J Biol Chem, Vol. 274, Issue 46, 32829-32834, November 12, 1999
Disruption of an Aromatase/Cyclase from the Oxytetracycline
Gene Cluster of Streptomyces rimosus Results in
Production of Novel Polyketides with Shorter Chain Lengths*
Hrvoje
Petkovi
§,
Arinthip
Thamchaipenet¶,
Li-Hong
Zhou,
Daslav
Hranueli
,
Peter
Raspor§,
Peter G.
Waterman, and
Iain S.
Hunter**
From the Department of Pharmaceutical Sciences,
University of Strathclyde, Glasgow, Scotland G1 1XW, United
Kingdom, § Biotechnical Faculty, University of Ljubljana,
Ljubljana, 61000 Slovenia, ¶ Department of Genetics, Kasetsart
University, Bangkok, 10900 Thailand, and PLIVA d.d., Research
Institute, Anti-infective Research, 10000 Zagreb, Croatia
 |
ABSTRACT |
Oxytetracycline is a polyketide antibiotic made
by Streptomyces rimosus. From DNA sequencing, the gene
product of otcD1 is deduced to function as a bifunctional
cyclase/aromatase involved in ring closure of the polyketide backbone.
Although otcD1 is contiguous with the ketoreductase gene,
they are located an unusually large distance from the genes encoding
the "minimal polyketide synthase" of the oxytetracycline gene
cluster. A recombinant, disrupted in the genomic copy of
otcD1, made four novel polyketides, all of shorter chain
length (by up to 10 carbons) than oxytetracycline. All four novel
structures contained the unusual carboxamido group, typical of
oxytetracycline. This implies that the carboxamido group is present at
the start of biosynthesis of oxytetracycline, a topic that has been
debated in the literature. Loss of the cyclase protein has a profound
influence on the length of polyketide chain assembled, implying that
OtcD1 plays a greater role in the overall integrity of the quaternary
structure of the polyketide complex than hitherto imagined.
| |
INTRODUCTION |
Oxytetracycline (OTC)1
is a broad spectrum antibiotic made by Streptomyces rimosus
(1). It is a member of the "polyketide" class of secondary
metabolites biosynthesized by condensation of coenzyme A derivatives of
metabolic precursors (2). The backbone of the antibiotic, consisting of
19 carbon atoms, is thought to be derived from an aminated starter unit
(most likely malonamyl-CoA), to which eight acetyl (malonyl-CoA)
extender units are added sequentially (Fig.
1). Molecular genetic analysis of OTC
biosynthesis revealed evidence that the pathway genes are clustered
together on the chromosome of S. rimosus (3, 4), and that
the tetracyclic backbone is assembled by a (so-called) "Type II"
polyketide synthase (PKS) (5), in which this multienzyme complex
carries out the iterative assembly of the backbone from its precursors
and subsequently folds and cyclizes the nascent polyketide. Type II
PKSs that catalyze the biosynthesis of different chemical structures
have a common genetic architecture: their genes share substantial
homology (6).
The minimal Type II PKS has been defined (7) as the three genes which,
together, encode the minimal number of gene products that can direct
biosynthesis of a polyketide chain. The minimal PKS consists of two
translationally coupled "ketosynthase" genes, KS
and
KS
, and an acyl carrier protein to which the growing
polyketide chain is attached. KS contains the essential amino acid residues involved in catalytic condensation of the co-acylated derivative of the extender units, whereas the
KS partner (which is highly similar but non-identical to
KS) is thought to influence the number of iterations of
the condensation cycle and hence the length of the polyketide chain,
justifying its original designation as the chain length factor (8).
Matrix experiments, in which the KS and acyl carrier protein elements have been exchanged among different PKSs, have reinforced this view.
When expressed alone in Streptomyces coelicolor using the elegant trancriptionally activated pRM5/CH999 system, the two KS
subunits and acyl carrier protein from the otc gene cluster directed the biosynthesis of an incorrectly cyclized 20-carbon backbone, without the terminal carboxamido group that is typical of
oxytetracycline (9). This implied that the minimal PKS for OTC
biosynthesis could not direct synthesis of the natural 19-carbon backbone on its own.
After the nascent polyketide chain has been formed, it must be folded
and cyclized to take up its final structure. Two different types of
gene products have been associated with this process: "ketoreductases" (KR), which are responsible for reduction of one
of the keto groups within the chain; and "cyclases" (CYC), which
are responsible for carbon-carbon bond formation resulting in the
closed rings of the final products (6). A KR step is not obligatory for
formation of some polyketides, e.g. the unreduced polyketide, tetracemonycin, folds and is cyclized without a KR (10).
However, cyclases are ubiquitous in polyketide gene clusters. They fall
into two classes: (a) those with "didomain" architecture (typified by frenolicin; see Ref. 11) in which the N-terminal segment
of the gene has likely been duplicated to form the C-terminal second
domain, and (b) those with "monodomain" architecture
(typified by whiE- ORFVI; see Ref. 12) containing a single
cyclase domain. In some instances a single cyclase domain may be linked
to a second domain of unrelated function e.g.
tcmN, which has an O-methyltransferase activity
at its C terminus (13). Some polyketide structures may require more
than one cyclase to fold them correctly (14).
Similarity among DNA sequences of different Type II PKSs has been used
to identify (e.g. see Ref. 15) and clone homologous sequences from other streptomycete species. Biosynthesis of
actinorhodin has become the paradigm for study of Type II PKSs. The
genes responsible for the assembly of the backbone of actinorhodin are
contiguous on the chromosome of S. coelicolor, an
architecture that is present in most other homologous gene clusters.
When actI (the minimal PKS) and actIII (KR) of
the actinorhodin cluster were used as probes with the cloned
otc cluster, two non-contiguous regions of homology were
identified (16): otcY1 hybridized with actI (and
DNA sequencing has shown it to constitute the otc minimal PKS (17)), whereas otcY2 lay 10 kb away and hybridized to
actIII. This paper is concerned with the analysis of the DNA
adjacent to the otcY2 locus. It describes the
characterization of a bifunctional cyclase gene, otcD1, its
subsequent manipulation to produce a recombinant blocked in that
function, and characterization of the polyketide metabolites made by
the recombinant.
| |
EXPERIMENTAL PROCEDURES |
Strain Maintenance and Cultivation--
S. rimosus R6
is used for the commercial production of OTC. It was maintained and
grown in fermentation medium as described previously (18).
General DNA Techniques--
The methods used were described by
Sambrook et al. (19) for Escherichia coli and by
Hopwood et al. (20) for Streptomyces. DNA was
introduced into S. rimosus by an electrotransformation technique (21). DNA was sequenced by the Sanger method (22). Templates
were prepared by subcloning the DNA into vectors that contained
universal and reverse primer binding sites. If necessary, the region of
sequence determination was extended with custom primers.
Isolation of Metabolites--
1.5 liters of fermentation medium
was saturated with NaH2PO4 and extracted with
ethyl acetate (3 × 1 liter). After evaporation to dryness, the
extract was applied to a silica gel flash chromatography column and
eluted with chloroform containing increasing amounts of methanol.
Fractions containing the desired compounds were purified further on
a Sephadex LH-20 column and eluted in a chloroform/methanol gradient.
Structural Elucidation of Metabolites--
This was undertaken
primarily by NMR (400 MHz) using, in addition to standard
one-dimensional experiments, the two-dimensional homonuclear correlated
spectroscopy-45 technique for direct 1H-1H
coupling (23), the hydrogen-carbon correlation with a bilinear rotation
decoupling pulse technique for 1-bond (1J)
heteronuclear (1H-13C interactions) (24), and
the HMBC4 pulse sequence for 2-bond
(2J) and 3-bond (3J)
interactions (23).
| |
RESULTS |
DNA Sequence of otcD1 and Analysis of the Deduced Gene
Product--
When the entire otc cluster was digested with
various restriction enzymes and subjected to Southern blots using
actIII (KR; Ref. 25) as a probe, a SacI fragment
of approximately 700 base pairs gave the strongest signal (HB in Fig.
2;
Sac29-Sac31 in the
restriction map of the otc cluster described in Ref. 26), indicating that it likely contained the major part of the gene encoding
the functional equivalent the KR for OTC biosynthesis. Analysis of this
DNA sequence and the segment adjacent (Sac31 to
Sph28) revealed two open reading frames: the
incomplete C terminus of the ActIII homologue, OtcY2-ORF2, and the
complete open reading frame of otcD1 (Fig. 2). The DNA
sequence of otcD1 and its flanks has been deposited at
GenBankTM.
View larger version (8K):
[in this window]
[in a new window]
|
Fig. 2.
The genetic architecture of the
otc cluster around the locus
otcY2. Open boxes denote the DNA
fragments identified previously in complementation studies
(otcD, otcY2; Ref. 3) or hybridization
experiments (see text, HB; Ref. 16). Arrows denote the
direction of transcription of the genes, which are shown in
shaded boxes, with their designations below. The restriction
sites around otcY2 that are relevant to this work are shown
(K, KpnI; Sp, SphI;
S, SacI). Numbering of restriction sites is from
Hunter and Hill (26). The figure is not drawn to scale. The
broken lines illustrate >8 kb gaps between the
otcY2 locus and the otcY1 locus containing the
genes for the minimal PKS (see text) on one flank and the
otrA resistance gene on the other flank.
|
|
The deduced polypeptide, OtcD1, was predicted to contain 317 amino
acids with Mr of 34635. Using the TBLASTN
software (27), OtcD1 showed substantial end-to-end similarity with
other deduced cyclases from polyketide biosynthetic clusters. The N-
and C-terminal halves of OtcD1 have reasonable similarity to each other
(Fig. 3) indicating a didomain
architecture, and aromatase/cyclase (ARO/CYC) function.
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 3.
Comparison of the N-terminal and C-terminal
peptide sequences of OtcD1. The PILEUP program (43) was used.
Identical residues are shown in capital letters.
|
|
Construction of a S. rimosus Recombinant Disrupted in
otcD1--
The restriction site Sac29 lies
within the open reading frame of otcD1 (Fig. 2).
ErmE, encoding erythromycin resistance (28), was subcloned
into the Sac29 site of an E. coli
plasmid-based copy of the gene and this disrupted copy of
otcD1 (otcD1::ermE) was
homogenotized into the chromosome of S. rimosus, using an unstable bifunctional construct, pGLW121, based on the
thiostrepton-resistant streptomycete vector pIJ487 (29). Transformation
of pGLW121 into S. rimosus R6 was selected by resistance to
thiostrepton and erythromycin. After two subcultures under
non-selective conditions, transformants were screened to identify
isolates (erythromycin-resistant, thiostrepton-sensitive) that had
the disrupted copy of otcD1 introduced into the chromosome
by a double crossover and lost the vector sequences from the cell. The
genotypes of such isolates were confirmed by Southern blotting (30),
and strain ZGL3 was chosen for further work.
Characterization of Metabolites Made by the otcD1::ermE
Recombinant, ZGL3--
Three compounds were separated from the
Sephadex column with 6% (v/v) methanol in chloroform, LH-1 (15.2 mg),
followed by LH-2 (8.5 mg), and finally LH-3 (6.7 mg). A further
compound, LH-4 (7.3 mg), was obtained by direct flash chromatography of the fermentation broth by elution with 8% (v/v) methanol in chloroform.
LH-1 failed to yield a molecular ion by electron impact mass
spectroscopy but the 13C NMR spectrum clearly resolved the
presence of 9 carbons: two sp3 (both methylenes)
and seven sp2 (two methines and five
quaternary). The 1H NMR spectrum confirmed the presence of
two methylenes and two olefinic methines and in addition revealed broad
resonances for OH and NH2. The NMR chemical shift data are
listed in Table I. The IR spectrum
revealed a number of absorption bands between 1,716 and 1,639 cm
1, which suggested a polyketide.
View this table:
[in this window]
[in a new window]
|
Table I
NMR chemical shift data for LH-1 and comparison with published data for
SEK4 (33)
All spectra were recorded in Me2SO-d6. Data for a SEK4
was modified to take into account the change in structure from a
4-oxo-2-hydroxypyran to a 2-oxo-4-hydroxypyran (31).
|
|
The structure was resolved largely on the basis of two-dimensional
homonuclear and heteronuclear NMR. The two olefinic protons exhibited
meta-coupling to each other with one (
5.26;
c 89.0) in an unusually electrophilic environment
reminiscent of H-3 in a 2,4-oxygenated pyran as found in SEK4 (Fig.
4, reported in Ref. 33) (see Table I).
This was confirmed by the HMBC experiment, the correlations observed
being shown in Fig. 5 and which allowed LH-1 (Fig. 4) to be identified as
6-(4-amino-2,4-butadionyl)-4-hydroxypyran-2-one (Fig.
6).2
The original report of SEK4 (33) proposed a 2-hydroxypyran-4-one system, but this was later revised to 4-hydroxypyran-2-one (31) on the
basis of NMR studies. We support this revision as the C-4 resonance at
c 170.2 is at variance with that found in the
pyran-4-one system of flavones and would be anticipated at
~c 176.0 (32).
View larger version (22K):
[in this window]
[in a new window]
|
Fig. 5.
The long-range 1H-13C
coupling observed in the NMR spectra of LH-1, LH-2, LH-3, and LH-4,
which were useful in characterizing the structures.
|
|
The second compound to be eluted again failed to show a molecular ion
by electron impact mass spectroscopy but the 13C NMR
spectrum (Table II) revealed 17 atoms as
three methylenes, four olefinic methines, and 10 quaternary carbons.
The 1H NMR revealed resonances for amide protons ( 6.73, 7.09) and hydroxyl protons ( 7.13, 10.40, 11.50). The presence of
the 6-substituted 4-hydroxypyran-2-one moeity was identified by
comparison with LH-1. The 6-substituent was established as a methylene
( 3.07, 2.99; c 44.0) by observation in an HMBC experiment (Fig.
5b) of the long range coupling of the methylene protons to
C-6 and C-5 of the pyran ring. Likewise, further resonances were
compatible with the presence of the methylenecarboxamide starter group
of this compound.
View this table:
[in this window]
[in a new window]
|
Table II
1H NMR and 13C NMR chemical shift data for compounds
LH-2, LH-3, and LH-4
All spectra were run in Me2SO · d6.
|
|
The intervening structure was established by further analysis of the
HMBC experiment (Fig. 5b). Thus, the methylene of the methylenecarboxamide was coupled with two meta-coupled
protons of an aromatic ring, one of which ( 6.21, c
103.2) must be placed between the two oxygens. From the 6-methylene
substituent of the pyran-4-one substituent there was a coupling with a
carbon resonating at 100.2 ppm. This quaternary carbon was also
correlated with the protons of another methylene and with the proton of
the hydroxyl resonating at 7.13, and it was concluded that it must
be an sp3 carbon carrying two oxygen
substituents. This additional methylene revealed further couplings to a
carbonyl at c 190.3, which is indicative of a
flavanone-like dihydropyran system (31). On the basis of these
observations, it was concluded that the bicyclic was a
2,7-dihydroxy-2,3 dihydropyran-4-one substituted at C-5 by the
methylenecarboxamide and at C-2 with the
6-methylene-2-hydroxypyran-4-one to give structure LH-2 (Figs. 4 and
6),
2-[2,7-dihydroxy-2-(4-hydroxy-6-oxo-4H-pyran-2-methyl-yl)-4- oxochroman-5-yl]-acetamide.
LH-3 exhibited many features in common with LH-2 and the presence of
the 6-methylene-4-hydroxypyran-2-one and methylenecarboxamide groups
could clearly be observed from the NMR spectra (Table II). The
13C spectrum again revealed 17 carbon atoms but in this
case there were three sp2 methines and no
methylenes associated with the intervening substructure. A carbonyl
resonance at c 177.7 was indicative of a
benzopyran-4-one (31) and there were no signals associated with a
benzodihydropyran-4-one, which suggested conversion of the latter to
the former through dehydration to give 7-hydroxybenzopyran-4-one.
Somewhat surprisingly, however, an HMBC experiment (Fig. 5c)
showed that this compound was not the dehydration product of LH-2. The
methylene protons of the methylenecarboxamide group correlated with the
oxygen-bearing C-2 carbon of the benzopyran, whereas the methylene
linked to the 4-hydroxypyran-2-one correlated with a methine (C-6) and
two quaternary carbons (C-4a, C-5) in the benzenoid ring of the
benzopyran. The structure of LH-3 (Fig. 6) must, therefore, be
2-[7-hydroxy-5-(4-hydroxy-6-oxo-4H-pyran-2- methyl-yl)-4-oxo-4H-chromen-2-yl]-acetamide.
The 13C NMR spectrum of LH-4 revealed only 15 carbons, 9 of
which gave chemical shift values comparable to the
2-hydroxy-2,3-dihydrobenzopyran ring system that had already been
identified in LH-2, whereas two others were in agreement with expected
values for a methylenecarboxamide substituent. However, unlike LH-2,
the long range correlations observed in the HMBC spectrum of this
compound (Fig. 5d) revealed that the methylenecarboxamide
occurred at the 2-position of the 2,3-dihydropyran-4-one. This left a
4-carbon unit to be placed at C-5 of the benzopyran. This moeity
consisted of a methylene, a highly shielded sp2
methine ( 4.65; c 86.8) and two quaternary
sp2 carbons. The absence of direct H-H coupling
between methylene and methine protons required that they were not
adjacent, whereas long range H-C coupling between H-6 of the benzopyran
and the methylene carbon meant the latter was linked to C-5 of the
pyran. NMR spectra failed to provide conclusive data to support the
structure of this side chain but given the polyketide origin of LH-4 a
3-hydroxybut-2-enoic acid substituent is plausible and is not
contradicted by the chemical shift values (Table II). The tentative
structure proposed for LH-4, as depicted in Fig. 6, is
4-(2-carbamoyl-2,7-dihydroxy-4-oxo-chroman-5-yl)-3-hydroxy-but-2-enoic acid.
| |
DISCUSSION |
otcD1 Encodes a Didomain Cyclase/Aromatase--
The initial
observation made by Southern blotting, that the minimal PKS and KR for
OTC biosynthesis were non-contiguous (16), has been confirmed by DNA
sequencing and extended now to show that the cyclase gene,
otcD1, lies just downstream of the KR (Fig. 2). For most
polyketide clusters (e.g. actinorhodin; see Ref. 34), the
genes responsible for the assembly, folding, and cyclization of the
backbone are contiguous, although there are some exceptions (e.g. for frenolicin (11) and nogalacin (35)). However, it is notable that the KR and cyclase genes for OTC biosynthesis are
located a considerable distance (10 kb) from the minimal PKS (Fig. 2,)
with genes encoding a variety of disparate functions (e.g.
methylase, CoA ligase, and aminotransferase; see Ref. 26) between the
two locations. Thus in the otc cluster, the KR and cyclase
genes are expressed on a different mRNA from the PKS genes, which
implies that expression of both mRNAs will need to be tightly co-ordinated.
The gene product, OtcD1, shows end-to-end homology with putative gene
products from other aromatic polyketide gene clusters when comparisons
are made using the TBLASTN software (27). Most notable (56% identity,
75% similarity) is the similarity with the putative aromatase/cyclase
of the chlortetracycline biosynthetic cluster (36). There is also good
homology with gene products from other polyketide clusters,
e.g. SnoE of the nogalamycin cluster (53%, 64%) (35),
Gris-Orf4 from the griseucin producer (49%, 61%) (37), DpsF from the
daunorubicin producer (48%, 60%) (38), and ActVII (45%, 59%) (34)
from S. coelicolor, the producer of actinorhodin.
The N- and C-terminal segments of the deduced OtcD1 polypeptide share
homology with each other (26% identity, 35% similarity (Fig. 3)). It
is therefore reasonable to deduce that otcD1 encodes a
didomain ARO/CYC, which is able to catalyze formation of the carbon-carbon bond that closes a ring and subsequently dehydrates the
ring to aromatize it. Set against the existing literature (for review,
see Ref. 14), the gene product would be responsible for the closing and
aromatizing of the "left" ring (Fig.
1) of OTC.
ZGL3, a Strain Disrupted in otcD1, Makes Novel Polyketides of
Shorter Chain Length--
The four compounds isolated from the
S. rimosus ZGL3 strain are all classical polyketides
containing 9, 15, or 17 carbon atoms. They share a common biosynthetic
origin, in which three (LH-1), six (LH-4), or seven (LH-2 and LH-3)
acetate units have been added to a methylenecarboxamide precursor (Fig.
6). Thus, they have been made from a common starter unit through
different iterative cycles of the PKS complex.
Disruption of otcD1 has thus resulted in loss of specificity
of chain length: normally the 19-carbon backbone of OTC is made. The
"genetic surgery" performed on the otc cluster to
generate the ZGL3 recombinant was some 10 kb distant from the minimal
PKS (Fig. 2), which is implicated in
specificity of chain length. As otcD1 is transcribed on a
different mRNA, expression of the minimal PKS should not have been
altered through any polar effect.
Our current understanding of chain length elongation in polyketide
biosynthesis (e.g. see Ref. 39) stems from studies with the
pRM5/CH999 system in S. coelicolor (7). The host strain (CH999) is deleted for the entire actinorhodin cluster and the plasmid
(pRM5) includes transcriptional activation machinery to ensure high
expression of polyketide genes that can be added back to the null
strain in different combinations and contexts. By design, this system
investigates the molecular interactions among a small number of
different gene products. Using hybrid PKS gene sets, the pRM5/CH999
system has shown (7, 8) that the KS subunit is a major
determinant of chain length. However, specificity is not as absolute
as, for example, the frenolicin KS duo direct biosynthesis of a mixture
of octa- and nonaketides.
The data reported here adopts a different strategy to investigate gene
function, to ablate selectively a gene within a chromosomally located
cluster and investigate the result. It is presumed that all of the
other gene products within the cluster will be expressed normally with
the possible exception of those whose genes lie downstream on the same
transcript and may be subject to polarity. In the ZGL3 recombinant,
disruption of otcD1 has completely "derailed" the
capability of the PKS to synthesize polyketides of the usual chain
length. This is most likely due to a change in quaternary interactions
among the subunits of the PKS complex in the absence of the
cyclase/aromatase polypeptide to act as a structural partner. It is not
unprecedented that interactions between a cyclase/aromatase and minimal
PKS can influence the fidelity of determination of chain length:
(a) when the frenolicin minimal PKS was coexpressed with
TcmN, the tetracenomycin monodomain cyclase, a nonaketide was produced
whereas the minimal PKS alone produced an octaketide (39);
(b) when the C-terminal domain of the ARO/CYC of the
griseolin polyketide cluster was co-expressed with the minimal PKS from the tetracenomycin (tcm) cluster, a polyketide of 18 carbons
was made, in addition to the usual 20-carbon structure that the
tcm gene products specify (40); (c) when the
whiE minimal PKS was expressed alone it made an 18-carbon
product, but a 20-carbon polyketide was made when the whiE
CYC/ARO was co-expressed (37). However, the remarkable feature of the
products made by the ZGL3 recombinant is the series of polyketide
molecules with a wide spectrum of chain lengths, which are shorter, in
the case of LH-1 much shorter, than the wild-type strain. These
products form a "quasi-homologous" series of chain length.
Expression of the minimal PKS from the otc cluster in the
pRM5/CH999 system resulted in a polyketide (RM20b, Fig. 4) with a
backbone of 20 carbon atoms (9). The structural similarity between
RM20b and LH-3 and SEK4, whose synthesis is directed by the minimal PKS
from the actinorhodin pathway, is immediately apparent. The
act system directs synthesis of a 16-carbon structure. On
its own, the otc minimal PKS directs synthesis of a
20-carbon polyketide, whereas in wild-type S. rimosus the
19-carbon backbone of OTC is made. In the absence of otcD1, the backbone is reduced to 17 carbons or less. Thus, OtcD1 (as a
representative cyclase/aromatase) may have a greater role to play in
the determination of chain length than hitherto imagined.
The Novel Polyketides Have Aminated Termini, Implying That
They Are All Derived from the Same Aminated Precursor--
Each of the
four new structures made by strain ZGL3 has an aminated group,
consistent with malonamyl-CoA being used as the starter unit for
biosynthesis of these polyketides. In the literature (e.g.
see Ref. 26), it has been hypothesized that malonamyl-CoA is the
starter of OTC but, equally, malonyl-CoA could also act as starter with
the amination taking place subsequent to completion of the backbone.
The elegant work of Thomas and Williams (41, 42) used stable isotopic
labeling to elucidate the nature of extender units and folding pathway
of the backbone of OTC but did not address the issue of the starter
unit. The truncated compounds made by ZGL3 are different from OTC in
three-dimensional structure and also different from each other. It is
unlikely that they would all be substrates for a single enzyme that
performs an amination after the backbone is formed. This implies that
the carboxamido group is present at the start of biosynthetic process
and that malonamyl-CoA is very likely the natural starter unit for the OTC PKS.
The presence of the carboxamido group in each of the four structures
helps also to resolve the issue of whether truncated polyketide
products made by recombinant strains are the result of premature
termination of the iterative process of chain extension or degradation
products. All polyketide structures reported previously have been
derived solely from acetate units, so it was impossible to come to a
decision on this issue. Indeed, previous reports (e.g. see
Ref. 31) have included the caveat that the structures could, formally,
be the result of degradation of "full-length" chains. The
carboxamido group in each of the four novel structures reported here
provides a unique "tag" to indicate the point from which chain
inititation is made, and affords the conclusion that the structures are
the result of different numbers of iterations of chain extension.
Novel Polyketides Are the Result of Different Folding
Patterns--
Although LH-2 and LH-3 are derived from the same
progenitor compound, their final structures are the result of different
folding/cyclization patterns. LH-3 is the result of a cyclization
pattern (Fig. 6) for the first ring (C12/C7) that is typical for
cyclization of most aromatic polyketides, including actinorhodin (7).
LH-2 has the unusual C10/C15 cyclization pattern (Fig. 6). Formation of
the LH-2 and LH-3 pair thus parallels the situation with SEK4 and
SEK4b, which are similarly derived from a single progenitor through
different folding routes. The folding pattern of LH-3 is replicated in
the 15 carbon compound, LH-4. It is likely that each of the 4 compounds
has cyclized spontaneously in the absence of the otcD1 gene product.
The Role of KR in the otcD1::ermE Strain, ZGL3--
Each
of the novel polyketides, LH-1-LH-4, has a structure that implies that
no reduction of a keto group by the KR of the otc cluster
has taken place in the ZGL3 recombinant. This is a curious result
because the gene encoding the KR is upstream of otcD1. The
genetic construction to make the recombinant should have left the KR
intact. Three possibilities could plausibly explain this outcome: (i)
the KR protein cannot integrate into the PKS assembly when the
otcD1 gene product is missing (as is the case in the ZGL3
recombinant); (ii) although the gene encoding the KR is proximal to
otcD1 on the polycistronic mRNA, its level of expression
may be reduced by instability of the mutant mRNA containing the
disrupted ARO/CYC (otcD1::ermE); or (iii) the
nascent polyketide backbones of the progenitor structures of LH-1-LH-4
are not substrates for the otcKR. Interestingly, when the
otc minimal PKS was co-expressed in S. coelicolor
using the pRM5/CH999 system along with the KR (actIII) from
the actinorhodin pathway, then the resulting polyketide had
no reduced keto group (9). In this case, ActIII is either incapable of
forming a catalytic partnership with the otc minimal PKS or
the 20-carbon product formed by this minimal PKS is not a substrate
for ActIII.
| |
Addendum |
After submission of this manuscript, a paper was
published (Shen et al. (1999) Proc. Natl. Acad. Sci.
U. S. A. 96, 3622-3627), which shows that ectopic
expression of the minimal PKS for WhiE, the polyketide-derived spore
pigment of S. coelicolor results in a large number of
compounds with different chain length. The authors deduce that the
minimal PKS enzyme complex must rely on the stabilizing effects of
additional subunits (i.e. the cyclase) to ensure that the
chain reaches its full length.
| |
FOOTNOTES |
*
This work was supported by ALIS LINK Grant CRO007 (to
I. S. H. and D. H.) from the British Council and the
Ministry of Science and Technology, Croatia and by Grant 058407 from
the Ministry of Science and Technology, Croatia (to D. H.). We
thank the Governments of Slovenia (Ministry of Science and Technology
Award P38-0259-0490-007/13542/94 (to P. R.)) and Thailand for the
awards of a studentship and fellowship (to H. P. and A.T.,
respectively).The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement" in accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) AF087137.
**
To whom correspondence should be addressed: Dept. of Pharmaceutical
Sciences, University of Strathclyde, Royal College Building, Glasgow,
Scotland G1 1XW, UK. Tel.: 44-141-548-4111; Fax: 44-141-548-4124; E-mail: i.s.hunter@strath.ac.uk.
2
The numbering system used in the figures and
tables in this paper is non-IUPAC but follows that of Fu et
al. (33) in reflecting the biosynthesis of cyclic compounds from
the non-cyclic precursor.
| |
ABBREVIATIONS |
The abbreviations used are:
OTC, oxytetracycline;
HMBC, heteronuclear multiple bond connectivity;
PKS, polyketide synthase;
KR, ketoreductase;
CYC, cyclase;
ARO, aromatase;
kb, kilobase(s).
| |
REFERENCES |
| 1.
|
Chopra, I.,
Hawkey, P. M.,
and Hinton, M.
(1992)
J. Antimicrobial. Chemother.
29,
245-277[Free Full Text]
|
| 2.
|
Gatenbeck, S.
(1962)
Biochim. Biophys. Res. Commun.
6,
4-6[CrossRef][Medline]
[Order article via Infotrieve]
|
| 3.
|
Butler, M. J.,
Friend, E. J.,
Hunter, I. S.,
Kaczmarek, F. S.,
Sugden, D. A.,
and Warren, M.
(1989)
Mol. Gen. Genet.
215,
231-238[CrossRef][Medline]
[Order article via Infotrieve]
|
| 4.
|
Binnie, C.,
Warren, M.,
and Butler, M. J.
(1989)
J. Bacteriol.
171,
887-895[Abstract/Free Full Text]
|
| 5.
|
Hopwood, D. A.,
and Sherman, D. H.
(1990)
Annu. Rev. Genet.
24,
37-66[CrossRef][Medline]
[Order article via Infotrieve]
|
| 6.
|
Hopwood, D. A.,
and Khosla, C.
(1992)
in
Secondary Metabolites: Their Function and Evolution
(Chadwick, D. J.
, and Whelan, J., eds), Vol. 171
, pp. 88-112, Ciba Foundation Symposium, John Wiley and Sons, Chichester, UK
|
| 7.
|
McDaniel, R.,
Ebert-Khosla, S.,
Fu, H.,
Hopwood, D. A.,
and Khosla, C.
(1994)
Proc. Natl. Acad. Sci. U. S. A.
91,
11542-11546[Abstract/Free Full Text]
|
| 8.
|
McDaniel, R.,
Ebert-Khosla, S.,
Hopwood, D. A.,
and Khosla, C.
(1993)
J. Am. Chem. Soc.
115,
11671-11675[CrossRef]
|
| 9.
|
Fu, H.,
Ebert-Khosla, S.,
Hopwood, D. A.,
and Khosla, C.
(1994)
J. Am. Chem. Soc.
116,
6443-6444[CrossRef]
|
| 10.
|
Hutchinson, C. R.,
and Fujii, I.
(1995)
Annu. Rev. Microbiol.
49,
201-238[CrossRef][Medline]
[Order article via Infotrieve]
|
| 11.
|
Bibb, M. J.,
Sherman, D. H.,
Omura, S.,
and Hopwood, D. A.
(1994)
Gene
142,
31-39[CrossRef][Medline]
[Order article via Infotrieve]
|
| 12.
|
Davis, N. K.,
and Chater, K. F.
(1990)
Mol. Microbiol.
4,
1679-1691[Medline]
[Order article via Infotrieve]
|
| 13.
|
Summers, R. G.,
Wendt-Pienkowski, E.,
Motamedi, H.,
and Hutchinson, C. R.
(1992)
J. Bacteriol.
174,
1810-1820[Abstract/Free Full Text]
|
| 14.
|
Hopwood, D. A.
(1997)
Chem. Rev.
97,
2465-2497[CrossRef][Medline]
[Order article via Infotrieve]
|
| 15.
|
Malpartida, F. M.,
Hallam, S. E.,
Kieser, H. M.,
Motamedi, H.,
Hutchinson, C. R.,
Butler, M. J.,
Sugden, D. A.,
Warren, M.,
McKillop, C.,
Bailey, C. R.,
Humphreys, G. O.,
and Hopwood, D. A.
(1987)
Nature
325,
818-821[CrossRef][Medline]
[Order article via Infotrieve]
|
| 16.
|
Butler, M. J.,
Binnie, C.,
Hunter, I. S.,
Sugden, D. A.,
and Warren, M.
(1990)
Dev. Ind. Microbiol.
31,
41-50
|
| 17.
|
Kim, E. S.,
Bibb, M. J.,
Butler, M. J.,
Hopwood, D. A.,
and Sherman, D. H.
(1994)
Gene
141,
141-142[CrossRef][Medline]
[Order article via Infotrieve]
|
| 18.
|
Hranueli, D.,
Pigac, J.,
and Ve ligaj, M.
(1979)
J. Gen. Microbiol.
114,
295-303[Medline]
[Order article via Infotrieve]
|
| 19.
|
Sambrook, J.,
Fritsch, E. F.,
and Maniatis, T.
(1989)
Molecular Cloning: A Laboratory Manual
, 2nd Ed.
, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
|
| 20.
|
Hopwood, D. A.,
Bibb, M. J.,
Chater, K. F.,
Kieser, T.,
Bruton, C. J.,
Kieser, H. M.,
Lydiate, D. J.,
Smith, C. P.,
Ward, J. M.,
and Schrempf, H.
(1985)
Genetic Manipulation of Streptomyces; A Laboratory Manual
, The John Innes Foundation, Norwich, UK
|
| 21.
|
Pigac, J.,
and Schrempf, H.
(1995)
Appl. Environ. Microbiol.
61,
352-356[Abstract]
|
| 22.
|
Sanger, F.,
Nicklen, S.,
and Coulson, A. R.
(1977)
Proc. Natl. Acad. Sci. U. S. A.
74,
5463-5467[Abstract/Free Full Text]
|
| 23.
|
Williams, D. H.,
and Fleming, I.
(1997)
Methods in Organic Chemistry
, 5th Ed.
, McGraw-Hill Publishing Co., London, UK
|
| 24.
|
Sanders, J. K. M.,
and Hunter, B. K.
(1987)
Modern NMR Spectrospcopy: A Guide for Chemists
, Oxford University Press, Oxford, UK
|
| 25.
|
Hallam, S. E.,
Malpartida, F.,
and Hopwood, D. A.
(1988)
Gene
74,
305-320[CrossRef][Medline]
[Order article via Infotrieve]
|
| 26.
|
Hunter, I. S.,
and Hill, R. A.
(1997)
in
Bio/Technology of Antibiotics
(Strohl, W. R., ed), 2nd Ed.
, pp. 659-682, Marcel Decker, Inc., New York
|
| 27.
|
Altschul, S. F.,
Madden, T. L.,
Schaffer, A. A.,
Zhang, J.,
Zhang, Z.,
Miller, W.,
and Lipman, D. J.
(1997)
Nucleic Acids Res.
25,
3389-3402[Abstract/Free Full Text]
|
| 28.
|
Bibb, M. J.,
Janssen, G. R.,
and Ward, J. M.
(1985)
Gene
38,
215-226[CrossRef][Medline]
[Order article via Infotrieve]
|
| 29.
|
Ward, J. M.,
Janssen, G. R.,
Kieser, T.,
Bibb, M. J.,
Buttner, M. J.,
and Bibb, M. J.
(1986)
Mol. Gen. Genet.
203,
468-478[CrossRef][Medline]
[Order article via Infotrieve]
|
| 30.
|
Southern, E. M.
(1975)
J. Mol. Biol.
98,
503-517[CrossRef][Medline]
[Order article via Infotrieve]
|
| 31.
|
Yu, T. W.,
Shen, Y.,
McDaniel, R.,
Floss, H. G.,
Khosla, C.,
Hopwood, D. A,
and Moore, B. S.
(1998)
J. Am. Chem. Soc.
120,
7749-7759[CrossRef]
|
| 32.
|
Agarwal, P. K.
(1989)
Carbon-13 NMR of Flavonoids
, Elsevier, Amsterdam
|
| 33.
|
Fu, H.,
Ebert-Khosla, S.,
Hopwood, D. A.,
and Khosla, C.
(1994)
J. Am. Chem. Soc.
116,
4166-4168[CrossRef]
|
| 34.
|
Fernandez-Moreno, M. A.,
Martinez, E.,
Boto, L.,
Hopwood, D. A.,
and Malpartida, F. J.
(1992)
J. Biol. Chem.
267,
19278-19290[Abstract/Free Full Text]
|
| 35.
|
Ylihonko, K.,
Tuikkanen, J.,
Jussila, S.,
Cong, L.,
and Mantsala, P.
(1996)
Mol. Gen. Genet.
251,
113-120[Medline]
[Order article via Infotrieve]
|
| 36.
| Ryan, M., Lotvin, J., Strathy, N., and Fantini, S. (1996) U. S. Patent
5589385-A
|
| 37.
|
Yu, T. W.,
Bibb, M. J.,
Revill, W. P.,
and Hopwood, D. A.
(1994)
J. Bacteriol.
176,
2627-2634[Abstract/Free Full Text]
|
| 38.
|
Ye, J.,
Dickens, M. L.,
Plater, R.,
Li, Y.,
Lawrence, J.,
and Strohl, W. R.
(1994)
J. Bacteriol.
176,
6270-6280[Abstract/Free Full Text]
|
| 39.
|
Kramer, P. J.,
Zawada, R. J.,
McDaniel, R.,
Hutchinson, C. R.,
Hopwood, D. A.,
and Khosla, C.
(1997)
J. Am. Chem. Soc.
119,
635-639[CrossRef]
|
| 40.
|
Zawada, R. J. X.,
and Khosla, C.
(1997)
J. Biol. Chem.
272,
16184-16188[Abstract/Free Full Text]
|
| 41.
|
Thomas, R.,
and Williams, D. J.
(1983)
J. Chem. Soc. Chem. Commun.
7,
443-444[CrossRef]
|
| 42.
|
Thomas, R.,
and Williams, D. J.
(1983)
J. Chem. Soc. Chem. Commun.
12,
128-130[CrossRef]
|
| 43.
|
Devereux, J.,
Haeberli, P.,
and Smithies, O.
(1984)
Nucleic Acids Res.
12,
387-395
|
Copyright © 1999 by The American Society for Biochemistry and Molecular Biology, Inc.
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
|
|
 
W. Zhang, K. Watanabe, C. C. C. Wang, and Y. Tang
Investigation of Early Tailoring Reactions in the Oxytetracycline Biosynthetic Pathway
J. Biol. Chem.,
August 31, 2007;
282(35):
25717 - 25725.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
H. Petkovic, J. Cullum, D. Hranueli, I. S. Hunter, N. Peric-Concha, J. Pigac, A. Thamchaipenet, D. Vujaklija, and P. F. Long
Genetics of Streptomyces rimosus, the Oxytetracycline Producer
Microbiol. Mol. Biol. Rev.,
September 1, 2006;
70(3):
704 - 728.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
W. Zhang, B. D. Ames, S.-C. Tsai, and Y. Tang
Engineered Biosynthesis of a Novel Amidated Polyketide, Using the Malonamyl-Specific Initiation Module from the Oxytetracycline Polyketide Synthase
Appl. Envir. Microbiol.,
April 1, 2006;
72(4):
2573 - 2580.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
N. Peric-Concha, B. Borovicka, P. F. Long, D. Hranueli, P. G. Waterman, and I. S. Hunter
Ablation of the otcC Gene Encoding a Post-polyketide Hydroxylase from the Oxytetracyline Biosynthetic Pathway in Streptomyces rimosus Results in Novel Polyketides with Altered Chain Length
J. Biol. Chem.,
November 11, 2005;
280(45):
37455 - 37460.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
