Hydroxyquinone O-methylation in mitomycin biosynthesis

Article abstract of DOI:10.1021/ja0700193

Mitomycins are bioreductively activated DNA-alkylating agents. One member of this family, mitomycin C, is in clinical use as part of combination therapy for certain solid tumors. The cytotoxicity displayed by mitomycins is dependent on their electrochemical potential which, in turn, is governed in part by the substituents of the quinone moiety. In this paper we describe studies on the biogenesis of the quinone methoxy group present in mitomycins A and B. An engineered Streptomyces lavendulae strain in which the mmcR methyltransferase gene had been deleted failed to produce the three mitomycins (A, B, and C) that are typically isolated from the wild type organism. Analysis of the culture extracts from the mmcR-deletion mutant strain revealed that two new metabolites, 7-demethylmitomycin A and 7-demethylmitomycin B, had accumulated instead. Production of mitomycins A and C or mitomycin B was selectively restored upon supplementing the culture medium of a S. lavendulae strain unable to produce the key precursor 3-amino-5-hydroxybenzoate with either 7-demethylmitomycin A or 7-demethylmitomycin B, respectively. MmcR methyltransferase obtained by cloning and overexpression of the corresponding mmcR gene was shown to catalyze the 7-O-methylation of both C9β- and C9α-configured 7-hydroxymitomycins in vitro. This study provides direct evidence for the catalytic role of MmcR in formation of the 7-OMe group that is characteristic of mitomycins A and B and demonstrates the prerequisite of 7-O-methylation for the production of the clinical agent mitomycin C.

Full text of DOI:10.1021/ja0700193

Published on Web 04/27/2007
Hydroxyquinone O-Methylation in Mitomycin Biosynthesis
Sabine Gru¨schow, Leng-Chee Chang,† Yingqing Mao,‡ and David H. Sherman*
Contribution from the Life Sciences Institute, Department of Medicinal Chemistry, UniVersity of
Michigan, Ann Arbor, Michigan 48109
Received January 2, 2007; E-mail: davidhs@umich.edu
Abstract: Mitomycins are bioreductively activated DNA-alkylating agents. One member of this family,
mitomycin C, is in clinical use as part of combination therapy for certain solid tumors. The cytotoxicity
displayed by mitomycins is dependent on their electrochemical potential which, in turn, is governed in part
by the substituents of the quinone moiety. In this paper we describe studies on the biogenesis of the quinone
methoxy group present in mitomycins A and B. An engineered Streptomyces lavendulae strain in which
the mmcR methyltransferase gene had been deleted failed to produce the three mitomycins (A, B, and C)
that are typically isolated from the wild type organism. Analysis of the culture extracts from the mmcR-
deletion mutant strain revealed that two new metabolites, 7-demethylmitomycin A and 7-demethylmitomycin
B, had accumulated instead. Production of mitomycins A and C or mitomycin B was selectively restored
upon supplementing the culture medium of a S. lavendulae strain unable to produce the key precursor
3-amino-5-hydroxybenzoate with either 7-demethylmitomycin A or 7-demethylmitomycin B, respectively.
MmcR methyltransferase obtained by cloning and overexpression of the corresponding mmcR gene was
shown to catalyze the 7-O-methylation of both C9â- and C9R-configured 7-hydroxymitomycins in vitro.
This study provides direct evidence for the catalytic role of MmcR in formation of the 7-OMe group that is
characteristic of mitomycins A and B and demonstrates the prerequisite of 7-O-methylation for the production
of the clinical agent mitomycin C.
organism, provide further support for this notion.5-8 Given the
influence of the quinone substituents on the activity of mito-
Introduction
The natural product mitomycin C (1, Figure 1) has been in
clinical use as an anticancer agent for over five decades and is
still part of combination therapy of certain solid tumors.1
Mitomycins are selectively activated in hypoxic cells because
of their requirement for reductive activation to exert their full
potential.2 Numerous biological agents such as diaphorase or
glutathione have been identified as participants in performing
one- or two-electron reduction of the quinone portion of
mitomycin to give the highly reactive semiquinone or hydro-
quinone species, respectively. The hydroquinone is a very short-
lived molecule (T1/2 < 1 s) that rapidly loses the C9a
substituent.3,4 The mitosene derivative formed in this way
subsequently acts as a DNA alkylating agent with concomitant
opening of the aziridine ring and displacement of the carbamoyl
substituent. It is now well-established that the cytotoxic activity
displayed by mitomycins correlates with the electrochemical
potential of the quinone ring. Thus, the reactivity of mitomycins
is governed in part through functional group modulation at the
quinone moiety. Indeed, substitutions at the C6 and C7 positions,
achieved through derivatization of naturally occurring mitomy-
cins or through directed precursor feeding in the producing
mycins, we focused our attention on the biogenesis of the C7-
methoxy group. In this article, we show that methylation of the
C7-hydroxy group of R- and â-mitomycins is performed by the
MmcR methyltransferase whose gene resides within the mito-
mycin biosynthetic gene cluster.9
The actinomycete Streptomyces laVendulae NRRL 2564
produces mitomycin C (1), mitomycin A (2), and mitomycin B
(3, Figure 1). These compounds differ in the stereochemistry
of the C9-substituent, in the nature of the C7-functional group,
and finally in their methylation pattern. Mitomycins with C9â-
configuration, 1 and 2, constitute the major components of the
metabolite profile, whereas the C9R-mitomycin 3 is only
produced as a minor constituent. Investigations into the bio-
synthesis of mitomycins started in the early 1960s, and by 1980
D-glucosamine and 3-amino-5-hydroxybenzoic acid (AHBA)
were established as the biosynthetic precursors that comprise
the mitosane skeleton (Scheme 1).10,11 AHBA is also a precursor
(5) Claridge, C. A.; Bush, J. A.; Doyle, T. W.; Nettleton, D. E.; Moseley, J.
E.; Kimball, D.; Kammer, M. F.; Veitch, J. J. Antibiot. 1986, 39, 437-
446.
(6) Sami, S. M.; Iyengar, B. S.; Tarnow, S. E.; Remers, W. A.; Bradner, W.
T.; Schurig, J. E. J. Med. Chem. 1984, 27, 701-708.
(7) Arai, H.; Kanda, Y.; Ashizawa, T.; Morimoto, M.; Gomi, K.; Kono, M.;
Kasai, M. J. Med. Chem. 1994, 37, 1794-1804.
† Current address: University of Minnesota, Duluth, MN 55812.
‡ Current address: Raybiotech Inc., Norcross, GA 30092.
(1) Bradner, W. T. Cancer Treat. ReV. 2001, 27, 35-50.
(2) Tomasz, M. Chem. Biol. 1995, 2, 575-579.
(8) Iyengar, B. S.; Sami, S. M.; Remers, W. A.; Bradner, W. T.; Schurig, J. E.
J. Med. Chem. 1983, 26, 16-20.
(3) Rao, G. M.; Lown, J. W.; Plambeck, J. A. J. Electrochem. Soc. 1977, 124,
195-198.
(9) Mao, Y. Q.; Varoglu, M.; Sherman, D. H. Chem. Biol. 1999, 6, 251-263.
(10) Hornemann, U.; Kehrer, J. P.; Nunez, C. S.; Ranieri, R. L. J. Am. Chem.
Soc. 1974, 96, 320-322.
(4) Rao, G. M.; Begleiter, A.; Lown, J. W.; Plambeck, J. A. J. Electrochem.
Soc. 1977, 124, 199-202.
9
6470
J. AM. CHEM. SOC. 2007, 129, 6470-6476
10.1021/ja0700193 CCC: $37.00 © 2007 American Chemical Society

Hydroxyquinone O-Methylation
A R T I C L E S
Figure 1. Structures of mitomycins produced by S. laVendulae NRRL 2564 and S. laVendulae ∆mmcR.
Scheme 1
inthebiosynthesisoftheansamycinandansamitocinantibiotics,12-14
and its biosynthesis via the amino-shikimate pathway has been
extensively studied.12,15-19 The aminosugar D-glucosamine is
incorporated intact, providing the nitrogen atom of the aziridine
ring (Scheme 1). AHBA is hydroxylated and further oxidized
to the quinone while its carboxy group is reduced to the C6-
methyl group after condensation with D-glucosamine.20 Mean-
while, all the N- and O-linked methyl groups are derived from
S-adenosylmethionine (AdoMet),21,22 and the carbamoyl group
is derived from carbamoyl phosphate.23,24 Mitomycin A has been
suggested to be the direct precursor of mitomycin C (Scheme
1) based on the initial accumulation of mitomycin A in the
culture broth and its subsequent decrease in concentration with
a corresponding increase in the levels of mitomycin C.25 Few
details are currently available regarding enzymatic assembly of
mitomycins. This is partially due to the unique structural
composition of the metabolites, the relatively low levels of
mitomycins produced by wild type S. laVendulae, and the
inherent lability of biosynthetic intermediates.
information provided considerable insight into the regulation
and cellular self-resistance of these important DNA alkylating
agents.26-29 Subsequent work on dissecting the enzymatic steps
responsible for the assembly of the molecule has focused on
late stage functional group modification.30 In this paper, we
describe the identification of two new biosynthetic intermediates
that accumulate in an mmcR gene deletion mutant strain of S.
laVendulae. These compounds, 7-demethylmitomycin A (4) and
7-demethylmitomycin B, (5) were shown by in ViVo and in Vitro
studies to be converted to the C9â-configured 2 and the C9R-
configured 3, respectively, by MmcR, the 7-O-methyltransferase
responsible for assembly of mitomycin A and mitomycin B.
Experimental Section
Materials. Mitomycins were obtained as a kind gift from Kyowa
Hakko Kogyo Ltd. (Tokyo, Japan). Buffer components and reagents
were purchased from Sigma-Aldrich (St. Louis, MO) or Fisher Scientific
and used without further purification. S-[Methyl-14C]-adenosyl-L-
methionine was obtained from ARC Inc. (St. Louis, MO) at 0.1 mCi/
mmol.
Analytical Methods. A ThermoFinnigan LTQ linear ion-trap
instrument equipped with electrospray source and Surveyor HPLC
system was routinely used for LC-MS analysis. For liquid chromatog-
raphy, a 10 mM ammonium acetate in water (solvent A)/methanol
(solvent B) system was used. Separations were carried out with a Waters
XBridge C18 (3.5 µm, 2.1 × 150 mm) column at a flow rate of 208
µL/min and a linear gradient of 5-100% solvent B in 24.2 min starting
1.2 min after sample injection. For mass spectrometry, the capillary
temperature was set to 200 °C with the source voltage at 4.5 kV, the
capillary voltage at 45 V, and the tube lens at 85 V. The normalized
collision energy for fragmentation was 25% (arbitrary unit). Spectra
A complete genetic characterization of the mitomycin bio-
synthetic gene cluster was reported several years ago.9 This
(11) Anderson, M. G.; Kibby, J. J.; Rickards, R. W.; Rothschild, J. M. J. Chem.
Soc., Chem. Commun. 1980, 1277-1278.
(12) Kim, C. G.; Yu, T. W.; Fryhle, C. B.; Handa, S.; Floss, H. G. J. Biol.
Chem. 1998, 273, 6030-6040.
(13) August, P. R.; Tang, L.; Yoon, Y. J.; Ning, S.; Mu¨ller, R.; Yu, T. W.;
Taylor, M.; Hoffmann, D.; Kim, C. G.; Zhang, X. H.; Hutchinson, C. R.;
Floss, H. G. Chem. Biol. 1998, 5, 69-79.
(14) Yu, T. W.; Bai, L. Q.; Clade, D.; Hoffmann, D.; Toelzer, S.; Trinh, K. Q.;
Xu, J.; Moss, S. J.; Leistner, E.; Floss, H. G. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 7968-7973.
(15) Yu, T. W.; Mu¨ller, R.; Mu¨ller, M.; Zhang, X. H.; Draeger, G.; Kim, C. G.;
Leistner, E.; Floss, H. G. J. Biol. Chem. 2001, 276, 12546-12555.
(16) Guo, J. T.; Frost, J. W. J. Am. Chem. Soc. 2002, 124, 10642-10643.
(17) Arakawa, K.; Mu¨ller, R.; Mahmud, T.; Yu, T. W.; Floss, H. G. J. Am.
Chem. Soc. 2002, 124, 10644-10645.
(26) Sheldon, P. J.; Johnson, D. A.; August, P. R.; Liu, H. W.; Sherman, D. H.
J. Bacteriol. 1997, 179, 1796-1804.
(18) Guo, J. T.; Frost, J. W. J. Am. Chem. Soc. 2002, 124, 528-529.
(19) Eads, J. C.; Beeby, M.; Scapin, G.; Yu, T. W.; Floss, H. G. Biochemistry
1999, 38, 9840-9849.
(20) Rickards, R. W.; Rukachaisirikul, V. Aust. J. Chem. 1987, 40, 1011-1015.
(21) Kirsch, E. J.; Korshalla, J. D. J. Bacteriol. 1964, 87, 247-255.
(22) Bezanson, G. S.; Vining, L. C. Can. J. Biochem. 1971, 49, 911-918.
(23) Hornemann, U; Cloyd, J. C. J. Chem. Soc., Chem. Commun. 1971, 301-
302.
(24) Hornemann, U.; Eggert, J. H. J. Antibiot. 1975, 28, 841-843.
(25) Wakaki, K.; Huromo, H.; Tomioka, K.; Shimizu, G.; Kato, E.; Kamada,
H.; Kudo, S.; Fujimoto, Y. Antibiot. Chemother. 1958, 8, 228-240.
(27) Martin, T. W.; Dauter, Z.; Devedjiev, Y.; Sheffield, P.; Jelen, F.; He, M.;
Sherman, D. H.; Otlewski, J.; Derewenda, Z. S.; Derewenda, U. Structure
2002, 10, 933-942.
(28) August, P. R.; Rahn, J. A.; Flickinger, M. C.; Sherman, D. H. Gene 1996,
175, 261-267.
(29) Belcourt, M. F.; Penketh, P. G.; Hodnick, W. F.; Johnson, D. A.; Sherman,
D. H.; Rockwell, S.; Sartorelli, A. C. Proc. Natl. Acad. Sci. U.S.A. 1999,
96, 10489-10494.
(30) Varoglu, M.; Mao, Y. Q.; Sherman, D. H. J. Am. Chem. Soc. 2001, 123,
6712-6713.
9
J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007 6471

A R T I C L E S
Gru¨schow et al.
were recorded in positive ion mode. Varian Mercury 300 MHz or
Bruker DRX 500 MHz instruments were used to record NMR spectra.
Spectra were calibrated to residual solvent at 1.93 ppm for acetonitrile
or 2.52 ppm for DMSO. A Beckman-Coulter DU 640 spectrophotometer
was used for the determination of extinction coefficients. Absorption
bands below 230 nm were not considered.
final concentration of 1 mg/mL. Aliquots of the crude extract (20 µL)
were analyzed by LC-MS as described above.
S. laVendulae strain DHS5349 (mitA::tsr) was used for complemen-
tation with 4 and 5. For mitomycin production, 10 mL of Nishikohri
medium was inoculated with 2.5% starter culture and incubated with
shaking at 30 °C. Individual cultures were supplemented with 0.1 M
solutions of either 4 or 5 in DMSO, or with DMSO alone after 72, 96,
100 h (20 µL each). Cells were removed after 115 h of culture time,
and the broth was extracted with ethyl acetate without adjusting the
pH. Analysis of the extracts was performed as described above.
Cloning, Protein Expression, and Purification. The mmcR gene
was PCR-amplified from S. laVendulae NRRL 2564 genomic DNA
using the primers 5′-atcatctcatatgaccgtggagcagacccc-3′ and 5′-attcaagct-
tcaggccctgcggatctcg-3′ to introduce NdeI and HindIII restriction sites,
respectively. The digested PCR product was ligated into pET-28b(+)
(Novagen, Madison, WI) to allow for expression of MmcR as a His-
tagged fusion protein. Sequencing of the construct revealed the
following deviations from the reported DNA sequence: deletion of
bases 411-415 and of base 460 of the mmcR gene (Pro138-Arg153 to
Asp138-Glu151), and transition of base 1013 (Cys338 to Tyr336). Repeated
cloning of the gene confirmed the deletions but revealed the transition
to be a PCR error. The mmcR gene sequence has been revised in the
original database entry (AF127374). Expression of the Cys338Tyr
mutant resulted in soluble but inactive enzyme. The expression construct
containing the wild type mmcR sequence was obtained through PCR
amplification from genomic DNA using Pfu as the DNA polymerase.
His-tagged MmcR was expressed in E. coli BL21(DE3) and purified
to apparent homogeneity in one step using Ni-NTA resin. The enzyme
was transferred into storage buffer containing 20% glycerol using PD-
10 columns (Amersham Biosciences). Protein concentrations were
determined by the Bradford method using BSA as the standard.
Approximately 20 mg of His6-MmcR was isolated from a 0.5 L culture.
Enzyme Assays and Kinetics. For conversion assays, the reaction
mixture contained 5 µM MmcR, 1.5 mM substrate (4 or 5), and 0.25
mM S-adenosylmethionine (AdoMet) in 75 mM sodium phosphate
buffer at pH 7.4. As a negative control, the enzyme was omitted. After
2 h of incubation at 30 °C, an equal volume of 1 M sodium citrate (pH
6) was added and the product was extracted with ethyl acetate. The
products were analyzed by LC-MS as described above.
For kinetic analysis, the reaction mixture contained 1 µM MmcR
and 0.3 mM S-[methyl-14C]-adenosylmethionine (26 µCi/mL) in 75 mM
sodium phosphate (pH 7.4). The enzyme solution was equilibrated at
30 °C, and the reaction was started through the addition of 0.05-5
mM 4, or 0.1-10 mM 5. For the determination of kinetic parameters
for the demethylation of AdoMet, a fixed concentration of 4 was used
(5 mM) and the concentration of AdoMet was varied from 1.5 to 300
µM. For the investigation of reaction conditions, additives were
introduced from a 10X stock solution and with saturating substrate
concentrations. The reactions were terminated by the addition of 1 M
sodium citrate (pH 6) within the range of linear product formation,
and the products were extracted as described above. Products were
resolved on TLC (silica 60 F254, 15% MeOH in dichloromethane). The
TLC plate was exposed to a phosphoimager screen overnight alongside
standards. Radiolabeled products were detected using a Typhoon 9410
variable mode imager and quantified using ImageQuant TL software
(both Amersham Biosciences). Curves were fitted using Prism 4.0a
software.
Bacterial Strains, Culture Conditions, and DNA Manipulation.
E. coli XL1-Blue and BL21(DE3) (Novagen, Madison, WI) were grown
at 37 °C in Luria-Bertani broth or on Luria-Bertani agar supplemented
with 50 µg/mL of kanamycin. Strains of S. laVendulae were cultured
in R2YE broth without antibiotics unless otherwise stated. For
mitomycin production, S. laVendulae was cultured in Nishikohri
medium as previously described.9,31 Standard methods for DNA
isolation and manipulation were performed as described by Sambrook
et al.32 S. laVendulae genomic DNA was isolated using the Wizard
Genomic DNA Purification Kit (Promega, Madison, WI) according to
the manufacturer’s instructions. PCR was carried out in an iCycler
Thermal Cycler (Bio-Rad, Hercules, CA) with PfuTurbo DNA poly-
merase (Stratagene, La Jolla, CA). Restriction enzymes and T4 DNA
ligase were purchased from New England Biolabs (Ipswich, MA) and
used according to the manufacturer’s recommendations. DNA sequenc-
ing was performed by the University of Michigan DNA Sequencing
Core facility using an ABI Model 3730 sequencer.
Preparation of mmcR Deletion Mutant. S. laVendulae DHS9505
was prepared in a two-step procedure Via an insertional mutation in
the mmcR gene. The target gene and flanking sequences were subcloned
into pUC119 and a 1.4 kb ApaL1-HindIII fragment from pFD66633
containing the aphII gene for kanamycin resistance was inserted into
an EcoNI site within mmcR. The insert was transferred into pKC113934
and conjugated into wild type S. laVendulae.35 Transconjugants were
selected on AS1 plates,36 overlaid with apramycin, kanamycin, and
nalidixic acid, and followed by propagation on R5T37 plates at 37 °C
for several generations. Insertional disruption mutants were selected
based on the phenotype changing from apramycin- and kanamycin-
resistant to apramycin-sensitive and kanamycin-resistant. For the
preparation of the mmcR deletion construct, flanking DNA was PCR-
amplified using the following primers: 5′-gcgaattcgtagggggtgccctc-3′
introducing an EcoRI site and 5′-gctctagaggtcatacgggggttcct-3′ intro-
ducing an XbaI site for the 5′-flanking region, and 5′-gctctagatgaaac-
cgcccctcctga-3′ introducing an XbaI site and 5′-gcaagcttctcgagcaccgct-
tggctctcctccctgcc-3′ introducing a HindIII site for the 3′-flanking
region. After digestion with the appropriate restriction enzymes, the
inserts were ligated into pUC119 and transferred into pKC1139 for
conjugation into the S. laVendulae mmcR::aphII strain as described
above. Deletion mutants were selected based on the phenotype
changing from apramycin and kanamycin resistant to apramycin- and
kanamycin-sensitive. The in-frame deletion of mmcR from the chromo-
some was verified by Southern blot hybridization (see Supporting
Information).
Analysis of mmcR Deletion Mutant and Chemical Complemen-
tation. S. laVendulae NRRL 2564 and DHS9505 were cultured in R2YE
for 3 days. This starter culture was used to inoculate 50 mL of
Nishikohri medium, and the main culture was incubated at 30 °C for
4 days. The pH of the fermentation broth was adjusted through addition
of an equal volume of 1 M sodium citrate (pH 6) before extraction
with ethyl acetate. The organic solvent was removed, and the residue
was dissolved in methanol/10 mM ammonium acetate 1:1 to give a
Preparation of 7-Demethylmitomycin A (4) and 7-Demethylmi-
tomycin B (5). The syntheses of 4 and 5 were adapted from Garrett38
and Beijnen et al.39 as described below.
7-Demethylmitomycin A (4): A solution of 10 mg/mL mitomycin
C (1, 51 µmol) in 50 mM aqueous NaOH was stirred at room
temperature for 4 h. Unreacted 1 was removed by extraction with two
(31) Nishikohri, K.; Fukui, S. Eur. J. Appl. Microbiol. 1975, 1, 129-145.
(32) Sambrook, J.; Russel, D. W. Molecular Cloning - A Laboratory Manual,
3rd ed.; Cold Spring Harbor Laboratory Press: Woodbury, NY, 2001.
(33) Denis, F.; Brzezinski, R. Gene 1992, 111, 115-118.
(34) Bierman, M.; Logan, R.; O’Brien, K.; Seno, E. T.; Rao, R. N.; Schoner,
B. E. Gene 1992, 116, 43-49.
(35) Mao, Y. Q.; Varoglu, M.; Sherman, D. H. J. Bacteriol. 1999, 181, 2199-
2208.
(36) Baltz, R. H. DeV. Ind. Microbiol. 1980, 21, 43-54.
(37) Mazodier, P.; Petter, R.; Thompson, C. J. Bacteriol. 1989, 171, 3583-
3585.
(38) Garrett, E. R. J. Med. Chem. 1963, 6, 488-501.
(39) Beijnen, J. H.; Denhartigh, J.; Underberg, W. J. M. J. Pharm. Biomed.
Anal. 1985, 3, 59-69.
9
6472 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007

Hydroxyquinone O-Methylation
A R T I C L E S
volumes of ethyl acetate (five times). The reaction mixture was then
adjusted to pH 5-6 by addition of 0.1 M aqueous citric acid, the product
was extracted four times with two volumes of ethyl acetate each, and
the solvent was removed under a stream of nitrogen gas to give 84%
1
of the title compound 4. H NMR (300 MHz, CD3CN) δ 1.93 (s, 3H,
6-CH3), 2.77 (dd, J ) 1.9, 4.4, 1 H, H2), 2.88 (d, J ) 4.4, 1H, H1),
3.16 (s, 3H, 9a-OCH3), 3.43 (dd, J ) 1.6, 12.9, 1H, H3′), 3.51 (dd, J
) 4.4, 10.7, 1H, H9), 3.98 (d, J ) 12.9, 1H, H3), 4.19 (t, J ) 10.7,
1H, H10′), 4.64 (dd, J ) 4.4, 10.7, 1H, H10), 5.28 (br. s, 2H, CONH2).
13C NMR (DEPT135, HMBC; 500/125 MHz, DMSO-d6) δ 7.1 (6-CH3),
31.5 (C2), 35.3 (C1), 42.8 (C9), 49.0 (OCH3), 49.5 (C3), 60.6 (C10),
106.2 (C9a), 110.5 (C8a), 111.3 (C6), 153.4 (C5a), 156.4 (CONH2),
156.5 (C7), 181.4 (C5). TLC (silica 60, 15% methanol in dichlo-
romethane) Rf 0.20. LC-MS tRet 7.7 min; m/z 336.0 (M + H+); MS/
MS m/z 243.1 (100%). HRMS (ESI+) m/z 358.1012 (M + Na+,
C15H17N3O6Na requires m/z 358.1015). UV-vis (MeOH) λmax 330 nm
(ꢀ 8217 M-1 cm-1), 530 nm (ꢀ 523 M-1 cm-1).
7-Demethylmitomycin B (5): A solution of 10 mg/mL mitomycin
B (3, 40 µmol) in 50 mM aqueous NaOH was stirred at room
temperature for 2 h. Unreacted 3 was removed by extraction with two
volumes of ethyl acetate (four times). The pH of the reaction mixture
was then adjusted to pH 6 by addition of 0.1 M aqueous citric acid,
the product was extracted four times with two volumes of ethyl acetate
each, and the solvent was removed under a stream of nitrogen gas to
Figure 2. Analysis of S. laVendulae wild type, mutant, and complemented
extracts. The extracted ion chromatograms for m/z 335, 336, and 350 are
shown.
mitomycin C. A more detailed sequence analysis of MmcR is
given in the Supporting Information.
In ViWo Experiments. An mmcR in-frame deletion mutant
in S. laVendulae (DHS9505) was prepared by first inserting a
kanamycin resistant marker into the gene and subsequently
removing the resistance cassette together with the whole mmcR
gene by a second round of homologous recombination. Extracts
from the resulting mutant strain DHS9505 were analyzed by
LC-MS and found to be devoid of mitomycins 1, 2, and 3,
produced by wild type S. laVendulae. However, upon closer
examination of the DHS9505 metabolite profile, two new
compounds were identified to be accumulated in the mmcR
deletion mutant but were absent in the wild type strain (Figure
2). The major compound (4) had a retention time of 8.0 min,
and the minor compound (5) eluted at 3.9 min. In good
agreement with the expected function of MmcR as the 7-O-
methyltransferase, both compounds exhibited a molecular ion
peak of m/z 336 corresponding to mitomycins 2 or 3 with the
loss of a methyl group. For further examination of the structures
of 4 and 5, MS/MS experiments were performed. Unlike the
molecular ion peaks, the product ions obtained from 4 and 5
were distinct from each other. Fragmentation of the molecular
ion peak of 4 yielded a single product ion with m/z 243
corresponding to loss of carbamate and methanol from the from
the C10 and C9a positions, respectively. Analogous product ions
were also observed for mitomycin A (2) and mitomycin C (1)
that both possess the 9a-methoxy substituent. Hence, the
analytical data were consistent with 4 corresponding to 7-de-
methylmitomycin A (Figure 1). Fragmentation of the molecular
ion peak of 5, on the other hand, gave rise to three product
ions with m/z 247, 257, and 275 corresponding to loss of
carbamate, loss of carbamate and water, and loss of carbamate
and carbon monoxide, respectively. The observed series of
product ions is typical for mitomycins with a 9a-hydroxy
substituent, and an analogous fragmentation pattern was also
observed for mitomycin B (3). The minor metabolite 5 was
therefore assigned as 7-demethylmitomycin B (Figure 1). The
identity of both compounds was further substantiated through
comparison to semisynthetic standards of 4 and 5 (see above)
that matched the LC-MS characteristics of the mitomycin
intermediates obtained from the ∆mmcR mutant strain DHS9505
(Supporting Information). It is also noteworthy that the ratio of
1
give 72% of the title compound 5. H NMR (500 MHz, CD3CN) δ
1.68 (s, 3H, 6-CH3), 2.27 (s, 3H, NCH3), 2.34 (d, J ) 4.8, 1H, H1),
2.43 (dd, J ) 2.0, 4.7, 1H, H2), 3.43 (dd, J ) 2.0, 13.0, 1H, H3′), 3.64
(dd, J ) 3.1, 8.5, 1H, H9), 3.91 (d, J ) 13.0, 1H, H3), 4.34 (dd, J )
8.5, 11.0, 1H, H10′), 4.57 (dd, J ) 3.0 10.9, 1H, H10), 5.17 (br s, 2H,
CONH2). 13C NMR (125 MHz, DMSO-d6) δ 7.3 (6-CH3), 43.0 (NCH3),
43.6 (C2), 43.9 (C9), 48.2 (C1), 48.3 (C3), 59.5 (C10), 100.1 (C9a),
110.3 (C8a), 110.9 (C6), 152.5 (C5a), 156.8 (CONH2), 157.0 (C7),
174.9 (C8), 181.9 (C5). TLC (silica 60, 15% methanol in dichlo-
romethane) Rf 0.08. LC-MS tRet 3.8 min; m/z 336.0 (M + H+); MS/
MS m/z 247.1 (30%), 257.1 (30%), 275.2 (100%). HRMS (ESI+) m/z
358.1011 (M + Na+, C15H17N3O6Na requires m/z 358.1015). UV-vis
(MeOH) λmax ) 332 nm (ꢀ ) 14290 M-1 cm-1), 550 nm (ꢀ ) 1207
M-1 cm-1).
Results
Identification of the Mitomycin 7-O-Methyltransferase.
The mitomycin gene cluster contains three genes encoding
AdoMet-dependent methyltransferases.9 Two of these genes,
mitM and mitN, are located adjacent to one another. They share
significant homology to each other and to other methyltrans-
ferases involved in secondary metabolism. A functional copy
of mitM is required for the production of mitomycin C,9 but so
far, in Vitro experiments with heterologously expressed enzyme
have only revealed an aziridine N-methyltransferase function.30
Given the absence of a methyl group on the aziridine ring of 1,
this finding was surprising and the function of MitM continues
to be under investigation in our laboratory. The third putative
methyltransferase gene, mmcR, was also shown to be required
for the production of 1.9 Its deduced amino acid sequence shows
significant overall similarity to methyltransferases involved in
the biosynthesis of numerous bacterial natural products such
as enediyne and type II polyketide synthase-derived antibiotics,
as well as to plant methyltransferases involved in the biosyn-
thesis of flavonoids. The common structural element to all of
these natural products is the presence of methylated phenols or
methoxy-substituted quinone moieties. This made MmcR a very
good candidate for the installment of the 7-O-methyl group of
mitomycin A that is thought to be the direct precursor of
9
J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007 6473

A R T I C L E S
Gru¨schow et al.
4 to 5 is approximately the same as the ratio for the C9â-
mitomycins 1 and 2 to the C9R-mitomycin 3. The overall
production yields of 4 and 5 in S. laVendulae DHS9505 are
estimated to be in the same range as those of mitomycins in
the wild type organism (between 1 and 4 mg/L culture).
In order to investigate whether 4 and 5 were chemically
competent intermediates in the mitomycin biosynthetic pathway
in ViVo, we performed feeding experiments with a S. laVendulae
strain deficient in mitomycin production. The strain selected
for these experiments (DHS5349) carries an insertional mutation
of the mitA gene.9 MitA (AHBA synthase) is the final enzyme
in the biosynthesis of AHBA and had been shown to be essential
for mitomycin biosynthesis.9,35 The insertional inactivation of
mitA leads to a polar effect on downstream genes that act after
assembly of AHBA. Hence, complementation with AHBA does
not restore mitomycin production in this strain. However, mmcR
should be expressed as part of a distinct transcript separate from
mitA based on analysis of the gene cluster. As had been
previously shown, mitomycins 1, 2, or 3 were absent from the
culture medium of DHS5349.9 However, when the strain was
supplemented with the 9â-configured 4, production of mito-
mycins 1 and 2 was fully restored. Conversely, supplementing
the DHS5349 culture medium with the 9R-configured 5 resulted
in production of 3 as the only mitomycin (Figure 2). These
results demonstrate that 4 and 5 are able to serve as substrates
in ViVo for the biosynthesis of mitomycins with C9â- or C9R-
stereochemistry, respectively. Rather unexpectedly, a fourth
compound (6) was also dependent on AHBA production and
its biosynthesis could only be restored upon addition of 4, but
not 5 to the fermentation medium. Both molecular ion peak
and fragmentation pattern were identical to that of 2, suggesting
6 to be an isomer of 2. Two possible structures, 9-epi-mitomycin
A and albomitomycin A (Figure 1), could be assigned based
on these findings. The former compound has not been previously
isolated; however, the latter product is a known mitomycin
derivative.40 Albomitomycin A forms from 2 in protic solvents
through an intramolecular Michael addition of the aziridine
nitrogen to C5a of the quinone moiety. Furthermore, this
compound is reported to yield the same MS/MS spectrum as
2,40 providing a clear rationale for assignment of 6 as albomi-
tomycin A.
In Vitro Assays. In order to characterize MmcR further we
cloned the corresponding mmcR gene into pET-28b for heter-
ologous expression as a His6-fusion protein. Upon closer
inspection of the gene sequence, deviations from the previously
reported sequence were apparent. As the same base pair changes
occurred from two independent PCR reactions with genomic
DNA as template and with two different DNA polymerases,
the reported mmcR gene sequence requires minor corrections.
The gene sequence of mmcR has been revised in the original
database entry. His-tagged MmcR was expressed in E. coli and
purified to near homogeneity using affinity chromatography.
Compounds 4 and 5 were prepared as authentic standards from
1 and 3, respectively, through base hydrolysis. The separation
of products from starting material was facilitated through the
presence of the enolic hydroxy group in 4 and 5 that is
deprotonated at basic pH. Starting material could thus be
removed through ethyl acetate extraction of the basic reaction
Figure 3. In Vitro assays with heterologously expressed MmcR. Extracted
ion chromatograms for m/z 336 and 350 are shown. A: 4 + AdoMet +
MmcR; B: authentic standard of 2; C: 5 + AdoMet + MmcR; D: authentic
standard of 3.
mixture, and the products were recovered through extraction
of the remaining mixture after neutralization. A semisynthetic
method was chosen over isolation from fermentation extract due
to the instability of 4 and 5 in solution41 and the relatively low
quantities obtained from S. laVendulae DHS9505 cultures.
The proposed activity of MmcR as a 7-O-methyltransferase
was confirmed by incubating 4 with the enzyme in the presence
of AdoMet. Analysis of the reaction mixture revealed the
presence of one major product that matched 2 in retention time,
molecular mass, and MS/MS pattern (Figure 3), and the same
minor product 6 that was also found in the S. laVendulae wild
type extracts (Figure 2). Isolation of 6 from the reaction mixture
initially yielded a colorless compound as expected for albomi-
tomycin A.40 Repeated dissolving of 6 in water or methanol
resulted in the reappearance of the purple color characteristic
for 2. As 2 and 6 exist in equilibrium, our previous assignment
of 6 as albomitomycin A was therefore further substantiated.
With 5 as the substrate, a single product corresponding to 3 in
retention time, molecular ion peak, and MS/MS pattern was
observed (Figure 3). No products were detected when the
enzyme was omitted (data not shown). Furthermore, none of
the other two methyltransferases found in the mitomycin cluster
(MitM and MitN) were capable of methylating 4 (data not
shown).
Next, we determined the kinetic parameters for the MmcR-
catalyzed methylation of 4 and 5 using [14C]-AdoMet. The
conversion of 4 to 2 followed Michaelis-Menten kinetics with
a Km value of 170 ( 25 µM and a kcat value of 0.103 ( 0.004
min-1 with saturating AdoMet (Figure 4A). Using 5 as the
substrate, no saturation was observed up to 10 mM substrate
(Figure 4C). Instead, a biphasic linear increase in the rate of
product formation with increasing substrate concentration was
observed. The specificity constant kcat/Km for the conversion of
5 to 3 was calculated from the slope of the linear regression up
to 0.6 mM substrate concentration and determined to be 0.458
( 0.002 M-1 s-1. This is approximately 20-fold lower than the
specificity constant for the conversion of 4 to 2. In order to
investigate the integrity of the active site, the kinetic parameters
for AdoMet were determined under saturating conditions of 4
(Figure 4B). A Km value of 3.8 ( 0.5 µM for AdoMet was
(40) Kono, M.; Saitoh, Y.; Shirhata, K.; Arai, Y.; Ishii, S. J. Am. Chem. Soc.
1987, 109, 7224-7225.
(41) Beijnen, J. H.; Lingeman, H.; van Munster, H. A.; Underberg, W. J. J.
Pharm. Biomed. Anal. 1986, 4, 275-295.
9
6474 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007

Hydroxyquinone O-Methylation
A R T I C L E S
Figure 4. Activity of the MmcR-catalyzed methylation reactions. (A) Conversion of 4 to 2 with saturating concentration of AdoMet. (B) Conversion of 4
to 2 with saturating concentration of 4. (C) Conversion of 5 to 3 with saturating concentration of AdoMet. (D) Effects of divalent cations on MmcR activity.
obtained, suggesting that the ability of MmcR to bind its cofactor
was not impaired, and hence the integrity of the active site was
unlikely to be affected by the presence of the His6-fusion tag.
The kcat value obtained for AdoMet was 0.110 ( 0.003 min-1
and is in agreement with the value obtained for 4 with saturating
AdoMet.
potentially catalyze methytransfer to a C7 hydroxyl group on
the reduced, dihydroquinone variants of 4 or 5. However, these
compounds are not available for in Vitro assays because of their
very limited half-life.3,4
Discussion
Over the past several years, we have been dissecting the
catalytic function of individual gene products involved in
mitomycin biosynthesis in an effort to assign a specific role to
the more than 40 enzymes believed to be involved in its
assembly, resistance, and regulation.9 The timing and sequence
of biosynthetic steps from AHBA and D-glucosamine to
mitomycins 1, 2, and 3 remain largely unknown and even
difficult to predict due to the unique structure of the mitosane
skeleton. An unusual feature in the biosynthesis of mitomycins
is the presence of two series of products that differ in the
stereochemistry of the C9-substituent. It is not known to what
extent the biosynthetic pathways leading to mitomycins with
C9â-configuration and C9R-configuration are shared or at what
stage the two pathways might diverge. Interestingly, the
difference in C9-configuration coincides with individual me-
thylation patterns. We have identified from in ViVo studies two
methyltransferase genes, mitN and mitM, that are required for
the aziridine-N-methylation of C9R-mitomycins and the 9a-O-
methylation of C9â-mitomycins, respectively (N. Sitachitta et
al., unpublished). This work suggests that methylation of the
right side of the mitosane skeleton is performed by enzymes
dedicated to a single branch of the biosynthetic pathway.
The third methyltransferase gene, mmcR, is required for the
formation of mitomycins bearing C9R- or C9â-configurations.
An mmcR deletion mutant of S. laVendulae was devoid of the
three major mitomycins usually found in the wild type organism
(1, 2, and 3). However, two new natural product metabolites,
7-demethylmitomycin A (4) and 7-demethylmitomycin B (5),
were found to be accumulated in that strain. Mitomycin C (1),
the only mitomycin with a 7-amino substituent, was not
We further investigated the effect of different reaction
conditions on the efficiency of MmcR using 4 as the substrate.
MmcR has a relatively broad pH optimum between 7 and 7.5
(data not shown). Increasing the salt concentration from 50 to
500 mM resulted in a gradual decrease in activity to 60% of
the initial value, whereas the addition of nonionic detergent did
not significantly affect enzymatic activity (data not shown). As
some methyltransferases depend on the presence of divalent
cations for optimal activity,42-44 a series of metals were tested.
Notably, the presence of 1 mM EDTA resulted in a slight but
significantly higher activity, whereas Mg2+ or Ca2+ did not alter
the reactivity of MmcR. Interestingly, both Mn2+ and Fe2+
resulted in a 4-fold decrease of enzyme activity, and the presence
of Zn2+ or Cu2+ completely inhibited the methylation reaction
(Figure 4D). A very similar pattern for the influence of metal
ions on enzyme activity was found for the rifamycin 27-O-
methyltransferase.45 However, no reaction conditions were found
that substantially improved the methylation reaction catalyzed
by MmcR with 4 as the substrate.
All methyltransferases involved in the methylation of aromatic
substrates reported to date act on phenolic compounds. As such
MmcR is the first methyltransferase to accept a hydroxyquinone
as substrate. This also raises the possibility that MmcR could
(42) Shafiee, A.; Motamedi, H.; Chen, T. Eur. J. Biochem. 1994, 225, 755-
764.
(43) Kreuzman, A. J.; Turner, J. R.; Yeh, W. K. J. Biol. Chem. 1988, 263,
15626-15633.
(44) Ibdah, M.; Zhang, X. H.; Schmidt, J.; Vogt, T. J. Biol. Chem. 2003, 278,
43961-43972.
(45) Xu, J.; Mahmud, T.; Floss, H. G. Arch. Biochem. Biophys. 2003, 411, 277-
288.
9
J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007 6475

A R T I C L E S
Gru¨schow et al.
identified in S. laVendulae ∆mmcR extracts, indicating that the
7-methoxy group is required for formation of 1. The production
of C9â-configured (1, 2) and C9R-configured (3) mitomycins
could be selectively restored by supplementing the culture
medium of an AHBA-deficient strain with either 4 or 5,
respectively. These results conclusively demonstrated that
MmcR is responsible for the 7-O-methylation of both C9â- and
C9R-mitomycins, and also that the partitioning of mitomycins
into the C9â- and C9R-branches is largely independent of the
function of this methyltransferase. However, the question
remained whether the main metabolic flux in mitomycin
biosynthesis proceeded through 4 and 5.
In order to address this question, in Vitro experiments with
heterologously expressed MmcR were conducted. The 7-O-
demethylated compounds 4 and 5 were converted to the
expected 7-O-methylated products 2 and 3, respectively. Metal
ions were not required for activity and, in fact, tended to
decrease enzyme efficiency. The conversion of 4 to 2 followed
Michaelis-Menten behavior, whereas no saturation was ob-
served for the analogous reaction with the C9R-configured 5.
The specificity constant obtained for the latter reaction was 20
times lower than for the former, indicating a significant
preference of MmcR for C9â-substituted mitomycins. Methyl-
transferases typically have similar Km values for both the methyl
acceptor and the methyl donor, AdoMet. This is not the case
for the MmcR-catalyzed conversion of 4 to 2. The Km value
for the substrate 4 is 45 times higher than that obtained for
AdoMet and approximately 3-20 times higher than what is
typicalformethyltransferasesinvolvedinsecondarymetabolism.45-49
Moreover, the rate of methylation is relatively slow. Similar
rates have been reported for other methyltransferases,46,47 but
typical rates of methylation are significantly faster.45,48,49
Thus, the late stage biosynthetic scheme for mitomycin
assembly as shown in Scheme 1 is consistent with the results
obtained from mmcR gene deletion and chemical complemen-
tation experiments. However, the kinetic evaluation of MmcR
suggests that optimal enzyme conditions have yet to be
established, another component of the methyltransferase is
required for optimal catalysis, or an alternative mitomycin
precursor is the preferred physiological substrate for MmcR.
In this latter respect, it can be envisaged that methylation of
the hydroxyquinone moiety might occur at an earlier stage in
mitomycin biosynthesis, possibly before branching to the C9R-
and C9â-mitomycins from a common precursor X (Scheme 1).
This hypothesis would be supported by the presence of only
one 7-O-methyltransferase gene in the cluster and the preference
toward substrates with C9â-configuration displayed by heter-
ologously expressed MmcR. Further investigation of additional
biosynthetic steps will shed more light on the assembly of
mitomycin natural products, and on the possible role of earlier
stage intermediates as substrates for MmcR.
Acknowledgment. We thank Kyowa Hakko Kogyo Ltd. for
generous gifts of mitomycin B and mitomycin C. Dr. Courtney
C. Aldrich is gratefully acknowledged for assistance with
substrate synthesis and Dr. Douglas A. Burr for assistance with
NMR. This study was supported by NIH grant R01 CA81172
and the J. G. Searle Professorship to D.H.S.
Supporting Information Available: Mass spectrometry data
for 1-6, Southern blot analysis, MmcR sequence comparison.
This material is available free of charge via the Internet at
http://pubs.acs.org.
(46) Freel Meyers, C. L.; Oberthu¨r, M.; Xu, H.; Heide, L.; Kahne, D.; Walsh,
C. T. Angew. Chem., Int. Ed. 2004, 43, 67-70.
JA0700193
(47) Lavid, N.; Wang, J.; Shalit, M.; Guterman, I.; Bar, E.; Beuerle, T.; Menda,
N.; Shafir, S.; Zamir, D.; Adam, Z.; Vainstein, A.; Weiss, D.; Pichersky,
E.; Lewinsohn, E. Plant Physiol. 2002, 129, 1899-1907.
(48) Pacholec, M.; Tao, J.; Walsh, C. T. Biochemistry 2005, 44, 14969-14976.
(49) Wang, J.; Pichersky, E. Arch. Biochem. Biophys. 1999, 368, 172-180.
9
6476 J. AM. CHEM. SOC. VOL. 129, NO. 20, 2007