SAR exploration at the C-3 position of tetrahydro-β-carboline sstr3 antagonists

Article abstract of DOI:10.1016/j.bmcl.2016.02.022

MK-4256, a tetrahydro-β-carboline sstr3 antagonist, was discontinued due to a cardiovascular (CV) adverse effect observed in dogs. Additional investigations revealed that the CV liability (QTc prolongation) was caused by the hERG off-target activity of MK-4256 and was not due to sstr3 antagonism. In this Letter, we describe our extensive SAR effort at the C3 position of the tetrahydro-β-carboline structure. This effort resulted in identification of 5-fluoro-pyridin-2-yl as the optimal substituent on the imidazole ring to balance sstr3 activity and the hERG off-target liability.

Full text of DOI:10.1016/j.bmcl.2016.02.022

Accepted Manuscript
SAR Exploration at the C-3 Position of Tetrahydro-β-carboline sstr3 Antago-
nists
Shuwen He, Peter H. Dobbelaar, Liangqin Guo, Zhixiong Ye, Jian Liu, Tianying
Jian, Quang Truong, Shrenik K. Shah, Wu Du, Hongbo Qi, Raman K. Bakshi,
Qingmei Hong, James D. Dellureficio, Edward Sherer, Alexander Pasternak,
Zhe Feng, Mikhail Reibarkh, Melissa Lin, Koppara Samuel, Vijay B. Reddy,
Stan Mitelman, Sharon X. Tong, Gary G. Chicchi, Kwei-Lan Tsao, Dorina
Trusca, Margaret Wu, Qing Shao, Maria E. Trujillo, Guillermo Fernandez,
Donald Nelson, Patricia Bunting, Janet Kerr, Patrick Fitzgerald, Pierre
Morissette, Sylvia Volksdorf, George J. Eiermann, Cai Li, Bei B. Zhang,
Andrew D. Howard, Yun-Ping Zhou, Ravi P. Nargund, William K. Hagmann
PII:
S0960-894X(16)30134-2
DOI:
http://dx.doi.org/10.1016/j.bmcl.2016.02.022
BMCL 23573
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
1 December 2015
7 February 2016
9 February 2016
Please cite this article as: He, S., Dobbelaar, P.H., Guo, L., Ye, Z., Liu, J., Jian, T., Truong, Q., Shah, S.K., Du, W.,
Qi, H., Bakshi, R.K., Hong, Q., Dellureficio, J.D., Sherer, E., Pasternak, A., Feng, Z., Reibarkh, M., Lin, M., Samuel,
K., Reddy, V.B., Mitelman, S., Tong, S.X., Chicchi, G.G., Tsao, K-L., Trusca, D., Wu, M., Shao, Q., Trujillo, M.E.,
Fernandez, G., Nelson, D., Bunting, P., Kerr, J., Fitzgerald, P., Morissette, P., Volksdorf, S., Eiermann, G.J., Li, C.,
Zhang, B.B., Howard, A.D., Zhou, Y-P., Nargund, R.P., Hagmann, W.K., SAR Exploration at the C-3 Position of
Tetrahydro-β-carboline sstr3 Antagonists, Bioorganic & Medicinal Chemistry Letters (2016), doi: http://dx.doi.org/
10.1016/j.bmcl.2016.02.022
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Bioorganic & Medicinal Chemistry Letters
SAR Exploration at the C-3 Position of Tetrahydro-β-carboline sstr3
Antagonists
Shuwen He , Peter H. Dobbelaar, Liangqin Guo, Zhixiong Ye, Jian Liu, Tianying Jian, Quang Truong,
Shrenik K. Shah, Wu Du, Hongbo Qi, Raman K. Bakshi, Qingmei Hong, James D. Dellureficio, Edward
Sherer, Alexander Pasternak, Zhe Feng, Mikhail Reibarkh, Melissa Lin, Koppara Samuel, Vijay B. Reddy,
Stan Mitelman, Sharon X. Tong, Gary G. Chicchi, Kwei-Lan Tsao, Dorina Trusca, Margaret Wu, Qing
Shao, Maria E. Trujillo, Guillermo Fernandez, Donald Nelson, Patricia Bunting, Janet Kerr, Patrick
Fitzgerald, Pierre Morissette, Sylvia Volksdorf, George J. Eiermann, Cai Li, Bei B. Zhang, Andrew D.
Howard, Yun-Ping Zhou, Ravi P. Nargund, and William K. Hagmann
Early Development and Discovery Sciences, Merck Research Laboratories, 2000 Galloping Hill Road, Kenilworth, New Jersey 07033, United States
ARTICLE INFO
ABSTRACT
Article history:
Received
Revised
Accepted
Available online
MK-4256, a tetrahydro-β-carboline sstr3 antagonist, was discontinued due to a cardiovascular
(CV) adverse effect observed in dogs. Additional investigations revealed that the CV liability
(QTc prolongation) was caused by the hERG off-target activity of MK-4256 and was not due to
sstr3 antagonism. In this report, we describe our extensive SAR effort at the C3 position of the
tetrahydro-β-carboline structure. This effort resulted in identification of 5-fluoro-pyridin-2-yl as
the optimal substituent on the imidazole ring to balance sstr3 activity and the hERG off-target
liability.
Keywords:
Somatostatin receptor subtype 3
2015 Elsevier Ltd. All rights reserved.
Antagonist
Tetrahydro-β-carboline
hERG
SAR
Recently, we disclosed that antagonism of somatostatin
receptor type 3 (sstr3) has the potential to be a novel mechanism
for the treatment of type 2 diabetes.1 For chemical matters, we
focused on the non-peptidic imidazolyl tetrahydro-β-carboline
class originally reported by Poitout et al. and Novartis
scientists.2,3 Our initial effort led to the identification of tool
compound 1, a potent and selective sstr3 antagonist with
adequate mouse pharmacokinetic (PK) properties (Figure 1) 4.
Profiling of compound 1 provided the desired pharmacological
proof of concept: this compound not only enhances glucose
stimulated insulin secretion in β-pancreatic cells but also reduces
glucose excursion in a mouse glucose tolerance test (GTT).
Figure 1. Structures of β-carboline sstr3 antagonists
However, the development of MK-4256 was discontinued due
to the potential risk of adverse cardiovascular (CV) effects. In a
conscious CV dog telemetry model, MK-4256 exhibited a dose-
dependent prolongation of the QTc interval (QT interval
corrected for the heart rate). The subsequent evaluation of
carboxylic acid 2 established that the QTc prolongation was
caused by the blockade of hERG channel and was not due to
sstr3 antagonism (Figure 1).6 However, the poor PK profile of 2
precluded its further progression.
Further optimization of this series led to the identification of our
5
first clinical candidate MK-4256 (Figure 1).
MK-4256
possesses excellent sstr3 potency and subtype selectivity, a good
pharmacokinetic (PK) profile in preclinical species, and superior
efficacy in a mouse GTT assay.
———
 Corresponding author. Tel.: 908-740-0881.
E-mail: shuwen_he@merck.com

This critical conclusion that sstr3 antagonism alone does not
elicit QTc prolongation re-aligned the subsequent medicinal
chemistry effort to focus on minimizing the hERG off-target
activity. This report describes the results of our extensive SAR
study at the C-3 position of the tetrahydro-β-carboline series.
potencies comparable to 1 in sstr3 binding assay and functional
assay. More interestingly, 6a seems to have less hERG off-target
liability than 1 based on the hERG binding assay.
Table 2. Pharmacokinetics of 6a compared with compound 1
The synthesis for the compounds followed the procedures
reported previously (Scheme 1).1-5 Boc-protected tryptophan or
its analog 3 was coupled with an α-bromo-methyl ketone to
furnish ester intermediate 4 which was then cyclized to form the
imidazole functionality in intermediate 5. After removing the
Boc protecting group under acidic conditions, a Pictet-Spengler
cyclization with the corresponding aldehyde or ketone gave the
tetrahydro-β-carboline products. For cases where there were
isomers present, they were separated by chromatography on a
silica gel column or a column with chiral stationary phase (e.g.,
Chiralpak® columns) to afford single isomers.
compd/species
6a
1
rat
mouse
4/10
mouse
1/2
dose, mg/kg, (iv/po)
oral AUCN, µM•h/(mg/kg)
iv CL (ml/min/kg)
iv t1/2 (h)
1/2
0.28
72
0.31
77
1.25
15
0.96
13.2
49
1.3
4.6
58
1.6
1.6
45
Vdss (L/kg)
F (%)
Compounds were dosed as 1 mg/mL solution in ethanol:PEG400:water
(2:23:75) for 6a and as 0.2 mg/mL solution in ethanol:PEG400:water
(10:30:60) for 1.
Based on its favorable in vitro profile, 6a was further
evaluated in rodent pharmacokinetics (PK) studies (Table 2). In
mice, 6a exhibits ~5 folds higher plasma clearance than
compound 1. The oral exposure of 6a is also inferior to that of
compound 1. In a rat PK study, again, the plasma clearance of 6a
is high and the oral exposure is poor. It appears that adding an
angular methyl group at the C3 position is detrimental to the
rodent PK profiles. Therefore, we abandoned this approach and
instead focused on the compounds without the C-3 angular
methyl.
We next synthesized and evaluated several compounds with
non-aromatic substitutions on the imidazole ring (Table 3).
Saturation of the phenyl to a cyclohexyl ring (7) is detrimental to
sstr3 activity as indicated in both binding and functional assays
(about 20-30 fold potency loss). Compound 8 with a t-butyl
group on the imidazole ring also has reduced activities. A bulky
1-admantyl group (9) regains potency in binding assay but
suffers large loss in the functional assay. Based on this result, we
focused on the designs with aromatic substitutions on the
imidazole.
Scheme 1. Reagents and conditions: (i) Cs2CO3, DMF, rt (ii) NH4OAc, p-
xylene, reflux (iii) a. 4 M HCl in dioxane or TFA, CH2Cl2, rt; b. Pictet-
Spengler cyclization with a ketone or an aldehyde.
Table 1. Profiles of compounds with an angular methyl at C3 position
Table 3. Compounds with non-aromatic substitutions on the imidazole ring
compd
human sstr3
binding IC50, nM cAMP antagonism EC50
human sstr3 functional
hERG binding
IC50, nM a
,
nM (% inh.) b
14
23
22 (87%)
17 (76%)
ND
380 (n=2)
2175 (n=4)
2946 (n=2)
1634 (n=2)
530 (n=2)
1
6a
6b
6c
6d
compd
R
human sstr3
human sstr3
functional cAMP
hERG binding IC50
nM
,
603
binding IC50
,
nM)
antagonism EC50
nM (% inh.)
,
>1333
102
ND
-Ph
14
230
150
25
22 (87%)
380 (n=2)
136 (n=2)
2944 (n=2)
506 (n=2)
1
7
8
9
>1000 (88%)
-cyclohexyl
-t-Bu
691 (61%)
a. n = denotes number of tests; b. ND = not determined.
1008 (110%)
>1000 (63%)
Our early SAR work focused on using compound 1 as the
benchmark. The analogs with an angular methyl group at the C3
position were among the first few compounds prepared. Starting
with racemic N-Boc amino acid 3 (R1 = Me, Scheme 1), all four
possible stereoisomers (6a-d) were obtained. Table 1 lists the in
vitro profile of the four isomers together with compound 1 as a
comparison. Isomer 6a is the most potent isomer affording
-1-admantyl
We surveyed various substitutions on the phenyl group
attached to the imidazole ring (Table 4). Among the mono-
fluoro substitutions (10-12), compound 10 with a para-fluoro

m,p-difluoro
o,p-difluoro
1.8
23
6.4 (101%)
37 (61%)
350 (n=2)
634 (n=2)
26
27
gives the best sstr3 potency (about 4-5 folds better than
compound 1). The hERG binding affinity of 10 is almost
identical to that of compound 1. Compound 13 with a larger
chloro substitution at the para position suffers a loss in sstr3
a. ND = Not determined.
potency as well as selectivity against hERG.
A cyano
substitution at meta position (15) seems to be better than at para
position (14). However, both are less potent in sstr3 assays than
compound 1. Both methyl and CF3 at para position (16 and 17)
have reduced sstr3 potency than compound 1. Compound 18
with a para hydroxyl group shows similar potency to compound
1. Methylation of the para-hydroxyl group reduces sstr3 potency
(19). However, migration of the methoxyl group to the meta
position afforded 20 which restores sstr3 potency.
Trifluoromethoxyl group at either meta and para position causes
a loss in sstr3 potency, especially in sstr3 functional activity (21
and 22). A para-phenyl substitution reduces sstr3 potency
significantly (23). Polar groups are not tolerated (such as 24 with
a carboxyl acid at the meta position and 25 with a methyl sulfone
at the para position). Two compounds with di-substitution were
also evaluated. Compound 26 with m,p-difluoro substitution has
a profile similar to compound 10 while 27 with o,p-difluoro gives
reduced sstr3 potency. This study suggested that substitution
tolerance in this area is narrow and a para-fluoro in this series is a
preferred substituent (compound 10).
We further profiled compound 10 in in vivo studies. In a
mouse GTT study, compound 10 reduces glucose excursion by
79% at 10 mg/kg oral dose comparable to the positive control
compound 1, which reduces glucose excursion by 75% at the
same dose. Compound 10 was also tested for pharmacokinetics
in mice (Table 5). The compound shows reduced plasma
clearance and much higher oral exposure than compound 1. The
bioavailability is also improved for compound 10. The para-
fluoro-phenyl moiety was later incorporated into MK-4256, our
first clinical candidate.
Table 5. Pharmacokinetics in mouse of 10 compared with compound 1
compd/species
10
1
oral AUCN,
µM•h/(mg/kg)
iv CL (ml/min/kg)
5.58
5.9
3.6
1.8
82
1.25
15
Table 4. Compounds with a substituted phenyl on the C-3 imidazole ring a
iv t1/2 (h)
Vdss (L/kg)
F (%)
1.6
1.6
45
Compounds were dosed to mice at 1 mg/kg for iv and 2 mg/kg for p.o.as
0.2 mg/mL solution in ethanol:PEG400:water (10:30:60) for 1 and in
ethanol:PEG400:water (4:16:80) for 10.
With compounds 1 and 10 as the benchmarks, we carried out
additional “methyl scan” beyond compounds 6a-6d. Compounds
33, 34, 36 and 38 were prepared via slightly modified routes
(Scheme 2). Starting from 28, a known intermediate, bis-Boc
protected intermediate 30 was obtained. Methylation of the
imidazole nitrogen gave 31, which was subjected to the Pictet-
Spengler cyclization. The resulted product was separated to give
33 and 34. Preparation of 36 was straightford by starting from
compd
X
Human sstr3
binding IC50 functional cAMP
human sstr3
hERG binding IC50
nM
,
(nM)
antagonism EC50
nM (% inh.)
,
H
p-F
14
2.8
22 (87%)
5.7 (119%)
88 (112%)
30 (82%)
380 (n=2)
377 (n=2)
258 (n=2)
495 (n=2)
144 (n=2)
ND
1
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
m-F
8.2
Boc-1-methyl-D-tryptophan.
Preparation of 38 with an
o-F
18
additional methyl substitution on the imidazole carbon was
identical to the general method described in Scheme 1 except
that 2-bromo-1-phenylpropan-1-one was used during the
alkylation step.
p-Cl
8.1
50 (93%)
p-CN
141
37
ND
m-CN
p-Me
435 (96%)
113 (83%)
1517 (75%)
26 (80%)
ND
33
101 (n=2)
158 (n=2)
467 (n=2)
69 (n=2)
247 (n=2)
130 (n=5)
58 (n=2)
17 (n=2)
839 (n=2)
ND
p-CF3
p-OH
p-OMe
m-OMe
p-OCF3
m-OCF3
p-Ph
26
34
127
4.6
181 (82%)
17 (108%)
>1000 (16%)
>1000 (34%)
>1000 (46%)
ND
639
43
391
>1333
>1333
m-CO2H
p-SO2Me
>1000 (18%)
Scheme 2. Reagents and conditions: (i) Boc2O, DMAP, MeCN, rt (ii)
NH4OAc, p-xylene, reflux under a Dean-Stark trap, 3h (iii) Cs2CO3, MeI,

DMF, rt (iv) follwing the general procedure shown in Scheme 1 (v) 2-bromo-
1-phenylpropan-1-one, Cs2CO3, DMF, rt.
significantly less potent than compound 1. The last compound
41 with a triazole replacement also shows reduced sstr3
activities.
The results from this effort indicate that the
As indicated by the data shown in Table 6, the “methyl scan”
does not provide any advantage. Methylation of the imidazole
nitrogens (33 and 34) drastically reduces the potency on human
sstr3. This result is consistent with an early observation by
Poitout group that a thiazole replacement of the imidazole ring
causes a significant loss of sstr3 activity.2 In contrast, compound
36 with methylation of the indole NH retains the sstr3 potency to
some extent. However, compared with 10, compound 36 is still
less potent by about 10 folds. The results above indicate that
these NH’s may have key interactions with the receptor. Finally,
a methyl group was attached to the imidazole ring to test the
impact of restricting the rotation of the phenyl group. However,
the resulted compound 38 suffers significant loss in potency
compared to the benchmark compound 1.
unsubstituted imidazole with the substitution pattern in 1 and 10
seems be optimal.
The preparation of compounds above (39-41) focused on
constructing
the
corresponding
hetereocycle-containing
precursors to the Pictet-Spengler cyclization. The method for the
precursor 47 for compound 39 with a pyrazole replacement is
shown in Scheme 3. Bis-Boc protected aldehyde 44 was obtained
through several steps from D-tryptophan methyl ester HCl salt.
Stannous chloride mediated condensation of aldehyde 44 with
diazoketone 48 provided the 1,3-diketone 45, which cyclized
with hydrazine to furnish pyrazole 46. 7 The Pictet-Spengler
cyclization of 47 with tetrahydro-2H-pyran-4-carbaldehyde
afforded compound 39.
Table 6. Activity of the compounds from the “methyl scan” a
compd
Human sstr3
Human sstr3 functional cAMP
antagonism EC50, nM (% inh.)
binding IC50, nM
14
2.8
22 (87%)
5.7 (119%)
>1000 (39% )
ND
1
10
33
34
36
38
225
1333
26
150 (83%)
>1000 (18%)
573
a. ND = Not determined.
Table 7. Compounds with a substituted phenyl on the C-3 imidazole ring
Scheme 3. Reagents and conditions: (i) Boc2O, Et3N, DMAP, THF, rt (ii)
NaBH4, LiCl, THF, EtOH, rt, 16 h (iii) a. (COCl)2, DMSO, CH2Cl2, -78 °C; b.
Et3N, -78 °C to rt (iv) SnCl2, intermediate 48, CH2Cl2, rt (v) hydrazine, EtOH,
rt, 16 h (vi) HCl, MeOH (vii) a. (COCl)2, DMF (cat.), CH2Cl2, 0 °C to rt; b.
CH2N2 , Et3N, THF, 0 °C to rt.
Similarly, compound 40 was derived from precursor 52
containing an imidazole moiety (Scheme 4). Diazo ketone 50
was prepared from bis Boc protected ester 42. Oxidation of the
diazo ketone 50 with dimethyl dioxirane gave glyoxal 51
cleanly. 8 Condensation of glyoxal with benzaldehyde under
excess NH4OAc provided intermediate 52, which was
subsequently converted to compound 40.
compd
R
X, Y, Z
Human sstr3 Human sstr3 functional
binding IC50
,
cAMP antagonism
nM
EC50, nM
(% inh.)
H
F
X = NH, Y = N,
Z = CH
14
22 (87%)
1
X = NH, Y = N,
Z = CH
2.8
5.7 (119%)
10
F
X = CH, Y = NH,
Z = N
280
487
938 (79%)
405 (60%)
39
40
H
X = N, Y = CH,
Z = NH
F
X = N, Y = NH,
Z = N
282
201 (81%)
41
Scheme 4. Reagents and conditions: (i) LiOH, THF, H2O, rt (ii) a. ClCO2(i-
Bu), Et3N, THF, -20 °C to -10 °C, 30 min; b. CH2N2 in ether was added, -
10 °C to rt, 3 h (iii) dimethyl dioxirane, acetone, rt, 10 min (iv) PhCHO,
NH4OAc, MeOH, reflux, 16 h.
We also attempted to replace the imidazole ring with an
analogous five-membered hetereocycle possessing a NH moiety
(Table 7). The activities of the compounds (39-41) are all
inferior to the original imidazole compounds 1 and 10.
Compound 39 with a pyrazole ring suffered ~10 fold loss in
potency compared with the imidazole compoud 10. Compound
40 with an alternative substitution on the imidazole is also
The preparation of compound 41 again focused on the
preparation of triazole intermediate 55 (Scheme 5). Acyl
hydrazine 54 readily available from Boc-D-Trp 53 was
condensed with imidothioate 57 to furnish intermediate 55 after

10
11 (72%)
ND
87 (n=4)
ND
61
62
re-protection of the amino group to facilitate the isolation.
Triazole 55 was then converted to compound 41 following the
general Pictect-Spengler cyclization protocol decribed above.
133
109
163
193
782
>1333
786
>1333
39
196 (119%)
432 (94%)
ND
4026 (n=2)
7518 (n=2)
2701 (n=2)
ND
63
64
65
4499 (106%)
>1000 (10%)
>1000 (48%)
>1000 (19%)
67 (106%)
199 (91%)
3.1 (76%)
66
Scheme 5. Reagents and conditions: (i) a. HOBt, EDC, DMF, rt, 2 h; b.
hydrazine, 0 °C to rt, 16 h (ii) a. intermediate 57, EtOH, reflux 48 h; b. Boc2O,
Et3N, rt, 16 h (iii) MeI, acetone, rt, 16 h.
ND
67
4393 (n=2)
ND
68
Heterocyclic substitutions replacing the phenyl ring on the
imidazole were then explored (Table 8). These compounds were
prepared according the general method shown in Scheme 1.
Compound 58 with 3-thiophenyl gives slight loss in sstr3 and a
slight increase on hERG off-target liability. Compounds 59 and
60, with additional substituents on the thiophenyl offer good sstr3
binding potency but with lessened sstr3 antagonist activity.
Compound 61 with 3-benzothiophene substitution has excellent
sstr3 binding and functional activity but suffers from a stronger
binding with hERG. Closely related compound 62 with 3-indolyl
substitution gives poor sstr3 binding activity. Compounds 63,
64, and 65 with pyridyl substitutions are less potent in sstr3
binding assay. However, they all have reduced activity in hERG
binding assay. Heterocycles with two hetereo atoms, either
diazine or thiazole (66-69) all result in large loss in sstr3 activity.
Given that compounds 10 and 13 with para-F and para-Cl on a
phenyl ring offered good sstr3 binding activity, we prepared 70
and 71 by changing phenyl to pyridyl. Compound 70 with fluoro
substituent provides better profile than 71 with chloro
substituent. Since compound 20 with meta-methoxy on phenyl
ring gives good sstr3 activity, we also prepared the pyridyl
version: compound 72, which shows good sstr3 activity and less
hERG binding compared to compound 1. From this effort, we
identified 5-fluoropyridin-2-yl (70) and 6-methoxypyridin-2-yl
(72) as the most suitable substituents on the imidazole ring.
69
1742 (n=2)
465 (n=2)
1025 (n=2)
70
74
71
8.0
72
a. ND = not determined.
The SAR accumulated above was incorporated into an effort
of modifying MK-4256 (Table 9). In all cases, the Pictet-
Spengler cyclization gave two diastereomers differing at the C-1
position. Compounds 73a-76a were determined to have the
configuration at C-1 identical to MK-4256 while 73b-76b have
the C-1 configurations opposite to MK-4256. In general, 73a-
76a are more potent on sstr3 than the corresponding
diatereomers, 73b-76b. Compound 73a with an unsubstituted
phenyl has an in vitro profile similar to MK-4256. 74a with a
pyridyl substituent gives reduced sstr3 activity but with an
improved hERG off-target profile consistent with compound 63
discussed above. Addition of a fluoro substitution resulted in
75a with good sstr3 activity and less hERG binding than MK-
4256. The methoxy pyridyl compound 76a also has good sstr3
activity but its hERG off-target profile is slightly inferior to 75a.
Besides its potent activity on sstr3, 75a also possesses excellent
selectivities over other sstr subtypes. The binding potencies on
sstr1, sstr2, sstr4, and sstr5 are 2.3 µM, >10 µM, 3.2 µM, and 10
µM, respectively.
Compound 75a incorporating 5-fluoro-
pyridin-2-yl on the imidazole ring is the best compound overall
from this SAR effort.
Table 8. Compounds with a heterocycle on the imidazole ring a
Table 9. Profiles of MK-4256 analogs varying substitutions on imidazole a
compd
-R
Human sstr3
binding IC50 functional cAMP
human sstr3
hERG binding IC50
nM
,
(nM)
antagonism EC50
nM (% inh.)
,
-Ph
14 +S.D.
15
22 (87%)
380 (n=2)
146 (n=2)
1
48 (129%)
58
compd
-R
Human sstr3
binding IC50 functional cAMP
human sstr3
hERG binding IC50
nM
,
(nM)
0.66
antagonism EC50
nM (% inh.)
,
17
34
206 (97%)
522 (n=2)
ND
59
60
0.95 (83%)
1701 (n=2)
MK-
4256
138 (102%)

1.4
0.46 (91%)
1817 (n=2)
iv t1/2 (h)
Vdss (L/kg)
F (%)
1.2
1.2
28
1.1
2.4
22
6.1
0.86
62
73a
2.6
9.4
3.3 (96%)
5.5 (69%)
1753 (n=2)
5503 (n=2)
73b
74a
Compound 75a was evaluated for hERG off-target activity
since MK-4256 was discontinued due to this off-target activity.
Table 11 lists the in vitro data with MK-4256 as the benchmark.
The IC50 of 75a was 8.5 μM in hERG binding assay. Consistent
with the binding data, in PatchXpress functional assay, the IC50
of 75a was 12 μM. The in vitro binding and functional assays
represent approximately 3-5 fold improvement over MK-4256.
24
77 (55%)
2.5 (85%)
4895 (n=2)
8544 (n=4)
74b
75a
3.6
6.9
2.1
2.9
7.4 (142%)
4.8 (117%)
17 (101%)
6481 (n=2)
3619 (n=2)
ND
75b
76a
Table 11. Comparison of in vitro hERG off-target activity of 75a with MK-
4256
76b
species
75a
MK-4256
a. ND = not determined
hERG binding assay,
8.5
12
1.7
3.4
IC50 (μM)
Compound 75a was profiled in a mouse GTT efficacy model
(Figure 2). 75a was shown to dose-dependently reduce glucose
excursion. At 1 mg/kg dose, 75a reduces glucose excursion by
87%. The plasma drug level at 2.5 h time point post dosing is 180
nM. The efficacy of 75a observed is similar to that of MK-4256.
hERG PatchXpress,
IC50 (μM)
To further assess the hERG off-target liability of 75a, this
compound was evaluated in the anesthetized vagus intact CV dog
model. The result of this study is summarized in Table 12
together with the data of MK-4256 obtained previously.
Disappointingly, even at the lowest dose of 10 mg/kg with Cmax
of 32 μM, compound 75a elicits QTc interval prolongation by
10%. The QTc interval prolongation effect worsens as the dose
and the plasma exposure increase. The result clearly indicates
that 75a causes QTc interval prolongation in a dose-dependent
manner and its overall effect is very similar to MK-4256.
Therefore, the 3-5 fold improvement of 75a over MK-4256 as
observed from in vitro assays did not translate into a meaningful
differentiation of the in vivo CV dog results. Based on this key
data, the advancement of 75a was discontinued. The faltering of
compound 75a resulted in a major change in the strategy of the
subsequent medicinal chemistry effort. Instead of continuing
with incremental improvements regarding off-target hERG
activity, the later medicinal chemistry effort switched to
eliminating hERG activity altogether.
Figure 2. 75a dose dependently reduces glucose excursion in the mouse GTT
model
Table 12. Comparison of 75a and MK-4256 in anesthetized, vagus-intact, iv
CV dog model
MK-4256
Plasma Cmax
75a
With the favorable properties discussed above, 75a was
profiled in PK studies in preclinical species (Table 10). 75a
shows reasonable plasma clearance and bioavailability in mice
and rats. The PK profile in dog is excellent with low plasma
clearance, a long half-life and an excellent oral exposure.
Dose
QTc
Plasma Cmax
QTc
(mg/kg
iv)
(uM)
increase
(μM)
increase
10
18
6%
8%
32
72
10%
12%
Table 10. Pharmacokinetics of 75a in preclinical species
20
45
92
species
mouse
rat
dog
30
12%
125
17%
Dose (iv/po)
(mg/kg)
4/10
0.85
1/2
0.5/1
13.3
In summary, we have described an extensive SAR study at the
C-3 position of tetrahydro-β-carboline class of sstr3 antagonists
using compound 1 and MK-4256 as the benchmarks. From this
study, we identified 5-fluoro-pyridin-2-yl as the optimal
substituent on the imidazole ring to balance sstr3 activity and
hERG off-target liability. As a representative, compound 75a
oral AUCN,
µM•h/(mg/kg)
0.28
iv CL
(ml/min/kg)
12
27
1.6

shows not only excellent sstr3 activity and good PK profiles but
also a 3-5 fold improvement on hERG liability over MK-4256
according to in vitro hERG assays. However, in the in vivo CV
dog model, 75a still elicits QTc prolongation to a similar extent
and at the similar plasma drug concentrations compared to MK-
4256.
Further advancement of 75a was put on hold.
Subsequently, the medicinal chemistry team embarked on a new
SAR direction tackling the C-1 position of the tetrahydro-β-
carboline scaffold with an objective of eliminating hERG off-
target liability completely. The 5-fluoro-pyridin-2-yl moiety
discussed above turned out to be the cornerstone of this effort
and this fragment was incorporated in nearly all the new
compounds prepared thereafter. The SAR effort at the C-1
position eventually led to our second clinical candidate MK-
1421, which features this 5-fluoro-pyridin-2-yl fragment.9

References and notes
1 Pasternak, A.; Feng, Z.; deJesus, R.; Ye, Z.; He, S.; Dobbelaar, P. H.; Bradley, S. A; Chicchi, G. G.; Tsao, K.; Trusca, D.;
Eiermann, G. J.; Feng, Y.; Wu, M.; Shao, Q.; Zhang, B.; Nargund, R. P.; Mills, S. G.; Howard, A. D.; Yang, L.; Zhou, Y. ACS
Med. Chem. Lett. 2012, 3, 289.
2 Poitout, L.; Roubert, P.; Contour-Galcera, M.-O.; Moinet, C.; Lannoy, J.; Pommier, J.; Plas, P.; Bigg, D.; Thurieau, C. J. Med.
Chem. 2001, 44, 2990.
3 Novartis AG (Troxler, T. J.; Hurth, K.; Hoyer, D.) β-Carboline Derivatives and its Pharmaceutical Use Against Depression and
Anxiety. U.S. 6,861,430 B2, March 1, 2005.
4 Pasternak, A.; Feng, Z.; de Jesus, R.; Ye, Z.; He, S.; Dobbelaar, P.; Bradley, S. A.; Chicchi, G. G.; Tsao, K.-L.; Trusca, D.;
Eiermann, G. J.; Li, C.; Feng, Y.; Wu, M.; Shao, Q.; Zhang, B. B.; Nargund, R.; Mills, S. G.; Howard, A. D.; Yang, L.; Zhou,
Y.-P. ACS Med. Chem. Lett. 2012, 3, 289.
5 He, S.; Ye, Z.; Truong, Q.; Shah, S.; Du, W.; Guo, L.; Dobbelaar, P. H.; Lai, Z.; Liu, J.; Jian, T.; Qi, H.; Bakshi, R. K.; Hong,
Q.; Dellureficio, J.; Pasternak, A.; Feng, Z.; deJesus, R.; Yang, L.; Reibarkh, M.; Bradley, S. A.; Holmes, M. A.; Ball, R. G.;
Ruck, R. T.; Huffman, M. A.; Wong, F.; Samuel, K.; Reddy, V. B.; Mitelman, S.; Tong, S. X.; Chicchi, G. G.; Tsao, K.-L.;
Trusca, D.; Wu, M.; Shao, Q.; Trujillo, M. E.; Eiermann, G. J.; Li, C.; Zhang, B. B.; Howard, A. D.; Zhou, Y.-P.; Nargund, R.
P.; Hagmann, W. K. ACS Med. Chem. Lett. 2012, 3, 484.
6 He, S.; Lai, Z.; Ye, Z.; Dobbelaar, W.; Shah, S. K.; Truong, Q.; Du, W.; Guo, L.; Liu, J.; Jian, T.; Qi, H.; Bakshi, R. K.; Hong,
Q.; Dellureficio, J.; Reibarkh, M.; Samuel, K.; Reddy, V. B.; Mitelman, S.; Tong, S. X.; Chicchi, G. G.; Tsao, K.-L.; Trusca, D.;
Wu, M.; Shao, Q.; Trujillo, M. E.; Fernandez, G.; Nelson, D.; Bunting, P.; Kerr, J.; Fitzgerald, P.; Morissette, P.; Volksdorf, S.;
Eiermann, G. J.; Li, C.; Zhang, B.; Howard, A. D.; Zhou, Y.-P.; Nargund, R. P.; Hagmann, W. K. ACS Med. Chem Lett. 2014, 5,
748.
7 Padwa, A.; Hornbuckle, S. F.; Zhang, Z.; Zhi, L. J. Org. Chem. 1990, 55, 5297.
8 Groarke, M.; Hartzoulakis, B.; McKervey, M. A.; Walker, B.; Williams, C. H. Bioorg. Med. Chem. Lett. 2000, 10, 153.
9 Shah, S. K.; He, S.; Guo, L.; Truong, Q.; Qi, H.; Du, W.; Lai, Z.; Liu, J.; Jian, T.; Hong, Q.; Dobbelaar, P.; Ye, Z.; Sherer, E.;
Feng, Z.; Yu, Y.; Wong, F.; Samuel, K.; Madiera, M.; Karanam, B. V.; Reddy, V. B.; Mitelman, S.; Tong, S. X.; Chicchi, G. G.;
Tsao, K.-L.; Trusca, D.; Feng, Y.; Wu, M.; Shao, Q.; Trujillo, M. E.; Eiermann, G. J.; Li, C.; Pachanski, M.; Fernandez, G.;
Nelson, D.; Bunting, P.; Morissette, P.; Volksdorf, S.; Kerr, J.; Zhang, B. B.; Howard, A. D.; Zhou, Y.-P.; Pasternak, A.;
Nargund, R. P.; Hagmann, W. K. ACS Med. Chem. Lett. 2015, 6, 513.

Graphical Abstract
To create your abstract, type over the instructions in the template box below.
Fonts or abstract dimensions should not be changed or altered.
SAR Exploration at the C-3 Position of Tetrahydro-β-
carboline sstr3 Antagonists
Leave this area blank for abstract info.
Shuwen He, Peter H. Dobbelaar, Liangqin Guo, Zhixiong Ye, Jian Liu, Tianying Jian, Quang Truong, Shrenik K. Shah, Wu Du, Hongbo Qi, Raman K.
Bakshi, Qingmei Hong, James D. Dellureficio, Edward Sherer, Alexander Pasternak, Zhe Feng, Mikhail Reibarkh, Melissa Lin, Koppara Samuel, Vijay B.
Reddy, Stan Mitelman, Sharon X. Tong, Gary G. Chicchi, Kwei-Lan Tsao, Dorina Trusca, Margaret Wu, Qing Shao, Maria E. Trujillo, Guillermo
Fernandez, Donald Nelson, Patricia Bunting, Janet Kerr, Patrick Fitzgerald, Pierre Morissette, Sylvia Volksdorf, George J. Eiermann, Cai Li, Bei B.
Zhang, Andrew D. Howard, Yun-Ping Zhou, Ravi P. Nargund, and William K. Hagmann
MK-4256, a tetrahydro-β-carboline sstr3 antagonist, was discontinued due to a cardiovascular (CV)
adverse effect observed in dogs. Additional investigations revealed that the CV liability (QTc
prolongation) was caused by the hERG off-target activity of MK-4256 and was not due to sstr3
antagonism. In this report, we describe our extensive SAR effort at the C3 position of the tetrahydro-β-
carboline structure. This effort resulted in identification of 5-fluoro-pyridin-2-yl as the optimal
substituent on the imidazole ring to balance sstr3 activity and hERG off-target liability.