This document summarizes the current status of research on enediyne natural products, which are a class of antitumor antibiotics with an unusual bicyclo or biocyclo core structure. It reviews methods for cloning the biosynthetic gene clusters responsible for enediyne production. It then presents a unified paradigm for enediyne biosynthesis involving polyketide synthase enzymes, accessory enzymes, and tailoring enzymes. Finally, it discusses strategies for applying combinatorial biosynthesis techniques to engineer novel enediyne analogs with the goal of developing new antitumor agents.

1. Current Medicinal Chemistry, 2003, 10, 2317-2325 2317
0929-8673/03 $41.00+.00 © 2003 Bentham Science Publishers Ltd.
Enediyne Natural Products: Biosynthesis and Prospect Towards Engineering
Novel Antitumor Agents
Ben Shen1,2*, Wen Liu1 and Koichi Nonaka1
1Division of Pharmaceutical Sciences and 2Department of Chemistry, University of
Wisconsin, Madison, WI 53705, USA
Abstract: This review gives a brief account on the current status of enediyne
biosynthesis and the prospective of applying combinatorial biosynthesis methods to
the enediyne system for novel analog production. Methods for cloning enediyne
biosynthetic gene clusters are first reviewed. A unified paradigm for enediyne
biosynthesis, characterized with (a) the enediyne PKS, (b) the enediyne PKS accessory
enzymes, and (c) tailoring enzymes, is then presented. Strategies and tools for novel
enediyne analog production by combinatorial biosynthesis are finally discussed. The
results set the stage to decipher the molecular mechanism for enediyne biosynthesis and
lay the foundation to engineer novel enediynes by combinatorial biosynthesis for
future endeavors.
Keywords: Biosynthesis, C-1027, calicheamicin, combinatorial biosynthesis, enediyne, polyketide synthase.
INTRODUCTION
Neocarzinostatin (NCS), the first member of the enediyne
family of antitumor antibiotics, was originally discovered as
a macromolecular antitumor antibiotic from the culture
filtrates of a Streptomyces carzinostaticus strain in 1965 [1].
Although it became clear shortly after its discovery that all
biological activities of NCS resided in a nonprotein
chromophore, the NCS chromophore structure (1) was not
elucidated until 1985 [2], revealing an unprecedented
bicyclo[7,3,0]dodecadiynene system, i.e., the 9-membered
enediyne core, decorated with an aminosugar and a naphthoic
acid moiety. However, this seminal work was
underappreciated in the subsequent two years, and it was the
discovery of the calicheamicins (2) from a Micromonospora
echinospora strain [3,4] and the esperamicins (3) from an
Actinomadura verrucosospora strain [5,6] in 1987 that
grabbed the attention of chemists and biologists alike to the
enediynes. In contrast to 1, structural elucidation of 2 and 3
unveiled a novel biocyclo[7,3,1]tridecadiynene system, i.e.,
the 10-membered enediyne core, decorated with several
deoxysugar moieties. Since then, the enediyne family of
natural products has been the focus of intense research
activity in the fields of chemistry, biology, and medical
sciences because of their highly unusual molecular
architectures, biological activities, and modes of actions [7-
13].
Over twenty enediyne natural products are currently
known, and new members of the enediyne family are
continuously to be discovered with the newest addition, the
shishijimicins (4) from a marine ascidian Didemnum
proliferum species, reported in early 2003 [14]. The
enediyne natural products could be classified into two sub-
*Address correspondence to this author at the Division of Pharmaceutical
Sciences, School of Pharmacy, University of Wisconsin-Madison, 777
Highland Ave., Madison, WI 53705, USA; Tel: +1-608-263-2673; Fax:
+1-608-262-5345; E-mail: bshen@pharmacy.wisc.edu
categories according to the enediyne core structures.
Members of the 9-membered enediyne core sub-category are
chromoproteins consisting of an apo-protein and the
enediyne chromophore, with N1999A2 (5) from
Streptomyces sp. AJ9493 as the only exception that was
isolated as a chromophore alone [15]. The apo-protein acts as
a stabilizer and specific carrier for the otherwise unstable
chromophore and its transport and interaction with target
DNA. Due to their intrinsic reactivity, only three other
chromoprotein chromophore structures, in additional to 1
and 5, are currently known—kedarcidin (6) from
Actinomycete L585-6 [16], C-1027 (7) from Streptomyces
globisporus [17-20], maduropeptin (8) from Actinomadura
madurae [21]. Members of the 10-membered enediyne core
sub-category are in general more stable, all of which were
isolated as discrete small molecules. In addition to 2, 3, 4,
other members of this sub-category whose structures have
been elucidated include dynemicin (9 ) from
Micromonospora chersina sp. nov. No. M965-1 [22,23] and
namenamicin (10) from a marine ascidian Polysyncraton
lithostrotum species [24] (Figure 1).
Members of both sub-categories of enediynes share a
common mechanism of action, despite their structural
difference. The enediyne core undergoes an electronic
rearrangement (Bergman or Myers rearrangement) to form a
transient benzenoid diradicals that damage DNA by
abstracting hydrogen atoms from the deoxyribose moiety on
both strands. Subsequent reactions of the resultant
deoxyribose carbon-centered radicals with molecular oxygen
initiate a process that leads in both single-strand and double-
strand DNA cleavage [7-13,25-27]. This novel mechanism of
DNA damage has important implications for the
development of the enediynes into clinical anticancer drugs
[25-36]. Although the natural enediynes have seen limited
use as clinical drugs mainly because of substantial toxicity,
various polymer-based delivery systems or enediyne-
antibody conjugates have shown great clinical success and/or
promise in anticancer chemotherapy [8,10,12,13,37,38]. For

2. 2318 Current Medicinal Chemistry, 2003, Vol. 10, No. 21 Shen et al.
O
O
O
O
O
CH3HN
O
HO
CH3
HO
O
O
OCH3
H3C
OH
O
O
O
H2C
O
O
N
OCH3
R
Cl
NH2
O
O
OH
OH
(CH3 )2 N
H3C
CH3
O
O
HO
N
H
Cl
O
H3 CO
O
OH
NH
CH3
OH
O
OHO O
CH3HO
H3C
OO
OOO
N
O
Cl
O
H3C
N(CH3)2
OH
O
OH
H3C
HO
H3C
i-PrO
H3CO
OCH3
OH
O
NH
OH
Cl
HO
O
O
O
OCH3
HOH2C
OH
A
(1)
R = OH (7)
R = H (11)
(8)
(6)
(5)
22
example, the poly(styrene-co-maleic acid)-conjugated NCS
was approved in Japan in 1993 and has been marketed since
1994 for use against hepatoma [10]. A CD33 monoclonal
antibody (mAB)-calicheamicin conjugate was approved in
US in 2000 and has been marked under the trade name of
Mylotarg to treat acute myeloid leukemia [37]. Several
antiheptoma mAB-C-1027 conjugates have also been
prepared and showed to display high tumor specificity and
to exert a strong inhibitory effect on the growth of
established tumor xenografts [38]. These examples clearly
demonstrate that the enediynes can be developed into
powerful drugs when their extremely potent cytotoxicity is
harnessed and delivered onto the target tumor cells. A great
challenge will be to develop ways to make new enediynes
for mechanistic and clinical studies.
Access to complex natural products such as the enediynes
and their analogs by total synthesis poses a monumental
challenge to synthetic chemists, and yet tour de force effort
by synthetic chemists has led to the total syntheses of
almost every member of the enediyne family of natural
products as well as a myriad of analogs. Significant progress
in developing the enediynes into clinical agents has been
made towards (a) improving cancer cell specificity, (b)
developing efficient delivery systems to the tumor targets,

3. Biosynthesis and Prospect Towards Engineering Novel Antitumor Agents Current Medicinal Chemistry, 2003, Vol. 10, No. 21 2319
(Fig. 1). contd.....
HNOH
OH
O
O OH
CO2H
OCH3
H
O
CH3
O
O
SSSCH3
NHCO2CH3
HO O
OHO
O
N
H
O
CH3 O
OCH3
NHEt
OH
CH3
S
O
H3C OCH3
OH
I OCH3
O
CH3O
HO
H3C
H3C
HO
O
O
O
SSSCH3
NHCO2CH3
HO
O
OHO
O
N
H
O
CH3 O
OCH3
NHPr-i
OH
CH3
SCH3
H3CO O
H3CO NH
O
O
OCH3O
O
SCH3
O
SSSCH3
NHCO2CH3
HO O
HOO
O
O
CH3 O
OCH3
NHPr-i
OH
CH3
SCH3
OH
B
O
SCH3
O
SSSCH3
NHCO2CH3
HO O
HOO
O
CH3
OCH3
NHPr-i
N N
H
OH
O
(2)
(3)
(9)
(10)
(4)
Fig. (1). The chromophore structures of naturally occurring 9-membered enediyne chromoproteins: neocarzinostatin (1), C-1027 (7),
maduropetin (8), kedarcidin (6), and N1999A2 (5) and as well as the engineered deshydroxy-C-1027 (11). (B) Structural
representation of naturally occurring 10-membered enediynes: calicheamicin r1
I (2), esperamicin A1 (3), dynemicin A (9),
namenamicin (10), and shishijimicin A (4).
and (c) designing triggers and trapping the enediynes as
prodrugs [11-13]. However chemical total synthesis has very
limited practical value for complex natural products such as
the enediynes, and analog generation by chemical
modification of the natural products can only access to
limited functional groups, often requiring multiple
protection and deprotection steps.

4. 2320 Current Medicinal Chemistry, 2003, Vol. 10, No. 21 Shen et al.
Complementary to organic synthesis, genetic
manipulations of genes governing secondary metabolism—
an emerging technology also known as combinatorial
biosynthesis—offer a promising alternative to preparing
complex natural products and their analogs biosynthetically
[39-46]. Specific structural alteration in the presence of other
functional groups can often be achieved, and the target
molecules will be produced by a recombinant organism that
is amenable for large-scale fermentation, thereby lowering
the production cost and reducing the environment concern
associated with conventional chemical synthesis. The
success of this approach depends on (a) the cloning and
genetic and biochemical characterization of the biosynthetic
pathways of the target metabolites and (b) the development
of strategies, methods, and expedient tools for combinatorial
manipulation of natural product biosynthetic gene clusters.
The enediynes offer an outstanding opportunity to
investigate the genetic and biochemical basis for the
biosynthesis of structurally complex natural products and to
apply the combinatorial biosynthesis principle and methods
to enediyne biosynthesis for generating novel enediyne
analogs, some of which could potentially be developed into
anticancer drugs.
In this review, we provide a brief account on the current
status of enediyne biosynthesis and the prospective of
applying combinatorial biosynthesis methods to the
enediyne system for novel analog production. We first
review the strategies available for cloning enediyne
biosynthetic gene clusters from various organisms. We then
examine the genetic and biochemical data for the
biosynthesis of 7 and 2, as examples for the 9- and 10-
membered enediynes, respectively, to illustrate a common
polyketide pathway for the enediyne family of natural
products, characterized by an iterative type I polyketide
synthase (PKS). We finally discuss challenges and possible
solutions for combinatorial biosynthesis of enediyne natural
products with the production of deshydroxy-C-1027 (11) to
demonstrate the feasibility of this approach. Early studies on
enediyne biosynthesis by isotope feeding experiments and
on cloning and characterization of the apo-proteins can be
seen in a previous review [47].
METHODS FOR CLONING OF ENEDIYNE
BIOSYNTHETIC GENE CLUSTERS
Antibiotic production genes are often clustered in one
region of the microbial chromosome, consisting of
structural, resistance, and regulatory genes. This clustering
phenomenon has become the cornerstone from which
strategies for cloning microbial natural product biosynthetic
gene clusters were developed [39,48-51]. Practically, once
one gene is cloned and confirmed to encode the production
of the target molecule, chromosomal walking from this
locus is virtually certain to lead to the identification of the
entire biosynthetic gene cluster (ranging in size from 20-200-
kb [48]). Recent technological advancement in DNA
sequencing has dramatically reduced the time and effort in
determining the DNA sequence once the cluster is cloned.
However, it remains to be a great challenge to
unambiguously identify, localize, and confirm the target
gene cluster from a genome of 9 Mbp (the approximate size
of genome of the antibiotic-producing actinomycetes), a
daunting task that is further complicated by the fact that
most of these organisms contain multiple secondary
metabolite biosynthetic gene clusters [49-51].
dNDP-Glucose 4,6-Dehydratase as A Probe
This strategy was exemplified by the cloning of the
biosynthetic gene cluster for 7 from S. globisporus [52,53]
and should be applicable to gene clusters of both the 9- and
10-membered enediynes. Except for 5 and 9, enediynes with
known structures all contain at least one deoxyhexose
moiety. The biosynthesis of various deoxyhexoses shares a
common key intermediate, 4-keto-6-deoxyglucose nucleoside
diphosphate, the formation of which from glucose
nucleoside diphosphate is catalyzed by the dNDP-glucose
4,6-dehydratase [54,55]. Taking advantage of the high amino
acid sequence homology observed among dNDP-glucose
4,6-dehydratases from actinomycetes, Bechthold and co-
workers developed a general approach for cloning dNDP-
glucose 4,6-dehydratase genes by PCR with primers
designed according to the conserved regions [56]. We
adopted this PCR method to clone the biosynthetic gene
cluster for 7 from S. globisporus [52,53]. We first cloned a
putative dNDP-glucose 4,6-dehydratase gene, sgcA, from S.
globisporus by PCR and confirmed its involvement in 7
biosynthesis by gene disruption and complementation [52].
We then screened the S. globisporus genomic library using
sgcA as a probe and localized the entire biosynthetic gene
cluster for 7 to an 85-kb contiguous region of the S.
globisporus chromosome [53]. We finally sequenced and
genetically characterized the cloned sgc cluster, revealing that
sgcA is indeed clustered with the other structural genes, the
previously cloned cagA gene that encodes the C-1027 apo-
protein [57], as well as resistance and regulatory genes for 7
production [53].
Apo-Protein as A Probe
This strategy depends on the availability of the apo-
protein, many of which have been previously cloned and
characterized [47], and its applicability, therefore, is limited
only to the chromoprotein enediynes as exemplified by the
cloning of the biosynthetic gene cluster for 1 from S.
carzinostaticus [48]. We initially attempted to localize the 1
biosynthetic gene cluster by the PCR method [56] to clone
the dNDP-glucose-4,6-dehydratase as a probe, following the
same strategy as we did for 7 [52]. Although we were
successful in amplifying a putative dNDP-glucose-4,6-
dehydratase gene from S. carzinostaticus, inactivation of it
failed to abolish 1 production, auguring against its
involvement in 1 biosynthesis. We then reasoned that the
apo-protein gene could be used as a probe to clone the
chromoprotein enediyne gene cluster, viewing the apo-
protein as a resistance mechanism via drug sequestering.
This hypothesis gained further support after we confirmed
that the cagA gene, encoding the CagA aproprotein for 7,
was indeed clustered with the rest of its biosynthesis,
resistance, and regulatory genes [52,53]. We then took
advantage of the previously cloned ncsA gene, which
encodes the NcsA apoprotein for 1 [58,59], as a probe to
screen the S. carzinostaticus genomic library and localized

5. Biosynthesis and Prospect Towards Engineering Novel Antitumor Agents Current Medicinal Chemistry, 2003, Vol. 10, No. 21 2321
the biosynthetic gene cluster for 1 to a 95-kb contiguous
region of the S. carzinostaticus chromosome. Our
confidence that the cloned gene cluster encoded 1
biosynthesis was further increased after identifying another
copy of the dNDP-glucose-4,6-dehydratase gene, different
from the one amplified by PCR, within this locus. The co-
localization of ncsA and a dNDP-glucose-4,6-dehydratase
gene, reminiscent to that for the 7 biosynthetic gene cluster
[53], was viewed as an evidence of clustering between
biosynthesis and resistance genes. DNA sequence of the 95-
kb region finally unveiled the ncs cluster, consisting of the
ncsA gene, a dNDP-glucose-4,6-dehydratase gene, as well as
other biosynthesis, resistance, and regulatory genes, a
genetic organization that is highly homologous to that of 7
[48,53].
Resistance Gene as A Probe
This strategy was exemplified by the cloning of the
biosynthetic gene cluster for 2 from M. echanospora [60,61]
and should be applicable to gene clusters of both the 9- and
10-membered enediynes. To clone the 2 biosynthetic gene
cluster, Thorson and co-workers first screened a M .
echanospora genomic library with probes designed
according to the conserved regions of both type I and type II
PKSs on the assumption that the enediyene core is of
polyketide origin. Since it is known that most
actinomycetes contain multiple polyketide pathways [49-
51,62,63], the resultant clones were further screened with
probes designed according to the conserved regions of both
glucose-1-phosphate nucleotidyltransferase and dNDP-
glucose 4,6-dehydratase targeting the biosynthesis of the
deoxyhexose moieties of 2. While these preliminary
screenings were effective in identifying putative clones for
the 2 biosynthetic locus, it was the finding that some of the
clones, positive towards both the PKS and deoxyhexose
probes, also conferred 2 resistance that ultimately convinced
these researchers that they had localized the biosynthetic
gene cluster for 2 [47]. They subsequently mapped the 2
resistance gene to calC and characterized CalC as a non-
heme iron metalloprotein, although the precise mechanism
of how CalC confers 2 resistance remains to be established
[60]. DNA sequencing of these clones finally unveiled the
entire cal gene cluster, consisting of PKS genes,
deoxyhexose biosynthesis genes, the calC resistance gene, as
well as other biosynthesis, resistance, and regulatory genes.
Inactivation of the calE8 enediyne PKS gene abolished 2
production, unambiguously confirming the involvement of
the cloned cal gene cluster in 2 biosynthesis [61].
Genome Scanning Method
This strategy was exemplified by the cloning of the
biosynthetic gene clusters for 9 from M. chersina strain
M956-1, macromomycin (the precise structure remains
unknown) from Streptomyces macromyceticus strain M480-
M1, as well as several putative novel enediynes whose
structures are yet to be determined from other actinomycete
strains that previously were not known as enediyne
producers. Since this method does not rely on specific
biosynthetic machinery, it should be applicable to gene
clusters of both the 9- and 10-membered enediynes as well
as other natural products [48]. Farnet and co-workers
developed this so-called “high-throughput genome scanning
method to discover metabolic loci” on the basis that genes
encoding secondary metabolite biosynthesis in
microorganisms are clustered [48]. Similar to the so-called
“perfect probes” approach to identify and clone type I PKS
gene clusters in a genome [63], this method consists
essentially three steps: (a) shotgun DNA sequencing to
generate short (700 bp) random genome sequence tags
(GSTs) from a genomic library, (b) sequence comparison of
GSTs to a database of microbial gene clusters known to be
involved in natural product biosynthesis to identify
candidate genes, and (c) genomic library screening with
selected GTSs as probes for clones containing genes of
interest as well as neighboring ones that together may
constitute a biosynthetic gene cluster. This method provides
an efficient way to survey the biosynthetic potential for a
given microorganism and to access specific biosynthetic
gene cluster of interest. Using this method, Farnet and co-
workers cloned and sequenced several enediyne biosynthetic
loci from various microorganisms, including not only these
that were known to be enediyne producers such as M .
cherisina strain M956-1 for 9 and Streptomyces
macromyceticus strain M480-M1 for macromomycin but
also ones that previously were not known as enediyne
producers, revealing several novel enediyne biosynthetic
gene clusters. It came as a surprise that the enediyne PKS
loci were found at a very high frequency (~15%) among the
organisms surveyed, suggesting that the ability of these
organisms to produce enediynes or other bioactive natural
products might have been greatly underestimated. Guided by
these genetic data, enediyne production by these organisms,
including those that were not known as enediyne producers
previously, were demonstrated under selective growth
conditions and detected by the enediyne-specific biochemical
induction assays, supporting the involvement of these
clusters in enediyne biosynthesis [48].
Enediyne PKS Gene as A Probe
This strategy was exemplified by the cloning of the
biosynthetic gene clusters for 3 from A. verrucosopora, 8
from A. madurea, and 9 from M. chersina and should be
applicable to gene clusters of both 9- and 10-membered
enediynes. The cloning and characterization of 7 and 2, as
model systems for the 9- and 10-membered enediynes,
respectively, revealed a common polyketide pathway for
enediyne biosynthesis characterized by the enediyne PKS,
iterative type I PKSs that exhibited head-to-tail sequence
homology and have identical domain organization (Figure 2)
[53,61,64]. Taking advantage of the remarkable similarity in
both sequence and domain organization of the enediyne
PKSs and their associated accessory enzymes, we developed
a PCP-based approach to access the enediyne PKS loci using
a set of primers designed according to various conserved
region of these enzymes. The effectiveness of this approach
was demonstrated by the cloning of the enediyne PKS and
its accessory enzymes for 3, 8, and 9 biosynthesis from the
corresponding producer, respectively [65]. The availability of
the enediyne PKS loci has now set the stage to use them as
probes to clone the entire biosynthetic gene clusters.

6. 2322 Current Medicinal Chemistry, 2003, Vol. 10, No. 21 Shen et al.
Fig. (2). (A) Domain organization and sequence comparison between representative enediyne PKSs for a 9-membered enediyne core
such as 7 (SgcE) and a 10-membered enediyne core such as 2 (CalE8). aa, amino acid (B) A unified paradigm for enediyne
biosynthesis featuring the enediyne PKSs to account for the linear unsaturated polyketide intermediates, enediyne PKS accessory
enzymes to branch into the 9- or 10-membered enediyne cores, and tailoring enzymes to imbue additional structural diversity. Atoms
that were incorporated intact from the acyl CoA precursors to the enediyne cores are shown in bold.
A UNIFIED PARADIGM FOR ENEDIYNE
BIOSYNTHESIS INVOLVING AN ITERATIVE
TYPE I PKS
Although feeding experiments with 13C-labeled
precursors clearly established that both the 9- and 10-
membered enediyne cores were derived (minimally) from
eight head-to-tail acetate units [47,64-66], controversy
remained until the recent cloning and characterization of the
biosynthetic gene clusters of 7 [53] and 2 [61] whether the
enediyne cores are assembled by de novo polyketide
biosynthesis or degradation from a fatty acid precursor.
Elucidation of the 7 and 2 biosynthetic gene clusters and
pathways, together with the cloning and sequence analysis of
the enediyne PKS loci for 1, 3, 8, and 9, now clearly
suggest a common polyketide pathway for the biosynthesis
of both 9- and 10-membered enediyne cores, despite the fact
that their incorporation pattern by 13C-labeled acetate feeding
experiments are distinct [47,64-66]. Common to all
enediyne biosynthetic gene clusters known to date is the
enediyne PKS—an unprecedented type I PKS consisting of a
ketoacyl synthase (KS), an acyltransferase (AT), an acyl
carrier protein (ACP), a ketoreductase (KR), a dehydratase
(DH), and a C-terminal domain (TD) (also proposed to
encode a phosphopantetheinyl transferase) (Figure 2).
Flanking the enediyne PKS gene are a group of five to ten
genes that are organizationally conserved among all
characterized enediyne gene clusters. These genes encode the
enediyne PKS accessory enzymes that are unusual
oxidoreductases or proteins of unknown functions that are

7. Biosynthesis and Prospect Towards Engineering Novel Antitumor Agents Current Medicinal Chemistry, 2003, Vol. 10, No. 21 2323
only associated with enediyne biosynthesis. Also identified
within the enediyne clusters are genes that encode a myriad
of pathway-specific tailoring enzymes for the biosynthesis of
the other structural moieties and their attachment to the
enediyne cores [48,53,61,64].
Although these findings fell short of revealing the precise
mechanism for enediyne core biosynthesis, they have
provided sufficient data to formulate a unified paradigm for
enediyne biosynthesis. The enediyne PKSs could be
envisaged to catalyze the assembly of a linear
polyunsaturated intermediate from the acyl CoA precursors
in an iterative process. The nascent intermediate is
subsequently desaturated to furnish the two yne groups and
cyclized to afford either a 9- or 10-membered enediyne core
intermediate. The enediyne PKS accessory enzymes serve as
the excellent candidates for these processing reactions.
Preliminary phylogenetic analysis of the enediyne PKSs and
their accessory enzymes showed that they could be grouped
into two distinct sub-families according to the 9- or 10-
membered enediyne core structures. This finding suggested
that it might be even possible to predict the enediyne core
structure on the basis of sequence analysis of its biosynthetic
gene cluster [65]. Further decoration of the enediyne cores
with the other structural moieties affords the structural
diversity of the enediyne family of natural products, and the
pathway specific tailoring enzymes are responsible for these
modifications. This emerging paradigm for enediyne
biosynthesis, consisting of (a) enediyne PKS-catalyzed
synthesis of a linear polyketide intermediate, (b) enediyne
PKS accessory enzymes-mediated processing of the enediyne
core, and (c) pathway specific tailoring enzymes-driven
structural diversification of the enediyne natural products,
has now laid the foundation to ask the fundamental
mechanistic questions for enediyne biosynthesis and to
explore the potential of engineering novel enediyne
compounds by applying combinatorial biosynthesis methods
to the enediyne biosynthetic machinery.
COMBINATORIAL BIOSYNTHESIS OF NOVEL
ENEDIYNES
A prerequisite for applying combinatorial biosynthesis
methods to engineer novel natural products is to have a
genetic system in place for the producing organism. Four
general methods are available for introducing plasmid DNA
into a Streptomyces species: (a) polyethyleneglycol (PEG)-
mediated protoplast transformation, (b) electroporation, (c)
E. coli-Streptomyces conjugation, and (d) phage infection
[69]. However, the applicability of each method to a given
organism, cannot be predicted a priori, and the effectiveness
of these methods to a particular species has to be determined
empirically. In spite of these difficulties, significant progress
has been made in the past a few years for several enediyne
producing organisms. We have developed efficient genetic
systems for both the 7-producing S. globisporus and 1-
producing S. carzinostaticus strains by either PEG-mediated
protoplast transformation or E. coli- Streptomyces
conjugation. These methods were successfully applied to
carry out in vivo genetic characterization of 7 or 1
biosynthesis by gene disruption, replacement, or
complementation experiments, setting the stage to engineer
the 9-membered enediyne biosynthetic pathways [48,52,53].
Thorson and co-workers have successfully developed a
genetic system for M. echinospora ssp. calichensis by PEG-
mediated protopolast transformation and applied it to
inactivate the calE8 enediyne PKS gene. The resultant calE8
mutant lost its ability to produce 2, demonstrating the
feasibility of manipulating the biosynthesis of 10-membered
enediynes such as 2 in vivo [61].
Complementary to manipulating natural product
biosynthesis in their native producing organisms, progress
has also been made to develop bacterial artificial
chromosome (BAC) that can shuttle between E. coli and
Streptomyces species for expression of the target gene cluster
in a heterologous host. The entire biosynthetic gene cluster
could be cloned directly from chromosomal DNA [70-73] or
reconstructed from previously isolated overlapping cosmids
[74-76]. The resultant BAC could be mobilized into
genetically amenable host strains, such as Streptomyces
coelicolor or Streptomyces lividans, allowing genetic
analysis and manipulation of biosynthetic gene clusters that
are otherwise intractable genetically in their native producers.
While holding a great promise, this strategy has yet to
demonstrate efficient expression of the entire gene cluster
and production of the encoded natural product in the
heterologous host. Attempts to clone the entire biosynthetic
clusters of 1 or 7 into BAC for heterologous expression were
not successful so far. In contrast, we have recently
discovered that the biosynthetic gene clusters for 1 and 7
reside on giant linear plasmids. By protoplast fusion, we
successfully mobilized the giant linear plasmid for 1
biosynthesis from S. carzinosaticus into S. lividans and
confirmed 1 production in the heterologous host [Zhang, J.;
Nonaka, K.; Shen, B. unpublished data]. These findings
provided an exciting alternative to manipulate enediyne
biosynthesis in one of the best-characterized Strteptomyces
hosts.
Although the engineering of novel enediynes with altered
enediyne core has to wait for the fundamental understanding
of how the enediyne PKS and its accessory enzymes catalyze
the enediyne core biosynthesis, the proof-of-principle for
combinatorial biosynthesis of novel enediynes has already
been demonstrated by the production of deshydroxy-C-1027
(11) [53]. Characterization of the 7 biosynthetic gene cluster
and elucidation of the 7 biosynthetic pathway established
that the C22-OH group of the β-amino acid moiety of 7 is
introduced by the SgcC hydroxylase. Inactivation of the
sgcC gene resulted in the production of a novel
chromoprotein complex, the chromophore of which was
confirmed to be 11. It was noteworthy that 11 is not only
biologically active as judged by bioassay with 7 as a
positive control but also at least fivefold more stable than 7
at 25 °C in respect to undergoing the Bergman cyclization
[53]. The latter property could be potentially explored in
developing 7 into clinically useful enediyne drugs. Taking
together, these results not only demonstrated the feasibility
of generating novel enediynes by manipulating the genes
governing their biosynthesis but also suggested that the
tailoring enzymes could be as equally rewarding as the
enediyne PKS and its accessory enzymes as the targets for
engineering novel enediynes. Given the relatively conserved
enediyne core structures yet distinct periphery moieties
among the enediynes, it is tempting to speculate that some

8. 2324 Current Medicinal Chemistry, 2003, Vol. 10, No. 21 Shen et al.
of the tailoring enzymes could cross-talk among the various
enediyne pathways, further expanding the enediyne structural
diversity accessible by combinatorial biosynthesis methods.
CONCLUSION AND PROSPECTIVE
The studies of enediyne biosynthesis have enjoyed a
renaissance in the past a few years. The enediyne family of
natural products has seen a steady growth. Specific and
general methods to access enediyne biosynthetic gene
clusters have been developed. Biosynthetic gene clusters for
three 9-membered enediynes (1, 7, 8) and three 10-membered
enediynes (2, 3, 9) have been cloned, respectively. A unified
paradigm for enediyne biosynthesis has emerged, featuring
an unprecedented iterative type I PKS and its accessory
enzymes as well as a myriad of tailoring enzymes. Genetic
systems for representative 9- and 10-membered enediyne
producing organisms have been developed, and genetic
characterization of enediyne gene clusters in vivo has begun
to unveil new mechanistic insights into enediyne
biosynthesis. Judicial application of combinatorial
biosynthesis methods to the enediyne biosynthetic
machinery has demonstrated the production of novel
enediynes. It is becoming increasingly clear that access to
the genetic information for natural product biosynthesis is
no longer the bottleneck for combinatorial biosynthesis. A
great challenge in studying enediyne biosynthesis lies in the
revelation of the fundamental mechanism of how the
enediyne PKS and its accessory enzymes assemble the
enediyne cores from the acyl CoA precursors. Although
sequence analysis and functional comparison among the
various enediyne gene clusters have enabled us to formulate
a few rudimentary hypotheses, innovation in both
biochemistry concepts and experimental designs are needed
to make the breakthrough in deciphering the precise
mechanism for the enediyne core biosynthesis. Outcomes
from these studies will greatly accelerate the pace to realize
the full potential of combinatorial biosynthesis of enediyne
natural products. We remain optimistic that enediyne
biosynthesis will continue to thrive in the coming years,
particularly in mechanistic enzymology and in generation of
novel enediynes by combinatorial biosynthesis methods.
ABBREVIATION
ACP = Acyl carrier protein
AT = Acyl transferase
DH = Dehydratase
KR = Ketoreductase
KS = Ketoacyl synthase
Mab = Monoclonal antibody
NCS = Neocarzinostatin
PKS = Polyketide synthase
TD = C-terminal domain
ACKNOWLEDGEMENTS
Enediyne biosynthesis described from our laboratory was
supported in part by National Institutes of Health grant
CA78747. B.S. is a recipient of National Science
Foundation CAREER Award MCB9733938 and National
Institutes of Health Independent Scientist Award AI51687.
REFERENCES
[1] Ishida, N.; Miyazaki, K.; Kumagai, K.; Rikimaru, M. J. Antibiot.,
1965, 18, 68.
[2] Edo, K.; Mizugaki, M.; Koide, Y.; Seto, H.; Furihata, K.; Otake,
N.; Ishida, N. Tetrahedron Lett., 1985, 26, 331.
[3] Lee, M.D.; Dunne, T.S.; Siegel, M.M.; Chang, C.C.; Morton, G.O.;
Borders, D.B. J. Am. Chem. Soc., 1987, 109, 3464.
[4] Lee, M.D.; Dinne, T.S.; Chang, C.C.; Ellestad, G.A.; Siegel, M.M.;
Morton, G.O.; McGahren, W.J.; Borders, D.B. J. Am. Chem. Soc.,
1987, 109, 3466.
[5] Golik, J.; Clardy, J.; Dubay, G.; Groenewold, G.; Kawaguchi, H.;
Konishi, M.; Krishnan, B.; Ohkuma, H.; Saitoh, K.; Doyle, T. W.
J. Am. Chem. Soc., 1987, 109, 3461.
[6] Golik, J.; Dubay, G.; Groenewold, G.; Kawaguchi, H.; Konishi,
M.; Krishnan, B.; Ohkuma, H.; Saitoh, K.; Doyle, T.W. J. Am.
Chem. Soc., 1987, 109, 3462.
[7] Nicolaou, K.C.; Dai, W.-M. Angew. Chem. Int. Ed. Engl., 1991,
30, 1387.
[8] Doyle, T.W.; Border, D.B. Enediyne Entibiotics as Antitumor
Agents. Marcel-Dekker: New York, 1995.
[9] Smith, A.L.; Nicolaou, K.C. J. Med. Chem., 1996, 39, 2103.
[10] Maeda, H.; Edo, K.; Ishida, N. Eds. Neocarzinostatin: The Past,
Present, and Future of an Anticancer Drug. Springer-Verlag:New
York, 1997.
[11] Xi, Z.; Goldberg, I.H. in Comprehensive Natural Products
Chemistry, Barton, D.; Nakanish, K.; Meth-Cohn, O., Eds.
Elesiver:New York, 1999, vol 7, p.553.
[12] Thrson, J.S.; Sievers, E.L.; Ahlert, J.; Shepard, E.; Whitwam, R.E.;
Onwueme, K.C.; Ruppen, M. Curr. Pharm. Design., 2000, 6, 1841.
[13] Jones, G.B.; Fouad, F.S. Curr. Pharm. Des., 2002, 8, 2415.
[14] Oku, N.; Matsunaga, S.; Fusetani, N. J. Am. Chem. Soc., 2003,
125, 2044.
[15] Ando, T.; Ishii, M.; Kajiura, T.; Kameyama, T.; Miwa, K.;
Sugiura, Y. Tetrahedron Lett., 1998, 39, 6495.
[16] Leet, J.E.; Schroeder, D.R.; Hofstead, S.J.; Golik, J.; Colson, K.L.;
Huang, S.; Klohr, S.E.; Doyle, T.W; Matson, J.A. J. Am. Chem.
Soc., 1992, 114, 7946.
[17] Minami, Y.; Yoshida, K.; Azuma, R.; Saeki, M.; Otani, T.
Tetrahehron Lett., 1993, 34, 2633.
[18] Yoshida, K.; Minami, Y.; Azuma, R.; Saeki, M.; Otani, T.
Tetrahehron Lett., 1993, 34, 2637.
[19] Iida, K.; Fukuda, S.; Tanaka, T.; Hirama, M.; Imajo, S.; Ishiguro,
M.; Yoshida, K.; Otani, T. Tetrahedron Lett., 1996, 37, 4997.
[20] Otani, T.; Yoshida, K.; Sasaki, T.; Minami, Y. J. Antibiot., 1999,
52, 415.
[21] Schroeder, D.R.; Colson, K.L.; Klohr, S.E.; Zein, N.; Langley,
D.R.; Lee, M.S.; Matson, J.A.; Dolye, T.W. J. Am. Chem. Soc.,
1994, 116, 9351.
[22] Konishi, M.; Ohkuma, H.; Matsumoto, K.; Tsuno, T.; Kamei, H.;
Miyaki, T.; Oki, T.; Kawaguchi, H. J. Antibiot., 1989, 42, 1449.
[23] Myers, A.G.; Fraley, M.E.; Tom, N.J.; Cohen, S.B.; Madar, D.J.
Chem. Biol., 1995, 2, 33.
[24] McDonald, L.A.; Capson, T.L.; Krishnamurth, G.; Ding, W.-D.;
Ellestad, G.A.; Bernan, V.S.; Maiese, W.M.; Lassota, P.;
Discafani, C.; Kramer, R.A.; Ireland, C.M. J. Am. Chem. Soc.,
1996, 118, 10898.
[25] Stassinopoulos, A.; Ji, J.; Gao, X.; Goldberg, I.H. Science, 1996,
272, 1943.
[26] Ikemoton, N.; Kumar, R.A.; Ling, T.T.; Elleastad, G.A. Proc. Natl.
Acad. Sci. USA, 1995, 92, 10506.
[27] Dedon, P.C.; Goldberg, I.H. Chem. Res. Toxicol., 1992, 5, 311.
[28] SugiuraY.; Shiraki, T.; Konishi, M.; Oki, T. Proc. Natl. Acad. Sci.
USA., 1990, 87, 3831.
[29] Zein, N.; Closon, K.L.; Leet, J.E.; Schroeder, D.R.; Solomon, W.;
Dolye, T.W.; Casazza, A.M. Proc. Natl. Acad. Sci. USA., 1993,
90, 2822.
[30] Nicolaou, K.C.; Stabila, P.; Esmaeli-Azad, B.; Wrasidlo, W.; Hiatt,
A. Proc. Natl. Acad. Sci. USA., 1993, 90, 3142.
[31] Kappen, L.S.; Goldberg, I.H. Science, 1993, 261, 1319.

9. Biosynthesis and Prospect Towards Engineering Novel Antitumor Agents Current Medicinal Chemistry, 2003, Vol. 10, No. 21 2325
[32] Yu, L.; Goldberg, I.H.; Dedon, P.C. J. Biol. Chem., 1994, 269,
4144.
[33] Hensens, O.D.; Chin, D.H.; Stassinopoulos, A.; Zink, D.L.;
Kappen, L.S.; Goldberg, I.H. Proc. Natl. Acad. Sci. USA., 1994,
91, 4534.
[34] Ho, S.N.; Boyer, S.H.; Schreiber, S.L.; Danishefsky, S.J.;
Crabtree, G.R. Proc. Natl. Acad. Sci. USA., 1994, 91, 9203.
[35] Schor, N.F.; Tyuina, Y.Y.; Fabisiak, J.P.; Tyurin, V.A.; Lazo, J.S.;
Kagan, V.E. Brain Res., 1999, 831, 125.
[36] Dziegielewski, J.; Beerman, T.A. J. Biol. Chem., 2002, 277,
20549.
[37] Sielvers, E.L.; Appelbaum, F.R.; Spielberger, R.T.; Forman, S.J.;
Flowers, D.; Smith, F.O.; Shannon-Dorcy, K.; Berger, M.S.;
Bernstein, I.D. Blood, 1999, 93, 3678.
[38] Brukner, I. Curr. Opin. Oncol. Endocr. Met. Invest. Drugs., 2000,
2, 344.
[39] Hopwood, D.A. Chem. Rev., 1997, 97, 2465.
[40] Cane, D.E.; Walsh, C.T.; Khosla, C. Science, 1998, 6, 319.
[41] Shen, B. Curr. Top. Chem., 2000, 209, 1.
[42] Staunton, J.; Wessman, K.J. Nat. Prod. Rep., 2001, 18, 380.
[43] Strohl, W.R. Met. Engineer., 2001, 3, 4.
[44] Du, L.; Shen, B. Curr. Opinion Drug Discov. Dev., 2001, 4, 215.
[45] Rodriguez, E.; McDaniel, R. Curr. Opinion Microbiol., 2001, 4,
526.
[46] Walsh, C.T. ChemBioChem, 2002, 3, 124.
[47] Thorson, J.S.; Shen, B.; Whitwam, R.E.; Liu, W.; Li, Y.; Ahlert, J.
Bioorg. Chem., 1999, 27, 172.
[48] Zazopoulos, E.; Huang K.; Staffa, A.; Liu, W.; Bachmann, B.O.;
Nonaka, K.; Ahlert, J.; Thorson, J.S.; Shen, B.; Farnet, C.M. Nat.
Biotechnol., 2003, 21, 187.
[49] Bentley, S.D.; Chater, K.F.; Cerdeno-Tarraga, A.M.; Challis, G.L.;
Thomson, N.R.; James, K.D.; Harris, D.E.; Quail, M.A.; Kieser,
H.; Harper, D.; Bateman, A.; Brown, S.; Chandra, G.; Chen, C.W.;
Collins, M.; Cronin, A.; Fraser, A.; Goble, A.; Hidalgo, J.;
Hornsby, T.; Howarth, S.; Huang, C.H.; Kieser, T.; Larke, L.;
Murphy, L.; Oliver, K.; O'Neil, S.; Rabbinowitsch, E.;
Rajandream, M.A.; Rutherford, K.; Rutter, S.; Seeger, K.;
Saunders, D.; Sharp, S.; Squares, R.; Squares, S.; Taylor, K.;
Warren, T.; Wietzorrek, A.; Woodward, J.; Barrell, B.G.;
Parkhill, J.; Hopwood, D.A. Nature, 2002, 417, 141.
[50] Omura, S.; Ikeda, H.; Ishikawa, J.; Hanamoto, A.; Takahashi, C.;
Shinose, M.; Takahashi, Y.; Horikawa, H.; Nakazawa, H.;
Osonoe, T.; Kikuchi, H.; Shiba, T.; Sakaki, Y.; Hattori, M. Proc.
Natl. Acad. Sci. USA, 2001, 98, 12215.
[51] Ikeda, H.; Ishikawa, J.; Hanamoto, A.; Shinose, M.; Kikuchi, H.;
Shiba, T.; Sakaki, Y.; Hattori, M.; Omura, S. Nat. Biotechnol.,
2003, 21, in press.
[52] Liu, W.; Shen, B. Antimicrob. Agents Chemother., 2000, 44, 382.
[53] Liu, W.; Christenson, S. D.; Standage, S.; Shen, B. Science, 2002,
297, 1170.
[54] Trefzer, A.; Salas, J.A.; Bechthold, A. Nat. Prod. Rep., 1999, 16,
283.
[55] He, X.M.; Liu, H.-w. Annu. Rev. Biochem., 2002, 71, 701.
[56] Decker, H.; Gaisser, S.; Pelzer, S.; Schneider, P.; Westrich, L.;
Wohlleben, W.; Bechthold, A. FEMS Microbiol. Lett., 1996, 141,
195.
[57] Sakata, N.; Ikeno, S.; Hori, M.; Hamada, M.; Otani, T. Biosci.
Biotech. Biochem., 1992, 56, 1592.
[58] Maeda, H.; Glaser, C.B.; Kuromizu, K.; Meienhofer, J. Arch.
Biochem. Biophys., 1974, 164, 379.
[59] Gibson, B.W.; Herlihy, W.C.; Samy, T.S.A.; Hahm, K.S.; Maeda,
H.,; Meienhofer, J.; Biemann, K. J. Biol. Chem., 1984, 259, 10801.
[60] Whitwam, R. E.; Ahlert, J.; Holman, T.R.; Ruppen, M.; Thorson,
J.S. J. Am. Chem. Soc., 2000, 122, 1556.
[61] Ahlert, J.; Shepard, E.; Lomovskaya, N.; Zazopoulos, E.; Staffa
A.; Bachmann, B. O.; Huang, K.; Fonstein L.; Czisny, A.;
Whitwam, R.E.; Farnet, C.M.; Thorson, J.S. Science, 2002, 297,
1173.
[62] Ruan, X.; Stassi, D.; Lax, S.A.; Katz, L. Gene, 1997, 203, 1.
[63] Santi, D.V.; Siani, M.A.; Julien, B.; Kupfer, D.; Roe, B. Gene,
2000, 247, 97.
[64] Shen. B. Curr. Opinion Chem. Biol., 2003, 7, 285.
[65] Liu, W.; Ahlert, J.; Gao, Q.; Wendt-Pienkowski, E.; Shen, B.
Thorson, J.S.; submitted.
[66] Hensens, O.D.; Ginger, J.-L.; Goldberg, L.H. J. Am. Chem. Soc.,
1989, 111, 3295.
[67] Tokiwa, Y.; Miyoshi-Saitosh, M.; Kobayashi, H.; Sunaga, R.;
Konishi, M.; Oki, T.; Iwasaki, S. J. Am. Chem. Soc., 1992, 114,
4107.
[68] Lam, K.S.; Veitch, J.A.; Golik, J.; Krishnan, B. J. Am. Chem. Soc.,
1993, 115, 12340.
[69] Kieser, T.; Bibb, M.J.; Buttner, M.J.; Chater, K.F.; Hopwood, D.A.
Practical Streptomyces Genetics. The John Innes
Foundation:Norwich, U.K., 2000.
[70] Gillespie, D.E.; Brady, S.F.; Bettermann, A.D.; Cianciotto, N.P.;
Liles, M.R.; Rondon, M.R.; Clardy, J.; Goodman, R.M.;
Handelsman, J. Appl. Environ. Microbiol., 2002, 68, 4301.
[71] Brady, S.F.; Chao, C.J.; Clardy, J. J. Am. Chem. Soc., 2002, 124,
9968.
[72] Zengler, K.; Toledo, G.; Rappe, M.; Elkins, J.; Mathur, E.J.; Short,
J.M.; Keller, M. Proc. Natl. Acad. Sci., 2002, 99, 15681.
[73] Miao, V.; Cöeffet-LeGal, M.-F.; Penn, J.; Whiting, A.; Parr, I.;
Bouchard, M.; Silva, C.; Doekel, S.; Alexander, D.; Martin, S.;
Gibson, T.; Brost, R.; Sinneman, S., Henton-Jones, D.; Chapple,
J.P.; Daavies, J.E.; Baltz, R.H.; Brian, P. Abstract S123, Society for
Industrial Microbiology Annual Meeting, August 11-15, 2002,
Philadelphia, PA.
[74] Sosio, M.; Giusino, F.; Cappellano, C.; Bossi, E.; Puglia, A.M.;
Donadio, S. Nat. Biotechnol., 2000, 18, 343.
[75] Sosio, M.; Bossi, E.; Donadio, S. Nucleic Acids Res., 2001, 37, e37.
[76] Donadio, S.; Monciardini, P.; Aldunia, R.; Mazza, P.; Chiocchini,
C.; Cavaletti, L.; Sosio, M.; Puglia, A.M. J. Biotechnol., 2002, 99,
187.