Biosynthesis of pyrrolylpolyenes in Auxarthron umbrinum

Article abstract of DOI:10.1039/b813236d

The biosynthesis of the pyrrolylpolyene rumbrin (1) in the fungus Auxarthron umbrinum was elucidated using feeding studies with labelled precursors. Incorporation of stable isotopes from [¹⁵N]-proline, [¹³C]-methionine and [¹³

Full text of DOI:10.1039/b813236d

View Article Online / Journal Homepage / Table of Contents for this issue
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Biosynthesis of pyrrolylpolyenes in Auxarthron umbrinum
Benjamin R. Clark and Cormac D. Murphy*
Received 31st July 2008, Accepted 24th September 2008
First published as an Advance Article on the web 31st October 2008
DOI: 10.1039/b813236d
The biosynthesis of the pyrrolylpolyene rumbrin (1) in the fungus Auxarthron umbrinum was
elucidated using feeding studies with labelled precursors. Incorporation of stable isotopes from
[15N]-proline, [13C]-methionine and [13C]-acetate confirmed that these were the precursors of the pyrrole
moiety, methyl groups, and backbone of rumbrin, respectively. Label-dilution experiments with
pyrrole-2-carboxylate confirmed it was a direct precursor in the biosynthesis of rumbrin.
Both 3- and 4-chloropyrrolecarboxylates were also accepted as precursors in polyene
production.
Introduction
Results and discussion
Rumbrin (1) is an unusual pyrrolylpolyene produced by a small
number of fungi from the order Onygenales. First isolated in
1993 from an Auxarthron umbrinum strain,1,2 it has since been
isolated a number of times along with several isomers, in addi-
tion to closely related polyenes including the auxarconjugatins3
and gymnoconjugatins.4 Rumbrin possesses several interesting
biological properties, including anticancer4 and cytoprotective
activities.1 Despite its intriguing structure, very little is known
about the biosynthesis of rumbrin. A single feeding study has
been carried out using [1,2-13C2]-acetate, confirming the polyketide
nature of the polyene; however the study was conducted for
structural elucidation purposes, and the biosynthetic implications
were not addressed.2 More recently, molecular techniques have
been used to probe the biosynthesis of other fungal polyenes,
showing them to be biosynthesised by iterative polyketide synthase
(PKS) systems.5,6 However, these polyenes lack the highly unusual
3-chloropyrrole moiety and potent biological activities of rumbrin.
In this study we describe investigations into the biosynthesis of
rumbrin using feeding studies with both isotopically labelled and
unlabelled precursors.
The commercially available strain A. umbrinum DSM-3193 was
used for the studies. Polyene production was assessed by HPLC
and MS analysis under a range of culture conditions and was found
to be best using the rich medium employed by Alvi et. al. 7 with pro-
duction beginning after approximately four days growth. Polyene
production can also be visually monitored by the development of
a deep red colour in the fungal mycelium. While some strains used
in previous studies have produced almost exclusively chlorinated
polyenes,8 the strain used in the current investigation produced
both non-chlorinated and chlorinated products, in a ratio of ~3:1,
as determined by HPLC and MS analysis. The major product
was dechloroisorumbrin (2), though rumbrin (1) and a range
of isomers and related metabolites such as the auxarconjugatins
were also detectable (Fig. 1a). Monitoring the production in A.
umbrinum cultures over a one month period showed no significant
change in the composition of the polyenes over time. Given the
relatively high production of 2, most of the biosynthetic studies
focused on this as a model for production of all polyenes.
School of Biomolecular and Biomedical Sciences, Centre for Synthesis
and Chemical Biology, Ardmore House, University College Dublin,
Belfield, Dublin 4, Republic of Ireland. E-mail: cormac.d.murphy@ucd.ie;
Fax: (+353 1) 716 1183; Tel: (+353 1) 716 1311
Fig. 1 HPLC analysis (DAD, 440 nm) of polyenes produced by A.
umbrinum DSM-3193 cultures. (a) Unsupplemented medium, (b) medium
plus 2 mM pyrrolecarboxylate (4), (c) plus 2 mM 5, (d) plus 2 mM 6.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 111–116 | 111
©

View Article Online
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Table 1 1H and 13C NMR data (500 MHz, d6-DMSO) and %13C
incorporation for auxarconjugatin D (3) from [1-13C]-acetate.
Stable isotope labelling studies
It was thought that the polyene backbone of rumbrin was likely
to be derived from acetate, with the pyrrole moiety originating
from proline and the methyl groups from methionine. Thus, [1-
13C]-acetate, [15N]-L-proline, and [methyl-13C]-L-methionine were
added to growing cultures of A. umbrinum DSM-3193 in order
to determine whether they were incorporated into the polyene.
Precursors were introduced to the fungal culture after three days,
immediately prior to the commencement of polyene biosynthesis.
After seven days the cultures were filtered, the mycelium extracted
with MeOH-CH2Cl2 (3:1), and the extracts analysed by electro-
spray mass spectrometry (ESI(+)MS) and HPLC. Incorporation
of precursors was assessed by comparison of peak heights at m/z
324 and 325 (dechloroisorumbrin, M+1 and M+2) in the mass
spectrum.
dH (m, J (Hz))
dC
%13Ca,b
N-1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
OMe
11.16 (br s)
6.85 (dd, 3.8, 2.6)
6.07 (dd, 3.0, 2.6)
6.24 (br s)
120.7
109.4
110.1
130.5
125.4
123.2
137.4
129.8
139.2
129.7
135.3
121.1
158.5
100.5
170.8
88.3
9.4
1.1
0.8
1.0
5.2
1.2
6.2
1.1
6.9
1.2
5.6
1.6
5.1
1.1
4.9
1.1
5.6
1.1
6.54 (d, 15.6)
6.65 (dd, 15.6, 11.2)
6.55 (dd, 14.1, 11.2)
6.35 (dd, 14.1, 11.4)
6.73 (dd, 14.5, 11.4)
6.43 (dd, 14.5, 11.3)
7.04 (dd, 15.2, 11.3)
6.27 (d, 15.2)
6.22 (d, 2.0)
5.58 (d, 2.0)
3.81 (s)
When [15N]-L-proline was added to the culture medium at
concentrations of 2 and 5 mM, isotope incorporation of 14%
and 43% was observed, confirming proline as a precursor. In
comparison, feeding with a biosynthetically unrelated amino
acid such as [15N]-glycine gave a much lower incorporation
at the same concentations (5.1% and 8.1%, respectively). This
incorporation presumably arose from enrichment of the cellular
nitrogen pool by deamination of glycine, for example via the
actions of serine hydroxymethyltransferase, which converts two
162.6
56.2
a Calculated by comparison of the 13C peak heights relative to a 13C
spectrum of unlabelled 3. The OMe peak was used as an internal standard
and defined as 1.1%. b Numbers in bold are the enriched positions.
However, two further signals at 120.7 and 125.4 ppm, C-2 and C-6,
were also enhanced. This was initially perplexing, as this portion
of the molecule was believed to be proline-derived. On further
consideration the origin of these labels became clear: proline is
biosynthesised from glutamate, which is in turn derived from a-
ketoglutarate, an intermediate in the acetate-fuelled TCA cycle.
Thus, C-1 of acetate is incorporated into proline at the C-1 and
C-5 positions, and subsequently into the polyene as C-6 and C-2
(Scheme 1).
The incorporation of these precursors is consistent with what
is already known about production of fungal polyketides. Given
the polyketide nature of the polyene, incorporation of acetate
was expected, though the high retention of label in the proline-
derived portion of the molecule was surprising, and was probably
a consequence of the high concentration of labelled acetate used
in the experiment.
+
molecules of glycine to serine with the liberation of NH4 , and
+
serine dehydratase, which converts serine to pyruvate and NH4 .
Mass spectral evidence suggested that [methyl-13C]-methionine
(3 mM) was also incorporated into dechlororumbrin (2), since
both single and double isotope incorporation of 13.7% and 5.5%,
respectively, were detected. In order to determine the location
of the 13C labels, the crude fungal extract was purified by solvent
partitioning and reversed phase SPE to yield a mixture of polyenes.
13C NMR of this crude polyene mixture revealed enhancements in
clusters of 13C resonances at 8.7, 12.8, 55.5 ppm, corresponding
to the 13-methyl, 17-methyl, and 16-O-methyl resonances of the
polyenes, respectively. For dechloroisorumbrin (2), percentage
incorporation at each position was 8.5%, 6.1%, and 10.7%. A
cluster of peaks at 21.7 ppm, corresponding to the 13-methyl group
of a 12Z-polyene,2 was also enhanced (8.3%).
The incorporation of [methyl-13C]-L-methionine is consistent
with previous studies. Pendant methyl groups on fungal polyke-
tides are invariably derived from the S-methyl group of methio-
nine, via S-adenosyl methionine.9,10 The production of polyenes
with a variable degree of methylation by A. umbrinum DSM-
3193 suggests that methylation by the hypothetical PKS cluster
is unpredictable, and that a methyltransferase module may be
“skipped” during the biosynthetic process.11
The biosynthesis of pyrrole-containing natural products has
not previously been investigated in fungi, and this represents
the first investigation of the biosynthesis of a pyrrole-containing
metabolite. Our observations that proline is the biosynthetic origin
of the pyrrole moiety has also been recognized for several bacterial
pyrrole-containing natural products.12
[1-13C]-Acetate (25 mM) was also incorporated into dechloro-
isorumbrin, presumably as malonate, with multiple 13C-labeled
acetates incorporated into a single molecule. In the mass spectrum,
species containing a single label (17.9%), double label (12.3%),
triple label (11.9%), four labels (8.9%) and >4 labels (4.8%)
were all detected. 13C NMR of the partially purified polyenes (as
above) gave a complex mixture of peaks that prevented a rigorous
analysis. Thus, further purification was carried out using reversed-
phase HPLC to yield several purified polyene metabolites. Due
to difficulties in the purification of dechloroisorumbrin (2),
the purified polyene selected for further 13C analysis was the
didemethyl polyene auxarconjugatin D (3).† Enhancement of
the 13C NMR signals at 137.4, 139.2, 135.3, 158.5, 170.7, and
162.6 ppm revealed the incorporation of the C-1 of acetate at
C-8, C-10, C-12, C-14, C-16, and C-18, as expected (Table 1).
Incorporation of pyrrole-2-carboxylates
With the origins of the atoms established, the next step was to
see if any more advanced intermediates could be incorporated.
Biosynthesis of bacterial pyrrole-containing antibiotics, such
† The structure of 3, a new metabolite, was determined by comparison
of 1H and 13C NMR shifts (Table 1) with those of related metabolites, in
particular auxarconjugatins B and C.3
112 | Org. Biomol. Chem., 2009, 7, 111–116
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Table 2 Relative proportions of chlorinated and non-chlorinated
polyenes produced upon addition of various pyrrole-2-carboxylic acids
to the culture medium
Pyrrole-2-carboxylic Total polyenesa Non-chlorinated Chlorinated
acid(s) (2 mM)
(mg/L)
polyene (%)
polyene (%)
None
4
106
174
111
123
135
194
77
88
61
76
48
80
23
12
39
24
52
20
5
5+4
6
6+4
a Determined by integration of peak areas at 440 nm.
as pyoluteorin and coumermycin A1 proceeds with pyrrolyl-
2-carboxyl (originally proline) tethered to a peptidyl carrier
protein.13,14 Nevertheless, experiments were conducted to examine
to possibility of free pyrrole-2-carboxylate (4) being incorporated
into the polyenes. Thus, a label dilution experiment, where both
[15N]-L-proline and pyrrole-2-carboxylate (2 mM each) were added
to an A. umbrinum culture, was conducted. Control experiments
with no substrate, [15N]-L-proline only, and 4 only were carried
out in parallel. Cultures were worked up and analysed as before.
The results were striking: while the proline-only extract showed
excellent incorporation (22%), as before, the label-dilution extract
showed a greatly reduced incorporation of [15N]-proline (2.6%),
indicating that 4 is a direct precursor in the biosynthesis of
the polyenes. Furthermore, in both cultures to which pyrrole-2-
caboxylate (4) had been added, drastic changes were observed in
the HPLC profile of the polyenes, with a noticeable increase in
the production of 2 and other non-halogenated polyenes, while
production of chlorinated polyenes was not enhanced (Fig. 1b,
Table 2). This effect was further investigated by culturing A.
umbrinum in medium containing varying amounts of pyrrole-2-
carboxylate (0, 1, 3, and 10 mM), which showed that polyene
production was enhanced further with higher concentrations
of pyrrolecarboxylate (Fig. 2) – a five-fold increase in polyene
production was observed with the addition of 10 mM of 4.
Scheme 1 Biosynthetic origin of atoms in dechlororumbrin (2).
Scheme
2 Synthesis of chloropyrroles. Reagents and conditions:
(i) CuSO4, H2O, hn, rt, 18 h, 33%; (ii) Ag2O, H2O/MeOH, rt, 18 h, 85%;
(iii) SO2Cl2, CHCl3, 0–25◦C, 1 h; (iv) NaOH, rt, 30 min, 35% over 2 steps.
in the presence of CuSO415,16 to yield 3-chloro-2-formyl pyrrole (8)
in 35% yield. 8 was subsequently oxidised with Ag2O to give the
final product.17 4-Chloropyrrole-2-carboxylate (6) was synthesised
using literature methods from 2-trichloroacetylpyrrole (9).14,18
An analogous feeding study was conducted as before, revealing
that addition of 5 slightly enhanced the proportion of chlorinated
polyenes (Table 2), although the overall amount of polyenes did
not noticeably increase. In contrast, the addition of 6 significantly
boosted production of chlorinated metabolites, and led to the
appearance of a new peak (16.7 min, Fig. 1). While this compound
was not isolated, based on its mass spectrum (m/z 358/360,
M+H+), characteristic pyrrolylpolyene UV-vis spectrum, and on
biosynthetic arguments, we propose the structure 10. Previous
studies have shown that the substitution of the pyrrole ring has
a pronounced effect on the cytotoxic and anticancer activity of
these polyenes,8 and this discovery opens the way for further SAR
studies.
Fig. 2 Increased polyene production by A. umbrinum DSM-3193 with
addition of pyrrole-2-carboyxlate (4). Average of duplicates.
The confirmation of pyrrole-2-carboxylate as a precursor of
dechloroisorumbrin prompted investigation of the possible incor-
poration of chlorinated pyrroles into the polyenes. Both 3- and
4-chloropyrrole-2-carboxylate (5, 6) were synthesized for use in
feeding studies (Scheme 2). 5 was synthesized in two steps from
4-chloro-pyridine-N-oxide (7), which was subjected to photolysis
Label-dilution studies using chloropyrroles and [15N]-L-proline
were conducted as before. In an experiment in which 2 mM of
both 5 and [15N]-L-proline were added to a growing culture of
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 111–116 | 113
©

View Article Online
Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
A. umbrinum, 15N incorporation in dechloroisorumbrin (2) re-
mained high (15.3%) while incorporation into rumbrin (1) was
greatly reduced (3.1%). Similar results were observed for 6, with
moderate incorporation into 2 detected (9.1%), but almost none
into the corresponding chlorinated polyene (2.1%).
Given the low incorporation of 5 relative to 4 and 6, we suggest
that while 5 is accepted as a substrate, it is not the natural precursor
for rumbrin (1), and thus chlorination does not occur on the free
pyrrole-2-carboxylate itself but at a later stage in the biosynthetic
pathway.
substitution. In contrast, the chlorine atom in rumbrin (1) occurs
at C-3 of the pyrrole unit (C-4 of rumbrin), suggesting that
in this instance halogenation occurs after polyketide extension
and formation of the polyene chain. Whether this occurs on the
finished polyene or on an ACP-bound intermediate species is
unknown. For completeness the cell-free extract was also assayed
for haloperoxidase activity, but none was detected.
Conclusions
In the biosynthesis of bacterial pyrrole-containing metabolites
such as pyoluteorin and clorobiocin, the prolyl group is oxidised
to a pyrrolecarboxyl moiety as a peptidyl carrier protein-bound
species.13,19 In clorobiocin biosynthesis, this is then transferred via
several protein-bound intermediates to the growing metabolite.20
In contrast, the facile incorporation of pyrrole-2-carboxylate (4)
into rumbrin suggests that free 4, rather than an enzyme-bound
derivative, is taken up directly by the PKS as the starter unit for
rumbrin biosynthesis. Similarly, we propose that the oxidation of
proline to pyrrolecarboxylate occurs on the free amino acid and
not a PCP-tethered derivative. The extension of the polyketide
via the incorporation of acetate units is presumably catalyzed in
the classical PKS fashion (Scheme 3), with subsequent reduction,
dehydration, and methylation yielding the mature polyene.
Precursor-directed feeding studies have illustrated the origin of
the atoms in the cytotoxic fungal polyene rumbrin (1), which is
biosynthesed from proline, methionine, and acetate, probably by
an iterative polyketide synthase enzyme complex. This represents
the first biosynthetic study of a fungal pyrrolyl polyene. Label
dilution experiments also confirmed pyrrole-2-carboxylate (4)
as a direct precursor. This suggests that the incorporation of
the pyrrole proceeds via a different mechanism to that which
that occurs in the biosynthesis of bacterial pyrrole-containing
metabolites. The chlorinated pyrrole 5 was also accepted as a
substrate, though its low incorporation leads us to believe that
it is not a natural precursor. Remarkably, the fungus was also able
to incorporate the “un-natural” chloropyrrole 6 into the polyenes
in high amounts. Further studies on the incorporation of other
substituted pyrrolecarboxylates are in progress, and may allow the
production of polyenes with improved biological properties.
Experimental
Auxarthron umbrinum DSM-3193 was obtained from the German
Collection of Microorganisms and Cell Cultures (DSMZ). All
chemicals were purchased from Sigma-Aldrich, except for labelled
substrates, purchased from Cambridge Isotope Laboratories. All
solvents were HPLC grade and used without further purification.
Analytical TLC was performed on pre-coated plates (0.2 mm
silica gel 60 on aluminium, Machery-Nagel). Compounds were
visualised by staining with anisaldehyde dip and heating on a hot
plate. Flash chromatography was performed on silica gel 60 (230–
400 mesh, Merck). UV irradiation was conducted using a Phillips
TUV 30W/G30 T8 Longlife Hg UV lamp.
HPLC was carried out using a Varian Prostar system consisting
of two solvent delivery modules (210), DAD detector (335),
autoinjector (410) and fraction collector (710). NMR was con-
ducted using Varian Inova 300, 400, and 500 MHz spectrometers,
and referenced to residual proton signals in the solvent (for
DMSO, dH 2.50 and dC 39.5 ppm). IR spectra were recorded
on a Varian 3100 Excalibur Series FT-IR spectrometer, and UV-
vis spectra on a Thermo-Spectronic Helios-b spectrophotometer.
Low resolution mass spectra were obtained using a Micromass
Quattro mass spectrometer, and high resolution spectra on a
Micromass LCT time-of-flight mass spectrometer, both coupled
to a Waters Alliance 2695 solvent delivery system.
Scheme 3 Proposed biosynthesis of 2.
Biochemical chlorination assays
Halogenation of the pyrrole-containing bacterial metabolite py-
oluteorin occurs via the actions of a FADH2-dependent halo-
genase, PltA.14 Thus the chlorination of rumbrin was further
probed by incubating cell extracts of A. umbrinum DSM-3193
with NADH, NaCl plus either dechlororumbrin or pyrrole-2-
carboxylate. However, no chlorinated product was detected by
HPLC. However, it should be noted that the solubility of 2 in
aqueous environments is very poor. In many halogenated pyrrole-
containing natural products, C-4 of the pyrrole ring is the primary
halogenation location. In many cases, this can be attributed
to the presence of an electron withdrawing carbonyl group at
C-2, which deactivates the C-3 and C-5 positions to electrophilic
General procedure for feeding studies
A. umbrinum DSM-3193 was grown in 250 mL flasks containing
25 mL of a medium consisting of 2% glucose, 1.5% Pharmame-
dia (cottonseed flour), 0.5% yeast extract, 0.4% CaCO3, 0.3%
(NH4)2SO4, and 0.003% ZnSO4.7H2O, at 28 ◦C on a rotary
114 | Org. Biomol. Chem., 2009, 7, 111–116
This journal is
The Royal Society of Chemistry 2009
©