Cooperation of two bifunctional enzymes in the biosynthesis and attachment of deoxysugars of the antitumor antibiotic mithramycin

Full text of DOI:10.1002/anie.201205414

Angewandte
Chemie
DOI: 10.1002/anie.201205414
Biosynthesis
Cooperation of Two Bifunctional Enzymes in the Biosynthesis and
Attachment of Deoxysugars of the Antitumor Antibiotic
Mithramycin**
Guojun Wang, Pallab Pahari, Madan K. Kharel, Jing Chen, Haining Zhu, Steven G. Van Lanen,
and Jꢀrgen Rohr*
Glycosylation reactions catalyzed by glycosyltransferases
(GTs) are important structural modifications that are often
necessary for the biological activities of natural products.[1]
Studying these is an increasingly important and rigorously
pursued research field. Although GTs usually display strict
substrate and regiospecificities, some examples were found
showing surprising tolerance, which can be exploited in
combinatorial biosynthetic approaches.[1b,c]
Mithramycin (MTM), a highly glycosylated aureolic acid
anticancer agent, consists of a tricyclic polyketide-derived
core with two alkyl side chains and five sugar moieties (three
d-olivoses, one d-oliose, and a d-mycarose) attached as
disaccharide (sugars A and B) and trisaccharide (sugars C–E)
chains at the 6- and 2-positions, respectively (Scheme 1).[2]
The sugar chains are important for the ability of MTM to bind
and cross-link DNA, thereby blocking replication and tran-
scription;[3] considerable efforts have been made to elucidate
the biosynthetic assembly of the MTM sugar chains. Gene
inactivation experiments suggested that MtmGIV catalyzes
the first glycosylation (attaching sugar C), followed by the
Scheme 1. Overview of the biosynthesis and structure of MTM. Sug-
transfer of sugar D by MtmGIII.[4] After the assembly of the
ars A, B, and C: d-olivose; sugar D: d-oliose; sugar E: d-mycarose.
trisaccharide, MtmGI and MtmGII then transfer the two d-
olivoses (sugars A and B) of the disaccharide.[5] However, an
obvious GT candidate for the transfer of the d-mycarose
moiety (sugar E) of the trisaccharide was missing. Although
MtmGIV was suggested for this role,[6] such a dual function of
a GTusing two different donor and two significantly different
acceptor substrates would be unprecedented, and has
remained speculation. Furthermore, the biosynthetic pathway
leading to d-olivose, the major monosaccharide building
block of MTM, remained incomplete until recently, when
a reconstitution of the biosynthesis of TDP (thymidine
diphosphate)-d-olivose utilizing enzymes of the MTM path-
way or functionally equivalent homologues (Scheme 2)[7]
revealed that MtmC catalyzes the 4-ketoreduction step.
[*] Dr. G. Wang, Dr. P. Pahari,[+] Dr. M. K. Kharel,
Prof. Dr. S. G. Van Lanen, Prof. Dr. J. Rohr
Department of Pharmaceutical Sciences, College of Pharmacy
University of Kentucky
789 South Limestone Street, Lexington, KY 40536 (USA)
E-mail: jrohr2@email.uky.edu
Dr. J. Chen, Prof. Dr. H. Zhu
Department of Biochemistry, University of Kentucky
Lexington, KY 40536 (USA)
[+] Present address: Synthetic Organic Chemistry Division
CSIR North East Institute of Science and Technology
Jorhat, Assam, 785006 (India)
[**] This work was supported by the NIH grant CA 091901 to J.R.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201205414.
Scheme 2. Biosynthetic pathway of the MTM sugars. Note the dual
role of MtmC as C-methyltransferase and 4-ketoreductase.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Previously, MtmC was only recognized as SAM (S-adenosyl-
methionine)-dependent C-methyltranferase involved in the
biosynthesis of the branched sugar d-mycarose (sugar E).[6] In
addition to MtmC, MtmTIII acts as a unique 4-ketoreductase
of the MtmC product in the biosynthesis of the d-mycarose
building block (Scheme 2).[6]
To determine the exact functions of MtmGIV, MtmC, and
MtmTIII in the assembly of the trisaccharide chain, we now
reconstituted the pathway of d-mycarose in vitro. MtmC was
overproduced in Escherichia coli as described previously,[7]
purified to near homogeneity (see Supporting Information,
Figure S1), and tested for its C-methyltransferase activity by
using TDP-4-keto-d-olivose (2; Scheme 2) and SAM as
substrates. Substrate 2 was prepared from TDP-4-keto-6-
deoxy-d-glucose (1) by using the enzyme pair OleV (an
MtmV homologue from the oleandomycin pathway)/OleW,[8]
together with an NADPH regeneration system consisting of
glucose-6-phosphate and glucose-6-phosphate dehydrogenase
(G6PD).[7]
attempted to overproduce MtmGIV in E. coli, which led to
a soluble but inactive protein. We also failed to obtain
functional MtmGIV from S. lividans TK64, but succeeded by
overexpressing the mtmGIV gene in the S. argillaceus
DmtmGIV mutant strain M3G4. This mutant strain produces
premithramycinone (6, Schemes 1 and 3),[4] and complemen-
tation with mtmGIV completely restored the MTM produc-
tion, suggesting that functional MtmGIV was produced
(Supporting Information, Figure S3). Using this comple-
mented strain, MtmGIV was purified to near homogeneity
(Supporting Information, Figure S4).
First, we wanted to demonstrate the glycosyltransferase
activity of MtmGIV by monitoring the transfer of d-olivose
(sugar C) onto 6 (the first glycosylation event; Scheme 3).
Incubations of MtmGIV with 6 and pure 3 resulted in the
production of a new peak (Figure 2, trace a) with a [MÀH]À
ion at m/z 543.1 (Supporting Information, Figure S5), and
HRMS (Supporting Information, Table S1) and 1H NMR
spectra (Supporting Information, Figure S6) that are consis-
tent with structure of premithramycin A1 (7), thus confirming
the enzyme activity and the role of MtmGIV in initiating the
trisaccharide assembly. As expected, the control proteins
prepared from M3G4 cells harboring only the empty vector
failed to glycosylate 6. By monitoring this glycosylation,
MtmGIV was found to optimally react at pH 8 (Supporting
Information, Figure S7). Interestingly, MtmGIV was also able
to transfer multiple d-olivose units iteratively onto 6, result-
ing in diolivosyl- and triolivosylpremithramycinone (10 and
11; Scheme 3) when excess amounts of 3 were present
(Figure 2, trace b). The identity of 10 and 11 was confirmed
by HRMS (Supporting Information, Table S1).
Substrate 2 disappeared when MtmC was incubated with
pure 2 and SAM, yet no unique TDP-sugar product was
detectable (Supporting Information, Figure S2). In contrast,
replacement of pure 2 by crude preparations (OleV/OleW
reactions for in situ generation of 2) did reveal a new product
that co-eluted with TDP-d-olivose (3; Figure 1, trace b). The
Figure 1. HPLC traces of sugar reactions catalyzed by MtmC. a) Crude
2 without MtmC; b) crude 2+MtmC+SAM+NADPH; c) adding
MtmTIII and NADPH into (b); d) adding NaBH4 into (b);
e) MtmC+pure 2+NADPH. A254: absorbance at 254 nm.
new product could not be reduced, neither by MtmTIII nor
chemically with NaBH4 (Figure 1, traces c and d), excluding
a ketosugar as a possibility. The product was confirmed to be 3
1
by H NMR spectroscopy (see the Supporting Information),
and its generation can be explained by the 4-ketoreductase
activity of MtmC. This reduction also occurred when MtmC
was incubated with pure 2 and NADPH (Figure 1, trace e).
Although we were unable to detect a methylated MtmC
product starting with pure 2 and SAM, the disappearance of 2
suggested that a reaction had occurred, but leading to an
unstable product. Possibly, the axial 3-OH group generated
concomitantly with the 3-methylation reaction can attack the
proximal phosphate to cleave the phosphodiester bond,
thereby removing the thymidine chromophore (Supporting
Information, Figure S2), which is analogous to a mechanism
we and others identified recently.[9]
Figure 2. HPLC traces of MtmGIV-catalyzed glycosylations using 6 (a–
c) or 8 (d,e) as the acceptor substrate. a) MtmGIV+6+3; b) Mtm-
GIV+6+excess amount of 3 (0.8 mm); c) MtmGIV+6+2;
d) MtmGIV+8+in situ generated 5; e) MtmGIV+8+4 generated
in situ. The conversion rate [%] is shown in parentheses. A420
:
absorbance at 420 nm.
As some earlier isolated MTM derivatives contained
ketosugars in C and E positions,[6] we wanted to confirm that
MtmGIV can also transfer ketoolivose (2). Thus, the activity
of MtmGIV was tested with 6 and 2. Two new peaks were
observed using LC-MS, with retention times tret = 20.3 and
16.3 min (Figure 2, trace c). Corresponding [MÀH]À ions
We hoped to circumvent this instability of the methylated
sugar by using MtmGIV-coupled reactions, and therefore
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
k
cat = (2.2 Æ 0.1) minÀ1 with
respect to
3 and KM =
(0.21 Æ 0.04) mm and kcat
=
(0.11 Æ 0.02) minÀ1
with
respect to
2 (Supporting
Information, Figure S8).
After confirming the
function of MtmGIV in
transferring d-olivose onto
6, we next tested whether
a
mixture of MtmC,
MtmTIII, and MtmGIV
could produce d-mycarose
in situ from 2 and transfer it
to premithramycin A2 (8;
Scheme 3). When MtmC,
MtmTIII, and MtmGIV
were incubated with 2,
acceptor substrate 8, SAM,
and NADPH, the HPLC
analysis revealed
peak that co-eluted with
premithramycin A3 (9;
a new
Figure 2, trace d). The LC-
MS yielded an [MÀH]À ion
at m/z 831.3 (Supporting
Information,
Figure S5)
consistent with the molecu-
lar formula C41H52O18 of 9,
which was further corrobo-
rated by HRMS (Support-
ing Information, Table S1)
and 1H NMR (Supporting
Information,
Figure S9).
These data confirm that
MtmGIV indeed catalyzes
two glycosylation events in
MTM
biosynthesis
by
Scheme 3. Proposed pathway for the assembly of the MTM trisaccharide chain and structures of non-naturally
glycosylated premithramycins.
attaching two structurally
distinct sugars onto two sig-
nificantly different acceptor
were m/z 541.1 and 559.1 (Supporting Information, Fig-
ure S5), respectively, which is consistent with a molecular
formula C27H26O12 of 4-keto-d-olivosyl-premithramycinone
(12; Scheme 3) and its hydrated form C27H28O13 (12a), which
was corroborated by HRMS (Supporting Information,
Table S1). Thus, MtmGIV was able to transfer both the 4-
keto and the reduced sugar to the acceptor substrate 6,
confirming for the first time in vitro such a sugar donor
substrate flexibility of MtmGIV. To probe whether the 4-
ketoreduction of sugar C in MTM biosynthesis occurs before
or following the MtmGIV catalysis, kinetic studies were
carried out. The results showed a complete conversion of 3
into product within 0.5 h, while the transfer of 2 required at
least 8 h (Supporting Information, Figure S8), suggesting the
reduced sugar 3 to be the preferred sugar donor for MtmGIV,
which was further confirmed by single-substrate kinetic
experiments. Both substrates followed typical Michaelis–
Menten kinetic patterns with KM = (0.25 Æ 0.04) mm and
molecules. Moreover, MtmGIV seemed to closely cooperate
with MtmC in its glycosylation events, triggering us to further
investigate MtmC. The C-methyltransferase activity of MtmC
was further tested by omitting MtmTIII and NADPH from
the reaction mix that now consisted of MtmC, MtmGIV, 2, 8,
and SAM. This enzyme mixture yielded two new peaks with
t
ret. = 24.3 and 19.8 min (Figure 2, trace e). LC-MS analysis of
these new peaks revealed [MÀH]À ions at m/z 829.3 and 847.4
(Supporting Information, Figure S5), respectively, consistent
with the molecular formula C41H50O18 of 4-keto-d-mycarosyl-
premithramycin A2 (13; Scheme 3) and its hydrated form
C41H52O19 (13a), which was further confirmed by HRMS
(Supporting Information, Table S1). These results clearly
confirmed the transfer of a methyl group onto 2 to afford
keto-d-mycarose. The same glycosylation reactions occur
when premithramycin A2’ (8a, Scheme 3) was used as the
acceptor substrate (Supporting Information, Figure S10).
Indirectly, these results also confirmed the function of
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
MtmTIII in vitro, which had previously been assigned as a 4-
ketoreductase for the d-mycarose biosynthesis based on an
mtmTIII inactivation experiment,[6] which led to the accumu-
lation of 4E-ketomithramycin.
Given the unusual substrate flexibility of MtmGIV in
MTM biosynthesis, we wanted to more closely examine its
substrate specificity by using alternative acceptors and TDP–
sugar donors. The attachment of sugar C to 6 was first
examined by using a panel of TDP-activated sugars (Support-
ing Information, Figure S11). A new peak with tret. = 16.2 min
appeared when TDP-d-quinovose was used as donor
(Figure 3, trace a), with an [MÀH]À ion at m/z 558.9 (Sup-
for which we checked the transfer to both the potential
disaccharidal acceptor substrates 8 and 8a. Here, MtmGIV
was unable to use any other TDP-sugars besides 4 and 5
(Supporting Information, Table S2). Most surprising was the
fact that MtmGIV failed to recognize TDP-d-digitoxose,
although MTM analogues with d-digitoxose as the E sugar
have been reported.[2b,11]
Following up on our observation that MtmGIV can
catalyze multiple, sequential transfers of d-olivose to gen-
erate mono-, di- and triglycosylated MTMs onto the acceptor
substrate 6, we also checked whether MtmGIV would tolerate
monosaccharide 7 (Scheme 3) as an acceptor substrate. In the
presence of 3 and 7, MtmGIV rapidly generated compounds
10 and 11 (Figure 4 trace a; Supporting Information, Fig-
ure S12). Incubation of MtmGIV with 7 and the ketosugar 2
Figure 3. HPLC traces of substrate flexibility studies of MtmGIV using
6 as the acceptor substrate. a) MtmGIV+6+TDP-d-quinovose;
b) MtmGIV+6+TDP-4-keto-d-digitoxose generated in situ;
c) MtmGIV+6+TDP-d-digitoxose generated in situ.
Figure 4. HPLC traces of substrate flexibility studies of MtmGIV using
7 as the acceptor substrate. a) MtmGIV+7+3; b) MtmGIV+7+2;
c) MtmGIV+7+3 and 5 generated in situ by adding MtmC, MtmTIII,
SAM, and NADPH into (b); d) MtmGIV+7+3 generated in situ by
adding only MtmC and NADPH into (b).
porting Information, Figure S5), consistent with the molec-
ular formula C27H28O13 of d-quinovosylpremithramycinone
(14), which was further supported by HRMS (Supporting
Information, Table S1). Notably, two TDP-sugars with axial 3-
OH groups were also prepared, TDP-4-keto-d-digitoxose and
TDP-d-digitoxose (Supporting Information, Figure S11). The
former sugar was generated in situ by incubating 1 with OleV
and EryBII (an earlier identified 3-ketoreductase from the
erythromycin pathway).[10] LC-MS analysis revealed two new
peaks eluted at 20.4 and 16.6 min, respectively (Figure 3,
trace b), with [MÀH]À ions at m/z 541.0 and 559.0 (Support-
ing Information, Figure S5) that are consistent with the
molecular formula C27H26O12 of 4-keto-d-digitoxosyl-premi-
thramycinone (15) and its hydrated form C27H28O13 (15a;
Supporting Information, Table S1). The latter sugar was
generated by utilizing the same conditions as above but
including MtmTIII and NADPH. LC-MS revealed a new
peak with tret. = 17.2 min (Figure 3, trace c) with an [MÀH]À
ion at m/z 543.2 (Supporting Information, Figure S5), consis-
tent with the molecular formula C27H28O12 of d-digitoxosyl-
premithramycinone (16). The identities of 15 and 16 were
confirmed by HRMS (Supporting Information, Table S1). In
total, our results suggested that MtmGIV has an unusually
relaxed substrate specificity towards the stereochemical
configuration at the 3-position of sugar donors, but rigidly
controls that of the 4-position and prefers 2,6-deoxygenated
d-sugars.
also yielded two new products with tret. = 20.5 min and
16.9 min (Figure 4, trace b). LC-MS revealed two [MÀH]À
ions at m/z 671.2 and 689.1 (Supporting Information, Fig-
ure S5), respectively, as confirmed by HRMS (Supporting
Information, Table S1), which were consistent with the
molecular formula C33H36O15 of 4-keto-d-olivosyl-premithra-
mycin A1 (17, m/z 672.2) and its hydrated form C33H38O16
(17a, m/z 690.2). Next, we tested the ability of MtmGIV to
transfer d-mycarose onto 7 by using 5 generated in situ.
Instead of the expected d-mycarosyl-premithramycin A1
(m/z 702.2), LC-MS revealed two peaks with tret. = 17.8 and
20.7 min (Figure 4, trace c) with [MÀH]À ions at m/z 673.2
and 817.2, respectively (Supporting Information, Figure S5).
The former was consistent with a disaccharide product
containing an additional d-olivose that is formed by MtmC-
catalyzed reduction of the ketosugar 2 instead of a d-
mycarose formed by MtmC-catalyzed methylation. Indeed,
the same product was obtained when MtmTIII and SAM
were omitted from the enzyme mixture (Figure 4, trace d).
The latter peak, analyzed by HRMS (Supporting Informa-
tion, Table S1), was consistent with the molecular formula of
C40H50O18 for the trisaccharidal compound 18 generated by
sequential transfer of d-olivose and d-mycarose. These results
clearly indicated that MtmGIV itself is able to assemble
trisaccharide-containing products, and that MtmC can pro-
The same panel of TDP-activated sugars was tested for
the second glycosyltransfer reaction catalyzed by MtmGIV,
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
9814 – 9859; c) R. W. Gantt, P. Peltier-Pain, J. S. Thorson, Nat.
Prod. Rep. 2011, 28, 1811 – 1853.
vide parallel formation of both d-olivose and d-mycarose,
from 2 through its 4-ketoreductase and C-methyltransferase
activities, respectively.
[2] a) J. Rohr, C. Mꢁndez, J. A. Salas, Bioorg. Chem. 1999, 27, 41 –
54; b) L. E. NfflÇez, S. E. Nybo, J. Gonzalez-Sabin, M. Pꢁrez, N.
Mꢁnendez, A. F. BraÇa, M. He, F. Morís, J. A. Salas, J. Rohr, C.
Mꢁndez, J. Med. Chem. 2012, 55, 5813 – 5825.
[3] M. Sastry, R. Fiala, D. J. Patel, J. Mol. Biol. 1995, 251, 674 – 689.
[4] G. Blanco, E. Fernandez, M. J. Fernandez, A. F. BraÇa, U.
Weissbach, E. Künzel, J. Rohr, C. Mꢁndez, J. A. Salas, Mol. Gen.
Genet. 2000, 262, 991 – 1000.
[5] M. Nur-e-Alam, C. Mꢁndez, J. A. Salas, J. Rohr, ChemBioChem
2005, 6, 632 – 636.
[6] L. L. Remsing, J. Garcia-Bernardo, A. M. Gonzalez, E. Künzel,
U. Rix, A. F. BraÇa, D. W. Bearden, C. Mꢁndez, J. A. Salas, J.
Rohr, J. Am. Chem. Soc. 2002, 124, 1606 – 1614.
[7] G. Wang, M. K. Kharel, P. Pahari, J. Rohr, ChemBioChem 2011,
12, 2568 – 2571.
[8] I. Aguirrezabalaga, C. Olano, N. Allende, L. Rodriguez, A. F.
BraÇa, C. Mꢁndez, J. A. Salas, Antimicrob. Agents Chemother.
2000, 44, 1266 – 1275.
[9] a) P. Guyett, J. Glushka, X. Gu, M. Bar-Peled, Carbohydr. Res.
2009, 344, 1072 – 1078; b) X. Chi, P. Pahari, K. Nonaka, S. G.
Van Lanen, J. Am. Chem. Soc. 2011, 133, 14452 – 14459.
[10] F. Lombó, M. Gibson, L. Greenwell, A. F. BraÇa, J. Rohr, J. A.
Salas, C. Mꢁndez, Chem. Biol. 2004, 11, 1709 – 1718.
[11] a) I. Baig, M. Pꢁrez, A. F. BraÇa, R. Gomathinayagam, C.
Damodaran, J. A. Salas, C. Mꢁndez, J. Rohr, J. Nat. Prod. 2008,
71, 199 – 207; b) M. Pꢁrez, I. Baig, A. F. BraÇa, J. A. Salas, J.
Rohr, C. Mꢁndez, ChemBioChem 2008, 9, 2295 – 2304.
[12] a) C. Zhang, C. Albermann, X. Fu, J. S. Thorson, J. Am. Chem.
Soc. 2006, 128, 16420 – 16421; b) W. Lu, C. Leimkuhler, M.
Oberthür, D. Kahne, C. T. Walsh, Biochemistry 2004, 43, 4548 –
4558; c) A. Luzhetskyy, M. Fedoryshyn, C. Dürr, T. Taguchi, V.
Novikov, A. Bechthold, Chem. Biol. 2005, 12, 725 – 729.
[13] a) P. C. Dorrestein, S. G. Van Lanen, W. Li, C. Zhao, Z. Deng, B.
Shen, N. L. Kelleher, J. Am. Chem. Soc. 2006, 128, 10386 – 10387;
b) Y. Chen, K. Fan, Y. He, X. Xu, Y. Peng, T. Yu, C. Jia, K. Yang,
ChemBioChem 2010, 11, 1055 – 1060; c) Y. H. Chen, C. C. Wang,
L. Greenwell, U. Rix, D. Hoffmeister, L. C. Vining, J. Rohr,
K. Q. Yang, J. Biol. Chem. 2005, 280, 22508 – 22514; d) A. Mayer,
T. Taguchi, A. Linnenbrink, C. Hofmann, A. Luzhetskyy, A.
Bechthold, ChemBioChem 2005, 6, 2312 – 2315; e) L. Zhu, B.
Ostash, U. Rix, M. Nur-e-Alam, A. Mayers, A. Luzhetskyy, C.
Mꢁndez, J. A. Salas, A. Bechthold, V. Fedorenko, J. Rohr, J. Org.
Chem. 2005, 70, 631 – 638; f) J. He, M. Müller, C. Hertweck, J.
Am. Chem. Soc. 2004, 126, 16742 – 16743; g) N. Tibrewal, T. E.
Downey, S. G. Van Lanen, E. Ul Sharif, G. A. OꢀDoherty, J.
Rohr, J. Am. Chem. Soc. 2012, 134, 12402 – 12405; h) B. D. Ames,
T. P. Korman, W. Zhang, P. Smith, T. Vu, Y. Tang, S. C. Tsai,
Proc. Natl. Acad. Sci. USA 2008, 105, 5349 – 5354.
In conclusion, we have functionally assigned two enzymes
that closely cooperate for the assembly of the MTM
trisaccharide chain; both have a dual role. MtmGIV is
responsible for two distinct sugar transfers, leading to
sugars C and E in MTM. It closely cooperates with MtmC,
which was shown to use a single substrate 2 yet remarkably
catalyzes either a reduction or a methyl transfer reaction
using the appropriate co-substrate (NADPH or SAM)
depending on which of the two MtmGIVactions is supported.
This way MtmC generates the two building blocks of MTMꢀs
trisaccharide chain, TDP-d-olivose and TDP-d-mycarose,
respectively, which serve as sugar donors for MtmGIV.
Although a few iteratively acting GTs, such as AveBI,
AknK, LanGT1, and LanGT4 had been reported,[12] they all
utilized only one sugar donor substrate. To our knowledge,
MtmGIV is the first characterized GT able to recognize two
distinct acceptor and two distinct donor substrates in a given
metabolite pathway. The discovery of these dual functional
enzymes explains two previously missing activities that are
essential for MTM biosynthesis and that were not readily
predicted from the bioinformatic analysis of the gene
products. Taken together, the results have allowed us to
propose a more complete biosynthetic pathway for MTM, and
also provide the impetus for generating new glycosylated
MTMs by exploiting the substrate flexibility of MtmGIV.
Apart from the mentioned GTs, also iteratively acting
ketosynthases or cyclases/aromatases, and most recently
a reductase/methyltransferase were identified previously.
Such multifunctional enzymes are likely to be just the tip of
the iceberg and will continue to be discovered in secondary
metabolism. Only a few of these have been previously studied
biochemically.[9a,13]
Received: July 9, 2012
Published online: && &&, &&&&
Keywords: bifunctional enzymes · glycosyltransferase ·
.
4-ketoreductase · c-methyltransferase · mithramycin
[1] a) V. Kren, T. Rezanka, FEMS Microbiol. Rev. 2008, 32, 858 –
889; b) C. J. Thibodeaux, C. E. MelanÅon III, H. W. Liu, Angew.
Chem. 2008, 120, 9960 – 10007; Angew. Chem. Int. Ed. 2008, 47,
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Biosynthesis
Two bifunctional enzymes cooperate in
the assembly and the positioning of two
sugars, d-olivose and d-mycarose, of the
anticancer antibiotic mithramycin. MtmC
finishes the biosynthesis of both sugar
building blocks depending on which
MtmGIV activity is supported. MtmGIV
transfers these two sugars onto two
structurally distinct acceptor substrates.
The dual function of these enzymes
explains two essential but previously
unidentified activities.
G. Wang, P. Pahari, M. K. Kharel, J. Chen,
H. Zhu, S. G. Van Lanen,
J. Rohr*
&&&& — &&&&
Cooperation of Two Bifunctional Enzymes
in the Biosynthesis and Attachment of
Deoxysugars of the Antitumor Antibiotic
Mithramycin
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!