New insights into the first oxidative phenol coupling reaction during vancomycin biosynthesis

Article abstract of DOI:10.1016/j.bmcl.2007.11.093

OxyB catalyzes the first oxidative phenol coupling reaction in vancomycin biosynthesis. OxyB is a P450 hemoprotein whose activity is strictly dependent upon the presence of molecular oxygen. Here, it was shown that label from ¹⁸O₂ is

Full text of DOI:10.1016/j.bmcl.2007.11.093

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3081–3084
New insights into the first oxidative phenol coupling
reaction during vancomycin biosynthesis
Nina Geib, Katharina Woithe, Katja Zerbe, Dong Bo Li and John A. Robinson*
Department of Chemistry, Institute of Organic Chemistry, University of Zu¨rich, Winterthurerstrasse 190, 8057 Zu¨rich, Switzerland
Received 6 October 2007; revised 21 November 2007; accepted 21 November 2007
Available online 28 November 2007
Abstract—OxyB catalyzes the first oxidative phenol coupling reaction in vancomycin biosynthesis. OxyB is a P450 hemoprotein
whose activity is strictly dependent upon the presence of molecular oxygen. Here, it was shown that label from 18O2 is not incor-
porated into the monocyclic product during catalysis by OxyB. In addition, it was shown that OxyB can convert a model hexapep-
tide substrate containing (R)-Tyr6, instead of (S)-Tyr6, covalently linked as a C-terminal thioester to a peptidyl carrier protein
(PCP-7S) derived from the vancomycin non-ribosomal peptide synthetase (NRPS), into the corresponding epimeric monocyclic
product. The binding of this epimeric hexapeptide-PCP conjugate to the Fe(III) form of OxyB, as monitored by UV–vis spectros-
copy, revealed a Kd = 35 5 lM. Thus, the enzyme reveals a surprising lack of stereospecificity in the binding and transformation of
these epimeric substrates.
Ó 2007 Elsevier Ltd. All rights reserved.
Vancomycin (Fig. 1), a glycopeptide produced by Amy-
colatopsis orientalis, is a clinically important antibiotic
that acts by inhibiting cell wall biosynthesis in Gram-po-
sitive bacteria.1 During vancomycin biosynthesis, key
steps include rigidification of the heptapeptide aglycone
backbone. These are accomplished by the actions of
three cytochrome P450 hemoproteins OxyA, OxyB,
and OxyC that catalyze three consecutive oxidative phe-
nol coupling reactions.2–5 The first coupling, which ulti-
mately links the phenol rings in the side chains of Hpg4
((R)-4-hydroxyphenylglycine) and Cht6 ((2S,3R)-m-
chloro-b-hydroxytyrosine) (C-O-D ring), is catalyzed
by OxyB. The second aryl-ether bridge is formed by
OxyA and links the side chains of amino acids Cht2
((2R,3R)-m-chloro-b-hydroxytyrosine) and Hpg4 (D-
O-E ring), and the last coupling catalyzed by OxyC
connects the aromatic side chains of Hpg5 and Dpg7
((S)-3,5-dihydroxyphenylglycine) (AB ring).
Figure 1. Structure of vancomycin.
to a recombinant peptide carrier protein (PCP) domain
from the vancomycin non-ribosomal peptide synthetase
(NRPS) (Fig. 3).6,7 It was also demonstrated that OxyB
activity is strictly dependent upon the presence of redox
proteins that deliver electrons to the hemoprotein, as
well as upon molecular oxygen. The model substrate,
1-PCP-7S (Fig. 2), binds to OxyB in the absence of other
ancillary proteins with a dissociation constant of
17 5 lM, as determined by UV–vis difference spectros-
copy.7 Furthermore, we showed that OxyB is also able
We have shown previously that a model linear hexapep-
tide (1) (Fig. 2) related to the vancomycin aglycone is
converted efficiently into a monocyclic product in vitro
by OxyB only when the peptide is linked as a thioester
Keywords: Glycopeptide; Antibiotic; Cytochrome P450; Peptide; Bio-
synthesis; Enzyme.
*
Corresponding author. Tel.: +41 44635 4242; fax: +41 44635 6833; e-
mail: robinson@oci.uzh.ch
0960-894X/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2007.11.093

3082
N. Geib et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3081–3084
Figure 2. Activation of hexapeptides 1 and 2 as thioesters. Reagents and conditions: (i) peptide (1 equiv), PyBOP (1.2 equiv), i-Pr2EtN (1.2 equiv),
PhSH (2.4 equiv), DMF, rt, 15 min; (ii) coenzyme A (4 equiv), phosphate buffer, pH 8.5, rt, 2 h; (iii) apo-PCP-7S, B. subtilis Sfp, MgCl2, Tris–HCl
buffer, pH 7.5, 37 °C, 30 min.
and heptapeptide intermediates whilst they are still at-
tached to the NRPS assembly line.7
In this work, we report labeling studies using 18O2 as the
oxygen source for OxyB, in order to probe the mecha-
nism of this oxidative coupling reaction. Furthermore,
we describe the transformation by OxyB of an epimeric
model hexapeptide (2) conjugated to PCP-7S, where (S)-
tyrosine at position 6 has been replaced by (R)-tyrosine
(Fig. 2).
Monooxygenase reactions catalyzed by P450 hemopro-
teins, such as the well-known camphor hydroxylase
P450cam, not only require molecular oxygen they also
usually incorporate one oxygen atom from molecular
oxygen into the product.8 Since, the incorporation of
one oxygen atom from O2 into the product cannot be ru-
led out a priori during the OxyB reaction, we set out to
test this using O-18 labeled oxygen. We again used the
model hexapeptide-PCP conjugate (1-PCP-7S) as sub-
strate and a reduction system consisting of recombinant
spinach ferredoxin, Escherichia coli flavodoxin reduc-
tase, and NADPH.7 Additionally, glucose-6-phosphate
and glucose-6-phosphate dehydrogenase were used for
NADPH regeneration. After incubation of these redox
proteins and OxyB with the substrate and subsequent
cleavage of the peptide products from the PCP with
hydrazine (Fig. 3), the resulting peptide hydrazides were
purified and analyzed by analytical HPLC. As reported
earlier,7 under the standard assay conditions, the sub-
strate 1-PCP-7S is converted almost quantitatively into
the expected monocyclic product-hydrazide 3 as well
as a minor monocyclic product-hydrazide, assigned as
the Tyr6 epimer 4 (vide infra). For the 18O labeling
studies, the assay solutions were degassed by freeze-
thaw cycles under vacuum, and the assay was performed
in an anaerobic glove box. Under these conditions no
Figure 3. OxyB catalyzed conversion of the PCP-bound hexapeptides
1 and 2 into monocyclic products 3 and 4.
to catalyze a comparable phenol coupling reaction on a
model heptapeptide-PCP-7S conjugate, thus raising the
possibility that OxyB may act upon both linear hexa-

N. Geib et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3081–3084
3083
significant product formation was observed. Upon
introduction of 18O2 (95% 18O2, Cambridge Isotope
Laboratories Inc.), the catalytic activity of OxyB was re-
stored. However, electrospray MS analysis of the iso-
lated major monocyclic hexapeptide hydrazide 3 gave
m/z 896 [M+H]+, indicating that within the limits of
detection no label from 18O2 had been incorporated into
the product.
The lack of 18O incorporation during the phenol cou-
pling reaction catalyzed by OxyB is of interest when
considering possible mechanisms for this transforma-
tion. We have described evidence that binding of sub-
strate to OxyB induces a spin-state shift toward the
high spin Fe(III) form.7 Since both electrons and molec-
ular oxygen are required for catalysis by OxyB, then in
analogy to the P450cam catalytic cycle, adding electrons
and oxygen to the heme Fe in OxyB should lead to the
so-called compound-0 intermediate (Fig. 4). Although
there is no evidence so far that this intermediate is in-
deed formed on OxyB, it seems reasonable to suggest
this. From here onwards, the mechanism of the oxida-
tive coupling becomes unclear. Even in the case of the
very well-studied P450cam, there is still no broad con-
sensus on the reaction mechanism following the forma-
tion of compound-0.9,10 However, should the OxyB
mechanism involve oxygen atom transfer to one aro-
matic ring, to form an arene epoxide (Fig. 4), a type
of process that is thought to occur (but has not been
proven) during oxidation of aromatic hydrocarbons by
P450 enzymes (e.g. in the liver),8 followed by rearrange-
ment to the macrocyclic product, then this must occur
without retention of the oxygen atom derived from
molecular oxygen. Clearly other pathways cannot pres-
ently be ruled out.
Figure 5. HPLC chromatograms showing peptide hydrazide products
of OxyB reaction following hydrazine cleavage using 1-PCP-7S (black)
and 2-PCP-7S (gray) as substrates. Peak 1, monocyclic product (4);
peak 2, unreacted linear peptide; peak 3, monocyclic product (3) (see
Fig. 3).
pursue this observation, we decided here to test directly
as substrate the epimeric hexapeptide 2, possessing (R)-
Tyr instead of (S)-Tyr at position 6, linked as a thioester
to PCP-7S. This hexapeptide was synthesized and con-
verted into the corresponding PCP-7S conjugate by a
standard procedure (Fig. 2).7 OxyB assays using this
substrate resulted in almost complete conversion into
two products, one major and one minor, identified by
HPLC analysis (Fig. 5). Interestingly, the two observed
peaks exhibited identical retention times but inverse
intensities to those observed when using the epimeric
model hexapeptide 1-PCP-7S as substrate. A full charac-
terization by MS and NMR of the major and minor
products obtained from transformation of peptide 2-
PCP-7S revealed that the expected C-O-D macrocycle
had been formed in both. Moreover, 1H 2D COSY,
TOCSY, and NOESY spectra of these major and minor
products (Fig. 5, peaks 1 and 3, respectively) proved that
the minor monocyclic product (peak 3) of this conver-
sion was identical to the major monocyclic hexapeptide
hydrazide (3) derived from assays using 1-PCP-7S.
Additionally, the major monocyclic product (peak 1,
Fig. 5) was identical to the minor monocyclic product
(4) derived from 1-PCP-7S.
As mentioned above, the standard assay for OxyB af-
fords mainly the expected monocyclic product, and also
around 10–20% of a minor monocyclic product that we
had earlier assigned as the Tyr6 epimer.7 In order to
The binding of the hexapeptide 2-PCP-7S conjugate to
OxyB was also monitored by UV–vis difference spec-
troscopy. Upon titration of OxyB with hexapeptide 2-
PCP-7S, the expected type I binding spectrum (Fig. 6)
showing a peak at 392 nm, a trough at 427 nm and an
isosbestic point at 413 nm, was observed. The concen-
tration dependence of the changes in absorbance fitted
well a binding equation describing a 1:1 interaction, with
a Kd of 35 5 lM, indicating a slightly weaker interac-
tion with OxyB compared to the standard hexapeptide
conjugate (1-PCP-7S: Kd = 17 5 lM).7 This binding
behavior mirrors nicely the results from the OxyB activ-
ity assays, in which both substrates show excellent turn-
over into monocyclic products under otherwise
comparable assay conditions.
Figure 4. The OxyB catalytic cycle should proceed through
a
compound-0 like intermediate. The subsequent steps, however, remain
to be established (see text).

3084
N. Geib et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3081–3084
alyst might prove useful in the preparation of other re-
lated macrocyclic peptides.
Acknowledgments
This work was supported by the Swiss National Science
Foundation and the European Union 6th Framework
Program.
References and notes
1. Hubbard, B. K.; Walsh, C. T. Angew. Chem. Int. Ed. Engl.
2003, 42, 730.
2. Bischoff, D.; Bister, B.; Bertazzo, M.; Pfeifer, V.;
Stegmann, E.; Nicholson, G. J.; Keller, S.; Pelzer, S.;
Wohlleben, W.; Sussmuth, R. D. Chembiochem 2005, 6,
267.
3. Bischoff, D.; Pelzer, S.; Bister, B.; Nicholson, G. J.;
Stockert, S.; Schirle, M.; Wohlleben, W.; Jung, G.;
Sussmuth, R. D. Angew. Chem., Int. Ed. 2001, 40, 4688.
4. Bischoff, D.; Pelzer, S.; Holtzel, A.; Nicholson, G. J.;
Stockert, S.; Wohlleben, W.; Jung, G.; Sussmuth, R. D.
Angew. Chem., Int. Ed. 2001, 40, 1693.
5. Sussmuth, R. D.; Pelzer, S.; Nicholson, G.; Walk, T.;
Wohlleben, W.; Jung, G. Angew. Chem., Int. Ed. 1999, 38,
1976.
6. Zerbe, K.; Woithe, K.; Li, D. B.; Vitali, F.; Bigler, L.;
Robinson, J. A. Angew. Chem., Int. Ed. 2004, 43, 6709.
7. Woithe, K.; Geib, N.; Zerbe, K.; Li, D. B.; Heck, M.;
Fournier-Rousset, S.; Meyer, O.; Vitali, F.; Matoba, N.;
Abou-Hadeed, K.; Robinson, J. A. J. Am. Chem. Soc.
2007, 129, 6887.
Figure 6. Difference UV–vis spectra of hexapeptide 2-PCP-7S binding
to OxyB. The concentration dependence of the spectral changes and
the wavelengths of the minima and maxima are shown.
These complementary and interlocking results obtained
with peptides 1 and 2 serve first to confirm the proposed
structure of the minor product obtained from assays
with the normal model hexapeptide 1-PCP-7S, described
earlier.7 They also demonstrate, perhaps surprisingly,
that the enzyme OxyB can transform efficiently the epi-
meric hexapeptide 2-PCP-7S into a monocyclic product,
apparently without the change in sense of chirality in
residue-6 having a substantial effect on either binding,
the shift in heme spin-state equilibrium, or turnover,
compared to 1-PCP-7S. Presently, it seems most likely
that the minor epimers seen to arise in these assays are
formed during the preparation of the peptide-PCP con-
jugates, and/or during peptide cleavage from the PCP
using hydrazine. Finally, these results indicate that a
more extensive study of the mechanism and substrate
specificity of OxyB is warranted, not least since this cat-
8. OrtizdeMontellano, P. R. Cytochrome P450: Structure,
Mechanism and Biochemistry, 3rd ed.; Kluwer Academic/
Plenum Publishers: New York, 2005.
9. Altun, A.; Shaik, S.; Thiel, W. J. Am. Chem. Soc. 2007,
129, 8978.
10. Groenhof, A. R.; Ehlers, A. W.; Lammertsma, K. J. Am.
Chem. Soc. 2007, 129, 6204.