New, potent, and selective peptidic oxytocin receptor agonists

Article abstract of DOI:10.1021/jm500365s

Mothers of preterm babies frequently have difficulty establishing or maintaining lactation, thought to be due to interference with the milk ejection reflex. Administration of exogenous oxytocin can produce alveolar contraction and adequate breast emptying resulting in establishment of successful lactation. The natural hormone oxytocin is not receptor-selective and may cause hyponatremia via V₂ receptor mediated antidiuresis. We have designed a series of potent oxytocin analogues containing N-alkylglycines in position 7 with excellent selectivity versus the related V_1a, V_1b, and V₂ vasopressin receptors and short half-life: agonists 31 ([2-ThiMeGly⁷]dOT), 47 (carba-6-[Phe²,BuGly ⁷]dOT), 55 (carba-6-[3-MeBzlGly⁷]dOT), and 57 (carba-1-[4-FBzlGly⁷]dOT) have EC₅₀ values at hOTR < 0.1 nM, selectivity ratios versus related human vasopressin receptors of >2000, IC₅₀ at hV_1aR > 500 nM, and total clearance in rats in the range of 60-80 mL min^-1 kg^-1. Compound 57 (FE 202767) is currently in clinical development for the treatment of preterm mothers requiring lactation support.

Full text of DOI:10.1021/jm500365s

Subscriber access provided by SEOUL NATL UNIV
Article
New, Potent and Selective Peptidic Oxytocin Receptor Agonists
Kazimierz Wisniewski, Sudarkodi Alagarsamy, Robert Galyean, Hiroe Tariga, Dorain Thompson, Brian Ly,
Halina Wisniewska, Steve Qi, Glenn Croston, Regent Laporte, Pierre J-M. Riviere, and Claudio D. Schteingart
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm500365s • Publication Date (Web): 29 May 2014
Downloaded from http://pubs.acs.org on June 6, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 45
Journal of Medicinal Chemistry
1
1
2
3
4
5
6
7
8
9
New, Potent and Selective Peptidic Oxytocin
Receptor Agonists
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Kazimierz Wiśniewski,* Sudarkodi Alagarsamy, Robert Galyean, Hiroe Tariga,
Dorain Thompson, Brian Ly, Halina Wiśniewska, Steve Qi, Glenn Croston,
Regent Laporte, Pierre JꢀM. Rivière and Claudio D. Schteingart
Ferring Research Institute Inc., 4245 Sorrento Valley Blvd., San Diego, CA 92121, USA
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 45
2
1
2
3
4
ABSTRACT
5
6
7
8
9
Mothers of preterm babies frequently have difficulty establishing or maintaining
lactation, thought to be due to interference with the milk ejection reflex. Administration
of exogenous oxytocin can produce alveolar contraction and adequate breast emptying
resulting in establishment of successful lactation. The natural hormone oxytocin is not
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
receptorꢀselective and may cause hyponatremia via V2 receptor mediated antidiuresis. We
have designed a series of potent oxytocin analogues containing Nꢀalkylglycines in
position 7 with excellent selectivity versus the related V1a, V1b, and V2 vasopressin
receptors and short halfꢀlife: agonists 31 ([2ꢀThiMeGly7]dOT), 47 (carbaꢀ6ꢀ
[Phe2,BuGly7]dOT), 55 (carbaꢀ6ꢀ[3ꢀMeBzlGly7]dOT) and 57 (carbaꢀ1ꢀ[4ꢀFBzlGly7]dOT)
have EC50 values at the hOTR < 0.1 nM, selectivity ratios versus related human
vasopressin receptors > 2000, IC50 at hV1aR > 500 nM, and total clearance in rats in the
range of 60ꢀ80 mL/min/kg. Compound 57 (FE 202767) is currently in clinical
development for the treatment of preterm mothers requiring lactation support.
ACS Paragon Plus Environment

Page 3 of 45
Journal of Medicinal Chemistry
3
1
2
3
4
INTRODUCTION
5
6
7
8
9
Of the four million births annually in the U.S.A, approximately 12% are preterm (born
prior to 37 weeks) with 1.4% being very low birthweight;1 incidences of 4 ꢀ 6 % are
reported in some European countries, and higher rates occur in less developed countries.2
Babies born preterm are at risk for a variety of complications resulting in morbidity and
mortality and in most cases require specialized treatment in hospital and may be subject
to adverse long term health and neurodevelopmental outcomes.2 It is well established that
early feeding of the preterm infant with mother’s own milk has significant advantages
over formula, and even over pasteurized donor milk, with lower incidence of necrotizing
enterocolitis, sepsis, infections, and readmissions3, 4 and shorter length of hospital stay.5
However, a significant proportion of preterm mothers experience difficulty or fail to
establish successful lactation.6 One contributing factor is that preterm infants are
typically treated in an intensive care unit and cannot nurse at the breast, requiring that
their mothers initiate and maintain lactation by manual and mechanical milk expression.
Although milk expression and frequent emptying of the breast may not be necessary to
initiate lactation, it is necessary for its successful maintenance. The hormone oxytocin
(OT), 1, is released into the circulation by the posterior pituitary in short pulses in
response to suckling and acts on the OTR on myoepithelial cells surrounding the breast
alveoli, causing them to contract and empty their contents (the “milk ejection reflex”). It
is believed that stress and other factors are responsible for the inhibition of oxytocin
release from the pituitary resulting in incomplete removal of milk and eventual downꢀ
regulation of milk production.7 Administration of exogenous oxytocin by the intranasal
route to preterm mothers produced significant increase in milk production in one study,8
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 45
4
1
2
but not in another at a lower dose.6 A nasal spray containing oxytocin at a concentration
of 40 IU/mL, Syntocinon, is approved for lactation support in a small number of
3
4
5
6
7
8
9
European countries. However, 1 is not selective versus the related receptors hV1a and hV2
(Table 1). It is markedly potent at the hV2 receptor of the kidneys to produce antidiuresis,
which may result in inappropriate water conservation9 and hyponatremia. Guidelines for
OT use have been created to prevent hyponatremia during labor induction10 and several
cases of severe hyponatremia by use of the nasal spray have been reported.11ꢀ13
We set out to design a new peptidic oxytocin agonist which would be safe and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
effective for lactation support in preterm mothers. As the compound would ideally be
selfꢀadministered multiple times a day before each breast pumping session, the intranasal
route of administration appeared the most convenient. Because only low doses can be
absorbed by this route, the agonist should have high potency to be efficacious. It should
also be selective versus the hV2 receptor in order to minimize the risk of hyponatremia in
the event of excessive or too frequent drug use by patients. Eliminating activity at the V1a
receptor would prevent local vasoconstriction at the site of application, which would be
exposed to a relatively high concentration of drug in the administered solution. Finally,
although the effects of peripheral V1b receptors are less well characterized, we decided to
introduce selectivity against this receptor in order to avoid potential peripheral effects on
the hypothalamicꢀpituitaryꢀadrenal axis or on the pancreas.14 The compound should have
high clearance and very short halfꢀlife to mimic the very short duration of contractile
action of 1 on the alveolar myoepithelium sufficient to cause milk letꢀdown while
avoiding accumulation of drug in the circulation after multiple daily administrations.
Thus, we aimed at obtaining compounds with human clearance similar to that of
ACS Paragon Plus Environment

Page 5 of 45
Journal of Medicinal Chemistry
5
1
2
oxytocin, around 20 mL/min/kg.15 Consideration of basic allometric relationship
between body weight and clearance suggested that the desired compounds should have
much higher clearance in rats (see Results section).
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Extensive research has been focused on identifying OTR antagonists suitable for
treatment of premature labor,16ꢀ20 while efforts to design new, clinically useful OTR
agonists have been relatively sparse21, 22 and have resulted in a sole synthetic peptide
analogue of 1, carbetocin (carbaꢀ1ꢀ[Tyr(Me)2]dOT, 2, Fig. 1) approved in several
countries for the treatment of postpartum bleeding.23, 24 The Gly7 modified analogues
proposed by Manning’s group25 are used as rOTRꢀselective pharmacological tools.
Recently, orally available, nonꢀpeptidic antagonists26, 27 and agonists28 for the OTR have
been discovered.
Figure 1. Structures of 1 (OT) and, 2 (carbetocin)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 45
6
1
2
3
4
5
6
7
8
9
The structure of 1 (Fig.1) was elucidated by du Vigneaud and his research team in the
early 1950's,29, 30 and it was shown to be a nonapeptide comprising a 20ꢀmembered ring,
formed by a disulfide bridge between the Cys1 and Cys6 residues, and a Cꢀterminal
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
tripeptide amide. The cyclic part of the hormone molecule is believed to interact with the
second extracellular loop and upper portions of transmembrane domains 3, 4 and 6 of the
OTR whereas the Cꢀterminal part of 1 binds to the first extracellular loop and the Nꢀ
terminal domain of the receptor.31ꢀ33 A similar binding profile has been postulated for 2
but not for antagonist molecules.34
Indeed, the biological activity of 1 has been shown to be very sensitive to changes in
the Cꢀterminal part of the molecule. OT analogues with Gly7 or Sar7 displayed an
improved selectivity profile versus the related vasopressin receptors25, 35, 36 while the
introduction of bulkier and conformationally restricted residues in position 7 resulted in
compounds with different degrees of antagonistic activity.37ꢀ39 The replacement of the
Leu8 residue with basic amino acids17 or the Gly9 residue with a variety of natural or
unnatural amino acids40 led to potent OTR antagonists when combined with DꢀAaa2 and
other modifications.41, 42
Here we report the synthesis, in vitro pharmacological and pharmacokinetic evaluation
of a series of peptidic OT agonists with Nꢀalkylglycine residues (ꢀNR2ꢀCH2ꢀCOꢀ) in
position 7 as the key modification which resulted in the discovery of potent and selective
analogs with high systemic clearance that could be used clinically for lactation support.
All new compounds were prepared as position 1 desamino analogs containing either an
unaltered disulfide bridge or its monocarba modifications. Some compounds were also
prepared with a Phe2 modification known to improve the selectivity profile of 1.43, 44
ACS Paragon Plus Environment

Page 7 of 45
Journal of Medicinal Chemistry
7
1
2
3
4
5
6
7
RESULTS AND DISCUSSION
8
9
The peptides were assembled on resin by standard solid phase peptide synthesis
methods using either Fmoc or hybrid Boc/Fmoc strategies. To prepare compound 5, its
cisHyp7 analogue was assembled and the hydroxyl group was replaced on resin with the
Sꢀacetyl moiety under Mitsunobu reaction conditions.45 After cleavage and cyclization,
the acetyl group was removed by hydrolysis46 and the resulting sulfhydryl function was
methylated in situ. For compounds 9-50, 52-56, 58, 59 and 61-65 the Nꢀalkylglycine
residues in position 7 were introduced by a twoꢀstep procedure used in the chemistry of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
peptoids:47 the resinꢀbound Cꢀterminal LeuꢀGly dipeptide was acylated with bromoacetic
acid and the product treated with an appropriate primary amine. Analogues 51, 57 and 60
were prepared with FmocꢀNꢀalkylglycines synthesized by a modified literature
procedure.48 Peptides 1, 3 and 9-36 were cleaved from the Rink amide resin and cyclized
with iodine in aqueous TFA. Analogues 2, 4, 6-8 and 51-65 were cleaved from the Rink
amide resin and cyclized in DMF with HBTU/DIPEA. Compounds 37-50 were cyclized
on MBHA resin with BOP/DIPEA and cleaved with HF/anisole. All analogues were
purified by preparative HPLC as described in the experimental section.
Compounds 1, 2, and desaminooxytocin (3) were chosen as reference compounds for
this study. To assess potency and selectivity the compounds were tested in in vitro
transcriptional reporter gene assays at the human OT/vasopressin family of receptors as
described in the Experimental Section. Compounds 1-3 were tested as agonists at the
OTR, V1aR, V1bR and V2R receptors, and as antagonists at the V1aR, V1bR and V2R.
Consistent with published data,49, 50 3 exhibited higher potency and improved selectivity
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 45
8
1
2
for the OTR as compared to 1. Compound 251 was less potent at the OTR, but more
3
4
5
6
7
8
selective versus related receptors than 3. Analogues 1-3 did not exhibit antagonism at the
hV1aR (Table 1), hV1bR and hV2R (data not shown). To assess their pharmacokinetic
properties the compounds were injected by the intravenous route to male Spragueꢀ
Dawley rats, plasma concentrations determined by LCꢀMS, and the values of systemic
clearance calculated as described in the Experimental Section. The CL value of 1
obtained, 21±2.2 mL/min/kg, is similar to that reported previously in rats by Lundin.52
The CL of reference compounds 2 and 3 was only slightly higher (Table 1).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Structure and pharmacological profiles of reference compounds 1-3 and initial
leads 4ꢀ8.
In vitro biological activity
Structurea
IC50
Rat
iv clearance
(mL/min/kg)
EC50 receptor (nM)
Selectivity vs.
receptorb
(nM)c
X
R1
R2
R3
hOT hV2
hV1a
hV1b
hV2
hV1a
hV1b
hV1a
1a
2a
3a
4a
S
OH
ꢀ(CH2)3ꢀ
ꢀ(CH2)3ꢀ
ꢀ(CH2)3ꢀ
2.3
7.3
10
240
3
4
100
>10000d
21
22
27
21
CH2
S
OMe
OH
0.70
0.10
0.98
170
3.5
41e
>10000f
180
240
35
58
>14000 >10000d
>10000d
21 e
210
>10000
1800
CH2
OMe
ꢀCH2ꢀCH(OMe)ꢀCH2ꢀg
690
>10000f
>10000
700
>10000 1300
ACS Paragon Plus Environment

Page 9 of 45
Journal of Medicinal Chemistry
9
1
2
3
4
5
6
7
8
9
5a
6a
7a
8a
CH2
CH2
CH2
CH2
OMe
Et
ꢀCH2ꢀCH(SMe)ꢀCH2ꢀg
ꢀCH2ꢀCH(OH)ꢀCH2ꢀi
ꢀCH2ꢀCH(OMe)ꢀCH2ꢀg
0.82
0.06
0.21
0.37
22
670 e
40
>1000h
>1000h
>100
>10000f
1100
810
>1200
>16000
>470
>2700
0.01
>12000 670
38
660
18000
17
25
Et
450
450 e
0.05
0.20
>10000f
>10000
4.3
2100
1200
0.002
0.003
>47000 55
20
OMe
H
H
>1000h
0.24
>27000 1400
45
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.20
0.15
NTj
NTj
NTj
7.5
AVP
72e
>1000h
11
>13
dDAVP
a For compound 1 R0 is NH2 and for compounds 2-8 R0 is H; b EC50 (receptor)/EC50
(hOTR) ratio. If no significant agonism was observed at the highest concentration tested
selectivity is >highest conc. tested/EC50 (hOTR); the selectivity ratios are rounded down
to the nearest values with two significant figures; c IC50 at hV1a receptor stimulated with 2
nM AVP; d no significant antagonism up to 10000 nM, the highest concentration tested; e
partial agonist, efficacy < 70%; f no significant agonism up to 10000 nM, the highest
concentration tested; g the middle carbon atom is of R configuration; h no significant
agonism up to 1000 nM, the highest concentration tested; i the middle carbon atom is of S
j
configuration; not tested.
Using compound 2 as a starting point, our initial effort was focused on obtaining
compounds with high potency at the hOTR, improved in vitro selectivity versus the other
human receptors, and higher CL in rats by introducing a substituent on the pyrrolidine
ring in position 7. Compounds were prepared with Oꢀmethyl tyrosine (4, 5, R1 = OMe) or
4ꢀethylphenylalanine (6, 7, R1 = Et) in position 2. The addition of a substituent in
position 4 of the Pro7 pyrrolidine ring led to analogs 4ꢀ7 with no activity at the hV1R, and
somewhat improved receptor selectivity profile versus the hV2R as compared to 2, but
with only modest if any increase in rat CL.
Both the Gly7 and the Sar7 modifications of OT have been reported to yield analogues
with good potency and selectivity for the OTR.35, 36, 53, 54 Indeed, when we prepared
compound 8, the Gly7 analogue of 2, it displayed an in vitro selectivity profile
comparable to those of compounds 4ꢀ7, but interestingly showed substantially higher CL
in male rats (45 mL/min/kg) than compounds 1-7. Preliminary work in our laboratories
indicated that replacement of Pro7 by various Nꢀalkyl αꢀsubstituted amino acids was not
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 45
10
1
2
3
4
well tolerated (data not shown). Therefore, a new drug discovery program was initiated
5
6
7
8
based on 3, which is more potent at the hOTR than 2, where Pro7 was replaced by glycine
Nꢀalkylated with an array of chemically diverse R2 groups to produce open ring
compounds 9ꢀ36 (Table 2).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Structures and pharmacological profiles of disulfide bridge compounds 9-36.
In vitro biological activity
Rat
iv clearance
(mL/min/kg)
IC50
EC50 receptor (nM)
Selectivity vs. receptora
R2
(nM)b
hOT
hV2
hV1a
hV1b
hV2
hV1a
hV1b
hV1a
0.05
0.01
4.2
9.9
>10000c
>10000c
700
660
84
>200000
14000
66000
>10000d
>10000d
36
59
cyclopropyl
9
990
>1000000
10
0.06
0.16
0.01
0.13
0.01
1.3
35
>10000c
85e
1000e
500
580
11
>160000
530
16000
3100
>10000d
>10000d
>10000d
>10000d
>10000d
2000
63
nꢀbutyl
11
12
13
14
15
16
17
18
19
1.8
77
NTf
cyclobutyl
pentyl
19e
370
7700
84
1900
37000
7500
94
11
>10000c
12e
980e
>76000
1200
61
cyclopentyl
hexyl
100
57
470
10000
43
47000
>7600
93000
>25000
2300
46
>10000c
>10000c
>10000c
>10000c
>10000c
2800e
>10000c
260
>7600
>330000
>25000
>90000
NTf
cyclohexyl
0.03
0.39
0.11
80
2600
250
300
1800
47
100
34
3400
28
>10000d
83
benzyl
0.12
68
>10000
290
560
>83000
2400
1900
81
20
ACS Paragon Plus Environment

Page 11 of 45
Journal of Medicinal Chemistry
11
1
2
3
4
5
6
7
8
0.01
0.15
0.01
30
84
82
>10000c
>10000c
>10000c
110
200
140
3000
560
>1000000
>66000
11000
1300
>10000d
1400
91
67
24
21
22
23
8200
>1000000
14000
880
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
0.06
78
>10000c
440
1300
>160000
7300
970
73
24
1.3
180
170
380
>10000c
>10000c
>10000c
1800e
1300e
2600
130
130
820
>7600
>7600
>21000
1300
1000
5600
2200
1300
3200
NTf
NTf
43
25
26
1.3
0.46
27
28
29
30
31
0.12
0.12
0.02
66
>10000c
>10000c
>10000c
590
770
720
550
>83000
>83000
>500000
4900
6400
36000
270
730
350
94
63
48
180
47
1500
2300
0.01
0.11
0.02
0.07
57
>10000c
>10000c
>10000c
>10000c
240
5700
560
>1000000
>90000
24000
8400
1200
390
71
67
65
39
62
930e
32
33
34g
35
36
86
600
4300
3200
>500000
>140000
30000
280
230
>10000c
>140000 1500
19
10
1400e
>10000c
>10000c
>10000c
>10000c
73
>520
>520
1300
NTf
NTf
1400
140
>1000
>1000
>10000d
a EC50 (receptor)/EC50 (hOTR) ratio. If no significant agonism was observed at the
highest concentration tested selectivity is >highest conc. tested/EC50 (hOTR); the
selectivity ratios are rounded down to the nearest values with two significant figures; b
IC50 at hV1a receptor stimulated with 2 nM AVP; c No significant agonism at the highest
concentration tested – 10000 nM; d No significant antagonism at the highest
concentration tested ꢀ 10000 nM; e partial agonist, efficacy < 70%; f Not tested; g
Compound 34 is the faster running diastereoisomer separated by preparative HPLC.
In vitro profile of position 7 modified desaminooxytocin analogues 9-36 (Table 2).
Replacement of Pro7 in 3 by Nꢀalkylated glycines produced remarkable and very different
changes in potency in in vitro functional assays at the four receptors studied. The change
resulted in complete elimination of activity at the hV1aR for all compounds in Table 2
with notable exceptions when R2 was cyclobutyl (12), nꢀpentyl (13), and nꢀhexyl (15),
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 45
12
1
2
3
4
5
6
which exhibited potencies similar to the reference 3. Compound 3 is already fairly
selective for the hOTR over hV1bR; replacement of Pro7 in 3 by Nꢀalkylated glycines
7
8
9
yielded compounds with similar or lower potency at the hV1bR and consequently similar
or better selectivity. The R2 groups producing the largest loss in potency at the hV1bR
were the bulky cyclohexyl in 16, and some but not all of the polar groups, for example in
compounds 18, 25 - 27, and 34 – 36.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 3 is not very selective for hOTR (EC50 0.10 nM), over hV2R (EC50 3.5 nM),
and replacement of Pro7 by Nꢀalkylated glycines resulted in modest to good
improvements in selectivity (Table 2). Considerable reductions in potency at hV2R
occurred only with the pyridylmethyl R2 substituents in 25 - 27, the charged R2 groups in
35 and 36, and with the 2ꢀtetrahydrofurylmethyl group in 34. Potency at the hV2R was
maintained as in 3, only with the small cycloalkyl substituents in 9, 12, and 14.
Replacement of Pro7 in 3 by Nꢀalkylated glycines resulted in compounds with similar
potency at the hOTR (Table 2) for most R2 groups tested, with some interesting
exceptions. The small linear alkyl R2 groups nꢀbutyl (11), nꢀpentyl (13), and nꢀhexyl (15)
increased potency at the hOTR whereas the corresponding cyclic substituents (12, 14, 16)
did not. However, the small cyclic cyclopropyl (9) and cyclopropylmethyl (10) R2 groups
were well tolerated and actually also improved potency. The insertion of an oxygen atom
into a straight alkyl substituent resulted in analogue 17 (R2 = 3ꢀoxabutyl) with higher
potency at the hOTR as compared to 11 (R2 = nꢀbutyl). The location of the oxygen atom
in R2 seemed to be important as the 3ꢀhydroxypropyl compound in 18 was about 10ꢀfold
less potent than 17.
ACS Paragon Plus Environment

Page 13 of 45
Journal of Medicinal Chemistry
13
1
2
3
4
5
6
7
8
9
Nꢀbenzyl compounds 19-24 were generally equipotent to 3 at the hOTR, but the
4ꢀfluorobenzyl (23, EC50 0.01 nM) and 3ꢀmethylbenzyl (21, EC50 0.01 nM) analogues
were the most potent in this series. The pyridylmethyl analogues of 25-27 were less
potent agonists at the hOTR than the benzyl analogue of 3 (19), paralleling the effect
observed at the hV2R. Increasing the distance from the phenyl group to the glycine
nitrogen with an additional methylene group (R2 = 2ꢀphenylethyl) in 28 resulted in no
loss of potency at the hOTR relative to 19 (R2 = Bzl). Surprisingly, the compounds in
which the pyridyl ring was also separated from the peptide backbone by two carbon
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
atoms (29, 30) were markedly more potent than their picolyl counterparts 25-27, with the
4ꢀpyridylethyl analogue 30 being substantially more potent (EC50 = 0.02 nM) than 3. The
introduction of the heterocycloalkyl R2 groups 2ꢀthiophenylmethyl (31) and
2ꢀthiophenylethyl (33) yielded potent hOTR agonists whereas the oxygen containing
analogues with R2 2ꢀfurylmethyl (32) and 2ꢀtetrahydrofurylmethyl (34) had similar
potency to 3. The presence of a positive charge/tertiary amino group in the nonꢀaromatic
heterocyclic R2 substituents had a deteriorating effect on potency at the hOTR of
compounds 35 and 36, as seen for the hV2R.
The above observations on the potency of the various analogues at the hOTR can be
interpreted as resulting from specific interactions between the R2 group and a lipophilic
binding site in the receptor. It is also possible that an additional conformational factor
might contribute to the potency of the analogues. The conformation of the Cys6ꢀPro7
bond in 1 in water at pH 3 is 90% trans,55 and it has been suggested that agonists at the
OTR require a trans conformation around this bond with potency increasing
proportionally to their trans content.56 In the sarcosyl7 derivative of OT the proportion of
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 45
14
1
2
Cys6ꢀPro7 trans conformer is somewhat decreased to about 75%.57 Work in the peptoid
field showed that in acetyl Nꢀalkylglycine model compounds the proportion of cis
3
4
5
6
7
8
9
conformer is not very different when the Nꢀalkyl group is methyl, ethyl or benzyl, but it
increases when it is branched in the αꢀcarbon to the nitrogen.58 Preliminary examination
of the 1HꢀNMR spectra of compounds 10, 16, 23 and 57 suggests that all the analogues
contain similar proportions of cis and trans isomers and that this is an unlikely cause of
the differences in potency obtained (data not shown). Overall, the strategy of replacing
Pro7 by Nꢀalkylated glycines resulted in considerable progress towards obtaining
selective hOTR agonists. Activity at the hV1aR was eliminated in most analogues, and
potency at the hV1bR and hV2R decreased, which when coupled to the substantial
increase in potency in some analogues resulted in compounds with considerably
increased selectivity for the hOTR.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The R2 modifications that resulted in the most potent and/or most selective (especially
versus the hV2R) dOT analogues were 3ꢀoxabutyl, 3ꢀhydroxypropyl, nꢀbutyl, nꢀpentyl,
benzyl, 3ꢀmethylbenzyl, 4ꢀfluorobenzyl, phenethyl, and 2ꢀthienylmethyl. A final series
of compounds 37-65 (Table 3) was prepared where these R2 groups were combined with
one of the two possible replacements of sulfur by a methylene group (“carba” alterations)
in the disulfide bridge, and in some cases, with Tyr2 replaced by Phe2 in an attempt to
further improve selectivity versus the hV2R.59
ACS Paragon Plus Environment

Page 15 of 45
Journal of Medicinal Chemistry
15
1
2
3
4
Table 3. Structures and pharmacological profiles of monocarba analogues 37-65.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In vitro biological activity
b
Rat
iv clearance
(mL/min/kg)
Structure
IC50
EC50 receptor (nM)
Selectivity vs. receptora
(nM)
X
Y
R1
R2
hOT hV2
hV1a
hV1b
hV2
hV1a
hV1b
hV1a
CH2
S
S
OH
OH
H
0.11
0.16
0.96
0.85
0.01
0.13e
0.86
70
>10000c
>10000c
>10000c
>10000c
>10000c
>10000c
>10000c
3200
630
>90000
>62000
>10000
>11000
>1000000
>76000
>11000
29000
>10000d
990
20
22
32
21
18
13
22
37
38
39
40
41
42
CH2
S
52
830
320
5100
CH2
S
1100
1300
500e
150
2600
>10000c
>10000c
>10000c
>10000c
>10000c
1100
1500
50000
1100
3000
>10000
>11000
>1000000
>76000
>11000
>10000d
2500
CH2
S
H
CH2
S
OH
OH
H
1400
CH2
CH2
1800
S
>10000d
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
CH2
S
S
OH
OH
H
nꢀbutyl
nꢀbutyl
nꢀbutyl
nꢀbutyl
nꢀpentyl
nꢀpentyl
nꢀpentyl
benzyl
0.12
0.14
0.23
0.09
0.11
0.25
0.07
0.04
0.05
0.40
0.08
0.02
0.56
0.08
0.04
73
>10000c
>10000c
>10000c
>10000c
86e
820
600
> 83000
>71000
>43000
>110000
780
6800
>10000d
>10000d
300
36
NTf
NTf
61
45
95
83
51
68
43
70
61
63
65
58
CH2
S
18
450
120
3200
CH2
S
2000e
1200e
110
2100e
1000e
140
36
>10000c
>10000c
500
8600
13000
1000
8400
14000
3500
720
>43000
>110000
4500
CH2
S
H
740
CH2
S
OH
OH
H
>10000d
>10000d
>10000d
1400
CH2
CH2
S
>10000c
>10000c
>10000c
>10000c
>10000c
>10000c
71e
3600e
1000
100
>40000
>140000
>250000
>200000
>25000
>120000
3500
14000
14000
2500
S
CH2
S
OH
OH
H
CH2
CH2
S
benzyl
100
2000
>10000d
S
benzyl
720
210
55
440
1800
2600
2700
2800
4100
3500
1100
1600
CH2
S
OH
OH
H
84
1000
>10000d
>10000d
480
CH2
CH2
S
87
4300
S
1600e
330
140
>10000c
>10000c
>10000c
620
>17000
>120000
>250000
1100
CH2
S
OH
OH
180
2200
1200
CH2
91e
2200
>10000d
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 45
16
1
2
3
4
5
6
7
8
9
S
CH2
S
H
0.23
0.04
0.05
0.78
0.09
0.14
0.30
1400
160
100
3000e
250
30
>10000c
>10000c
>10000c
>10000c
>10000c
>10000c
>10000c
750e
450
6000
4000
2000
3800
2700
210
>43000
>250000
>200000
>12000
>110000
>71000
>33000
3200
11000
3200
1900
4600
1300
2800
1300
68
75
52
62
48
53
49
59
60
61
62
63
64
65
CH2
S
OH
OH
H
490
CH2
CH2
S
160
640
S
1500 e
330
CH2
S
OH
OH
H
420
2100
380
CH2
CH2
190
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
S
760
850
2500
1200
a EC50 (receptor)/EC50 (hOTR) ratio. If no significant agonism was observed at the
highest concentration tested selectivity is >highest conc. tested/EC50 (hOTR); the
selectivity ratios are rounded down to the nearest values with two significant figures; b
IC50 at hV1a receptor stimulated with 2 nM AVP; c No significant agonism at the highest
concentration tested ꢀ 10000 nM; d No significant antagonism at the highest concentration
tested ꢀ 10000 nM; e partial agonist, efficacy < 70%; f Not tested.
In vitro profile of position 7 modified monocarba analogues 37-65 (Table 3).
In the Tyr2 series (R1 = OH), the potency of the carba analogues at the hOTR generally
decreased slightly relative to their disulfide bridge counterparts, but there were
substantial gains in selectivity versus the hV2R, especially in the carbaꢀ1 (X = CH2, Y =
S) analogues. The most striking improvement over a disulfide analogue was seen for the
3ꢀhydroxypropyl compound 41 that showed remarkably higher potency as hOTR agonist
(EC50 0.01 nM) and notably lower potency as hV2R agonist (EC50 500 nM) than
compound 18. The benzyl analogues 51 and 52 and the phenethyl analogues 60 and 61
showed only a slightly better in vitro pharmacological profile than their disulfide
counterparts 19 and 28, respectively, while the 2ꢀthienylmethyl carba analogues 63 and
64 were less potent and selective than 31. As expected in the Phe2 series, the potency at
the hOTR was markedly reduced as compared to the corresponding Tyr2 compounds with
the noteworthy exception of 47 and 50 where R2 was a straight hydrocarbon chain. The
EC50 values at the hV2R for all Phe2 compounds were in the micromolar range resulting
in high hV2 selectivities (>1000) with peptide 47 showing the most remarkable in vitro
ACS Paragon Plus Environment

Page 17 of 45
Journal of Medicinal Chemistry
17
1
2
pharmacological profile among Phe2 analogues. None of the analogues in Table 3 were
active at the hV1aR (EC50 > 10000 nM) except for the combination of carbaꢀ6 and 3ꢀ
methylbenzyl R2 in 55 (EC50 71 nM) and of carbaꢀ1 and nꢀpentyl in 48 (EC50 86 nM),
illustrating the exquisite sensitivity of this receptor to subtle changes in structure. Overall,
compounds 9ꢀ65 showed superior in vitro pharmacological profile as compared to 1.
Doseꢀresponse curves for compound 57 and 1 in reporter gene assays at the hOTR, hV2R,
hV1aR and hV1bR are shown in Figure 2.
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. Representative in vitro doseꢀresponse curves for functional activation of the
hOTR and related receptors by compound 57 (left) and parent hormone oxytocin (1,
right). Data are expressed as percentage of the maximum response (%MPE) to the
internal control agonist used in each reporter gene assay (carbetocin for hOTR, AVP for
hV1aR and hV1bR, dDAVP for hV2R).
Compounds 9-65 as vasopressin receptors antagonists.
With the exception of compounds 12, 13, 15, 48 and 55, none of the analogues in
Tables 2 and 3 exhibited agonism at the hV1aR up to the highest concentration tested
(10,000 nM). To rule out the possibility that the compounds might still have affinity for
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 45
18
1
2
3
4
5
6
7
8
9
the hV1aR they were tested as antagonists against 2 nM concentration of AVP. Although
none of the compounds was found to be a potent hV1aR antagonist, some peptides where
an aromatic or heteroaromatic ring was linked to the peptide backbone by an ethyl spacer
(28, 30, 33, 60, and 62) or a methylene spacer (56 and 64) displayed IC50 values below
500 nM. A subset of compounds 9-65 was also tested as hV2R and hV1bR antagonists and
none of the compounds showed any significant antagonistic activity at either receptor
(data not shown).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Structure-pharmacokinetics relationship.
In addition to high potency and selectivity for the hOTR, our objective was to identify
peptides which would likely have short duration of action in humans. The halfꢀlife of
endogenous oxytocin, 1, released during nursing appears to be very short60 and obviously
adequate for its purpose. Therefore, the clearance of 1 in nonꢀpregnant women, about 20
mL/min/kg15, 61, was selected as the target value for the desired compound in humans.
Assuming that the CL of (at least some) the analogues prepared follows classic allometric
scaling (whole body CL = a•BW0.75 where a is allometric coefficient and BW is body
weight),62 the target CL in rats falls in the range of 60ꢀ80 mL/min/kg. However, the CL
of 1 itself does not scale allometrically as its value is similar in various species (rat,
Cynomolgus monkey, humans, data not shown). Indeed, not only 1, but reference
compounds 2 and 3 had CL value in male rats in the range of 21ꢀ27 mL/min/kg (Table 1),
similar to 1 in humans. The CL values in female and postpartum rats (data not shown)
were not different from male rats, and therefore these were used throughout the program.
The CL values in male rats for compounds 9 – 65 are shown in Tables 2 and 3. Seventeen
ACS Paragon Plus Environment

Page 19 of 45
Journal of Medicinal Chemistry
19
1
2
3
4
5
6
analogues had rat CL in the desired range of 60 – 80 mL/min/kg and of those, compounds
24, 31, 33, 47, 55, 57, 59, 60 and 62 showed the best combined in vitro/PK
7
8
9
pharmacological profiles and were deemed potential drug candidates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3 shows the correlation between total male rat CL and the capacity factor k'
determined by reversed phase HPLC on a C18 column using a TFA acetonitrileꢀwater
based gradient, a surrogate of compound lipophilicity (see Experimental for details). For
all four compound series, disulfide bridged (Table 2), carbaꢀ1, carbaꢀ6, and [Phe2, carbaꢀ
6] analogues (Table 3), there is a reasonably good direct correlation between CL and
lipophilicity. In the disulfide bridged series, the compounds with the highest k' values
containing hydrophobic R2 groups such as nꢀpentyl (13), benzyl (19), methylbenzyl (20,
21) and 2ꢀphenylethyl (28) displayed the highest CL values, with the maximum value of
94 mL/min/kg for 13 and 28. At the other extreme, compound 18 with the polar
hydroxypropyl R2 group had the lowest retention time (k' = 7.60) and the lowest CL of 28
mL/min/kg in the series. Compounds 27, 29 and 30 followed a different trend line, with
higher CL and low k' values. Unlike all other compounds in Figure 3, which are neutral
peptides, these contain a pyridyl ring that is protonated in the TFA HPLC buffer,
resulting in much shorter retention times. Under physiological conditions the compounds
are not expected to be protonated and their CL values are indeed similar, but somewhat
lower, than their nonꢀpyridyl counterparts 11 and 28. Two compounds in the disulfide
bridged series are outside the general correlation, the Nꢀhexyl analogue 15 and the
fluorobenzyl analogue 23, which have lower CL than expected from their k' value.
For otherwise identical sequences, the CL values increased in the order: CL (carbaꢀ6
analogues) ≤ CL (carbaꢀ1 analogues) < CL (disulfide bridge analogues). The CL values
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 45
20
1
2
for compounds where R2 is 2ꢀphenylethyl (61, 60 and 28), 3ꢀmethylbenzyl (55, 54, 21)
and 3ꢀhydroxypropyl (42, 41, 18) followed the trend remarkably well. Peptide 42, with
the lowest k' value of all compounds tested, had CL 13 mL/min/kg, which is lower than
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
that of the reference compounds 1 - 3. On the other hand, the disulfide analogue 23 (R2 =
4ꢀfluorobenzyl), displayed a considerably lower CL (24 mL/min/kg) than its carba
counterparts 57 and 58, which otherwise follow the trend within their respective series.
The relationship of pharmacokinetic parameters such as CL and volume of distribution
with lipophilicity and/or charge is relatively well described in the literature for small
molecules.63, 64 There appears to be a positive correlation between the lipophilicity of
compounds expressed as octanolꢀwater distribution coefficient logD and the extraꢀrenal
clearance of free, plasma proteinꢀunbound fraction; and there is an inverse relationship
between the lipophilicity and the renal component of the freeꢀfraction clearance.59, 60 On
the other hand, for compounds with high plasma protein binding, CL negatively
correlates with hydrophobicity.65 These correlations reflect passive partitioning of the
molecules into hepatocytes where they undergo metabolism by cythochrome P450
enzymes as the main mechanism of disposition. Such systematic correlations between
pharmacokinetic parameters and lipophilicity appear not to have been previously
performed for peptides. Peptides are not expected to diffuse passively into cells, and
typically are not metabolized by cytochromes. The fairly good correlation between rat
clearance and capacity factor shown in Figure 3 was actually surprising. The disposition
of peptides is generally considered to occur by passive renal filtration, proteolysis by
enzymes in plasma and in tissues, and by carrier mediated uptake by the liver and other
tissues. We expected to find that the SAR for the CL of the analogues would be complex
ACS Paragon Plus Environment

Page 21 of 45
Journal of Medicinal Chemistry
21
1
2
3
4
5
6
and exhibit features reflecting recognition by proteases and/or transporters. For small
molecules, very good correlations have been found between LogP and the logarithm of
7
8
9
the reversed phase HPLC capacity factor k', as both reflect the change in free energy for
partition between an aqueous and a nonꢀpolar phase.66
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Effect of lipophilicity expressed as k' (k' = HPLC capacity factor, see
experimental section) on systemic clearance (CL) in male rats for different series of OT
analogues: disufide bridge analogues (diamonds), carbaꢀ1 analogues (triangles), carbaꢀ6
analogues (squares) and Phe2 carbaꢀ6 analogues (circles). Compound numbers for
analogues mentioned in the PK discussion are shown in the graph.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 45
22
1
2
3
4
5
6
7
8
9
Figure 3 shows an approximately linear correlation between CL and k', and a priori
there is no reason to assume that a more complex function of k' would provide a better
correlation. Since at this time the main mechanism of disposition of the peptides is
unknown, any rationale for the empirical relationship found would be speculative.
Binding to plasma proteins can probably be excluded as this would result in an inverse
correlation between CL and lipophilicity by any mechanism that eliminates or
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
metabolizes the free peptide fraction. A full pharmacokinetic characterization of selected
compounds from this collection in several animal species will be reported elsewhere.
CONCLUSION
We have designed, synthesized and pharmacologically characterized a series of hOTR
agonists modified in position 2, 7 and at the disulfide bridge. Replacement of Pro7 by Nꢀ
alkyl glycines resulted in potent hOTR agonists devoid of hV1aR activity and with
improved selectivity against the hV1bR and hV2R, and high rat clearance. Replacement of
one sulfur atom in the disulfide bridge further improved selectivity versus the hV2R. A
correlation between rat clearance and lipophilicity was found and exploited to obtain
compounds with high rat clearance, which would be expected to have short halfꢀlife in
humans suitable for multiple daily administrations. Several compounds such as 31 ([2ꢀ
ThiMeGly7]dOT, 47 (carbaꢀ6ꢀ[Phe2,BuGly7]dOT), 55 (carbaꢀ6ꢀ[3ꢀMeBzlGly7]dOT) and
57 (carbaꢀ1ꢀ[4ꢀFBzlGly7]dOT), showing the desired pharmacological profile (high
potency as OTR agonists, high selectivity for the hOTR particularly versus the hV2R, no
antagonist activity at the vasopressin receptors and high CL in rats) have been identified.
Based on the in vitro and pharmacokinetic data reported in this paper and other
ACS Paragon Plus Environment

Page 23 of 45
Journal of Medicinal Chemistry
23
1
2
3
4
considerations, compound 57, FE 202767, has been selected as a clinical candidate for
5
6
the lactation support indication.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
Synthesis
General. Amino acid derivatives and resins were purchased from Novabiochem,
Bachem, Peptide International and RSP Amino Acids. FmocꢀCys((CH2)3COOtBu)ꢀOH
(the carbaꢀ1 derivative)67 and FmocꢀHcy((CH2)2COOtBu)ꢀOH (the carbaꢀ6 derivative)68
were synthesized according to published procedures. FmocꢀNꢀalkylglycines: Fmocꢀ
BzlGlyꢀOH, Fmocꢀ4FBzlGlyꢀOH and FmocꢀPhEtGlyꢀOH were prepared from ethyl
bromoacetate by an improved (i.e., without chromatography) literature method.48 Fmocꢀ
Hyp(Me)ꢀOH was synthesized from BocꢀHypꢀOtBu by a three step procedure consisting
of Oꢀmethylation with MeI in DME in the presence of NaH, removal of protecting groups
with neat TFA and the Fmoc group introduction with FmocꢀOSu.
Other chemicals and solvents were purchased from SigmaꢀAldrich and VWR.
Analytical HPLC was performed on a Waters 600 Liquid Chromatograph with
Empower 2 software package using a Vydac C18, 5 ꢁm, 4.6 x 250 mm column at a flow
rate of 2 mL/min. Preparative HPLC was performed on a Waters 2000 Liquid
Chromatograph using a 15 ꢁm PrepPak 47 x 300 mm cartridge at a flow rate of 100
mL/min. Final purity of analogues was assessed on a 1100 Agilent Liquid
Chromatograph using the following analytical method: column – Vydac C18, 5 µm, 2.1 x
250 mm; column temperature – 40°C; flow rate – 0.3 mL/min; solvent A – 0.01%
aqueous TFA; solvent B – 70% CH3CN, 0.01% TFA; gradient – 0ꢀ20% B in 1 min., then
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 45
24
1
2
3
4
5
6
7
8
9
20ꢀ40% B in 20 min., then held at 100% B for 5 min.; when necessary the first two
segments of the gradient were adjusted for compound lipophilicity; UV detection at
214 nm. The purity of all analogues exceeded 95% (see Table 4 in Supporting
Information). HPLC capacity factors (k’) were calculated using the following equation:
k’ = (t – t0)/t0 where t is retention time of a test compound and t0 is retention time of
unretained species. For capacity factor calculations the retention times were determined
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
on a 1200rr Agilent Liquid Chromatograph using an Agilent Zorbax SBꢀC18, 1.8 ꢁm, 4.6
x 50 mm column at a flow rate of 1.5 mL/min. Solvent A was 0.05% aqueous TFA and
solvent B was 90% CH3CN, 0.045% TFA. Mass spectra were recorded on a Finnigan
MAT spectrometer.
Peptide synthesis.
Disulfide bridge analogues 9-36. All compounds were assembled on Rink amide resin
by Fmoc strategy. The following derivatives were employed: FmocꢀGlyꢀOH, FmocꢀLeuꢀ
OH, FmocꢀCys(Trt)ꢀOH, FmocꢀAsn(Trt)ꢀOH, FmocꢀGln(Trt)ꢀOH, FmocꢀIleꢀOH and
FmocꢀTyr(tBu)ꢀOH and Mpa(Trt)ꢀOH. The Cys derivative was coupled using DIC in
DCM with a 4ꢀfold excess and all other couplings were mediated by DIC/HOBt with a 3ꢀ
fold excess of reagents. The Nꢀalkylglycine residues in position 7 were introduced by a
twoꢀstep procedure comprising the acylation of the resinꢀbound dipeptide with
bromoacetic acid/DIC and the displacement of the bromine atom with a suitable primary
amine.47 The linear peptides were cleaved with the TFA/TIS/H2O 95/2.5/2.5 (v/v/v)
cocktail and the ring formation was achieved by an iodine oxidation of crude analogues
dissolved in 10% aqueous TFA containing up to 20% acetonitrile. The crude products
ACS Paragon Plus Environment

Page 25 of 45
Journal of Medicinal Chemistry
25
1
2
3
4
were purified by preparative HPLC in acetic acid buffer. If necessary, an additional
5
6
7
8
purification using a triethylammonium phosphate buffer was performed prior to the acetic
acid run. The fractions with purity exceeding 97% were pooled and lyophilized.
Compound 34 was synthesized using racemic 2ꢀaminomethyl furane and was separated
by preparative HPLC as a faster running diastereoisomer. No attempt to determine the
absolute configuration of the C1 carbon in the furane ring was undertaken.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Carba analogues 37-50. All compounds were synthesized by a hybrid Boc/Fmoc
strategy on MBHA resin. The following derivatives were employed: BocꢀGlyꢀOH, Bocꢀ
LeuꢀOH, FmocꢀCys((CH2)3COOtBu)ꢀOH (for carbaꢀ1 analogues) or Fmocꢀ
Hcy((CH2)2COOtBu)ꢀOH (for carbaꢀ6 analogues), FmocꢀAsn(Trt)ꢀOH, FmocꢀGln(Trt)ꢀ
OH, FmocꢀIleꢀOH and BocꢀTyr(tBu)ꢀOH or BocꢀPheꢀOH. The carba derivatives were
coupled using DIC in DCM with a 4ꢀfold excess and all other couplings were mediated
by DIC/HOBt with a 3ꢀfold excess of reagents. The Nꢀalkylglycine residues in position 7
were introduced by the twoꢀstep procedure as described above for compounds 9-36. The
resin bound peptides were treated with the TFA/TIS/H2O 95/2.5/2.5 (v/v/v) cocktail to
remove the acid sensitive protecting groups and were subsequently cyclized with
BOP/DIPEA in DMF. The crude cyclic peptides were cleaved from the resin with
HF/anisole 10/1 (v/v) and purified by preparative HPLC as described above.
Carba-1 analogues 51, 57 and 60. The peptides were assembled by Fmoc chemistry
on Rink amide resin. The following derivatives were employed: FmocꢀGlyꢀOH, Fmocꢀ
LeuꢀOH, FmocꢀBzlGlyꢀOH (for compound 51) or Fmocꢀ4FBzlGlyꢀOH (for compound
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 45
26
1
2
3
4
57) or FmocꢀPhEtGlyꢀOH (for compound 60), FmocꢀCys((CH2)3COOtBu)ꢀOH, Fmocꢀ
5
6
7
8
Asn(Trt)ꢀOH, FmocꢀGln(Trt)ꢀOH, FmocꢀIleꢀOH and BocꢀTyr(tBu)ꢀOH. The couplings
were performed as described above. The linear peptides were cleaved with the
TFA/TIS/H2O 95/2.5/2.5 (v/v/v) cocktail and cyclized with HBTU/DIPEA in DMF. The
crude products were purified as described above.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compounds 52-56, 58, 59 and 61-65. These compounds were prepared by methods
comprising various steps used in the synthesis of analogues 9-51, 57 and 60.
Analogues 2, 4 and 6-8. The peptides were assembled by Fmoc chemistry on Rink
amide resin. The following derivatives were employed: FmocꢀGlyꢀOH, FmocꢀLeuꢀOH,
FmocꢀProꢀOH (for compound 3) or FmocꢀHyp(Me)ꢀOH (for compounds 4, 6 and 7) or
FmocꢀGlyꢀOH (for compound 8), FmocꢀCys((CH2)3COOtBu)ꢀOH, FmocꢀAsn(Trt)ꢀOH,
FmocꢀGln(Trt)ꢀOH, FmocꢀIleꢀOH and BocꢀTyr(Me)ꢀOH (for compounds 2, 4 and 8) or
FmocꢀPhe(Et)ꢀOH (for compounds 6 and 7). The couplings were mediated by DIC/HOBt
with 3ꢀfold excess of reagents. The cleavage, cyclization and purification were performed
as described for analogues 51, 57 and 60.
Analogue 5. The peptide was assembled by Fmoc chemistry on Rink amide resin using
the following derivatives: FmocꢀGlyꢀOH, FmocꢀLeuꢀOH, FmocꢀcisHypꢀOH, Fmocꢀ
Cys((CH2)3COOtBu)ꢀOH, FmocꢀAsn(Trt)ꢀOH, FmocꢀGln(Trt)ꢀOH, FmocꢀIleꢀOH and
BocꢀTyr(Me)ꢀOH. The couplings were mediated by DIC/HOBt with 3ꢀfold excess of
reagents. The resin bound peptide was converted to the Pro(SAc)7 analogue upon
ACS Paragon Plus Environment

Page 27 of 45
Journal of Medicinal Chemistry
27
1
2
treatment with CH3COSH/TPP/DIAD45 in DME and cleaved with the TFA/TIS/H2O
3
4
5
6
7
8
95/2.5/2.5 (v/v/v) cocktail. The crude linear peptide was cyclized with PyBOP/DIPEA in
DMF and purified by HPLC. The Sꢀacetyl group was converted to the Sꢀmethyl group by
hydrolysis in MeOH with an equivalent of 1M NaOH46 and in situ methylation of the
resulting sulfhydryl group with methyl iodide. The peptide was then purified as described
above.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Biological methods
In vitro receptor assays. Agonist activity of compounds at the human oxytocin OTR
was determined in a transcriptional reporter gene assays (RGA) using CHOꢀK1 cells
stably expressing the human OTR transfected with a reporter DNA construct containing
intracellular calcium responsive promoter elements regulating expression of the firefly
luciferase. Two days following transfection, cells were treated with appropriate doses of
peptides, incubated at 37°C for 5 hours, lysed in the presence of luciferin and the total
luminescence was measured.
To determine receptor selectivity, compounds were also tested in luciferaseꢀbased
transcriptional reporter assays in HEK293 cells expressing the human V2R, human V1aR
or human V1bR.69, 70 AVP was used as an internal control for the vasopressin V1aR and
V1bR assays, dDAVP was used as an internal control for the V2R assays and carbetocin,
was used as an internal control for the OTR assays. The assays were standardized by
including these controls in every experiment. Doseꢀresponse curves were analyzed using
a oneꢀsite, four parameter model from Xlfit (IDBS) to estimate EC50 and efficacy values.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 45
28
1
2
3
4
5
6
7
8
9
Agonist potency determined from multiple independent experiments is reported as the
geometric mean EC50, first calculating pEC50 values (pEC50 = ꢀlog(EC50 in M) for each
individual value, and then determining the arithmetic mean, and standard error of the
mean (SEM) of pEC50 values. Average pEC50 values are transformed back to the normal
numeric scale to report as EC50. Compounds in these assays were diluted with
logarithmic spacing between concentrations, requiring the use of the geometric mean for
potency values.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Selectivity values are given as ratios of the EC50 values at the receptor of interest to
the corresponding EC50 values at the OTR.
Antagonist activity of compounds at human V1aR, human V2R and human V1bR were
determined in luciferaseꢀbased transcriptional reporter assays in cells expressing the
human V1aR, human V1bR and human V2R. Antagonism was tested against receptors
activated with reference agonists at concentrations which lie between the respective EC50
and EC90 (2 nM AVP at human V1aR, 50 nM AVP at human V1bR, 1 nM dDAVP at
human V2R). Interassay reproducibility was controlled using known receptor antagonists
(HOꢀPhaaꢀDꢀTyr(Me)ꢀPheꢀGlnꢀAsnꢀArgꢀProꢀArgꢀNH271 for human V1aR and human
V1bR, and AdAcꢀDꢀTyr(Et)ꢀPheꢀValꢀAsnꢀAbuꢀProꢀArgꢀArgꢀNH272 for human V2R) as
internal controls. Compounds were tested in at least three independent experiments.
Doseꢀresponse curves were analyzed using a oneꢀsite, four parameter model from Xlfit
(IDBS) to estimate IC50 and efficacy values. pIC50 values are calculated from IC50 (pIC50
= ꢀlog (IC50 in M) for each individual value, then determining the arithmetic mean and
SEM of pIC50. Average pIC50 values are then transformed into the normal numeric scale
to report the geometric mean IC50. Compounds in these assays were diluted with
ACS Paragon Plus Environment

Page 29 of 45
Journal of Medicinal Chemistry
29
1
2
3
4
5
6
7
8
9
logarithmic spacing between concentrations, requiring the use of the geometric mean for
potency values. Antagonist potency is presented as geometric means of IC50 values in
nanomol/L (nM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Determination of clearance in rat
Animals. Singleꢀdose pharmacokinetic profiles of OT analogues were investigated
following i.v. bolus adminstration in male SpragueꢀDawley rats (230 – 300 g). Animals
were obtained from Harlan Laboratories Inc. (Indianapolis USA) with chronic jugular
vein and carotid artery catheters inserted surgically. The rats were given free access to
food (18% protein Rodent Diet, Harlan Teklad, Madison USA) and water. They were
housed in a conventional animal facility in either static caging (Alternative Design
Manufacturing & Supply Inc., Siloam Springs USA) or individually ventilated caging
(LabProducts Inc., Seaford USA) with appropriate air flow under controlled
environmental conditions (20ꢀ22ºC, 12hr light/dark cycle).
All animal procedures were approved by the Ferring Research Institute Institutional
Animal Care and Use Committee and were in accordance with the Guide for the Care
and Use of Laboratory Animals published by the National Research Council.
Dosing and sampling. Analogues were dissolved in water for injection or sterile
saline. Each rat was given a single dose of 0.1 to 0.2 mg/kg of test compounds in cassette
mode through the jugular vein.73 Cassettes were constituted of 1 to 8 analogues with each
compound at a concentration of 0.1 to 0.2 mg/mL. Blood samples (250 ꢁL) were
collected from the carotid artery catheter into Microtainer EDTA Tubes (BD Diagnostics,
ACS Paragon Plus Environment