Synthesis and Biological Investigation of δ12-Prostaglandin J3 (δ12-PGJ3) Analogues and Related Compounds

Article abstract of DOI:10.1021/jacs.6b02075

A series of δ¹²-prostaglandin J₃ (δ¹²-PGJ₃) analogues and derivatives were synthesized employing an array of synthetic strategies developed specifically to render them readily available for biological investigations. The synthesized compounds were evaluated for their cytotoxicity against a number of cancer cell lines, revealing nanomolar potencies for a number of them against certain cancer cell lines. Four analogues (2, 11, 21, and 27) demonstrated inhibition of nuclear export through a covalent addition at Cys528 of the export receptor Crm1. One of these compounds (i.e., 11) is currently under evaluation as a potential drug candidate for the treatment of certain types of cancer. These studies culminated in useful and path-pointing structure-activity relationships (SARs) that provide guidance for further improvements in the biological/pharmacological profiles of compounds within this class.

Full text of DOI:10.1021/jacs.6b02075

Article
pubs.acs.org/JACS
Synthesis and Biological Investigation of Δ12-Prostaglandin J3 (Δ12-
PGJ3) Analogues and Related Compounds
K. C. Nicolaou,*,† Kiran Kumar Pulukuri,†,§ Stephan Rigol,†,§ Philipp Heretsch,†,§ Ruocheng Yu,†
Charles I. Grove,† Christopher R. H. Hale,† Abdelatif ElMarrouni,† Verena Fetz,‡ Mark Bronstrup,‡
̈
Monette Aujay,# Joseph Sandoval,# and Julia Gavrilyuk#
†Department of Chemistry, BioScience Research Collaborative, Rice University, 6100 Main Street, Houston, Texas 77005, United
States
‡Department of Chemical Biology, Helmholtz Centre for Infection Research, Inhoffenstraße 7, 38124 Braunschweig, Germany
#Stemcentrx Inc., 450 East Jamie Court, South San Francisco, California 94080, United States
S
* Supporting Information
ABSTRACT: A series of Δ12-prostaglandin J3 (Δ12-PGJ3)
analogues and derivatives were synthesized employing an array
of synthetic strategies developed specifically to render them
readily available for biological investigations. The synthesized
compounds were evaluated for their cytotoxicity against a
number of cancer cell lines, revealing nanomolar potencies for
a number of them against certain cancer cell lines. Four
analogues (2, 11, 21, and 27) demonstrated inhibition of nuclear
export through a covalent addition at Cys528 of the export
receptor Crm1. One of these compounds (i.e., 11) is currently under evaluation as a potential drug candidate for the treatment of
certain types of cancer. These studies culminated in useful and path-pointing structure−activity relationships (SARs) that provide
guidance for further improvements in the biological/pharmacological profiles of compounds within this class.
improved over those of Δ12-PGJ3 (1), providing motivation for
1. INTRODUCTION
As described in a separate article1 and a previous communica-
further molecular designs along these structural types.
2.2. Synthesis of Δ12-PGJ3 Ester (3−5 and 10), Amide
tion,2 we developed a number of synthetic strategies for the
Analogues (6−9) and Hydroxy Derivative 12. Ester
synthesis of the reported antileukemic agent Δ12-prostaglandin J3
derivatives of Δ12-PGJ3 (i.e., 3−5 and 10, Scheme 2) were
(Δ12-PGJ3, 1, Figure 1).3,4 In this article, we describe the
synthesized through their TBS−ether precursors (46−49,
application of these synthetic technologies for the synthesis of a
respectively) prepared from TBS−ether 452 by EDCI/DMAP-
series of designed analogues and derivatives of this newly
facilitated esterification as summarized in Scheme 2. The latter
discovered prostanoid whose biological profile3,4 elevated it to a
were desilylated by exposure to aq HF as shown in Scheme 2. In
lead compound for drug discovery. We also describe the results
addition to the esters and in a similar manner, we synthesized a
of a number of biological investigations with these compounds,
series of amide analogues (6−9, Scheme 3) from TBS−ether 452
including their evaluation as cytotoxic agents against a series of
by standard methods (Boc2O/NH4HCO3 or EDCI/HOBt) via
cancer cell lines, as well as aspects of their mechanism of action.
their TBS−ether derivatives (50−53, respectively, Scheme 3).
Our motivation for the design and synthesis of the targeted
The C1-hydroxy analogue 12 was also prepared from OPMB
molecules (2−44, ent-1, ent-2, ent-11, Figure 1) was to improve
the stability and potency of the parent compound (Δ12-PGJ3, 1)
and to discover new molecular entities for further development
as potential drug candidates for cancer chemotherapy.
derivative 542 through desilylation (aq HF) and deprotection
(DDQ) via intermediate 55 as shown in Scheme 4.
2.3. Synthesis of ent-Δ12-PGJ3 Analogues ent-1, ent-2,
and ent-11. To test the importance of the absolute
configuration of Δ12-PGJ3 (1) to its biological activity, we set
out to prepare its enantiomer (ent-Δ12-PGJ3, ent-1) and its
derivatives, methyl ester (ent-2) and lactone (ent-11). As seen in
Scheme 5, we took advantage of our menthol-chiral approach1 to
this structural type to generate the required enantiomeric enone
ent-57 (generated from 56). ent-57 was converted, through an
2. RESULTS AND DISCUSSION
2.1. Synthesis of Δ12-PGJ3 Methyl Ester 2 and Lactone
11. In an effort to improve membrane permeability and stability,
we synthesized the methyl ester (2) and lactone (11) derivatives
of Δ12-PGJ3 through the use of TMSCHN2 (93% yield) and
MNBA/DMAP (Shiina method,6 71% yield), respectively, as
summarized in Scheme 1. As will be discussed below, both the
chemical stability and potency of derivatives 2 and 11 were
Received: February 24, 2016
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 1. Molecular structures of Δ12-prostaglandin J3 (Δ12-PGJ3, 1) and designed analogues 2−44, ent-1, ent-2, and ent-11.
Scheme 1. Synthesis of Δ12-PGJ3 Methyl Ester 2 and Lactone
Scheme 2. Synthesis of Δ12-PGJ3 Ester Analogues 3−5 and
a
a
11
10
a
Reagents and conditions: (a) TMSCHN2 (2 M in Et2O, 1.5 equiv),
a
3:2 C6H6/MeOH, 25 °C, 30 min, 93%; (b) MNBA (1.4 equiv),
DMAP (6.0 equiv), CH2Cl2, 25 °C, 17 h, 74%; MNBA = 2-methyl-6-
nitrobenzoic anhydride, DMAP = 4-dimethylaminopyridine.
Reagents and conditions: (a) 46: EDCI (2.0 equiv), DMAP (0.05
equiv), i-PrOH (1.5 equiv), CH2Cl2, 0 °C, 6 h, 79%; 47: EDCI (2.0
equiv), DMAP (0.05 equiv), C10H21OH (1.5 equiv), CH2Cl2, 0 to 25
°C, 12 h, 76%; 48: EDCI (2.0 equiv), DMAP (0.05 equiv), C20H41OH
(1.5 equiv), CH2Cl2, 0 to 25 °C, 12 h, 71%; 49: EDCI (1.5 equiv),
DMAP (0.1 equiv), PhSO2(CH2)2OH (1.5 equiv), CH2Cl2, 0 °C, 6 h,
86%; (b) 3: HF·NEt3 (50 equiv), MeCN, 0 to 25 °C, 48 h, 80%; 4: HF
(50% aq, 100 equiv), MeCN, 0 °C, 30 min, 79%; 5: HF (50% aq, 100
equiv), MeCN, 0 °C, 30 min, 76%; 10: HF (50% aq, 50 equiv),
MeCN, 0 °C, 30 min, 76%; EDCI = N-[3-(dimethylamino)propyl]-
N′-ethylcarbodiimide.
aldol reaction with enantiomeric aldehyde ent-58,5 to product 59
(mixture of diastereomers), which was transformed (MsCl,
DMAP) to cross-dienone ent-54 (single isomer) in 49% yield for
the two steps. The rest of the synthesis of ent-1, ent-2, and ent-11
proceeded through the same steps and in similar yields as
described in a separate article for the synthesis of 11 and
summarized here in Scheme 5 (through intermediates 60−ent-
45).
shown in Scheme 6. Thus, intermediate 57 (obtained from the
corresponding menthol-enone through DIBAL-H reduction1)
was deprotected (DDQ, 94% yield), the resulting hydroxy enone
(62) was silylated (TBSCl, 89% yield), and the resulting product
(63) was coupled with conjugated aldehyde 645 under the aldol
conditions (LDA, −78 °C) to afford a mixture of alcohols (65).
The latter was treated with MsCl in the presence of DMAP (0 to
2.4. Synthesis of Dehydrated Δ12,14-PGJ3 Analogue 13
and Its Methyl Ester 14. To illuminate the effect of higher
conjugation of the enone system of Δ12-PGJ3, and for
comparison reasons, we prepared the dehydrated counterparts
of the parent compounds 1 and 2, namely 15-dehydroxy-Δ12,14
-
PGJ3 (13) and its methyl ester analogue 14 from enone 571 as
B
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 3. Synthesis of Δ12-PGJ3 Amide Analogues 6−9
Scheme 5. Synthesis of ent-Δ12-PGJ3 Analogues ent-1, ent-2,
a
and ent-11
a
Reagents and conditions: (a) Boc2O (2.6 equiv), (NH4)HCO3 (2.4
equiv), pyridine (1.2 equiv), CH2Cl2, 0 to 25 °C, 6 h, 68%; (b) 51:
EDCI (2.0 equiv), HOBt (2.0 equiv), H2NMe (2.0 M in THF, 2.0
equiv), CH2Cl2, 0 to 25 °C, 2 h, 78%; 52: EDCI (2.0 equiv), HOBt
(2.0 equiv), H2Ni-Pr (1.0 M in THF, 2.0 equiv), CH2Cl2, 0 to 25 °C, 2
h, 71%; 53: EDCI (1.5 equiv), HOBt (1.5 equiv), HNMe2 (1.0 M in
THF, 1.2 equiv), CH2Cl2, 0 to 25 °C, 3 h, 71%; (c) 6: HF (50% aq,
100 equiv), MeCN, 0 °C, 30 min, 68%; 7: HF (50% aq, 100 equiv),
MeCN, 0 °C, 30 min, 78%; 8: HF (50% aq, 100 equiv), MeCN, 0 °C,
30 min, 75%; 9: HF (50% aq, 100 equiv), MeCN, 0 °C, 30 min, 84%;
Boc2O = di-tert-butyl dicarbonate, HOBt = hydroxybenzotriazole.
a
Scheme 4. Synthesis of Δ12-PGJ3 Hydroxy Analogue 12
a
Reagents and conditions: (a) DIBAL-H (1.5 equiv), CH2Cl2, −10 °C,
30 min, 76%; (b) LDA (2.0 equiv); then ent-57 (1.0 equiv); then ent-
58 (1.2 equiv), THF, −78 °C, 30 min; (c) MsCl (2.0 equiv), DMAP
(10 equiv), CH2Cl2, 0 to 25 °C, 6 h, 49% for two steps; (d) DDQ (1.5
equiv), 16:1 CH2Cl2/H2O, 0 °C, 45 min, 87%; (e) PCC (2.0 equiv),
CH2Cl2, 25 °C, 2 h; (f) NaClO2 (3.0 equiv), NaH2PO4 (3.0 equiv), 2-
methyl-2-butene (10 equiv), t-BuOH, 25 °C, 30 min, 86% for two
steps; (g) HF (50% aq, 100 equiv), MeCN, 0 °C, 45 min, 92%; (h)
TMSCHN2 (2 M in Et2O), 3:2 C6H6/MeOH, 25 °C, 30 min, 93%; (i)
MNBA (1.4 equiv), DMAP (6.0 equiv), CH2Cl2, 25 °C, 17 h, 74%;
DIBAL-H = diisobutylaluminum hydride, LDA = lithium diisopropy-
lamide, MsCl = methanesulfonyl chloride, PCC = pyridinium
chlorochromate.
a
Reagents and conditions: (a) HF (50% aq, 50 equiv), MeCN, 0 °C, 2
h, 87%; (b) DDQ (1.5 equiv), 4:1 CH2Cl2/H2O, 0 °C, 1 h, 92%;
DDQ = 2,3-dichloro-5,6-dicyano-1,4-benzoquinone.
25 °C) to furnish conjugated enone 66 in 31% overall yield for
the two steps. Desilylation of enone 66 (aq HF, 87% yield) was
7
followed by stepwise oxidation (PCC, 68; NaClO2 ) to furnish
the targeted 15-deoxy-Δ12,14-PGJ3 (13), from which methyl ester
derivative 14 was generated by exposure to TMSCHN2 (90%
yield).
2.5. Synthesis of 15-Deoxy-Δ12-PGJ3 (15) and Its
Methyl Ester 16. The 15-deoxy-Δ12-PGJ3 analogues 15 and
16 (Scheme 7) were synthesized in order to test the importance
of the 15-hydroxy group of Δ12-PGJ3 for biological activity. To
this end, and as shown in Scheme 7, enone 571 was reacted with
aldehyde fragment 695 under our aldol conditions (LDA, −78
°C), and the resulting mixture of alcohols (70) was exposed to
the action of MsCl and DMAP to afford enone PMB derivative
71 in 61% overall yield for the two steps. The latter was oxidized
fully and in one step by treatment with oxo-piperidinium
tetrafluoroborate salt 728 to afford, in 45% yield, targeted
analogue 15. Treatment of the latter with TMSCHN2 furnished
methyl ester analogue 16 (90% yield).
2.8. Synthesis of Alkyne Δ12-PGJ3 Analogue 22, Its
Methyl Ester 23, and Lactone 24. In a similar fashion,
acetylenic Δ12-PGJ3 analogues 22 and 23 were synthesized from
key building blocks 86,5 and 571 through the standard sequence
via intermediates 88−93 as presented in Scheme 10. The 1,15-
lactone 24 was prepared from hydroxy acid 22 by exposure to
MNBA and DMAP as shown in Scheme 10.
2.9. Synthesis of Δ12-PGJ3 Analogue 25, Its Methyl
Ester 26, and Lactone 27. The synthesis of Δ12-PGJ3
analogues 25−27 carrying a trifluoromethyl phenyl residue at
the end of the “lower” chain from building blocks 57,1 lactol 94,12
and phosphonium salt 9513 is shown in Scheme 11. Thus, Wittig
reaction of the ylide derived from 95 with lactol 94 (n-BuLi) gave
an inconsequential mixture of olefinic isomers 96 [(E)/(Z) =
65:35, 72% yield], reduction of which (10% Pd/C, H2) led to
primary alcohol 97 (91% yield). The latter was oxidized (DMP,
83% yield) to aldehyde 98, which was coupled with enone 57
through an aldol reaction to afford coupling product 99 (mixture
of diastereomers), whose elaboration to analogues 25−27
proceeded through the standard sequence, and intermediates
100−103 as summarized in Scheme 11.
2.6. Synthesis of Δ12-PGJ3 Analogues 17 and 18. To test
the effect of “lower” side chain modifications, truncated Δ12-PGJ3
analogues 17 and 18 (TMSCHN2) were synthesized from key
building blocks, enone 571 and propionaldehyde (73), via
intermediates 74 and 75 as illustrated in Scheme 8.
2.7. Synthesis of Δ12-PGJ3 Analogue 19, Its Methyl
Ester 20, and Lactone 21. C20-trifluoro Δ12-PGJ3 analogues
19−21 were prepared from starting materials 76,9 77,10 and 571
through intermediates 78−85 by standard methods as
summarized in Scheme 9. The motivation for this modification
was the well-known benefits of fluorine residues not only to
pharmacological properties11 but also to block metabolic
oxidation at C20.
2.10. Synthesis of Δ12-PGJ3 Analogue 28, Its Methyl
Ester 29, and Lactone 30. Scheme 12 summarizes the
synthesis of Δ12-PGJ3 analogues 28−30. Inspired by docosahex-
aenoic acid (DHA) and a hypothesis for the plausible existence of
28 as a naturally existing substance, these analogues were chosen
C
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Scheme 6. Synthesis of 15-Deoxy-Δ12,14-PGJ3 Analogue 13
Scheme 8. Synthesis of Δ12-PGJ3 Analogues 17 and 18
a
and Its Methyl Ester 14
a
Reagents and conditions: (a) LDA (2.0 equiv), THF, −78 °C; then
57 (1.0 equiv); then C2H5CHO (1.5 equiv), 15 min; (b) MsCl (2.0
equiv), DMAP (10 equiv), CH2Cl2, 0 to 25 °C, 6 h, 63% for two steps;
(c) 4-(acetylamino)-2,2,6,6-tetramethyl-1-oxo-piperidinium tetrafluor-
oborate (6.0 equiv), 9:1 MeCN/H2O, 25 °C, 30 min, 57%; (d)
TMSCHN2 (2 M in Et2O, 1.5 equiv), 3:2 C6H6/MeOH, 25 °C, 30
min, 93%.
a
Reagents and conditions: (a) DDQ (1.5 equiv), 20:1 CH2Cl2/H2O, 0
°C, 2 h, 94%; (b) TBSCl (1.5 equiv), imid. (3.0 equiv), CH2Cl2, 0 to
25 °C, 4 h, 89%; (c) LDA (2.0 equiv), THF, −78 °C; then 63 (1.0
equiv); then 64 (1.2 equiv), 15 min; (d) MsCl (2.0 equiv), DMAP (10
equiv), CH2Cl2, 0 to 25 °C, 6 h, 31% for two steps; (e) HF (50% aq,
50 equiv), MeCN, 0 °C, 45 min, 87%; (f) PCC (2.0 equiv), CH2Cl2,
25 °C, 2 h; (g) NaClO2 (1.5 equiv), NaH2PO4 (1.5 equiv), 2-methyl-
2-butene (10 equiv), t-BuOH, 25 °C, 30 min, 67% for two steps; (h)
TMSCHN2 (2 M in Et2O, 1.5 equiv), 3:2 C6H6/MeOH, 25 °C, 30
min, 90%; TBSCl = tert-butyldimethylsilyl chloride, imid. = 1H-
imidazole.
Scheme 9. Synthesis of Δ12-PGJ3 Analogue 19, Its Methyl
a
Ester 20, and Lactone 21
Scheme 7. Synthesis of 15-Deoxy-Δ12-PGJ3 (15) and Its
a
Methyl Ester 16
a
Reagents and conditions: (a) LDA (2.0 equiv), THF, −78 °C; then
57 (1.0 equiv); then 69 (1.2 equiv), 15 min; (b) MsCl (2.0 equiv),
DMAP (10 equiv), CH2Cl2, 0 to 25 °C, 6 h, 61% for two steps; (c) 4-
(acetylamino)-2,2,6,6-tetramethyl-1-oxo-piperidinium tetrafluorobo-
rate (6.0 equiv), 9:1 MeCN/H2O, 25 °C, 30 min, 45%; (d)
TMSCHN2 (2 M in Et2O, 2.0 equiv), 3:2 C6H6/MeOH, 25 °C, 30
min, 90%.
a
Reagents and conditions: (a) 3,3,3-trifluoropropylphosphonium
iodide (2.0 equiv), NaHMDS (1.9 equiv), 0 to 25 °C, 1 h; then 76,
−78 to 25 °C, THF, 5 h, 88%; (b) DDQ (1.5 equiv), 20:1 CH2Cl2/
H2O, 0 to 25 °C, 5 h, 91%; (c) DMP (2.0 equiv), CH2Cl2, 0 to 25 °C,
2 h, 76%; (d) LDA (2.0 equiv), THF, −78 °C; then 57 (1.0 equiv);
then 80 (1.2 equiv), 15 min; (e) MsCl (2.0 equiv), DMAP (10 equiv),
CH2Cl2, 0 to 25 °C, 6 h, 43% for two steps; (f) DDQ (1.5 equiv), 20:1
CH2Cl2/H2O, 0 °C, 2 h, 92%; (g) PCC (2.0 equiv), CH2Cl2, 25 °C, 2
h; (h) NaClO2 (3.0 equiv), NaH2PO4 (3.0 equiv), 2-methyl-2-butene
(30 equiv), t-BuOH, 25 °C, 30 min, 84% for two steps; (i) HF (50%
aq, 100 equiv), MeCN, 0 °C, 45 min, 90%; (j) TMSCHN2 (2 M in
Et2O, 1.5 equiv), 3:2 C6H6/MeOH, 25 °C, 30 min, 60%; (k) MNBA
(1.4 equiv), DMAP (6.0 equiv), CH2Cl2, 25 °C, 17 h, 62%; NaHMDS
= sodium bis(trimethylsilyl)amide, DMP = Dess−Martin periodinane.
for biological comparison with their counterpart Δ12-PGJ3
compounds. The first task was the construction of key building
block 112 (Scheme 12a). Its synthesis began with TMS−
acetylene which reacted in the presence of n-BuLi with readily
available epoxide 10414 to afford alcohol 105 (88%), whose
desilylation (TBAF) led to terminal acetylene 106 (94% yield).
Coupling of 106 with propargyl bromide 107 in the presence of
CuI and K2CO3 furnished bis-acetylene 108, which was
selectively reduced with NaBH4−H2 in the presence of
Ni(OAc)2·4H2O15 to give bis-olefin 109 with both olefinic
bonds formed in the desired (Z)-geometry16 (77% overall yield
from 106). The hydroxy group of the latter intermediate was
protected as a TBS−ether (TBSCl, 91% yield), and the resulting
product (110) was sequentially treated with DDQ (66% yield)
and DMP (80% yield) to afford required aldehyde 112 through
intermediate 111. The construction of the other required
fragment, enone 118 (containing one less carbon in its side chain
than the corresponding enone for Δ12-PGJ3), its coupling with
segment 112, and elaboration of the product to the targeted
analogues 28−30 are shown in Scheme 12b.
Thus, cyclopentene methyl ester 1131 (inconsequential
mixture of diastereomers) was reduced with DIBAL to afford
aldehyde 114, which was reacted with the ylide derived from
phosphonium salt 115 (NaHMDS) to furnish, after desilylation
D
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 10. Synthesis of Alkyne Δ12-PGJ3 Analogue 22, Its
Scheme 11. Synthesis of Δ12-PGJ3 Analogue 25, Its Methyl
a
a
Methyl Ester 23, and Lactone 24
Ester 26, and Lactone 27
a
Reagents and conditions: (a) DDQ (1.5 equiv), 20:1 CH2Cl2/H2O, 0
to 25 °C, 2 h, 90%; (b) DMP (1.5 equiv), CH2Cl2, 0 to 25 °C, 2 h,
88%; (c) LDA (2.0 equiv), THF, −78 °C; then 57 (1.0 equiv); then
88 (1.2 equiv), 15 min; (d) MsCl (2.0 equiv), DMAP (10 equiv),
CH2Cl2, 0 to 25 °C, 6 h, 36% for two steps; (e) DDQ (2.0 equiv),
10:1 CH2Cl2/H2O, 0 °C, 2 h, 98%; (f) PCC (2.0 equiv), CH2Cl2, 25
°C, 40 min.; (g) NaClO2 (1.5 equiv), NaH2PO4 (1.5 equiv), 2-methyl-
2-butene (10 equiv), t-BuOH, 25 °C, 30 min, 80% for two steps; (h)
HF (50% aq, 100 equiv), MeCN, 0 °C, 30 min, 74%; (i) TMSCHN2
(2 M in Et2O, 2.0 equiv), 3:2 C6H6/MeOH, 25 °C, 0.5 h, 90%; (j)
MNBA (1.4 equiv), DMAP (6.0 equiv), CH2Cl2, 25 °C, 17 h, 50%.
a
Reagents and conditions: (a) n-BuLi (1.5 equiv), 95 (1.5 equiv),
THF, −78 °C, 2 h; then 94 (1.0 equiv), −78 °C, 30 min, 40 °C, 24 h,
72% [(E)/(Z) = 65:35]; (b) 10% Pd/C (10 wt%), H2, MeOH, 3 h,
91%; (c) DMP (2.0 equiv), CH2Cl2, 25 °C, 1 h, 83%; (d) LDA (2.0
equiv), THF, −78 °C; then 57 (1.0 equiv); then 98 (1.2 equiv), 15
min; (e) MsCl (2.0 equiv), DMAP (10 equiv), CH2Cl2, 0 to 25 °C, 6
h, 45% for two steps; (f) DDQ (1.5 equiv), 20:1 CH2Cl2/H2O, 0 °C, 2
h, 86%; (g) PCC (2.0 equiv), CH2Cl2, 25 °C, 2 h; (h) NaClO2 (3.0
equiv), NaH2PO4 (3.0 equiv), 2-methyl-2-butene (30 equiv), t-BuOH,
25 °C, 30 min, 72% for two steps; (i) HF (50% aq, 100 equiv), MeCN,
0 °C, 45 min, 84%; (j) TMSCHN2 (2 M in Et2O, 2.0 equiv), 3:2
C6H6/MeOH, 25 °C, 0.5 h, 86%; (k) MNBA (1.4 equiv), DMAP (6.0
equiv), CH2Cl2, 25 °C, 17 h, 61%.
(TBAF), hydroxy olefin 117 (mixture of diastereomers, 68%
overall yield for the three steps from 113). PCC oxidation of 117
led to enone 118 (92% yield) whose aldol reaction with aldehyde
112 furnished hydroxy enone 119 (mixture of diastereomers).
Exposure of the latter to MsCl and Et3N followed by treatment of
the resulting mixture of mesylates (120) to Al2O3 gave the
desired cross-conjugated dienone 121 in 30% overall yield for the
three steps from 118. Removal of the PMB group from 121
(DDQ, 66% yield) followed by PCC oxidation led to aldehyde
123, whose further oxidation with NaClO2 afforded TBS−ether
carboxylic acid 124 in 78% overall yield from 122. Desilylation of
the latter (aq HF) afforded analogue 28, from which methyl ester
29 was prepared by exposure to TMSCHN2 (72% yield).
Cyclization of 28 in the presence of MNBA and DMAP afforded
lactone 30 (62% yield). Incidentally, compound 28 may found to
be a naturally occurring secondary metabolite formed enzymati-
cally from DHA by analogy to Δ12-PGJ3, which originates from
EPA.
treated with MsCl and DMAP to yield cross-conjugated
chloroenone 129 in 45% yield for the two steps. Removal of
the PMB group (DDQ, 86% yield) followed by sequential
oxidation with PCC and NaClO2 gave carboxylic acid 132 via
aldehyde 131 in 84% yield for the two steps. Finally, the targeted
compound 10-chloro-Δ12-PGJ3 (31, aq HF, 88% yield), its
methyl ester (32, TMSCHN2, 56% yield) and lactone (33,
MNBA, DMAP, 73% yield) were generated from 132 by the
standard conditions as shown in Scheme 13.
2.11. Synthesis of Cl-Δ12-PGJ3 Analogue 31, Its Methyl
Ester 32, and Lactone 33. The 10-chloro series of analogues
(31−33, Scheme 13) were designed expecting possible improve-
ments in membrane permeability and potency. Their synthesis is
shown in Scheme 13. Thus, PMB−ether enone 571 was
selectively epoxidized with H2O2 and KOH to afford epoxy
ketone 125 whose exposure to LiCl and Amberlyst-15 furnished
a mixture of PMB−ether chloroenone 126 (24% yield) and
hydroxy chloroenone 127 (28% yield), the latter being easily
converted to the former through the action of PMB-
trichloroacetimidate and Sc(OTf)3 (80% yield). Chloroenone
126 underwent aldol reaction with aldehyde 58 (LDA, −78 °C),
and the resulting mixture of alcohols (128) so formed was
2.12. Synthesis of 8-Methyl-Δ12-PGJ3 Methyl Esters 34
and 36, and Lactones 35 and 37. Scheme 14 depicts the
synthesis of the 8-methyl-Δ12-PGJ3 methyl ester (34 and 36) and
lactone (35 and 37) analogues. The required 8-methyl (PG
numbering) enone 137 was prepared from (+)-menthol
vinylogous ester 1331 through a three-step sequence involving
methylation (LDA, DMI, MeI, 76% yield), alkylation with allylic
bromide 135 (LDA, DMI, 73% yield), and DIBAL-H-induced
removal of the chiral auxiliary (76% yield) affording racemic 137
via intermediate 136 (ca. 55:45 dr). Processing enone 137
through the standard aldol-mesylation/elimination-oxidation
sequence as summarized in Scheme 14 led to TBS−ether
carboxylic acid 141 (mixture of diastereomers ca. 66:33 dr). The
E
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 12. Synthesis of Δ12-PGJ3 Analogue 28, Its Methyl
Scheme 13. Synthesis of Cl-Δ12-PGJ3 Analogue 31, Its Methyl
a
a
Ester 29, and Lactone 30
Ester 32, and Lactone 33
a
Reagents and conditions: (a) H2O2 (2.0 equiv), KOH (10% in H2O,
0.2 equiv), MeOH, −20 °C, 6 h; (b) LiCl (10.0 equiv), Amberlyst-15
(200 wt%), MeCN, 25 °C, 24 h, 126 (24%) and 127 (28%); (c)
PMBOC(=NH)CCl3 (2.0 equiv), Sc(OTf)3 (0.1 equiv), CH2Cl2, 25
°C, 4 h, 80%; (d) LDA (2.0 equiv), THF, −78 °C; then 126 (1.0
equiv); then 58 (1.3 equiv), 15 min; (e) MsCl (2.0 equiv), DMAP (10
equiv), CH2Cl2, 0 to 25 °C, 12 h, 45% for two steps; (f) DDQ (2.0
equiv), 20:1 CH2Cl2/H2O, 0 °C, 2 h, 86%; (g) PCC (2.0 equiv),
CH2Cl2, 25 °C, 2 h; (h) NaClO2 (3.0 equiv), NaH2PO4 (3.0 equiv), 2-
methyl-2-butene (30 equiv), t-BuOH, 25 °C, 30 min, 84% for two
steps; (i) HF (50% aq, 100 equiv), MeCN, 0 °C, 45 min, 88%; (j)
TMSCHN2 (2 M in Et2O, 2.0 equiv), 3:2 C6H6/MeOH, 25 °C, 30
min, 56%; (k) MNBA (1.5 equiv), DMAP (4.0 equiv), CH2Cl2, 25 °C,
12 h, 73%.
a
Panel (a): Synthesis of aldehyde 112: Reagents and conditions: (a)
TMS−acetylene (1.5 equiv), n-BuLi (1.3 equiv), BF3·Et2O (1.3 equiv),
−78 to 25 °C, 2 h, 88%; (b) TBAF (1.0 M in THF, 1.2 equiv), THF, 0
to 25 °C, 2 h, 94%; (c) 107 (1.2 equiv), K2CO3 (1.3 equiv), CuI (1.3
equiv), NaI (1.3 equiv), DMF, 25 °C, 15 h; (d) Ni(OAc)2·4H2O (0.32
equiv), NaBH4 (0.77 equiv), 1,2-diaminoethane (3.6 equiv), H2,
EtOH, 25 °C, 18 h, 77% for two steps; (e) TBSCl (1.5 equiv), imid.
(3.0 equiv), CH2Cl2, 12 h, 91%; (f) DDQ (1.5 equiv), 10:1 CH2Cl2/
H2O, 0 °C, 105 min, 66%; (g) DMP (1.5 equiv), CH2Cl2, 0 to 25 °C,
1.5 h, 80%. Panel (b): Synthesis of Δ12-PGJ3 analogue 28, methyl ester
29, and lactone 30: Reagents and conditions: (a) DIBAL-H (1.1
equiv), CH2Cl2, −78 °C, 45 min; (b) 115 (1.5 equiv), NaHMDS (2.0
equiv), THF, 0 to 25 °C, 15.5 h; (c) TBAF (1.2 equiv), THF, 0 to 25
°C, 3 h, 68% for three steps; (d) PCC (1.9 equiv), CH2Cl2, 25 °C, 2 h,
92%; (e) LDA (2.2 equiv), then 112 (1.3 equiv), THF, −78 °C, 30
min; (f) MsCl (5.0 equiv), Et3N (10 equiv), CH2Cl2, 0 °C, 30 min;
(g) Al2O3 (21 equiv), CH2Cl2, 25 °C, 6 h, 30% for three steps; (h)
DDQ (1.7 equiv), 10:1 CH2Cl2/H2O, 0 °C, 105 min, 66%; (i) PCC
(2.0 equiv), CH2Cl2, 25 °C, 75 min; (j) 2-methyl-2-butene (10 equiv),
NaH2PO4 (1.5 equiv), NaClO2 (1.5 equiv), t-BuOH, 25 °C, 30 min,
78% for two steps; (k) aq HF (ca. 100 equiv), MeCN, 0 °C, 75 min,
81%; (l) TMSCHN2 (2 M in Et2O, 2.0 equiv), 3:2 C6H6/MeOH, 25
°C, 30 min, 72%; (m) MNBA (1.4 equiv), DMAP (6.0 equiv),
CH2Cl2, 25 °C, 17 h, 62%; TBAF = tetra-n-butylammonium fluoride.
65:35 dr) whose chromatographic separation proved challeng-
ing. The difficulties of separating the C8-epimers of carboxylic
acid 142 were reduced upon methylation (TMSCHN2, 82%
yield) to produce methyl ester analogues 34 (57%) and 36 (25%)
(chromatographically separated on silica gel) or upon lactone
formation (MNBA, DMAP, 59% yield) to afford lactones 35
(38%) and 37 (21%) (chromatographically separated on silica
gel). The absolute configuration of the C8-stereocenter in this
series of compounds was not discernible from the NMR spectral
data and, therefore, the depicted structures should be considered
interchangeable (i.e., 34 or 36; 35 or 37).
2.13. Synthesis of 15-Fluoro-Δ12-PGJ3 (38) and Its
Methyl Ester 39. The 15-fluoro-Δ12-PGJ3 analogues 38 and 39
were prepared from enone 571 and aldehyde ent-58,5 the latter
chosen so as to allow for the anticipated inversion of
configuration at the fluorination stage as shown in Scheme 15.
Building blocks 571 and ent-585 were united through the
standard aldol protocol to afford diastereomeric mixture 143,
which was converted to cross-conjugated enone 144 through
sequential treatment with MsCl, Et3N and Al2O3. The latter was
desilylated (3HF·Et3N) to give alcohol 145 whose exposure to
PhenoFluor at 80 °C furnished fluoride 146 (35% yield) with
(presumed) inversion of configuration at C15.17 Precursor 146
was converted in one step to the targeted 15-fluoro-Δ12-PGJ3
analogue 38 with piperidinium tetrafluoroborate 728 in 45%
yield. 15-Fluoro-Δ12-PGJ3 methyl ester (39) was prepared from
sluggishness and low yield of the aldol reaction in this case is
attributed to steric hindrance introduced by the C8 quaternary
center adjacent to the reactive site. Desilylation of 141 led to
hydroxy acid 142 as a mixture of diastereomers (75% yield, ca.
F
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 14. Synthesis of 8-Methyl-Δ12-PGJ3 Methyl Esters 34
Scheme 15. Synthesis of 15-Fluoro-Δ12-PGJ3 (38) and Its
a
a
and 36, and Lactones 35 and 37
Methyl Ester 39
a
Reagents and conditions: (a) LDA (2.0 equiv), THF, −78 °C; then
57 (1.0 equiv), 30 min; then ent-58 (1.2 equiv), 30 min, 86%; (b)
MsCl (5.0 equiv), Et3N (10 equiv), CH2Cl2, 0 °C, 5 min; (c) Al2O3
(42 equiv), CH2Cl2, 25 °C, 8 h, 65% for two steps; (d) 3HF·Et3N (23
equiv), THF, 0 to 25 °C, 3 d, 88%; (e) PhenoFluor (6.0 equiv), KF
(9.0 equiv), i-Pr2NEt (9.0 equiv), toluene, 80 °C, 2.5 h, 35%; (f) 4-
(acetylamino)-2,2,6,6-tetramethyl-1-oxo-piperidinium tetrafluorobo-
rate (6.0 equiv), 9:1 MeCN/H2O, 25 °C, 35 min, 45%; (g)
TMSCHN2 (2 M in Et2O, 1.5 equiv), 3:2 C6H6/MeOH, 25 °C, 35
min, 50%; PhenoFluor = 1,3-bis(2,6-diisopropylphenyl)-2,2-difluoro-
2,3-dihydro-1H-imidazole.
lines. Selected data are shown in Table 1 (for full data, see
Supporting Information). Compounds 1, 5, ent-1, 13, 17, 19, 22,
31, 34−36, and 40−43 did not advance beyond the one-dose
assay (10 μM) as they did not show significant cytotoxicities (see
Table 1), whereas the remaining compounds, having exhibited
promising potencies, were tested at lower concentrations. As
shown in Table 1, among the most potent of these analogues
against the tested cell lines were the following: dimethyl amide 9
(melanoma, GI50 = 0.356 μM); methyl esters 29 (melanoma,
GI50 = 0.210 μM), and 32 (leukemia, GI50 = 0.0893 μM);
lactones 11 (colon cancer, GI50 = 0.201 μM), ent-11 (colon
cancer, GI50 = 0.133 μM), 24 (breast cancer, GI50 = 0.203 μM),
a
Reagents and conditions: (a) LDA (1.05 equiv), THF, −78 °C; then
133 (1.0 equiv), 0.5 h, then MeI (1.3 equiv), DMI, 12 h, 76%; (b)
LDA (1.05 equiv), THF, −78 °C; then 134 (1.0 equiv); then 135 (1.3
equiv), DMI, 12 h, 73% (mixture of diastereomers, ca. 55:45 dr); (c)
DIBAL-H (1.5 equiv), CH2Cl2, −10 °C, 6 h, 76%; (d) LDA (2.0
equiv), THF, −78 °C; then 137 (1.0 equiv); then 58 (1.2 equiv), 30
min, 17% (57% brsm); (e) MsCl (2.0 equiv), DMAP (10 equiv),
CH2Cl2, 0 to 25 °C, 6 h, 67% for two steps (mixture of diastereomers,
ca. 67:33 dr); (f) DDQ (2.0 equiv), 20:1 CH2Cl2/H2O, 0 °C, 2 h, 88%
(mixture of diastereomers, ca. 68:32 dr); (g) PCC (2.0 equiv),
CH2Cl2, 25 °C, 2 h; (h) NaClO2 (3.0 equiv), NaH2PO4 (3.0 equiv), 2-
methyl-2-butene (10 equiv), t-BuOH, 25 °C, 30 min, 78% for two
steps (mixture of diastereomers, ca. 66:33 dr); (i) HF (50% aq, 50
equiv), MeCN, 0 °C, 45 min, 75% (mixture of diastereomers, ca. 65:35
dr); (j) TMSCHN2 (2 M in Et2O, 1.5 equiv), 3:2 C6H6/MeOH, 25
°C, 0.5 h, 82% (57% for 34; 25% for 36); (k) MNBA (1.4 equiv),
DMAP (6.0 equiv), CH2Cl2, 25 °C, 12 h, 59% (38% for 35; 21% for
37); DMI = 1,3-dimethyl-2-imidazolidinone.
33 (leukemia, GI50 = 0.0753 μM), and 44 (melanoma, GI50
=
0.084 μM). Interestingly, Δ12-PGJ3 macrolactone 11 demon-
strated no toxicity in athymic nude mice up to 200 mg/kg per
dose (intraperitoneally administered), a promising result that
allowed it to advance to in vivo efficacy studies in mice (NCI, see
Supporting Information).
A selected number of compounds were also assayed for their
cytotoxicities against cancer cell lines HEK 293T (human
embryonic kidney cell line), MES SA (uterine sarcoma cell line),
and MES SA DX (MES SA cell line with marked multidrug
resistance). The results are summarized in Table 2. The most
38 by treatment with TMSCHN2 (50% yield) as summarized in
Scheme 15.
Truncated compounds 40−43 (Figure 1) were also
synthesized (see Supporting Information) in order to test the
effects of shorter “top” side chains on the biological activity in the
series, while 17,18-dihydro lactone analogue 44 was constructed
(see Supporting Information) to probe the effect of another
“lower” side chain modification.
potent proved to be 10-chloro-Δ12-PGJ3 derivatives 32 (IC50
=
0.26 μM against HEK 293T) and 33 (IC50 = 0.32 μM against
MES SA DX); lactones 11 (IC50 = 0.074 μM against MES SA
DX) and 24 (IC50 = 0.31 μM against MES SA DX); amides 7
(IC50 = 0.56 μM against HEK 293T) and 9 (IC50 = 0.32 μM
against HEK 293T), and methyl ester 2 (IC50 = 0.57 μM against
MES SA).
2.15. Nuclear Export Studies. Compounds 2, 11, 21, and
27 were tested for their ability to inhibit nuclear export,
motivated by reported activities of the structurally related 15-
2.14. Biological Evaluation of the Δ12-PGJ3 Analogues.
The synthesized Δ12-PGJ3 (1) and Δ12-PGJ3 analogues (2−44,
ent-1, ent-2, and ent-11) were evaluated for their cytotoxicities
against the NIH-60 Screen panel of human cancer cell lines,
including leukemia, non-small cell lung, colon, central nervous
system, melanoma, ovarian, renal, breast, and prostate cancer cell
G
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a
Table 1. Selected NCI-60 Cytotoxicity Screen Human Cancer Cell Line Panel Data (GI50 in μM) for Δ12-PGJ3 (1) and Analogues
b
2−44, ent-1, ent-2, and ent-11
a
b
c
GI50 = Concentration of compound required to inhibit growth by 50%. See Supporting Information for complete NCI-60 screen data. Mean
d
growth % at 10 μM. Compound 37 was submitted to NCI after testing of compounds 34, 35, and 36 and was not chosen to be tested.
deoxy-Δ12,14-PGJ2.18 All compounds inhibited the nuclear export
of a translocation biosensor at a concentration of 50 μM (Figure
S1). This effect was confirmed by monitoring the intracellular
localization of the full length protein TFIIA fused to GFP (Figure
2A). The compounds exhibited comparable potencies to those
reported for 15-deoxy-Δ12,14-PGJ2 in both assays.18 The
inhibitory effect was completely revoked upon addition of
dithiothreitol (Figure S2), indicating a mechanism involving a
nucleophilic attack by a thiol (e.g., cysteine residue) on the cross-
conjugated enone moiety, most likely at the endocyclic C9
position.19 The export of both substrates, the translocation
biosensor and TFIIA, is mediated by the receptor Crm1. Cys528
of Crm1, located in the binding pocket for export substrates, is
the site of attack of covalently binding nuclear export inhibitors
like leptomycin B and ratjadones.20 To investigate whether 2, 11,
21, and 27 may also act through binding to Cys528, two peptides
were synthesized. The first one comprised residues 523−531 of
Crm1 (DLLGLCEQK), while the second contained a cysteine to
serine exchange at position 528 (DLLGLSEQK). For all four
compounds, covalent binding to DLLGLCEQK was detected by
H
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
values for esters 2−4 and 10 (compared to those for 1), with the
shorter O-chain (i.e., OMe, OiPr, compounds 2 and 3) being
optimal. These observations suggest cell permeability increase
imparted by the ester groups in compounds 2 and 3 as opposed
to the carboxylic acid moiety, although the length and steric
demand of the side-chain seems to also play a moderating role as
shown by the reversal of the trend with analogues 4, 5, and 10.
Amide groups at C1 also seem to be tolerated and, in some cases,
are even more potent than the corresponding methyl esters as
demonstrated by dimethyl amide analogue 9 (see Tables 1 and
2). The most significant potency enhancement, however, was
achieved through macrolactonization (C1 carboxylic acid with
C15 hydroxy moieties) as demonstrated within this series of
compounds (i.e., 11, ent-11, 24, 27, 30, 33, 37, and 44). These
results also demonstrate enhancement of potency in going from
the carboxylic acid to methyl ester to the macrolactone analogues
(e.g., ent-1 → ent-2 → ent-11; 19 → 20 → 21). Interestingly,
reduction of the carboxylic acid (1) to its primary alcohol
counterpart (i.e., 12) resulted in higher potencies than the parent
compound possibly due to increase of lipophilicity. It is plausible
that all these C1-carboxylate derivatives act as prodrugs,
generating the carboxylic acid once inside the cell, although
whether this group is essential for cytotoxicity is not known at
this time.
Table 2. Cytotoxicity Data Against Cancer Cell Lines HEK
a
293T, MES SA, and MES SA DX for Δ12-PGJ3 and Selected
b
Analogues (IC50 Values in μM)
The results with the enantiomeric Δ12-PGJ3 analogues ent-1
(acid), ent-2 (methyl ester), and ent-11 (lactone) are of interest
in that they demonstrate comparable cytotoxicity properties to
those of the natural absolute configurations (i.e., 1, 2, and 11),
except for the cases of ent-2 [showing lower potencies against
breast and ovarian cancer cells than its parent counterpart (2),
see Table 1] and ent-11 [which exhibited higher potencies than
its parent compound (i.e., 11) in the majority of the assays
performed (see Table 1)].
a
IC50 = 50% inhibitory concentration of compound against cell
growth; MES SA = uterine sarcoma cell line; MES SA DX = MES SA
cell line with marked multidrug resistance; HEK 293T = human
embryonic kidney cell line. See Supporting Information for further
details. These studies were carried out at Stemcentrx.
b
mass spectrometric analysis (Figures 2B, S3−S6, and Table
ST1). In contrast, no binding to DLLGLSEQK was observed.
Binding to Cys528 of Crm1 was validated in a cellular
environment by monitoring the decreased co-localization
between Crm1 and HIV-1Rev, but not between mutant
Crm1C528S and HIV-1Rev, upon compound treatment (Figure
2C). Similarly, expression of Crm1C528S-HA but not Crm1-HA
rescued nuclear export of the translocation biosensor upon
treatment with compound 2 (Figure S8). Thus, it was shown that
2, 11, 21, and 27 acted as inhibitors of nuclear export through
covalent bond formation at Cys528 of the export receptor Crm1.
The IC50 values representing the cytotoxic effects of the tested
compounds differ significantly from the concentrations needed
for inhibiting nuclear export. Because nuclear export inhibition
was determined after 1.5 h of incubation, whereas cytotoxicity
was assessed after 48 h of treatment, the higher cytotoxic potency
may arise from the extended time for the compound to exert an
effect. Nevertheless, nuclear export inhibition must account at
least in part for the compound’s cytotoxic effects, as this
consistency has been previously reported for various nuclear
export inhibitors (reviewed in ref. 21). It is also of note that Δ12-
PGJ3 was reported to activate the ATM/p53 pathway of
apoptosis in leukemia stem cells.3a This effect may also be
expected for the structurally related compounds described
herein, providing an additional mode of action for their cytotoxic
effects.
Of interest was also the fact that the C14−C15 dehydro-
derivative of Δ12-PGJ3 methyl ester analogue 14 exhibited
comparable cytotoxicities to those of the parent compound (i.e.,
2, see Table 1) but higher potencies than Δ12-PGJ3-C15-deoxy
analogue 16. More interesting was the observation that
truncation of the lower side chain by removing its C16−C20
segment and the C15-hydroxy group as in Δ12-PGJ3 methyl ester
analogue 18 resulted only in minor loss of activity compared to
that of Δ12-PGJ3 methyl ester (2). Even more gratifying was the
realization that replacement of the terminal methyl group of the
lower side chain with a trifluoromethyl group (i.e., compounds
19−21) led to an increase of potency in most of the tested cell
lines compared to those of the parent compounds (i.e., 1, 2, and
11, see Table 1). Furthermore, substitution of the C17−C18
olefinic bond with an acetylenic linkage (i.e., compounds 22−
24), installment of a m-trifluoromethyl aryl moiety (i.e.,
compounds 25−27), or replacement of the C15-hydroxy
group with a fluoride residue (i.e., compounds 38, 39) within
the Δ12-PGJ3 molecule led to similar or slightly higher potencies.
These data lead to the conclusion that a wide variety of
modifications of the lower side chain can be tolerated without
significant loss of activity. In fact, in some cases such changes may
lead to moderate increase in potency as demonstrated with
lactone 21 or even more significant enhancement of potency (see
below). In contrast to the lower side chain modifications,
changes in the top side chain such as truncation with attachment
point modification (to the cyclopentenone moiety, e.g.,
compounds 40−43) led to complete loss of activity as seen in
Table 1. A similar loss of potency was seen by introduction of a
2.16. Structure−Activity Relationships (SARs). From the
gathered biological data, a number of useful path-pointing
structure−activity relationships (SARs) were formed. Thus,
esterification at the C1 position of Δ12-PGJ3 led to increase of
potency as indicated by the GI50 (Table 1) and IC50 (Table 2)
I
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Figure 2. Compounds 2, 11, 21, and 27 inhibit Crm1-dependent nuclear export; HeLa cells transiently expressing TFIIA-GFP were treated with
compounds 2, 11, 21, and 27 (50 μM) for 1.5 h. The intracellular localization of TFIIA was evaluated by fluorescence microscopy after fixation (A). The
binding of 11 to a peptide comprising amino acids 523−531 of Crm1 (DLLGLCEQK; C-peptide) and a control peptide (DLLGLSEQK, S-peptide) was
evaluated by liquid chromatography−mass spectrometry. Depicted are extracted ion chromatograms for the doubly protonated ions of the peptide
(509.75 Da for C-peptide and 501.77 Da for S-peptide) and peptide−compound conjugate masses (666.85 Da for C-peptide−11 and 658.98 Da for S-
peptide−11) and the corresponding high resolution mass spectra (B); co-localization of Crm1 and HIV-1 Rev is disabled by the compounds (C). HeLa
cells were co-transfected with plasmids encoding for Crm1-GFP and HIV-1 Rev-BFP or Crm1C528S-GFP and HIV-1 Rev-BFP.22 The day after
transfection compounds 2, 11, 21, and 27 were applied at a concentration of 50 μM for 1.5 h. Cells were fixed and the localization of fluorescent signals
was analyzed by confocal microscopy.
methyl residue at the C8 position of Δ12-PGJ3 (i.e., compounds
34−37, see Tables 1 and 2).
A most notable enhancement of activity in this series of
compounds (see Figure 1) was observed upon introduction of a
chlorine residue at the C10 position of the Δ12-PGJ3 molecule
(analogues 31−33). Indeed, 10-chloro lactone 33 exhibited the
second highest potencies of the series in the NCI-60-Screen
proving itself more potent than 11. The most potent compound
within this series of analogues turned out to be lactone 44, which
lacks the Δ17,18-olefinic double bond in the lower side chain.
Figure 3 provides a summary of the structure−activity
relationships (SARs) as derived from these investigations.
J
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(2) Nicolaou, K. C.; Heretsch, P.; ElMarrouni, A.; Hale, C. R. H.;
Pulukuri, K. K.; Kudva, A. K.; Narayan, V.; Prabhu, K. S. Angew. Chem.,
Int. Ed. 2014, 53, 10443.
(3) (a) Hegde, S.; Kaushal, N.; Ravindra, K. C.; Chiaro, C.; Hafer, K.
T.; Gandhi, U. H.; Thompson, J. T.; van den Heuvel, J. P.; Kennett, M.
J.; Hankey, P.; Paulson, R. F.; Prabhu, K. S. Blood 2011, 118, 6909.
(b) Kudva, A. K.; Kaushal, N.; Mohinta, S.; Kennett, M. J.; August, A.;
Paulson, R. F.; Prabhu, K. S. PLoS One 2013, 8, e80622.
(4) Prabhu, K. S.; Paulson, R. F.; Hegde, S.; Kaushal, N.; Gandhi, U. H.
U.S. Patent US 8802680 B2, 2014.
(5) Nicolaou, K. C.; Heretsch, P.; Hale, C. R. H.; ElMarrouni, A.;
Pulukuri, K. K.; Yu, R.; Grove, C. U.S. Patent Application
WO2014US57426, 2015.
Figure 3. Structure−activity relationships (SARs) of Δ12-PGJ3 family of
compounds.
(6) Shiina, I.; Mukaiyama, T. Chem. Lett. 1994, 23, 677.
(7) Bal, B. S.; Childers, W. E., Jr.; Pinnick, H. W. Tetrahedron 1981, 37,
2091.
(8) Pradhan, P. P.; Bobbitt, J. M.; Bailey, W. F. J. Org. Chem. 2009, 74,
9524.
(9) Reddy, C. R.; Rao, N. N. Tetrahedron Lett. 2010, 51, 5840.
(10) Liu, Z.; Kumar, K. Synthesis 2010, 2010, 1905.
(11) Purser, S.; Moore, P. R.; Swallow, S.; Gouverneur, V. Chem. Soc.
Rev. 2008, 37, 320.
(12) Mohapatra, D. K.; Umamaheshwar, G.; Rao, M. M.; Umadevi, D.;
Yadav, J. S. RSC Adv. 2014, 4, 8335.
3. CONCLUSION
Inspired by a recent report on the isolation and biological
properties of Δ12-prostaglandin J3 (Δ12-PGJ3, 1)3 and enabled by
our synthetic approaches to this naturally occurring mole-
cule,1,2,5 the described chemistry resulted in the chemical
synthesis of a series of designed analogues (2−44, ent-1, ent-2,
and ent-11). Biological evaluation of these compounds led to the
identification of a number of cytotoxic agents against a variety of
cancer cell lines and established valuable structure activity
relationships (SARs). These investigations led to the identi-
fication of macrolactone 11 as a preclinical development drug
candidate by virtue of its promising chemical and pharmaco-
logical profiles. Due to their higher potency, compounds 33 and
44 (synthesized after compound 11 was selected for early
development) are also under consideration for further preclinical
development. Improvement of the biomedical properties of these
compounds are expected through further molecular fine-tuning
based on the established SARs in this study.
(13) Ermann, P.; Straub, H.; Treuner, U. D.; Drossard, J.-M. U.S.
Patent US 5266593 (A), 1993.
(14) (a) Frick, J. A.; Klassen, J. B.; Bathe, A.; Abramson, J. M.;
Rapoport, H. Synthesis 1992, 1992, 621. (b) Wullschleger, C. W.;
Gertsch, J.; Altmann, K.-H. Org. Lett. 2010, 12, 1120.
(15) (a) Brown, H. C.; Brown, C. A. J. Am. Chem. Soc. 1963, 85, 1005.
(b) Brown, C. A.; Ahuja, V. K. J. Chem. Soc., Chem. Commun. 1973, 553.
́
(16) Oger, C.; Bultel-Ponce, V.; Guy, A.; Balas, L.; Rossi, J.-C.; Durand,
T.; Galano, J.-M. Chem. - Eur. J. 2010, 16, 13976.
(17) Sladojevich, F.; Arlow, S. I.; Tang, P.; Ritter, T. J. Am. Chem. Soc.
2013, 135, 2470.
(18) Hilliard, M.; Frohnert, C.; Spillner, C.; Marcone, S.; Nath, A.;
Lampe, T.; Fitzgerald, D. J.; Kehlenbach, R. H. J. Biol. Chem. 2010, 285,
22202.
(19) Suzuki, M.; Mori, M.; Niwa, T.; Hirata, R.; Furuta, K.; Ishikawa,
T.; Noyori, R. J. Am. Chem. Soc. 1997, 119, 2376.
(20) (a) Kudo, N.; Matsumori, N.; Taoka, H.; Fujiwara, D.; Schreiner,
E. P.; Wolff, B.; Yoshida, M.; Horinouchi, S. Proc. Natl. Acad. Sci. U. S. A.
1999, 96, 9112. (b) Meissner, T.; Krause, E.; Vinkemeier, U. FEBS Lett.
2004, 576, 27.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/jacs.6b02075.
■
S
Experimental procedures and characterization data for all
compounds; NCI-60 cytotoxicity screen and HEK 293T,
MES SA, and MES A DX data (PDF)
(21) Turner, J. G.; Dawson, J.; Sullivan, D. M. Biochem. Pharmacol.
2012, 83, 1021.
(22) Fetz, V. Identifizierung chemischer Transportinhibitoren zur
AUTHOR INFORMATION
Corresponding Author
*kcn@rice.edu
■
Modulation (patho)biologischer Genprodukte. Dissertation, Johannes
Gutenberg-Universitat, Mainz, 2012.
̈
Author Contributions
§K.K.P., S.R., and P.H. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Drs. Lawrence B. Alemany and Quinn Kleerekoper
(Rice University) for NMR-spectroscopic assistance and Drs.
Christopher L. Pennington (Rice University) and Ian
Riddington (University of Texas at Austin) for mass-
spectrometric assistance. This work was supported by the
National Institutes of Health (USA) (Grant AI055475), the
Cancer Prevention & Research Institute of Texas (CPRIT), and
The Welch Foundation (Grant C-1819).
REFERENCES
■
(1) Nicolaou, K. C.; Pulukuri, K. K.; Yu, R.; Rigol, S.; Heretsch, P.;
Grove, C. I.; Hale, C. R. H; ElMarrouni, A. Chem.-Eur. J. 2016, DOI:
10.1002/chem.201601449.
K
DOI: 10.1021/jacs.6b02075
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX