Biosynthetic Studies of Telomycin Reveal New Lipopeptides with Enhanced Activity

Article abstract of DOI:10.1021/jacs.5b01794

Telomycin (TEM) is a cyclic depsipeptide antibiotic active against Gram-positive bacteria. In this study, five new natural telomycin analogues produced by Streptomyces canus ATCC 12646 were identified. To understand the biosynthetic machinery of telomycin and to generate more analogues by pathway engineering, the TEM biosynthesis gene cluster has been characterized from S. canus ATCC 12646: it spans approximately 80.5 kb and consists of 34 genes encoding fatty acid ligase, nonribosomal peptide synthetases (NRPSs), regulators, transporters, and tailoring enzymes. The gene cluster was heterologously expressed in Streptomyces albus J1074 setting the stage for convenient biosynthetic engineering, mutasynthesis, and production optimization. Moreover, in-frame deletions of one hydroxylase and two P450 monooxygenase genes resulted in the production of novel telomycin derivatives, revealing these genes to be responsible for the specific modification by hydroxylation of three amino acids found in the TEM backbone. Surprisingly, natural lipopeptide telomycin precursors were identified when characterizing an unusual precursor deacylation mechanism during telomycin maturation. By in vivo gene inactivation and in vitro biochemical characterization of the recombinant enzyme Tem25, the maturation process was shown to involve the cleavage of previously unknown telomycin precursor-lipopeptides, to yield 6-methylheptanoic acid and telomycins. These lipopeptides were isolated from an inactivation mutant of tem25 encoding a (de)acylase, structurally elucidated, and then shown to be deacylated by recombinant Tem25. The TEM precursor and several semisynthetic lipopeptide TEM derivatives showed rapid bactericidal killing and were active against several multidrug-resistant (MDR) Gram-positive pathogens, opening the path to future chemical optimization of telomycin for pharmaceutical application. (Chemical Equation Presented).

Full text of DOI:10.1021/jacs.5b01794

Subscriber access provided by NEW YORK UNIV
Article
Biosynthetic Studies of Telomycin Reveal
New Lipopeptides with Enhanced Activity
Chengzhang Fu, Lena Keller, Armin Bauer, Mark Brönstrup, Alexandre Froidbise,
Peter Hammann, Jennifer Herrmann, Guillaume Mondesert, Michael Kurz,
Matthias Schiell, Dietmar Schummer, Luigi Toti, Joachim Wink, and Rolf Müller
J. Am. Chem. Soc., Just Accepted Manuscript • Publication Date (Web): 04 Jun 2015
Downloaded from http://pubs.acs.org on June 6, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Biosynthetic Studies of Telomycin Reveal New Lipopeptides
with Enhanced Activity
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chengzhang Fu,† Lena Keller,†, ‡ Armin Bauer,§ Mark Brönstrup,‡,§,# Alexandre Froidbise,∥ Peter
Hammann,⊥ Jennifer Herrmann,†, ‡ Guillaume Mondesert,∥ Michael Kurz,§ Matthias Schiell,§ Dietmar
Schummer,§,∇ Luigi Toti,§ Joachim Wink,§,# Rolf Müller*,†, ‡
†Helmholtz Institute for Pharmaceutical Research Saarland, Helmholtz Centre for Infection Research, and Department of
Pharmaceutical Biotechnology, Saarland University, Building C 2.3, 66123, Saarbrücken, Germany
‡German Centre for Infection Research, Inhoffenstraße 7, 38124 Braunschweig, Germany
§SanofiꢀAventis Deutschland GmbH, R&D LGCR, Industriepark Höchst, 65926 Frankfurt am Main, Germany
∥Sanofi R&D, TSU Infectious Diseases, 195 Route d‘Espagne, 31036 Toulouse, France
⊥SanofiꢀAventis Deutschland GmbH, R&D TSU Infectious Diseases, Industriepark Höchst, 65926 Frankfurt am Main,
Germany
ABSTRACT: Telomycin (TEM) is a cyclic depsipeptide antibiotic active against Gramꢀpositive bacteria. In this study, five new
natural telomycin analogues produced by Streptomyces canus ATCC 12646 were identified. To understand the biosynthetic maꢀ
chinery of telomycin and to generate more analogues by pathway engineering, the TEM biosynthesis gene cluster has been characꢀ
terized from S. canus ATCC 12646: it spans approximately 80.5 kb and consists of 34 genes encoding fatty acid ligase, nonribosoꢀ
mal peptide synthetases (NRPSs), regulators, transporters and tailoring enzymes. The gene cluster was heterologously expressed in
Streptomyces albus J1074 setting the stage for convenient biosynthetic engineering, mutasynthesis and production optimization.
Moreover, inꢀframe deletions of one hydroxylase and two P450 monooxygenase genes resulted in the production of novel telomyꢀ
cin derivatives, revealing these genes to be responsible for the specific modification by hydroxylation of three amino acids found in
the TEM backbone. Surprisingly, natural lipopeptide telomycin precursors were identified when characterizing an unusual precurꢀ
sor deacylation mechanism during telomycin maturation. By in vivo gene inactivation and in vitro biochemical characterization of
the recombinant enzyme Tem25, the maturation process was shown to involve the cleavage of previously unknown telomycin preꢀ
cursorꢀlipopeptides, to yield 6ꢀmethylheptanoic acid and telomycins. These lipopeptides were isolated from an inactivation mutant
of tem25 encoding a (de)acylase, structurally elucidated and then shown to be deacylated by recombinant Tem25. The TEM precurꢀ
sor and several semisynthetic lipopeptide TEM derivatives showed rapid bactericidal killing and were active against several multiꢀ
drugꢀresistant (MDR) Gramꢀpositive pathogens, opening the path to future chemical optimization of telomycin for pharmaceutical
application.
INTRODUCTION
lacking an aspartic acid and hydroxylation of cisꢀ3ꢀhydroxylꢀ
proline called LLꢀAO341 β 1 revealed that specific inhibition
of DNA, RNA, protein, lipid or peptidoglycan synthesis by
LLꢀAO341 β 1 could not be detected. It was postulated that
LLꢀAO341 β 1 could cause cell lysis by membrane deꢀ
energization.10
Telomycin (1) is a peptide antibiotic produced by Streptoꢀ
myces canus C159, which was described as effective antibiotic
against Gramꢀpositive pathogens when initially isolated at the
BristolꢀMyers Company (New York, USA) in the 1950s.1ꢀ3 It
was demonstrated to inhibit Staphylococcus aureus (SA) with
a minimal inhibitory concentration (MIC) of 8 ꢁg/ml and even
penicillinꢀresistant SA with a slightly higher MIC.1 In the
1960s, a first partial structure of telomycin containing a perꢀ
mutation of three amino acids was published.4,5 Two nuclear
magnetic resonance (NMR) based studies on telomycin in
1973 provided the correct structure of the complex
molecule.6,7 Telomycin (1, Chart 1) is a cyclic depsipeptide
which is composed of eleven amino acids including five nonꢀ
proteinogenic ones. It is comprised of a nonapeptide lactone
The emergence of multiꢀdrug resistance (MDR) among
many clinically relevant pathogens is an area of unmet mediꢀ
cal need. Natural products play a pivotal role in antibiotic drug
discovery and scaffolds with novel MoA are urgently
needed.11,12 For several antibiotic classes, structural modificaꢀ
tions of the natural parent molecule have been required to
sufficiently optimize key parameters such as potency, efficacy,
pharmacokinetics and safety so as to identify viable developꢀ
ment candidates which were then successfully forwarded into
clinical development.13ꢀ14 In this study, telomycin was found to
inhibit the growth of resistant pathogens including methicillinꢀ
resistant Staphylococcus aureus (MRSA), and vancomycinꢀ
intermediate SA (VISA). The latter phenotype is seen with
many hardꢀtoꢀtreat nosocomial infections. Since vancomycin
4
ring formed between the Thr hydroxyl group and the Cꢀ
terminal carboxyl group.4ꢀ7 Only a few reports on this comꢀ
pound were published after the structure of telomycin was
elucidated,8ꢀ10 and to date, the mode of action (MoA) is still
unclear. However, studies on a truncated analog of telomycin
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 23
is a firstꢀline antibiotic to treat infections caused by MRSA, A systematic inꢀframe inactivation of decorating genes was
1
2
3
4
5
6
7
8
new antibiotics with novel MoA are needed that overcome
reduced vancomycin susceptibility among clinical isolates.14
Currently, only very limited alternatives are available includꢀ
ing the use of naturally occurring highly potent lipodepsipeꢀ
tides such as daptomycin, which is already approved for the
treatment of complicated infections caused by MDR pathoꢀ
gens.15
crucial for deducing the function of specific tailoring enzymes
which subsequently led to the production of several novel
purified and structurally elucidated telomycin derivatives.
Surprisingly, the natural precursor of telomycin exhibited 2ꢀ4ꢀ
fold enhanced activity against Gramꢀpositive pathogens comꢀ
pared to telomycin and showed no crossꢀresistance to clinicalꢀ
ly used antibiotics. Based on this knowledge, a few novel
semisynthetic lipopeptides were synthesized and tested for
biological activities as a starting point for a future program of
chemical optimization of telomycin.
In order to rationally optimize the fermentation yields of teꢀ
lomycin as prerequisite for embarking on a program of strucꢀ
tural modifications, we investigated its biosynthesis, which to
the best of our knowledge, has not been elucidated. The depꢀ
sipeptide structure of telomycin suggests that it is probably
synthesized by a nonꢀribosomal peptide synthetase (NRPS)
pathway. NRPSs are multimodular enzymes in which each
module is responsible for the incorporation of a specific amino
acid residue of the elongated peptide. These modules contain
different catalytic domains and act in a successive way to
elongate the peptide chain. A typical NRPS elongation module
is comprised of a condensation (C) domain, an adenylation (A)
domain and a thiolation (T) domain. The A domain selects the
amino acid and activates it as aminoacylꢀAMP which is then
attached to the adjacent T domain. The C domain performs the
condensation step that connects the upstream Tꢀbound pepꢀ
tidyl chain with the downstream Tꢀbound aminoacyl residue
through peptide bond formation.16
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
Bacterial Strains, Plasmids and DNA Manipulation. S.
canus ATCC 12646 was the wildꢀtype telomycin producer
strain used in this study. Escherichia coli DH10B competent
cells were prepared as the host for general subcloning and
plasmid manipulation. E. coli GB2005 cells were used as the
host for homologous recombination engineering.29,30 E. coli
ET12567/pUZ8002 served as the host for intergeneric transfer
of knock out plasmid constructs to streptomycetes by conjugaꢀ
tion. The pIJ774 plasmid was used as the template to amplify
the apramycin resistance gene aac(3)IV and the oriT cassette
to replace the target gene. Two genomic libraries of S. canus
ATCC 12646 with different size inserts were constructed
(detailed procedures were described in the Supporting Inforꢀ
mation). SuperCos 1 plasmid (Agilent Technologies, USA)
was used as the vector for cosmid genomic library construcꢀ
tion while pBACe3.6 plasmid was used as the vector for bacꢀ
terial artificial chromosome (BAC) genomic library construcꢀ
tion. Expression vector pASKꢀIBA6 (IBA, Göttingen, Germaꢀ
ny) was used to heterologously express Tem25 in E. coli BL21
(DE3).
Recently, several NRPS and/or polyketide synthases (PKS)
derived compounds have been shown to be processed by proꢀ
drug or precursor activation mechanisms to become the mature
compound. For example, in the biosynthesis of xenocoumacin,
a peptidase was found to hydrolyze the acylated Dꢀasparagine
residue from the precursor called prexenocoumacin leading to
the bioactive xenocoumacin.17 Similar mechanisms also could
be found in the biosynthesis of other compounds such as zwitꢀ
termicin and colibactin.17ꢀ20 Pyoverdine is a fluorescent siderꢀ
ophore produced by a NRPS pathway in fluorescent pseudoꢀ
monads, supposedly involved in infection in various disease
models of the opportunistic pathogen Pseudomonas aeruginoꢀ
sa.21 During pyoverdine biosynthesis, a periplasmic NTN
hydrolase (PvdQ) was discovered that is responsible for the
cleavage of the myristate or myristoleate moiety from the Nꢀ
terminal amino acid. After knockout of pvdQ larger pyoverꢀ
dine derivatives harboring an ester moiety are produced.22,23
The structure of PvdQ was characterized and the in vitro
cleavage of a fatty acid moiety from pyoverdine precursors
were shown.24 Didemnin, a complex cyclic depsipeptide, is
also thought to be synthesized by a precursor maturation
mechanism. The pentaaminoꢀalcohol zeamine I, a broad specꢀ
trum antibiotic, was discovered to be hydrolyzed from
prezeamine I (with weaker activity) by a putative acylpeptide
hydrolase.25 Didemnin X and Y are acyl–glutamine ester deꢀ
The genomic DNA of Streptomyces strains was extracted
and purified manually following the cetyltrimethylammonium
bromide (CTAB) procedure and pulsed field gel electrophoreꢀ
sis (PFGE) grade chromosomal DNA preparation procedure
from Streptomyces Practical Genetics.31 Taq DNA polymerase
(Thermo Scientific) or Phusion HighꢀFidelity DNA Polymerꢀ
ase (Thermo Scientific) was used in PCR experiments. PCR
products were cloned into pCR2.1 TOPO vector (Life Techꢀ
nologies) or pJET1.2/blunt vector (Thermo Scientific). All
standard transformations and Red/ET recombinations with E.
coli were done by electroporation. Plasmid DNA was isolated
by standard alkaline lysis method. Restriction enzymes,
Shrimp Alkaline Phosphatase, T4 DNA ligase and other comꢀ
mon molecular biology reagents were purchased from Thermo
Scientific. The end sequencing of plasmids was done by LGC
genomics (Berlin, Germany). All primers (listed in Table S1ꢀ3)
used in this paper were synthesized by SigmaꢀAldrich (Steinꢀ
heim, Germany).
rivatives of the
bioactive didemnin B. Didemnin B is
In Frame Gene Inactivation. To inactivate a specific gene,
the cosmid bearing the gene was engineered to exchange the
gene for the aac(3)IV and oriT cassette by Red/ET recomꢀ
bineering.29,30,32 The conjugation process which transferred the
engineered cosmid from E. coli ET12567/pUZ8002 to S.
canus ATCC 12646 was performed according to standard
procedures.33 Primers and experimental design can be found in
the Supporting Information (Table S1, Figure S6ꢀS18).
thought to be the product of a cleavage of these larger derivaꢀ
tives by unknown proteins.26 Lastly, acylated saframycin preꢀ
cursors were discovered by in vitro reconstitution experiments;
however, the chemical nature of these compounds and the
enzymes involved in their maturation are still elusive.27,28
In this study, a precursor maturation mechanism involved in
telomycin biosynthesis was unveiled through both in vivo and
in vitro experiments. Moreover, the cloning, sequencing, muꢀ
tational and biochemical analysis and heterologous expression
of the intact telomycin biosynthetic gene cluster are described.
Identification, Sequencing and Analysis of the Telomycin
Biosynthetic Gene Cluster. The genomic DNA of strain
ATCC 12646 was analyzed by wholeꢀgenome sequencing. In
ACS Paragon Plus Environment

Page 3 of 23
Journal of the American Chemical Society
Chart 1. Structures of telomycins.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the draft genome data, a candidate NRPS gene cluster was
to 75 µl/min before entering the maXis HRꢀToF mass specꢀ
trometer (BrukerDaltonics, Bremen, Germany) using the
standard ESI source. Mass spectra were acquired in centroid
mode ranging from 150 ꢀ 2000 m/z at 2 Hz scan speed. The
peptides were linearized prior to MS/MS fragmentation analyꢀ
sis using 50 µl of the crude extract containing the depsipeptide
dissolved in methanol or 50 µg pure depsipeptide in 50 µl
methanol. 50 µl of 1N NaOH was added and the sample was
shaken for 4h at 40°C. The solvent was evaporated under
vacuum and reꢀdissolved in methanol prior to MS measureꢀ
ments.
discovered. One NRPS gene was inactivated via gene reꢀ
placement. Two BAC clones P1G3 and P1N13 which harbor
the telomycin biosynthetic gene cluster were also shotgun
sequenced as the genome sequence was found to contain sevꢀ
eral errors. The precise restriction map of the NRPS gene
region was verified by digestion of BAC clone P1B12 using
Eco72I, Bsp119I, NotI, Eco47III and PstI. The restriction map
of the TEM gene cluster region was verified by digesting
cosmids covering the cluster by BamHI and NotI and compariꢀ
son of the predicted to real fragments. A gap region containing
highly repetitive sequences was closed by a series of subꢀ
cloning steps and primer walking from cosmid clone P1N6.
Initial prediction and analysis of the telomycin biosynthesis
Heterologous Expression and Purification of Acylase
Tem25. The acylase gene was firstly submitted to SignalP 4.1
server (http://www.cbs.dtu.dk/services/SignalP/) for the signal
peptide prediction analysis.37 The part of the gene not encodꢀ
ing a possible signal peptide was amplified by PCR from the
genomic DNA of S. canus ATCC 12646 using Phusion Highꢀ
Fidelity DNA Polymerase (Thermo Scientific). The primers
gene
cluster
was
done
by using
antiSMASH
(http://antismash.secondarymetabolites.org/).34 Detailed preꢀ
diction of NRPS A domain specificity was performed using
NRPSpredictor2
(http://nrps.informatik.uniꢀtuebingen.de)35
and alignments36. The functional prediction of open reading
frames (ORFs) encoding proteins was performed using protein
blast and/or blastx program (http://blast.ncbi.nlm.nih.gov) and
Pfam (http://pfam.xfam.org). Routine DNA analysis such as
ORF identification, primer design, restriction analysis was
done using Geneious software (Biomatters Ltd, Auckland,
New Zealand). The accession number of the sequence depositꢀ
ed in GenBank is KP756960.
used
in
this
PCR
were:
5’ꢀ
GAATTCGCGCCCGCCCCGTCCGACꢀ3’ (forward primer,
with
EcoRI
restriction
site)
and
5’ꢀ
CTGCAGGGGGGCGGTCAGGACGACGGTꢀ3’
(reverse
primer, with PstI restriction site). The PCR product was ligatꢀ
ed into pJET1.2/blunt vector and transformed to E. coli
DH10B. The plasmid pJET1.2/blunt::tem25 was extracted and
digested with EcoRI and PstI. The acylase gene fragment with
correct sticky ends was inserted into pASKꢀIBA6 vector genꢀ
erating the expression plasmid pTM25. Strain E. coli BL21
(DE3) carrying pTM25 was grown in LB medium at 37 oC and
induced by adding anhydrotetracycline to 200 ng/ml when the
OD600 of the culture reached 0.6. The cells were harvested
after 4 hours of further cultivation after induction and reꢀ
suspended in high sucrose concentration buffer P (100 mM
TrisꢀHCl pH 8.0, 500mM sucrose, 1mM EDTA). The reꢀ
suspended mixture was incubated on ice for 30 minutes and
centrifuged for 5 minutes at 13,000 rpm to produce clear lyꢀ
sate. Further sonication was applied only if cells lysis was not
sufficient. The recombinant acylase was purified using Strepꢀ
Trap HP column (GE Helthcare, USA) based on StrepꢀTactin
Characterization of Telomycin Derivatives. NMR specꢀ
tra were recorded on a Bruker AVANCE 500 spectrometer, a
Bruker AVANCE 700 spectrometer or a Bruker Ascend 700
spectrometer all equipped with a 5 mm TXI cryoprobe. The
detailed conditions could be found in the Supporting Inforꢀ
mation. LCHRMS data was performed on a Dionex Ultimate
3000 RSLC system using a Waters BEHC18, 100 x 2.1 mm,
1.7 µm dp column. Separation of 2 µl sample was achieved by
a linear gradient with (A) H2O + 0.1 % FA to (B) ACN + 0.1 %
FA at a flow rate of 600 µl/min and 45 °C. The gradient was
initiated by a 0.5 min isocratic step at 5 % B, followed by an
increase to 95% B in 18 min. UV spectra were recorded by a
DAD in the range from 200 to 600 nm. The LC flow was split
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 23
affinity chromatography. The filtrated clear lysate was loaded
Table 1. MIC values (µg/ml) of telomycin and analogues
(see Chart 1 and 2).
1
2
3
4
to the column preꢀwashed by buffer W (100 mM TrisꢀHCl pH
8.0, 150 mM NaCl, 1 mM EDTA) and then washed by buffer
W again after sample application. The protein was eluted by
buffer E (100 mM TrisꢀHCl pH 8.0, 150 mM NaCl, 1 mM
EDTA, 2.5 mM desthiobiotin).
S. aureus
DSMꢀ11822
(MRSA)
E. faeciꢀ
um DSMꢀ
20477
S. aureus
Newman
Telomycin
5
6
7
8
9
4
4
2
4
16
64
1
2
Enzyme Assays. The reaction mixtures of in vitro enzyme
assays were prepared in 100 ꢁl of 20 mM TrisꢀHCl (pH 7.4)
buffer, supplemented with 0.1 mM substrate and 0.2 ꢁM acylꢀ
o
9
>64
>64
>64
>64
2
>64
>64
>64
>64
1
>64
>64
>64
>64
16
ase. The reaction was performed at 30 C for 4 h. The assay
solution was then mixed with 300 ꢁl methanol and dried by
evaporation. The residue was dissolved in 100 ꢁl methanol
and centrifuged for 10 min at 21, 500 ×g to remove particles
before HPLCꢀESIꢀMS analysis.
10
11
12
7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chart 2. Chemical structures of semisynthetic telomycin
analogues (7, 13ꢀ15).
13
14
15
DAPa
4
2
32
2
1
16
2
1
1
0.5
0.25
ꢀ
a reference antibiotic daptomycin; ꢀ not determined.
Antibiotic Activity. Telomycin showed good antibiotic acꢀ
tivities against S. aureus as well as Enterococcus faecium. The
natural analogues 9ꢀ12 all lost activity, revealing the nonapepꢀ
tide lactone ring as critical for biological activity, whereas the
loss of one hydroxylation found in 2 only led to a minor deꢀ
crease of activity against S. aureus. Compound 7 and 13ꢀ15 all
showed comparable and by tendency improved activities comꢀ
pared to 1 (Table 1).
The antibacterial activity of telomycin and its semisynthetic
analogues was also evaluated in a panel of resistant Gramꢀ
positive bacteria. Intriguingly, all tested derivatives were
active on MRSA, VISA and VRE in the low µg/ml range. We
also assessed possible crossꢀresistance with structurally related
daptomycin and could demonstrate that telomycins are only by
factor 4ꢀ16 less active on an isolate that displays highꢀlevel
daptomycin resistance (DRSA; > 1000ꢀfold change in MIC
compared to WT). Furthermore, the MIC of daptomycin on a
telomycin resistant SA strain (TRSA; > 32ꢀfold change in
MIC of 15 compared to WT) did not increase when compared
to SA Newman WT (Table 2).
RESULTS and DISCUSSION
Natural Telomycin Derivatives. From a 16 liter fermentaꢀ
tion of S. canus ATCC 12646, compound 2 and four additional
new compounds (9ꢀ12) representing minor products or shunt
products were isolated besides 1. The structure of 2 was eluciꢀ
dated by extensive analysis of the 1D and 2D NMR data (Supꢀ
porting Information). In comparison to 1, 2 contained a methꢀ
ylene group instead of an oxygenated methine located in the
Cꢀterminal proline showing that it was missing the hydroxylaꢀ
tion of 11Pro (Chart 1). Two peaks of telomycins with the same
m/z and overlapping retention time could be observed by
HPLCꢀMS analysis (Figure 2); however, this phenomenon has
been observed in other prolineꢀcontaining peptides where the
peaks belong to the cisꢀtrans isomers of the same proline
compound.38ꢀ40 The structures of 9ꢀ12 were elucidated by the
1D and 2D NMR experiments (Chart 1) (Supporting Inforꢀ
mation). Compounds 9 and 10 are ringꢀopened telomycins
with different degrees of hydroxylation; they are probably
intermediates released spontaneously or by catalysis of the TE
domain. Compound 11 is a shunt product representing a premꢀ
aturely released intermediate while compound 12 is a macroꢀ
lactone formed between 2Ser and 11Pro instead of lactonization
More importantly, compound 15 showed excellent early
bactericidal activity versus SA (Figure S5A) and VRE (Figure
S5B) in a standard timeꢀkill assay. The bactericidal effect of
15 on SA is even slightly superior to that of daptomycin in
terms of killing rate (Figure S5A).
The cytotoxicity of 1, 7 and 13ꢀ15 on Chinese hamster ovaꢀ
ry (CHOꢀK1) cells was evaluated in standard growth inhibiꢀ
tion experiments and none of the tested compounds was active
at concentrations up to 40 µg/ml.
Identification and Characterization of the Telomycin
(TEM) Biosynthetic Gene Cluster. The structures of the
telomycins suggest that their biosynthetic route is based on
NRPS biochemistry. To find the corresponding biosynthetic
gene cluster, the genome of S. canus ATCC 12646 was seꢀ
quenced by Illumina sequencing technology and one candidate
NRPS gene cluster matching our expectations was found.
However, the substrate specificity predictions of the A doꢀ
mains in this NRPS gene cluster were not consistent with all
building blocks in telomycin. According to the initial genome
3
between Thr and 11Pro. The implications of the identification
of these shunt products with different length and hydroxylaꢀ
tion during the biosynthesis of 1 will be discussed below. In
all these telomycin analogs, the absolute stereochemistry of
the βꢀposition in the βꢀmethyltryptophan (βꢀMeTrp) moiety
could not be determined.
ACS Paragon Plus Environment

Page 5 of 23
Journal of the American Chemical Society
Table 2. Crossꢀresistance evaluation of telomycin and selected analogues (MIC in µg/ml; see Charts 1 and 2).
1
SA N315
(MRSA)
SA Mu50
(MRSA/VISA)
E. faecium DSMꢀ
2
3
SA WT
DRSAa
TRSAb
17050 (VRE)c
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
1
4
1
2
2
0.25
1
2
4
2
2
ꢀ
> 64
16
64
64
32
4
7
0.13
ꢀ
ꢀ
13
2
2
0.5
14
1
2
0.13
ꢀ
15
2
1
0.5
64
ꢀ
AMPd
CIPd
DAPd
ERYd
RIFd
TETd
VANd
> 64
0.25
0.13
> 64
0.01
2
> 64
16
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
2
0.25
> 64
> 0.64
64
≤ 0.005
5
ꢀ
0.5
ꢀ
8
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
0.03
ꢀ
ꢀ
ꢀ
1
8
ꢀ
ꢀ
> 64
a in vitro obtained daptomycinꢀresistant mutant (derivative of S. aureus WT); b in vitro obtained telomycin analogue 15ꢀresistant
c
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
mutant (derivative of S. aureus Newman); vancomycinꢀresistant Enterococcus; d reference antibiotics (AMP: ampicillin, CIP:
ciprofloxacin, DAP: daptomycin, ERY: erythromycin, RIF: rifampicin,TET: tetracycline, VAN: vancomycin); ꢀ not determined.
assembly this NRPS gene cluster exhibited only eight NRPS
modules and thus seemed unlikely to be responsible for teloꢀ
mycin production. To test whether telomycin biosynthesis
depends on this NRPS gene cluster, the NRPS gene tem22 was
inactivated using an inꢀframe gene inactivation strategy (Figꢀ
ure S9). The resulting mutant ꢁtem22 lost the ability to proꢀ
duce 1 and 2, which proved the involvement of this NRPS
gene cluster in telomycin biosynthesis (Figure S21). Meanꢀ
while, cosmid and BAC genomic libraries of S. canus ATCC
12646 were constructed. To deduce possible misꢀassemblies
causing a gap in the gene cluster, we screened these two geꢀ
nomic libraries with PCR probes targeting three genes located
in the 5’ end, the middle and the 3’ end of the gene cluster
(Tab. S2). Through a 2ꢀdimensional highꢀthroughput screening
strategy (Supporting Information), five cosmids (pCP2K22,
pCP1G10, pCP1N6, pCP3A6, pCP2G24) spanning approxiꢀ
mately 120 kb and three BAC plasmids (pBP1B12, pBP1G3,
pBP1L2) overlapping about 104 kb were identified that covꢀ
ered the whole NRPS gene cluster (Figure 1). This physical
genomic presentation was characterized by end sequencing
and classical restriction mapping (Figure 1, S3 and S4).
93.2% DNA identity and more than 850 bp of their middle
parts are exactly identical in sequence, probably because they
are both responsible for threonine activation. As discovered
before in hormaomycin biosynthesis, a recombinatorial exꢀ
change of a short DNA region approximately 420 bp led to
different A domain functionality. However, the A domain of
module 2 (A2) and A domain of module 5 (A5) in telomycin
biosynthetic gene cluster share as high as 99.4% DNA identity
except for only ten nucleotides difference. The Stachelhausꢀ
codes of A2 and A5 are exactly the same (Table S4), seven
amino acids outside the binding pocket are different (Figure
S2B). However, these two A domains are expected to activate
different amino acids: Ser and Ala, respectively. These feaꢀ
tures suggest point mutations to be involved in the evolution
of NRPS function as already described for the myxochromide
pathway.41 In addition, the C domains of module 3 and module
6 share as much as 93.1% DNA identity. Most likely, these
repeats led to the wrong assembly of C3 and C6 and the deleꢀ
tion of the region in between them in the initial assemblies.
The corrected gene cluster sequence was then verified by
classical physical restriction mapping (Figure 1, S3, and S4).
By using different restriction enzymes to digest these cosꢀ
mids and BAC plasmids, we found a region larger than 10 kb
missing in the genome assembly (Figure S1). To close this gap,
the BAC plasmid pBP1L2 covering the whole gene cluster,
was sequenced again by Illumina sequencing technology.
Assembly of the resulting primary data revealed that only a
7.9 kb fragment of the gap was identified and added to the
sequence still leaving a gap approximately 5 kb in size. The
sequence of this fragment was determined by a series of subꢀ
cloning and primer walking steps based on cosmid pCP1N6.
The completely closed gap was found to be approximately 11
kb in size and an additional 2 kb fragment of wrong sequence
assembly found immediately downstream of the gap was also
revised (Figure S1). Taking a closer look at this region, we
found that it contains highly repetitive DNA segments which
the employed genome assembly algorithm obviously failed to
assemble correctly, probably due to the relatively short read
length of Illumina sequencing technology (Figure S2A). The
first two newly identified A domains of module 3 and 4 share
To determine the boundaries of the TEM cluster, a series of
ORFs within the gene cluster were inactivated by gene reꢀ
placement. The simultaneous inactivation of two genes tem1
and tem2 related to sugar metabolism dramatically impaired
the production of telomycins which suggests that these genes
affect the biosynthesis of telomycin (Figures S7 and S21). In
contrast, inactivation of a hypothetical protein encoding gene
orf(ꢀ1) did not have any influence on telomycin production
which indicates that this gene is located outside of the TEM
cluster (Figures S6 and S21). As to the 3’ end boundary, we
inactivated the genes tem33 and tem34 exhibiting hypothetic
functions and the production of telomycins disappeared in
both mutants (Figures S16, S17 and S21). However, the inacꢀ
tivation of gene orf(+1) and orf(+2) simultaneously did not
affect the production of telomycins (Figures S18 and S21). In
addition to the functional assignment of putative gene products
within the cosmid overlapping region, these in vivo gene inacꢀ
tivation results revealed the TEM biosynthetic gene cluster to
span 80,533 bp with a high GC content (73.0%) and to consist
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 23
Table 3. Deduced functions of ORFs encoded in the Telomycin biosynthetic gene cluster.
1
Size
(aa)a
Predicted function in TEM
biosynthesis
Best match accession no.
2
3
4
Gene
Best match protein/organism
hypothetical protein
(identity/similarity % )a
orf(ꢀ1)
433
CAJ87953 (69/76)
hypothetic protein/Streptomyces coelicolor A3(2)
hydrolase/Streptomyces coelicolor A3(2)
5
6
7
8
tem1
tem2
tem3
tem4
tem5
1051
401
273
280
426
NP_624423 (93/95)
NP_624422 (93/95)
NP_624421 (99/99)
NP_624420 (98/99)
NP_624419 (96/98)
ABC transporter
ABC transporter
ABC transporter
integral membrane permease/Streptomyces coelicolor A3(2)
integral membrane permease/Streptomyces coelicolor A3(2)
extracellular binding protein/Streptomyces coelicolor A3(2)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
tem6
tem7
572
297
347
434
567
267
hypothetical protein/Streptomyces coelicolor A3(2)
putative glucokinase /Streptomyces coelicolor A3(2)
NP_624418 (93/96)
NP_624417 (91/93)
NP_624416 (96/97)
NP_624415 (92/94)
GAF44958 (61/75)
YP_004335958 (39/53)
transcriptional regulator
transcriptional regulator
tem8
LacI family transcriptional regulator/Streptomyces coelicolor A3(2)
hypothetical protein/Streptomyces coelicolor A3(2)
tem9
tem10
tem11
putative hydrolase/Rhodococcus wratislaviensis NBRC 100605
LuxR family transcriptional regulator/Pseudonocardia dioxanivorans CB1190
transcriptional regulator
tryptophan dehydrogenase
tem12
314
NADPH:quinone reductase/Amycolatopsis mediterranei
KDO08548 (85/93)
tem13
tem14
tem15
tem16
tem17
202
477
352
160
306
conserved hypothetical protein/Streptomyces ambofaciens ATCC 23877
transcriptional regulator, partial/Streptomyces venezuelae ATCC 10712
anthranilate phosphoribosyltransferase/Streptomyces toyocaensis
secreted protein/Streptomyces sp. ML694ꢀ90F3
CAJ89303 (58/66)
AAF01064 (67/75)
KES07266 (63/77)
BAP34762 (73/84)
YP_007746804 (51/62)
transcriptional regulator
transcriptional regulator
ArsR family transcriptional regulator/Streptomyces albus J1074
tem18
tem19
tem20
tem21
tem22
tem23
tem24
582
89
fatty acid CoA ligase
acyl carrier protein
NRPS
long chain fatty acid CoA ligase/uncultured bacterium esnapd4
peptidyl carrier protein/Streptomyces albus
NRPS/uncultured bacterium esnapd4
AGS49393 (63/72)
AEZ53971 (49/60)
AGS49395 (52/63)
AGS49328 (50/62)
AGS49397 (54/66)
AEA30275 (63/78)
AAU34213 (79/87)
4742
5824
2401
419
74
NRPS
NRPS/uncultured bacterium esnapd2
NRPS
NRPS/uncultured bacterium esnapd4
P450 monooxygenase
MbtH
P450 monooxygenase/Streptomyces sp. Acta 2897
MbtHꢀlike protein/Streptomyces hygroscopicus
tem25
tem26
824
268
cyclic lipopeptide acylase
methyltransferase
hypothetic protein/uncultured bacterium esnapd4
AGS49399 (56/68)
AGS49400 (67/77)
betaꢀketoadipate enolꢀlactone hydrolase/uncultured bacterium esnapd4
tem27
tem28
tem29
tem30
tem31
tem32
tem33
tem34
orf(+1)
orf(+2)
715
916
405
272
314
287
296
667
837
247
methyltransferase/Streptomyces flocculus
hypothetical protein/uncultured bacterium esnapd4
AFW04573 (39/54)
AGS49410 (62/75)
P450 monooxygenase
ABC transporter
P450 monooxygenase/Kutzneria albida DSM 43870
AHI00608 (41/56)
ABC transporter/Streptomyces coelicolor A3(2)
NP_627437 (41/58)
ABC transporter
ABC transporter/Kribbella flavida DSM 17836
YP_003381220 (71/80)
PDB: 4P7W_A (31/48)
YP_007592609 (29/41)
YP_008013999 (33/45)
YP_007523698 (38/52)
YP_003299905 (68/79)
proline hydroxylase
Chain A, Lꢀprolineꢀbound Lꢀproline Cisꢀ4ꢀhydroxylase/Mesorhizobium loti MAFF303099
hypothetical protein/Streptomyces davawensis JCM 4913
hypothetical protein/Amycolatopsis orientalis HCCB10007
ABC transporter integral membrane protein/Streptomyces davawensis JCM 4913
ABC transporter/Thermomonospora curvata DSM 43183
a amino acids
of 34 ORFs from tem1 to tem34. These ORFs were initially
characterized by bioinformatic analysis and their putative
functions were assigned by BLAST searches. The 34 genes
can be classified in subcategories according to their functions
in the biosynthesis of telomycin. First of all, in the center of
the gene cluster there are three large NRPS genes (tem20ꢀ22).
At the 5’ꢀend of the NRPS genes, there are genes responsible
for fatty acid activation and attachment (tem18, 19) and the
remaining genes apart from several genes of unknown funcꢀ
tion are identified as encoding tailoring enzymes, transporters,
and regulatory proteins (Table 3).
Heterologous Expression of the TEM Biosynthetic Gene
Cluster. After the integration of BAC plasmids harboring the
TEM cluster into the genome of heterologous hosts, several
resulting apramycinꢀresistant mutants were fermented to test if
any telomycin was produced. The HPLCꢀESIꢀMS analysis
revealed one telomycin derivative produced in low amount in
the mutants of Streptomyces albus J1074 which carry the
ACS Paragon Plus Environment

Page 7 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
Figure 1. The gene organization and restriction map of the TEM gene cluster as well as the proposed biosynthetic pathway of telomycin.
(A) Restriction map of the telomycin biosynthetic gene cluster. The dashed blue line indicates the gap containing highly repetitive DNA
sequence. The restriction map of the whole DNA region covered by cosmids and BACs is found in the Supporting Information. (B) Tem18
and Tem19 initially activate a 6ꢀmethylheptanoate and transfer it to the carrier protein where Tem20, Tem21 and Tem22 build up the
peptide chain. The three hydroxylation reactions performed by Tem23, Tem29 and Tem32 most likely take place postꢀNRPS assembly.
The complex βꢀmethylation of tryptophan is probably achieved prior to the activation by the respective A domain. After the mature
lipopeptide is released and cyclized, the 6ꢀmethylheptanoyl moiety is cleaved off by Tem25 to yield telomycin.
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 23
inserts pBP1G3HE and pBP1L2HE, respectively (Figure S19 according to the chemical structure of telomycin. The C doꢀ
1
2
3
4
5
6
7
8
and S22). This compound was identified as 2 based on the
same retention time, molecular mass and fragmentation pattern
(tR=3.99 min, molecular mass [M+H]+ at m/z of 1256.6, Figure
S22). In the mutants of Streptomyces coelicolor A(3)2 and
Streptomyces lividans TK64, we did not identify any derivaꢀ
tive of 1. The production level of telomycins was very low,
probably because of complex regulation process involved in
biosynthesis of TEM, as suggested by numerous regulators
located in the TEM gene cluster. Although the heterologous
expression of the TEM cluster in S. albus did not lead to a
high titer of telomycin production, it is still a promising platꢀ
form for genetic engineering. Further experiments such as
promoter exchange are currently in progress to improve teloꢀ
mycin production in S. albus J1074.
main located at the Nꢀterminus of module 1 of Tem20 shows
similarity to enzymes which acylate the first amino acid with
activated fatty acid moieties such as ACMS II47, DhbF48 and
PstB49. Thus, the presence of the Nꢀterminal starter C domain
implies that 1Asp might be acylated in an unkown precursor of
telomycin (see below).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Nonribosomal Peptide Synthetases. Three large genes
tem20, tem21 and tem22 are located in the center of the TEM
gene cluster and their deduced protein products are typical
NRPS enzymes exhibiting a modular organization. These three
giant enzymes Tem20, Tem21 and Tem22 contain 4, 5 and 2
modules, respectively, totaling 11 modules comprising C, A,
and T domains (Figure 1B).
Two additional epimerization (E) domains were found in
modules 4 and 9 (Figure 1B). Usually, an E domain would
convert an
However, the two amino acid residues incorporated by modꢀ
ules 4 and 9 are both ꢀamino acids, which still keep the
form in the presence of E domains. The amino acid incorpoꢀ
rated by module 4 is ꢀalloꢀThr which might be formed from
ꢀThr since the (2S, 3S) stereoisomer is rare in nature. As there
is no precedent of βꢀepimerization by an E domain, there
might be other enzymes involved in the configuration converꢀ
sion. However, the possibility that the βꢀepimerization is cataꢀ
lyzed by the E domains cannot be excluded at this moment.
The βꢀMeTrp residue of telomycin seems to be incorporated
by module 9. However, whether it represents the (2S,3S) or
(2S,3R) stereoisomer is still unclear since the absolute configꢀ
uration of βꢀposition of βꢀMeTrp in natural products has not
been determined yet. Hence, whether the epimerization doꢀ
main in module 9 is functional and what kind of function it has
remains elusive.
Lꢀamino acid to the respective D
ꢀamino acid.16
Figure 2. HPLCꢀMS analysis of telomycins in crude extracts
(smallꢀscale fermentation) of the wildꢀtype and tem25 geneꢀ
inactivated mutant strain of Streptomyces canus ATCC 12646.
Extracted ion chromatograms (EIC) of telomycins (1,
[M+H]+=1272.6, green, 2, [M+H]+=1256.6, blue, 7,
[M+H]+=1398.7, red, and 8, [M+H]+=1382.7, violet) were anaꢀ
lyzed for both mutant and wildꢀtype strains to test production.
Compounds found in the wildꢀtype and mutant strains are identiꢀ
fied by numbers and their structures as shown in Chart 1.
L
Lꢀ
L
L
Natural Precursor Maturation Processed by an Acylase.
Since telomycin is a cyclic peptide and not a lipopeptide, it is
surprising to find a starter C domain located at the N terminus
of Tem20 that cannot be functionally assigned. Upstream of
the first NRPS gene tem20, tem18 is located which putatively
encodes a fatty acid CoA ligase and tem19 encodes an acyl
carrier protein (ACP). According to the protein function preꢀ
diction, Tem18 and Tem19 should be involved in activation of
a fatty acid moiety and its loading onto the ACP. Tem18 has a
characteristic domain of fatty acylꢀAMP ligase (FAAL) which
belongs to the class I adenylate forming enzyme family, and
its homologues are assumed to be responsible for fatty acid
activation in lipopeptide biosynthesis such as laspartomycin50,
friulimicin51, taromycin A52 or glycopeptide biosynthesis, e.g.
for teicoplanins53. Most fatty acid moieties in these lipopepꢀ
tides are branched or unsaturated.
The substrate specificity of each A domain was predicted
according to the putative bindingꢀpocket constituents (the soꢀ
called Stachelhausꢀcode).36 However, the substrate predictions
for A5, A8, A9 and A10 are not consistent with the amino
acids present in 1 (Tab. S4). All close hits predicted for the
5Alaꢀactivating domain A5 are Serꢀactivating domains found
in the NRPS for CDA,42 enterobactin,43 nostopeptolide44 and
belomycin45 biosynthesis. A8 is expected to activate Trp or Zꢀ
ꢁTrp, but the closest prediction for A8 is Val. ZꢀꢁTrp was also
found in CDA, but the binding pocket code (DGWAVASVCK)
of the A domain from CDAꢀmodule 11 incorporating ZꢀꢁTrp
is quite different from the code of A8 in the TEM cluster.42
Since module 9 of the TEM cluster may serve to incorporate
βꢀMeTrp as found for the maremycin gene cluster, A9 of the
TEM cluster and the A domain from the maremycin cluster
were aligned.46 Surprisingly, the A domain from the maremyꢀ
cin cluster (code DGYLYGSFTK) is very different from the
TEM A9 code, implying different substrate are activated by
these A domains.46
All these points suggest that the TEM biosynthetic machinery
produces a typical lipopeptide. To verify our hypothesis that
there might be a tailoring step involved in forming the mature
peptide from a lipopeptide precursor, we turned our attention
to the remaining genes in the cluster that were not functionally
assigned to telomycin biosynthesis. Indeed, tem25 encodes a
putative cyclic lipopeptide acylase. The deduced product
Tem25 shares a characteristic domain with a family of betaꢀ
lactam acylases which includes penicillin G acylase (PGA)
and cephalosporin acylase (CA). These proteins all contain a
conserved Ntn (Nꢀterminal nucleophile) hydrolase fold and
belong to the Nꢀterminal (Ntn) hydrolase superfamily. PGA
Except for the Nꢀterminal C domain, all other domains are
found to be consistent with the expected number and order
ACS Paragon Plus Environment

Page 9 of 23
Journal of the American Chemical Society
and CA share low sequence similarity with each other, but the was found connected to the Nꢀterminal asparagine through a
1
2
3
4
5
6
7
8
structural similarities surrounding the active sites are very high.
However, regardless of the conserved active sites, they hydroꢀ
lyze the amide bonds of very different substrates.54 Tem25
displays 63% similarity to an acylase found in Streptomyces sp.
no. 6907 that hydrolyzes the palmitoyl moiety of the antifunꢀ
gal lipopeptide FR901379 produced by Coleophoma empetri
Fꢀ11899.55 It also shows 52% similarity with PvdQ, which is
critical in pyoverdine maturation and transportation.22,24
key HMBC correlation from Hꢀ2Asp1 to the carbonyl Cꢀ1MHA
(Supporting Information). Compound 8 showed the same
pattern as 2, identifying it as an acylated derivative missing
the hydroxylation of the Cꢀterminal proline residue (Figure
S24).
Since
7 and 8 are lipopeptides containing the 6ꢀ
methylheptanoyl moiety coupled to the amino group of the Nꢀ
terminal Asp of 1 and 2, respectively, the gene tem25 was thus
proposed as encoding the acylase which can hydrolyze the 6ꢀ
methylheptanoyl moiety of the lipopeptidic precursor to proꢀ
duce the nonꢀacylated peptide telomycins (Figure 1B).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To our surprise, testing of 7 unveiled an increased activity
compared to 1. It is remarkable that the biological activities
against S. aureus and methicillinꢀresistant Staphylococcus
aureus (MRSA) were increased for the lipopeptide 7 that bears
an acyl lipid side chain while the activity against E. faecium
was not improved (Table 1).
Several semisynthetic derivatives with different acyl chains
1
attached to the amino group of the Asp of TEM were preꢀ
pared (Scheme S1 and Supplemental Experimental Section). A
selection of these tool compounds (13, 14, and 15) (Chart 2
and Supporting Information) were assessed for their antibacteꢀ
rial activities, which confirmed the beneficial effect of a lipid
side chain on activity. The biological activity against S. aureus
and methicillinꢀresistant Staphylococcus aureus (MRSA) was
also increased for all semisynthetic lipopeptides 13, 14, and 15
but activity against E. faecium was only increased in the case
of 15. Compound 15 displayed a 4ꢀfold activity increase
against SA and MRSA and a 16ꢀfold activity increase against
E. faecium (Table 1).
Figure 3. SDSꢀPAGE analysis and in vitro characterization of
Tem25. A standard Tem25 assay was prepared in 100 ꢁl of 20
mM TrisꢀHCl (pH 7.4) buffer, supplemented with 0.1 mM subꢀ
strate and 0.2 ꢁM acylase. (A) The recombinant enzyme was
purified as described in Materials and Methods. The α and β
subunit are indicated by arrows. M, protein marker; lane 1, uninꢀ
duced cells; lane 2, induced cells; lane 3, lysate supernatant of
induced cells; lane 4, semiꢀpurified Tem25. (B) HPLC analysis of
(a) 7 without adding Tem25; (b) boiled Tem25 incubated with 7;
(C) Tem25 incubated with 7.
The question why a “more potent” precursor is hydrolyzed
to give a natural product with weaker antibiotic activity curꢀ
rently remains elusive. One possible answer is that telomycin
plays a different role in nature than acting as an antibiotic
against Gramꢀpositive bacteria.56,57 Moreover, we considered
that the deacylation step could constitute a detoxification or a
soꢀcalled selfꢀresistance mechanism. We analyzed the activity
of the acylase in a time course experiment of S. canus ATCC
12646 fermentation. The acylase activity could be detected
from the third day of fermentation which is when the telomyꢀ
cin production level is still low. Moreover, we never found the
production of the natural precursors 7 or 8 in the fermentation
of the wildꢀtype telomycin producer. It thus seems that the
acylase Tem25 is expressed as early as the NRPS genes. Howꢀ
ever, the concentration of telomycin in the natural environꢀ
ment is probably much lower than under laboratory conditions
and relatively low concentrations might already trigger the
expression of Tem25.
To probe whether tem25 plays a role in telomycin biosynꢀ
thesis, this gene was inactivated via replacement with the
aac(3)IVꢀoriT cassette (Figure S11). The corresponding muꢀ
tant Streptomyces canus ATCC 12646 ꢁtem25 stopped proꢀ
ducing 1 and 2 but accumulated larger and more hydrophobic
novel products. From a six liter fermentation of mutant strain
ꢁtem25, two new compounds were isolated and structurally
elucidated as lipopeptidic telomycin derivatives
7
([M+H]+=1398.7, tR=5.11 min) and 8 ([M+H]+=1382.7,
tR=5.33 min), both containing a 6ꢀmethylheptanoic acid resiꢀ
In vitro Characterization of Acylase Tem25. To confirm
the role of Tem25 in the biosynthesis of 1, Tem25 was characꢀ
terized in vitro. Since many PGAs and CAs from Streptomyces
were identified as secreted proteins, the deduced product of
tem25 was analyzed for the presence of a signal peptide. Inꢀ
deed, a cleavage site was predicted as APAꢀAP between resiꢀ
due 32 and 33, and the first 32 residues of Tem25 contain the
typical signal peptide N, H and C region indicative for an Nꢀ
terminal signal peptide.37,58 The part of tem25 without the
signal peptide encoding sequence was amplified from ATCC
12646 genomic DNA, and the corresponding recombinant
protein was heterologously expressed in E. coli and purified as
a Strepꢀtag fusion protein (Figure 3A). To test the in vitro
1
due (MHA) (Figure 2 and Chart 1). Comparing the H NMR
data with 1, compound 7 showed additional signals for methꢀ
ylene groups between 1.18 and 2.24 ppm, a methine at 1.52
ppm and two additional methyl groups at 0.86 ppm. The seꢀ
quential spin system of the 6ꢀmethylheptanoic acid residue
was deduced from COSY correlations starting from the two
methyl groups at δ 0.86 (Meꢀ7 and Meꢀ8MHA) attached to the
methine at δ 1.52 (Hꢀ6MHA) and comprising four sequential
methylene groups Hꢀ5 to Hꢀ2. Together with the previous
findings, HMBC correlations from Hꢀ3MHA and Hꢀ2MHA to
a carbonyl resonance at d 175.8 (Cꢀ1MHA) clearly identified
MHA as a branched chain fatty acid moiety. The fatty acid
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 23
hydrolytic activity of Tem25, the recombinant protein was
biosynthesis.59,60 A tem23 inactivation mutant was created and
1
2
3
4
5
6
7
8
incubated together with crude extracts from a ꢁtem25 fermenꢀ
tation. As expected, LCꢀMS analysis identified the accumulaꢀ
tion of 1 and 2 while 7 and 8 disappeared from the extract.
Furthermore, recombinant Tem25 was incubated with pure 7
in vitro, and the conversion from 7 to 1 was proven by LCꢀMS
analysis (Figure 3B). These data confirmed the acylase activiꢀ
ty of Tem25, which hydrolyzes 7 and 8 to release the 6ꢀ
methylheptanoyl moiety.
the crude extract from the fermentation broth was analyzed by
HPLCꢀESIꢀMS (Figures S10 and S20). We found that 1 and 2
were not produced by ꢁtem23. However, two new peptides 3
and 5 were discovered in the culture broth (Chart 1 and Figure
S20). Linearization of these peptides for MS/MS analysis
yielded the linearized analogues 16 and 17, respectively. The
fragmentation pattern of 16 proved that the hydroxylation of
the leucine residue was missing in 3 and the fragmentation
pattern of 17 showed that the hydroxyl groups of βꢀOHꢀLeu
and cisꢀ3ꢀOHꢀPro were both lost in 5 (Figure 5A). Peptide 3
and 5 thus lack the βꢀhydroxyl group of βꢀOHꢀLeu compared
to 1 and 2, respectively, revealing that Tem23 is the P450
monooxygenase responsible for βꢀhydroxylation of leucine as
found in telomycin.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Interestingly, Sky32 was shown to interact with PCPꢀbound
amino acids in skyllamycin biosynthesis by in vivo and in vitro
experiments.59,60 The specificity of Sky32 was demonstrated to
be mediated by a carrier protein loaded with the respective
amino acid, which is thought to be the reason why this P450
hydroxylated specific amino acids. A skyllamycin derivative
without any βꢀhydroxylated amino acid residues was discovꢀ
ered from the sky32 inactivation mutant.59,60 No natural dehyꢀ
droxyꢀskyllamycin derivatives could be isolated from the wildꢀ
type producer strain.59 However, natural telomycins 9 and 12
exhibiting Leu instead of βꢀOHꢀLeu moieties could be isolated
from the wildꢀtype telomycin producer in low yields. Furtherꢀ
more, the natural telomycin derivative 10 with only the βꢀOHꢀ
Leu but without the other two hydroxylated amino acids has
also been discovered. Because of these natural dehydroxyꢀ
telomycins and the accumulation of 3 and 5 in the ꢁtem23
mutant, it is likely that the βꢀhydroxylation of leucine hapꢀ
pened after the NRPS assembly. However, there is another
possibility which implies that the A domain and its downꢀ
stream C domain have relaxed substrate specificity. Hence,
these telomycins exhibiting different hydroxylation patterns
might also be formed during NRPS assembly.
Figure 4. HPLC analysis (UV absorption at 320 nm) of various
substrates incubated with recombinant Tem25. A standard Tem25
assay was prepared in 100 ꢁl of 20 mM TrisꢀHCl (pH 7.4) buffer,
supplemented with 0.1 mM substrate and 0.2 ꢁM acylase. (A) 13
without Tem25; (B) 13 with inactivated Tem25; (C) 13 with
Tem25; (D) 14 without Tem25; (E) 14 with inactivated Tem25; (F)
14 with Tem25; (G) 15 with Tem25; (H) 15 with inactivated
Tem25; (I) 15 with Tem25.
Hydroxylation of Two Proline Residues by a P450
Monooxygenase and A Hydroxylase. Apart from βꢀOHꢀLeu,
telomycin contains two more hydroxylated amino acids (transꢀ
3ꢀOHꢀ7Pro and cisꢀ3ꢀOHꢀ11Pro). Gene tem29 was annotated as
encoding a cytochrome P450 dependent monooxygenase. The
putative product of tem29 shows 58% similarity to a P450
monooxygenase of unknown function from Streptomyces
eurythermus involved in angolamycin biosynthesis.61 It also
exhibits 55% similarity to a P450 monooxygenase which
might be responsible for the hydroxylation of the macrolide
ring of tylactone and tylosin produced by Streptomyces fradiꢀ
ae.62 To probe the function of tem29, the gene was inactivated
and the resulting mutant ꢁtem29 lost the ability to produce 1
and 2 and accumulated new products 4 and 6 (Chart 1; Figures
S13 and S20). Compound 4 showed almost the same retention
time and m/z as 2 and 3. However, the hydrolysis of 4 proꢀ
duced 18, MS/MS analysis of which proved that 4 differs from
2. The transꢀ3ꢀOHꢀ7Pro residue was replaced by Pro in 4,
whereas the cisꢀ3ꢀOHꢀ11Pro was substituted for Pro in 2.
Compound 6 was also hydrolyzed to give 19. This served to
prove that both transꢀ3ꢀOHꢀ7Pro and cisꢀ3ꢀOHꢀPro were subꢀ
stituted with Pro in 6. The transꢀ3ꢀOHꢀ7Pro moiety found in
compound 1 and 2 was substituted with Pro to result in strucꢀ
tures 4 and 6, respectively (Figure 5B). These data revealed
that Tem29 is the P450 monooxygenase responsible for the
transꢀ3ꢀhydroxylation of 7Pro.
To provide some insight on the substrate specificity of
Tem25, the semisynthetic lipopeptides 13, 14 and 15 have
been employed as substrates of Tem25 (Chart 2). Using
HPLCꢀESIꢀMS analysis it was found that Tem25 can hydroꢀ
lyze 13 and 14 containing nꢀheptanoyl and nꢀoctanoyl moieties
almost completely, similar to the hydrolysis of 7 and 8. Comꢀ
pound 15, which carries the nꢀdodecanoyl moiety, could only
be partially hydrolyzed under standard condition (Figure 4).
These assays employing different substrates including semꢀ
isynthetic telomycins provided some hints as to the substrate
specificity of Tem25. The enzyme shows a preference for
amide bond hydrolysis preferring the connection between a
fatty acid moiety and telomycin, since almost complete hyꢀ
drolysis is observed for 7, 8, 13 and 14. The fatty acid chain
length seems to be important, as roughly only 60% of comꢀ
pound 15 was hydrolyzed under standard conditions.
Hydroxylation of Leucine. The deduced product of tem23
is a characteristic cytochrome P450 dependent enzyme. Acꢀ
cording to Blast searches, the putative gene product of tem23
shows 78% similarity with Sky32, a cytochrome P450
monooxygenase from Streptomyces sp. Acta2897. The latter
enzyme is responsible for three βꢀhydroxylations to form βꢀ
OHꢀPhe, βꢀOHꢀOꢀMeꢀTyr and βꢀOHꢀLeu in skyllamycin
ACS Paragon Plus Environment

Page 11 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Characterization of telomycin derivatives. HPLCꢀMS/MS fragmentation analysis of (A) 16 and 17 formed by hydrolysis of 3
and 5, respectively, (B) 18 and 19 formed by hydrolysis of 4 and 6, respectively, and (C) 20 formed by hydrolysis of 2.
Tem29 showed low similarity to other P450 monooxygenꢀ
ases. Since most of these homologues were assumed to be
involved in polyketide biosynthesis,61,62 it is hard to predict the
timing of Tem29 activity by a comparison with homologues.
Due to the presence of telomycin derivatives 4 and 6 in mutant
ꢁtem29 and due to the natural intermediates 9ꢀ12 (Chart 1), in
which the transꢀ3ꢀOHꢀ7Pro was substituted by Pro, the hyꢀ
droxylation probably occurred post NRPS assembly. In addiꢀ
tion, the Tem29 homologue TpdJ2 was reported to be involved
in the biosynthesis of the thiopeptide GE2270. One hypothetic
function of TpdJ2 is the βꢀhydroxylation of phenylalanine.
TpdJ2 would thus use the ribosomally synthesized thiopeptide
core as substrate, which further implied Tem29 could perform
the hydroxylation after the telomycin peptide intermediate was
released.63 As discussed before, the hydroxylation might also
take place earlier due to relaxed substrate specificity of the
NRPS enzymes.
hydroxylases, which are iron (II)/2ꢀoxoglutarate (2ꢀOG)ꢀ
dependent oxygenases. Tem32 shows 48% similarity with a
Lꢀ
proline cisꢀ4ꢀhydroxylase from Mesorhizobium loti
MAFF30309964 and 47% similarity with a proline 3ꢀ
hydroxylase (2ꢀoxoglutarate dependent) from Streptomyces sp.
strain TH165. To verify the hypothesis that Tem32 plays a
similar role in the synthesis of cisꢀ3ꢀOHꢀ11Pro in 1, a tem32
mutant ꢁtem32 was constructed and probed for telomycin
production (Figure S15). The HPLCꢀESIꢀMS analysis reꢀ
vealed that the production of 1 was lost while 2 was still proꢀ
duced (Chart 1 and Figure S20). Compound 2 was subjected to
base hydrolysis to yield the corresponding linearized peptide
20. The MS/MS fragmentation pattern analysis of 20 revealed
that in 2, the cisꢀ3ꢀOHꢀ11Pro was replaced by Pro (Figure 5C).
Therefore, Tem32 is assumed to be a proline 3ꢀhydroxylase
responsible for the formation of cisꢀ3ꢀOHꢀ11Pro.
Due to the discovery of natural telomycin derivatives with
unꢀhydroxylated 11Pro (such as 9ꢀ10 and 12) in addition to 2
and the abolishment of the production of 1 after the inactivaꢀ
The deduced product of tem32 has a characteristic domain
of aspartyl/asparaginyl betaꢀhydroxylases including proline
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 23
tion of tem32, it is possible that Tem32 catalyzes the hydroxꢀ dependent alcohol dehydrogenaseꢀlike family was inactivated.
1
2
3
4
5
6
7
8
ylation after the NRPS assembly. However, the homologue of
Tem32 found in streptomycetes acts as proline 3ꢀhydroxylase
(with 2ꢀoxoglutarateꢀdependent dioxygenase properties), hyꢀ
However, telomycin production was abolished after deletion
of tem12 (Figures S8 and S21). Thus, whether tem12 is inꢀ
volved in the formation of ZꢀꢁTrp remains an open question.
A detailed discussion of the function of these two genes can be
found in supporting information.
droxylating free Lꢀproline to form cisꢀ3ꢀOHꢀL
ꢀproline.65 Hence,
the cisꢀ3ꢀhydroxylation in telomycin biosynthesis might occur
on free proline or after the NRPS assembly.
The gene tem15 encodes an anthranilate phosphoribosylꢀ
transferase (TrpD) which is involved in tryptophan biosyntheꢀ
sis, and therefore is speculated to be involved in the producꢀ
tion of tryptophan used in telomycin biosynthesis.
Genes Putatively Involved in Resistance, Transport,
Regulation, and Substrate Biosynthesis of Unknown Funcꢀ
tion. The three genes tem3, tem4, and tem5 located at the 5’
end of the TEM cluster putatively encode ABC transporter
subunits and might be involved in the telomycin export (selfꢀ
resistance) and biosynthesis. Genes tem30 and tem31 also
encode putative ABC transporters and were therefore inactiꢀ
vated simultaneously resulting in the mutant strain Streptomyꢀ
ces canus ATCC 12646 ꢁtem30ꢀ31 (Figure S14). In several
repeats of fermentation of the mutant strain Streptomyces
canus ATCC 12646 ꢁtem30ꢀ31, we never detected any teloꢀ
mycin in the supernatant of the culture broth (Figure S21).
However, we detected trace amounts of telomycins in the cells
of the fermentation twice but not every time. We hypothesize
that this unsteady and low amount of production in the cells is
caused by transcription adaption (suppressor mutation) since
the initially synthesized amount of telomycin cannot be exꢀ
ported and thus leads to cell death.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The deduced product of tem24 shows homology to the
MbtH family of proteins which are prevalent in NRPS biosynꢀ
thetic pathways. MbtH proteins are found to be required for
the efficient biosynthesis of many secondary metabolites deꢀ
pending on NRPSs.67ꢀ69
There are additional genes found in the TEM cluster which
cannot be assigned to any function in telomycin biosynthesis.
For instance, the gene tem10 encoding a YtcJꢀlike metal deꢀ
pendent amidohydrolase, tem16 encoding a prokaryotic phosꢀ
pholipase A2 and tem28 encoding a representative of a family
of bacterial ferritinꢀlike proteins.
CONCLUSIONS
Novel lipopeptides were identified as natural precursors in
the telomycin biosynthesis. Surprisingly, the precursor of
telomycin displayed enhanced activity against Gramꢀpositive
pathogens versus telomycin itself. Telomycins are particularly
important antibiotics as they most likely exhibit a novel mode
of action as demonstrated by a lack of crossꢀresistance with
several clinically important antibiotics. We could also demonꢀ
strate that telomycin 15 displays favorable early bactericidal
activity on methicillinꢀsensitive S. aureus (MSSA) and VRE.
Interestingly, various genes encoding transcription regulaꢀ
tors were discovered in the 5’ region of the TEM cluster which
suggested a sophisticated regulation system for telomycin
production. The gene tem7, which putatively encodes a sugar
kinase, belongs to the ROK family of transcription regulators.
The deduced product of tem8 belongs to the LacI transcripꢀ
tional regulator family. Since tem6 is located close to tem7,
and since its putative product is a mannoseꢀ6ꢀphosphate isoꢀ
meraseꢀlike protein involved in sugar metabolism, we assume
that Tem6 is possibly functionally connected to Tem7. Altꢀ
hough the function of the gene product of tem9 is still elusive,
this protein structurally belongs to the 6 hairpin glycosidase
superfamily, which suggests Tem9 might be functional in
conjunction with Tem8. The gene tem11 encodes a putative
LuxR transcriptional regulator, which is a DNAꢀbinding proꢀ
tein. Interestingly, the putative product of tem13 seems to
contain a characteristic AmiR and NasR transcription antiterꢀ
mination regulator (ANTAR) domain representing a RNAꢀ
binding domain found in bacterial transcription antiterminaꢀ
tion regulatory proteins.66 The gene tem14 was annotated to
encode a GntR family transcriptional regulator, which has a
similar Nꢀterminal DNA binding domain but a heterogeneous
Cꢀterminal effectorꢀbinding domain. In addition tem17 enꢀ
codes a protein containing a characteristic domain of the ArsR
subfamily of helixꢀturnꢀhelix bacterial transcriptional regulaꢀ
tors.
The deacylase TEM25 was characterized by in vivo and in
vitro experiments, explaining the biochemistry behind the
unusual maturation process of telomycins: a novel telomycin
lipopeptide precursor was purified from a tem25 mutant, strucꢀ
turally elucidated and the enzyme Tem25 was found to deacꢀ
ylate this precursor to eventually form telomycin. However,
the natural function of this transformation currently remains
unclear and it will be interesting to study why the more active
lipopeptide is formed initially and then deacylated.
The biosynthetic gene cluster for the biosynthesis of the
nonꢀribosomal peptide telomycin was cloned, sequenced and
heterologously expressed. A series of tem genes were inactiꢀ
vated in the wild type producer strain to identify the cluster
boundaries and the precursor maturation related genes as well
as to elucidate the function of enzymes involved in nonꢀ
proteinogenic amino acid biosynthesis, including the hydroxꢀ
ylation of three amino acid residues, the doubleꢀbond forꢀ
mation of the first tryptophan residue and the methylation of
the second tryptophan residue. From these mutants, seven new
telomycin biosynthetic intermediates were identified, which
not only helped to elucidate the respective enzyme`s functions
but also expanded the chemical diversity of the telomycins.
Additionally, five new telomycin intermediates isolated from
the natural producer S. canus ATCC 12646 also provided hints
to the timing of nonꢀproteinogenic amino acid formation and
contributed to preliminary structure activity relationships. This
study, describing the synthesis of lipidated derivatives of
telomycin displaying enhanced antibacterial activity, the idenꢀ
tification of the biosynthetic gene cluster with an unusual
The gene tem27 encodes a biꢀfunctional enzyme in which
the Nꢀterminal part is an aminotransferase and the Cꢀterminal
part is a methyltransferase. This gene might be involved in the
βꢀmethylation of tryptophan according to the reported case in
maremycins biosynthesis.46 However, the inactivation of
tem27 led to the abolishment of all telomycins (Figures S12
and S21). Whether this βꢀMeTrp is formed by Tem27 or other
enzymes could not be assigned yet. There is another Trp resiꢀ
due in telomycin which represents a Zꢀ2,3ꢀdehydrotryptophan
(ZꢀꢁTrp). tem12 containing a characteristic domain belonging
to the medium chain reductase/dehydrogenase (MDR)/zincꢀ
ACS Paragon Plus Environment

Page 13 of 23
Journal of the American Chemical Society
(17) Reimer, D.; Pos, K. M.; Thines, M.; Grun, P.; Bode, H. B.
Nat. Chem. Biol. 2011, 7 (12), 888ꢀ890.
(18) Bian, X. Y.; Fu, J.; Plaza, A.; Herrmann, J.; Pistorius, D.;
Stewart, A. F.; Zhang, Y. M.; Muller, R. Chembiochem 2013, 14
(10), 1194ꢀ1197.
maturation process, the identification of tailoring genes as well
1
2
3
as the establishment of a heterologous expressions system,
now forms the basis for the further pharmaceutical developꢀ
ment of the antibiotic class of telomycins.
4
5
(19) Brotherton, C. A.; Balskus, E. P. J. Am. Chem. Soc. 2013,
135 (9), 3359ꢀ3362.
ASSOCIATED CONTENT
(20) Vizcaino, M. I.; Engel, P.; Trautman, E.; Crawford, J. M.
J. Am. Chem. Soc. 2014, 136 (26), 9244ꢀ9247.
(21) Visca, P.; Imperi, F.; Lamont, I. L. Trends Microbiol.
2007, 15 (1), 22ꢀ30.
(22) Hannauer, M.; Schafer, M.; Hoegy, F.; Gizzi, P.; Wehrung,
P.; Mislin, G. L. A.; Budzikiewicz, H.; Schalk, I. J. FEBS Lett. 2012,
586 (1), 96ꢀ101.
(23) Schalk, I. J.; Guillon, L. Environ. Microbiol. 2013, 15 (6),
1661ꢀ1673.
(24) Drake, E. J.; Gulick, A. M. ACS Chem. Biol. 2011, 6 (11),
1277ꢀ1286.
(25) Masschelein, J.; Clauwers, C.; Awodi, U. R.; Stalmans, K.;
Vermaelen, W.; Lescrinier, E.; Aertsen, A.; Michiels, C.; Challis, G.
L.; Lavigne, R. Chem. Sci. 2015, 6 (2), 923ꢀ929.
(26) Xu, Y.; Kersten, R. D.; Nam, S. J.; Lu, L.; AlꢀSuwailem, A.
M.; Zheng, H. J.; Fenical, W.; Dorrestein, P. C.; Moore, B. S.; Qian,
P. Y. J. Am. Chem. Soc. 2012, 134 (20), 8625ꢀ8632.
(27) Koketsu, K.; Watanabe, K.; Suda, H.; Oguri, H.; Oikawa,
H. Nat. Chem. Biol. 2010, 6 (6), 408ꢀ410.
(28) Koketsu, K.; Minami, A.; Watanabe, K.; Oguri, H.;
Oikawa, H. Curr. Opin. Chem. Biol. 2012, 16 (1ꢀ2), 142ꢀ149.
(29) Zhang, Y. M.; Buchholz, F.; Muyrers, J. P. P.; Stewart, A.
F. Nat. Genet. 1998, 20 (2), 123ꢀ128.
(30) Zhang, Y. M.; Muyrers, J. P. P.; Testa, G.; Stewart, A. F.
Nat. Biotechnol. 2000, 18 (12), 1314ꢀ1317.
(31) Kieser, T.; Bibb, M. J.; Buttner, M. J.; Chater, K. F.;
Hopwood, D. A. Preparation and analysis of genomic and plasmid
DNA. In Practical Streptomyces Genetics, John Innes Foundation:
Norwich, England, 2000; pp 161ꢀ210.
(32) Gust, B.; Challis, G. L.; Fowler, K.; Kieser, T.; Chater, K.
F. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (4), 1541ꢀ1546.
(33) Kieser, T.; Bibb, M. J.; Buttner, M. J.; Chater, K. F.;
Hopwood, D. A. Introduction of DNA into Streptomyces. In Practical
Streptomyces Genetics, John Innes Foundation: Norwich, England,
2000; pp 229ꢀ252.
(34) Blin, K.; Medema, M. H.; Kazempour, D.; Fischbach, M.
A.; Breitling, R.; Takano, E.; Weber, T. Nucleic Acids Res. 2013, 41
(W1), W204ꢀW212.
(35) Rottig, M.; Medema, M. H.; Blin, K.; Weber, T.; Rausch,
C.; Kohlbacher, O. Nucleic Acids Res. 2011, 39, W362ꢀW367.
(36) Stachelhaus, T.; Mootz, H. D.; Marahiel, M. A. Chem.
Biol. 1999, 6 (8), 493ꢀ505.
(37) Petersen, T. N.; Brunak, S.; von, H. G.; Nielsen, H. Nat.
Methods 2011, 8 (10), 785ꢀ786.
(38) Moore, A. W.; Jorgenson, J. W. Anal. Chem. 1995, 67
(19), 3464ꢀ3475.
(39) Husain, R. D.; McCandless, J.; Stevenson, P. J.; Large, T.;
Guthrie, D. J. S.; Walker, B. J. Chromatogr. Sci. 2002, 40 (1), 1ꢀ6.
(40) Pierson, N. A.; Chen, L.; Russell, D. H.; Clemmer, D. E. J.
Am. Chem. Soc. 2013, 135 (8), 3186ꢀ3192.
6
7
8
9
Supporting Information. Supplemental experimental section,
primer sequences, design and verification of mutants, and NMR
spectra and assignment. This material is available free of charge
via the Internet at http://pubs.acs.org.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
AUTHOR INFORMATION
Corresponding Author
rolf.mueller@helmholtzꢀhzi.de
Present Addresses
#HelmholtzꢀZentrum für Infektionsforschung, Inhoffenstr. 7,
38124 Braunschweig, Germany
∇Technische Hochschule Mittelhessen, Fachbereich MNI,
Wiesenstr. 14, 35390 Gießen, Germany
ACKNOWLEDGMENT
In this research, Chengzhang Fu was supported by the Alexander
von Humboldt Foundation.
REFERENCES
(1)
GOUREVITCH, A.; HUNT, G. A.; MOSES, A. J.;
ZANGARI, V.; PUGLISI, T.; LEIN, J. Antibiot. Annu. 1957, 5, 856ꢀ
862.
(2)
MISIEK, M.; FARDIG, O. B.; GOUREVITCH, A.;
JOHNSON, D. L.; HOOPER, I. R.; LEIN, J. Antibiot. Annu. 1957, 5,
852ꢀ855.
(3)
TISCH, D. E.; HUFTALEN, J. B.; DICKISON, H. L.
Antibiot. Annu. 1957, 5, 863ꢀ868.
(4)
Sheehan, J. C.; Stock, J. A.; Maeda, K.; Nakamura, S.;
Drummond, P. E.; Mania, D.; Gardner, J. N.; Sen, A. K. J. Am.
Chem. Soc. 1963, 85 (18), 2867ꢀ&.
(5)
Maeda, K. J. Am. Chem. Soc. 1968, 90 (2), 462ꢀ&.
(6) Kumar, N. G.; Urry, D. W. Biochemistry 1973, 12 (20),
3811ꢀ3817.
Sheehan, J. C.; Mania, D.; Nakamura, S.; Stock, J. A.;
(7)
Kumar, N. G.; Urry, D. W. Biochemistry 1973, 12 (22),
4392ꢀ4399.
(8)
Mori, H.; Shibasaki, T.; Uozaki, Y.; Ochiai, K.; Ozaki, A.
Appl. Environ. Microbiol. 1996, 62 (6), 1903ꢀ1907.
(9)
Gurovic, M. S. V.; Muller, S.; Domin, N.; Seccareccia, I.;
Nietzsche, S.; Martin, K.; Nett, M. Int. J. Syst. Evol. Microbiol. 2013,
63, 3812ꢀ3817.
(10) Oliva, B.; Maiese, W. M.; Greenstein, M.; Borders, D. B.;
Chopra, I. J. Antimicrob. Chemother. 1993, 32 (6), 817ꢀ830.
(11) Boucher, H. W.; Talbot, G. H.; Bradley, J. S.; Edwards, J.
E.; Gilbert, D.; Rice, L. B.; Scheld, M.; Spellberg, B.; Bartlett, J.
Clin. Infect. Dis. 2009, 48 (1), 1ꢀ12.
(12) Wright, G. D. Can. J. Microbiol. 2014, 60 (3), 147ꢀ154.
(13) Kirst, H. A. Expert Opin. Drug Discovery 2013, 8 (5),
479ꢀ493.
(14) Koch, G.; Yepes, A.; Forstner, K. U.; Wermser, C.;
Stengel, S. T.; Modamio, J.; Ohlsen, K.; Foster, K. R.; Lopez, D. Cell
2014, 158 (5), 1060ꢀ1071.
(15) Bionda, N.; Pitteloud, J. P.; Cudic, P. Future. Med. Chem.
2013, 5 (11), 1311ꢀ1330.
(41) Wenzel, S. C.; Meiser, P.; Binz, T. M.; Mahmud, T.;
Muller, R. Angew. Chem. Int. Ed Engl. 2006, 45 (14), 2296ꢀ2301.
(42) Hojati, Z.; Milne, C.; Harvey, B.; Gordon, L.; Borg, M.;
Flett, F.; Wilkinson, B.; Sidebottom, P. J.; Rudd, B. A. M.; Hayes, M.
A.; Smith, C. P.; Micklefield, J. Chem. Biol. 2002, 9 (11), 1175ꢀ1187.
(43) ShawꢀReid, C. A.; Kelleher, N. L.; Losey, H. C.; Gehring,
A. M.; Berg, C.; Walsh, C. T. Chem. Biol. 1999, 6 (6), 385ꢀ400.
(44) Hoffmann, D.; Hevel, J. M.; Moore, R. E.; Moore, B. S.
Gene 2003, 311, 171ꢀ180.
(45) Du, L. C.; Sanchez, C.; Chen, M.; Edwards, D. J.; Shen, B.
Chem. Biol. 2000, 7 (8), 623ꢀ642.
(46) Zou, Y.; Fang, Q.; Yin, H. X.; Liang, Z. T.; Kong, D. K.;
Bai, L. Q.; Deng, Z. X.; Lin, S. J. Angew. Chem. , Int. Ed. 2013, 52
(49), 12951ꢀ12955.
(16) Fischbach, M. A.; Walsh, C. T. Chem. Rev. 2006, 106 (8),
3468ꢀ3496.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 23
(47) Stindl, A.; Keller, U. J. Biol. Chem. 1993, 268 (14),
10612ꢀ10620.
1
(48) May, J. J.; Wendrich, T. M.; Marahiel, M. A. J. Biol.
Chem. 2001, 276 (10), 7209ꢀ7217.
(49) Heinzelmann, E.; Berger, S.; Muller, C.; Hartner, T.;
Poralla, K.; Wohlleben, W.; Schwartz, D. Microbiology 2005, 151,
1963ꢀ1974.
2
3
4
5
(50) Wang, Y.; Chen, Y.; Shen, Q. R.; Yin, X. H. Gene 2011,
483 (1ꢀ2), 11ꢀ21.
6
7
(51) Muller, C.; Nolden, S.; Gebhardt, P.; Heinzelmann, E.;
Lange, C.; Puk, O.; Welzel, K.; Wohlleben, W.; Schwartz, D.
Antimicrob. Agents Chemother. 2007, 51 (3), 1028ꢀ1037.
(52) Yamanaka, K.; Reynolds, K. A.; Kersten, R. D.; Ryan, K.
S.; Gonzalez, D. J.; Nizet, V.; Dorrestein, P. C.; Moore, B. S. Proc.
Natl. Acad. Sci. U. S. A. 2014, 111 (5), 1957ꢀ1962.
(53) Li, T. L.; Huang, F. L.; Haydock, S. F.; Mironenko, T.;
Leadlay, P. F.; Spencer, J. B. Chem. Biol. 2004, 11 (1), 107ꢀ119.
(54) Oinonen, C.; Rouvinen, J. Protein Sci. 2000, 9 (12), 2329ꢀ
2337.
(55) Ueda, S.; Shibata, T.; Ito, K.; Oohata, N.; Yamashita, M.;
Hino, M.; Yamada, M.; Isogai, Y.; Hashimoto, S. J. Antibiot. 2011,
64 (2), 169ꢀ175.
(56) Davies, J. J. Ind. Microbiol. Biotechnol. 2006, 33 (7), 496ꢀ
499.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(57) Yim, G.; Wang, H. H. M.; Davies, J. Int. J. Med.
Microbiol. 2006, 296 (2ꢀ3), 163ꢀ170.
(58) Kall, L.; Krogh, A.; Sonnhammer, E. L. Nucleic Acids Res.
2007, 35 (Web Server issue), W429ꢀW432.
(59) Pohle, S.; Appelt, C.; Roux, M.; Fiedler, H. P.; Sussmuth,
R. D. J. Am. Chem. Soc. 2011, 133 (16), 6194ꢀ6205.
(60) Uhlmann, S.; Sussmuth, R. D.; Cryle, M. J. ACS Chem.
Biol. 2013, 8 (11), 2586ꢀ2596.
(61) Schell, U.; Haydock, S. F.; Kaja, A. L.; Carletti, I.; Lill, R.
E.; Read, E.; Sheehan, L. S.; Low, L.; Fernandez, M. J.; Grolle, F.;
McArthur, H. A. I.; Sheridan, R. M.; Leadlay, P. F.; Wilkinson, B.;
Gaisser, S. Org. Biomol. Chem. 2008, 6 (18), 3315ꢀ3327.
(62) Mersondavies, L. A.; Cundliffe, E. Mol. Microbiol. 1994,
13 (2), 349ꢀ355.
(63) Morris, R. P.; Leeds, J. A.; Naegeli, H. U.; Oberer, L.;
Memmert, K.; Weber, E.; LaMarche, M. J.; Parker, C. N.; Burrer, N.;
Esterow, S.; Hein, A. E.; Schmitt, E. K.; Krastel, P. J. Am. Chem.
Soc. 2009, 131 (16), 5946ꢀ5955.
(64) Koketsu, K.; Shomura, Y.; Moriwaki, K.; Hayashi, M.;
Mitsuhashi, S.; Hara, R.; Kino, K.; Higuchi, Y. ACS Synth. Biol.
2015, 4 (4), 383ꢀ392.
(65) Mori, H.; Shibasaki, T.; Yano, K.; Ozaki, A. J. Bacteriol.
1997, 179 (18), 5677ꢀ5683.
(66) Shu, C. J.; Zhulin, I. B. Trends Biochem. Sci. 2002, 27 (1),
3ꢀ5.
(67) Davidsen, J. M.; Bartley, D. M.; Townsend, C. A. J. Am.
Chem. Soc. 2013, 135 (5), 1749ꢀ1759.
(68) Lautru, S.; OvesꢀCostales, D.; Pernodet, J. L.; Challis, G.
L. Microbiology 2007, 153, 1405ꢀ1412.
(69) Wolpert, M.; Gust, B.; Kammerer, B.; Heide, L.
Microbiology 2007, 153, 1413ꢀ1423.
ACS Paragon Plus Environment

Page 15 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
15
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chart 1. Structures of telomycins.
69x34mm (600 x 600 DPI)
ACS Paragon Plus Environment

Page 17 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chart 2. Chemical structures of semisynthetic telomycin analogues (7, 13-15).
55x54mm (600 x 600 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 18 of 23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. The gene organization and restriction map of the TEM gene cluster as well as the proposed
biosynthetic pathway of telomycin. (A) Restriction map of the telomycin biosynthetic gene cluster. The
dashed blue line indicates the gap containing highly repetitive DNA sequence. The restriction map of the
whole DNA region covered by cosmids and BACs is found in the Supporting Information. (B) Tem18 and
Tem19 initially activate a 6ꢀmethylheptanoate and transfer it to the carrier protein where Tem20, Tem21
and Tem22 build up the peptide chain. The three hydroxylation reactions performed by Tem23, Tem29 and
Tem32 most likely take place postꢀNRPS assembly. The complex βꢀmethylation of tryptophan is probably
achieved prior to the activation by the respective A domain. After the mature lipopeptide is released and
cyclized, the 6ꢀmethylheptanoyl moiety is cleaved off by Tem25 to yield telomycin.
109x130mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 19 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. HPLC-MS analysis of telomycins in crude extracts (small-scale fermentation) of the wild-type and
tem25 gene-inactivated mutant strain of Streptomyces canus ATCC 12646. Extracted ion chromatograms
(EIC) of telomycins (1, [M+H]+=1272.6, green, 2, [M+H]+=1256.6, blue, 7, [M+H]+=1398.7, red, and 8,
[M+H]+=1382.7, violet) were analyzed for both mutant and wild-type strains to test production.
Compounds found in the wild-type and mutant strains are identified by numbers and their structures as
shown in Chart 1.
53x51mm (300 x 300 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 20 of 23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. SDSꢀPAGE analysis and in vitro characterization of Tem25. A standard Tem25 assay was prepared
in 100 ꢁl of 20 mM TrisꢀHCl (pH 7.4) buffer, supplemented with 0.1 mM substrate and 0.2 ꢁM acylase. (A)
The recombinant enzyme was purified as described in Materials and Methods. The α and β subunit are
indicated by arrows. M, protein marker; lane 1, uninduced cells; lane 2, induced cells; lane 3, lysate
supernatant of induced cells; lane 4, semiꢀpurified Tem25. (B) HPLC analysis of (a) 7 without adding
Tem25; (b) boiled Tem25 incubated with 7; (C) Tem25 incubated with 7.
14x21mm (600 x 600 DPI)
ACS Paragon Plus Environment

Page 21 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. HPLC analysis (UV absorption at 320 nm) of various substrates incubated with recombinant
Tem25. A standard Tem25 assay was prepared in 100 µl of 20 mM Tris-HCl (pH 7.4) buffer, supplemented
with 0.1 mM substrate and 0.2 µM acylase. (A) 13 without Tem25; (B) 13 with inactivated Tem25; (C) 13
with Tem25; (D) 14 without Tem25; (E) 14 with inactivated Tem25; (F) 14 with Tem25; (G) 15 with
Tem25; (H) 15 with inactivated Tem25; (I) 15 with Tem25.
17x25mm (600 x 600 DPI)
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 22 of 23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Characterization of telomycin derivatives. HPLC-MS/MS fragmentation analysis of (A) 16 and 17
formed by hydrolysis of 3 and 5, respectively, (B) 18 and 19 formed by hydrolysis of 4 and 6, respectively,
and (C) 20 formed by hydrolysis of 2.
110x140mm (300 x 300 DPI)
ACS Paragon Plus Environment

Page 23 of 23
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
TOC graphic
105x44mm (300 x 300 DPI)
ACS Paragon Plus Environment