A Flavin-Dependent Decarboxylase-Dehydrogenase-Monooxygenase Assembles the Warhead of α,β-Epoxyketone Proteasome Inhibitors

Article abstract of DOI:10.1021/jacs.6b01619

The α,β-epoxyketone proteasome inhibitor TMC-86A was discovered as a previously unreported metabolite of Streptomyces chromofuscus ATCC49982, and the gene cluster responsible for its biosynthesis was identified via genome sequencing. Incorporation experiments with [¹³C-methyl]l-methionine implicated an α-dimethyl-β-keto acid intermediate in the biosynthesis of TMC-86A. Incubation of the chemically synthesized α-dimethyl-β-keto acid with a purified recombinant flavin-dependent enzyme that is conserved in all known pathways for epoxyketone biosynthesis resulted in formation of the corresponding α-methyl-α,β-epoxyketone. This transformation appears to proceed via an unprecedented decarboxylation-dehydrogenation-monooxygenation cascade. The biosynthesis of the TMC-86A warhead is completed by cytochrome P450-mediated hydroxylation of the α-methyl-α,β-epoxyketone.

Full text of DOI:10.1021/jacs.6b01619

Subscriber access provided by - Access paid by the | UCSB Libraries
Communication
A flavin-dependent decarboxylase-dehydrogenase-monooxygenase
assembles the warhead of #, #-epoxyketone proteasome inhibitors
Daniel Zabala, Joshua W. Cartwright, Douglas M. Roberts, Brian J. C. Law, Lijiang
Song, Markiyan Samborskyy, Peter F Leadlay, Jason Micklefield, and Gregory L. Challis
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.6b01619 • Publication Date (Web): 21 Mar 2016
Downloaded from http://pubs.acs.org on March 22, 2016
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 5
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
A flavin-dependent decarboxylase-dehydrogenase-monooxygenase
assembles the warhead of , -epoxyketone proteasome inhibitors
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
Daniel Zabala†,§, Joshua W. Cartwright†,§, Douglas M. Roberts†, Brian J. C. Law,# Lijiang Song†,
Markiyan Samborskyy‡, Peter F. Leadlay‡, Jason Micklefield# and Gregory L. Challis†,*
†Department of Chemistry, University of Warwick, Coventry CV4 7AL, U.K.
‡Department of Biochemistry, University of Cambridge, Cambridge CB2 1GA, U.K.
#School of Chemistry and Manchester Institute of Biotechnology, University of Manchester, Manchester M1 7DN, U.K.
Supporting Information Placeholder
By analogy with the pathway suggested by Kaysser and co-
ABSTRACT: The -epoxyketone proteasome inhibitor TMC-
workers for eponemycin biosynthesis,8 we propose that the nonri-
86A was discovered as a previously unreported metabolite of
bosomal peptide synthetase (NRPS) TmcG catalyzes the conden-
sation of butanoyl-TmcE 8 with L-serine and L-leucine to give N-
acyl-dipeptidyl thioester 9 (Figure 2). Elongation of N-acyl-
dipeptidyl thioester 9 with a malonyl-CoA-derived extender unit
by the modular polyketide synthase (PKS) TmcH yields the corre-
sponding -ketothioester. It was unclear whether the C-methyl
transferase (CMT) domain of TmcH catalyzes mono- or di-
Streptomyces chromofuscus ATCC49982 and the gene cluster
responsible for its biosynthesis was identified via genome se-
quencing. Incorporation experiments with [13C-methyl]L-
methionine implicated an -dimethyl--keto acid intermediate in
the biosynthesis of TMC-86A. Incubation of the chemically syn-
thesized -dimethyl--keto acid with a purified recombinant
flavin-dependent enzyme that is conserved in all known pathways
for epoxyketone biosynthesis resulted in formation of the corre-
sponding -methyl--epoxyketone. This transformation ap-
pears to proceed via an unprecedented decarboxylation-
dehydrogenation-monooxygenation cascade. The biosynthesis of
the TMC-86A warhead is completed by cytochrome P450-
mediated hydroxylation of the -methyl--epoxyketone.
methylation of this -ketothioester. To investigate this, we fed
[13C-methyl]-L-methionine to S. chromofuscus and examined its
incorporation into TMC-86A. 13C NMR spectroscopic analysis
showed that both methylene carbons of the -hydroxymethyl-
-epoxyketone moiety of TMC-86A are labeled (Figure 4 and
supporting information). This indicates that -ketothioester is
dimethylated by the CMT domain, yielding -dimethyl--
ketothioester 10, which is hydrolytically released from TmcH by
the thioesterase (TE) domain to give -dimethyl--keto acid 11
(Figure 2). A similar conclusion has recently been reached for
TMC-86A 1 belongs to a family of peptide natural products that
irreversibly inhibit the proteasome (Figure 1).1 Other members of
this family include eponemycin 2, epoxomicin 3, macyranone A 4
and landepoxins A/B 5/6 (Figure 1).2-5 The -epoxyketone
pharmacophore of these metabolites covalently modifies the
catalytically essential N-terminal Thr residue of the proteasome -
subunit.6 This inspired the development of carfilzomib 7, a syn-
thetic proteasome inhibitor that has recently been approved for the
treatment of multiple myeloma (Figure 1).7 Although several
biosynthetic gene clusters for metabolites belonging to this group
have recently been identified and genetically investigated,4,5,8 the
mechanism for assembly of the epoxyketone pharmacophore
remains unclear.
eponomycin biosynthesis on the basis of analogous incorporation
experiments.10, 11
During the course of a genomics-driven natural product discovery
program, we generated a near-complete genome sequence of the
herboxidiene-producer Streptomyces chromofuscus ATCC49982
and showed that TMC-86A 1 is a previously unknown metabolite
of this organism (see supporting information).9 Analysis of the S.
chromofuscus genome sequence identified the putative TMC-86A
biosynthetic gene cluster (BGC), which shows a high degree of
similarity to the epn gene cluster reported to direct the biosynthe-
sis of eponemycin in Streptomyces hygroscopiscus ATCC53709
(Figure 2).8 The tmcH gene in S. chromofuscus was replaced with
an apramycin resistance gene. TMC-86A production was abrogat-
ed in the resulting mutant (Figure 3), confirming the involvement
of tmc cluster in the biosynthesis of this metabolite.
Figure 1. Structures of natural and synthetic proteasome inhibi-
tors containing the -epoxyketone pharmacophore.
1
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 5
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: Organization of the S. chromofuscus TMC-86A and S. hygrosopicus eponomycin BGCs, and proposed pathway for TMC-86A
biosynthesis. Desaturation of the leucine side chain occurs after loading of L-leucine onto TmcG, but the timing of this reaction is unclear.
Comparison of the genes encoding tailoring enzymes in the tmc
cluster with those in other epoxyketone BGCs,4,5,8 identified tmcF
as the only tailoring gene that is conserved in all clusters. We
therefore hypothesized that TmcF and its homologs in the other
clusters are responsible for assembly of the epoxyketone pharma-
cophore. TmcF shows sequence similarity to flavin-dependent
dehydrogenases, which typically catalyze -desaturation of
acyl-CoA or acyl-ACP thioesters.12 To examine the role played by
TmcF in TMC-86A biosynthesis, we attempted to overproduce it
in Escherichia coli. However, it resided exclusively within insol-
uble inclusion bodies. We thus cloned and overexpressed epnF,
the homolog from the S. hygroscopicus eponomycin BGC,8 in E.
coli. The resulting N-terminal hexahistidine fusion protein was
produced in soluble form and was purified by Ni-affinity chroma-
tography. The identity of the purified protein was confirmed by
ESI-MS and peptide mass fingerprinting, and gel filtration chro-
matography showed that it is a homodimer. Solutions of the puri-
fied protein were bright yellow and UV-Vis spectroscopic analy-
sis revealed an absorbance maximum at ~450 nm, consistent with
the presence of a flavin cofactor (see supporting information).
Given the propensity of -keto acids to undergo spontaneous
decarboxylation and the possibility that the TE domain of TmcH
could catalyze this process,13,14 we initially surmised that isopro-
pyl ketone 12, resulting from decarboxylation of -dimethyl--
keto acid 11, was most likely to be the substrate of TcmF/EpnF.
We thus synthesized 4-amino-2,6-dimethyl-heptan-3-one 14 by
reacting the Weinreb amide of N-BOC-L-leucine 13 with iPrMgCl
and subsequent deprotection with HCl in 1, 4-dioxane.15 BOP-
mediated condensation of 14 with N-butanoyl-L-serine 16, derived
from DCC coupling of L-serine methyl ester 15 with butanoic
acid and subsequent saponification with H2O2/LiOH, gave isopro-
pyl ketone 12 (Scheme 1).16 NMR spectroscopic analysis indicat-
ed that this material is a mixture of diastereomers resulting from
epimerization of the ketone -stereocenter.
Figure 4. Sites of Incorporation of [13C-methyl]-L-methionine
into TMC-86A determined by 13C NMR spectroscopic analysis.
No detectable products were formed when 12 was incubated with
EpnF. We thus turned our attention to the synthesis of -
dimethyl--keto acid 11 (Scheme 2), the presumed product of
TcmH. EDCI/DMAP-mediated coupling of N-BOC-L-leucine
with Meldrum’s acid, followed by thermolysis in the presence of
BnOH gave -keto ester 17.17 Methylation of 17 with excess
K2CO3 and MeI gave the corresponding -dimethyl--ketoester,18
which was deprotected with TFA and coupled with N-butanoyl-L-
serine 16 to give the benzyl ester 18 of -dimethyl--keto acid 11.
Hydrogenolysis of 18 over Pd/C gave a mixture of 11 and its
decarboxylation product 12. The ratio of 11 to 12 obtained from
this reaction varied, but was ~9:1 in the best cases. Consistent
with the anticipated propensity of 11 to undergo spontaneous
decarboxylation, a significant proportion was converted to 12
during spectroscopic analysis.
Figure 3. Extracted ion chromatograms at m/z = 343.18, corre-
sponding to [M+H]+ for TMC-86A, from UHPLC-ESI-TOF-MS
analyses of wild type S. chromofuscus (top trace) and the tmcH
mutant (bottom trace).
2
ACS Paragon Plus Environment

Page 3 of 5
Journal of the American Chemical Society
Scheme 1. Synthesis of isopropyl ketone 12
To investigate whether -unsaturated ketone 21 can dissociate
1
2
3
4
5
6
7
from EpnF during the conversion of 11 to 19, we included N-
acetylcysteamine (NAC) in the reaction mixture. Only the product
resulting from opening of the epoxide in 19 with NAC could be
detected in LC-MS analyses (see supporting information). None
of the product resulting from conjugate addition of NAC to -
unsaturated ketone 21 was observed. Thus, 21 is likely unable to
dissociate from EpnF before undergoing epoxidation.
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
Incubation of a mixture of 11 and 12 with EpnF resulted in for-
mation of a new compound with a molecular formula correspond-
ing to epoxyketone 19 (Figure 5). This compound was absent in
control experiments from which the enzyme had been omitted, or
added after heat denaturation. High resolution MS/MS analyses
confirmed that 19 is the product of the EpnF-catalyzed reaction
(see supporting information). These findings are consistent with
the recent report that heterologous expression of epnF, epnG and
epnH in E. coli results in the production of the N-hexanoyl ana-
logue of 19.10
Figure 6. Proposed mechanism of the EpnF-catalyzed reaction.
Scheme 2. Synthesis of -dimethyl--keto acid 11
Hydroxylation of the methyl group in the -methyl-, -
epoxyketone moiety of 19 could be catalyzed by either of the
cytochrome P450s encoded by tmcI and tmcK. Sequence compari-
sons show that tmcK orthologs are only present in BGCs for
epoxyketones that contain a dehydroleucine residue. Thus, TmcK
is proposed to be involved in desaturation of the side chain of the
leucine residue incorporated into TMC-86A, but the timing of this
reaction remains unclear. One scenario involves TmcJ, which
contains MbtH, adenylation (A) and peptidyl carrier protein (PCP)
domains. Loading of L-Leu by the A domain of TmcJ onto the
adjacent PCP domain would afford the corresponding aminoacyl
thioester. TcmK-catalyzed desaturation of this thioester, followed
by hydrolytic release would generate 4, 5-dehydro-L-Leu, which
could be loaded onto the second PCP domain of TmcG by the
adjacent A domain. Other non-proteinogenic amino acid NRPS
substrates are known to be biosynthesized via a similar mecha-
nism.19 Alternatively, the second A domain of TmcG could load
L-Leu directly onto the adjacent PCP domain and TmcK could
catalyze side chain desaturation at either this point, or a later stage
in the assembly of TMC-86A. To discriminate between these
possibilities, we overproduced the second A-PCP didomain of
TmcG in E. coli as a soluble C-terminal hexahistidine fusion. The
substrate specificity of the A domain within the purified protein
was assessed using a pyrophosphate release assay. While the
TmcG A domain could adenylate 4, 5-dehydro-L-leucine, it was
found to greatly prefer L-leucine as a substrate (see supporting
information). Thus it appears that TmcK acts at some point after
loading of L-leucine onto the second PCP domain of TmcG. This
conclusion is supported by the observation that both eponemycin
and its leucine-containing derivative are produced when the epn
cluster is expressed in Streptomyces albus.11
The catalytic mechanism of acyl-CoA dehydrogenases typically
involves deprotonation of the thioester substrate by an active site
base to generate an enolate intermediate that transfers hydride
from the -carbon atom to the bound flavin cofactor.12 The obser-
vation that EpnF can utilize 11, but not 12, as a substrate indicates
that it employs a unique mechanism involving decarboxylation of
11 to generate the enolate intermediate 20. Elimination of hydride
from one of the -carbon atoms of 20 would reduce the flavin
cofactor, yielding -unsaturated ketone 21. Reaction of the
reduced flavin with molecular oxygen forms the flavin hydroper-
oxide, which could convert 21 to the epoxyketone 19 via a conju-
gate addition-1, 3-elimination mechanism (Figure 6).
By a process of elimination, we hypothesized that TmcI catalyzes
hydroxylation of the methyl group in the -methyl-, -
epoxyketone moiety of 19. To test this hypothesis, we overpro-
duced TmcI in E. coli as a soluble N-terminal hexahistidine fu-
sion. The identity of the purified protein was confirmed by ESI-
MS and it was shown to exist as a monomer containing a heme
prosthetic group (see supporting information). Incubation of a
mixture of 11 and 12 with EpnF, TmcI, spinach ferredoxin (Fd)
and ferredoxin reductase (FdR), and NADPH resulted in produc-
Figure 5. Extracted ion chromatograms at m/z = 329.19, corre-
sponding to [M+H]+ for 19, from UHPLC-ESI-TOF-MS analyses
of reactions containing a mixture of 11 and 12, and native EpnF
(top trace) or the heat-denatured enzyme (bottom trace).
3
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 5
tion of a compound with a molecular formula corresponding to -
1
2
3
4
hydroxymethyl-, -epoxyketone 22. This product was absent
from control reactions containing heat-denatured EpnF and TmcI
(Figure 7). Comparison of high resolution MS/MS spectra for 1
and the reaction product confirmed the structure of the latter as 22
(see supporting information).
Corresponding Author
G.L.Challis@warwick.ac.uk
Author Contributions
5
§These authors contributed equally.
6
7
8
9
Notes
The authors declare no competing financial interests.
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
This research was supported by the BBSRC and Syngenta through
a sLoLa to GLC/PFL/JM (BB/K002341/1), and the BBSRC
MIBTP Doctoral Training Partnership (BB/J014532/1). The
assistance of Dr Matthew Jenner with the measurement and inter-
pretation of protein mass spectra and of Dr Ivan Prokes with the
acquisition of NMR spectra is gratefully acknowledged.
Figure 7. Extracted ion chromatograms at m/z = 367.17, corre-
sponding to [M+Na]+ for 22, from UHPLC-ESI-TOF-MS anal-
yses of reactions containing a mixture of 11 and 12, spinach fer-
redoxin and ferredoxin reductase, NADPH and native (top trace)
or heat-denatured (bottom trace) EpnF and TmcI.
(1) Koguchi, Y.; Kohno, J.; Suzuki, S.; Nishio, M.; Takahashi, K.;
Ohnuki, T.; Komatsubara, S. J. Antibiot. (Tokyo) 1999, 52, 1069.
(2) Sugawara, K.; Hatori, M.; Nishiyama, Y.; Tomita, K.; Kamei, H.;
Konishi, M.; Oki, T. J. Antibiot. (Tokyo) 1990, 43, 8.
(3) Hanada, M.; Sugawara, K.; Kaneta, K.; Toda, S.; Nishiyama, Y.;
Tomita, K.; Yamamoto, H.; Konishi, M.; Oki, T. J. Antibiot. (Tokyo)
1992, 45, 1746.
(4) Keller, L.; Plaza, A.; Dubiella, C.; Groll, M.; Kaiser, M.; Müller, R.
J. Am. Chem. Soc. 2015, 137, 8121.
(5) Owen, J. G.; Charlop-Powers, Z.; Smith, A. G.; Ternei, M. a; Calle,
P. Y.; Reddy, B. V. B.; Montiel, D.; Brady, S. F. Proc. Natl. Acad. Sci. U.
S. A. 2015, 112, 4221.
(6) Groll, M.; Kim, K. B.; Kairies, N.; Huber, R.; Crews, C. M. J. Am.
Chem. Soc. 2000, 122, 1237.
(7) Kim, K. B.; Crews, C. M. Nat. Prod. Rep. 2013, 30, 600.
(8) Schorn, M.; Zettler, J.; Noel, J. P.; Dorrestein, P. C.; Moore, B. S.;
Kaysser, L. ACS Chem. Biol. 2014, 9, 301.
(9) The sequence of the TMC-86A biosynthetic gene cluster has been
depositied in the EMBL database under the accession number LN999909.
(10 ) Liu, J.; Zhu, X.; Zhang, W. ChemBioChem, 2015, 16, 2585.
(11) Zettler, J.; Zubeil, F.; Kulik, A.; Grond, S.; Kaysser, L.
ChemBioChem, 2016, doi: 10.1002/cbic.201500567.
(12) Ghisla, S.; Thorpe, C. Eur. J. Biochem., 2004, 271, 494.
(13) Poust, S.; Phelan, R. M.; Deng, K.; Katz, L.; Petzold, C. J.;
Keasling, J. D. Angew. Chem. Int. Ed. 2015, 54, 2370.
(14) Gu, L.; Wang, B.; Kulkarni, A.; Gehret, J. J.; Lloyd, K. R.;
Gerwick, L.; Gerwick, W. H.; Wipf, P.; Håkansson, K.; Smith, J. L.;
Sherman, D. H. J. Am. Chem. Soc. 2009, 131, 16033.
(15) Pereira, A. R.; Kale, A. J.; Fenley, A. T.; Byrum, T.; Debonsi, H.
M.; Gilson, M. K.; Valeriote, F. A.; Moore, B. S.; Gerwick, W. H.
ChemBioChem 2012, 13, 810.
(16) Schmidt, U.; Schmidt, J. Synthesis (Stuttgart), 1994, 3, 300.
(17) Lee, S. K.; Choi, K. H.; Lee, S. J.; Lee, J. S.; Park, J. Y.; Kim, B.
M.; Lee, B. J. Bioorg. Med. Chem. Lett. 2011, 21, 133.
(18) Giordano, A. ; Della Monica, C. ; Landi, F. ; Spinella, A. ; Sodano,
G. Tetrahedron Lett. 2000, 41, 3979.
In conclusion, using a genomics-led approach, we have identified
a TMC-86A biosynthetic gene cluster in Streptomyces chro-
mofuscus ATCC49982. Incorporation of [13C-methyl]L-
methionine into both of the methylene groups of the -
hydroxymethyl--epoxyketone warhead of TMC-86A impli-
cated an -dimethyl--keto acid intermediate in its biosynthesis.
This -dimethyl--keto acid was synthesized in eight steps and
shown to be converted to the corresponding -methyl--
epoxyketone by a flavin-dependent enzyme that is conserved in
all known pathways for epoxyketone biosynthesis. The isopropyl
ketone resulting from decarboxylation of the -dimethyl--keto
acid was also synthesized and shown not to be a substrate of the
enzyme. Thus we conclude that the pharmacophore of -
epoxyketone proteasome inhibitors is biosynthesised by a remark-
able decarboxylase-dehydrogenase-monooxygenase that employs
a catalytic flavin cofactor. To the best of our knowledge, no other
flavin-dependent enzyme is known to catalyze a similar sequence
of reactions. The -methyl--epoxyketone pharmacophore
initially formed in TMC-86A biosynthesis undergoes additional
elaboration via hydroxylation of the -methyl group. We have
shown that this is catalyzed by the cytochrome P450 TmcI. Hom-
ologs of TmcI are encoded by genes within the eponemyin 2 and
epoxomicin 3 BGCs,8 although in the latter the TmcI homolog
appears to be non-functional.11
Supporting Information
Details of experimental procedures, additional LC-MS, MS/MS,
NMR and protein characterization data. This material is available
free of charge via the Internet at http://pubs.acs.org.
(19) Walsh, C. T.; O'Brien, R. V.; Khosla, C. Angew. Chem. Int. Ed.
2013, 52, 7098.
4
ACS Paragon Plus Environment

Page 5 of 5
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
5
ACS Paragon Plus Environment