The origin of the β-lactam carbons of clavulanic acid

Full text of DOI:10.1039/a701934c

View Article Online / Journal Homepage / Table of Contents for this issue
The origin of the b-lactam carbons of clavulanic acid
Jan E. Thirkettle,a Jack E. Baldwin,a Jeff Edwards,b John P. Griffinb and Christopher J. Schofield*a
a Dyson Perrins Laboratory, South Parks Road, Oxford, UK OX1 3QY
b SmithKline Beecham Pharmaceuticals, Clarendon Road, Worthing, West Sussex, UK BN14 8QH
2
Pyruvate is the most likely primary metabolic source of the
incorporation levels to be determined. Retention levels of H
were analysed by the observation of splitting or isotopic shifts
of the doublets.
three b-lactam carbons of clavulanic acid.
Clavulanic acid 1 is a commercially important inhibitor of b-
lactamases, and is produced via fermentation of Streptomyces
clavuligerus. It is known that the primary metabolic precursor
of the five-carbon unit (Scheme 1) of clavulanic acid is
arginine,1 Pioneering studies by Elson et al.2 and Townsend
et al.3 have demonstrated that the precursor of the three-carbon
unit is a three-carbon primary metabolite (both pyruvate2 and
glycerate3 have been proposed) although no conclusive evi-
dence has yet been presented. Monocyclic b-lactam 2 has been
identified as an intermediate in clavulanic acid 1 biosynthesis
(Scheme 1) and there is evidence4 that it is precursed by the
acyclic compound 3, but the pathway from primary metabolism
to these putative intermediates is unknown.
The l- and d-[1,2-13C2, 2-2H]lactates 5a,b were syn-
thesised7,8 in 23% yield (and > 98% isotopic purity) from
[1-13C]acetic acid (Scheme 2). Notably Tris-HCl buffer was
used for the enzyme catalysed reductions of pyruvate to l- and
d-lactates rather than phosphate buffer9 which resulted in a
mixture of lactates with H and H at C-2. Synthesis of dl-
[1,2-13C2, 2,3,3-2H3]glycerate10‡ was achieved from [2H2]for-
maldehyde (Scheme 3) in an overall yield of 80% and > 98%
isotopic purity. dl-[1,2-13C2, 1,1-18O2]glycerate was syn-
thesised in a similar fashion. Resolution of the racemic labelled
glycerate was achieved by HPLC on a preparative C-18 column
using the method of Horikawa et al.11 In the lactate and
glycerate syntheses, hydrolysis of the intermediate nitriles to the
acids by literature conditions12,13 was low yielding. Use of
strong acid ion exchange resin however resulted in good yields
and trivial product purification.
The labelled precursors were fed to shake flask cultures of a
Streptomyces clavuligerus wild-type re-isolate (SC2) in a
minimal medium.5 Clavulanic acid 1 was isolated as its benzyl
ester after gel and silica chromatography and analysed by 13C
NMR spectroscopy acquired under fully quantitative condi-
tions. It was found that optimisation of the aquisition parameters
was required in order to obtain spectra which were suitable for
integration due to the large pertubations in signal intensity
caused by the 13C and 2H isotope effect on spin relaxation rates.
Samples were also examined by 2H NMR spectroscopy to
corroborate results. The spectra recorded for 1 derived from the
labelled lactates indicated that whilst both d- and l-lactate 5a,b
were specifically incorporated into 1 to similar levels (1.2 and
1
2
Feeding experiments using labelled precursors have appar-
ently demonstrated that glycerol,3,5 propionate,5 b-hydroxy-
propionate,6 pyruvate,3 and d- and l-glycerates3 are specifically
incorporated into the b-lactam ring of 1. It has been reported
that both l- and d-[2-3H, 1-14C]glycerate are specifically
incorporated into 1 to similar levels, but that only in the case of
the d-isomer is any 3H incorporated at C-6 of 1. A recent study3b
using racemic [1,2-13C2, 2,3,3,-2H3]glycerate, however, re-
ported that no incorporation of the C-2 label into C-6 of 1 was
observed. Other possible 3-carbon precursors such as malonic
acid semi-aldehyde, b-hydroxypropionate or acrylate cannot be
ruled out by these previous studies.
2
Herein we report studies using optically active 13C and H
labelled lactates and glycerates which imply that pyruvate 4 and
not glycerate is the primary metabolic precursor of the three
b-lactam carbons of 1.† 1,2-13C labelled species were used in
order to generate doublets in the 13C NMR spectra separating
the signals arising from labelled material from the relatively
intense singlet of the unlabelled material, allowing accurate
2
1.9% respectively), in neither case had any retention of H at
C-6 of 1 occurred. This implies that lactate 5a,b must pass via
pyruvate 4 before incorporation into 3. The spectra recorded for
C-6 C-5
OH
CO2H
NH
H2N
OH
O
O
HN
NH
H2N
NH
N
N
NH
O
O
CO2H
CO2H
'C3'
'C5'
1
2
3
+ L-arginine
GLYCOLYSIS
2–O3PO
?
OH
OH
HO
O
O
O
HO
O
2–O3PO
O
O
OH
OH
OH
OH
OH
OH
OH
4
5b
5a
6a
8
HO
O
•CO2
OH
O
O
HO
O
TCA CYCLE
OH
OH
7
6b
Scheme 1 Proposed biosynthesis of clavulanic acid 1 (2H labels not shown)
Chem. Commun., 1997
1025

View Article Online
the l- and d-glycerates 6a,b similarly showed no evidence for
an additional indication of the metabolic conversions that this
species must undergo before incorporation into 2. Cell extracts
from fermentations of labelled lactate, analysed by
13C-INADEQUATE NMR, showed that an accumulation of
[1,2-13C2]alanine could be detected indicating conversion of
lactate 5a,b to pyruvate 4. It is conceivable that lactate is
converted to glycerate before assimilation into 3, however,
metabolism via the TCA cycle will break the 1-2 13C–13C bond
and is thus ruled out. Additionally the direct conversion of
pyruvate 4 to phosphoenol pyruvate 8 has not been reported for
Streptomyces ssp. and would not explain the loss of all the 2-2H
label (and 20% of the 3-2H labels) in the labelled glycerate
fermentions which imply oxidation of C-2 to the ketone. We
conclude that the most likely primary metabolic source of the
b-lactam carbons of clavulanic acid 1 is pyruvate 4 not glycerate
6a,b or lactate 5a,b.
2
retention of H at C-6 of 1. In contrast, 80% of the clavulanic
acid 1 derived from either d- or l-labelled glycerate was
labelled with 2H at C-5. These results are consistent with results
reported by Pitlik and Townsend using racemic material.3b The
total carbon incorporation for d-glycerate 6a was 1.9% whilst
that for l-glycerate 6b was only 0.3%. The 2H labelling of C-5
of 1 is in good agreement with previous results.3,14,15 The intact
incorporation of the 1,2-13C labels from lactate and glycerate
together with the lack of incorporation of the C-2 2H label from
either substrate indicate that neither is the primary metabolic
precursor for the b-lactam carbons of 1. Instead the results
imply that pyruvate is the precursor. Oxidation of lactate 5a,b to
pyruvate 4 will result in the loss of the 2-2H label whilst
retaining the carbon skeleton. Glycolysis of glycerate 6a,b to
pyruvate 4 will similarly remove the 2-2H label and result in
some loss of the 3-2H label through enolisation. The low level
of incorporation of l-glycerate 6b probably reflects the
differences between metabolism of l- and d-glycerate in S.
clavuligerus. In mammalian systems it is known that
l-glycerate may be epimerised to d-glycerate via oxidation to
hydroxypyruvate16 7. The low incorporation of 18O (19%) into
the C-7 carbonyl of 1 from dl-[1,2-13C2, 1,1-18O2]glycerate is
We thank T. Claridge for advice on NMR, B. Barton, S. W.
Elson and H. Holms for helpful discussions and encouragement,
and the BBSRC/SmithKline Beecham Pharmaceuticals for a
CASE award to J. E. T.
Footnotes
* E-mail: christopher.schofield@dpl.ox.ac.uk
† This work was first presented at the Royal Australian Chemical Institute
15th national conference (2nd July 1996).
‡ The same material was prepared using an alternative route by Pitlik and
Townsend [ref. 3(b)].
HO
D
O
O
O
i
ii
iii
iv
v
13CN
HO
O
Br
O
HO
O
HO
O
H2N
O
Scheme 2 Synthesis of l- and d-[1,2-13C2, 2-2H]lactates 5a,b. Reagents and
conditions: i, PBr3; ii, Cu13CN; iii, HCl, Et2O then H2O; iv, DOWEX-50
(H+), H2O, 80 °C; v, d- or l-lactate dehydrogenase, NAD+, yeast alcohol
dehydrogenase, Tris-HCl, CD3CD2OD, pH 8.5, 25 °C.
References
1 B. P. Valentine, C. R. Bailey, A. Doherty, J. Morris, S. W. Elson,
K. H. Baggaley and N. H. Nicholson, J. Chem. Soc., Chem. Commun.,
1993, 1210.
2 S. W. Elson, in Recent Advances in the Chemistry of b-Lactam
Antibiotics, ed. G. I. Gregory, Royal Society of Chemistry, London,
1981, p. 142.
3 (a) C. A. Townsend and M. Ho, J. Am. Chem. Soc., 1985, 107, 1066; (b)
J. Pitlik and C. A. Townsend, Chem. Commun., 1997, 225.
4 S. W. Elson, K. H. Baggaley, M. Fulston, N. H. Nicholson, J. W. Tyler,
J. Edwards, H. Holms, I. Hamilton and D. Mosdale, J. Chem. Soc.,
Chem. Commun., 1993, 1211.
OD
OD
O
OD
D
OH OH
OH
i
ii
iii
iv
OD
O
13CN
D
D
D
D
D
13CN
D
O
D
D
D
D
D
D
Scheme 3 Synthesis of dl-[1,2-13C2, 2,3,3-2H3]glycerate 6a/b. Reagents
and conditions: i, K13CN, pH 8.5, 25 °C, D2O; ii, Pd–BaSO4, D2, pH 1.7,
D2O; iii, K13CN, pH 8.5, 25 °C, D2O; iv, AMBERLITE IR-120 (H+), D2O,
95 °C.
5 S. W. Elson and R. S. Oliver, J. Antibiot., 1978, 31, 586.
6 A. L. Gutman, V. Ribon and A. Boltanski, J. Chem. Soc., Chem.
Commun., 1985, 1627.
7 H. S. Anker, J. Biol. Chem., 1948, 176, 133.
8 R. C. Thomas, C. H. Wang and B. E. Christensen, J. Am. Chem. Soc.,
1951, 73, 5914.
9 S. S. Shapiro and D. Denis, Biochemistry USA, 1965, 4, 2283.
10 A. S. Serianni, H. N. Nunez and R. Barker, Carbohydr. Res., 1979, 72,
71.
Table 1 Incorporation data for labelled lactate and glycerate
Carbone
Retention Retention
incorpora- at C-6d
at C-5d
(%)
Precursor
tion (%)
(%)
11 R. Horikawa, H. Sakamoto and T. Tanimura, J. Liq. Chromatogr., 1986,
9, 537.
12 A. P. Doerschuk, J. Am. Chem. Soc., 1951, 73, 821.
13 H. J. Sallach, J. Am. Chem. Soc., 1952, 74, 2415.
14 A. Basak, S. P. Salowe and C. A. Townsend, J. Am. Chem. Soc., 1990,
112, 1654.
15 C. A. Townsend and S. Mao, J. Chem. Soc., Chem. Commun., 1987,
86.
16 F. Dickens and D. H. Williamson, Biochem. J., 1959, 72, 496.
l-[1,2-13C2, 2-2H]lactatea 5b
d-[1,2-13C2, 2-2H]lactatea 5a
1.9
1.2
< 2
< 2
< 1
< 5
< 2
n/a
n/a
80
80
80
dl-[1,2-13C2, 2,3,3-2H3]glycerate 6a/b 1.6
l-[1,2-13C2, 2,3,3-2H3]glycerateb 6b 0.3
d-[1,2-13C2, 2,3,3-2H3]glyceratec 6a 1.9
dl-[1,2-13C2, 1,1-18O2]glyceratec
1.8
(C-7 retn: 19% 18O)
a
b
c
d
> 98% ee.
> 90% ee.
> 95% ee. As a proportion of carbon
incorporation; levels measured by 2H NMR were within 5%. e Adjusted to
allow for the amount of 1 present before addition of labelled species, in
order to allow for more reliable comparison of labelling results between
different fermentation batches.
Received in Glasgow, UK, 19th February 1997; Com.
7/01934C
1026
Chem. Commun., 1997