Selected Mutations Reveal New Intermediates in the Biosynthesis of Mupirocin and the Thiomarinol Antibiotics

Article abstract of DOI:10.1002/anie.201611590

Thiomarinol and mupirocin are assembled on similar polyketide/fatty acid backbones and exhibit potent antibiotic activity against methicillin-resistant Staphylococcus aureus (MRSA). They both contain a tetrasubstituted tetrahydropyran (THP) ring that is essential for biological activity. Mupirocin is a mixture of pseudomonic acids (PAs). Isolation of the novel compound mupirocin P, which contains a 7-hydroxy-6-keto-substituted THP, from a ΔmupP strain and chemical complementation experiments confirm that the first step in the conversion of PA-B into the major product PA-A is oxidation at the C6 position. In addition, nine novel thiomarinol (TM) derivatives with different oxidation patterns decorating the central THP core were isolated after gene deletion (tmlF). These metabolites are in accord with the THP ring formation and elaboration in thiomarinol following a similar order to that found in mupirocin biosynthesis, despite the lack of some of the equivalent genes. Novel mupirocin–thiomarinol hybrids were also synthesized by mutasynthesis.

Full text of DOI:10.1002/anie.201611590

Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201611590
German Edition: DOI: 10.1002/ange.201611590
Synthetic Biology Very Important Paper
Selected Mutations Reveal New Intermediates in the Biosynthesis of
Mupirocin and the Thiomarinol Antibiotics
Shu-Shan Gao+, Luoyi Wang+, Zhongshu Song, Joanne Hothersall, Elton R. Stevens,
Jack Connolly, Peter J. Winn, Russell J. Cox, Matthew P. Crump, Paul R. Race,
Christopher M. Thomas, Thomas J. Simpson,* and Christine L. Willis*
Abstract: Thiomarinol and mupirocin are assembled on
similar polyketide/fatty acid backbones and exhibit potent
antibiotic activity against methicillin-resistant Staphylococcus
aureus (MRSA). They both contain a tetrasubstituted tetrahy-
dropyran (THP) ring that is essential for biological activity.
Mupirocin is a mixture of pseudomonic acids (PAs). Isolation
of the novel compound mupirocin P, which contains a 7-
hydroxy-6-keto-substituted THP, from a DmupP strain and
chemical complementation experiments confirm that the first
step in the conversion of PA-B into the major product PA-A is
oxidation at the C6 position. In addition, nine novel thiomar-
inol (TM) derivatives with different oxidation patterns deco-
rating the central THP core were isolated after gene deletion
(tmlF). These metabolites are in accord with the THP ring
Figure 1. Selected thiomarinols and pseudomonic acids.
formation and elaboration in thiomarinol following a similar
order to that found in mupirocin biosynthesis, despite the lack
of some of the equivalent genes. Novel mupirocin–thiomarinol
hybrids were also synthesized by mutasynthesis.
the trans-AT class of modular polyketide-derived antibiot-
ics.[4] The mup biosynthetic gene cluster was one of the first of
this class to be discovered,[5] but there are now more than 20
that have been at least partially characterized, with many
others known from genomic sequences.[4] Thiomarinols can be
regarded as comprising three main elements: a highly func-
tionalized polyketide-derived acid that is esterified by 8-
hydroxyoctanoic acid, which itself forms an amide with the
bicyclic amino acid derived pyrrothine. The major compo-
nents TM-A (1) and PA-A (7) differ in the lack of the C10/
C11 epoxide, the presence of a 4-hydroxy group, a C8 rather
than a C9 fatty acid, and the pyrrothine in TM-A.[6]
The 6,7-dihydroxytetrahydropyran (THP) ring is neces-
sary for the biological activity of mupirocin and thiomarinol
so elucidation of the biosynthetic mechanisms for ring
formation and further modification to the 6,7-diol is impor-
tant. Mutational analysis of mupirocin biosynthesis showed
somewhat counterintuitively a) that the 8-hydroxy group of
PA-B (8) has to be lost not gained, with feeding experiments
confirming that PA-B is an intermediate in the biosynthesis of
PA-A (7),[7] and b) that PA-C (9) is the product of a parallel
pathway that branches from the main pathway (to PA-A)
following failure to undergo epoxidation of the C10/C11
double bond.[8] Analyses of mutant strains[9] indicate that the
putative biosynthetic precursor 12 (Scheme 1) is cyclized and
esterified to give PA-B with the tetrahydropyran ring. In
addition, the identification of minor metabolites with C5 and
C7 fatty acid side chains is in accord with elaboration of the 9-
hydroxynonanoic acid moiety occurring by initial addition of
3-hydroxypropionate followed by three successive MmpB-
mediated chain elongations.[7b]
T
hiomarinols (1–6) are an unusual group of hybrid anti-
biotics produced by the marine bacterium Pseudoalteromonas
sp. SANK73390 (Figure 1).[1] They combine close analogues
of the clinically important agent mupirocin (a mixture of
pseudomonic acids (PAs) A–C, 7–9)[2] with the pyrrothine
subunit of the holomycin class of antibiotics. Both thiomar-
inols and mupirocin display potent activity against MRSA
(methicillin-resistant Staphylococcus aureus),[3] and belong to
[*] Dr. S.-S. Gao,[+] Dr. L. Wang,[+] Dr. Z. Song, Prof. R. J. Cox,
Prof. M. P. Crump, Prof. T. J. Simpson, Prof. C. L. Willis
School of Chemistry, University of Bristol
Bristol, BS8 1TS (UK)
E-mail: tom.simpson@bristol.ac.uk
chris.willis@bristol.ac.uk
Dr. P. R. Race
School of Biochemistry
University of Bristol
Bristol BS8 1TD (UK)
Prof. R. J. Cox
BMWZ, Leibniz-Universitꢀt Hannover
Schneiderberg 38, 30167 Hannover (Germany)
Dr. J. Hothersall, E. R. Stevens, J. Connolly, Dr. P. J. Winn,
Prof. C. M. Thomas
School of Biosciences, University of Birmingham
Edgbaston, Birmingham, B15 2TT (UK)
[+] These authors contributed equally to this work.
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
http://dx.doi.org/10.1002/anie.201611590.
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Angewandte
Communications
Chemie
Figure 2. HPLC trace of the crude extract of the mmpEDOR/DmupP
double-mutant strain of P. fluorescens NCIMB 10586.
parts of the pathway, thus confirming their roles as inter-
mediates in mupirocin biosynthesis. To investigate whether
the 7-hydroxy-6-ketone is also a true intermediate on the
pathway, desepoxymupirocin P (16), isolated from the
mmpEDOR/DmupP strain, was fed to cultures of the
mmpEDOR/DmupW strain, and LC-MS analysis showed
a peak with the same retention time as desepoxymupirocin P
but with Mw = 484. The metabolite was isolated, and NMR
spectroscopy revealed that PA-C had been produced, con-
firming that desepoxymupirocin P is indeed an intermediate
in PA-C (9) biosynthesis (Scheme 2).
Scheme 1. MupP acts as a dehydratase in mupirocin biosynthesis.
Conversion of PA-B (8) into PA-A (7; replacing 8-OH
with 8-H) has been shown to require MupU (CoA ligase) and
MacpE (ACP), MupO (P450), MupV (see below), MupC
(enoate reductase), and MupF (ketoreductase) via the
isolated intermediates mupirocin C1 (14) and mupirocin F1
(15; Scheme 1). Deletion of any of the mupO, mupU, mupV,
or macpE genes gave PA-B (8) as the major product along
with macrolactone 10.[8] The first step in the conversion of PA-
B (8) into PA-A (7) is MupU-catalyzed ligation of PA-B to
MacpE (Scheme 1) followed by oxidation (MupO). However,
no metabolite corresponding to the putative a-hydroxyketone
product 13 of MupO has been detected. Isolation and
identification of these epoxide-containing metabolites is
challenging as they are prone to rearrangements but they
are key to elucidating the remaining biosynthetic steps from
PA-B to PA-A. Indeed, PA-A itself is limited to mainly topical
applications owing to its instability in serum.[3]
Our first goal was to establish the proposed 7-hydroxy-6-
ketone 13 as a biosynthetic intermediate to PA-A. Analysis of
a DmupP strain revealed an unstable metabolite with Mw =
514, and detailed 1D and 2D NMR analysis confirmed the
structure of this key intermediate, which was named mupir-
ocin P (13).
We have shown previously that the OR domain of MmpE
is responsible for epoxidation.[8] Thus a double mutant, the
mmpEDOR/DmupP strain, was constructed. Gratifyingly, this
produces the less labile desepoxymupirocin P (16) as the main
metabolite along with a small amount of desepoxy-PA-B (17;
Figure 2). MupP had not been previously assigned a function.
Bioinformatic analysis indicates that it belongs to the
glyoxalase/bleomycin resistance protein/dioxygenase super-
family while it here appears to act overall as a dehydratase to
give enol ketone 14 (Scheme 1).
Scheme 2. Feeding 16 to either mmpEDOR/DmupW or DmupV strains
gives PA-C (9).
Reanalysis of the MupV amino acid sequence, which had
previously been identified as a putative oxidoreductase,
revealed a second domain with about 30% sequence identity
to many Pseudomonas a/b hydrolases. These contain
a GXCXG consensus sequence where cysteine replaces the
more common nucleophilic serine and include an amino
transferase involved in shuffling intermediates between
thiolation domains in the syringomycin biosynthetic pathway
in Pseudomonas syringae.[10] Thus MupV may act as a thio-
esterase or perhaps in some related ACP transfer capacity.
Previously, deletion of mupV had provided few clues as this
yielded PA-B (8) as the major product.[9c] However, in this
study, feeding desepoxymupirocin P (16) to the DmupV strain
gave PA-C in accord with a Mup V activity being required
before MupP leading to mupirocin P (13, and desepoxymu-
pirocin P (16); Schemes 1 and 2).
The thiomarinol (tml) biosynthetic gene cluster was
identified by complete genome sequencing of SANK73390,
which was found to harbor a 97 kb plasmid consisting almost
entirely of the thiomarinol biosynthetic genes.[11] These
include multimodular trans-AT polyketide synthase (PKS)
genes (tmpA, tmpC, and tmpD), putative fatty acid synthase
(FAS) genes, tmpB, and associated tailoring and resistance
genes (tmlA–tmlZ), most of which exhibit high homology to
counterparts in the mupirocin (mup) cluster (see the Sup-
porting Information). A non-ribosomal peptide synthetase
(NRPS) linked to a set of tailoring enzymes (holA–holH)
Previous mutasynthesis studies on mupirocin biosynthe-
sis[8] showed that PA-B (8), mupirocin C1 (14), and mupir-
ocin F1 (15) were all efficiently (15–50%) converted into PA-
A (7) by mutant strains of P. fluorescens blocked in earlier
2
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!

Angewandte
Communications
Chemie
similar to that recently shown to control holomycin biosyn-
mupirocin,[8,9] but reverse genetics of SANK73390 proved to
thesis in Streptomyces clavuligerus is also present.[12] We
produced mutant strains of SANK73390 in which the PKS and
NRPS parts of the cluster had been insertionally deactivated.
Analysis of WT, NRPS, and tmlU mutant strains led to the
isolation of marinolic acids A (18), A6 (20), and A4 (21)
lacking the pyrrothine (Figure 3).[13] Using a combination of
be problematic. We were however successful in making the
key DtmlF by the method previously reported[11] using
approximately 500 bp arms defining an in-frame deletion
from amino acid 10 to amino acid 327 in TmlF.
Analysis of extracts showed that, as anticipated, the
production of thiomarinols A and C (1 and 2, Figure 1) was
completely abolished, but the xenorhabdin acyl pyrrothine
metabolites were generated as normal. Further analysis of the
HPLC trace indicated that various new thiomarinol-related
metabolites were present albeit in minor amounts of 0.1–
1.2 mgLÀ1 (Figure 4). A major product among those minor
Figure 3. Marinolic acids and acyl pyrrothine metabolites isolated from
WT and mutant strains of Pseudoalteromonas sp. SANK73390.
genetic and isotopic labeling studies, we showed that the 8-
hydroxyoctanoic acid side chain is generated via successive
chain extensions of a C4 precursor (4-hydroxybutyrate)
derived from succinate to give marinolic acids 18, 20, and 21
and that pyrrothine is assembled from two molecules of
cysteine (HolA–HolD) prior to intact incorporation into
thiomarinol.[13] It has recently been shown that TmlU acts as
a substrate-selective CoA ligase that activates marinolic acid
as a thioester, which in turn is the substrate for amide
formation catalyzed by the acyl transferase HolE.[14] Analysis
of the PKS mutant revealed a series of acyl pyrrothines
designated as xenorhabdins.[13]
Marinolic acid A (18) has a similar structure to PA-A (7)
but lacks the epoxide and has an additional 4-hydroxy group.
The tml pathway, however, has only some of the homologues
of the mup genes, namely tmlW, T, O, P, C, and F, but
homologues of the standalone ACPs macpE, mupU, and
mupV are missing. The absence of these genes from the tml
cluster raises the question of how the same overall conversion
can be achieved with an apparently simpler gene set. We now
report results that indicate that a similar set of intermediates
are indeed involved in THP ring formation and further
modification in both mupirocin and thiomarinol biosynthesis.
Deletion of mupF (7-ketoreductase) in P. fluorescens
releases a number of intermediates including mupirocin C1
(14) and PA-B (8) and related shunt products in addition to
mupirocin F1 (15; Scheme 1).[9b] TmlF shows 33% amino acid
sequence identity to MupF so it is likely to act as a 7-
ketoreductase as does MupF in the final step in mupirocin
biosynthesis, which converts mupirocin F1 (15) into the
dihydroxytetrahydropyran PA-A (7). Thus deletion of tmlF
could release more intermediates and shunt products aside
from its own substrate analogous to the mupF deletion.
WT SANK73390 was cultivated on a modified marine
broth medium (see the Supporting Information for details),
which significantly improved the production of the total
thiomarinols. Using this medium, TM-A (1) could be
obtained in a yield of 70–100 mgLÀ1, compared to 10–
20 mgLÀ1 in previous studies.[6] Our aim was to carry out
a set of gene KOs to parallel those carried out with
Figure 4. Novel metabolites isolated from extracts of the DtmlF
mutant of Pseudoalteromonas sp. SANK73390.
metabolites was 8-hydroxythiomarinol C (22). Its structure
and that of other metabolites reported herein were deter-
mined by full NMR analysis and high-resolution ESI-MS (see
the Supporting Information for details). The presence of an 8-
hydroxy-substituted THP ring as in PA-B (8) suggests that
both biosynthetic pathways feature analogous pyran ring-
closing mechanisms, and will require similar mechanisms for
removal of the 8-hydroxy function. The only previous
thiomarinol derivative reported to contain an 8-hydroxy
group is TM-G (6, as a very minor metabolite of WT
SANK73390),[1c] but it lacks both the 4 and 6-hydroxy groups.
We have not observed 6 in any of our studies, and its structure
remains to be confirmed. Further detailed analysis of minor
components resulted in the isolation of metabolites 23–30.
These metabolites form two structurally distinct groups. The
first consists of 6,7-diketothiomarinol C (23), analogous to
mupirocin C1 (14) isolated from DmupC, its 4-hydroxy
analogue 24, and derivative 25, a 6,7-diketo analogue of
marinolic acid A6 lacking the 4-hydroxy moiety and the
pyrrothine. These all exist as the enol ketone tautomers
shown. The second group features the 6-hydroxy-7-keto THP
ring (7-ketothiomarinol C (26) and its 4-hydroxy analogue 27)
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
similar to mupirocin F1 (15) isolated from DmupF, and
products 28–30 derived from 26 and 27 by keto–enol
tautomerism and/or epimerization. This is consistent with
the biosynthetic sequence to thiomarinol C (2) being similar
to that towards mupirocin after the formation of the 8-
hydroxy-THP ring. Interestingly, the 4-hydroxy analogue of 8-
hydroxythiomarinol C was not detected.
When PA-A (7) was fed to the PKS mutant strain of
SANK73390 in which all of the tailoring genes are intact, it
was metabolized with both 4-hydroxylation and/or pyrrothine
addition taking place to give the novel mupirocin/thiomarinol
hybrid molecules 31–33 (Scheme 3).[6,13] In this study, we have
investigated biotransformations of further substrates isolated
from P. fluorescens, and the products were isolated and fully
characterized by spectroscopic methods.
It is interesting to compare the metabolism of 8-hydroxy-
THPs in P. fluorescens and Pseudoalteromonas. In the former,
desepoxy-PA-B (17) and PA-B (8) undergo efficient oxida-
tion, dehydration, and keto reduction to give the bioactive
6,7-diols PA-C (9) and PA-A (7). In contrast, in the PKS
mutant of Pseudoalteromonas, whilst further processing of the
6,8-dihydroxy-7-keto-THP ring of desepoxymupirocin P (16)
to the mature 6,7-dihydroxy-THP as well as pyrrothine
addition readily occur, upon feeding the 6,7,8-trihydroxy-
THP substrates 8 or 17, only pyrrothine addition was
apparent.
In conclusion, MupP has been identified as a dehydratase
involved in mupirocin biosynthesis. The DmupP strain of
P. fluorescens was cultured, and the novel 6-keto-7-hydroxy-
substituted mupirocin P (13) was isolated and fully charac-
terized. Results of chemical complementation studies feeding
desepoxymupirocin P (16) to the mmpEDOR/DmupW strain
blocked earlier in the biosynthetic pathway are in accord with
16 being an intermediate in the biosynthesis of the bioactive
6,7-diol PA-C (9) and the epoxy analogue mupirocin P (13)
was deduced to be a biosynthetic precursor of PA-A (7).
Hence the first step in the conversion of PA-B (8) into PA-A
(7) is oxidation at the C6 position. Furthermore, feeding 16 to
the DmupV strain gave PA-C in accord with MupV acting
before MupP.
Despite some divergences in the gene clusters, the
isolation of novel metabolites including 22, 23, and 26 from
cultures of the DtmlF mutant of SANK73390 point to some
similarities in THP ring processing in both thiomarinol and
mupirocin. Indeed, feeding desepoxymupirocin P (16) to
cultures of the PKS mutant not only led to addition of the
pyrrothine but also confirmed that further processing of the
THP ring occurred to give the 6,7-dihydroxylated products 36
and 37. In contrast, feeding desepoxy-PA-B (17) to the PKS
mutant strain simply led to addition of the pyrrothine, and no
THP-modified products were detected. PA-B (8) is efficiently
converted into PA-A (7) by mutant strains of P. fluorescens
blocked earlier in the biosynthetic pathway, indicating an
important difference between the two pathways.
Scheme 3. Transformation of mupirocin analogues by the PKS mutant
strain of Pseudoalteromonas SANK73390.
Differences are evident in the two gene clusters. Homo-
logues of the mupirocin standalone ACP MacpE, along with
MupU and MupV, are missing in SANK73390. TmpB,
however, contains extra KS and ACP domains compared to
MmpB, which is speculated to replace the apparently missing
activities.[13] Thus the tetrahydropyran-processing reactions
occurring on intermediates bound to MacpE may take place
while the intermediates are bound to the additional ACP on
TmpB, thus obviating the need for homologues of MacpE and
MupU. The lack of ring processing of PA-B (8) and desepoxy-
PA-B (17) fed to the PKS mutant of Pseudoalteromonas thus
may be rationalized by the inability of these substrates to be
loaded onto TmpB, providing key evidence to support our
proposed pathway to the marinolic acids. Structure–activity
studies as well as investigations to fully understand the
differences between the two pathways and to identify the
genes for 4- and 6-hydroxylation in thiomarinol and 6-
hydroxylation in mupirocin biosynthesis are in progress.
On feeding PA-C (9), with the mature 6,7-dihydroxypyran
ring, the pyrrothine was added to give the novel metabolite 36
(Mw = 638), and 4-hydroxylation yielded 37 (Mw = 654) in
similar amounts (Scheme 3). In contrast, feeding PA-B (8) or
desepoxy-PA-B (17) simply led to pyrrothine addition (34 and
35, respectively), and no C4-hydroxylated metabolites were
detected. Hence it is apparent that the presence of an 8-
hydroxy group inhibits 4-hydroxylation.
Desepoxymupirocin P (16) was also fed to cultures of the
DPKS mutant giving the three metabolites 35, 36, and 37, all
of which had the pyrrothine added. Importantly, it was
evident that further processing of the THP ring had occurred
leading to loss of the 8-hydroxy group and keto reduction
giving the fully mature 6,7-dihydroxylated product 36;
a minor product detected by LC-MS was the corresponding
4-hydroxylated metabolite 37. The 6,7,8-trihydroxy metabo-
lite 35 was formed by keto reduction and pyrrothine addition,
but no loss of the 8-hydroxy group was observed.
4
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!

Angewandte
Communications
Chemie
[6] A. C. Murphy, D. Fukuda, Z. Song, J. Hothersall, R. J. Cox, C. L.
Acknowledgements
Willis, C. M. Thomas, T. J. Simpson, Angew. Chem. Int. Ed. 2011,
50, 3271 – 3274; Angew. Chem. 2011, 123, 3329 – 3332.
[7] a) P. G. Mantle, M. De Langen, V. K. Teo, J. Antibiot. 2001, 54,
166 – 174; b) J. Hothersall, J. e. Wu, A. S. Raham, J. A. Shields, J.
Haddock, N. Johnson, S. M. Cooper, E. R. Stephens, R. J. Cox, J.
Crosby, C. L. Willis, T. J. Simpson, C. M. Thomas, J. Biol. Chem.
2007, 282, 15451 – 15461.
This work was funded by the BBSRC and EPSRC including
through BrisSynBio, the Bristol Centre for Synthetic Biology
(BB/L01386X/1), awards BB/I014039/1, BB/I014373/1, and
a BBSRC DTP studentship (BB/J014532/1). We thank the
EPSRC (EP/F066104/1) for LCMS equipment.
[8] S.-S. Gao, J. Hothersall, J. Wu, A. C. Murphy, Z. Song, E. R.
Stephens, C. M. Thomas, M. P. Crump, R. J. Cox, T. J. Simpson,
C. L. Willis, J. Am. Chem. Soc. 2014, 136, 5501 – 5507.
[9] a) S. M. Cooper, R. J. Cox, J. Crosby, M. P. Crump, J. Hothersall,
W. Laosripaiboon, T. J. Simpson, C. M. Thomas, Chem.
Commun. 2005, 1179 – 1181; b) R. W. Scott, A. C. Murphy, J.
Wu, J. Hothersall, R. J. Cox, T. J. Simpson, C. M. Thomas, C. L.
Willis, Tetrahedron 2011, 67, 5098 – 5106; c) S. M. Cooper, W.
Laosripaiboon, J. Hothersall, A. K. El-Sayad, C. Winfield, J.
Crosby, R. J. Cox, T. J. Simpson, C. M. Thomas, Chem. Biol.
2005, 12, 825 – 833.
Conflict of interest
The authors declare no conflict of interest.
Keywords: antibiotics · biosynthesis · mutasynthesis ·
polyketides · structure determination
[10] G. M. Singh, F. H. Vaillancourt, J. Yin, C. T. Walsh, Chem. Biol.
2007, 14, 31 – 40.
[11] D. Fukuda, A. S. Haines, Z. Song, A. Murphy, J. Hothersall,
E. R. Stephens, R. Gurney, C. Riemer, R. Marshall, R. J. Cox, J.
Crosby, C. L. Willis, T. J. Simpson, C. M. Thomas, PLoS ONE
2011, 6, e18031.
[12] a) B. Li, C. T. Walsh, Proc. Natl. Acad. Sci. USA 2010, 107,
19731 – 10735; b) S. Huang, Y. Zhao, Z, Qin, X. Wang, M.
Onega, L. Chen, J. He, Y. Yu, H. Deng, Process Biochem. 2011,
46, 811 – 816.
[13] A. C. Murphy, S. Gao, L.-C. Han, D. Fukuda, Z. Song, J.
Hothersall, R. J. Cox, J. Crosby, C. L. Willis, C. M. Thomas, T. J.
Simpson, Chem. Sci. 2014, 5, 397 – 402.
[14] Z. D. Dunn, W. J. Wever, N. J. Economou, A. A. Bowers, B. Li,
Angew. Chem. Int. Ed. 2015, 54, 5137 – 5141; Angew. Chem. 2015,
127, 5226 – 5230.
[1] a) H. Shiozawa, T. Kagasaki, T. Kinoshita, H. Haruyama, H.
Domon, Y. Utsui, K. Kodama, S. Takahashi, J. Antibiot. 1993, 46,
1834 – 1842; b) H. Shiozawa, T. Kagasaki, A. Torikata, N.
Tanaka, K. Fujimoto, T. Hata, Y. Furukawa, S. Takahashi, J.
Antibiot. 1995, 48, 907 – 909; c) H. Shiozawa, A. Shimada, S.
Takahashi, J. Antibiot. 1997, 50, 449 – 452.
[2] a) A. T. Fuller, G. Mellows, M. Woolford, G. T. Banks, K. D.
Barrow, E. B. Chain, Nature 1971, 234, 416 – 417; b) E. B. Chain,
G. Mellows, J. Chem. Soc. Perkin Trans. 1 1977, 294 – 309; c) E. B.
Chain, G. Mellows, J. Chem. Soc. Perkin Trans. 1 1977, 318 – 322.
[3] C. M. Thomas, J. Hothersall, C. L. Willis, T. J. Simpson, Nat. Rev.
Microbiol. 2010, 8, 281 – 289.
[4] J. Piel, Nat. Prod. Rep. 2010, 27, 996 – 1047; E. J. Helfrich, J. Piel,
Nat. Prod. Rep. 2016, 33, 231 – 316.
[5] A. K. El-Sayed, J. Hothersall, S. M. Cooper, E. Stephens, T. J.
Simpson, C. M. Thomas, Chem. Biol. 2003, 10, 419 – 430.
Manuscript received: November 28, 2016
Final Article published: && &&, &&&&
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
Synthetic Biology
A long and winding road: Novel metab-
olites from mutant strains of P. fluores-
cens and the marine bacterium Pseudoal-
teromonas were isolated and structurally
characterized. In combination with com-
plementation experiments, it was shown
that formation of the tetrahydropyran ring
and elaboration of thiomarinol follow
a similar order to that in mupirocin
biosynthesis despite the lack of some
genes.
S.-S. Gao, L. Wang, Z. Song, J. Hothersall,
E. R. Stevens, J. Connolly, P. J. Winn,
R. J. Cox, M. P. Crump, P. R. Race,
C. M. Thomas, T. J. Simpson,*
C. L. Willis*
&&& — &&&
Selected Mutations Reveal New
Intermediates in the Biosynthesis of
Mupirocin and the Thiomarinol
Antibiotics
6
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!