Design, synthesis and prostate cancer cell-based studies of analogs of the Rho/MKL1 transcriptional pathway inhibitor, CCG-1423

Article abstract of DOI:10.1016/j.bmcl.2009.11.056

We recently identified bis(amide) CCG-1423 (1) as a novel inhibitor of RhoA/C-mediated gene transcription that is capable of inhibiting invasion of PC-3 prostate cancer cells in a Matrigel model of metastasis. An initial structure-activity relationship study focusing on bioisosteric replacement of the amides and conformational restriction identified two compounds, 4g and 8, with improved selectivity for inhibition of RhoA/C-mediated gene transcription and attenuated cytotoxicity relative to 1. Both compounds were also capable of inhibiting cell invasion with equal efficacy to 1 but with less attendant cytotoxicity.

Full text of DOI:10.1016/j.bmcl.2009.11.056

Bioorganic & Medicinal Chemistry Letters 20 (2010) 665–672
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and prostate cancer cell-based studies of analogs
of the Rho/MKL1 transcriptional pathway inhibitor, CCG-1423
Chris R. Evelyn a, Jessica L. Bell b, Jenny G. Ryu b, Susan M. Wade a, Andrew Kocab a, Nicole L. Harzdorf b,
b,
H. D. Hollis Showalter b, Richard R. Neubig a, , Scott D. Larsen
*
*
a Department of Pharmacology, University of Michigan Medical School, University of Michigan, Ann Arbor, MI 48109, USA
b Department of Medicinal Chemistry, College of Pharmacy, University of Michigan, Ann Arbor, MI 48109, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
We recently identified bis(amide) CCG-1423 (1) as a novel inhibitor of RhoA/C-mediated gene transcrip-
tion that is capable of inhibiting invasion of PC-3 prostate cancer cells in a Matrigel model of metastasis.
An initial structure–activity relationship study focusing on bioisosteric replacement of the amides and
conformational restriction identified two compounds, 4g and 8, with improved selectivity for inhibition
of RhoA/C-mediated gene transcription and attenuated cytotoxicity relative to 1. Both compounds were
also capable of inhibiting cell invasion with equal efficacy to 1 but with less attendant cytotoxicity.
Ó 2009 Elsevier Ltd. All rights reserved.
Received 1 October 2009
Revised 12 November 2009
Accepted 16 November 2009
Available online 18 November 2009
Keywords:
Metastasis
RhoA
Prostate cancer
Serum response factor
Serum response element
Conformational restriction
Bioisostere
Cancer metastasis is a tremendous medical problem responsible
for thousands of deaths every year.1 Metastases arise when dysreg-
ulation of one or more cellular processes allow malignant cells to
escape the confines of the tissue of origin and establish themselves
in alternate sites. Signaling through RhoA/C is important for inva-
sion and metastasis of many cancers.2–4 In addition to the well-
known role of RhoA/C in cytoskeletal function, there is a less well
understood downstream action on gene transcription.5,6 The path-
way involved in this has recently been elucidated and several
components are related to cancer pathogenesis. The mitogenic G
protein coupled receptor ligands bombesin, thrombin, and
lysophosphatidic acid (LPA) and their receptors are well-known
cancer metastasis7 and for MKL1 and SRF in melanoma and breast
cancer metastases.8 These studies provide important support for
the idea that Rho signaling and specifically Rho-regulated gene tran-
scription may be exciting targets for cancer therapy.
We recently identified a compound CCG-1423 (1) that blocks
SRE-Luciferase gene transcription in response to activation of RhoA
and RhoC signaling pathways.9 Consistent with its role as a Rho/
SRF pathway inhibitor, 1 potently (<1 lM) inhibited LPA-induced
DNA synthesis in PC-3 prostate cancer cells. It also inhibited the
growth of RhoC-overexpressing melanoma lines (A375M2 and
SK-Mel-147) at nanomolar concentrations, but was less active on
related cell lines (A375 and SK-Mel-28) that express lower levels
of RhoC. Compound 1 inhibited Rho-dependent invasion by PC-3
mitogens and stimulate tumor invasion. The novel Ga12 family
of heterotrimeric G proteins (G12 and G13) activates RhoA and RhoC
through guanine exchange factors such as leukemia-associated
RhoGEF (LARG). Most relevant to the present work on Rho-tran-
scriptional mechanisms are the megakaryoblastic leukemia tran-
scriptional co-activator proteins (MKL1 & 2) which cooperate
with the transcription factor, serum response factor (SRF), to
increase expression of a number of genes potentially related to
cancer progression and metastasis.5,6 Exciting recent knockout
and siRNA data have shown a key in vivo role for RhoC in breast
prostate cancer cells, whereas it did not affect the Gai-dependent
invasion by the SKOV-3 ovarian cancer cell line. Thus, based on
its profile, 1 is a promising lead compound for the development
of novel pharmacologic tools to disrupt transcriptional responses
of the Rho pathway in cancer.
Cl
O
O
F3C
O
N
N
H
H
* Corresponding authors. Tel.: +1 734 323 1187 (S.D.L.), +1 734 764 8165 (R.R.N.).
E-mail addresses: rneubig@umich.edu (R.R. Neubig), sdlarsen@umich.edu (S.D.
Larsen).
CF3
1
(CCG-1423)
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.11.056

666
C. R. Evelyn et al. / Bioorg. Med. Chem. Lett. 20 (2010) 665–672
Cl
O
O
O
O
O
a
b
F3C
F3 C
N
L
OH
HN
L
OH
N
L
N
H
2
CF3
3
CF3
4
Scheme 1. Reagents and conditions: (a) 3,5-bis(CF3)PhCOCl, aq NaOH, rt, overnight; or 3,5-bis(CF3)PhCOCl, triethylamine (TEA), CH2Cl2; (b) 4-ClPhNH2, EDC, HOBt, DIPEA,
THF, rt, overnight.
Despite its favorable effects on cancer cell function, 1 did exhibit
some modest acute cellular toxicity toward PC-3 cells at 24 h as evi-
denced by some non-specific inhibition of gene expression
(TK-Renilla) and a parallel decrease in a WST-1 cell viability readout.
Consequently, we undertook initial molecular modifications of 1
with the goal of improving its potency and/or selectivity and
attenuating its cytotoxicity. Three structural features of the lead
were identified as potential areas of concern. First, the N–O bond
in the tether between the two carboxamides is susceptible to reduc-
tive cleavage by thiols, thereby giving 1 the potential to non-selec-
tively modify cysteine-containing proteins or to be cleaved by
glutathione. Second, the two carboxamides could be expected to
limit potency by impeding cell penetration.10,11 Finally, the rela-
tively flexible nature of the tether between the two aromatic rings
is likely not optimal for achieving both potency and selectivity.12
Our initial strategy to modify 1 thus included: removal of the N–
O bond, bioisosteric replacement of the amides,13 and conforma-
tional restriction14 of the tether between the aromatic rings. A lim-
ited survey of aromatic substitution was also undertaken to clarify
the role of the lipophilic substituents.
The synthetic routes to new analogs of 1 are presented in
Schemes 1–6. A preparation of bis(amides) 4 that was general for
a variety of amino acids 2 (acyclic or cyclic, Tables 1 and 4) is sum-
marized in Scheme 1. Acylation with bis(trifluoromethyl)benzoyl
chloride, either under Schotten–Baumann conditions or under
anhydrous conditions, afforded the mono(amides) 3 in good yields.
Final amidation with 4-chloroaniline was then effected with N-(3-
dimethylaminopropyl)-N0-ethylcarbodiimide hydrochloride (EDC)
and 1-hydroxybenzotriazole (HOBt) to afford bis(amides) 4. N-
Methylanilide 4e, benzylamide 4f, and indoline amide 4p were
made under similar conditions using N-methyl-4-chloroaniline,
4-chlorobenzylamine or 5-chloroindoline, respectively, in the final
amidation step. Applying the same chemistry, bis(amides) 5a–g of
aminoxyacetic acid (Table 2) with various aromatic substitution
Table 1
Effects of tether length and composition on transcription and cytotoxicity in transfected PC-3 cellsa
Cl
O
O
F3C
N
H
L
N
H
CF3
Compd
L
IC50 SRE.Lb
(lM)
% inh SRE.Lb (10, 100
lM)
% inh pRL-TKc (10, 100
l
M)
% inh WST-1d (10, 100
lM)
1
–OCH(CH3)–
–OCH2–
–CH2CH2–
–CH2–
–CH2CH2CH2–
–CH2CH2CH2CH2–
1.5
4.7
38
33
21
74, ND
71, 100
38, 64
45, 85
37, 79
48, ND
53, 89
0, 22
15, 25
5, 42
44, ND
42, 91
0, 10
0, 30
0, 12
5a
4a
4b
4c
4d
>100
a
b
c
For assay descriptions, see Ref. 20. All values are mean of P3 experiments, each run in triplicate.
Inhibition of Rho-pathway selective serum response element–luciferase reporter.
Inhibition of control pRL-thymidine kinase Renilla luciferase reporter.
Inhibition of mitochondrial metabolism of WST-1.
d
Table 2
Effects of aromatic substitution on transcription and cytotoxicity in transfected PC-3 cellsa
R2
O
O
R1
O
N
H
N
H
Compd
R1
R2
IC50 SRE.L (
lM)
% inh SRE.L (10, 100
l
M)
% inh pRL-TK (10, 100
lM)
% inh WST-1 (10, 100 lM)
5a
5b
5c
5d
5e
5f
3,5-Bis(CF3)
3,5-Bis(CF3)
3,5-Bis(CF3)
3-CF3
4-CF3
4-H
4-Cl
3-Cl
4-H
4-Cl
4-Cl
4-Cl
3,5-Bis(CF3)
4.7
5.9
36
27
29
71, 100
65, 100
13, 65
25, 86
26, 91
53, 89
51, 89
33, 59
5, 19
6, 0
42, 91
49, 97
0, 37
0, 58
0, 56
>100
8.6
5g
4-Cl
58, 100
19, 87
11, 96
a
Assays defined in Table 1.

C. R. Evelyn et al. / Bioorg. Med. Chem. Lett. 20 (2010) 665–672
667
Cl
Cl
O
b
a
F3C
H2N
Br •HBr
H2N
N
H
N
H
N
H
6
7
8
CF3
Scheme 2. Reagents and conditions: (a) 4-ClPhNH2, PhMe, reflux, 45 min, 28%; (b) 3,5-bis(CF3)PhCOOH, HOBt, EDC, DIPEA, THF, RT, overnight, 65%.
Cl
Cl
O
O
O
b
a
F3C
N
H
N
H
Br
Cl
Br
N
H
9
10
11
CF3
Scheme 3. Reagents and conditions: (a) 4-ClPhNH2, MeCN, reflux, 3 h, 94%; (b) 3,5-bis(CF3)PhCH2NH2, MeCN, reflux, 7 h, 99%.
patterns were also prepared. Monoamine 8 was prepared by alkyl-
ation of 4-chloroaniline with 3-bromopropylamine 6,15 followed
by acylation with bis(trifluoromethyl)benzoyl chloride (Scheme
2). The regioisomeric monoamine 11 was synthesized by alkylation
of bis(trifluoromethyl)benzylamine with bromide 10, which was
prepared by acylation of 4-chloroaniline with 3-bromopropionyl
chloride (Scheme 3). Bis(amine) 12 (Table 3) could be obtained
by exhaustive reduction of bis(amide) 4a (Table 1) with borane–
THF complex at reflux.
lactam 19 required a multi-step synthesis (Scheme 4). Deprotec-
tion of commercially available N-(Boc)-4-piperidinone 15, followed
by acylation, afforded benzamide 16. Oxime formation followed by
Beckman rearrangement generated the ring-expanded intermedi-
ate diazepanone 18. Final N-arylation of the lactam with 4-iodo-
chlorobenzene under Buchwald conditions16 then completed the
synthesis. Amide 16 could also be reductively aminated with
4-chloroaniline and sodium cyanoborohydride to provide
conformationally restricted amine 20 (Table 4).
Thiazole 13 and ether 14 (Table 5) could be made by simple
acylation of the respective commercially available amines with
bis(trifluoromethyl)benzoyl chloride. Conformationally restricted
Isoxazoline 23 was prepared by a cycloaddition of the
nitrile-oxide derived from oxime 2217 with N-(4-chlorophenyl)but-
3-enamide18 (Scheme 5). Triazoles 29 and 31 were prepared from
Table 3
Effects of amide modifications on transcription and cytotoxicity in transfected PC-3 cellsa
F3C
T
Cl
F3C
Compd
T
IC50 SRE.L (
lM)
% inh SRE.L (10, 100
38, 64
lM)
% inh pRL-TK (10, 100
0, 22
lM)
% inh WST-1 (10, 100
0, 10
lM)
4a
4e
4f
8
11
12
–CONHCH2CH2CONH–
–CONHCH2CH2CON(Me)–
–CONHCH2CH2CONHCH2
–CONHCH2CH2CH2NH–
–CH2NHCH2CH2CONH–
–CH2NHCH2CH2CH2NH–
38
>100
>100
5.1
8.1
9.1
70, 80
64, 100
65, 100
37, 35
12, 91
6, 90
0, 11
0, 92
0, 89
a
Assays defined in Table 1.
O
N
O
O
BOC
F3C
a, b
c
F3C
N
N
OH
O
N
CF3
16
CF3
17
15
O
O
F3C
F3C
e
N
d
N
O
O
N
NH
CF3
CF3
18
19
Cl
Scheme 4. Reagents and conditions: (a) TFA, CH2Cl2, 0 °C to rt, 3 h, quant.; (b) 3,5-bis(CF3)PhCOCl, aq NaOH, rt, overnight, 74%; (c) NH2OHꢀHCl, pyridine, 3 Å sieves, rt,
overnight, 72%; (d) Na2CO3, p-TsCl, acetone, RT, 3 h, 66%; (e) 4-I-PhCl, MeNHCH2CH2NHMe, CuI, Cs2CO3, dioxane, 100 °C, 24 h, 48%.

668
C. R. Evelyn et al. / Bioorg. Med. Chem. Lett. 20 (2010) 665–672
Table 4
Effects of conformational restriction on transcription and cytotoxicity in transfected PC-3 cellsa
Compd Structure
IC50 SRE.L (lM)
% inh SRE.L (10, 100
lM)
% inh pRL-TK (10, 100
l
M)
% inh WST-1 (10, 100 lM)
Cl
O
O
O
F3C
N
H
N
H
4a
38
38, 64
0, 22
0, 10
CF3
Cl
O
O
F3C
F3C
N
N
N
H
4g
4h
9.8
16
37, 78
17, 87
19, 34
0, 14
0, 39
CF3
H
N
0, 17
CF3
O
Cl
H
O
N
O
O
F3C
N
4i
4j
69
23, 83
12, 27
0, 22
Cl
CF3
O
F3C
N
Cl
16.3
30, 100
8, 78
0, 67
HN
CF3
Cl
O
O
N
H
F3C
4k
9.5
33, 86
0, 62
0, 14
N
CF3
CF3
CF3
O
O
F3C
Cl
N
H
4l
>100
O
N
H
H
N
F3C
F3C
N
H
4m
4n
4o
1.7
79, ND
0, ND
38, ND
O
Cl
Cl
O
O
>100
>100
HN
HN
F3C
F3C
O
NH
O
Cl
N
CF3
H

C. R. Evelyn et al. / Bioorg. Med. Chem. Lett. 20 (2010) 665–672
669
Table 4 (continued)
Compd Structure
IC50 SRE.L (lM)
% inh SRE.L (10, 100
l
M) % inh pRL-TK (10, 100
l
M)
% inh WST-1 (10, 100 lM)
O
O
Cl
F3C
N
H
N
4p
13
15, 40
0, 1
0, 0
CF3
O
N
H
N
CF3
4q
10
51, 100
30, 91
10, 88
Cl
O
N
CF3
O
O
N
F3C
F3C
19
20
>100
10.8
CF3
Cl
O
Cl
N
19, 54
0, 20
0, 0
N
H
CF3
a
Assays defined in Table 1. ND: not determined.
anilines 28 and 30, respectively, by azidation and copper-catalyzed
Click cyclization with alkyne amides 25 and 27 (Scheme 6).19
Effects on Rho-mediated gene expression, non-specific gene
expression, and cytotoxicity of all new compounds were deter-
mined in transiently transfected PC-3 cells (Tables 1–5).20 The data
in the tables are intended to illustrate both potency and efficacy.
This was done due to the observation that some compounds had
equivalent potencies (IC50s) against SRE.L, yet differed in their
maximal responses (efficacies). There are a number of possible
explanations for this, including differences in passive permeability
into the cells, efflux out of the cells, or even the complement of
Rho-pathway targets impacted. To optimize the likelihood for
eventual in vivo activity, we elected to consider both potency
and efficacy versus SRE.L in our structure–activity relationship
(SAR) analysis. Effects against pRL-TK and WST-1 at high and low
concentrations are included as an approximate indicator of selec-
tivity. IC50s often could not be calculated from the generally weak-
er dose response data against these selectivity targets, and
therefore are not included.
even at high dose, and improved selectivity versus Renilla, but
attenuated potency against SRE.L by over an order of magnitude.
Decreasing or increasing the length of the carbon chain by one car-
bon (4b and 4c) did not greatly impact the potency or efficacy rel-
ative to 4a, but lengthening to four carbons (4d) resulted in a total
loss of activity.
A brief survey of aromatic substitution is summarized in Table
2. Despite the diminished toxicity of 4a, we elected to use the more
potent unsubstituted aminoxyacetic acid template of 5a to mag-
nify any changes in activity. Moving the chloro group to the 3-po-
sition (5b) had no effect, but removing it altogether (5c) negatively
affected both potency and efficacy. Removing one of the trifluoro-
methyl groups (5d and 5e) was similarly detrimental to removing
the chloro group, and removing both trifluoromethyl groups (5f)
led to a complete loss of activity. Reversing the position of the
two aromatic rings (5g) gave a compound with activity similar to
5a. These limited data suggest that some degree of lipophilicity
on the aromatic rings is crucial for activity, possibly to facilitate
cell permeability. A more in-depth examination of aromatic substi-
tuent SAR, including whether electron deficiency is necessary, will
be the subject of future work.
Table 1 summarizes the impact of changes on the tether be-
tween the two carboxamides of compound 1. Removing the methyl
group (5a) had little effect on activity or selectivity. Replacement of
the oxygen with carbon (4a) indeed removed acute cytotoxicity,
Table 3 summarizes our initial foray into modification of the
secondary amides. N-Methylation of the 4-chloroaniline amide
O
N
F3C
CHNOH
F3C
CHO
F3C
a
b
NH
O
CF3
CF3
CF3
21
23
22
Cl
Scheme 5. Reagents and conditions: (a) HONH2ꢀHCl, MeOH, reflux, 5 h, 48%; (b) N-(4-chlorophenyl)but-3-enamide, NaOCl, CH2Cl2, 0 °C to rt, 24 h, 99%.

670
C. R. Evelyn et al. / Bioorg. Med. Chem. Lett. 20 (2010) 665–672
(4e), intended to remove a hydrogen bond donor and improve cell
Cl
O
O
a
permeability, resulted in a dramatic loss in activity. Replacement
of one of the anilides with a benzylamide (4f) also was detrimental.
Reducing the amides to the corresponding amines was considerably
more successful. Both monoamines 8 and 11, as well as diamine 12,
exhibited improved potency and efficacy against Rho-dependent
gene expression. Of particular interest was monoamine 8, which had
potency approaching that of the aminoxyacetic acid template 5a,
retained the moderate selectivity of 4a versus non-specific gene
expression, and showed very little cytotoxicity compared to the other
monoamine and the diamine, even at high concentrations. The other
two amines exhibited unacceptable levels of cytotoxicity.
Conformational restriction of the flexible bis(amide) tether was
explored as a way to potentially improve both potency and selec-
tivity (Table 4). In several cases, improved potency versus the cor-
responding acyclic analog was observed (e.g., 4g vs 4a, 4p vs 4e),
but the goal of improving selectivity versus non-specific gene
expression was not clearly achieved with any of the restricted ana-
logs. Among a series of closely related five- and six-membered ring
analogs (4g–4j), it is interesting to note that a ‘meta’-like disposi-
tion of the two aromatic rings (4g, 4j) was favored over ‘para’ or
‘ortho’ in closely related analogs (4h, 4i). This pattern is dramati-
cally reflected in the series of analogs constrained by a central aro-
matic ring (4l–4n), wherein only the meta analog 4m is active. The
inactivity of the vicinal substituted analog 4o and an analog with a
‘vicinal-like; disposition of the amides (19) is also consistent with
this pattern. Compound 20, a restricted analog of monoamine 8,
was inferior to the acyclic analog with regard to both potency
OH
N
H
24
25
O
F3C
b
N
H
H2N
•HCl
CF3
27
26
HN
O
N
Cl
N
N
F3C
NH2
F3C
c,d
CF3
CF3
CF3
29
28
NH2
e,f
H
N
Cl
Cl
F3C
N
N
30
N
O
31
Scheme 6. Reagents and conditions: (a) 4-ClPhNH2, HOBt, EDC, DIPEA, THF, rt,
overnight, 87%; (b) 3,5-bis(CF3)COCl, TEA, CH2Cl2, rt, overnight, 97%; (c) (i) TFA,
ꢁ10 °C, NaNO2; (ii) NaN3, ꢁ10 °C to rt, 2 h, 63%; (d) 25, CuSO4ꢀ5H2O,
L-ascorbic acid,
H2O/tBuOH, 1:1, rt, overnight, 19%; (e) (i) 6 M HCl, 0 °C, NaNO2; (ii) NaN3, 0 °C,
15 min., 56%; (f) 27, CuSO4ꢀ5H2O,
L-ascorbic acid, H2O/tBuOH, 1:1, rt, 48 h, 10%.
Table 5
Effects of amide replacement on transcription and cytotoxicity in transfected PC-3 cellsa
Compd
Structure
IC50 SRE.L (lM)
% inh SRE.L (10, 100
lM)
% inh pRL-TK (10, 100
lM)
% inh WST-1 (10, 100 lM)
Cl
O
O
O
F3C
N
H
N
4a
38
38, 64
0, 22
0, 10
H
CF3
F3C
N
N
H
Cl
13
14
4.1
8.9
50, 45
39, 38
0, 0
S
CF3
Cl
O
F3C
N
O
55, 86
70, 84
28, 65
51, 60
0, 38
0, 38
H
CF3
F3C
F3C
N
O
Cl
O
23
29
4.2
N
H
F3C
N
N
N
H
N
>100
O
Cl
F3
C
Cl
F
3C
O
31
>100
N
N
HN
N
F3
C
a
Assays defined in Table 1.

C. R. Evelyn et al. / Bioorg. Med. Chem. Lett. 20 (2010) 665–672
671
Table 6
efficacies against SRE.L at 10 and 100
in the Matrigel assay. Compound 4g, however, was somewhat less
toxic at the higher concentration, exhibiting an efficacy:toxicity
l
M, were similarly active
Effects of new analogs on cell invasion and cytotoxicitya
Compd % inh invasionb % inh WST-1c % inh invasionb % inh WST-1c
(10
µM)
(10
µM)
(100
µM)
(100 µM)
profile at 100
10 M. Also,
l
M that is superior to that of the original lead 1 at
1
71
13
14
6
54
0
l
a tight correlation between the selectivity for
4a
4b
4c
4d
4g
4h
4i
62
75
31
25
72
79
78
86
0
28
38
1
SRE.L:cytotoxicity and invasion:cytotoxicity can be seen in the ser-
ies of amines 8, 11 and 12. Monoamine 8 had the most favorable
ratio of efficacy to cytotoxicity in the transfected cell SRE-Lucifer-
ase studies, and this is reflected in the data in Table 6. This
0
0
15
20
14
12
63
2
0
1
0
23
36
44
70
1
0
compound in fact inhibits invasion at 10
approaching that of the lead 1, with no observable toxicity. At
100 M, nearly complete inhibition of invasion was achieved with
a lesser degree of toxicity than that induced by 1 at 10 M. Mono-
lM to an extent
0
4m
4n
4p
5a
8
31
0
l
l
0
0
0
0
50
54
47
30
18
3
14
0
88
84
96
86
8
78
27
95
97
0
amine 11 and diamine 12, on the other hand, were much less selec-
tive for efficacy versus toxicity in the transfected cells (Table 3),
and that is reflected in the data in Table 6 at the higher
concentration.
Correlation graphs between the average inhibition of the PC-3
Matrigel invasion assay versus the average inhibition of the
11
12
13
14
20
0
0
0
0
26
56
0
18
0
0
G
10
a
12QL-stimulated SRE.L-luciferase expression in PC-3 cells at
a
b
c
For assay descriptions, see Ref. 21.
Inhibition of invasion by cultured PC-3 cells into a Matrigel matrix.
Inhibition of mitochondrial metabolism of WST-1.
l
M and 100 M are presented in Figure 1. A positive correlation
l
can be seen at both concentrations, although there are a few
outliers that strongly inhibit SRE.L without inhibiting invasion.
Interestingly, there seems to be a threshold for inhibition of SRE.L
(about 50%) that must be achieved before effects on invasion are
observed.
In summary, an initial SAR survey of 1, an inhibitor of Rho-med-
iated gene expression, was undertaken with the goal of improving
selectivity and/or potency, while attenuating cytotoxicity. In addi-
tion to removing the obviously labile N–O bond, two strategies
were applied: (1) replacement of the secondary carboxamides to
improve cell permeability; and (2) conformational restriction to
improve potency and/or selectivity. These approaches afforded
compounds with improved biological profiles relative to the lead
1, albeit at a cost of 5–10-fold lower potencies. Of particular inter-
est are nipecotic amide 4g and monoamine 8, each of which are
and efficacy, perhaps another example of the disadvantageous
arrangement of the aromatic rings in a ‘para’-like orientation. Fi-
nally, reversal of the aromatic rings (4q vs 4h) led to a significant
loss in selectivity at high concentration versus Renilla and
cytotoxicity.
Table 5 presents data for analogs incorporating potential bio-
isosteric replacements for one of the amides. Three of these com-
pounds (13, 14 and 23) exhibited improved potency versus the
bis(amide) 4a, but selectivity versus Renilla and/or cytotoxicity suf-
fered. Triazoles 29 and 31 were completely inactive, perhaps sug-
gesting a lack of permeability into the cells.
In our original publication, we showed that 1 selectively inhib-
ited spontaneous PC-3 prostate cancer cell invasion through a
capable of inhibiting Ga12QL-stimulated SRE.L-luciferase expres-
sion with efficacy equal to that of 1, but with significantly less
attendant cytotoxicity. Furthermore, we have for the first time
demonstrated within this series of compounds a clear correlation
between inhibition of Rho-mediated gene expression and cell inva-
sion in a Matrigel matrix model of metastasis.
Future work will be aimed at expanding the diversity of the aro-
matic rings using one or both of the improved templates of 4g and
8, as well as constructing affinity reagents for the identification of
the molecular target(s) of this novel class of inhibitors of the Rho-
signaling pathway.
Matrigel matrix, but not the
Gai-dependent LPA-stimulated
SKOV-3 ovarian cancer cell invasion, in vitro.9 Therefore, we tested
several of our analogs for their ability to inhibit PC-3 prostate can-
cer cell invasion with better selectivity for invasion versus toxicity
(as measured by metabolism of WST-1) in comparison to 1. Data
for selected compounds is presented in Table 6.21 Overall, a fairly
consistent correlation was observed between inhibition of invasion
and inhibition of SRE.L in PC-3 cells. For example, conformationally
restricted analogs 4g–4i, which all possessed roughly equivalent
Invasion vs. SRE.L - 10uM
Invasion vs. SRE.L - 100uM
A
B
100
80
60
40
20
0
100
80
60
40
20
0
-20
-20
-20
0
20
40
60
80 100
-20
0
20
40
60
80 100
SRE.L (% inhibition)
SRE.L (% inhibition)
Figure 1. Correlation of Matrigel invasion and G
a12QL-stimulated SRE.L-luciferase expression inhibition in PC-3 cells. The data represent the average of experiments
performed three separate times for an n = 3 in duplicate and triplicate, respectively.

672
C. R. Evelyn et al. / Bioorg. Med. Chem. Lett. 20 (2010) 665–672
16. Klapars, A.; Huang, X. H.; Buchwald, S. L. J. Am. Chem. Soc. 2002, 124, 7421.
Acknowledgments
17. Liu, A.; Liu, A. J. Agric. Food Chem. 2008, 56, 6562.
18. Tellitu, I.; Urrejola, A.; Serna, S.; Moreno, I.; Herrero, M. T.; Dominguez, E.;
Supported by NIH grants R01GM39561 (RRN) and
F31CA113268-04 (CRE) and support from the UM Comprehensive
Cancer Center and the Prostate Cancer SPORE (P50CA069568).
SanMartin, R.; Correa, A. Eur. J. Org. Chem. 2007, 437.
19. Rostovtsev, V. V.; Green, L. G.; Fokin, V. V.; Sharpless, K. B. Angew. Chem., Int. Ed.
2002, 41, 2596.
20. PC-3 cells were seeded into 96-well plates at a cell density of 4 ꢂ 104 cells per
well 24 h prior to transient transfection with a Ga12Q231L activator expression
plasmid along with both the SRE.L luciferase and pRL-TK Renilla luciferase
reporter plasmids. The DNA plasmids were transfected using the lipid-based
LipofectAMINE 2000 (Invitrogen) transfection reagent at a concentration of
References and notes
1. Sawyer, T. K. Exp. Opin. Invest. Drugs 2004, 13, 1.
2. Sahai, E.; Marshall, C. J. Nat. Cell Biol. 2003, 5, 711.
3. van Golen, K. L.; Wu, Z. F.; Qiao, X. T.; Bao, L. W.; Merajver, S. D. Cancer Res.
2000, 60, 5832.
4. Liu, N.; Zhang, G.; Bi, F.; Pan, Y.; Xue, Y.; Shi, Y.; Yao, L.; Zhao, L.; Zheng, Y.; Fan,
D. J. Mol. Med. 2007, 85, 1149.
1 lL per lg of DNA in antibiotic-free, Opti-MEM I medium. The total amount of
DNA was kept constant by inclusion of the appropriate amount of the
pcDNA3.1-zeo plasmid. Six hours after transfection, the transfection mixture
was removed and cells were serum-starved overnight in DMEM medium
containing 0.5% FBS and 1% penicillin–streptomycin. Firefly and Renilla
luciferase activities were determined 19 h later using the dual-luciferase
assay kit (Promega) according to manufacturer’s instructions. Just before cell
lysis, the viability of the cells was measured using a WST-1 cell proliferation
reagent. Data are expressed as IC50 or percentage of inhibition (DMSO
alone = 0%) in Tables 1–5. Individual experiments were run in triplicate, and
the values in the tables are the mean of at least three separate experiments.
Because of the large degree of variability inherent in transient transfection
assays, standard error measurements have not been included.
5. Cen, B.; Selvaraj, A.; Burgess, R. C.; Hitzler, J. K.; Ma, Z.; Morris, S. W.; Prywes, R.
Mol. Cell Biol. 2003, 23, 6597.
6. Miralles, F.; Posern, G.; Zaromytidou, A. I.; Treisman, R. Cell 2003, 113, 329.
7. Hakem, A.; Sanchez-Sweatman, O.; You-Ten, A.; Duncan, G.; Wakeham, A.;
Khokha, R.; Mak, T. W. Genes Dev. 2005, 19, 1974.
8. Medjkane, S.; Perez-Sanchez, C.; Gaggioli, C.; Sahai, E.; Treisman, R. Nat. Cell
Biol. 2009, 11, 257.
9. Evelyn, C. R.; Wade, S. M.; Wang, Q.; Wu, M.; Iniguez-Lluhi, J. A.; Merajver, S. D.;
Neubig, R. R. Mol. Cancer Ther. 2007, 6, 2249.
10. Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Adv. Drug Delivery Rev.
1997, 23, 3.
11. Burton, P. S.; Conradi, R. A.; Ho, N. F.; Hilgers, A. R.; Borchardt, R. T. J. Pharm. Sci.
1996, 85, 1336.
12. Veber, D. F.; Johnson, S. R.; Cheng, H. Y.; Smith, B. R.; Ward, K. W.; Kopple, K. D.
J. Med. Chem. 2002, 45, 2615.
13. Lima, L. M.; Barreiro, E. J. Curr. Med. Chem. 2005, 12, 23.
14. Mann, A.; LeChatlier, H. L. In Practice of Medicinal Chemistry; Elsevier: London,
2003; pp 233–250.
15. Orelli, L. R.; Garcia, M. B.; Niemevz, F.; Perillo, I. A. Synth. Commun. 1999, 29,
1819.
21. Compounds were tested as follows: PC-3 cells (2 ꢂ 105) were transferred to 24-
well Matrigel inserts in low-serum DMEM medium (0.5% FBS) with DMSO or
chemical compounds in the upper chamber. Low-serum DMEM medium (0.5%
FBS) was added to the lower well and the invasion chambers were incubated at
37 °C in 5% CO2 for 24 h. Inserts were fixed in methanol for 10 min and then
stained for 60 min with 0.5% crystal violet in 20% methanol. After wiping the
top surface of the filter with cotton swabs to remove non-invaded cells, the
inserts were allowed to dry overnight. Inserts were incubated in 20% acetic
acid on a plate shaker for 15 min to extract the crystal violet stain. The number
of invaded cells was quantitated by measuring the absorbance of the extracted
crystal violet stain at a wavelength of 595 nm with the Victor2 plate reader.