A sweet origin for the key congocidine precursor 4-acetamidopyrrole-2- carboxylate

Article abstract of DOI:10.1002/anie.201201445

Feeding (Streptomyces) frenzy: Natural products belonging to the pyrrolamide family are defined by their pyrrole-2-carboxamide moiety. 4-acetamidopyrrole-2-carboxylate is identified as the key pyrrolamide congocidine precursor (see scheme) through feeding studies using Streptomyces ambofaciens. The biosynthetic pathway of congocidine starts with the carbohydrate N-acetylglucosamine and involves carbohydrate-processing enzymes. Copyright

Full text of DOI:10.1002/anie.201201445

Angewandte
Chemie
DOI: 10.1002/anie.201201445
Biosynthetic Pathway
A Sweet Origin for the Key Congocidine Precursor 4-
Acetamidopyrrole-2-carboxylate**
Sylvie Lautru,* Lijiang Song, Luc Demange, Thomas Lombꢀs, Hervꢁ Galons,
Gregory L. Challis, and Jean-Luc Pernodet
Pyrrole groups are found in nature in two primary metabolites
(heme and tryptophan) and in a large variety of secondary
metabolites (e.g. pyoluteorin, clorobiocin, congocidine, and
prodigiosins). The diversity of these metabolites is mirrored
in the multiple biosynthetic pathways leading to pyrrole
groups. Six different biosynthetic pathways have been char-
acterized to date,[1,2] involving various primary metabolite
precursors, such as amino acids (glycine, proline, serine,
threonine, and tryptophan),[1–10] dicarboxylic acids (malonate,
oxaloacetate, and succinate),[1,2,7,8] or N-(5’-phosphoribosyl)-
anthranilate.[1]
Pyrrolamides are a family of natural products, synthesized
by Streptomyces and related actinobacteria, that all contain
one or more pyrrole-2-carboxamide moieties in their struc-
ture. Most pyrrolamides, such as the well-characterized
congocidine (1, also called netropsin; Figure 1) and distamy-
cin, bind noncovalently in the DNA minor groove with some
sequence specificity,[11] but pyrronamycin B has been sug-
gested to bind DNA covalently.[12] This capacity to bind DNA
confers on them a variety of biological activities, such as
antiviral, antibacterial, antitumor, and anthelmintic activities,
but also renders them too toxic for clinical use. Nonetheless,
because molecules binding to DNA at specific positions can
manipulate the expression of genes involved in various
diseases (such as cancer), congocidine and distamycin have
prompted the synthesis of many structurally related mole-
cules that bind in the minor groove of DNA at various defined
sequences.[13]
The nature and biosynthetic origin of the pyrrolamide
pyrrole precursor are unknown. A retrobiosynthetic analysis
of pyrrolamide structures suggests 4-aminopyrrole-2-carbox-
ylate as the potential pyrrole precursor common to all
pyrrolamides. However, this remains to be established and
no biosynthetic pathway has been reported for the synthesis
of this molecule. We report herein that 4-acetamidopyrrole-2-
carboxylate (10) is the true precursor of congocidine and
propose for this compound a biosynthetic pathway starting
from N-acetylglucosamine-1-phosphate (2), involving carbo-
hydrate metabolizing enzymes, and differing entirely from
known pyrrole biosynthetic pathways.
We recently reported the identification, analysis, and
heterologous expression of the first pyrrolamide gene cluster;
the cgc gene cluster directs congocidine biosynthesis in
Streptomyces ambofaciens.[14] Sequence analyses of the pro-
teins encoded by the cgc cluster led us to propose a pathway
for congocidine biosynthesis involving a noncanonical non-
ribosomal peptide synthetase (NRPS) and three putative
precursors: guanidinoacetate, 3-aminopropionamidine, and 4-
aminopyrrole-2-carboxylate. However, the biosynthetic ori-
gins of these precursors could not easily be inferred from the
analysis of the cgc gene cluster. In particular, we could not
identify any gene encoding homologues of known pyrrole
biosynthetic enzymes, suggesting that the pyrrole groups in
pyrrolamides could be synthesized through an entirely new
pathway.
[*] Dr. S. Lautru, Dr. J.-L. Pernodet
Institut de Gꢀnꢀtique et Microbiologie
Universitꢀ Paris-Sud, CNRS, UMR8621, 91405 Orsay (France)
E-mail: sylvie.lautru@igmors.u-psud.fr
Dr. L. Song, Prof. G. L. Challis
Department of Chemistry, University of Warwick,
Coventry CV4 7AL (UK)
Dr. L. Demange, Prof. H. Galons
Laboratoire de Chimie et Biochimie Pharmacologiques et Toxico-
logiques UMR 8601 CNRS
Universitꢀ Paris Descartes, Sorbonne Paris Citꢀ, UFR Biomꢀdicale
des Saints Pꢁres
45 rue des Saints Pꢁres, 75006 Paris (France)
The cgc18 gene encodes an NRPS module that has been
proposed to recognize 4-aminopyrrole-2-carboxylate and
catalyze its ATP-dependent activation as the corresponding
peptidyl carrier protein (PCP)-bound thioester.[14] To inves-
tigate whether 4-aminopyrrole-2-carboxylate is indeed a pre-
cursor of congocidine, we used HPLC to compare the profile
of metabolites in the culture supernatants of the “wild type”
SPM110 Streptomyces ambofaciens strain and the CGCA004
mutant strain, in which cgc18 has been disrupted.[14] A new
compound accumulated in the culture supernatant of
CGCA004 (retention time = 12.3 min, Figure 2a), which we
expected to be 4-aminopyrrole-2-carboxylate. This metabo-
lite peak was not detected in the CGCA004 chromatogram of
our previous report[14] and on average, it was detected in
approximately half of the cultures of strains deleted in cgc
T. Lombꢁs
UMR 8638 CNRS-Paris Descartes, Facultꢀ des Sciences Pharma-
ceutiques et Biologiques
4 av. de l’Observatoire, 75006 Paris (France)
[**] We would like to thank Dr. Christophe Corre for helpful discussions,
Dr. Laurent Micouin for his help with the synthesis of the 4-
acetamidopyrrole-2-methanol, and Dr. Cyrille Kouklovsky, Gꢀraldine
Le Goff, and Dr. Robert Thaꢂ for their help with mass spectrometry.
This work was supported in part by the European Union through the
Integrated Project ActinoGEN (CT-2004-0005224) and the reinte-
gration grant BOPA (CT-2005-029154), by the Pꢃle de Recherche et
d’Enseignement Supꢀrieur UniverSud Paris and by the Bettencourt
Schueller Foundation.
Supporting information for this article (experimental details) is
available on the WWW under http://dx.doi.org/10.1002/anie.
201201445.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Figure 1. Congocidine gene cluster (a) and proposed pathway for biosynthesis of 4-acetamidopyrrole-2-carboxylate (10) from fructose-6-phosphate
by way of an intermediate 2 (b). [U-13C]Glycerol is proposed to be converted into [U-13C]glyceraldehyde-3-phosphate, which undergoes
transaldolase-catalyzed condensation with seduheptulose-7-phosphate with loss of erythrose-4-phosphate to yield fructose-6-phosphate bearing
13C labels at C-4, C-5, and C-6. The positions of the 13C labels are highlighted in red. The primary metabolic enzymes GlmSMU catalyze conversion
of fructose-6-phosphate into 2. Genes involved in 10 biosynthesis are indicated in green.
genes not involved in the biosynthesis of 10. It is likely that
this variability is related to the variability of secondary
metabolite production often seen in Streptomyces.[15] LC-MS
analysis of the CGCA004 culture supernatant showed that
this metabolite produces an m/z 168.9 ion rather than the
expected m/z 127.1 ion (Supporting Information, Figure S1a).
MS/MS analyses suggested that the accumulated metabolite
was 10 (Supporting Information, Figure S1a). This hypothesis
was confirmed by the preparation of a standard of 10, whose
retention time (Figure 2b), UV/Vis spectrum, and MS/MS
fragmentation pattern (Supporting Information, Figure S1b)
were identical to those recorded for the metabolite accumu-
lated in the CGCA004 culture supernatant.
We next confirmed that 10 is an intermediate in con-
gocidine biosynthesis and investigated the pathway for its
assembly. Among the 22 genes constituting the cgc gene
cluster, 13 could not be assigned a function in regulation of
gene expression, congocidine resistance, or assembly of the
putative precursors, guanidinoacetate, 3-aminopropionami-
dine, and 10.[14] These 13 genes encode proteins with similarity
to enzymes of known function (with the exception of cgc7)
and are likely to be involved in the biosynthesis of congoci-
dine precursors. Six of the genes encode proteins that are
similar to enzymes involved in carbohydrate biosynthesis or
transfer (Cgc8, putative NDP-hexose dehydrogenase; Cgc9,
putative NDP-hexose epimerase/dehydrogenase; Cgc10,
putative glycosyltransferase; Cgc11, putative hexose nucleo-
tidyltransferase; Cgc12, putative DegT/DnrJ/EryC1/StrS
aminotransferase; Cgc13, putative glycoside hydrolase). The
presence of these genes in the cgc cluster is intriguing because
congocidine does not contain a carbohydrate moiety. Yet, we
have previously shown that at least one of these genes, cgc10,
is involved in congocidine biosynthesis.[14] To determine
whether the other five genes are also involved in congocidine
biosynthesis, we constructed in-frame deletions of each within
a bacterial artificial chromosome (BAC; pCGC002) contain-
ing the entire cgc cluster. The resulting constructs were
heterologously expressed in S. lividans TK23. None of the
S. lividans strains with the genetically modified BACs pro-
duced congocidine, which would elute at 15.1 min, (Fig-
ure 2dand Supporting Information, Figure S3, left panels)
except the CGCL047 strain (Dcgc13), which produced
residual congocidine (15% of the “wild type” strain
CGCL006). This indicates that the genes encoding putative
carbohydrate biosynthesis/transfer enzymes are involved in
congocidine biosynthesis and, therefore, are likely to partic-
ipate in biosynthesis of a congocidine precursor.
We examined which of the carbon atoms of congocidine
are carbohydrate-derived using incorporation experiments
with labeled precursors. Cultures of S. ambofaciens SPM110
were pulse-fed with uniformly 13C-labeled glycerol ([U-
13C]glycerol; 4 ꢀ 100 mmol) and congocidine was purified by
cation exchange chromatography. The DEPTQ-135 NMR
spectrum (Supporting Information, Figure S4) of the isolated
congocidine showed coupled doublets flanking the natural
abundance signals for C-2/C-3, C-6/C-9, and C-12/C-15,
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
pyrrole-2-carboxamide units from fructose-6-phosphate by
way of a 2 intermediate (Figure 1).[17] To verify this hypoth-
esis, we next attempted to feed cultures of S. ambofaciens with
[15N,13C-1] N-acetylglucosamine. Unfortunately, the addition
of N-acetylglucosamine strongly suppresses congocidine pro-
duction (presumably because of DasR-mediated regula-
tion).[18] Thus, we were not able to obtain enough congocidine
to determine whether N-acetylglucosamine was incorpo-
rated.[19]
Taken together these data suggest that the putative
carbohydrate-processing enzymes encoded within the cgc
cluster are responsible for the assembly of 10 from 2. To
further investigate this hypothesis and examine whether 10 is
an intermediate in congocidine biosynthesis, we fed chemi-
cally synthesized 10 to cultures of the cgc11, cgc8, cgc9, cgc12,
and cgc10 deletion mutants. LC-MS analyses showed that
production of congocidine was restored in all of the mutants
(Supporting Information, Figure S3, central and right panels),
indicating that 10 is a congocidine precursor and that cgc11,
cgc8, cgc9, cgc12, and cgc10 are all involved in its biosynthesis.
When the cgc13 deletion mutant was fed with 10, congocidine
production was increased by a factor of 2.5 (Supporting
Information, Figure S3h). This result suggests that the
reduced congocidine production in the cgc13 deletion
mutant is due to impaired synthesis of 10.
To determine whether other cgc genes are involved in 10
biosynthesis, the six cgc genes with no assigned function (cgc3,
cgc4, cgc5, cgc6, cgc7, and cgc17) were individually deleted in
frame within the pCGC002 BAC and the resulting BACs
heterologously expressed in S. lividans TK23. The cgc4 dele-
tion mutant produced residual congocidine, whereas congo-
cidine production was abolished in the other deletion mutants
(Supporting Information, Figure S3 f,g, left panels; Figure S6,
left panels). The cgc4, cgc5, cgc6, and cgc7 deletion mutants
accumulated 10 (confirmed by MS analyses, Supporting
Information, Figure S6, right panels), indicating that these
genes are not involved in 10 biosynthesis. The cgc3 and cgc17
deletion mutants, however, did not produce 10. The involve-
ment of Cgc3 (a putative aldehyde dehydrogenase)[14] and
Cgc17 (a putative alcohol dehydrogenase)[14] in 10 biosynthe-
sis was confirmed by feeding 10 to the mutant strains, which
restored congocidine production (Supporting Information,
Figure S3 f,g, central and right panels).
The cgc3 deletion mutant accumulated a new metabolite
in the culture medium that was not found in the other mutants
(Supporting Information, Figure S3g). MS and NMR spec-
troscopic analyses of the purified metabolite (Supporting
Information, Figure S7) showed that it was 4-acetamidopyr-
role-2-carboxaldehyde (9), which is likely to be the substrate
of Cgc3 (Figure 3).
The cgc17 gene encodes a putative alcohol dehydrogenase
possibly involved in the oxidation of 4-acetamidopyrrole-2-
methanol (8) to yield 9 (Figure 1). To verify this hypothesis,
cultures of the cgc11, cgc8, cgc9, cgc12 cgc10, cgc17, and cgc3
deletion mutants were fed with chemically synthesized 8
(Supporting Information, methods and Figure S8). Produc-
tion of congocidine was restored for the first five mutant
strains (Supporting Information, Figure S9). In contrast,
feeding of 8 to the cgc17 and cgc3 deletion mutants did not
Figure 2. LC analysis of culture supernatents of mutants with deletions
in congocidine biosynthetic genes. Chromatograms from a) CGCA004
(disruption of cgc18); b) synthetic 10; c) CGCL048 (Dcgc14);
d) CGCL054 (Dcgc9), and e) CGCL045 (Dcgc3).
showing that a two-carbon fragment of glycerol was incorpo-
rated intact into C-2/C-3, C-6/C-9, and C-12/C-15 (see
Figure 1 for carbon atom numbering). The guanidinoacetate
precursor of congocidine has been suggested to originate from
glycine by way of transguanylation.[16] This is consistent with
the incorporation pattern seen for C-2/C-3 (Supporting
Information, Figure S5). More significant, however, is the
incorporation of two carbons from glycerol into the same
position (C-2 and the carboxamide group) in both 4-amino-
pyrrole-2-carboxamide units of congocidine. This incorpora-
tion pattern is consistent with derivation of the 4-amino-
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
The similarity of Cgc13 with glycoside hydrolases suggests
that Cgc13 could be involved in providing 2.
Our proposed pathway results in the biosynthesis of 10.
Yet, before condensation with the other congocidine precur-
sors can occur, this molecule must be de-acetylated to yield 4-
aminopyrrole-2-carboxylate. Among the proteins encoded by
the cgc gene cluster, we identified Cgc14, a predicted
amidohydrolase, as the enzyme likely to catalyze the de-
acetylation. Therefore, we constructed a cgc14 in-frame
deletion mutant (CGCL048). This mutant did not produce
congocidine but accumulated 10 (Figure 2c), consistent with
the proposed role for Cgc14. However a mutant strain
impaired in congocidine assembly (CGCA004) also accumu-
lates 10 and not 4-aminopyrrole-2-carboxylate. These obser-
vations suggest that 10, rather than 4-aminopyrrole-2-carbox-
ylate, may be the substrate of the Cgc18 adenylation domain
and that de-acetylation occurs after 10 has been loaded onto
the Cgc18 and Cgc19 PCPs (Figure 1). Consistent with this
hypothesis, feeding of the cgc14 deletion mutant with
commercially available 4-aminopyrrole-2-carboxylate did
not restore congocidine production (Supporting Information,
Figure S11). N-acetylation is a well characterized mechanism
of arylamine detoxification and maintaining the 4-amino-
pyrrole-2-carboxylate in its N-acetylated form when free in
solution would avoid the cytotoxic effects associated with
arylamines.[24]
Figure 3. Summary of COSY, HMBC, and NOESY correlations recorded
for 9 purified from the culture supernatant of the cgc3 mutant.
restore congocidine production. These results are consistent
with 8 being the substrate of Cgc17. Accumulation of 8 in the
cgc17 deletion mutant could not be detected using standard
LC-MS analysis because this compound is unstable and
poorly retained on reverse-phase HPLC columns.
Based on our data, we suggest the following pathway for
10 biosynthesis from 2 (Figure 1). Cgc11 catalyzes nucleotide
triphosphate (NTP)-dependent conversion of 2 into the
nucleotide diphosphate (NDP) derivative 3, which undergoes
Cgc8-catalyzed oxidation of the C-6 hydroxy group to
a carboxy group to yield NDP-2-acetamido-2-deoxygluco-
pyranuronic acid (4). Cgc9 catalyzes the decarboxylation of 4
to afford NDP-threo-2-acetamido-2-deoxy-pentopyran-4-
ulose (5). Reductive amination of the C-4 keto group of 5 is
catalyzed by the putative pyridoxal-5-phosphate-dependent
aminotransferase Cgc12. The next step involves intramolec-
ular nucleophilic attack of the C-4 amino group on C-1 to
displace the NDP and afford 4-acetamido-2,3-dihydropyr-
role-2-methanol (7). This conversion of a carbohydrate into
a pyrrole molecule is proposed to be catalyzed by Cgc10, an
enzyme that resembles a glycosyltransferase. Loss of water
from 7 to form 8 could then occur spontaneously (the
formation of the resulting pyrrole aromatic ring is expected
to be thermodynamically favorable). Finally, oxidation of the
hydroxy group of 8 to the aldehyde 9 and subsequently the
carboxylic acid 10 is catalyzed by Cgc17 and Cgc3, respec-
tively.
The initial steps of our proposed pathway are catalyzed by
enzymes (Cgc11, Cgc8, Cgc9, and Cgc12) that are homolo-
gous to enzymes catalyzing similar reactions in the biosyn-
thesis of aminodeoxypentose sugars found in enediyne
natural products such as maduropeptin, calicheamicin, and
AT2433 (Supporting Information, Figure S10).[20,21] Carbohy-
drates incorporated in these natural products can be activated
using UTP or TTP. By analogy, it appears likely that UTP or
TTP is the NTP used to activate 2. UDP-N-acetylglucos-
amine, derived from 2, is an intermediate in the biosynthesis
of bacterial peptidoglycan.[22,23] As the cgc11 deletion mutant
does not produce any congocidine, it is unlikely that UDP-N-
acetylglucosamine is an intermediate in congocidine biosyn-
thesis. Indeed, activation of 2 by a nucleotide triphosphate
other than UTP (most likely TTP) could constitute the first
committed step of 10 biosynthesis.
In conclusion, we have shown that 10 is the true
intermediate in the biosynthesis of congocidine. Interestingly,
the biosynthesis of 10 described herein is unprecedented in
that it involves enzymes and precursors from carbohydrate
metabolism. In particular, it bears no resemblance to the
biosynthesis of other pyrrole-2-carboxylates incorporated
into secondary metabolites such as clorobiocin,[6] coumermy-
cin,[4,6]
pyoluteorin,[3,25]
and
undecylprodiginine.[3,7,8]
Although this pathway is likely to be involved in the
biosynthesis of other members of the pyrrolamide family of
metabolites, it is not currently known whether it is also
utilized in the biosynthesis of other pyrrole-containing natural
products. However, database searches identified a putative 10
biosynthetic gene cluster in the genome of Micromonospora
carbonacea var. Africana ATCC 39149 (MCAG_05565 and
MCAG_05568 to MCAG_05575). Moreover, it is interesting
to note that this pathway, with the reductive amination
reaction replaced by a ketoreduction, would be perfectly
adapted to the biosynthesis of the furan precursors of
proximicins, analogues of congocidine recently isolated from
marine actinomycetes.[26]
Received: February 21, 2012
Revised: April 11, 2012
Published online: && &&, &&&&
Keywords: biosynthesis · congocidine · natural products ·
.
pyrrolamide · pyrrole
Our data show that Cgc13 is not essential for congocidine
biosynthesis. Nonetheless the reduction of congocidine pro-
duction in the cgc13 deletion mutant, and the partial
restoration of the production when the mutant is fed with
10, indicates that Cgc13 may participate in the supply of 10.
[1] C. T. Walsh, S. Garneau-Tsodikova, A. R. Howard-Jones, Nat.
Prod. Rep. 2006, 23, 517 – 531.
[2] S. Siebenberg, N. Burkard, A. Knuplesch, B. Gust, S. Grond, L.
Heide, ChemBioChem 2011, 12, 2677 – 2685.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
[3] M. G. Thomas, M. D. Burkart, C. T. Walsh, Chem. Biol. 2002, 9,
171 – 184.
[4] H. Xu, Z.-X. Wang, J. Schmidt, L. Heide, S.-M. Li, Mol. Genet.
Genomics 2002, 268, 387 – 396.
[5] S. Maharjan, N. Aryal, S. Bhattarai, D. Koju, J. Lamichhane, J. K.
Sohng, Appl. Microbiol. Biotechnol. 2012, 93, 687 – 696.
[6] S. Garneau, P. C. Dorrestein, N. L. Kelleher, C. T. Walsh,
Biochemistry 2005, 44, 2770 – 2780.
[7] S. Mo, P. K. Sydor, C. Corre, M. M. Alhamadsheh, A. E. Stanley,
S. W. Haynes, L. Song, K. A. Reynolds, G. L. Challis, Chem. Biol.
2008, 15, 137 – 148.
ration of two molecules of [U-13C]glycerol. However, fructose-
bisphosphate aldolase does not appear to be expressed/active in
Streptomyces species cultivated in rich growth media (see C.
Corre, G. L. Challis, ChemBioChem 2005, 6, 2166 – 2170 and
G. L. Challis, K. F. Chater, Chem. Commun. 2001, 935 – 936).
Thus, only incorporation of a single molecule of [U-13C]glycerol
into fructose-6-phosphate by way of the pentose phosphate
pathway is observed.
´
[18] M. A. Swia˛tek, E. Tenconi, S. Rigali, G. P. van Wezel, J.
Bacteriol. 2012, 194, 1136 – 1144.
[19] We tried to feed [U-13C]glucose to a mutant (CGCL022) that
accumulates 10. Some of the 10 was purified and analyzed by
HRMS (see Figure S12). As expected, 13C-5 10 was produced.
However, in addition to the 13C-5 labeled compound and to the
unlabeled compound, 13C-1, 13C-2, 13C-3, 13C-4 and 13C-7 labeled
molecules were also observed. This nonspecific incorporation
pattern can be expected with labeled glucose feeding experi-
ments, as many primary metabolites that could be incorporated
into the 2 precursor of 10 are derived from glucose.
[20] Q. Gao, C. Zhang, S. Blanchard, J. S. Thorson, Chem. Biol. 2006,
13, 733 – 743.
[8] A. E. Stanley, L. J. Walton, M. Kourdi Zerikly, C. Corre, G. L.
Challis, Chem. Commun. 2006, 3981 – 3983.
[9] T. Nishizawa, S. Grꢁschow, D.-H. E. Jayamaha, C. Nishizawa-
Harada, D. H. Sherman, J. Am. Chem. Soc. 2006, 128, 724 – 725.
[10] S. Kirner, P. E. Hammer, D. S. Hill, A. Altmann, I. Fischer, L. J.
Weislo, M. Lanahan, K. H. van Pꢂe, J. M. Ligon, J. Bacteriol.
1998, 180, 1939 – 1943.
[11] S. Neidle, Nat. Prod. Rep. 2001, 18, 291 – 309.
[12] A. Asai, Y. Sakai, H. Ogawa, Y. Yamashita, S. Kakita, K. Ochiai,
T. Ashizawa, A. Mihara, T. Mizukami, H. Nakano, J. Antibiot.
2000, 53, 66 – 69.
[13] P. B. Dervan, Bioorg. Med. Chem. 2001, 9, 2215 – 2235.
[14] M. Juguet, S. Lautru, F.-X. Francou, S. Nezbedovꢃ, P. Leblond,
M. Gondry, J.-L. Pernodet, Chem. Biol. 2009, 16, 421 – 431.
[15] S. Siebenberg, P. M. Bapat, A. E. Lantz, B. Gust, L. Heide, J.
Biosci. Bioeng. 2010, 109, 230 – 234.
[21] S. G. van Lanen, T.-J. Oh, W. Liu, E. Wendt-Pienkowski, B. Shen,
J. Am. Chem. Soc. 2007, 129, 13082 – 13094.
[22] D. Mengin-Lecreulx, J. van Heijenoort, J. Bacteriol. 1994, 176,
5788 – 5795.
[23] A. M. Gehring, W. J. Lees, D. J. Mindiola, C. T. Walsh, E. D.
Brown, Biochemistry 1996, 35, 579 – 585.
[16] M. E. Wildfeuer, The biosynthesis of netropsin, PhD thesis,
University of Delaware (USA), 1963.
[24] M. Payton, A. Mushtaq, T. W. Yu, L. J. Wu, J. Sinclair, E. Sim,
Microbiology 2001, 147, 1137 – 1147.
[17] Fructose-6-phosphate can also be formed from fructose-bisphos-
phate aldolase-mediated condensation of glyceraldehyde-3-
phosphate and dihydroxyacetone phosphate with subsequent
dephosphorylation. In this case, we would expect all six carbon
atoms of fructose-6-phosphate to be labeled through incorpo-
[25] B. Nowak-Thompson, N. Chaney, J. S. Wing, S. J. Gould, J. E.
Loper, J. Bacteriol. 1999, 181, 2166 – 2174.
[26] K. Schneider, et al., Angew. Chem. 2008, 120, 3302 – 3305;
Angew. Chem. Int. Ed. 2008, 47, 3258 – 3261.
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Biosynthetic Pathway
S. Lautru,* L. Song, L. Demange,
T. Lombꢁs, H. Galons, G. L. Challis,
J.-L. Pernodet
&&&& — &&&&
Feeding (Streptomyces) frenzy: Natural
products belonging to the pyrrolamide
family are defined by their pyrrole-2-
carboxamide moiety. 4-acetamidopyrrole-
2-carboxylate is identified as the key
pyrrolamide congocidine precursor (see
scheme) through feeding studies using
Streptomyces ambofaciens. The biosyn-
thetic pathway of congocidine starts with
the carbohydrate N-acetylglucosamine
and involves carbohydrate-processing
enzymes.
A Sweet Origin for the Key Congocidine
Precursor 4-Acetamidopyrrole-2-
carboxylate
6
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 6
These are not the final page numbers!