 31P NMR spectra were recorded from solutions in CDCl3 or D2O at
300 K on a Bruker DRX 400 (or Avance III 400) at 400.13, 100.61, and
161.98 MHz (or at 400.27, 100.64, and 162.03 MHz), respectively. Some
1H and 13C NMR spectra were recorded on a Bruker Avance 600 at
600.13 and 150.90 MHz, respectively. Chemical shifts were referenced to
residual CHCl3 (dH =7.24 ppm) or HDO (dH =4.80 ppm), CDCl3 (dC =
77.00 ppm), and external H3PO4 (85%, dP =0.00 ppm). IR spectra of
liquid samples were recorded from films on a silicon disc[20] or between
NaCl plates on a Perkin–Elmer 1600 FTIR spectrometer. IR spectra of
crystalline compounds were recorded from samples on an attenuated
total internal reflection (ATR) diamond on a Bruker Vertex 70 spectro-
AHCTUNGTRENNUNG
[1-13C1]9): (Æ)-2-(Triisopropylsilyloxy)propanal: Diisobutylaluminum hy-
dride (DIBAH) (1m solution in heptane, 6 mL, 6 mmol) was added drop-
wise to a stirred solution of (Æ)-2-(triisopropylsilyloxy)propanenitrile
((Æ)-8, 1.137 g, 5 mmol) in dry toluene (15 mL) at À788C under argon.
The reaction mixture was then allowed to warm to 108C over a period of
16 h. Et2O (30 mL), a saturated aqueous solution of NH4Cl (30 mL), and
H2SO4 (20 mL, 1.6m) were added and the mixture was stirred vigorously
for 0.5 h at ambient temperature. The organic phase was separated and
the aqueous phase was extracted with Et2O (3ꢃ30 mL). The combined
organic phases were washed with water, dried (MgSO4), and concentrat-
ed under reduced pressure. The crude product was dried by co-distilla-
tion with toluene and then maintained in vacuo (0.50 mbar) for 1 h to
yield the crude aldehyde (1.049 g) as an oil, which was immediately used
for the next step. Diethyl trimethylsilyl phosphite (1.37 mL, 1.265 g,
6.0 mmol) was added dropwise to a stirred solution of the above alde-
hyde (1.049 g) in dry toluene (20 mL) at À788C under argon. The reac-
tion mixture was then allowed to warm to 108C over a period of 16 h.
Volatile components were removed in a rotary evaporator at 0.5 mbar
for 1 h to yield 2.56 g of the crude phosphonates as an oil, which was re-
dissolved in dry ethanol (30 mL). Concentrated HCl (3 drops) was added
and the mixture was stirred for 1 h (TLC: hexanes/acetone, 3:1, Rf =0.38
for (Æ)-9 and 0.30 for (Æ)-10). The solution was concentrated under re-
duced pressure and the residue was redissolved in water (10 mL). The
aqueous layer was extracted with EtOAc (3ꢃ15 mL). The combined or-
ganic phases were dried (MgSO4) and concentrated under reduced pres-
sure. The residue ((Æ)-9/(Æ)-10, 7:1 by 31P NMR) was purified by flash
chromatography (hexanes/acetone, 5:1, Rf =0.28 for (Æ)-9 and 0.20 for
(Æ)-10) to give (Æ)-9 (1.172 g, 53% starting from nitrile) and (Æ)-10
(0.155 g, 7%, containing 7% of (Æ)-9) as colorless oils.
ACHTUNGTRENNUNGmeter. FID and HR-ESI mass spectra were recorded on a MAT 95S and
a MAT 900S (Finnigan MAT), respectively. TLC was carried out on
0.25 mm thick Merck plates of silica gel 60 F254. Spots were visualized by
UV and/or by dipping the plate into a solution of (NH4)6Mo7O24·4H2O
(23.0 g) and CeACHTUNGTRENNUNG(SO4)2·4H2O (1.0 g) in 10% aqueous H2SO4 (500 mL),
followed by heating with a heat gun. Flash (column) chromatography
was performed with Merck silica gel 60 (230–400 mesh). DMF and pyri-
dine were dried by refluxing over powdered CaH2, then distillation and
storage over molecular sieves (4 ꢄ). Dichloromethane and 1,2-dichloro-
ethane were dried by passing through aluminum oxide 90 active neutral
(0.063–0.200 mm, activity I) and stored over molecular sieves (3 and 4 ꢄ,
respectively). Hexane was dried by storage over molecular sieves (4 ꢄ).
Et2O was refluxed over LiAlH4, THF over potassium and distilled prior
to use. Melting points were determined on a Reichert Thermovar instru-
ment and are uncorrected.
Hydrolysis of cells and chromatography: Three aliquots of 2.0 mL of cell
suspension (containing approximately 0.5–0.6 g of dry matter) were
mixed with 6m HCl (6 mL, containing 2.0% (w/v) phenol) and the mix-
ture was heated at 958C for 8 h. After cooling, the hydrolysate was fil-
tered through filter paper and concentrated in a rotary evaporator at
408C. The solution was diluted with water (3.0 mL) and then freeze-
dried. The residue was redissolved in 3m aqueous KOH (1.0 mL) and
then Ac2O (3.0 mL) was added. After 2 h, the solution was diluted with
water to a volume of 40 mL and freeze-dried. The freeze-dried sample
was redissolved in water (0.8 mL) and injected onto a Kromasil C18
column of dimensions 250ꢃ10 mm. A linear gradient of 0.02m HCO2H
(eluent A) and MeOH/H2O, 8:2 (v/v) (eluent B) was applied. The eluent
composition was 10% B at the start of the gradient, 45% B at 12.0 min,
60% B at 13.0 min, and 100% B at 22.0 min. The flow rate was set at
5.0 mLminÀ1. Acetylated amino acids were detected by means of a
(Æ)-9: 1H NMR (400.27 MHz, CDCl3): d=4.32 (quintd, J=7.3, 6.2 Hz,
1H; PCCH), 4.22–4.12 (m, 4H; 2 ꢃ OCH2), 3.61 (td, J=6.2, 4.9 Hz, 1H;
PCH), 2.95 (dd, J=16.6, 4.9 Hz, 1H; OH), 1.35 (d, J=6.2 Hz, 3H; CH3),
1.33 (t, J=7.1 Hz; OCCH3), 1.32 (t, J=7.1 Hz, 3H; OCCH3), 1.14–
Chem. Eur. J. 2011, 17, 13341 – 13348
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13345

F. Hammerschmidt, J. P. Quinn et al.
1.04 ppm (m, 21H; 3 ꢃ iPr); 13C NMR (100.65 MHz, CDCl3): d=72.9 (d,
J=162.9 Hz), 68.3 (d, J=5.9 Hz), 62.6 (d, J=6.6 Hz), 62.3 (d, J=6.6 Hz),
21.5 (d, J=5.9 Hz), 18.11 (3C), 18.02 (3C), 16.5 (d, J=5.8 Hz, 2C),
12.7 ppm (3C); 31P NMR (161.98 MHz, CDCl3): d=22.2 ppm; IR (Si):
n˜ =3303, 2943, 2867, 1465, 1275, 1258, 1054, 1029, 971 cmÀ1; elemental
analysis calcd (%) for C16H37O5PSi (368.52): C 52.15, H 10.12; found: C
52.43, H 10.10.
66.5 (dd, J=36.9, 2.9 Hz; CHCP), 63.3 (d, J=6.8 Hz), 62.6 (d, J=
6.8 Hz), 19.1 (d, J=10.7 Hz; CH3), 16.42 (d, J=4.9 Hz; CH3), 16.39 ppm
(d, J=5.8 Hz; CH3); 31P NMR (161.98 MHz, CDCl3): d=24.8 ppm (d,
J=158.5 Hz; 13CHP); MS (FID): m/z (%): 213 [M+] (2.5), 214 (100).
A
1,2-dihydroxy-[1-13C1]propylphosphonic acid ((Æ)-3 and (Æ)-[1-13C1]3):
(Æ)-3:
1
(Æ)-10: H NMR (400.27 MHz, CDCl3): d=4.30 (qdd, J=6.3, 5.8, 2.8 Hz,
(Æ)-11 (0.368 g, 1.73 mmol), allylTMS (0.985 g, 1.38 mL, 8.65 mmol,
5 equiv), and bromotrimethylsilane (3.337 g, 2.88 mL, 21.8 mmol) in dry
1,2-C2H4Cl2 (8.0 mL) was kept under argon for 16 h at room temperature.
Volatile components were then removed in vacuo (0.50 mbar, ambient to
408C) and the residue was dissolved in a 1:1 mixture of EtOH/H2O
(10 mL). The solvent was removed under reduced pressure and the resi-
due was taken up in water (10 mL) and concentrated NH3 (1.0 mL) and
then lyophilized to leave a gummy residue of the ammonium salt of (Æ)-
3 (0.30 g, containing an unknown amount of ammonia and water; the
yield was assumed to be 95%). The spectroscopic data (1 H, 13C, and 31P)
were identical to those of the ammonium salt of the (1R,2R)-dihydroxy-
propylphosphonic acid derived from fosfomycin. If desired, the ammoni-
um salt can be purified by the method used for the conversion of the cy-
clohexylammonium salt of 1,2-dihydroxypropylphosphonic acid to the
ammonium salt.[10]
1H; PCCH), 4.21–4.10 (m, 4H; 2 ꢃ OCH2), 4.01 (dd, J=10.8, 2.8 Hz,
1H; PCH), 2.35 (brs, 1H; OH), 1.32 (d, J=6.3 Hz, 3H; CH3), 1.31 (t, J=
7.1 Hz, 6H; 2 ꢃ OCCH3), 1.10–1.00 ppm (m, 21H; 3 ꢃ iPr); 13C NMR
(100.65 MHz, CDCl3): d=72.9 (d, J=167.9 Hz), 68.6 (d, J=9.0 Hz), 62.7
(d, J=7.0 Hz), 62.3 (d, J=6.0 Hz), 18.4 (erroneously reported[10] to be a
doublet at 17.7), 18.3 (3C), 18.2 (3C), 16.67 (d, J=5.9 Hz), 16.64 (d, J=
5.9 Hz), 12.5 ppm (3C); 31P NMR (161.98 MHz, CDCl3): d=21.0 ppm;
IR (Si): n˜ =3306, 2943, 2867, 1465, 1242, 1258, 1141, 1096, 1054, 1028,
972 cmÀ1
;
HRMS (ESI): m/z: calcd for C16H37O5PSiNa [M+Na]+:
391.2046; found: 391.2052.
(Æ)-[1-13C1]9 and (Æ)-[1-13C1]10: In analogy to the preparation of the un-
labeled species, (Æ)-2-(triisopropylsilyloxy)propane-[1-13C1]nitrile ((Æ)-
[1-13C1]8), 1.60 g, 7.0 mmol) was converted into the crude aldehyde
(1.37 g), which was immediately transformed into labeled hydroxy-
phosphonates (Æ)-[1-13C1]9 (1.242 g, 48%, based on the starting nitrile)
and (Æ)-[1-13C1]10 (0.155 g, 6%, containing 16 mol% diethyl phosphite),
both with 99% 13C at C-1.
(Æ)-[1-13C1]3·
(NH3)x: The labeled 1,2-dihydroxypropylphosphonate (Æ)-
E
[1-13C1]11 (0.337 g, 1.59 mmol) was converted into the ammonium salt of
(Æ)-[1-13C1]3 following the procedure used for the unlabeled species.
1H NMR (400.13 MHz, D2O): d=4.05 (m, 1H; CHCP), 3.50 (ddd, J=
138.0, 9.5, 4.5 Hz, 1H; 13CHP), 1.31 ppm (t, J=5.5 Hz, 3H; CH3);
13C NMR (100.61 MHz, D2O): d=73.3 (d, J=141.5 Hz; 13CHP), 68.3 (d,
J=39.8 Hz; CCP), 18.8 ppm (d, J=8.4 Hz; CH3); 31P NMR (161.97 MHz,
D2O): d=17.0 ppm (d, J=141.7 Hz).
(Æ)-
ACHTUNGTRENNUNG
[1-13C1]9: 1H NMR (400.13 MHz, CDCl3): d=4.30 (m, 1H; CHCP),
4.15 (sext, J=7.0 Hz, 4H; 2 ꢃ OCH2), 3.59 (td, J=142.6, 6.0 Hz, 1H;
13CHP), 2.78 (brs, 1H; OH), 1.35 (t, J=6.0 Hz, 3H), 1.31 (t, J=7.0 Hz,
3H; OCCH3), 1.30 (t, J=7.0 Hz, 3H; OCCH3), 1.18–1.00 ppm (m, 21H;
3
ꢃ
iPr); 13C NMR (100.61 MHz, CDCl3): d=72.8 (d, J=162.3 Hz;
13CHP), 68.3 (dd, J=37.9, 5.8 Hz; C13CP), 62.6 (d, J=6.8 Hz; OCH2),
62.2 (d, J=6.8 Hz; OCH2), 21.4 (dd, J=5.8, 1.9 Hz; CH3), 18.09 and
18.00 (2s, 6C), 16.4 (d, J=5.8 Hz, 2C), 12.7 ppm (3C); IR (Si): n˜ =3303,
2-Methyl-2-propenyl 4-nitrobenzoate and 2-methyl-2-[1-13C1]propenyl 4-
nitrobenzoate (17 and [1-13C1]17): A solution of isopropenylmagnesium
bromide (0.5m in THF, 20 mL, 10.0 mmol) was added to a suspension of
paraformaldehyde (0.150 g, 5.0 mmol) in dry Et2O (2 mL) at À208C
under argon. The mixture was stirred overnight at room temperature.
After careful addition of H2O (10 mL) and dissolution of the precipitate
with dilute H2SO4, the organic phase was separated and the aqueous
phase was extracted with CH2Cl2 (2ꢃ10 mL). The combined organic solu-
tions were dried (Na2SO4) and filtered (the Na2SO4 was washed with dry
CH2Cl2), and most of the CH2Cl2 was distilled off at atmospheric pressure
over a short column. The resulting THF solution of the alcohol was
stirred at room temperature under argon for 30 min with molecular
sieves (3.5 g, 4 ꢄ). The mixture was then cooled to 08C, whereupon 4-ni-
trobenzoyl chloride (1.45 g, 7.81 mmol), dry pyridine (2.6 mL), and
DMAP (52 mg) were successively added. After stirring overnight at
room temperature, the solution was decanted from the molecular sieves
(which were rinsed with CH2Cl2), and the combined solutions were treat-
ed with H2O (30 mL) and concentrated HCl (about 4.5 mL, to dissolve
the precipitate). The organic layer was separated and the aqueous phase
was extracted twice with CH2Cl2. The combined organic phases were
dried (MgSO4) and concentrated under reduced pressure. The residue
was purified by flash chromatography (hexane/CH2Cl2, 1:1, Rf =0.50 for
hexane/EtOAc, 10:1; 0.55 for hexane/CH2Cl2, 1:1) to yield 4-nitroben-
zoate 17 (0.686 g, 62%) as yellowish crystals; m.p. 68–708C (hexane/a
few drops of 1,2-C2H4Cl2; lit.:[21] no data). (400.13 MHz, CDCl3): d=8.25
(AA’BB’ system, 4H; Har), 5.07 (brs, 1H; =CH), 5.00 (brs, 1H; =CH),
4.77 (brs, 2H; OCH2), 1.83 ppm (brs, 3H; CH3); (100.61 MHz, CDCl3):
d=164.33, 150.59, 139.31, 135.56, 130.72 (2C), 123.56 (2C), 113.76, 69.03,
19.55 ppm; IR (ATR): n˜ =3117, 1710, 1520, 1268, 1119 cmÀ1; elemental
analysis calcd (%) for C11H11NO4 (221.2): C 59.73, H 5.01, N 6.33; found:
C 59.85, H 4.79, N 6.28.
2943, 2867, 1465, 1275, 1258, 1054, 1029, 971 cmÀ1
(Æ)-
.
ACHTUNGTRENNUNG
[1-13C1]10: 1H NMR (400.13 MHz, CDCl3): d=4.30 (m, 1H; CHCP),
4.21–4.10 (m, 4H; 2 ꢃ OCH2), 4.01 (ddd, J=140.5, 10.5, 3.0 Hz, 1H;
13CHP), 2.55 (brs, 1H; OH), 1.32 (d, J=7.0 Hz, 3H; CH3), 1.31 (t, J=
7.0 Hz, 6H; 2 ꢃ OCCH3), 1.12–0.98 ppm (m, 21H; 3 ꢃ iPr); 13C NMR
(100.61 MHz, CDCl3): d=72.9 (d, J=161.7 Hz; 13CHP), 68.6 (dd, J=
37.9, 9.3 Hz; CCP), 62.7 (d, J=6.9 Hz), 62.3 (d, J=6.1 Hz), 18.1, 18.02,
and 17.98 (2s, 6C), 16.42 (d, J=5.4 Hz), 16.40 (d, J=5.4 Hz), 12.3 ppm
(3C); 31P NMR (161.98 MHz, CDCl3): d=22.3 ppm (d, J=161.7 Hz;
13CHP); MS (FID): m/z (%): 369 [M+] (1), 370 (100).
A
(Æ)-diethyl 1,2-dihydroxy-[1-13C1]propylphosphonate ((Æ)-11 and (Æ)-
[1-13C1]11): A solution of TIPS-protected diol (Æ)-9 (0.654 g, 1.77 mmol)
in dry CH3CN (15 mL), water (7 drops), and HF (40%, 10 drops) was
left at room temperature in a small plastic bottle until the starting mate-
rial had been consumed (22 h) and then concentrated under reduced
pressure in a round-bottomed glass flask. The residue was flash-chroma-
tographed (EtOAc/MeOH, 20:1, Rf =0.14) to give diol (Æ)-11 (0.368 g,
98%) as a viscous oil. 1H NMR (400.27 MHz, CDCl3): d=4.27–4.08 (m,
5H; 2 ꢃ OCH2, PCCH), 3.65 (brt, J=6.5 Hz, 1H; CHP), 3.35–3.00 (two
broad overlapping s, 2H; OH), 1.34 (t, J=7.0 Hz, 3H; CH3), 1.33 (t, J=
7.0 Hz, 3H; CH3), 1.29 ppm (dd, J=6.2, 1.5 Hz, 3H; CH3); 13C NMR
(100.65 MHz, CDCl3): d=71.4 (d, J=159.1 Hz), 66.4, 63.46 (d, J=
7.3 Hz), 62.49 (d, J=7.3 Hz), 19.0 (d, J=11.7 Hz), 16.47 (d, J=5.1 Hz),
16.44 ppm (d, J=5.1 Hz); 31P NMR (162.03 MHz, CDCl3): d=23.3 ppm;
IR (Si): n˜ =3317, 2979, 2940, 2867, 1462, 1391, 1275, 1053, 1029, 971 cmÀ1
;
elemental analysis calcd (%) for C7H17O5P (212.18): C 39.62, H 8.08;
found: C 39.47, H 7.83.
Similarly, 13C-labeled paraformaldehyde (0.162 g, 5.23 mmol) gave
(Æ)-[1-13C1]11: Similarly, TIPS-protected diol (Æ)-[1-13C1]9 (1.12 g,
3.03 mmol) was converted into diol (Æ)-[1-13C1]11 (0.592 g, 92%).
1H NMR (400.13 MHz, CDCl3): d=4.16 (m, 5H; 2 ꢃ OCH2, PCCH),
3.90 (ddd, J=9.0, 2.5, 2.0 Hz, 1H; OH), 3.64 (tdd, J=142.0, 9.0, 3.0 Hz,
1H; 13CHP), 3.54 (d, J=4.5 Hz, 1H; OH), 1.32 (t, J=7.0 Hz, 3H; CH3),
1.31 (t, J=7.0 Hz, 3H; CH3), 1.27 ppm (ddd, J=5.0, 4.5, 1.5 Hz, 3H;
CH3); 13C NMR (100.61 MHz, CDCl3): d=71.8 (d, J=159.4 Hz; 13CHP),
[1-13C1]17 (0.789 g, 68%), 99% C-13 labeled, m.p. 62–638C (hexane).
1
N
Har), 5.06 (d, J=6.8 Hz, 1H; =CH), 4.99 (d, J=12.13 Hz, 1H; =CH), 4.76
(d, J=147.8 Hz, 2H; O13CH2), 1.82 ppm (d, J=4.29 Hz, 3H; Me);
13C NMR (100.61 MHz, CDCl3): d=164.33 (d, J=2.3 Hz; CO), 150.59
(Car), 139.30 (d, J=46.0 Hz; C=), 135.56 (Car), 130.73 (2 ꢃ CHar), 123.56
13346
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13341 – 13348

Biodegradation of Fosfomycin in R. huakuii PMY1 Cells
FULL PAPER
(2 ꢃ CHar), 113.75 (d, J=3.1 Hz; CH2=), 69.03 (13CH2), 19.55 ppm (d, J=
5.4 Hz; Me).
lution.[9] Either filter-sterilized (0.22 mm) fosfomycin (5 mm) or the re-
spective organophosphonate under test as a potential pathway intermedi-
ate was added as the sole carbon and phosphorus source. Microbial
growth was measured by the increase in absorbance at 650 nm using an
ATI-Unicam PU 8625 UV/Vis spectrophotometer, while phosphate re-
lease into the culture supernatant was monitored according to the
method of Fiske and SubbaRow.[22]
Feeding of (Æ)-[1-13C1]3: Culture (500 mL, in a 2 L flask) with (Æ)-
[1-13C1]3 (7.5 mm) was shaken for 11 days; the release of Pi amounted to
2.2 mmol. The cells were harvested by centrifugation, then sterilized and
hydrolyzed.
2-Oxo-propyl 4-nitrobenzoate and 2-oxo-[1-13C1]propyl 4-nitrobenzoate
(18 and [1-13C1]18): 4-Nitrobenzoate 17 (0.665 g, 3.0 mmol) was dissolved
in a mixture of dry CH2Cl2 (15 mL) and dry MeOH (15 mL), and the so-
lution was cooled to À788C and subjected to ozonolysis. Thereafter, tri-
phenylphosphane (3.1 mmol) was added to the cold solution, which was
then stirred overnight and allowed to warm to room temperature. The
solvent was evaporated under reduced pressure and the crude product
was purified by flash chromatography (hexane/EtOAc, 5:1, then 1:1; Rf =
0.61 for 1:1) to yield 18 (0.647 g, 96%) as yellowish crystals; m.p. 96–
988C (hexane/1,2-C2H4Cl2).
Enzyme assays: Cell extracts of fosfomycin-grown cells were prepared by
sonication as previously described.[9] Both acetol dehydrogenase and
methylglyoxal dehydrogenase were assayed according to the method of
Taylor et al.[15] by measuring the production of NADH at 340 nm. The
assay mixture (1 mL) consisted of glycine-NaOH buffer pH 10 (90 mmol),
NAD (0.5 mmol), and either acetol (2 mmol) or methylglyoxal (2 mmol) as
substrate. The reaction was initiated by the addition of 20 mL of enzyme
solution (containing 1–1.5 mgmLÀ1 of cell-extract protein) and the mix-
ture was incubated for 30 min at 208C. Acetol monooxygenase was as-
sayed by the method of Hartmans and de Bont[16] through measurement
of NADPH consumption at 340 nm. The reaction mixture (1 mL) consist-
ed of potassium phosphate buffer (0.3 mmol) pH 8.0, NADPH
(0.8 mmol), and acetol (8 mmol). The reaction was initiated as described
above. One enzyme unit is defined as the amount leading to either the
consumption of 1 nmol of substrate or the formation of 1 nmol of product
per min at 208C. For all enzyme assays, controls containing no crude cell
extract or substrate were prepared to evaluate the degree of background
activity. Protein concentration was determined by the method of Brad-
ford using bovine serum albumin as standard.[23]
19: 1H NMR (400.13 MHz, CDCl3): d=8.27 (AA’BB’ system, 4H; Har),
4.93 (s, 2H; OCH2), 2.23 ppm (s, 3H; CH3); 13C NMR (100.61 MHz,
CDCl3): d=200.23 (CO), 163.98 (CO2), 150.81 (Car), 134.59 (Car), 131.03
(2CHar), 123.62 (2CHar), 69.14 (OCH2), 26.07 ppm (CH3); IR (ATR): n˜ =
1743, 1720, 1617, 1347, 1272, 1103 cmÀ1; elemental analysis calcd (%) for
C10H9NO5 (223.2): C 53.82, H 4.06, N 6.28; found: C 53.72, H 3.90, N
6.20.
Similarly, 4-nitrobenzoate [1-13C1]17 (2.39 g, 10.80 mmol) was converted
into [1-13C1]18 (2.206 g, 91%); m.p. 958C (hexane/CH2Cl2), 98% 13C at
C-1. 1H NMR (400.13 MHz, CDCl3): d=8.25 (AA’BB’ system, 4H; Har),
4.93 (d, J=147.3 Hz, 2H; O13CH2), 2.22 ppm (d, J=2.3 Hz, 3H; Me);
13C NMR (100.61 MHz, CDCl3): d=200.24 (d, J=40.6 Hz; CO), 163.96
(d, J=2.3 Hz; CO2), 150.78 (Car), 135.00 (d, J=1.3 Hz; Car), 131.01 (2 ꢃ
CHar), 123.60 (2 ꢃ CHar), 69.12 (O13CH2), 26.04 ppm (d, J=18.4 Hz; Me).
Hydroxyacetone and 1-hydroxy-[1-13C1]acetone (5 and [1-13C1]5): Ester
18 (2.01 g, 9.0 mmol) was dissolved in dry CH2Cl2 (28 mL) and the solu-
tion was cooled to À208C under argon. Then, a solution of NaOMe in
MeOH, which was prepared by dissolving Na (70 mg) in dry MeOH
(2.8 mL), was added. After stirring for exactly 20 min at À208C, small
pieces of dry ice (about 0.4 g) were added to neutralize the solution,
which was immediately applied to a flash chromatography column (about
55 g of silica gel, 2ꢃ37 cm, CH2Cl2 as eluent). Fractions containing the
product (TLC: hexane/EtOAc, 1:1, Rf =0.26) were collected (150–
200 mL) and the solvent was evaporated under reduced pressure (rotary
evaporator, flask not in the water bath) until only a small amount was
left. The residue was transferred to a small flask using Et2O and bulb-to-
bulb distilled (90–100 mmHg/1208C final bath temperature) to yield hy-
droxyacetone (5) (0.347 g, 52%) as a liquid containing a small amount of
Acknowledgements
We gratefully acknowledge financial support by the Austrian Science
Fund FWF (project no. P10566-CHE). Dr. S. Prechelmacher and Dr. B.
Peric Simov are thanked for the preparation of 1,2-dihydroxy-
[1-13C1]propylphosphonic acid.
1
impurities. The H and 13C NMR spectra were identical to those of an au-
[1] For a review, see: H. Seto, T. Kuzuyama, Nat. Prod. Rep. 1999, 16,
589–596.
[2] F. M. Kahan, J. S. Kahan, P. J. Cassidy, H. Kropp, Ann. N. Y. Acad.
Sci. 1974, 235, 364–386.
[3] For a review, see: W. W. Metcalf, W. A. van der Donk, Ann. Rev.
Biochem. 2009, 78, 65–94.
[4] a) H. M. Seidel, S. Freeman, C. H. Schwalbe, J. R. Knowles, J. Am.
Chem. Soc. 1990, 112, 8149–8155; b) M. S. McQueney, S.-L. Lee,
W. H. Swartz, H. L. Ammon, P. S. Mariano, D. Dunaway-Mariano, J.
Org. Chem. 1991, 56, 7121–7130.
[5] F. Hammerschmidt, J. Chem. Soc. Perkin Trans. 1 1991, 1993–1996.
[6] R. D. Woodyer, G. Li, H. Zhao, W. A. van der Donk, Chem.
Commun. 2007, 359–361.
[7] a) L. J. Higgins, F. Yan, P. Liu, H.-w. Liu, C. L. Drennan, Nature
2005, 437, 838–844; b) K. McLuskey, S. Cameron, F. Hammersch-
midt, W. N. Hunter, Proc. Natl. Acad. Sci. USA 2005, 102, 14221–
14226.
[8] a) A. K. White, W. W. Metcalf, Ann. Rev. Microbiol. 2007, 61, 379–
400; b) J. P. Quinn, A. N. Kulakova, N. A. Cooley, J. W. McGrath,
Environ. Microbiol. 2007, 9, 2392–2400; c) N. G. Ternan, J. W.
McGrath, G. McMullan, J. P. Quinn, World J. Microbiol. Biotechnol.
1998, 14, 635–647.
thentic sample.
Analogously, labeled ester [1-13C1]17 (2.18 g, 9.72 mmol) was converted
into labeled hydroxyacetone [1-13C1]5 (0.430 g, 59%), 97% C-1 labeled.
1H NMR (400.13 MHz, CDCl3): d=4.19 (d, J=131.4 Hz; O13CH2), 3.07
(brs, 1H; OH), 2.11 ppm (d, J=2.3 Hz, 3H; Me); 13C NMR
(100.61 MHz, CDCl3): d=207.1 (d, J=37.6 Hz; CO), 68.6 (O13CH2),
25.2 ppm (d, J=16.1 Hz; CH3).
N-Acetylalanine: 1H NMR (400.13 MHz, D2O): d=4.28 (q, J=7.5 Hz,
1H; CHCO2), 1.96 (s, 3H; CH3CO), 1.35 ppm (d, J=7.5 Hz, 3H; CH3; J-
ACHTUNGTRENNUNG
(13C,H)=129.5 Hz); 13C NMR (100.61 MHz, D2O): d=177.2, 174.4, 49.1,
21.9, 16.4 ppm; HCO2H: 8.2 ppm.
N-Acetylvaline: 1H NMR (400.13 MHz, D2O): d=4.27 (d, J=6.0 Hz,
1H; CHCO2), 2.23 (ꢀoct, J=6.5 Hz, 1H; CH), 2.10 (s, 3H; CH3), 1.01
(d, J=6.5 Hz, 3H; CH3), 1.00 ppm (d, J=6.5 Hz, 3H; CH3); 13C NMR
(100.61 MHz, D2O): d=176.1, 174.9, 59.0, 30.1, 21.9, 18.6, 17.5 ppm.
N-Acetylmethionine: 1H NMR (600.13 MHz, D2O): d=4.55 (dd, J=9.6,
4.5 Hz, 1H; CHCO2), 2.71–2.58 (m, 2H; SCH2), 2.27–2.19 (m, 1H; CH),
2.14 (s; SCH3 or CH3CO; JACHTNUTGRNEUNG
(13C,H)=136.9 Hz), 2.09–2.03 (m, 1H; CH),
2.08 ppm (s; SCH3 or CH3CO); 13C NMR (100.61 MHz, D2O): d=175.5,
174.4, 51.7, 29.4, 21.6, 14.1 ppm.
Microorganism and culture conditions: Cells of Rhizobium huakuii
PMY1[9] were grown in batch cultures at 308C on an orbital shaker at
100 rpm in minimal salt medium (pH 7.0) of the following composition
(per liter): KCl, 0.20 g; MgSO4·7H2O, 0.20 g; CaCl2·2H2O, 1.0 mg;
NH4Cl, 1.0 g; ferric ammonium citrate, 1.0 mg; phosphate-free yeast ex-
tract (0.05 g),[9] and 1 mL each of trace element solution and vitamin so-
[9] J. W. McGrath, F. Hammerschmidt, J. P. Quinn, Appl. Environ. Mi-
crobiol. 1998, 64, 356–358.
[10] J. W. McGrath, F. Hammerschmidt, W. Preusser, J. P. Quinn, A.
Schweifer, Org. Biomol. Chem. 2009, 7, 1944–1953.
[11] E. Y. Lau, Z. E. Newby, T. C. Bruice, J. Am. Chem. Soc. 2001, 123,
3350–3357.
Chem. Eur. J. 2011, 17, 13341 – 13348
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
13347

F. Hammerschmidt, J. P. Quinn et al.
[12] K. L. Fillgrove, S. Pakhomova, M. E. Newcomer, R. N. Armstrong,
J. Am. Chem. Soc. 2003, 125, 15730–15731.
[18] K. Faber, Biotransformations in Organic Chemistry, 5th ed. Springer,
Berlin, 2004, p. 273.
[13] G. Cainelli, M. Panunzio, E. Bandini, G. Martellim, G. Spunta, Tet-
rahedron 1996, 52, 1685–1698.
[19] J. W. Frost, S. Loo, M. L. Cordeiro, D. Li, J. Am. Chem. Soc. 1987,
109, 2166–2171.
[14] A. L. Lehninger, Biochemistry, 2nd. ed., Worth Publishers, New
York, 1978.
[15] D. G. Taylor, P. W. Trudgill, R. E. Cripps, P. R. Harris, J. Gen. Micro-
biol. 1980, 118, 159–170.
[20] W. Mikenda, Vib. Spectrosc. 1992, 3, 327–330.
[21] C. R. Schmid, Tetrahedron Lett. 1992, 33, 757–760.
[22] C. H. Fiske, Y. Subbarow, J. Biol. Chem. 1925, 66, 375–400.
[23] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
[16] S. Hartmans, J. A. M. de Bont, FEMS Microbiol. Lett. 1986, 36, 155–
158.
[17] a) A.-M. Lacoste, C. Dumora, A. Cassaigne, Life Sci. Adv.: Bio-
chem. 1989, 8, 97–111; b) S.-L. Lee, T. W. Hepburn, W. H. Swartz,
H. L. Ammon, P. S. Mariano, D. Dunaway-Mariano, J. Am. Chem.
Soc. 1992, 114, 7346–7354.
Received: March 9, 2011
Revised: August 22, 2011
Published online: October 20, 2011
13348
www.chemeurj.org
ꢂ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 13341 – 13348