Nahuoic acid A produced by a Streptomyces sp. isolated from a marine sediment is a selective SAM-competitive inhibitor of the histone methyltransferase SETD8

Article abstract of DOI:10.1021/ol303416k

The histone lysine monomethyltransferase SETD8 is an epigenetic regulator of cell cycle progression. Nahuoic acid A (1), a polyketide produced in culture by a Streptomyces sp. obtained from a tropical marine sediment, is the first known selective SAM-competitive inhibitor of SETD8. The structure of nahuoic acid A (1) has been elucidated by chemical transformation and detailed analysis of spectroscopic data.

Full text of DOI:10.1021/ol303416k

ORGANIC
LETTERS
2013
Vol. 15, No. 2
414–417
Nahuoic Acid A Produced by a Streptomyces
sp. Isolated From a Marine Sediment Is a
Selective SAM-Competitive Inhibitor of the
Histone Methyltransferase SETD8
David E. Williams,† Doralyn S. Dalisay,†,^ Fengling Li,‡ James Amphlett,†
Wisanu Maneerat,† Miguel Angel Garcia Chavez,§ Yan Alexander Wang,§
Teatulohi Matainaho, Wenyu Yu,‡ Peter J. Brown,‡ Cheryl H. Arrowsmith,‡
Masoud Vedadi,‡ and Raymond J. Andersen*,†
Departments of Chemistry and Earth, Ocean & Atmospheric Sciences, University of
British Columbia, 2036 Main Mall, Vancouver, BC, Canada V6T 1Z1, Structural
Genomics Consortium, University of Toronto, 101 College Street, Ontario, Canada M5G
1L7, and Department of Pharmacology, University of Papua New Guinea, Port
Moresby, Papua New Guinea
raymond.andersen@ubc.ca
Received December 13, 2012
ABSTRACT
The histone lysine monomethyltransferase SETD8 is an epigenetic regulator of cell cycle progression. Nahuoic acid A (1), a polyketide produced
in culture by a Streptomyces sp. obtained from a tropical marine sediment, is the first known selective SAM-competitive inhibitor of SETD8. The
structure of nahuoic acid A (1) has been elucidated by chemical transformation and detailed analysis of spectroscopic data.
Protein methyltransferases (PMTs) transfer methyl groups
from S-adenosylemethionine (SAM) to lysine or arginine
residues in histones and other chromatin-associated sub-
strates.1 These proteins play essential roles in epigenetic
regulation of gene expression and chromatin-dependent
signaling. Mutation and aberrant regulation of PMTs are
linked to many diseases, especially cancer,2 and finding
PMT inhibitors is considered to be a new frontier for drug
discovery. Targeting the common (SAM) cofactor binding
site of PMTs is a particularly attractive strategy for cancer
therapy.1
SETD8, a PMT that monomethylates lysine 20 of histone
H4 (H4K20), plays a central role in the silencing of euchro-
matic genes.3,4 SETD8 also monomethylates lysine 382 of
p53/TP53, thereby repressing p53/TP53-target genes,5
and monomethylates lysine 248 of proliferating cell
† Departments of Chemistry and Earth, Ocean & Atmospheric Sciences,
University of British Columbia.
(3) Nishioka, K.; Rice, J. C.; Sarma, K.; Erdjument-Bromage, H.;
Werner, J.; Wang, Y.; Chuikov, S.; Valenzuela, P.; Tempst, P.; Steward,
R.; Lis, J. T.; Allis, C. D.; Reinberg, D. Mol. Cell 2002, 9, 1201–1213.
(4) Fang, J.; Feng, Q.; Ketel, C. S.; Wang, H.; Cao, R.; Xia, L.;
Erdjument-Bromage, H.; Tempst, P.; Simon, J. A.; Zhang, Y. Curr. Biol.
2002, 12, 1086–1099.
‡ University of Toronto.
§ Department of Chemistry, University of British Columbia.
University of Papua New Guinea.
^ Institute of Biological Chemistry, Washington State University.
(1) Arrowsmith, C. H.; Bountra, C.; Fish, P. V.; Lee, K.; Schapira,
M. Nat. Rev. Drug Discovery 2012, 11, 384–400.
(2) Chi, P.; Allis, C. D.; Wang, G. G. Nat. Rev. Cancer 2010, 10, 457–469.
(5) Shi, X.; Kachirskaia, I.; Yamaguchi, H.; West, L. E.; Wen, H.; Wang,
E. W.; Dutta, S.; Appella, E.; Gozani, O. Mol. Cell 2007, 27, 636–646.
r
10.1021/ol303416k
Published on Web 12/28/2012
2012 American Chemical Society

nuclear antigen (PCNA), enhancing the interaction
between PCNA and the flap endonuclease FEN1.6
SETD8 is overexpressed in various types of cancer,
and aberrant monomethylation by SETD8 may lead to
human carcinogenesis.6 Potent SETD8 inhibitors would
serve as useful chemical probes7 to further investigate the
cellular effects of SETD8 inhibition in both normal and
diseased cells and as lead strucures for development of
anticancer therapeutics. However, no SETD8 inhibitors
that target the SAM cofactor site have been reported in the
literature to date.
A screen of a library of marine organism extracts and pure
marine natural products revealed that the highly hydroxy-
lated polyketide nahuoic acid A (1) is a selective SAM-
competitive SETD8 inhibitor in vitro. Nahuoic acid A (1) is
produced in culture by a Streptomyces sp. (isolate RJA2928)
obtained from a marine sediment collected near the passage
Padana Nahua in Papua New Guinea. The highly modified
tetrapeptides padanamide A (actinoramide A8) and padana-
mide B are produced by cultures of the same Streptomyces sp.
(RJA2928).9 Details of the isolation, structure elucidation,
and SETD8 inhibitory activity of nahuoic acid A (1) are
reported below.
Nahuoic acid A (1) was obtained as an optically active
oil that gave a [M þ Na]þ ion in the HRESIMS at m/z
545.3467 appropriate for a molecular formula of C30H50O7,
requiring six sites of unsaturation. The 1H and 13C NMR
spectra obtained for the samples of nahuoic acid A (1) ob-
tained via C8 HPLC with or without added TFA (Table 1,
Supporting Information) were identical in all respects
except that five exchangeable proton resonances seen as
distinct peaks in the 1H NMR spectrum of the first C8 HPLC
sample (eluent, H2O/MeCN 18:7) coalesced into broad
signals in the second HPLC sample due to the presence of
trace amounts of TFA from the eluent. Therefore, the
initial structural analysis of nahuoic acid A (1) was carried
out using the NMR data obtained for the sample of 1 with
distinct OH resonances and the structural arguments were
subsequently confirmed on the sample containing TFA.
The 1H/13C/gCOSY60/gHSQC/gHMBC NMR data
obtained for 1 (Table 1, Supporting Information) identi-
fied resonances that could be assigned to 30 carbon atoms,
with 44 protons bound to carbon (5 ꢀ C, 14 ꢀ CH, 3 ꢀ
CH2, 8ꢀ CH3) and sixexchangeableprotons, inagreement
with the ESIHRMS measurement. Five of the six ex-
changeable proton signals were assigned to secondary OH
protons [δ 4.36 (bs, OH-9), 4.30 (d, J = 4.9 Hz, OH-21),
4.29 (d, J = 4.7 Hz, OH-19), 4.26 (bd, J = 4.7 Hz, OH-17),
and 4.21 (d, J = 5.8 Hz, OH-7)] that each correlated to a
methine resonance [δ 3.47 (H-9), 3.12 (H-21), 3.77 (H-19),
3.56 (H-17), and 3.51 (H-7)] in the gCOSY60 experiment.
The five methine proton resonances correlated in the gHSQC
to carbon resonances with chemical shifts appropriate for
carbinol methines [δ72.8 (C-9), 78.6 (C-21), 70.0 (C-19), 67.1
(C-17), and 66.3 (C-7)]. A final exchangeable proton
resonance at δ 11.78 was assigned to a carboxylic acid,
whose carbon appeared at δ 169.1 (C-1).
Production culturesofRJA2928 were grown aslawns on
solid agar containing marine medium at rt for 14 days (see
Supporting Information). Mature cultures were extracted
by soaking cells and medium together in two separate
aliquots of EtOAc. Concentration of the combined EtOAc
extracts in vacuo, followed first by partition of the residue
between H2O and EtOAc and then Sephadex LH-20 chro-
matography (eluent, 4:1 MeOH/CH2Cl2) of the EtOAc
soluble material, gave fractions containing the padanamides
and nahuoic acid A (1). Further purification by step gradient
Si gel flash chromatography gave fractions highly enriched
Three trisubstituted alkenes [δ 1H/13C 6.41 (d, J = 10.7
Hz, H-3)/ 124.4 (C-2); 149.0 (C-3); 5.15 (bs, H-11)/132.5
(C-11); 132.7 (C-12); 4.97 (m, H-15)/ 135.1 (C-14); 125.4
(C-15)], four methyl doublets (1H/13C δ 0.85 (J = 7.2 Hz,
H-25)/18.2 (C-25); 0.83 (J = 6.5 Hz, H-30)/18.5 (C-30);
0.76 (J = 6.5 Hz, H-23)/19.5(C-23); 0.76 (J = 6.8 Hz,
H-29)/7.2 (C-29)], four methyl singlets (δ 1H/13C 1.58 (H-
24)/13.0 (C-24); 1.46 (H-27)/21.8 (C-27); 1.01 (H-26)/27.6
(C-26); 1.40 (H-28)/12.4 (C-28)], three aliphatic methy-
lenes [δ 1H/13C 2.04 (H-16), 1.91 (H-160)/37.0(C-16); 1.87
(H-8eq), 1.62 (H-8ax)/40.2 (C-8); 1.18 (H-18), 1.42 (H-180)/
41.7 (C-18)], six additional methines [δ1H/13C 3.54 (H-4)/
36.1 (C-4); 2.23 (H-13) /56.8 (C-13); 1.92 (H-6)/38.3 (C-6);
1.63 (H-22)/30.3 (C-22); 1.52 (H-5)/50.1 (C-5); 1.39 (H-
20)/39.9 (C-20)], and one aliphatic quaternary carbon (δ
41.1, C-10) could also be readily identified from the NMR
data. The carboxylic acid and three alkenes accounted for
only four of the six sites of unsaturation required by the
molecular formula, requiring that nahuoic acid A (1) had
to contain two rings.
1
in the H NMR signals for nahuoic acid A. Pure nahuoic
acid A (1) was obtained initially from C8 reversed-phase
HPLC by elution with 18:7 H2O/MeCN. Subsequently, it
was found that HPLC using the same column but with 7:3
(0.05% TFA/H2O)/MeCN as eluent was more efficient.
(6) Takawa, M.; Cho, H. S.; Hayami, S.; Toyokawa, G.; Kogure, M.;
Yamane, Y.; Iwai, Y.; Maejima, K.; Ueda, K.; Masuda, A.; Dohmae,
N.; Field, H. I.; Tsunoda, T.; Kobayashi, T.; Akasu, T.; Sugiyama, M.;
Ohnuma, S. I.; Atomi, Y.; Ponder, B. A.; Nakamura, Y.; Hamamoto, R.
Cancer Res. 2012, 72, 3217–3227.
€
(7) (a) Frye, S. V. Nat. Chem. Biol. 2010, 6, 159–161. (b) Muller, S.;
Brown, P. J. Clin. Pharmacol. Ther. 2012, 92, 689–693.
Six fragments shown in bold in Figure 1A could be
unambiguously identified from the gCOSY60 data. These
fragments could be connected via the series of HMBC
correlations [Me-29 (δ 0.76) to C-19 (δ 70.0); Me-28 (δ
1.40) to C-13 (δ 56.8); Me-27 (δ 1.46) to C-13 (δ 56.8);
(8) Padanamide A and actinoramide A have the same structure and
were reported simultaneously. See: Nam, S. J.; Kauffman, C. A.; Jensen,
P. R.; Fenical, W. Tetrahedron 2011, 67, 6707–6712.
(9) Williams, D. E.; Dalisay, D. S.; Patrick, B. O.; Matainaho, T.;
Andrusiak, K.; Deshpande, R.; Myers, C. L.; Piotrowski, J. S.; Boone,
C.; Yoshida, M.; Andersen, R. J. Org. Lett. 2011, 13, 3936–3939.
Org. Lett., Vol. 15, No. 2, 2013
415

Scheme 1. Derivatization of Nahuoic Acid A (1)
observed between Me-26 (δ 1.01) and each of H-6 (δ 1.92),
H-8ax (δ 1.62), and H-9 (δ 3.47) were consistent with a chair
conformation and established that Me-26 was axial, H-5 was
equatorial, Me-25 was equatorial, OH-7 was equatorial, and
OH-9 was axial on the C-5 to C-10 ring. H-4 (δ 3.54) was
coupled to each of H-3 (δ 6.41), H-5 (δ 1.52), and H-13
(δ 2.23) with a J g 8.9 Hz indicating diaxial relationships
between H-4 and both H-5 and H-13, and an anti relationship
between H-4 and H-3. tROESY correlations between the
three methine protons H-3 (δ 6.41), H-5 (δ 1.52) and H-13
(δ 2.23) confirmed the diaxial relationship between H-5 and
H-13 and the cis relationship between both H-5 and H-13 and
C-3. tROESY correlations observed between Me-24 (δ 1.58)
and H-4 (δ 3.54), between Me-28 (δ 1.40) and H-16/H-160
(δ 1.91/2.04), and between H-15 (δ 4.97) and H-13 (δ 2.23)
demonstrated that the Δ2,3 and Δ14,15 olefins both had the E
configuration.
In order to facilitate the determination of the relative
configuration of the C-17 to C-23 segment of nahuoic acid A,
the natural product 1 was treated with 2,2-dimethoxypro-
pane and PPTS, and without purification, the putative
acetonides 1a were subsequently reacted with p-bromo-
phenacylbromide and Et3N in DMF to give a major
product ester 2 (Scheme 1) (see Supporting Information
for MS and NMR assignments). The geminal methyl
carbons (Me-40 and Me-41) in the C-19/C-21 acetonide
had chemical shifts of δ 19.9 and 30.4 demonstrating that
the C-19 and C-21 alcohols in 1 were syn.10
Analysis of the J couplings using a seriesof homonuclear
decoupling experiments, combined with the tROESY and
gCOSY60 NMR data for ester 2, yielded the relative
configurations at C-17, C-19, C-20, and C-21 as shown
in Figure 2. tROESY correlations observed between the
axial geminal methyl (C-40: δ 1.41) of the acetonide and
both H-19 (δ 4.31) and H-21 (δ 3.21), and between H-19
and H-21, confirmed that the conformation of the dioxane
ring, as predicted by Rychnovsky et al., was a chair10 and
that H-19 and H-21 were axial and C-18 and C-22 were
equatorial. Small couplings (<2 Hz) were observed be-
tween H-20 (δ 1.33) and both H-19 and H-21, indicating
Figure 1. Selected gCOSY60, gHMBC, and tROESY correla-
tions observed for Nahuoic acid A (1).
Me-26 (δ 1.01) to C-11 (δ 132.5), C-10 (δ 41.1), C-9
(δ 72.8); H-8eq (δ 1.87) to C-9 (δ 72.8), C-10 (δ 41.1);
Me-25 (δ 0.85) to C-5 (δ 50.1); Me-24 (δ 1.58) to C-1
(δ 169.1)] summarized with arrows in Figure 1A, establish-
ing the linear carbon backbone of the polyketide chain
from C-1 to C-23 and the locations of the three trisubsti-
tuted alkenes, the branching methyls, the carboxylic acid,
and the secondary alcohols. Vicinal coupling observed
between H-4 (δ 3.54) and H-13 (δ 2.23) in the gCOSY60
identified a bond between C-4 and C-13. gHMBC correla-
tions observed between Me-26 (δ 1.01) and both C-5
(δ 50.1) and C-10 (δ 41.1) established a bond between
C-10 and C-5. The C-4/C-13 and C-5/C-10 bonds created
the two rings of a decalin ring system from the linear
polyketide backbone.
The relative configuration about the decalin ring system
in 1 was determined from the J coupling data obtained for
1 containing TFA, in which no coupling to the exchange-
able OH protons was observed, and the series of tROESY
correlations illustrated in Figure 1B. A tROESY correla-
tion between Me-26 (δ 1.01) and H-5 (δ 1.52) established
that the decalin ring junction was cis. The H-7 resonance
at δ 3.51 appeared as a triplet of doublets with the large
coupling of 11.0 Hz, indicating that the ring encompassing
C-5 to C-10 was in a chair conformation with H-6, H-7, and
H-8ax all axial as shown in Figure 1B. tROESY correlations
(10) Rychnovsky, S. D.; Rogers, B.; Guang Yang, G. J. Org. Chem.
1993, 58, 3511–3515.
416
Org. Lett., Vol. 15, No. 2, 2013

which matched the experimentally obtained data (see
Supporting Information for details).
Nahuoic acid A (1) inhibited SETD8 activity with an
IC50 value of 6.5 ( 0.5 μM and a Hill slope of 1.6
(Supporting Information). Interestingly, nahuoic acid A
(1) had no significant inhibitory effect on the activity of
other protein methyltransferases such as G9a, EHMT1,
SETD7, SUV39H2, SUV420H1, SUV420H2, DOT1L,
PRMT3, and PRMT5 and MLL complexes. Kinetic anal-
ysis of the inhibition of SETD8 activity by nahuoic acid A (1)
indicated that it is a competitive inhibitor of SAM binding
with a Ki value of 2 ( 0.3 μM (Supporting Information), but
a noncompetitive inhibitor with respect to the binding of the
peptide substrate (Supporting Information).
The pentahydroxylated and hexamethylated linear C-1
to C-23 carbon backbone of nahuoic acid A (1) appears to
have a PKS origin, but its carbon skeleton is without
precedent among known natural products.11 A Dielsꢁ
Alderase12 may be involved in the formation of the decalin
ring fragment. To the best of our knowledge, nahuoic acid
A (1) is the first natural product known to inhibit SETD813
and the first selective SETD8 inhibitor known to be a
competitive inhibitor of SAM binding. Nahuoic acid A (1)
has the potential to be a useful tool to aid the investiga-
tion of the roles of SETD8 in cells and to act as a probe
for defining a selective SAM-competitive binding site on
SETD8 that can guide the design of more potent syn-
thetic inhibitors for therapeutic1 or cell biology re-
search applications.
Figure 2. tROESY data for 1 and 2.
that H-20 was equatorial. H-19 (δ 4.31) had a small coupl-
ing (J < 2 Hz) to the H-18b resonance (δ 1.38) and their
correlation in the gCOSY60 spectrum was weak. In con-
trast, the coupling constant between H-19 and H-18a
(δ 1.84) was 10.1 Hz and their gCOSY60 correlation was
strong, indicating that these protons were anti. Similarly a
9.6 Hz coupling between H-21 and H-22 (δ 1.74) required
these to be anti. Me-29 (δ 0.99) showed tROESY correla-
tions to both H-18a and H-22 in agreement with the above
assignment. A 2.5 Hz coupling between H-18a and H-17
(δ 4.10) required a dihedral angle of near 60ꢀ and tROESY
correlations between H-16a (δ 2.15) and H-18a and be-
tween H-16b (δ 2.28) and H-18b (δ 1.38) were consistent
with the extended zigzag conformation and the relative
configurations for the carbon chain extending from C-17
to C-21 as shown in Figure 2. A weak tROESY correlation
observed between Me-40 and H-17, and the small coupling
(J = 3.4 Hz) between H-16a and H-17 provided additional
support for the assigned relative configurations. tROESY
correlations observed between H-19 (δ 3.77), and both
H-21 (δ 3.12) and 17-OH (δ 4.26) in nahuoic acid A (1)
suggests that the underivatized natural product adopts an
analogous extended zigzag conformation.
Acknowledgment. Financial support was provided by
NSERC (RJA and YAW) and CCSRI (RJA). The Struc-
tural Genomics Consortium is a registered charity (number
1097737) that receives funds from CIHR, Eli Lilly Canada,
Genome Canada, GlaxoSmithKline, the Ontario Ministry of
Economic Development and Innovation, the Novartis
Research Foundation, Pfizer, Abbott, Takeda and the
Wellcome Trust.
Supporting Information Available. Experimental details;
1D and 2D NMR spectra for 1 - 4; bioactivity data for 1;
details of Mosher ester analysis. This material is availablefree
of charge via the Internet at http://pubs.acs.org.
The complete relative configuration and the absolute
configuration of nahuoic acid A (1) was determined by a
modified Mosher ester analysis. Nahuoic acid A (1) was
first converted into the Mosher esters 3 and 4 as shown
in Scheme 1. A standard Mosher ester analysis of the 1H
NMR data collected for 3 and 4 showed that the absolute
configuration of 1 is 4R,5S,6R,7R,9R,10R,13R,17S,19R,
20S,21S. The Mosher ester absolute configuration assign-
ment was confirmed by calculating the CD spectrum for 1,
(11) Fiscbach, M. A.; Walsh, C. T. Chem. Rev. 2006, 106, 3468–3496.
(12) Oikawa, H.; Tokiwano, T. Nat. Prod. Rep. 2004, 21, 321–352.
(13) Reinberg, D.; Trojer, P.; Sbardella, G. PCT WO 2007149782.
Kodama, T.; Tanaka, T.; Kawamura, T.; Wada, Y.; Sigiyama, A.;
Fukunishi, Y. Japan Patent 2010/062402 (PCT WO 2011010715).
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 2, 2013
417