6-Deoxyerythronolide B synthase thioesterase-catalyzed macrocyclization is highly stereoselective

Article abstract of DOI:10.1021/ol300707j

Macrocyclic polyketide natural products are an indispensable source of therapeutic agents. The final stage of their biosynthesis, macrocyclization, is catalyzed regio- and stereoselectively by a thioesterase. A panel of substrates were synthesized to test their specificity for macrocyclization by the erythromycin polyketide synthase TE (DEBS TE) in vitro. It was shown that DEBS TE is highly stereospecific, successfully macrocyclizing a 14-member ring substrate with an R configured O-nucleophile, and highly regioselective, generating exclusively the 14-member lactone over the 12-member lactone.

Full text of DOI:10.1021/ol300707j

ORGANIC
LETTERS
2012
Vol. 14, No. 9
2278–2281
6-Deoxyerythronolide B Synthase
Thioesterase-Catalyzed Macrocyclization
Is Highly Stereoselective
Atahualpa Pinto,†,‡ Meng Wang,†,‡ Mark Horsman,† and Christopher N. Boddy*,†
Departments of Chemistry and Biology, Centre for Catalysis Research and Innovation,
University of Ottawa, Ottawa, ON, Canada T1N 6N5, and Department of Chemistry,
Syracuse University, Syracuse, New York 13244, United States
cboddy@uottawa.ca
Received March 20, 2012
ABSTRACT
Macrocyclic polyketide natural products are an indispensable source of therapeutic agents. The final stage of their biosynthesis,
macrocyclization, is catalyzed regio- and stereoselectively by a thioesterase. A panel of substrates were synthesized to test their specificity
for macrocyclization by the erythromycin polyketide synthase TE (DEBS TE) in vitro. It was shown that DEBS TE is highly stereospecific,
successfully macrocyclizing a 14-member ring substrate with an R configured O-nucleophile, and highly regioselective, generating exclusively
the 14-member lactone over the 12-member lactone.
Macrocyclic polyketide natural products have been
developed into successful pharmaceuticals in nearly all
therapeutic areas.1 The macrocyclic nature of these com-
pounds is critical to their potent and selective biological
activities as it reduces their conformational flexibility,
effectively lowering the entropic cost of binding, which
leads to high affinity and selectivity of binding.2
During the biosynthesis of macrocyclic polyketides, the
macrolactone linkage is formed by a thioesterase (TE)3
found at the C-terminus of the last modular type I poly-
ketide synthase (PKS) in the biosynthetic pathway.4 This
step is essential for turnover of the mature polyketide from
the PKS machinery to which it has been covalently bound.
While essential for successful biosynthesis, TE-catalyzed
macrocyclization is still very poorly understood.5
TEs catalyze macrolactone formation by a two-step
mechanism (Figure1). The activesite serineisfirst acylated
by a linear polyketide thioester. This acyl-enzyme inter-
mediate then undergoes an intramolecular attack leading
to macrolactone formation and its release from the en-
zyme. The acylation step is well-characterized. Based on
kinetic and high-resolution structural data, it is clear that
(5) (a) Boddy, C. N.; Schneider, T. L.; Hotta, K.; Walsh, C. T.;
Khosla, C. J. Am. Chem. Soc. 2003, 125, 3428. (b) Aldrich, C. C.;
Venkatraman, L.; Sherman, D. H.; Fecik, R. A. J. Am. Chem. Soc. 2005,
127, 8910. (c) He, W. G.; Wu, J. Q.; Khosla, C.; Cane, D. E. Bioorg. Med.
Chem. Lett. 2006, 16, 391. (d) Wang, M.; Zhou, H.; Wirz, M.; Tang, Y.;
Boddy, C. N. Biochemistry 2009, 48, 6288.
(6) (a) Wang, M.; Boddy, C. N. Biochemistry 2008, 47, 11793.
(b) Giraldes, J. W.; Akey, D. L.; Kittendorf, J. D.; Sherman, D. H.;
Smith, J. L.; Fecik, R. A Nat. Chem. Biol. 2006, 2, 531. (c) Akey, D. L.;
Kittendorf, J. D.; Giraldes, J. W.; Fecik, R. A.; Sherman, D. H.; Smith,
J. L. Nat. Chem. Biol. 2006, 2, 537.
† University of Ottawa.
‡ Syracuse University.
(1) Li, J. W.-H.; Vederas, J. C. Science 2009, 325, 161.
(2) (a) Kopp, F.; Marahiel, M. A. Nat. Prod. Rep. 2007, 24, 735.
(b) Parenty, A.; Moreau, X.; Campagne, J. M. Chem. Rev. 2006, 106,
911.
(3) Kohli, R. M.; Walsh, C. T. Chem. Commun. 2003, 297.
(4) Michael, A.; Fischbach, M.; Walsh, C. T. Chem. Rev. 2006, 106,
3468.
r
10.1021/ol300707j
Published on Web 04/22/2012
2012 American Chemical Society

this step is rate-determining in vitro and highly substrate
tolerant with hydrophobic enzymeÀsubstrate interactions
controlling substrate specificity.6
polyketides generated by both de novo engineering of
polyketide synthases and chemical synthesis.
Characterization of the macrocyclization activity of
PKS TEs requires substrate analogs that vary the stereo-
chemistry and position of the natural product’s functional
groups. Due to the synthetic complexity of these sub-
strates,5b these essential studies have not progressed. We
thus focused on identifying synthetically tractable sub-
strates that can be used to systematically probe TE macro-
cyclization activity. Herein we disclose the first non-native
TE substrate that undergoes macrocyclization by a type I
PKS TE. With the complete stereoisomer library of this
substrate, our study shows macrocyclization to be exqui-
sitely stereoselective and unpredictable in its selectivity,
highlighting the need for high-resolution characterization
of the enzymeÀsubstrate interactions in this enzyme.
Having demonstrated the ability of DEBS TE to rapidly
hydrolyze short amide-containing thioester substrates,9 we
designed four amide-based stereoisomeric substrates for
the DEBS TE, 2À5 (Figure 1). The amide was selected to
mimic the ketonefound in the nativesubstrate. This ketone
has been proposed to play a role in controlling the sub-
strate’s conformation and helping to properly position the
nucleophilic intramolecular hydroxy group in the active
site channel.5b,c High-resolution structural characteriza-
tion of a nonhydrolyzable acyl-enzyme intermediate of the
related PIK TE suggests that the ketone interacts with a
hydrophilic barrier consisting of bulk water and a gluta-
mine at the exit from the enzyme’s substrate channel,
inducing a turn in the substrate.6c This turn is proposed
to place the intramolecular nucleophilic alcohol in close
proximity to both the catalytic base and the electrophilic
carbonyl, favoring macrocyclization over hydrolysis.6c
The presence of the ketone in TE substrates however has
been shown to lead to hemiketal formation, complicating
in vitro analysis.5b,c Replacement of the ketone with an
amide preserves the carbonyl for interaction with the
proposed hydrophobic barrier, prevents hemiketal forma-
tion, and facilitates fragment coupling during substrate
synthesis. The presence of the C11, C13 diol in the sub-
strate further enabled the possibility of 14- and 12-member
macrolactone formation. Both ofthese ring sizes have been
generated by DEBS TE.5b,c,7 Lastly, the native ethyl sub-
stituent at C13 was replaced with a benzyl chromophore.
This modification was not anticipated to substantially im-
pact TE activity because the benzyl substituent, along with
other sterically more demanding groups, has been shown to
be tolerated at this position in vivo by the DEBS TE.7b,10
This panel of substrates was thus been designed to
provide insight into the role of stereochemistry on macro-
cyclization activity and the regiochemistry of macro-
cyclization.
Figure 1. DEBS TE-catalyzed macrocyclization of 6-deoxyerythro-
nolide B (1) and synthetic analogs 2À5whichweredesignedtoprobe
the stereo- and regioselectivity of DEBS TE macrocyclization.
To date PKS TE-catalyzed macrocyclization has been
observed only with native substrates and very close
analogs.5,7 For example in vivo the TE from the 6-deoxy-
erythronolide B biosynthetic pathway (DEBS TE) cata-
lyzes the formation of its native 14-member ring, 1, as well
as the non-native but highly homologous 16-, 12-, and
8-member rings.7
In vitro macrolactonization has only been observed in
four studies.5 The TEs from the pikromycin and epothi-
lone biosynthetic pathways (PIK TE and Epo TE) have
successfully generated their native products, 10-deoxy-
methonolide and epothilone C respectively.5aÀc The DEBS
TE has also been shown to macrocyclize the precursor to
10-deoxymethonolide,5c and the TE from the zearalenone
pathway has successfully generated a close analog of
zearalenone.5d All other substrates investigated in vitro
have failed to yield detectable macrocyclic products, de-
monstrating rigorous substrate specificity for macrocyc-
lization.8 Identifying the molecular basis controlling TE-
catalyzed macrocyclization is a current and serious gap
in our understanding of polyketide biosynthesis, limit-
ing our ability to design TEs with tunable substrate toler-
ance and regioselectivity for the macrocyclization of new
(7) (a) Xue, Q.; Ashley, G.; Hutchinson, C. R.; Santi, D. V. Proc.
Natl. Acad. Sci. U.S.A. 1999, 96, 11740. (b) Jacobsen, J. R.; Hutchinson,
C. R.; Cane, D. E.; Khosla, C. Science 1997, 277, 367. (c) Kao, C. M.;
McPherson, M.; McDaniel, R. N.; Fu, H.; Cane, D. E.; Khosla, C.
J. Am. Chem. Soc. 1997, 119, 11339. (d) Kao, C. M.; Luo, G. L.; Katz, L.;
Cane, D. E.; Khosla, C. J. Am. Chem. Soc. 1995, 117, 9105. (e) Jacobsen,
J. R.; Cane, D. E.; Khosla, C. Biochemistry 1998, 37, 4928.
(9) Wang, M.; Opare, P.; Boddy, C. N. Bioorg. Med. Chem. Lett.
2009, 19, 1413.
(8) (a) Aggarwal, R.; Caffrey, P.; Leadlay, P. F.; Smith, C. J.;
Staunton, J. J. Chem. Soc., Chem. Commun. 1995, 1519. (b) Weissman,
K. J.; Smith, C. J.; Hanefeld, U.; Aggarwal, R.; Bycroft, M.; Staunton,
J.; Leadlay, P. F. Angew. Chem., Int. Ed. 1998, 37, 1437. (c) Gokhale,
R. S.; Hunziker, D.; Cane, D. E.; Khosla, C. Chem. Biol. 1999, 6, 117.
(10) (a) Hunziker, D.; Wu, N.; Kenoshita, K.; Cane, D. E.; Khosla,
C. Tetrahedron Lett. 1999, 40, 635. (b) Pfeifer, B. A.; Admiraal, S. J.;
Gramajo, H.; Cane, D. E.; Khosla, C. Science 2001, 291, 1790.
(c) Kinoshita, K.; Pfeifer, B. A.; Khosla, C.; Cane, D. E. Bioorg. Med.
Chem. Lett. 2003, 13, 3701.
Org. Lett., Vol. 14, No. 9, 2012
2279

Steady state kinetic characterization of TE-catalyzed
consumption of each substrate provided the turnover
number, Michaelis constant, and, when the enzyme could
not be saturated, the specificity constant (2, kcat = 0.11 (
0.02 minÀ1,Km =1.5(0.6 mM, kcat/Km =1.2(0.5 MÀ1 sÀ1;
Scheme 1. Synthesis of Enantioenriched Substrate 3
3, kcat = 0.21 ( 0.08 minÀ1, Km = 2.2 ( 1.6 mM, kcat
/
Km = 1.6 ( 0.6 MÀ1 sÀ1; 4, kcat = 0.6 ( 0.1 minÀ1, Km =
1.8 ( 1 mM, kcat/Km = 5.6 ( 3.1 MÀ1 sÀ1; 5, kcat/Km
0.42 ( 0.04 MÀ1 À1). The observed specificity constants
=
s
are comparable to the known range determined for DEBS
TE-catalyzed hydrolysis of N-acetylcysteamine (NAC)
thioesters (0.04À31 MÀ1 s
)
À1 6a,8c,9,14 and macrocyclization
(1.5 ( 0.8 MÀ1 À1).5b,c They also agree well with the
s
specificity constant for the DEBS TE-catalyzed hydrolysis
of the highly analogous N-Boc-7-aminoheptanoate NAC
thioester (1.47 ( 0.07 MÀ1 sÀ1).9 The observation that the
pseudo-first-order rate constant for macrocyclization is
Substrates 2À5 were synthesized as shown in Scheme 1.
The Brown allylation11 was used to set the C13 stereo-
center, generating the two enantioenriched homoallylic
alcohols 7 and 8. Following protection and epoxidation,
the Jacobsen hydrolytic kinetic resolution12 was used to set
the C11 stereocenters, ultimately providing access to all
four stereoisomeric protected epoxyalcohols 9, 12À14.
Regioselective vinyl cuprate addition to each enantioen-
riched epoxide, followed by the oxidative cleavage of
the olefin under KMnO4/NaIO4 conditions, generated
the required carboxylic acids which were coupled with
the 7-aminoheptanoate derivative 18.9,13 Deprotection of
the silyl ethers in the presence of the thioester using HF in
acetonitrile completed the synthesis of all four desired
substrates 2À5.
The DEBS TE proved to be highly substrate specific for
macrocyclization (Figure 2). LCMS analysis of 18 h
incubations of 5 μM DEBS TE (50 mM phosphate, pH
7.4) with 2.5 mM 2À5 showed the exclusive macrocycliza-
tion of substrate 2. Syntheses of the authentic macrocycli-
zation standards (see Supporting Information (SI))
enabled unambiguous identification of this enzymatic
productasthe 14-memberring macrocycle19, demonstrat-
ing that the DEBS TE-catalyzed a highly regioselective
macrocyclization. The remaining three substrates, 3À5,
underwent hydrolysis, generating the free acids 21À23.
(11) Brown, H. C.; Jadhav, P. K. J. Am. Chem. Soc. 1983, 105, 2092.
(12) Tokunaga, M.; Larrow, J. F.; Kakiuchi, F.; Jacobsen, E. N.
Science 1997, 277, 936.
(13) Moutevelis-Minakakis, P.; Neokosmidi, A.; Filippakou, M.;
Stephens, D.; Dennis, E. A.; Kokotos, G. J. Pept. Sci. 2007, 13, 634.
(14) (a) Lu, H.; Tsai, S. C.; Khosla, C.; Cane, D. E. Biochemistry
2002, 41, 12590. (b) Sharma, K. K.; Boddy, C. N. Bioorg. Med. Chem.
Lett. 2007, 17, 3034.
Figure 2. LCMS traces (254 nm with observed m/z values) and
product distribution from 18 h incubations of DEBS TE (5 μM
in 50 mM phosphate pH 7.4) with substrates 2À5 (2.5 mM).
2280
Org. Lett., Vol. 14, No. 9, 2012

comparable to those observed for hydrolysis suggests that
active site acylation is also rate determining for macro-
cyclization.6a Because the enzymeÀsubstrate interactions
governing hydrolysis versus macrocyclization are likely
developed after the initial acylation step, the kinetic para-
meters cannot provide insight into these interactions.
A working hypothesis in the community has been that
stereochemistry of the nucleophilic alcohol plays a major
role in controlling macrocyclization activity. This is sup-
ported by the observation that all 16-, 14-, 12-, and
8-member rings produced by the DEBS TE have an
R configured O-nucleophile.5b,c,7 While the successful
macrocyclization of 2 is in agreement with this hypothesis,
the hydrolysis of 3 is not. The hydrolysis of 3 is even more
surprising since its stereochemistry at C11 and C13
matches that of the native substrate. To confirm that 3
was not macrocyclized by the DEBS TE and then subse-
quently ring opened, as is seen with the Epo TE,5a we
incubated the 14-member ring macrocyclic product of 3
(see SI for its synthesis) with the DEBS TE. After pro-
longed incubation, no ring opening of the macrocycle was
observed, confirming that the DEBS TE catalyzes the
direct hydrolysis of 3.
though it is the most homologous of our analogs to the
native substrate, demonstrates that additional key enzymeÀ
substrate interactions play a role in macrocyclization.
Presumably since hydrophobic interactions have been
shown to play a major role in substrate recognition during
acyl-enzyme intermediate formation,6 hydrophobic inter-
actions between the enzyme and the substrate’s methyl
groups may playanimportant role in macrocyclization. As
additional functional groups and side chains are added to
the substrate, the role of the innate conformational pre-
ference of the varying substrates for macrocyclization will
become increasingly important and may also ultimately
play a role in controlling the exquisite selectivity of the
macrocyclization reaction.2b,6c Ultimately, understanding
the interactions that control TE macrocyclization versus
hydrolysis will require high-resolution, structural data of
nonhydrolizable analogs of substrates like 2À5 bound in
the TE active site.
Acknowledgment. We thank L. Barriault (University of
Ottawa) and A. Saleem (University of Ottawa) for assis-
tance with GC and HPLC analysis respectively. This work
was supported by NSERC, the University of Ottawa, and
an NSF Graduate Research Fellowship to A.P.
This first systematic study of the role of stereochemistry
on PKS TE-catalyzed macrocyclization reveals that the
DEBS TEs can macrocyclize analogs that differ substan-
tially from its native substrate. While this study agrees with
the hypothesis that the carbonyl at C9 and the stereo-
chemistry of the nucleophilic alcohol are critical for
macrocyclization,5b,c,7 our data show they are not suffi-
cient. The inability of 3 to undergo macrocyclization, even
Supporting Information Available. Experimental pro-
cedures and characterization data for all new compounds.
This material is available free of charge via the Internet at
http://pubs.acs.org.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 9, 2012
2281