Angewandte
Chemie
1
1
from isonovobiocin by H NMR spectroscopy. The H NMR
spectrum (500 MHz, [D6]dimethyl sulfoxide (DMSO), 238C)
obtained for the NovN product was identical to the spectrum
of novobiocin and showed peaks at the chemical shifts
characteristic for noviose resonances (d = 5.16 (dd, J2,3 = 3.1,
Materials and Methods
Preparation of pNovP-pET28a and pNovN-pET37b overexpression
constructs: The genes encoding NovPand NovN were amplified from
Streptomyces spheroides (ATCC23965) genomic DNA and pXHN,[18]
respectively. Amplification of novP was accomplished by using the
foward primer 5’-GCGTATCATATGGCACCTATCGTGGAAACCGCG-3’
and the reverse primer 5’-GCCATTAAGCTTTTACCGGGTGCGTTGCCAG
TAG-3’, which include NdeI/HindIII restriction sites. Amplification of
novN was accomplished by using the foward primer 5’-CCGAATCA
TATGCTCATTCTTGGCCTGAACGGG-3’ and the reverse primer 5’-
GCGTTACTCGAGTCACGGTGCGGATCCGGCACC-3’, with NdeI/XhoI
restriction sites. The amplified genes were inserted into pET28a
(novP) and pET37b (novN) vectors after restriction digests, and
pNovP-pET28a and pNovN-pET37b were expressed in E. coli TOP10
competent cells.
J
3,4 = 9.8 Hz, 1H; 3-H), 4.08 ppm (m, 1H; 2-H)), which are
distinctly different from those recorded in the literature for
isonovobiocin (d = 4.18 (3-H), 4.93 ppm (2-H)).[10]
Further confirmation of novobiocin formation was
obtained by comparison of the antibiotic potencies of the
NovN reaction product and commercially available novobio-
cin. The minimum inhibitory concentrations (MICs) of each
compound against two bacterial strains (Enterococcus fae-
cium and Staphylococcus aureus) were determined and found
to be comparable (data not shown). Previous reports have
established that the 3’-O-carbamoyl moiety is essential for
antibiotic activity;[11] loss of activity is observed with the
descarbamoylnovobiocin precursor. We also observed an
MIC for descarbamoylnovobiocin (5) that was more than 100-
fold higher than that determined for novobiocin.
Kinetic parameters were determined for NovN by using
purified descarbamoylnovobiocin (5) obtained from a large-
scale tandem NovM/NovPincubation. NovN catalyzes the
carbamoylation of 5 with kcat = 4.1 Æ 0.2 minÀ1. Km values of
4.6 Æ 1.3 mm and 5.1 Æ 0.4 mm were measured for the sub-
strates descarbamoylnovobiocin and carbamoyl phosphate,
respectively.
Surprisingly, the carbamoyltransferase activity of NovN is
dependent upon Mg-ATP. The regulatory role of ATP in the
activation of the aspartate N-carbamoyltransferases in a
number of bacterial systems has been well studied.[12–16]
Activation of O-carbamoyltransferases by ATPin secondary
metabolite biosynthesis has been reported in only one case:
O-carbamoylation of 3-hydroxymethylcephem in Streptomy-
ces clavuligerus.[17] The role of ATPin the activation of
carbamoyltransferase CmcH was investigated in this study,
but the authors were unable to conclude whether ATPacts as
an effector or as a substrate in the transfer of a carbamoyl
moiety to 3-hydroxymethylcephem. In general, regulation of
late-stage transformations by ATPin secondary metabolite
biosynthesis is not well documented. The role of ATP
activation in this final step of novobiocin biosynthesis remains
to be elucidated.
Overproduction and purification of NovPand NovN: Purified
pNovP-pET28a and pNovN-pET37b plasmids were transformed into
BL21 (DE3) competent E. coli cells. Transformants harboring the
pNovP-pET28a and pNovN-pET37b constructs were grown in Luria–
Bertani (LB) medium supplemented with kanamycin (50 mgmLÀ1).
[5]
NovPwas overproduced and purified as described for NovM. For
the overproduction of NovN, cells were grown at 258C to an optical
density of around 0.35. The temperature was reduced to 208C and the
cells were shaken for 1 h. NovN production was induced with
isopropyl-b-d-thiogalactopyranoside (60 mm) and the cells were
shaken overnight at 208C. NovN was isolated from cell extracts as
described for NovM.[5]
Characterization of NovP: Reactions were carried out at ambient
temperature and reaction mixtures contained desmethyldescarba-
moylnovobiocin (4) and SAM in tricine (75 mm, pH 8.5), NaCl
(100 mm), bovine serum albumin (BSA; 1 mgmLÀ1), and DMSO
(5%). Reactions were initiated by the addition of NovPand
terminated at specified time points by quenching in methanol (2
reaction volume) at 48C. For the determination of Km and kcat, NovP
and SAM were added to final concentrations of 500 nm and 500 mm,
respectively, and the concentration of desmethyldescarbamoylnovo-
biocin was varied (2.5–50 mm). The reaction was terminated after 10
and 20 min for each concentration. Each experiment was carried out
in triplicate.
Quenched aliquots were incubated at À208C for 30 min and then
centrifuged (5 min at 13000 rpm) to remove precipitated protein. The
supernatant was analyzed by analytical reverse-phase HPLC
(CH3CN/H2O (60:40), 0.1% TFA, 1 mLminÀ1
) monitored at
340 nm. Product formation was confirmed by LCMS (5: calcd for
C30H35NO10: m/z 570.23 [M+H]+; found: 570.20) on a Shimadzu
LCMS-QP8000a instrument. The desmethyldescarbamoylnovobiocin
(4; Rt = 7.0 min) and descarbamoylnovobiocin (5; Rt = 12.5 min)
HPLC peak areas were measured and the product concentration
was calculated as a percent of the total peak area.
The availability of the last three enzymes in the novobio-
cin biosynthetic pathway, NovM, NovP, and NovN, in purified,
active forms through E. coli expression will allow evaluation
of the possibility of generating an aminocoumarin library in
which variation of each component, as well as sugar
decoration results in novobiocin analogues that can be
evaluated for antibiotic activity. Moreover, tandem NovM/
NovPincubations have been carried out on a 30-mg scale,
which provides sufficient material to study the regioselective
chemical carbamoylation and/or acylation of descarbamoyl
novobiocin (5) at the 3’-position. We anticipate that variation
of the 3’ moiety will provide insight into the structural
requirements for gyrase inhibition and antibiotic activity.
Characterization of NovN: Reactions (75 mL total volume) were
carried out at ambient temperature and reaction mixtures contained
descarbamoylnovobiocin (5), carbamoyl phosphate, and ATP(4 m m)
in Tris (75 mm, pH 9.0), MgCl2 (10 mm), BSA (1 mgmLÀ1), and
DMSO (5%). Reactions were initiated by the addition of NovN
(200 nm) and terminated by quenching the reaction mixture with
methanol (150 mL) at 48C. Km and kcat were determined with
descarbamoylnovobiocin (5) and carbamoyl phosphate as the varia-
ble substrates. The determination of the Km value for descarbamoyl-
novobiocin (5) was accomplished at a constant concentration of
carbamoyl phosphate (500 mm) and over a range of descarbamoylno-
vobiocin (5) concentrations (5–50 mm). The reaction was quenched
after 5 min for each concentration. The determination of the Km value
for carbamoyl phosphate was accomplished at a constant concen-
tration of descarbamoylnovobiocin (5; 50 mm) and over a range of
carbamoyl phosphate concentrations (1–14 mm). The reaction was
quenched after 10 min for each concentration. Each experiment was
carried out in triplicate.
Angew. Chem. Int. Ed. 2004, 43, 67 –70
ꢀ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
69