Discovery and biological characterization of capromorelin analogues with extended half-lives

Article abstract of DOI:10.1016/S0960-894X(02)00734-5

New tert-butyl, picolyl and fluorinated analogues of capromorelin (3), a short-acting growth hormone secretagogue (GHS), were prepared as part of a program to identify long-acting GHSs that increase 24-h plasma IGF-1 levels. Compounds 4c and 4d (ACD LogD values ≥2.9) displayed extended plasma elimination half-lives in dogs, primarily due to high volumes of distribution, but showed weak GH secretagogue activities in rats (ED₅₀s>10 mg/kg). A less lipophilic derivative 4 (ACD LogD=1.6) exhibited a shorter canine half-life, but stimulated GH secretion in two animal species. Repeat oral dosing of 4 in dogs for 29 days (6 mg/kg) resulted in a significant down-regulation of the post dose GH response and a 60 and 40% increase in IGF-1 levels relative to pre-dose levels at the 8- and 24-h post dose time points. Compound 4 (CP-464709-18) has been selected as a development candidate for the treatment of frailty.

Full text of DOI:10.1016/S0960-894X(02)00734-5

Bioorganic & Medicinal Chemistry Letters 12 (2002) 3279–3282
Discovery and Biological Characterization of Capromorelin
Analogues with Extended Half-Lives
Philip A. Carpino,* Bruce A. Lefker, Steven M. Toler, Lydia C. Pan, John R. Hadcock,
Marianne C. Murray, Ewell R. Cook, Joseph N. DiBrino, Shari L. DeNinno,
Kristin L. Chidsey-Frink, William A. Hada, John Inthavongsay, Sharon K. Lewis,
F. Michael Mangano, Michelle A. Mullins, David F. Nickerson, Oicheng Ng,
Christine M. Pirie, John A. Ragan, Colin R. Rose, David A. Tess, Ann S. Wright,
Li Yu, Michael P. Zawistoski, John C. Pettersen, Paul A. DaSilva-Jardine,
Theresa C. Wilson and David D. Thompson
Pfizer Global Research & Development, Groton Labs, MS8220-3004, Eastern Point Rd, Groton, CT 06340, USA
Received 23 May 2002; accepted 20 August 2002
Abstract—New tert-butyl, picolyl and fluorinated analogues of capromorelin (3), a short-acting growth hormone secretagogue
(GHS), were prepared as part of a program to identify long-acting GHSs that increase 24-h plasma IGF-1 levels. Compounds 4c
and 4d (ACD LogD values ꢀ2.9) displayed extended plasma elimination half-lives in dogs, primarily due to high volumes of dis-
tribution, but showed weak GH secretagogue activities in rats (ED50s>10 mg/kg). A less lipophilic derivative 4 (ACD LogD=1.6)
exhibited a shorter canine half-life, but stimulated GH secretion in two animal species. Repeat oral dosing of 4 in dogs for 29 days
(6 mg/kg) resulted in a significant down-regulation of the post dose GH response and a 60 and 40% increase in IGF-1 levels relative
to pre-dose levels at the 8- and 24-h post dose time points. Compound 4 (CP-464709-18) has been selected as a development can-
didate for the treatment of frailty.
# 2002 Elsevier Science Ltd. All rights reserved.
Over the past decade, two classes of growth hormone
secretagogues (GHSs) have undergone extensive clinical
evaluation. Short-duration GHSs stimulate growth
hormone (GH) release after single or repeat adminis-
tration and produce transient elevations in plasma
insulin-like growth factor-1 (IGF-1) levels.1 Compounds
such as pralmorelin (1)2 have been reported to increase
growth velocity in GH-deficient children.3 Long-dura-
tion GHSs increase the amplitudes and frequencies of
endogenous GH peaks and produce sustained elevations
of plasma IGF-1 levels.4 The post-dose GH response of
this class of drugs is attenuated due to negative feed-
back inhibition by IGF-1. One long-acting GHS, ibuta-
moren (2, MK-077, Fig. 1),5 has been shown to: (1)
increase markers of bone turnover in elderly adults and
obese young males;6 (2) increase energy expenditure and
fat-free mass in obese subjects;7 and (3) reverse diet-
induced catabolism in healthy volunteers.8 Both short-
and long-acting GHSs exhibit high affinities for the
human type 1a growth hormone secretagogue receptor
(hGHS-R1a), also known as the receptor for the new
gastrointestinal hormone ghrelin.
We have recently described the pharmacological char-
acterization of a short-duration pyrazolinone-piperidine
(PP) dipeptide GHS capromorelin (3, CP-424391-18).9
Because of the growing clinical precedent established
with long-acting GHSs such as ibutamoren, we sought
an orally active derivative of 3 with an extended half-life
(t1/2) in a beagle dog model used to predict GH/IGF-1
activity in humans. In order to manipulate the pharma-
cokinetic (PK) parameters that determine t1/2, potential
metabolic sites in 3 were blocked and physicochemical
properties of 3 were altered. Herein, the syntheses, bio-
logical activities, and PK properties of new tert-butyl,
picolyl, and fluorinated PP dipeptidyl analogues are
described. This work led to the discovery of 4, a long-
duration GHS that increases 24-h plasma IGF-1 levels
in dogs.
*Corresponding author. Fax: +1-860-715-8067; e-mail: philip_a_
carpino@groton.pfizer.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00734-5

3280
P. A. Carpino et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3279–3282
Figure 1. Structures of selected GHSs.
Scheme 1. Syntheses of 4 and 4a–d. Reagents and conditions: (a) Boc2O, DMAP, CH2Cl2, 0 ꢁC; (b) PhCH2Br or 2-PyrCH2Cl, NaH, DMF; (c)
R1NHNH2, EtOH, reflux; (d) toluene, cat. HOAc (if necessary), reflux; (e) TFA, 0 ꢁC; (f) HOAt, EDC, NEt3, CH2Cl2, 0 ꢁC; (g) concd HCl, EtOH;
(h) (l)-tartaric acid, MeOH; (i) crystallization from hot EtOAc.
The picolyl PP dipeptide GHS 4 was prepared from com-
mercially available methyl 4-oxo-3-piperidinecarboxylate
(5) as shown in Scheme 1. Treatment of 5 with di-tert-
butyldicarbonate (Boc2O) provided 6 which was alky-
lated with 2-picolyl chloride using sodium hydride as a
base to give 7. Condensation of the keto ester with
2,2,2-trifluoroethylhydrazine yielded the pyrazolinone-
piperidine heterocycle 8. Subsequent acid-catalyzed
removal of the N-tert-butoxycarbonyl (N-Boc) protect-
ing group gave 9. Cross-coupling of the racemic PP
amine with the dipeptide Boc-Aib-(d)-O-2,4-di-F-Bn-
Ser-OH (10a, see Scheme 2 for preparation) using 1-[3-
Ser-OH (10c),13 separating the diastereomers, and then
removing the N-Boc protecting groups under acidic
conditions. Compounds 9a–c were prepared in an ana-
logous manner to 9 using benzyl bromide in the initial
alkylation step and an appropriately substituted alkyl
hydrazine in the condensation step. The reaction of 7a
with tert-butylhydrazine in ethanol gave a hydrazone
intermediate which underwent condensation in toluene
only in the presence of acetic acid.
Pharmacokinetic evaluations of capromorelin analogues
4 and 4a–d were carried out in male beagle dogs. Plasma
elimination half-life, plasma clearance (CL) and volume
of distribution (Vd) values were measured following
intravenous (iv) administration of drug (1 mg/kg). Bind-
(dimethylamino)propyl]-3-ethylcarbodiimide
hydro-
chloride (EDC) and 1-hydroxy-7-azabenzotriazole
(HOAt)10 provided a mixture of two diastereomers,
(3aR)-11 and (3aS)-11. Addition of HOAt to the reac-
tion mixture minimized racemization of the dipeptide
subunit. The desired, less polar (3aR)-11 diastereomer
was de-protected with aqueous hydrochloric acid and
the product (3aR)-12 was treated with (l)-tartaric acid
to yield GHS 4.11 The stereochemistry of the 3a-carbon
center in the PP ring of 4 was established by X-ray
crystallographic analysis.12 Fluorinated and tert-butyl
analogues 4a–d (HCl salts) were obtained by coupling
the appropriately substituted PP intermediates 9a–c
with either 10b (see Scheme 2) or Boc-Aib-(d)-O-Bn-
Scheme 2. Syntheses of 10a,b. Reagents and conditions: (a) NaH, 3,4-
di-F-PhCH2Br or 2,4-di-F-PhCH2Br, DMF, 0 ꢁC; (b) concd HCl,
EtOH.

P. A. Carpino et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3279–3282
3281
ing affinities (Ki values) were determined via competitive
binding assays against [125I]-ghrelin (Perkin–Elmer
NEN-Life Sciences) in membranes prepared from
HEK293 cells over-expressing the hGHS-R1a. Func-
tional activities (in vitro GH release) were measured in
rat pituitary cell cultures using rat specific double anti-
body radioimmunoassays (NIDDK reagents). The in
vivo GHS activities of selected analogues were assessed
in an anesthesized Wistar rat model following iv
administration and in a beagle dog model following oral
administration. The IGF-1 response was measured in a
beagle dog model using previously described protocols.
ation of xenobiotics.14 The phenyl ring on the 3a-benzyl
side chain in the PP moiety, the last remaining site for
chemical manipulation, was replaced by polar, weakly
basic heterocycles. Compound 4 is a close-in derivative
of 4d with a picolyl group in place of the benzyl group
on the PP ring, a different fluorine substitution pattern
on the (d)-O-Bn-Ser group in the dipeptide subunit, and
a reduced calculated logD value (see Table 1). In dogs,
the plasma clearance of 4 was low in comparison to
more lipophilic analogues. Volume of distribution was
also small even though multiple fluorine substituents
were present throughout the molecule. Despite a
decrease in Vd, the half-life of 4 was only slightly
shorter than the half-lives of related analogues without
picolyl groups. Intrinsic clearances of 4, 4c, and 4d in
dog microsome preparations were not measured, so the
specific effects of reducing lipophilicity on metabolic
clearance were not determined.
The short half-life of capromorelin in the dog was
attributed to moderate Vd and moderate-to-high
plasma CL. Major routes of hepatic metabolism included
N-de-methylation of the PP ring, oxidation and de-
benzylation of the (d)-O-Bn-Ser side chain, and oxida-
tion of one of the methyl groups on the a-aminoisobu-
tyramide (Aib) subunit (data not shown). Replacement
of the methyl group on the PP ring in capromorelin with
a tert-butyl group as in 4a led to an increase in the
canine t1/2, primarily because of a decrease in plasma
clearance (see Table 1). Introduction of two fluorine
substituents on the phenyl ring in the (d)-O-Bn-Ser
moiety of capromorelin (see 4b) also resulted in an
extension of the half-life, but now because of an
increase in Vd. Incorporation of lipophilic substituents
at metabolic sites on both the PP ring and the side chain
in the dipeptide moiety provided compounds such as 4c
and 4d with moderately long half-lives due to large
volumes of distribution. Plasma clearances for many of
the compounds were often unpredictable, probably
because potential sites of metabolism in the Aib moiety
were not blocked. Modification of the Aib group was
not undertaken because of the restrictive structure–
activity relationships in the N-terminal region of the
dipeptide.
Selected capromorelin analogues with extended half-
lives (4 and 4b–d) exhibited varying GHS activities in
the primary screens (see Table 2). The picolyl derivative
4 and capromorelin were equipotent in the binding and
pituicyte cell culture assays, but 4 was 2-fold less active
in the anesthesized rat model. The fluorinated com-
pound 4b was approximately 3-fold less potent than
capromorelin in the in vitro assays, but 6-fold less active
in the anesthesized rat model. The most lipophilic ana-
logues 4c and 4d exhibited subnanomolar affinities in
the binding assay, with Ki values similar to ghrelin, the
endogenous GHS-R1a ligand. These two analogues stim-
ulated GH release in pituicyte cell cultures with EC50
values less than 5 nM, but surprisingly showed no in
vivo activity in the anesthesized rat model.
Compounds 4, 4c, and 4d possessed comparable half-
lives in dogs (t1/2sꢀ3.7 h) and showed equivalent EC50
values in rat pituicyte cell cultures, but only 4 with a
picolyl group and a low calculated logD value exhibited
GHS activity in an in vivo rat model. The unbound
(free) fraction of 4 in rat plasma was high (39%), pro-
viding a possible explanation for its robust activity. The
unbound fractions of 4c and 4d were not measured,
however, high protein binding could have reduced the
amount of free drug capable of interacting with the
GHS-R1a. Very large Vds in the rat as in the dog could
also have limited the concentrations of these drugs in
the plasma compartment.
In an attempt to lower plasma clearance values in ana-
logues with tert-butyl or trifluoroethyl groups on the PP
ring and fluorine groups in the dipeptide moiety, the
overall lipophilicity of the compounds was reduced by
decreasing the calculated octanol–water distribution
coefficient (ACD logD). Lipophilic compounds tend to
exhibit high metabolic clearances because of the large
hydrophobic binding sites in many of the cytochrome
P450 enzymes that are involved in the biotransform-
Table 1. Canine t1/2, Vd and CL values and calculated lipophilicities
Table 2. Binding affinities and GHS activities
Ki (nM)a,d
EC50 (nM)b,e
ED50 (mg/kg)c,e
a
Compd
t1/2
(h)
Vd
(L/kg)
CL
(mL/min/kg)
ACD
LogDb,c
Compd
Ghrelin
Capromorelin (3)
4
4b
4c
4d
0.42
7
5
18
0.34
0.45
Capromorelin (3)
4
4a
4b
4c
4d
1.3
3.7
2.5
2.7
3.9
4.7
2
19
5.3
13
17
19
13
1.7
1.6
2.9
1.7
2.9
3.0
3
3
10
4
0.05
0.1
0.3
1.5
2.2
3.5
5.9
5.1
>10
>10
2
aCompetitive binding assay, [125I]-ghrelin as radioligand.
bGH release in pituitary cell cultures.
cGH release in anesthesized rat model.
dAverage of duplicate experiments.
aPlasma elimination half-life (0–8 h).
bCalculated using Advanced Chemical Development ACD/LogD
Suite.
cpH=7.0.
eAverage of triplicate experiments.

3282
P. A. Carpino et al. / Bioorg. Med. Chem. Lett. 12 (2002) 3279–3282
With a moderately long canine half-life, compound 4
behaved similarly to other long-duration GHSs in dog
models. Compound 4 stimulated canine GH secretion
after a single 1 mg/kg oral dose, with a mean peak
height of 60Æ17 ng/mL. Repeat oral administration of
4 (6 mg/kg) for 29 days resulted in a down-regulation of
the post dose GH response, with peak GH levels
declining from 46 ng/mL on day 1 to 12 ng/mL on day
29. The serum IGF-1 levels increased 60% above pre-
dose levels at an 8-h post dose time point, though the
levels declined to 40% above pre-dose levels at the 24-h
post-dose time point.
References and Notes
1. Yang, L.; Morriello, G.; Patchett, A. A.; Leung, K.; Jacks,
T.; Cheng, K.; Schleim, K. D.; Feeney, W.; Chan Wanda, W.-
S.; Chiu, S.-H. L.; Smith, R. G. J. Med. Chem. 1998, 41, 2439.
2. d-Alanyl-3-(2-naphthalenyl)-d-alanyl-l-alanyl-l-tryptophyl-
d-phenylalanyl-l-lysinamide.
3. Mericq, V.; Cassorla, F.; Salazar, T.; Avila, A.; Iniguez, G.;
Bowers, C. Y.; Merriam, G. R. J. Clin. Endocrinol. Metab.
1998, 83, 2355.
4. Hickey, G. J.; Jacks, T. M.; Schleim, K.-D.; Frazier, E.;
Chen, H. Y.; Krupa, D.; Feeney, W.; Nargund, R.; Patchett,
A. A.; Smith, R. G. J. Endocrinol. 1997, 152, 183.
5. 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. J.; Jacks, T.; Schleim, K.; Pong, S.-
S.; Chaung, L.-Y. P.; Chen, H. Y.; Frazier, E.; Leung, K. H.;
Chiu, S.-H.; Smith, R. G. Proc. Natl. Acad. Sci. U.S.A. 1995,
92, 7001.
6. Svensson, J.; Ohlsson, C.; Jansson, J.-O.; Murphy, G.;
Wyss, D.; Krupa, D.; Cerchio, K.; Polvino, W.; Gertz, B.;
Baylink, D.; Mohan, S.; Bengtsson, B.-A. J. Bone Miner. Res.
1998, 13, 1158.
7. Svensson, J.; Lonn, L.; Jansson, J.-O.; Murphy, G.; Wyss,
D.; Krupa, D.; Cerchio, K.; Polvino, W.; Gertz, B.; Boseaus,
I.; Sjostrom, L.; Bengtsson, B.-A. J. Clin. Endocrinol. Metab.
1998, 83, 362.
8. Murphy, M. G.; Plunkett, L. M.; Gertz, B. J.; He, W.;
Wittreich, J.; Polvino, W. M.; Clemmons, D. R. J. Clin.
Endocrinol. Metab. 1998, 83, 320.
9. 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.; DaSilva-Jardine,
P. A.; Thompson, D. D. Endocrine 2001, 14, 121.
10. Carpino, L. A. J. Am. Chem. Soc. 1993, 115, 4397.
11. Physicochemical and spectral data for 4: mp 188–189 ꢁC;
1H NMR (MeOH-d4) d 8.28 (d, 1H), 7.59 (t, 1H), 7.41–7.39
(m, 1H), 7.18–7.13 (m, 1H), 6.92 (t, 1H), 5.20 (t, 1H), 4.56 (bs,
3H), 4.36 (s, 2H), 4.31–4.25 (m, 1H), 4.13–4.06 (m, 1H), 3.78
(d, 2H), 3.21 (t, 1H), 3.18–2.96 (m, 2H), 2.65–2.55 (m, 2H),
1.57 (d, 6H); HRMS m/z 611 (M+ +1).
Pharmacokinetic properties of 4 were measured in
female Sprague–Dawley rats and in male and female
beagle dogs. In rats, after an iv dose of 1 mg/kg of free
base equivalents (1 mgA/kg), mean values for Vd, CL,
and t1/2 were 4.9 L/kg, 46 mL/min/kg and 1.9 h. Fol-
lowing a 1 mgA/kg oral dose, maximum plasma con-
centrations (Cmax) of 88 ng/mL were achieved at 0.25 h
(Tmax). Bioavailability of 4 was 59%, with an apparent
oral t1/2 of 3.9 h due to slow absorption. In the dog,
following a 1 mgA/kg oral dose, Cmax was 260 ng/mL,
with a Tmax of 1.3 h. Bioavailability was 77%, with an
apparent oral t1/2 of 4.2 h.
In conclusion, introduction of tert-butyl and fluorinated
alkyl groups on the PP ring in capromorelin (3) and
fluorine substituents on the phenyl group in the Aib-(d)-
O-Bn-Ser subunit provided new dipeptidyl analogues
(4c and 4d) with large Vds, moderate CLs, and long
half-lives in dogs, but weak GH activities in rat models.
Modulation of the lipophilicity of a close-in analogue of
these compounds by replacing a benzyl group with a 2-
picolyl group led to a decrease in plasma CL and Vd in
the dog, but no significant change in t1/2. Compound 4
possessed a suitably extended canine t1/2, but unlike
more lipophilic analogues, stimulated GH secretion in
two animal species. Repeat oral administration of 4
in dogs increased 24-h plasma IGF-1 levels and resulted
in a down-regulation of the post-dose GH reponse.
Compound 4 is an orally active, long-duration GHS
that has been selected as a development candidate for
the treatment of frailty.
12. Busch, F. R.; Chiu, C. K.-F.; Meltz, C. N.; Post, R. J.;
Rose, P. R. EP 1031575, 2000.
13. Dorziotis, I.; Houpis, I.; Molina, A.; Volante, R. US
Patent 6 046 333, 2000.
14. van de Waterbeemd, H.; Smith, D. A.; Beaumont, K.;
Walker, D. K. J. Med. Chem. 2001, 44, 1313.