Introduction of a non-natural amino acid into a nonribosomal peptide antibiotic by modification of adenylation domain specificity

Article abstract of DOI:10.1002/anie.201202043

Calcium-dependent antibiotics (CDA) are cyclic lipopeptides assembled by nonribosomal peptide synthetase (NRPS) enzymes. Active site modification of the 3-methyl glutamate activating adenylation (A) domain of the CDA NRPS enables the incorporation of synthetic 3-methyl glutamine into CDA. This provides the first example of how A-domains can be engineered to introduce synthetic non-natural amino acids into nonribosomal peptides. Copyright

Full text of DOI:10.1002/anie.201202043

Angewandte
Chemie
DOI: 10.1002/anie.201202043
Biosynthetic Engineering
Introduction of a Non-Natural Amino Acid into a Nonribosomal
Peptide Antibiotic by Modification of Adenylation Domain
Specificity**
Jenny Thirlway, Richard Lewis, Laura Nunns, Majid Al Nakeeb, Matthew Styles, Anna-
Winona Struck, Colin P. Smith, and Jason Micklefield*
Nonribosomal peptides are a group of structurally diverse
secondary metabolites, which include a number of important
therapeutic agents and agrochemicals.[1,2] The modular archi-
tecture of the nonribosomal peptide synthetases (NRPS),
responsible for the assembly of these complex peptides,
provides an attractive platform from which to engineer the
biosynthesis of “non-natural” nonribosomal peptides.[1] For
example, NRPS module and domain exchanges have been
used to produce cyclic lipopeptide variants of the clinically
important antibiotic daptomycin.[3] Active-site modification
of adenylation domains, which are responsible for activating
substrate amino acids, has also been explored as an alter-
native approach for engineering NRPS assembly lines.[4–6]
Using the archetypal structure of an A-domain, PheA from
gramidicin S synthetase,[7] and multiple sequence alignments
of known A-domains, it is possible to identify the key active-
site residues that are responsible for recognition of substrate
amino acids.[8] Whilst these structural insights have been very
effective in guiding mutagenesis approaches to alter the
substrate specificity of isolated A-domains in vitro,[9] there
have been relatively few examples of how this approach can
be used to engineer new nonribosomal peptide products
in vivo.[4–6] Moreover, only conservative changes involving the
exchange of proteinogenic amino acids of similar size and
polarity, within peptide structures, have been effected to date
using A-domain modifications in vivo. For instance, a directed
evolution approach involving saturation mutagenesis of
specific A-domain residues resulted in the production of
andrimid analogues where the natural Val residue had been
replaced by similar hydrophobic amino acids Leu, Ile, Ala,
and Phe.[6]
Previously we have been investigating the biosynthesis
and biosynthetic engineering of calcium-dependent antibiot-
ics (CDA),[4,10,11] members of the acidic lipopeptide family
which also includes daptomycin.[12] Here we show how it is
possible to change the specificity of an A-domain within the
CDA NRPS assembly line to incorporate a synthetic non-
natural amino acid into the decapeptide lactone core of CDA
(Figure 1). Our strategy focused on altering the specificity of
the module 10 A-domain of CdaPS3 so that it preferentially
incorporates (2S,3R)-3-methyl glutamine (mGln) and Gln
over the natural substrates (2S,3R)-3-methyl glutamic acid
(mGlu) and Glu. Multiple sequence alignments of the
predicted active site of the module 10 A-domain and related
Glu- and Gln-activating A-domains (Figure 1C) indicate that
Glu-activating A-domains tend to possess a basic residue, Lys
or His, at either position 239 or 278, which presumably
stabilizes the side-chain carboxy group of the substrate
through electrostatic interactions. On the other hand, the
glutamine-activating A-domains tend to differ by the pres-
ence of a Gln rather than a basic residue, at the same relative
positions 239 or 278. Furthermore, in vitro studies have shown
that by changing Lys239 to a Gln residue within the
glutamate-activating A-domain of the surfactin synthetase,
SrfA, the specificity of this isolated A-domain is then changed
from Glu to Gln.[5] In light of this, mutants of the CDA NRPS
module 10 A-domain containing two single amino acid
changes, Lys278Gln and Gln236Glu, as well as a double
mutant comprising both mutations were constructed. It was
envisaged that the complementary Gln236Glu mutation
might further aid the recognition of glutamine or mGln
substrates, with electrostatic interactions disfavoring interac-
tion of glutamate or mGlu substrates.
Accordingly, a 1.8 kb DNA fragment spanning the A-
domain of module 10 was generated and the desired muta-
tions incorporated by site-directed mutagenesis. The three
mutant DNA fragments were each transferred to pMAH,[13]
which operates in Streptomyces coelicolor as a suicide vector,
to generate three plasmids pKQ, pQE, and pKQQE. The
wild-type strain S. coelicolor MT1110 was chosen as a host as
it tends to produce non-phosphorylated CDAs which reduces
the complexity of product analysis. In addition, the previously
described mutant strain MT1110 DglmT[11] was used as
a second host strain, because it lacks the glmT gene
(SCO3215) that encodes a methyltransferase required for
mGlu biosynthesis. MT1110 DglmT is unable to produce the
[*] Dr. J. Thirlway,[+] L. Nunns, Dr. M. Al Nakeeb, M. Styles,
Dr. A.-W. Struck, Prof. J. Micklefield
School of Chemistry & Manchester Interdisciplinary Biocentre
The University of Manchester
131 Princess Street, Manchester M1 7DN (UK)
E-mail: j.micklefield@manchester.ac.uk
Dr. R. Lewis,[+] Prof. C. P. Smith
FHMS, University of Surrey
Guildford, Surrey, GU27XH (UK)
[+] These authors contributed equally to this work.
[**] This work was supported by the BBSRC (BB/C503662, BB/C507210,
and PhD studentships to L.N. and M.S.). We also thank Dr. Barrie
Wilkinson (Biotica) for his valuable advice and Prof. Gareth Morris’s
group (University of Manchester) for assistance with NMR experi-
ments.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201202043.
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
has little effect on CDA production compared with the
parental strains (Supporting Information, Figure S1). On the
other hand the Lys278Gln mutation did have a significant
effect. In the case of MT1110-KQ a new major product with
a retention time of 6.7 min was evident, with a minor amount
of CDA3a and CDA4a (retention time 7.0 min) that are both
produced as the major products in the parental strain,
MT1110 (Figure S2). The new product exhibited protonated,
sodiated and potassiated singly charged ions which are
consistent with a lipopeptide variant CDA3a-10Q (Fig-
ure 1A) possessing a glutamine residue at position 10. The
new product was subsequently purified by HPLC and
subjected to high-resolution mass spectrometry, confirming
the proposed molecular formula (Table 1).
Table 1: HRMS of the new cylic lipopeptides.
CDAs
Ion
Formula
Observed[a] Calculated
[M+H]+
C66H78N15O25
1480.5309
740.7686
1494.5513
747.7768
1480.5288
740.7681
1494.5444
747.7759
+
CDA3a-10Q
[M+2H]2+ C66H79N15O25
2+
CDA4a-10mQ [M+H]+
[M+2H]2+ C67H81N15O25
C67H80N15O25
+
2+
[a] Observed masses are all within 5 ppm of calculated mass values.
During isolation of the new product CDA3a-10Q,
a second product was observed which possesses a slightly
shorter retention time and exhibits m/z 749.77 and 1498.53,
which corresponds to the [M+2H]2+ and [M+H]+ ions of
a linear form of CDA3a-10Q (1, Scheme 1) which is proposed
to arise from hydrolysis of the lactone ring or failure of the
thioesterase domain to effect complete cyclization. Similar
byproducts have been identified during production of
CDAs[11] and other related lipopeptides.[12c] The proposed
linear variant 1, was subjected to tandem MS yielding
a product ion spectrum (Figure S4). From the products ions,
complete sequence data was obtained (Table 2) confirming
the presence of Gln at position 2 in place of Glu. The MT1110
DglmT-KQ mutant also produces CDA3a-10Q as the major
product, with CDA3a as a minor component (Figure S3).
Neither of the strains containing double point mutations,
MT1110-KQQE or MT1110 DglmT-KQQE, produced any
CDA lipopeptides as observed by LC-MS analysis. In light of
this, the MT1110 DglmT-KQ strain was selected for further
studies aimed at incorporating mGln into CDA.
Figure 1. A) Structures of CDA lipopeptides. B) Organization of the
CDA NRPS: C, condensation domain (red); adenylation domain with
cognate substrate inset (blue); T, thiolation domain; E, epimerization
domain (green); Te, thioesterase domain (orange). C) Alignment of
the module 10 A-domain of CdaPS3 with selected Glu and Gln
activating A-domains: gramicidin S synthetase A (GrsA); surfactin
synthetase A (SrfA); fengycin synthetase A (FenA); lichenysin synthe-
tase A (LicA); tyrocidine synthetase C (TycC). [a] Activates mGlu as
well as Glu.
mGlu precursor, thereby circumventing any competition with
the preferred natural substrate for the module 10 A-domain.
The host strains were then transformed with the plasmids
pKQ, pQE and pKQQE and using
Initially Boc-protected (2S,3R)-3-methyl glutamine was
synthesised (Scheme S1). However, removal of the Boc group
the standard double crossover
methodology several independent
mutant strains were generated for
each of the desired mutations.
Mutant strains derived from at
least two independent transforma-
tions were obtained in all cases and
these strains were grown in liquid
culture and then analyzed by LC-
MS for CDA production. This indi-
cates that the Gln236Glu mutation
Scheme 1. Linear variants of CDA3a-10Q 1 and CDA4a-10mQ 2.
2
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Table 2: MS-MS product ion series observed for CDA3a-10Q linear
peptide 1.
doubly charged and singly charged molecular ions in MS that
are consistent with a CDA derivative possessing a mGln
residue CDA4a-10mQ (Figure S5). This product was isolated
by HPLC and subjected to HRMS, which revealed doubly
charged and singly charged molecular ions that were con-
sistent with the proposed molecular formula of CDA4a-10mQ
(Table 1). The linear forms of CDA3a-10Q 1 and CDA4a-
10mQ 2 were also evident in the culture supernatant (Fig-
ure S6). Tandem MS analysis of CDA4a-10mQ (Figure 1)
resulted in a product ion spectrum (Figure S7 and Table 3)
y Ion[a]
Observed
Calcd
b Ion
Observed
Calcd
y10
y9
y8
y7
y6
1281.433
1180.415
994.313
879.289
764.264
615.217
500.191
443.171
313.130
185.071
1281.444
1180.397
994.317
879.290
764.263
615.216
500.189
443.167
313.130
185.071
b10
b9
b8
b7
b6
b5
b4
b3
b2
b1
–
1296.466
1168.407
1038.369
981.348
866.321
717.273
602.246
487.219
301.140
200.092
1168.421
1038.359
981.361
866.308
717.268
602.243
487.223
–
y5
y4
y3
y2[b]
y1[b]
Table 3: MS-MS product ion series observed for CDA4a-10mQ.[a]
–
[a] The masses of the y ion series are consistent with C-terminal acylium
ions, which are typically generated upon fragmentation of peptides
possessing N-terminal acyl groups. [b] Observed masses for y2 and y1
ions are consistent with a loss of Gln at this position.
y Ion
Observed
Calcd
b Ion
Observed
Calcd
y10
y9
y8
y7
y6
1295.250
–
1295.460
1194.412
1008.333
893.306
778.279
629.231
514.204
457.183
327.145
185.071
b10
b9
b8
b7
b6
b5
b4
b3
b2
b1
–
1310.481
1168.407
1038.369
981.348
866.321
717.273
602.246
487.219
301.140
200.092
1168.422
1038.495
981.211
866.319
717.255
602.247
487.212
301.099
200.084
1008.233
893.238
778.254
629.214
514.193
–
proved problematic and any (2S,3R)-3-methyl glutamine
(mGln) that may have been produced was subsequently
transformed into 3-methyl glutamic acid during work-up and
purification. Similar observations have been made previously
during the synthesis of other glutamine analogues.[14] We
suggest that any mGln initially produced on Boc-deprotection
undergoes acid-catalyzed cyclization to generate a pyrogluta-
mate (lactam) intermediate which is subsequently hydrolyzed
to give 3-methyl glutamic acid. In light of this we chose to
synthesize the Gly-mGln dipeptide 7 (Scheme 2), which we
y5
y4
y3
y2[b]
y1[b]
327.146
185.069
[a] The product ions series is derived from the doubly charged cyclic
parent ion [M+2H]2+. The y ion series masses are consistent with C-
terminal acylium ions. [b] Observed masses for y2 and y1 ions are
consistent with the presence of mGln at position 10.
that clearly indicates a peptide sequence with a mGln residue
at position 10, consistent with the proposed structure of
CDA4a-10mQ. The major product CDA3a-10Q was also
purified and subjected to additional characterization by
1H NMR spectroscopy. TOCSY and NOESY experiments,
and comparison of NMR data with the parent lipopeptide
CDA3a, provide further evidence to support the structural
assignment obtained using tandem MS (Tables S2 and S3 and
Figures S10–S17).
In summary, we have shown that introduction of a single
Lys278Gln mutation into the module 10 A-domain of the
CDA NRPS results in a novel glutamine-containing CDA
variant. Moreover, feeding the dipeptide Gly-mGln to
a Lys278Gln mutant lacking the natural mGlu precursor
(MT1110 DglmT-KQ) also leads to the incorporation of
synthetic mGln into CDA. The fact that the mGln-containing
CDA4a-10mQ was produced at lower levels than the natural
CDA4a from the parental strain may be due to the facile
hydrolysis of the amide side chain of the mGln precursor
released upon proteolysis. Nevertheless, (2S,3R)-configured
mGln has not been identified in nature to date and
consequently the work presented here provides the first
example of how active-site modification of an adenylation
domain within an NRPS can be used to introduce a synthetic
non-natural amino acid into a nonribosomal peptide natural
product. The combination of the mutasynthesis strategy
described here and directed evolution methods for selection
of A-domains with altered specificity[6] may prove to be
a powerful approach to expand the structural diversity of
nonribosomal peptides which are too complex for effective
Scheme 2. Synthesis of Gly-mGln: a) SOCl2, MeOH; 30 min at À208C
to 58C under N2 (91%); b) Cbz-Gly-OSu, NaHCO3 (10%) aqueous
solution, 1,4-dioxane; 30 min at RT (78%); c) 28% aqueous NH4OH;
18 h at RT (65%); d) H2/10% Pd/C in MeOH.
anticipated would be more stable to side-chain amide
hydrolysis, during fermentation, but would be susceptible to
proteolysis inside the cell releasing mGln for incorporation by
the CDA NRPS assembly line. The synthesis of the dipeptide
7 was achieved starting with mGlu 3,[11] which was selectively
esterified to give the monomethyl ester 4. Coupling with Cbz-
protected Gly gave dipeptide 5. Ammoniolysis of the ester
group of 5 provided Cbz-Gly-mGln 6 which was deprotected
by hydrogenolysis. The Gly-mGln dipeptide 7, which proved
to be stable to hydrolysis, was then fed to MT1110 DglmT-KQ
grown in liquid culture. LC-MS analysis of the culture
supernatant revealed production of CDA3a-10Q as the
major product, with an additional new product that exhibited
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[7] E. Conti, T. Stachelhaus, M. A. Marahiel. P. Brick, EMBO J.
1997, 16, 4174 – 4183.
synthetic modification. Ultimately such engineering
approaches might be useful for fine tuning the physicochem-
ical properties and biological activities of complex nonribo-
somal peptides, which are urgently required for pharmaceut-
ical, agrochemical, and other applications.
[8] T. Stachelhaus, H. D. Mootz, M. A. Marahiel, Chem. Biol. 1999,
6, 493 – 505; G. L. Challis, J. Ravel, C. A. Townsend, Chem. Biol.
2000, 7, 211 – 224; C. Rausch, T. Weber, O. Kohlbacher, W.
Wohlleben, D. H. Huson, Nucleic Acids Res. 2005, 33, 5799 –
5808.
[9] a) C.-Y. Chen, I. Georgiev, A. C. Anderson, B. R. Donald, Proc.
Natl. Acad. Sci. USA 2009, 106, 3764 – 3769; b) B. Villiers, F.
Hollfelder, Chem. Biol. 2011, 18, 1290 – 1299.
Received: March 14, 2012
Revised: May 18, 2012
Published online: && &&, &&&&
[10] a) Z. Hojati, C. Milne, B. Harvey, L. Gordon, M. Borg, F. Flett,
B. Wilkinson, C. P. Smith, J. Micklefield, Chem. Biol. 2002, 9,
1175 – 1187; b) A. Powell, M. Borg, B. Amir-Heidari, J. M.
Neary, J. Thirlway, B. Wilkinson, C. P. Smith, J. Micklefield, J.
Am. Chem. Soc. 2007, 129, 15182 – 15191; c) C. Mahlert, F. Kopp,
J. Thirlway, J. Micklefield, M. A. Marahiel, J. Am. Chem. Soc.
2007, 129, 12011 – 12018; d) A. Powell, M. Al Nakeeb, B.
Wilkinson, J. Micklefield, Chem. Commun. 2007, 2683 – 2685;
e) B. Amir-Heidari, J. Thirlway, J. Micklefield, Org. Biomol.
Chem. 2008, 6, 975 – 978; f) B. Amir-Heidari, J. Thirlway, J.
Micklefield, Org. Lett. 2007, 9, 1513 – 1516; g) J. Neary, A.
Powell, L. Gordon, C. Milne, F. Flett, B. Wilkinson, C. P. Smith, J.
Micklefield, Microbiology 2007, 153, 768 – 776; h) B. Amir-
Heidari, J. Micklefield, J. Org. Chem. 2007, 72, 8950 – 8953.
[11] C. Milne, A. Powell, M. Al Nakeeb, C. P. Smith, J. Micklefield, J.
Am. Chem. Soc. 2006, 128, 11250 – 11259.
[12] a) R. H. Baltz, Curr. Opin. Chem. Biol. 2009, 13, 144 – 151; b) L.
Robbel, M. A. Marahiel, J. Biol. Chem. 2010, 285, 27501 – 27508;
c) R. H. Baltz, V. Miao, S. K. Wrigley, Nat. Prod. Rep. 2005, 22,
717 – 741; d) J. Micklefield, Chem. Biol. 2004, 11, 887 – 888.
[13] G. Bucca, A. M. E. Brassington, G. Hotchkiss, V. Mersinias, C. P.
Smith, Mol. Microbiol. 2003, 50, 153 – 166.
[14] a) R. Dave, B. Badet, P. Meffre, Amino Acids 2003, 24, 245 – 261;
b) P. Meffre, R. H. Dave, J. Leroy, B. Badet, Tetrahedron Lett.
2002, 43, 6279.
Keywords: adenylation domain · antibiotics ·
biosynthetic engineering · natural products ·
nonribosomal peptides
.
[1] a) S. A. Sieber, M. A. Marahiel, Chem. Rev. 2005, 105, 715 – 738;
b) C. T. Walsh, Science 2004, 303, 1805 – 1810; c) B. Wilkinson, J.
Micklefield, Nat. Chem. Biol. 2007, 3, 379 – 386; d) M.-C. Wu, B.
Law, B. Wilkinson, J. Micklefield, Curr. Opin. Biotechnol. 2012,
DOI: 10.1016/j.copbio.2012.03.008.
[2] a) J. H. Lee, B. S. Evans, G. Li, N. L. Kelleher, W. A. van der
Donk, Biochemistry 2009, 48, 5054 – 5056; b) D. Schwartz, N.
Grammel, E. Heinzelmann, U. Keller, W. Wohlleben, Antimi-
crob. Agents Chemother. 2005, 49, 4598 – 4607; c) M. Ongena, P.
Jacques, Trends Microbiol. 2008, 16, 115 – 125.
[3] K. T. Nguyen, D. Ritz, J.-Q. Gu, D. Alexander, M. Chu, V. Miao,
P. Brian, R. H. Baltz, Proc. Natl. Acad. Sci. USA 2006, 103,
17462 – 17467; S. Doekel, M.-F. Coꢀffet-Le Gal, J.-Q. Gu, M.
Chu, R. H. Baltz, P. Brian, Microbiology 2008, 154, 2872 – 2880.
[4] G. C. Uguru, C. Milne, M. Borg, F. Flett, C. P. Smith, J.
Micklefield, J. Am. Chem. Soc. 2004, 126, 5032 – 5033.
[5] K. Eppelmann, T. Stachelhaus, M. A. Marahiel, Biochemistry
2002, 41, 9718 – 9726.
[6] B. S. Evans, Y. Chen, W. W. Metcalf, H. Zhao, N. L. Kelleher,
Chem. Biol. 2011, 18, 601 – 607.
4
www.angewandte.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Biosynthetic Engineering
Calcium-dependent antibiotics (CDA) are
cyclic lipopeptides assembled by non-
ribosomal peptide synthetase (NRPS)
enzymes. Active site modification of the
3-methyl glutamate activating adenyla-
tion (A) domain of the CDA NRPS
enables the incorporation of synthetic 3-
methyl glutamine into CDA. This provides
the first example of how A-domains can
be engineered to introduce synthetic
“non-natural” amino acids into nonribo-
somal peptides.
J. Thirlway, R. Lewis, L. Nunns,
M. Al Nakeeb, M. Styles, A.-W. Struck,
C. P. Smith,
J. Micklefield*
&&&& — &&&&
Introduction of a Non-Natural Amino
Acid into a Nonribosomal Peptide
Antibiotic by Modification of Adenylation
Domain Specificity
Angew. Chem. Int. Ed. 2012, 51, 1 – 5
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!