Synthesis of potential early-stage intermediates in the biosynthesis of FR900482 and mitomycin c

Article abstract of DOI:10.1021/ol802631c

Beyond the identification of 3-amino-5-hydroxybenzoic acid (AHBA) and D-glucosamine as biosynthetic precursors to mitomycin C (5) and FR900482 (6), little is known about the pathway Nature uses to prepare these antitumor antibiotics. To gain some insight

Full text of DOI:10.1021/ol802631c

ORGANIC
LETTERS
2009
Vol. 11, No. 4
791-794
Synthesis of Potential Early-Stage
Intermediates in the Biosynthesis of
FR900482 and Mitomycin C
Stephen Chamberland,† Sabine Gru¨schow,‡ David H. Sherman,‡ and
Robert M. Williams*,†,§
Department of Chemistry, Colorado State UniVersity, Fort Collins, Colorado 80523,
Life Sciences Institute and Departments of Medicinal Chemistry, Chemistry,
Microbiology & Immunology, The UniVersity of Michigan, 210 Washtenaw AVenue,
Ann Arbor, Michigan 48109-2216, and UniVersity of Colorado Cancer Center,
Aurora, Colorado 80045
rmw@lamar.colostate.edu
Received November 14, 2008
ABSTRACT
Beyond the identification of 3-amino-5-hydroxybenzoic acid (AHBA) and D-glucosamine as biosynthetic precursors to mitomycin C (5) and
FR900482 (6), little is known about the pathway Nature uses to prepare these antitumor antibiotics. To gain some insight into their biosynthesis,
amino acids 1 and 2 as well as C-2 N-acetylated derivatives 3 and 4 were prepared. Preparation of these putative biosynthetic intermediates
and N-acetylcysteamine thioester analogues 28 and 29 should enable confirmation of their involvement in FR900482 and mitomycin C biosynthesis.
Mitomycin C (MMC, Figure 1) has been employed success-
fully as a cancer chemotherapeutic for nearly three decades,
but use of this drug often causes harmful side effects that
are associated with hematotoxicity.1-3 One of the modes of
action by which MMC displays cytotoxicity is through a
bioreductive DNA alkylation and an interstrand cross-linking
reaction, a process that has been thoroughly studied.3 A more
recently isolated structural relative of MMC is FR900482,
which has been shown to exhibit even more potent biore-
ductive mitosene-based DNA cross-linking activity than
MMC.4,5 In contrast to MMC, FR900482 and several
semisynthetic relatives (such as FK317) lack the quinone
moiety responsible for the production of adventitious and
† Colorado State University.
‡ The University of Michigan.
§ University of Colorado Cancer Center.
Figure 1. Structures of the mitomycins, FR900482, and congeners.
(1) Rajski, S. R.; Williams, R. M. Chem. ReV. 1998, 98, 2723–2795
.
(2) Denny, W. A. In Cancer Chemotherapeutic Agents; Foye, W. A.,
Ed.; American Chemical Society: Washington, DC, 1995; pp 491-493.
harmful superoxide during bioreductive MMC activation.6,7
We have also demonstrated that FR900482 and its derived
(3) Tomasz, M. Chem. Biol. 1995, 2, 575–579
.
(4) Wolkenberg, S. E.; Boger, D. L. Chem. ReV. 2002, 102, 2477–2495
.
10.1021/ol802631c CCC: $40.75
Published on Web 01/22/2009
2009 American Chemical Society

semisynthetic relative, FK317, cross-link the HMG-A1
oncoproteins to chromosomal DNA in human cells.5,8 The
interesting biological activity and unique structures of these
agents have also attracted considerable interest in the total
synthesis of FR900482 and derivatives.9
Scheme 1. Unified Biosynthesis of FR900482 and MMC
In addition to shared aspects of their mechanism of
biological activity, the structural similarity between MMC
and FR900482 suggests that these antitumor antibiotics share
a common biosynthetic origin. Indeed, biogenetic studies
revealed that these agents originate from 3-amino-5-hydroxy-
benzoic acid (AHBA) and D-glucosamine (Figure 2).10
Figure 2. Incorporation of AHBA and D-glucosamine into mito-
mycin C and FR900482.
Beyond these early investigations, little is known regarding
the details of how these simple precursors are assembled into
the complex architectures characteristic of these structurally
unique agents.11
jugate (7).15 Once this substance is formed, N-glycosylation
of activated D-glucosamine derivative 8 with ACP-bound
amine 7 may afford the first committed biosynthetic inter-
mediate, N-glycoside 9.11 Another enzyme critical for
mitomycin biosynthesis, MitB, has been implicated as the
glycosyl transferase responsible for the in vivo assembly of
N-glycoside 9.12,13 Next, ring-opening and reduction of
N-glycoside 9 would afford the corresponding amine 10.12,13
An alternative pathway for joining the two halves may
involve the biosynthetic equivalent of an intermolecular
Friedel-Crafts alkylation of ACP-ligated AHBA with a
suitably activated derivative of D-glucosamine, but this
possibility appears unlikely.12,13 Of particular intrigue in the
overall biogenesis of MMC and FR900482 is the striking
difference in the stereochemistry of the C9 (MMC number-
A unified biogenesis for the formation of MMC and
FR900482 has been partially formulated by Sherman and
co-workers and is further elaborated herein (Scheme 1).12
A preliminary step in the biogenesis of MMC and FR900482
likely involves MitE, an acyl AMP ligase, and MmcB, a
discrete acyl carrier protein.13,14
We propose that MitE catalyzes the reaction between
AHBA-AMP and MmcB to provide an AHBA-ACP con-
(5) Williams, R. M.; Rajski, S. R.; Rollins, S. B. Chem. Biol. 1997, 4,
127–137.
(6) Masuda, K.; Makamura, T.; Shimomura, K.; Shibata, T.; Terano,
H.; Kohsaka, M. J. Antibiot. 1988, 41, 1497–1499.
(7) Sartorelli, A. C.; Pristos, C. A. Cancer Res. 1986, 46, 3528–3532.
(8) (a) Beckerbauer, L.; Tepe, J. J.; Eastman, R. A.; Mixter, P.; Williams,
R. M.; Reeves, R. Chem. Biol. 2002, 8, 427–441. (b) Beckerbauer, L.; Tepe,
J. J.; Cullison, J.; Reeves, R.; Williams, R. M. Chem. Biol. 2000, 7, 805–
812.
(11) Namiki, H. N.; Chamberland, S.; Gubler, D. A.; Williams, R. M.
Org. Lett. 2007, 9, 5341–5344.
(9) For total syntheses of FR900482, see: (a) Suzuki, M.; Kambe, M.;
Tokuyama, H.; Fukuyama, T. J. Org. Chem. 2004, 69, 2831–2843. (b) Judd,
T. C.; Williams, R. M. J. Org. Chem. 2004, 69, 2825–2830, and references
cited therein.
(12) Sherman, D. H.; Mao, Y.; M.; Varoglu, M.; He, M.; Sheldon, P.
Mitomycin biosynthetic gene cluster. U.S. Patent 6,495,348, December 17,
2002.
(10) (a) Floss, H. G. Nat. Prod. Rep. 1997, 14, 433–452. (b) Fujita, T.;
Takase, S.; Otsuka, T.; Terano, H.; Kohsaka, M. J. Antibiot. 1988, 41, 392–
394. (c) Kibby, J. J.; Rickards, R. W. J. Bacteriol. 1981, 34, 605–607. (d)
Anderson, M. G.; Kibby, J. J.; Rickards, R. W.; Rothschild, J. M. J. Chem.
Soc., Chem. Commun. 1980, 1277–1278. (e) Hornemann, U.; Eggert, J. H.
J. Antibiot. 1975, 28, 841–843. (f) Hornemann, U.; Kehrer, J. P.; Nunez,
C. S.; Ranieri, R. L. J. Am. Chem. Soc. 1974, 96, 320–322. (g) Hornemann,
U.; Aikman, M. J. J. Chem. Soc., Chem. Commun. 1973, 88–89. (h)
Hornemann, U.; Cloyd, J. C. J. Chem. Soc., Chem. Commun. 1971, 301–
302.
(13) (a) Mao, Y.; Varoglu, M.; Sherman, D. H. Chem. Biol. 1999, 6,
251–263. (b) Mao, Y.; Varoglu, M.; Sherman, D. H. J. Bacteriol. 1999,
181, 2199–2208.
(14) Sherman, D. H. Personal communication, 2007.
(15) Sherman, D. H.; Li, S. Unpublished results, 2007.
(16) (a) Kalmanowa, M.; Sokolowski, J. Rocz. Chem., Ann. Soc. Chim.
Pol. 1976, 50, 1175–1189. (b) Bertho, A.; Koziollek, D. Chem. Ber. 1959,
92, 627–637. (c) Bertho, A.; Koziollek, D. Chem. Ber. 1954, 87, 934–940.
(d) Weygand, F.; Perkow, W.; Kuhner, P. Chem. Ber. 1951, 84, 594–602.
(e) Kuhn, R.; Strobele, R. Chem. Ber. 1937, 70B, 773–787.
792
Org. Lett., Vol. 11, No. 4, 2009

ing) and C7 (FR900482 numbering) benzylic stereocenter
in MMC and FR900482, which has potentially powerful
biosynthetic implications.
To confirm whether the more plausible N-glycosylation
is responsible for the union of precursors 7 and 8, we
prepared AHBA/D-glucosamine adduct 1 (Scheme 2). Lit-
Scheme 3
.
Attempts To Prepare SNAC Thioester Analogues 21
and 26 of Amines 13 and 15, Respectively
Scheme 2
.
Synthesis of Putative Biosynthetic Intermediates 1
and 2
isolate the pure product. Assembly and deprotection of
N-Boc-protected and reduced amine 28 readily provided the
free amine 23; however, acid-sensitive N-glycoside 26 was
incompatible with standard conditions employed to deprotect
the Boc-carbamate, as well as a plethora of alternative
conditions that were explored (Scheme 5). To circumvent
this problem, an alternative approach was necessary; namely,
coupling amino acid 1 to N-acetylcysteamine using DCC
afforded the desired thioester 22, albeit in low yield (Sch-
eme 4).
erature precedent for the synthesis of unprotected N-aryl
glycosides is extremely limited.16 Nonetheless, these methods
were amenable to the preparation of compound 1. In practice,
condensation of N-Cbz-D-glucosamine (17)17 and AHBA18
under Brønsted acid catalysis afforded N-glycoside 18
exclusively in the ꢀ-configuration (Scheme 2). Subsequent
cleavage of the N-Cbz moiety using catalytic hydrogenolysis
with palladium(II) oxide cleanly provided the requisite amino
acid 1.19 Putative biosynthetic intermediate 2 was prepared
in a two-step, one-flask N-glycosylation/reductive amination
reaction sequence enlisting N-Cbz-D-glucosamine (17) and
AHBA·HCl in the presence of sodium cyanoborohydride to
provide amine 19.20 Unmasking the amine functionality using
catalytic hydrogenolysis provided amino acid 2 in a fashion
analogous to that applied to the preparation of amine 1.18
To facilitate chemoenzymatic analysis of the proposed
ACP-complexed biosynthetic intermediates with cloned,
functionally expressed biosynthetic enzymes MitB, MitE, and
putative reductase MitF, we prepared N-acetylcysteamine
(SNAC) thioester analogues of compounds 1 and 2 (Scheme
3).12,13,21 Although SNAC ester derivatives of Cbz-protected
N-aryl glycosides 18 and 19 could be prepared, the thioester
moiety was incompatible with standard Pd- or Pt-mediated
hydrogenolysis conditions used to deprotect the Cbz group.
Furthermore, repeated efforts to reveal the C-2 amino
functionality in analogues possessing a photolabile Nvoc (6-
nitroveratryl) carbamate were similarly unsuccessful.
Scheme 4. SNAC Thioester Preparation
In addition to complex N-aryl glycosides of 2-amino-2-
deoxyglucose, we considered biosynthetic intermediates
bearing an acetyl residue on the C-2 nitrogen atom. Recent
studies on the biosynthesis of the antibiotics teicoplanin and
butirosin suggest that a common strategy for incorporating
(17) van der Berg, R. J. B. H.; Noort, D.; van der Marel, G. A.; van
Boom, J. H.; Benschop, H. P. J. Carbohydr. Chem. 2002, 21, 167–188.
(18) Herlt, A. J.; Kibby, J. J.; Rickards, R. W. Aust. J. Chem. 1981, 34,
1319–1324.
Because purification and handling of these polar reaction
products proved to be quite onerous, we found in the tert-
butoxycarbonyl (Boc) group a potential masking agent for
the C-2 amino functionality that could undergo facile
cleavage without the production of byproducts that require
separation, thereby obviating extensive manipulation to
(19) Brieskorn, C. H.; Hamm, R. Arch. Pharm. (Weinheim) 1986, 319,
673–682.
(20) Kolter, T.; Dahl, C.; Giannis, A. Liebigs Ann.Chem. 1995, 625–
629.
(21) (a) Ehmann, D. E.; Trauger, J. W.; Stachelhaus, T.; Walsh, C. T.
Chem. Biol. 2000, 7, 765–772. (b) Yue, S.; Duncan, J. S.; Yamamoto, Y.;
Hutchinson, C. R. J. Am. Chem. Soc. 1987, 109, 1253–1255. (c) Cane, D. E.;
Yang, C.-C. J. Am. Chem. Soc. 1987, 109, 1255–1257.
Org. Lett., Vol. 11, No. 4, 2009
793

Scheme 5
.
Attempts To Prepare SNAC Thioesters 22 and 23
from N-Boc-D-glucosamine (24)
Scheme 6
.
Preparation of SNAC Thioester Analogues 36 and 37
Incorporating N-Acetyl-D-glucosamine
FR900482 and the mitomycins in Streptomyces spp.13a We
are currently examining whether compound 22 or 30 is a
product of MitB catalysis and if compound 23 or 31,
respectively, follows in the biosynthetic pathway via MitF.
In conclusion, we have prepared several potential sub-
strates believed to be early biosynthetic intermediates to
FR900482 and the mitomycins in a straightforward fashion.
Along with enabling ongoing biosynthetic studies, the
availability of these putative biosynthetic intermediates may
reveal additional, unforseen aspects of the biochemical
production of mitomycin C and FR900482. For example,
access to the various substrates described herein and
more advanced potential biosynthetic intermediates may
elucidate the role of enzymes implicated in the pathway, but
whose function has not yet been ascertained. Studies to assess
these issues are in progress, and we will report our findings
in due course.
a glucosamine moiety into secondary metabolites utilizes the
abundant primary metabolite uridine diphospho-N-acetyl-D-
glucosamine (UDP-GlcNAc).22 Consequently, we have reas-
sessed the function of early MMC biosynthetic enzyme MitC
and now believe this protein may be involved in deacetylation
of an early adduct of AHBA and D-glucosamine.14,21 These
findings, in conjunction with the 30% sequence identity found
at the C-terminus of putative mitomycin glycosyltransferase
MitB with UDP-GalNAc-polypeptide N-acetylgalactosami-
nyltransferase, stimulated our desire to analyze N-acetylated
sugar derivatives.13b The presence of an N-acetyl group at
the C-2 position prior to biosynthetic N-glycosylation is also
chemically sound, since neighboring group participation can
increase the rate of synthetic glycosylation reactions. Scheme
6 illustrates the facile approach used to prepare 2-N-acetyl-
2-deoxy-D-glucosamine analogues.
Acknowledgment. We gratefully acknowledge the Na-
tional Institutes of Health (CA51975). Mass spectra were
obtained on instruments supported by NIH Shared Instru-
mentation Grant GM49631. We dedicate this paper to
Thomas A. and Margaret A. Vanderslice Professor of
Chemistry T. Ross Kelly of Boston College on the occasion
of his 65th birthday. We are grateful to Dr. John Greaves of
the University of California, Irvine, for mass spectrometric
data.
The requisite substrates 3 and 4 were prepared as detailed
in Scheme 6 and converted into their corresponding SNAC
esters 30 and 31 in synthetically useful yields.
To establish whether amino acids 1 and 2 or N-acetylated
analogues 3 and 4 are biosynthetic intermediates leading to
the production of mitomycin C and FR900482, we must now
assess how these potential substrates interact with the
biosynthetic machinery of the respective Streptomyces spp.,
the organisms from which these antitumor agents were
isolated. Sherman and co-workers have identified many of
the genes believed to be responsible for the biosynthesis of
Supporting Information Available: Experimental pro-
cedures and characterization of all new compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(22) (a) Truman, A. W.; Huang, F.; Llewellyn, N. M.; Spencer, J. B.
Angew. Chem., Int. Ed. 2007, 46, 1462–1464. (b) Truman, A. W.; Robinson,
L.; Spencer, J. B. ChemBioChem 2006, 7, 1670–1675.
OL802631C
794
Org. Lett., Vol. 11, No. 4, 2009