Optimization of 1H-tetrazole-1-alkanenitriles as potent orally bioavailable growth hormone secretagogues

Article abstract of DOI:10.1016/j.bmcl.2008.01.100

1H-Tetrazole-1-alkanenitrile SR-9g exhibits a >10-fold in vivo potency enhancement over the lead nitrile 1 and has acceptable oral bioavailability in rats and dogs. An enantiospecific synthesis of 1H-tetrazole-1-alkanenitrile nitriles 9 has been developed.

Full text of DOI:10.1016/j.bmcl.2008.01.100

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 2067–2072
Optimization of 1H-tetrazole-1-alkanenitriles as potent
orally bioavailable growth hormone secretagogues
a,
a
b
b
*
´
´
Andres S. Hernandez, Stephen G. Swartz, Dorothy Slusarchyk, Mujing Yan,
R. Krishna Seethala,b Paul Sleph,b Gary Grover,b Kenneth Dickinson,b Leah Giupponi,b
Timothy W. Harper,c W. Griffith Humphreys,c Daniel A. Longhi,b Neil Flynn,b
Brian J. Murphy,b David A. Gordon,b Scott A. Biller,a
Jeffrey A. Robla and Joseph A. Tinoa,*
aDiscovery Chemistry, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543, USA
bMetabolic Disease Biology, Bristol-Myers Squibb Pharmaceutical Research Institute, Princeton, NJ 08543, USA
cPharmaceutical Candidate Optimization, Bristol-Myers Pharmaceutical Research Institute, Princeton, NJ 08543, USA
Received 11 September 2007; revised 25 January 2008; accepted 25 January 2008
Available online 31 January 2008
Abstract—1H-Tetrazole-1-alkanenitrile SR-9g exhibits a >10-fold in vivo potency enhancement over the lead nitrile 1 and has
acceptable oral bioavailability in rats and dogs. An enantiospecific synthesis of 1H-tetrazole-1-alkanenitrile nitriles 9 has been
developed.
ꢀ 2008 Elsevier Ltd. All rights reserved.
Both growth hormone-releasing hormone (GHRH)1
and the more recently discovered growth hormone
secretagogues (GHS)2 induce growth hormone (GH)3
release from the pituitary upon binding, respectively,
to the GHRH receptor4 and the GHS receptor (GHS-
R1a). The earliest reported GHS were oligopeptides,
called growth hormone-releasing peptides (GHRPs),
such as GHRP-6.5 Several orally bioavailable small pep-
tidomimetic GHS6 were synthesized before ghrelin was
identified as the endogenous GHS receptor ligand.7
44471111 and the two tetrazoles 1 and 2 previously re-
ported by us.12
The differences in polarity and lipophilicity between a
cyano and a phenyl group of tetrazoles 1 and 2 suggest
that the propanonitrile and phenylpropyl moieties may
project into different binding pockets. We sought to coa-
lesce the two structures to generate a high-affinity ago-
nist. Accordingly a set of 1H-tetrazole analogs 9a–9d
were prepared bearing a variety of branched alkyl chains
capped with phenyl and cyano in the hope of identifying
the spacers that would maintain the optimal orientation
of each substituent. This paper describes the SAR eluci-
dation for these analogs, culminating with the identifica-
tion of GHS agonists exhibiting >10-fold greater in vivo
potency compared to that of 1.
A series of small molecule orally active GHS described
in the literature8 contain a dipeptide formed from
O-benzyl-D-serine or its carbon- analogue, R-2-amino-
5-phenylpentanoic acid, and 2-methyl alanine as a
common feature. Structure–function studies9 of ghrelin
suggest this dipeptide may correspond to the N-terminus
of the GHRPs and ghrelin. Members of the above series
shown in Figure 1 are MK-677,6b CP-424391,10 LY
Mitsunobu promoted alkylation of the chiral S-tetrazole
6 with racemic alcohols 3–5 generated 1H-tetrazoles
7a–7d along with the corresponding isomeric 2H-tetra-
zoles 8a–8d.12 As outlined in Scheme 1, alcohol 3 was
readily obtained from a commercially available ketone;
alcohols 4 and 5 were prepared as shown from the indi-
cated esters. Following separation of the tetrazole regio-
isomeric mixture by column flash chromatography,13
TFA-mediated cleavage of the BOC-protecting group of
Keywords: Growth hormone secretagogues; Tetrazole; Enantiospecific
synthesis.
*
Corresponding authors. Tel.: +1 609 818 7127; fax: +1 609 818 6810
(A.S.H); e-mail: andres.hernandez@bms.com
0960-894X/$ - see front matter ꢀ 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2008.01.100

´
A. S. Hernandez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2067–2072
2068
NH2
NH2
O
O
NaBH4
58%
NH
3
O
NH
N
O
HO
O
CN
CN
O
N
O
N S
O
O
N
Ca(BH4)2
93%
NC
N
NC
4
5
O
HO
MK-677
EtO2C
CP-424391
NH2
1. LiCN, 67%
O
NH
2. Ca(BH4)2, 46%
MeO2C
HO
CN
N
N
O
N
N
O
H
NHBOC
O
O
NH
LY 444711
BnO
N
N
NH
N
NH2
NH2
O
O
6
NH
NH
O
O
1. ROH (3,4 or 5), Ph3P, DEAD
2. Regioisomer Separation
N
N
N
N
N
N
N
N
CN
NHR2
NHR2
O
1
2
O
Figure 1. GHSs sharing a dipeptide structural motif.
NH
NH
BnO
N
BnO
+
1H-tetrazoles 7a, 7c and 7d yielded chiral mixtures of SR
and SS diastereomeric tetrazoles 9a, 9c and 9d. Hydroge-
nation (5% Pt/C) of the diastereomeric mixture 7a fol-
lowed by BOC cleavage generated saturated analogs 9b.
Although the mixture of SR and SS diastereomers 9a
and 9b were separated into single epimers by reverse phase
HPLC, the separation of the diastereomeric mixtures 9c
and 9d was not attempted given their lack of in vitro activ-
ity relative to 9a and 9b. The BOC-protected mixture of
2H-tetrazole 8a was separated into single chiral diastereo-
mers by reverse phase HPLC and then converted to single
epimers 10a. The reduced 2H analogs 10b were synthe-
sized from single epimers 8a by hydrogenation (5% Pt/
C) followed by BOC cleavage. TFA treatment converts
2H-tetrazole 8c and 8d to the chiral diastereomeric mix-
ture of SR and SS 2H-tetrazoles 10c and 10d.
N
N
N R
N
N
N
N
R
7, R2= BOC
9, R2= H
8 R2= BOC
TFA
TFA
10, R2= H
Scheme 1. Preparation of alcohols 3, 4 and 5, and synthesis of
tetrazole-alkanenitriles 9 and 10.
of 9a and 9b vastly exceeds that of the other. In other
words, all the activity resides in just one diastereomer
of the diastereomeric pair comprising 9a or 9b. The
observation that 9c and 9d exhibit comparable
in vitro agonist activity to 1 and 2 suggests that their
spacers allows either the phenyl or the cyano but not
both to assume the optimal orientation.
In agreement with previous SAR,12 the 2H-tetrazole
series is consistently less potent than the corresponding
1H regioisomer. The analogs in this series exhibit the
following order in decreasing potency: 10a > 10b ꢀ
10d > 10c. The difference in activity between epimers
of a same 2H regioisomer 10 is far smaller than the dif-
ferences found for the corresponding 1H regioisomer 9.
This observation infers that neither 2H diastereomer
binds in an optimal fashion.
Table 1 summarizes the results for nitrile analogs 9
and 10 in the in vitro binding and functional assays.12
Since functional activity of these full agonists was
10- to 75-fold greater than binding affinity for the
GHS receptor, the SAR discussion will focus on the
functional data. Only diastereomer 1 of 1H-tetrazoles
9a and 9b offered any promise of meeting the criteria
for enhanced activity, as they exhibited 60- and 40-fold
greater agonist potency relative to 1. This finding is
consistent with the existence of two binding sites, since
only the spacers of 9a and 9b are sufficiently long to
allow the tethered phenyl and cyano moieties to simul-
taneously assume the preferred conformation of 1 and
2. As would be expected given the three-dimensional
nature of the binding site, the binding affinity and
in vitro potency for one of the chiral diastereomers
Evaluation of the more active single chiral diastereomers
of 9a and 9b in the acute IV rat model12 revealed that
the respective 60- and 40-fold enhanced in vitro potency
relative to the parent 1 collapsed to only a 3-fold
14
improvement in ED250 at best (Table 2). ClogP in-
creased from À0.3 for 1 to approximately 2 upon incor-

´
A. S. Hernandez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2067–2072
2069
Table 1. In vitro activity of tetrazole nitriles 9 and 10
NH2
NH
N R
NH2
NH
O
O
BnO
BnO
N
N
N
N
N
N
R
N
9
10
a,b
EC50 (nM)
Compound
R
Substitution
Diastereomer
NA
Kia (nM)
1
1H
2940
30
CN
2
1H
NA
200
17
9a
1H
1H
2H
2H
1
2
1
2
12
0.5
62
737
453
2510
10a
19
202
N
9b
1H
1H
2H
2H
1
2
1
2
63
0.8
1380
2560
3940
139
150
244
10b
N
N
9c
10c
1H
2H
1/1 mix
1/1 mix
298
>5000
22
2470
9d
10d
1H
2H
1/1 mix
1/1 mix
611
>5000
27
302
N
a See Ref. 12 for detailed description.
b All compounds are full agonists.
poration of a 2-phenylethene/2-phenylethyl substituent
in 9a and 9b. This 200-fold predicted increase in lipo-
philicity may have adversely impacted ADME proper-
ties thereby accounting for the fact that the in vivo
potency of 9a and 9b relative to 1 does not reflect the
in vitro differential in functional activity (Table 2). In
an attempt to modulate the increase in liphophilicity
while enhancing in vitro potency, analogs 9e–9h were
prepared containing heteroatoms or polar functionality.
11 in 77% yield. Mitsunobu reaction of 11 with phenols
12e, 12f, 12h or 2-hydroxypiridine (12g) provided the
aryl ethers 13e, 13f, 13h or the heteroaryl ether 13g in
44–80%. The conversion to 13g was particularly ineffi-
cient because of competitive N-alkylation. Cyanide
addition to aziridines 13 followed by BOC cleavage fur-
nished the 3-amino-butanonitriles 14 in an average 71%
yield. Subsequent conversion of the racemic amines 14
to a mixture of SR and SS diastereomeric 1H-tetrazole
nitriles 9e–9h is summarized in Scheme 2. The benzyl-
oxyphenol 15h was quantitatively converted to its re-
lated phenol by hydrogenolysis before BOC cleavage.
Following separation by reverse phase HPLC, evalua-
tion of the individual diastereomers revealed that the
faster eluting epimer was consistently more active.15
A previously developed regiospecific route12 was em-
ployed to yield only the 1H-tetrazoles of analogs 9e–
9h. At this stage the SAR was not sufficiently developed
to confidently state whether the SR or SS diastereomer
was preferred. Consequently, ( )-methyl aziridine-2-car-
boxylate was utilized to prepare racemic 3-amino-bu-
tanonitrile 14 thereby insuring formation of both
diasteremers (Scheme 2).
The absolute stereochemistry of the more active epimer
was assigned following enantiospecific conversion of
chiral amines 14 to individual epimers of 9 using the
methodology of Scheme 2. The enantiospecific synthesis
of amines R-14 from readily available oxazolidine R-16
Methyl aziridine-2-carboxylate was BOC-protected and
then reduced with calcium borohydride to afford alcohol

´
A. S. Hernandez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2067–2072
2070
Table 2. Correlation between in vitro activity, in vivo activity and liphophilicity for 1H-tetrazole nitriles 9
NH2
O
NH
BnO
N
N
N R
N
9
a,b
EC50 (nM)
c
Compound
R
Diastereomer: (SR/SS)
Kia (nM)
ED250 (lmol/kg)
ClogP
À0.3
1
NA
2940
30
4.8
CN
9a
1
2
12
737
0.5
62
>2
ND
1.9
N
9b
9e
1
2
63
1380
0.8
139
1.7
ND
2.0
1.4
N
1
2
18
696
0.6
64
ND
ND
O
F
N
9f
9f
9g
1:SR
2:SS
1:SR
7.5
197
40
0.4
19
0.87
ND
0.33
1.3
1.3
0.7
O
O
O
N
N
N
0.4
N
N
9g
9h
2:SS
1:SR
503
5.7
19
ND
0.7
0.6
O
O
N
0.3
<0.15
OH
OH
N
9h
2:SS
240
23
ND
0.6
O
N
a See Ref. 12 for detailed description.
b All compounds are full agonists.
c In vivo activity of 1H-tetrazoles in the acute anesthetized IV rat model. See Ref. 12 and 14 for detailed description.

´
A. S. Hernandez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2067–2072
2071
ether R-17g in 76, 67 and 49% yields, respectively. Acid
methanolysis of oxazolidines R-17 and subsequent aziri-
dine formation induced by Mitsunobu activation of the
hydroxy function generated enantiomers R-13 in an
average 70% yield. Cyanide addition to R-13 and BOC
removal yielded enantiomerically pure amines R-14.
Condensation of N-Boc-O-benzyl-D-serine with amines
R-14f, R-14g or R-14h following the reaction sequence
outlined in Scheme 2 yielded nitriles SR-9f, SR-9g or
SR-9h, which spectroscopic analysis confirmed to be
the more active epimer 1 of 9f, 9g and 9h, respectively.
To further support this conclusion, an analogous reac-
tion sequence beginning with oxazolidine S-16, 2-
hydroxypyridine (12g) and N-Boc-O-benzyl-D-serine
yielded SS-9g, which was identical to epimer 2 of 9g.
Based on these structural correlations, we conclude that
the R configuration at the epimeric center is preferred
for activity.
OH
CO2Me
1. (BOC)2O, Et3N
BOCN
HN
2. Ca(BH4)2
77%
11
OH
12e, X= C-F
12f, X= C-H
12g, X= N
X
12h, X= C-OBn
Ph3P,
DEAD
1. KCN, DMSO
X
X
O
O
2. TFA, thioanisole
H2N
BOCN
N
14
13
NHR1
1. NMM, iBuO2CCl,
(+/-)-14
2. TMSN3, Ph3P,
O
NHBOC
CO2H
DEAD
NH
X
BnO
BnO
Analysis of analogs 9e, 9f, 9g and 9h (Table 2) revealed
that the impact of decreasing the calculated lipophilicity
20-fold had little impact on in vitro potency as exempli-
fied by the near equivalence of 9h and 9a. However, an
increase in polarity appears to have positively impacted
ADME properties resulting in a respective 15- and 30-
fold enhanced in vivo activity for SR-9g and SR-9h rel-
ative to nitrile 1.
O
3. TFA, thioanisole
4. BOC-Aib, EDAC,
HOAT
N
N
N
N
N
15, R1= BOC
9, R1= H
TFA
Scheme 2. Preparation of 3-aminobutanenitriles 14 and synthesis of
1H-tetrazole nitriles 9e–9h.
Further characterization of SR-9g and SR-9h revealed
SR-9h to be a more potent inhibitor of CYPs 2D6 and
3A4 than SR-9g (Table 3). Consequently, despite being
2-fold less potent in the in vivo model, SR-9g was cho-
sen as a lead compound for additional pharmacokinetic
studies. The oral bioavailability of 1H-tetrazole-1-
alkanenitrile SR-9g was 24% in rats and 55% in dogs,
somewhat less than the corresponding values of 56%
and 76% for the parent nitrile 1.
CO2Me
1. LiBH4, 97%
2.Ph3P, DEAD
BOCN
X
O
O
BOCN
R
-16
12f, X= C-H
12g, X= N
OH
O
X
12h, X= C-OBn
R
-17
1. p-TsOH,
MeOH
2. Ph3P,
DEAD
In summary, an enantiospecific route for the synthesis of
nitriles 9 has been developed. The potency of nitrile 1
has been optimized by introducing an 2-arylethyl moiety
which provides an additional interaction with the GHS
receptor. A precisely defined spatial relationship is
required between the newly added aryl and the existing
cyano group to achieve synergistic improvement in
activity and otherwise potency is unaffected or even
diminished. Heteroatoms were included in the 2-aryleth-
yl moiety to maintain similar physicochemical proper-
ties, such as lipophilicity, to those exhibited by nitrile
1. Only in vitro functional activity of polar analogs pro-
portionally translated to in vivo activity. These modifi-
cations led to 1H-tetrazole-1-alkanenitrile SR-9g which
exhibited >10-fold improved potency in vivo than the
lead nitrile 1 with acceptable oral bioavailability in rats
and dogs.
1. KCN, DMSO
X
X
2. TFA, thioanisole
O
O
H2N
BOCN
N
R
-14
R
-13
Scheme 3. Enantiospecific route to 3-aminobutanenitriles R-14.
and the appropriate phenol is depicted in Scheme 3.
Reduction of the methyl ester R-16 and condensation
of the resulting alcohol with phenols 12f, 12h or 2-
hydroxypiridine (12g), under Mitsunobu conditions,
afforded the aryl ethers R-17f, R-17h and the hetero aryl
Table 3. Cytochrome P450 profile for SR-9g and SR-9h
Compound
1A2 IC50 (lM)
2C9 IC50 (lM)
2C19 IC50 (lM)
2D6 IC50 (lM)
3A4/BFC IC50 (lM)
3A4/Bz-RES IC50 (lM)
SR-9g
SR-9h
>100
>66
>100
>66
33
42
>100
5.1
59
9.8
11
5.5

´
A. S. Hernandez et al. / Bioorg. Med. Chem. Lett. 18 (2008) 2067–2072
2072
8. Carpino, P. A. Exp. Opin. Ther. Patents 2002, 12, 1599.
9. Dodge, J. A.; Heiman, M. L. Annu. Rep. Med. Chem.
2003, 38, 81.
Acknowledgments
We thank Dr. William Washburn, Dr. Peter T.W.
Cheng and Dr. Robert Zahler for their suggestions on
the manuscript.
10. Pan, L. C.; Carpino, P. A.; Lefker, B. A.; Ragan, J. A.;
Toler, S. M.; Pettersen, J. C.; Nettleton, D. O.; Ng, O.;
Pirie, C. M.; Chidsey-Frink, K.; Lu, B.; Nickerson, D. F.;
Tess, D. A.; Mullins, M. A.; MacLean, D. B.; Da Silva-
Jardine, P. A.; Thompson, D. D. Endocrine 2001, 14, 437.
11. (a) Seyler, D. E.; Dodge, J. A.; Osborne, J. J.; Cox, K. L.;
Viswanath, D.; Wilmot, A. F.; Keaton, M. J.; Heiman, M.
L.; Bryant, H. U.; Cutler, G. B., Jr. Drug Dev. Res. 2000,
49, 260; (b) Lugar, C. W.; Clay, M. P.; Lindstrom, T. D.;
Woodson, A. L.; Smiley, D.; Heiman, M. L.; Dodge, J. A.
Bioorg. Med. Chem. Lett. 2004, 14, 5873.
References and notes
1. Rivier, J.; Spiess, J.; Thorner, M.; Vale, W. Nature 1982,
300, 276.
2. (a) Bowers, C. Y. In Growth Hormone Secretagogues;
Bercu, B. B., Walker, R. F., Eds.; Springer-Verlag: New
York, 1996; pp 9–28; (b) Bowers, C. Y.; Reynolds, G. A.;
Durham, D.; Barrera, C. M.; Pezzoli, S. S.; Thorner, M.
O. J. Clin. Endocrinol. Metab. 1990, 70, 975.
´
12. Hernandez, A. S.; Cheng, P. T. W.; Musial, C. M.; Swartz,
S. G.; George, R. J.; Grover, G.; Slusarchyk, D.; Seethala,
R. K.; Smith, M.; Dickinson, K.; Giupponi, L.; Longhi,
D. A.; Flynn, N.; Murphy, B. J.; Gordon, D. A.; Biller, S.
A.; Robl, J. A.; Tino, J. A. Bioorg. Med. Chem. Lett. 2007,
17, 5928.
3. Reichlin, S.; Saperstain, R.; Jackson, I. M. D.; Boyd, A.
E.; Patel, Y. Ann. Rev. Physiol. 1976, 38, 389.
4. Howard, A. D.; Feighner, S. D.; Cully, D. F.; Arena, J. P.;
Liberator, P. A.; Rosenblum, C. I.; Hamelin, M.; Hreniuk,
D. L.; Palyha, O. C.; Anderson, J.; Paress, P. S.; Diaz, C.;
Chou, M.; Liu, K. K.; McKee, K. K.; Pong, S.-S.; Chaung,
L.-Y. P.; Elbrecht, A.; Dashkevicz, M.; Heavens, R.; Rigby,
M.; Sirinathsinghji, D. J. S.; Dean, D. C.; Melillo, D. G.;
Patchett, A. A.; Nargund, R. P.; Griffin, P. R.; DeMartino,
J. A.; Gupta, S. K.; Schaeffer, J. M.; Smith, R. G.; Van der
Ploeg, L. H. T. Science 1996, 273, 974.
13. 1H NMR easily distinguished between the 1H and 2H
isomers by comparison of the Dd between the methyls of
the 2-aminoisobutyric fragment. The Dd is significantly
larger in the 1H isomer than in the 2H isomer.
14. A dose response for MK-677 was performed in order to
benchmark tetrazole analogs versus a well studied refer-
ence. At the highest dose utilized, 5.23 lmol/kg, there was
no evidence of a plateau in the response. Higher doses
could not be achieved as toxicity was observed and
therefore an ED50 could not be determined. Since the
maximal tolerated dose of MK-677 effects a 500% increase
in GH, in vivo potencies for tetrazole based GHS are
expressed as ED250 or the dose at which a 250% increase in
plasma GH is achieved. The ED250 of MK-677 is
0.46 lmol/kg.
15. Analytical HPLC conditions: Column: Zorbax SB C18,
4.6 · 75 mm; Grad. T: 8 min; Flow R.: 2.5 mL/min.;
Solvent Grad.: 0–100% B; Wave: 220 nm (A = 10% MeOH
À90% H2O À0.2% H3PO4; B = 90% MeOH À10% H2O
À0.2% H3PO4). SR-9g RT/SS-9g RT = 5.30 min/5.65 min.
5. Bowers, C. Y.; Momany, F. A.; Reynolds, G. A.; Hong,
A. Endocrinology 1984, 114, 1537.
6. (a) DeVita, R. J.; Schoen, W. R.; Fischer, M. H.;
Fontier, A. J.; Pisano, J. M.; Wyvratt, M. J.; Cheng, K.;
Chan, W.-S.; Butler, B. S.; Hickey, G. J.; Jacks, T. M.;
Smith, R. G. Bioorg. Med. Chem. Lett. 1994, 4, 2229;
(b) Patchett, A. A.; Nargund, R. P.; Tata, J. R.; Chen,
M.-H.; Barakat, K. J.; Johnston, D. B. R.; Cheng, K.;
Chan, W. W.-S.; Butler, B.; Hickey, G.; Jacks, T.;
Schleim, K.; Pong, S.-S.; Chaung, L.-Y. P.; Chen, H.
Y.; Frazier, E.; Leung, K. H.; Chiu, S.-H. L.; Smith, R.
G. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 7001.
7. Kojima, M.; Hosoda, H.; Date, Y. Nature 1999, 402, 656.