Auxotrophic-precursor directed biosynthesis of nonribosomal lipopeptides with modified tryptophan residues

Article abstract of DOI:10.1039/b718766c

Feeding 5-hydroxy and 5-fluorotryptophan to a Streptomyces coelicolor Trp-auxotrophic strain WH101 results in the production of a number of new calcium-dependent antibiotics (CDAs) possessing modified Trp residues. It is anticipated that this method could be used to modulate the biological properties of Trp-containing nonribosomal peptide natural products, or to generate analogues with useful fluorescent properties for studying biological mechanisms of action.

Full text of DOI:10.1039/b718766c

View Article Online / Journal Homepage / Table of Contents for this issue
www.rsc.org/obc
Volume 6 | Number 6 | 21 March 2008 | Pages 941–1128
ISSN 1477-0520
COMMUNICATION
Jason Micklefield et al.
2WT\XRP[ꢀBRXT]RT
Auxotrophic-precursor directed
biosynthesis of nonribosomal
lipopeptides with modified
tryptophan residues
In this issue...
1477-0520(2008)6:6;1-B

COMMUNICATION
www.rsc.org/obc | Organic & Biomolecular Chemistry
Auxotrophic-precursor directed biosynthesis of nonribosomal lipopeptides
with modified tryptophan residues†
Bagher Amir-Heidari,‡ Jenny Thirlway and Jason Micklefield*
Received 4th December 2007, Accepted 18th January 2008
First published as an Advance Article on the web 11th February 2008
DOI: 10.1039/b718766c
Feeding 5-hydroxy and 5-fluorotryptophan to a Streptomyces
coelicolor Trp-auxotrophic strain WH101 results in the pro-
duction of a number of new calcium-dependent antibiotics
(CDAs) possessing modified Trp residues. It is anticipated
that this method could be used to modulate the biological
properties of Trp-containing nonribosomal peptide natural
products, or to generate analogues with useful fluorescent
properties for studying biological mechanisms of action.
Despite this, there are very few examples where amino acid aux-
otrophs have been exploited in the precursor directed biosynthesis
of natural products.5
Previously, we used the Trp-auxotrophic WH101 strain of
Streptomyces coelicolor6 to determine the stereochemical course
of Trp dehydrogenation during the biosynthesis of Z-2ꢀ,3ꢀ-
dehydrotryptophan (Z-DTrp) found at the C-terminal position
11 of the calcium-dependent antibiotics (CDA) (Fig. 2).7 CDA
is a nonribosomal lipopeptide, which in addition to Z-DTrp,
contains a number of other nonproteinogenic and D-amino acids,
including D-Trp at position 3, as well as an N-terminal trans-
2,3-epoxyhexanoyl fatty acid side chain.8 CDA shares a similar
structure, and probably a related calcium-dependent antimicrobial
mechanism of action, as the commercial lipopeptide antibiotic
daptomycin.9 As a result of this, there has been considerable
interest in developing new methods for the engineered, and
precursor directed, biosynthesis of new lipopeptide variants.10
Following on from this, it was envisaged that the S. coelicolor
WH101 strain, which can incorporate high levels (ca. 60%) of
deuterated Trp into CDA when grown on minimal medium,7 might
be used in the directed biosynthesis of CDAs possessing modified
Trp-derived residues.
For many years auxotrophic strains of bacteria and fungi,
which are unable to produce a specific proteinogenic amino
acid, have been exploited to produce proteins containing amino
acid analogues.1 For example tryptophan analogs including 7-
azatryptophan (7AW) 1 and 5-hydroxytryptophan (5HW) 2
(Fig. 1) have been incorporated into proteins in Trp-auxotrophic
bacteria, grown on minimal media with depleted levels of L-Trp.2
Such Trp analogues (1 and 2) have distinct fluorescence properties,
which can be used to probe the structure, function and dynamics
of selected proteins.2 Indeed shorter peptides (domains) have been
similarly produced, possessing modified Trp residues, and then
subsequently assembled into larger multidomain proteins using
expressed protein ligation.2c The biological properties of small ri-
bosomally encoded peptides, including the disulfide bridged cyclic
peptide compstatin, can also be improved by incorporation of
Trp analogues using an E. coli Trp-auxotroph.3 Finally, precursor
directed biosynthesis, involving feeding amino acid analogues
to wild-type bacterial and fungal strains, has also resulted in
the production of secondary metabolite analogues incorporating
modified amino acid residues.4 For example fluorinated analogs
of the nonribosomal peptide actinomycin D were produced by
feeding 5-fluorotryptophan (5FW) 3 to Streptomyces parvullus.4b
Fig. 2 Calcium-dependent antibiotics (CDAs) produced by the wild type
S. coelicolor: CDA1, R9 = OPO3H2 and R10 = H; CDA2, R9 = OPO3H2
and R10 = CH3; CDA3, R9 = OH, R10 = H; CDA4, R9 = OH, R10 = CH3.
The a-series contain Z-DTrp (R11 = p-bond) and the b-series contain L-Trp
(R11 = H,H) at position 11.
Fig. 1 Structures of 7-azatryptophan 1, 5-hydroxytryptophan 2 and
5-fluorotryptophan 3 analogues used in this study.
School of Chemistry and Manchester Interdisciplinary Biocentre, The
University of Manchester, 131 Princess Street, Manchester, UK M1 7DN.
E-mail: Jason.micklefield@manchester.ac.uk; Fax: +44 (0)161 306 8918;
Tel: +44 (0)161 306 4509
† Electronic supplementary information (ESI) available: Experimental
details, additional results and figures. See DOI: 10.1039/b718766c
Accordingly, the WH101 strain was grown in the minimal
medium SV2a, estimated to contain low levels of L-Trp (ca.
6.0 lg mL−1),7 which was supplemented with D/L-7-azatryptophan
(7AW) 1 (75.0 lg mL−1). Following fermentation, the culture
supernatants were analysed by ESI-LC-MS.7 Detailed analysis
of the LC-MS failed to show the presence of any CDAs with
‡ Current address: Pharmaceutical Technology Research Center, Kerman
Medical University, Kerman, Iran.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 975–978 | 975
©

molecular weights consistent with 7-azatryptophan containing
variants. Instead, wild-type CDA4b was shown to be the major
product by comparison with an authentic sample. In the absence
of 7AW 1, the WH101 strain, under identical conditions, produces
predominantly the a-series CDA4a with a minor amount of
CDA3a, which both contain C-terminal Z-DTrp residues. Thus,
whilst 7AW 1 is not incorporated into CDA, it apparently inhibits
the production of a-series CDA in favour of CDA4b, which
possesses a C-terminal L-Trp, rather than a Z-DTrp residue. One
possible reason for this could be that 7AW inhibits the putative
Trp dehydrogenase/oxidase enzyme that is predicted to catalyse
the oxidation of the C-terminal Trp side chain.7 Indeed, Trp
analogues with modifications in the indole ring have been shown
to inhibit the L-tryptophan oxidase (LTO) from Chromabacterium
violaceum,11 which shares some mechanistic similarity to the
enzyme responsible for Z-DTrp in CDA.7
Following this, S. coelicolor WH101 was grown in SV2a,
supplemented with 5-hydroxytryptophan (5HW) 2 (37.5 lg mL−1),
and the culture supernatants were analysed by LC-MS (Fig. 3).
This revealed that CDA4a was the major product (Rt = 6.98).
However, in addition, three new products were evident that are
not produced by the WH101 strain in the absence of 5HW 2.
Furthermore, these new products do not correspond with the
retention times or MS molecular ions that are observed for any
of the known CDAs produced by S. coelicolor. Indeed, two of
the products (Rt = 5.96 and 6.19 min) exhibit molecular ions
that are exactly 16 Da higher than CDA4a. This is consistent
with the single incorporation of 5HW 2 at either position 3 or 11
of CDA4a. The low quantities of these products and difficulties
associated with their separation prevented isolation of sufficient
quantities of peptides for the assignment of the regiochemistry
of the 5HW incorporation. However, a fourth additional minor
product was also evident with a shorter retention time (Rt =
5.19 min) and molecular ions that are 32 Da higher than CDA4a.
The increased polarity and MS of this product is thus consistent
with a new CDA analogue that has incorporated two 5HW
molecules at both position 3 and 11 of CDA4a. Furthermore,
the molecular ions of the three new products in this case are
all consistent with 2ꢀ,3ꢀ-dehydrogenation of the Trp or 5HW
residues, at position 11, having taken place. This suggests that the
putative CDA Trp dehydrogenase/oxidase biosynthetic enzyme
is not inhibited by 5HW and indeed can accept a 5HW residue
as a substrate, presumably oxidising this to the 2ꢀ,3ꢀ-dehydro-5-
hydroxytryptophan residue.
In addition to this, 5-fluorotryptophan (5FW) 3 (37.5 lg mL−1)
was fed to the WH101 auxotroph in a similar fashion. In this
case, there is evidence of very low-level production of CDA4b and
CDA3b, however, none of the a-series wild-type lipopeptides is
observed above the threshold level of detection. Instead, the major
products from these feedings are a group of four closely eluting
compounds, which are clearly not evident in fermentations in the
absence of 5FW (Fig. 4). In addition, these products have longer
retention times and do not possess molecular ions in the MS that
correspond with any known CDAs produced by WH101 or any
other known S. coelicolor strain. The products with retention times
of 7.41 and 7.52 min, exhibit molecular ions that are consistent
with the single incorporation of 5FW into CDA3b and CDA4b,
respectively. In addition, the two remaining products with longer
retention times of 7.62 and 7.71 min, exhibit molecular ions in the
Fig. 3 (A) HPLC trace and (B) MS of the products resulting from feeding
5-hydroxytryptophan (5HW) 2 to S. coelicolor WH101. Rt = 7.01 min,
CDA4a: m/z 1495.6 ([M + H]+ C67H79N14O26 requires 1495.5); 1517.6 ([M
+ Na]+ C67H78N14O26Na requires 1517.5); 1533.6 ([M + K]+ C67H78N14O26K
requires 1533.5). Rt = 6.19 and 5.96 min, single incorporation (5HW)
CDA4a regioisomers:m/z 1511.6 ([M + H]+ C67H79N14O27 requires 1511.5);
1533.6 ([M + Na]+ C67H78N14O27Na requires 1533.5); 1549.6 ([M + K]+
C67H78N14O27K requires 1549.5). Rt = 5.17 min, double incorporation
product, (5HW)2CDA4a: m/z 1527.6 ([M + H]+ C67H79N14O28 requires
1527.5); 1549.6 ([M + Na]+ C67H78N14O28Na requires 1549.5); 1565.6 ([M
+ K]+ C67H78N14O28K requires 1549.5).
MS that are consistent with the double incorporation of 5FW into
both positions 3 and 11 of CDA3b and CDA4b, respectively. Given
that the incorporation of 5FW 3 into CDA is apparently higher
than 5HW 2, this would appear to suggest that the 5FW is a better
substrate for the NRPS enzymes that activate and condense the
Trp precursors. On the other hand, the absence of any wild-type
a-series CDA or any probable a-series CDA variants possessing
5FW residues suggests that 5FW residues are not substrates for
the Trp-dehydrogenase/oxidase and probably 5FW also inhibits
the dehydrogenation of the natural Trp residues in the wild-type
CDAs.
Conclusion
In summary, using the Trp-auxotrophic S. coelicolor WH101
strain, it is possible to effect the precursor directed biosynthesis
976 | Org. Biomol. Chem., 2008, 6, 975–978
This journal is
The Royal Society of Chemistry 2008
©

was approved for use in 2003 and is now widely used in the
clinic, possesses an L-Trp residue as well as a Trp derived L-
kynurenine (Kyn) residue. Indeed, the Streptomyces roseosporus
industrial daptomycin production strain gives ca. 100 fold higher
titres of lipopeptides compared to S. coelicolor. It is thus likely
that a S. roseosporus Trp-auxotroph could be used to produce
significant quantities of daptomycins possessing modified Trp and
Kyn residues. In addition to potentially improving the biological
properties of the lipopeptide, those analogs possessing 5HW will
exhibit fluorescence excitation at wavelengths longer than the
absorbance of Trp and thus their fluorescence can be selectively
excited.2a The distinct fluorescent properties of such lipopeptides
could thus be used to further probe the structure and dynamics
of lipopeptides. In addition, fluorescence could be used to study
the interactions of lipopeptides with specific molecular targets
within bacterial cell membranes. This could assist in efforts to
further elucidate the antimicrobial mechanisms of action of these
lipopeptides, which remains a subject of considerable debate and
uncertainty.9,12
Acknowledgements
This work was supported by the BBSRC through research grants
(36/B12126 and BB/C503662). B. Amir-Heidari is the recipient
of a PhD scholarship from the Iranian Ministry of Health
and Medical Education. We also thank Prof. D. A. Hodgson
(University of Warwick) for S. coelicolor WH101.
Notes and references
Fig. 4 (A) HPLC trace and (B) MS of the products resulting from
feeding 5-fluorotryptophan (5FW) 3 to S. coelicolor WH101. Rt
7.71 min, double incorporation product (5FW)2CDA4b: m/z 1533.5
([M
H]+ C67H79F2N14O26 requires 1533.5), 1555.5 ([M Na]+
C67H78F2N14Na1O26 requires 1555.5), 1571.5 ([M + K]+ C67H78F2K1N14O26
requires 1571.5); Rt 7.62 min, double incorporation product
=
1 (a) L. Wanga and P. G. Schultz, Chem. Commun., 2002, 1–11; (b) N.
Budisa, C. Minks, J. Medrano, J. Lutz, R. Huber and L. Moroder, Proc.
Natl. Acad. Sci. U. S. A., 1998, 95, 455–1459; (c) E. D. Fenstert and
H. S. Anker, Biochemistry, 1969, 8, 269–1274.
2 (a) S. M. Twine and A. G. Szabo, Methods Enzymol., 2003, 360, 104–
127; (b) S. Barlati and O. Ciferri, J. Bacteriol., 1970, 101, 166–172; (c) V.
Muralidharan, J. Cho, M. Trester-Zedlitz, L. Kowalik, B. T. Chait, D. P.
Raleigh and T. W. Muir, J. Am. Chem. Soc., 2004, 126, 14004–14012.
3 M. Katragadda and J. D. Lambris, Protein Expression Purif., 2006, 47,
289–295.
4 (a) R. Thiericke and J. Rohr, Nat. Prod. Rep., 1993, 10, 265–289; (b) A.
Kawashima, H. Seto, I. Kato, A. Yasuda, K. Uchida and N. Take,
J. Antibiot., 1985, 38, 1625–1628; (c) F. V. Ritacco, E. I. Graziani,
M. Y. Summers, T. M. Zabriskie, K. Yu, V. S. Bernan, G. T. Carter
and M. Greenstein, Appl. Environ. Microbiol., 2005, 71, 1971–1976;
(d) C. Nilanonta, M. Isaka, P. Kittakoop, S. Trakulnaleamsai, M.
Tantichareon and Y. Thebtaranonth, Tetrahedron, 2002, 58, 3355–
3360.
+
+
=
(5FW)2CDA3b: m/z 1519.5 ([M + H]+ C66H77F2N14O26 requires 1519.5),
1541.5 ([M + Na]+ C66H76F2N14Na1O26 requires 1541.5), 1557.5 ([M + K]+
C66H76F1K1N14O26 requires 1557.5); Rt = 7.51 min, single incorporation
product (5FW)1CDA4b: m/z 1515.5 ([M + H]+ C67H80F1N14O26 requires
1515.5), 1537.5 ([M + Na]+ C67H79F1N14Na1O26 requires 1537.5), 1553.6
([M + K]+ C67H79F1K1N14O26 requires 1537.5); Rt = 7.41 min, single incor-
poration product (5FW)1CDA3b: m/z 1501.6 ([M + H]+ C66H78F1N14O26
requires 1501.5); 1523.6 ([M + Na]+ C66H77F1N14Na1O26 requires 1523.5),
1539.4 ([M + K]+ C66H78F1K1N14O26, requires 1539.5).
5 A Cephalosporium acrenroniurn mutant blocked in its lysine path-
way before a-aminoadipate, has been shown to incorporate L-S-
carboxymethylcysteine analogues of a-aminoadipate leading to modi-
fied penicillins: H. Troonen, P. Roelants and B. Boon, J. Antibiot., 1976,
29, 1258–1267.
of CDA analogues possessing modified Trp residues by feeding
5-hydroxy and 5-fluorotryptophan analogues (2 and 3). Interest-
ingly, the 7-azatryptophan analogues cannot be incorporated into
CDA in this way. The typical low yields for production of CDAs
and difficulties associated with separating the products, including
regioisomers, complicated by the fact that CDA possesses two Trp
derived residues, prevented the isolation of sufficient quantities
of products for biological testing. Nevertheless, it is possible that
this approach might be used to modify Trp residues found in
other natural products, particularly nonribosomal peptides, which
could lead to new products with improved biological activities.
For example, the commercial lipopeptide daptomycin, which
6 F. Barona-Gomez and D. A. Hodgson, EMBO Rep., 2003, 4, 296–
300.
7 (a) B. Amir-Heidari, J. Thirlway and J. Micklefield, Org. Lett., 2007,
9, 1513–1516; (b) B. Amir-Heidari and J. Micklefield, J. Org. Chem.,
2007, 72, 8950–8953.
8 (a) C. Milne, A. Powell, J. Jim, M. Al Nakeeb, C. P. Smith and J.
Micklefield, J. Am. Chem. Soc., 2006, 128, 11250–11259; (b) C. Mahlert,
F. Kopp, J. Thirlway, J. Micklefield and M. A. Marahiel, J. Am. Chem.
Soc., 2007, 129, 12011–12018.
9 (a) J. Micklefield, Chem. Biol., 2004, 11, 887–888; (b) L.-J. Ball, C. M.
Goult, J. A. Donarski, J. Micklefield and V. Ramesh, Org. Biomol.
Chem., 2004, 2, 1872–1878.
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 975–978 | 977
©

10 (a) G. C. Uguru, C. Milne, M. Borg, F. Flett, C. P. Smith and J.
Micklefield, J. Am. Chem. Soc., 2004, 126, 5032–5033; (b) V. Miao, M. F.
Coeffet-LeGal, K. Nguyen, P. Brian, J. Penn, A. Whiting, J. Steele, D.
Kau, S. Martin, R. Ford, T. Gibson, M. Bouchard, S. K. Wrigley and
R. H. Baltz, Chem. Biol., 2006, 13, 269–276; (c) K. T. Nguyen, D. Ritz,
J.-Q. Gu, D. Alexander, M. Chu, V. Miao, P. Brian and R. H. Baltz,
Proc. Natl. Acad. Sci. U. S. A., 2006, 103, 17462–17467.
11 R. Genet, P.-H. Be´netti, A. Hammadi and A. Me´nez, J. Biol. Chem.,
1995, 270, 23540–23545.
12 (a) J. A. Silverman, N. G. Perlmutter and H. M. Shapiro, Agents
Chemother., 2003, 47, 2538–2544; (b) S. K. Straus and R. E. W.
Hancock, Biochim. Biophys. Acta, 2006, 1758, 1215–1223; (c) J. N.
Steenbergen, J. Alder, G. M. Thorne and F. P. Tally, J. Antimicrob.
Chemother., 2005, 55, 283–288.
978 | Org. Biomol. Chem., 2008, 6, 975–978
This journal is
The Royal Society of Chemistry 2008
©