4-[4-[(2-hydroxybenzoyl)amino]phenyl]butyric acid as a novel oral delivery agent for recombinant human growth hormone

Article abstract of DOI:10.1021/jm960038f

A series of N-acetylated, non-α, aromatic amino acids was prepared and shown to promote the absorption of recombinant human growth hormone (rhGH) from the gastrointestinal tract. Seventy compounds in this family were tested in vivo in rats. Of the compoun

Full text of DOI:10.1021/jm960038f

J . Med. Chem. 1996, 39, 2571-2578
2571
4-[4-[(2-Hyd r oxyben zoyl)a m in o]p h en yl]bu tyr ic Acid a s a Novel Or a l Deliver y
Agen t for Recom bin a n t Hu m a n Gr ow th Hor m on e
Andrea Leone-Bay,* Koc-Kan Ho, Rajesh Agarwal, Robert A. Baughman, Kiran Chaudhary, Frenel DeMorin,
Lise Genoble, Campbell McInnes, Christine Lercara, Sam Milstein, Doris O’Toole, Donald Sarubbi,
Bruce Variano, and Duncan R. Paton
Emisphere Technologies, Inc., Hawthorne, New York, 10532
Received J anuary 11, 1996X
A series of N-acetylated, non-R, aromatic amino acids was prepared and shown to promote the
absorption of recombinant human growth hormone (rhGH) from the gastrointestinal tract.
Seventy compounds in this family were tested in vivo in rats. Of the compounds tested, 4-[4-
[(2-hydroxybenzoyl)amino]phenyl]butyric acid was identified as a preclinical candidate and
was used to demonstrate the oral delivery of rhGH in primates. A significant positive correlation
was found between the relative log k′ of the delivery agents, as determined by HPLC on an
immobilized artificial membrane (IAM) column, and serum rhGH concentrations following oral
or colonic dosing in rats. Structure-activity relationships have also been developed on the
basis of electronic effects and hydrogen-bonding characteristics of the aromatic amide
substituents.
In tr od u ction
be administered by subcutaneous or intramuscular
injection; however, subcutaneous injection is preferable
because it is less painful.8 rhGH therapy has been
approved for a number of medical indications, but the
clearest clinical indication for rhGH is replacement
therapy in pediatric isolated growth hormone defi-
ciency.9 The clinical manifestation of this condition is
a lack of growth as evidenced by a height that is
significantly below that expected on the basis of age and
sex. Growth hormone therapy increases growth and
improves bone density in this population of patients
with relatively few adverse side effects.10 Daily subcu-
taneous administration is recommended because it has
proven to be more effective than the three times/week
or once weekly dosing schedules used previously.11 The
daily dosing schedule also more closely mimics the
natural, circadian cycle of normal, endogenous growth
hormone release.
Dramatic progress in the field of biotechnology during
the past decade has resulted in a large increase in the
number of commercially available protein drugs.1 While
these new drugs have enormous therapeutic potential,
they represent a major clinical challenge for oral
delivery2 because the function of the gastrointestinal
tract, by design, is to degrade proteins3 and prevent
their absorption.4 Acid-induced hydrolysis in the stom-
ach, enzymatic degradation throughout the gastrointes-
tinal tract, and bacterial fermentation in the colon are
among the many barriers that can prevent the oral
delivery of proteins and large peptides. Physical bar-
riers to delivery include poor solubility in the intestinal
environment and lack of permeation through the epi-
thelial cells. The latter can exclude the passage of
compounds across the tissue on the basis of size, charge,
and/or lipophilicity. Given these barriers, it is not
surprising that the oral delivery of protein and peptide
drugs has been considered impossible or, at best, dif-
ficult. Taken together, these characteristics result in
minimal oral bioavailibility of protein drugs5 and neces-
sitate administration by injection in order to achieve
therapeutic concentrations of the drug in the blood.
Human growth hormone is a protein drug that has
been used to treat growth hormone deficiency in chil-
dren since 1958.6 Endogenously, it is a heterogeneous
mixture of polypeptides whose main component is a
single-chain, 191 amino acid protein (MW 22 124)
incorporating two intramolecular disulfide bonds and
a free amino terminus.7 The drug was obtained exclu-
sively from human, cadaveric, pituitary extracts until
1985 when recombinant human growth hormone (rhGH)
became one of the first products of the biotechnology
industry. Ample supplies of rhGH are currently ob-
tained from various bacterial strains and mammalian
cells that have been modified by the addition of the gene
for human growth hormone.6
Because rhGH is most commonly used in pediatric
patients, a noninjectable dosage form would greatly
facilitate patient compliance. To date, only two studies
in this area have been reported, and both have met with
limited success. Ron et al.12 have developed a controlled-
release system for long-term rhGH therapy that uses a
biodegradable polymer implant. Implantation of this
polyanhydride polymer impregnated with bovine growth
hormone (bGH) into the lower abdominal cavity of rats
caused the release of biologically active bGH into the
circulation for up to 3 weeks. In the other study, serum
levels of rhGH were reported following intranasal
administration in humans.13 These studies notwith-
standing, oral delivery of rhGH would still be preferred
to parenteral, but, to our knowledge, has not yet been
achieved.
As part of continuing efforts to design and develop
oral drug delivery agents, a family of novel compounds
has been prepared that promotes the gastrointestinal
absorption of rhGH in rats and primates. These com-
pounds were selected from the third generation of
delivery agents designed and synthesized in our labo-
ratories. In a previous report,14 the first generation of
drug delivery agents has been presented. These mate-
Currently, all of the FDA-approved growth hormone
preparations are parenteral formulations. The drug can
X Abstract published in Advance ACS Abstracts, J une 1, 1996.
S0022-2623(96)00038-6 CCC: $12.00 © 1996 American Chemical Society

2572 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Leone-Bay et al.
an orally dosed solution of 2 (300 mg/kg) and rhGH (6
mg/kg) and a colonically dosed solution of 2 (25 mg/kg)
and rhGH (1 mg/kg). The serum rhGH concentrations
are similar for both routes of administration in spite of
the significantly lower doses of both delivery agent and
drug used in the colonic study.
F igu r e 1. Generic structural representation of N-acylated
non-R-amino acid oral drug delivery agents.
The data compiled from the rat in vivo studies,
reported in Table 2, suggest a number of interesting
relationships between structure and activity. For ex-
ample, consider the homologous series of compounds 1,
29, 35, and 44. An increase in the activities of these
delivery agents appears to correlate with an increase
(from 0 to 3) in the length of the amide carbon chain.
Figure 3 is a plot of the peak rhGH serum concentra-
tions versus the log relative k′, as measured by chro-
matography on an immobilized artificial membrane
(IAM) column.20 The ability of this group of compounds
to promote the absorption of rhGH increases with
increasing carbon chain length and, by implication,
increasing lipophilicity in the amide portion of the
molecule.
rials, which formed microspheres, were highly complex,
structurally uncharacterized mixtures of thermally
condensed R-amino acids called proteinoids.14 Subse-
quently, microsphere formation and oral delivery of
salmon calcitonin in rats and primates was demon-
strated by the use of structurally defined N-acylated
R-amino acids.15 These latter compounds were also
shown to promote the oral delivery of both salmon
calcitonin and interferon-R in rats and primates when
dosed as an aqueous solution instead of a microsphere
suspension.16 Herein is reported the successful oral
delivery of rhGH in rats and primates from aqueous
solutions using a series of N-acylated, non-R-amino acids
as novel drug delivery agents.
Resu lts a n d Discu ssion
We have also identified a correlation between the
electronic properties of diverse substituents at the ortho
position of the amide aromatic ring and the efficiency
of drug delivery. The data indicate a general trend
toward improved activity for delivery agents bearing an
electron-withdrawing ortho substituent. For example,
22 (Y ) 2-OCF3) is significantly more efficient at
promoting the absorption of rhGH than is 20 (Y )
2-OCH3). For the various ortho-substituted compounds,
1 (2-H), 8 (2-Cl), 9 (2-F), 13 (2-CH3), 20 (2-OCH3), and
22 (2-OCF3), activity increases with increasing σ0
values21 (Figure 4). Two additional compounds in this
series that do not follow this general trend are 2 (2-
OH) and 19 (2-NO2). Delivery agents 2 and 19 show
greater and lesser activities, respectively, than predicted
by the σ0 values. This finding is not unexpected because
the greatest incidence of failures in the Hammett
equation are encountered in correlations involving
molecules bearing substituents capable of either accept-
ing or donating an electron pair.21 In this case, both 2
and 19 have ortho substituents that are capable of
hydrogen bonding. Hydrogen bonding between the
hydroxyl hydrogen and the amide carbonyl in 2 results
in the formation of a 6-membered ring. The hydrogen
bonding between the nitro oxygens and the NH of the
amide bond in 19 also appears to be a favorable
interaction.
Syn th etic Design . A family of oral drug delivery
agents has been prepared comprising 70 compounds
generically represented by the chemical formula shown
in Figure 1. All of the compounds described herein are
derivatives of either 4-aminobenzoic acid (n ) 0), 2-(4-
aminophenyl)acetic acid (n ) 1), 3-(4-aminophenyl)-
propionic acid (n ) 2), or 4-(4-aminophenyl)butyric acid
(n ) 3). These compounds were designed as part of
an effort to expand upon our earlier leads, which
were N-acylated, R-amino acids.16 Of the current com-
pounds, 2 has been identified as a preclinical candidate
(Table 1).
Syn th etic P r oced u r es. The compounds listed in
Table 1 were prepared by standard techniques17-19 in
either aqueous or organic solvents. In general, aqueous
reaction conditions included the dissolution of a non-R-
amino acid in aqueous sodium hydroxide followed by the
addition of the appropriate acid chloride. After the
mixture was stirred at room temperature for about 2 h,
the product was isolated by precipitation from the
acidified reaction mixture and purified by recrystalli-
zation. A representative process in an organic solvent
was the dissolution of equimolar amounts of a non-R-
amino acid and triethylamine in tetrahydrofuran, fol-
lowed by the addition of a selected acid chloride. After
about 4 h of stirring at room temperature, the reaction
mixture was quenched, and the product was purified
by recrystallization. All of the compounds reported
herein were prepared by either of these two methods
with the exception of compounds 15, 16, 21, 22, 31, 34,
39, 40, 51, and 70 whose syntheses are described in the
Experimental Section.
In an effort to use an electronic parameter that
accounts for the effects of hydrogen bonding, σ- values
were substituted for the σ0 values of 2 and 19. This
manipulation did not improve the correlation between
the electronic activity and the in vivo data of these two
compounds.
The effects of meta and para substituents on the
amide aromatic ring were also investigated. In general,
any substituent other than hydrogen at the para posi-
tion resulted in substantially reduced (6, 7, 33, 54, 61,
and 63) or highly variable (4) protein delivery. Electron-
withdrawing substituents at the meta position de-
creased activity (11, 12, and 21); however, the highly
electron-donating meta, N,N-dimethyl substituent (15)
increased activity.
Lea d Id en tifica tion a n d Str u ctu r e-Activity Re-
la tion sh ip s. Each of the compounds in Table 1 was
evaluated in a rat model for its ability to promote the
absorption of rhGH. During the course of our screening
studies, it was found that rhGH absorption following
intracolonic dosing was comparable to oral gavage, but
at lower doses of both delivery agent and drug. The
bioavailability of rhGH relative to subcutaneous injec-
tion is 3-5% orally and 8-10% colonically. Subse-
quently, colonic administration replaced dosing by oral
gavage. Figure 2 is a comparison of the results from
The trend of increased lipophilicity that results in the
increased delivery of drug previously mentioned is also

Novel Oral Delivery Agent for rhGH
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2573
Ta ble 1. Protein Oral Delivery Agents
compd
Y
X
n
m
Z
compd
Y
X
n
m
Z
1
2
3
4
5
6
7
8
9
H
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
SO2
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
3
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
1
1
1
0
1
1
0
2
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
cyclohexyl
cyclohexyl
cycloheptyl
cyclohexyl
cyclohexyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
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
61
62
63
64
65
66
67
68
69
70
H
F
2-CH3
2-OCH3
2-F
H
H
H
H
H
H
H
H
H
2-OH
2-NO2
2,6-(OH)2
2-OCH3
4-OCH3
H
2-OH
H
H
H
H
4-CH3
2-OH
4-F
H
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CdO
SO2
CdO
CdO
CdO
CdO
CdO
CdO
SO2
CdO
CdO
CdO
CdO
CdO
CdO
3
3
3
3
3
3
3
3
3
3
3
3
3
3
2
2
2
2
2
vinyl
vinyl
vinyl
vinyl
vinyl
CHMe
CHEt
CHEt
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
cyclohexyl
phenyl
phenyl
phenyl
phenyl
cyclohexyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
phenyl
cyclohexyl
phenyl
cyclohexyl
phenyl
phenyl
phenyl
cyclohexyl
phenyl
phenyl
phenyl
phenyl
cyclohexyl
2-OH
2,3-Ph
4-Ph
3,4-Ph
4-OCH3
4-F
2-Cl
2-F
2,4-(OH)2
3-CF3
3-Cl
2-CH3
2,6-(OH)2
3-N(CH3)
3,4-OCH2O
2,6-diCH3
2-COOH
2-NO2
2-OCH3
3-NO2
2-OCF3
H
H
4-CH3
H
2-COOH
H
H
3
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
(CH2)2O
(CH2)2CdO
(CH2OH)2
3
4
0
0
0
vinyl
vinyl
0
0
0
1
1
2
0
0
0
0
1
1
2
vinyl
1
1
1
1
1
0
0
0
0
0
0
0
0
0
0
CdO
CdO
CdO
CdO
CdO
CdO
CdO
CH2
2-OH
2-OH
2-F
4-i-Bu
H
H
H
H
CdO
CdO
CH2
2-OCH3
H
H
vinyl
3
3
H
CdO
showed 36 to be at least 2 times more effective in
promoting the gastrointestinal absorption of rhGH, even
at a lower dose. This same trend is observed with
intracolonic dosing of rats (Table 2) with aqueous
solutions of rhGH (1 mg/kg) and either 20 or 39 (25 mg/
kg) where 39 showed twice the activity of 20.
On the basis of the results of the rodent studies,
delivery agent 2 was chosen as a preclinical candidate
and used to demonstrate oral rhGH delivery in a
primate model. The pharmacokinetic profile after oral
dosing of four cynomolgus monkeys with aqueous solu-
tions of 2 (800 mg/kg) and rhGH (6 mg/kg) via naso-
gastric gavage is presented in Figure 5.
Once 2 was identified as the most active delivery
agent in this series of compounds, the specific intramo-
lecular hydrogen-bonding properties of this molecule
and its relationship to drug delivery was more closely
studied. Compounds 14, 30, and 31 were prepared and
tested. Although structurally similar, none of these
compounds has the ability to participate in the particu-
lar intramolecular hydrogen-bonding interactions that
are available to 2. Compound 14, although it bears the
necessary o-hydroxy group, is sterically encumbered by
an additional hydroxy substituent at position 6 on the
aromatic ring. Therefore, only 2 can assume a planar
conformation as a result of the through-space interac-
tion between the phenolic hydroxyl hydrogen and the
amide carbonyl forming a 6-membered ring (Figure 6).
The substitution of an amine function in 31 for the
F igu r e 2. Comparison of oral and colonic administration of
2 and rhGH in rats. The triangles represent the response
following intracolonic instillation of an aqueous solution of 2
(25 mg/kg) and rhGH (1 mg/kg). The squares represent the
response following oral administration of an aqueous solution
of 2 (300 mg/kg) and rhGH (6 mg/kg).
evident when the efficacy of amide delivery agents and
their vinylogs is evaluated. This contention is supported
by the in vivo activities of two pairs of compounds,
namely, 1/36 and 20/39. Compound 36 is the vinylogous
amide of 1, and 39 is the vinylogous amide of 20. Oral
dosing (Table 2) of aqueous solutions of rhGH (6 mg/
kg) and either 1 (500 mg/kg) or 36 (400 mg/kg) in rats

2574 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Leone-Bay et al.
Ta ble 2. Delivery of Recombinant Human Growth Hormone (rhGH) in Rats
compound dose rhGH dose
route of
administration
peak [rhGH]
(ng/mL)
compound dose rhGH dose
route of
administration
peak [rhGH]
(ng/mL)
compd
(mg/kg)
(mg/kg)
compd
(mg/kg)
(mg/kg)
1
500
300
25
300
25
300
300
400
300
300
25
300
25
200
25
25
25
25
25
25
25
25
25
25
25
6
6
1
6
1
6
6
6
6
6
1
6
1
6
1
1
1
1
1
1
1
1
1
0.8
1
1
6
6
6
6
6
3
6
6
6
1
1
6
6
6
OGa
OG
ICb
OG
IC
OG
OG
OG
OG
OG
IC
OG
IC
OG
IC
IC
IC
IC
IC
IC
IC
IC
IC
IC
IC
21.7 ( 19.5a 36
400
200
25
25
25
300
300
400
300
25
25
25
25
25
25
25
25
25
25
400
25
500
300
300
300
400
25
300
25
500
25
500
500
300
25
300
300
0
6
6
1
1
1
6
6
6
6
1
1
1
1
1
1
1
1
1
1
6
1
6
6
6
6
6
1
6
1
6
1
6
6
6
1
6
6
6
1
OG
OG
IC
OG
IC
OG
OG
OG
OG
IC
IC
IC
IC
IC
IC
IC
IC
IC
IC
OG
IC
OG
OG
OG
OG
OG
IC
OG
IC
OG
IC
OG
OG
OG
IC
OG
OG
OG
IC
49.7 ( 25.0
31.1 ( 14.4
16.4 ( 17.8
32.9 ( 8.2
7.7 ( 4.2
2.3 ( 2.6
5.9 ( 6.6
0.5 ( 0.7
17.9 ( 6.9
44.1 ( 8.2
0.6 ( 0.7
2.5 ( 2.6
26.0 ( 5.7
8.4 ( 0.5
37.2 ( 17.5
17.6 ( 14.7
9.3 ( 3.3
61.7 ( 17.8
5.0 ( 5.6
6.5 ( 4.6
5.6 ( 6.3
0
5.4 ( 3.0
17.5 ( 3.5
49.7 ( 13.5
64.7 ( 5.5
5.9 ( 7.8
39.9 ( 34.4
28.2 ( 4.5
0
2.8 ( 3.2
52.2 ( 11.4
43.4 ( 16.6
20.8 ( 5.8
16.9 ( 14.4
17.6 ( 14.5
0
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
9.08 ( 2.9
20.9 ( 11.5
50.7 ( 14.5
0
7.7 ( 4.2
22.3 ( 5.5
14.9 ( 6.8
17.8 ( 10.7
33.5 ( 12.7
56.8 ( 15.7
28.0 ( 20.3
2.6 ( 3.1
63.8 ( 7.4
41.6 ( 18.5
92.8 ( 27.5
0
9.0 ( 7.6
15.7 ( 8.0
12.6 ( 10.3
0
25
IC
500
250
500
400
500
200
500
300
300
25
OG
OG
OG
OG
OG
OG
OG
OG
OG
IC
0
1.0 ( 1.2
2.2 ( 2.5
13.7 ( 14.1
0
0
0
3.3 ( 4.1
30.4 ( 12.9
11.9 ( 13.3
29.3 ( 19.3
0
24
25
26
27
28
29
30
31
32
33
34
35
64
65
24.2 ( 14.5
8.1 ( 2.6
0
66
67
68
32.5 ( 7.7
7.3 ( 3.0
7.8 ( 5.0
20.4 ( 17.0
12.8 ( 3.9
69
70
none
25
IC
200
200
300
OG
OG
OG
0
0
a
b
Oral gavage. Intracolonic instillation. c Mean ( SD.
F igu r e 4. Correlation of σ to peak serum concentrations of
rhGH after intracolonic dosing of rats with an aqueous solution
of delivery agent at 25 mg/kg and rhGH at 1 mg/kg. Com-
pounds 2, 8, 9, 13, 19, 20, and 22 comprise an homologous
series in which the ortho substituent varies. A correlation
coefficient of 0.90 was obtained from linear regression analysis
when 2 and 19 were excluded.
F igu r e 3. Correlation of log relative k′ from the IAM column
to the peak serum rhGH concentrations after orally dosing rats
with an aqueous solution of delivery agent at 300 mg/kg and
rhGH at 6 mg/kg. Compounds 1, 29, 35, and 44 comprise an
homologous series in which the amide carbon chain length
increases from 0 to 3. A correlation coefficient of 0.98 was
obtained from linear regression analysis.
Consequently, neither 30 nor 31 is likely to be planar
structures. Perhaps the most interesting compound in
this group is 14: the additional aromatic hydroxyl
function in the 6 position presents the possibility of
amide in 2 prevents intramolecular hydrogen bonding
completely, and intramolecular hydrogen bonding in 30
must occur through a less favorable 7-membered ring.

Novel Oral Delivery Agent for rhGH
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2575
F igu r e 5. Pharmacokinetic profile following oral administra-
tion of an aqueous solution of 2 (800 mg/kg) and rhGH (6 mg/
kg) in conscious cynomolgus monkeys.
F igu r e 7. Correlation of log relative k′ from IAM to peak
serum concentrations of rhGH after intracolonic dosing of rats
with an aqueous solution of delivery agent at 25 mg/kg and
rhGH at 1 mg/kg. Compounds 2, 50, 56, and 62 comprise a
homologous series in which the acid carbon chain length
increases from 1 to 4. A correlation coefficient of 0.98 was
obtained from linear regression analysis.
for study were 50, 56, and 62. The observation that 2
was significantly more active than 30 appeared to rule
out the effectiveness of higher amide homologues. As
expected, based on the previous observation that in-
creased delivery agent lipophilicity within a structural
series leads to improved drug delivery (Figure 3), a
similar relationship exists in the series 62, 56, 50, and
2 in which the rhGH serum concentrations increase as
the length of the acid carbon chain increases from 1-4
(Figure 7).
Mech a n istic Stu d ies. The promising in vivo data
generated significant speculation in our laboratories
with regard to the mechanism of action of these novel
protein delivery agents, including classical penetration
enhancement and/or enzyme inhibition. The latter was
deemed unlikely based on our previous report on similar
compounds in which the delivery agents were shown to
be extremely inefficient inhibitors of proteolytic en-
zymes.15
In order to address the issue of the delivery agents
acting as penetration enhancers and causing drug
absorption as a result of alteration in membrane
structure (i.e., damage), histopathological studies were
conducted on rats dosed orally or intracolonically with
2. Groups of five or six rats were given aqueous
solutions of 2 at a dose of 300 mg/kg (oral) or 25 mg/kg
(colonic) and sacrificed at 30, 60, and 120 min postdos-
ing. Necropsy of the gastrointestinal tract that included
the stomach, duodenum, jejunum, and ileum of each
animal, followed by histological examination, indicated
no untoward pathology. These studies suggest that
drug transit across the intestinal membrane is not the
result of mucosal damage.
F igu r e 6. Comparison of the intramolecular hydrogen-
bonding capabilities of 2, 30, 31, and 14. Hydrogen bonding
between the phenolic hydrogen and the amide carbonyl occurs
through a 6-membered ring in 2. The same interaction in 30
produces a 7-membered ring. The substitution of an amine
for the amide in 31 prevents the hydrogen-bonding interaction
completely. The additional hydroxyl in 14 allows for two
intramolecular hydrogen-bonding interactions.
forming two 6-membered rings via intramolecular hy-
drogen bonding by the interaction of one hydroxyl
hydrogen with the amide carbonyl and the other hy-
droxyl hydrogen with the amide nitrogen. Coincidental
formation of these two hydrogen bonds would force the
aromatic rings of 14 into an orthogonal orientation.
Examination of the in vivo data for 2, 14, 30, and 31
indicates that 2 is the most effective in promoting the
oral absorption of rhGH. This suggests that hydrogen-
bonding substituents that impose an overall planar
structure may be a requirement for efficient oral protein
delivery.
Encouraged by these findings, additional studies
aimed at examining the mechanism(s) of drug delivery
were started. For example, the efficiency of compound
2 in mediating rhGH delivery in rats was studied as a
function of both dose and dose volume. Figure 8 shows
the effects of varying the dose of 2 from 25 to 10 mg/kg.
This data suggests that the minimum effective delivery
agent dose is 20 mg/kg in the rat model under these
To gain additional information on the lead compound,
homologs of 2 that varied only in the length of the acid
carbon chain were examined. The compounds chosen

2576 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Leone-Bay et al.
increasing solute concentration at concentrations below
saturation. In this case, the more concentrated dosing
solutions provide more favorable conditions for the
intermolecular association of rhGH with 2. Therefore,
one might postulate that some type of “complex” be-
tween the protein and the delivery agent could be
responsible for efficient drug absorption.
On the basis of this hypothesis, studies designed to
identify techniques that would facilitate characteriza-
tion of the interaction between 2 and rhGH and its role,
if any, in drug delivery were initiated. Many of the
conventional spectroscopic techniques commonly used
to study intermolecular interactions were found not to
be applicable here because of interference from the
delivery compounds at the wavelengths of interest or
interaction with the specific probes designed to evaluate
intermolecular phenomena. However, preliminary data
from differential scanning calorimetry (DSC), capillary
zone electrophoresis (CZE), and molecular modeling
experiments may prove useful in characterizing the
nature of the interaction between the delivery agent and
rhGH. Comprehensive studies using these techniques
are underway.
F igu r e 8. Effect of varying the dose of delivery agent 2 on
the delivery of rhGH (1 mg/kg) in rats following intracolonic
administration of 1 mL/kg of dosing solution. The squares
represent the response at a delivery agent dose of 50 mg/kg.
The circles represent the response at a delivery agent dose of
25 mg/kg. The inverted triangles represent the response at a
delivery agent dose of 20 mg/kg. The upright triangles
represent the response at a delivery agent dose of 10 mg/kg.
Con clu sion
We have prepared a series of 70 N-acylated, non-R-
amino acids and demonstrated their use in two species
as oral delivery agents for rhGH. Structure-activity
analyses showed that increasing lipophilicity within
structurally related groups of compounds enhanced
their ability to promote protein absorption from the
gastrointestinal tract in a rat model. The effects of
substituents in the aromatic amide portion of these
compounds with varying electronic properties were also
examined, and electron withdrawing substituents at the
ortho position were found to be the most efficacious.
Preliminary mechanistic studies suggest that the for-
mation of a drug/delivery agent complex is necessary
to achieve oral drug delivery. Future studies will be
focused on characterizing the mechanism by which these
materials promote protein absorption following oral
administration.
F igu r e 9. Effect of varying dose volume on the delivery of
rhGH (1 mg/kg) in rats using delivery agent 2 (25 mg/kg). The
triangles represent the response following intracolonic instil-
lation of 0.5 mL/kg of dosing solution. The circles represent
the response following intracolonic instillation of 1.0 mL/kg
of dosing solution. The squares represent the response fol-
lowing intracolonic instillation of 1.5 mL/kg of dosing solution.
Exp er im en ta l Section
Ch em istr y. NMR spectra were recorded at 300 MHz in
either D2O or DMSO-d6. Combustion analyses were performed
by Microlit Laboratories, Madison, NJ , and are within accept-
able limits (C, H, N ( 0.4%). Thin layer chromatography was
performed using E. Merck Kieselgel 60 F-254 plates. Reac-
tions were monitored by high-pressure liquid chromatography
on a Vydac 25 × 4.6 mm Protein and Peptide column using a
gradient of 0-50% acetonitrile in water with 0.1% trifluoro-
acetic acid. Melting points were performed using a Mel-Temp
II from Laboratory Devices.
Gen er a l P r oced u r es for th e P r ep a r a tion s of Deliver y
Agen ts. With the exception of 18, 27, 31, and 34, one of the
following three procedures was used to prepare the compounds
described herein. The preparation of 4-(4-(salicyloylamino)-
phenyl)butyric acid (2) is given as a representative example.
Meth od A. P r ep a r a tion of 2. A 1 L round bottom flask
fitted with a magnetic stirrer was charged with 4-(4-ami-
nophenyl)butyric acid (50.0 g, 0.28 mol, 1.17 equiv) and 2 M
aqueous sodium hydroxide (300 mL). Finely ground O-
acetylsalicyloyl chloride (47.7 g, 0.24 mol, 1.00 equiv) was
added portionwise over 1 h to the stirred solution. After the
addition, the reaction mixture was stirred for 2.5 h at ambient
temperature, and the pH of the solution was kept at ca. 10 by
the addition of 10 M sodium hydroxide. The solution was then
acidified with 1 M hydrochloric acid to pH 4.5 to obtain a pale
unoptimized dosing conditions. Similar studies were
carried out in which the dose of compound 2 was held
constant at 25 mg/kg and the dose of rhGH was varied
over the range 6, 3, 1, 0.5, and 0.25 mg/kg. Under these
dosing conditions, the minimum effective dose of rhGH
was determined to be 0.5 mg/kg.
To determine the effect of dose volume on rhGH
delivery in the presence of 2, the doses of 2 and rhGH
were held constant at 25 and 1 mg/kg, respectively,
while the dose volume was varied from 0.5 to 1.5 mL/
kg. It is important to note that the concentration of the
delivery agent and the rhGH in the dosing solutions will
increase as the dosing volume decreases because the
total dose is held constant. The data presented in
Figure 9 show that the efficiency of drug delivery
increased with decreasing dose volume. One possible
explanation for this observation is that intermolecular
associations between molecules in solution increase with

Novel Oral Delivery Agent for rhGH
J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13 2577
pink solid. The solid was filtered, washed with 1 M hydro-
chloric acid (3 × 100 mL) and water (100 mL), and air-dried.
It was redissolved in boiling acetone (ca. 500 mL), decolorized
with activated charcoal (3 g), and filtered. Water (1.5 L) was
added to the filtrate to induce the formation of a brown oil.
The brown oil solidified upon stirring at room temperature for
10 min. The crude solid was collected by filtration and
recrystallized from 70% methanol-water (v/v) to afford 2 as a
g, 0.028 mol) was added followed by concentrated hydrochloric
acid (0.5 mL). The reaction mixture was heated to reflux
under nitrogen for 12 h and then cooled to ambient temper-
ature. Sodium cyanoborohydride (2.1 g, 0.33 mol) was added
and the mixture stirred for 4 h. The reaction was quenched
by the addition of concentrated hydrochloric acid (10 mL), and
the methanol was evaporated in vacuo. The residue was
purified by column chromatography (40% ethyl acetate/hex-
anes) on silica gel to afford 34 as a colorless solid (1.4 g, 19%):
1
tan solid (35.1 g, 49%): mp 159-163 °C; H NMR (300 MHz,
1
DMSO-d6) δ 7.74 (1H, dd), 7.38 (2H, d), 7.21 (3H, m), 6.67 (1H,
m), 6.57 (1H, m), 2.48 (2H, t), 2.07 (2H, t), 1.71 (2H, m). Anal.
Calcd for C17H17NO4: C, 68.22; H, 5.72; N, 4.68. Found: C,
68.29; H, 5.60; N, 4.60.
mp 55-58 °C; H NMR (300 MHz, DMSO-d6) δ 7.3-7.4 (m,
3H), 7.2 (m, 1H), 6.8 (d, 2H), 6.5 (d, 2H), 6.1 (s, 1H), 4.2 (s,
2H), 2.4 (t, 2H), 2.2 (t, 2H), 1.7 (q, 2H). Anal. Calcd for C17H19
-
NO2: C, 75.79; H, 7.12; N, 5.18. Found: C, 75.55; H, 7.49; N,
4.70.
Compounds 1, 3-6, 10, 14, 23-26, 28-30, 33, 41-43, 47-
48, 50, 55, 57-61, and 63-67 were prepared by this process.
Meth od B. P r ep a r a tion of 2. A 2 L three-neck round
bottom flask was fitted with a magnetic stirrer and two
addition funnels under an argon atmosphere. A suspension
of 4-(4-aminophenyl)butyric acid (50.0 g, 0.28 mol, 1.17 equiv)
in ethyl acetate (700 mL) was added to the flask. A solution
of O-acetylsalicyloyl chloride (55.60 g, 0.28 mol, 1.00 equiv) in
ethyl acetate (250 mL) was charged to one of the addition
funnels and added dropwise over 1 h. Triethylamine (28.20
g, 0.28 mol, 1.00 equiv) was subsequently charged to the second
funnel and added dropwise over 15 min. The reaction mixture
was stirred at ambient temperature for 3 h, and the solvent
was evaporated in vacuo, giving a residual brown oil. Water
(600 mL) was added to the residue followed by sodium
hydroxide (2 M, 500 mL), and the mixture was stirred at
ambient temperature for 3 h. The resultant brown solution
was acidified with 2 M hydrochloric acid (ca. 1 L). After the
mixture was cooled in an ice bath for 1 h, a yellow solid formed
and was collected by filtration. The solid was washed with
water (3 × 1.5 L) and recrystallized from 50% ethanol-water
(v/v) to give 2 as a tan solid (56.60 g, 68%): mp 167-170 °C;
1H NMR (see method A). Anal. Calcd for C17H17NO4: C,
68.22; H, 5.72; N, 4.68. Found: C, 67.87; H, 5.85; N, 4.69.
Compounds 7-9, 11-13, 17, 20, 28, 32, 35-38, 44-46, 49,
52-54, 56, 62, and 68-69 were prepared by this process.
Meth od C. P r ep a r a tion of 2. A 2 L round bottom flask
equipped with a magnetic stirrer and a reflux condenser was
charged with a suspension of 4-(4-aminophenyl)butyric acid
(50.0 g, 0.28 mol, 1.17 equiv) in dichloromethane (560 mL).
Chlorotrimethylsilane (62.36 g, 0.57 mol, 2.05 equiv) was
added in one portion, and the mixture was heated to reflux
for 1 h under argon. The reaction mixture was allowed to cool
to room temperature and was placed in an ice bath (internal
temperature < 10 °C). The reflux condenser was replaced with
an additional funnel containing triethylamine (42.50 g, 0.42
mol, 1.50 equiv). The triethylamine was added dropwise over
15 min and a yellow solid formed during the addition. The
funnel was replaced by another addition funnel containing a
solution of O-acetylsalicyloyl chloride (55.60 g, 0.28 mol, 1.00
equiv) in dichloromethane (100 mL). The solution was added
dropwise over 30 min. The reaction was stirred in the ice bath
for another 30 min and at ambient temperature for 1 h. The
dichloromethane was evaporated in vacuo to give a brown oil.
The brown oil was cooled in an ice bath, and an ice-cold
solution of 2 M sodium hydroxide (700 mL) was added. The
ice bath was removed, and the reaction mixture was stirred
for 2 h to afford a clear brown solution. The solution was
acidified with 2 M sulfuric acid (400 mL) and stored at ca. 5
°C for 1 h. A yellow solid formed and was collected by
filtration. The solid was washed with water (3 × 100 mL) and
recrystallized from 50% ethanol-water (v/v) to afford 2 as tan
needles (65.7 g, 78%): mp 174-176 °C; 1H NMR (see method
A). Anal. Calcd for C17H17NO4: C, 68.22; H, 5.72; N, 4.68.
Found: C, 68.27; H, 5.70; N, 4.62.
Compound 31 was prepared by this process.
P r ep a r a tion of 70. A 500 mL round bottom flask equipped
with a magnetic stirrer and a reflux condenser was charged
with cyclohexanebutyric acid (17.0 g, 0.10 mol) and chloroform
(200 mL). N-Hydroxysuccinamide (12.7 g, 0.11 mol) and N,N-
dicyclohexylcarbodiimide (22.7 g, 0.11 mol) were added, and
the mixture was stirred at ambient temperature for 12 h. A
white precipitate formed and was removed by filtration.
Glacial acetic acid (5 mL) was added to the filtrate and stirred
for 3 h. The solution was washed with saturated sodium
bicarbonate (2 × 100 mL) and water (2 × 100 mL). The
combined aqueous extracts were back-extracted with chloro-
form (50 mL). The combined chloroform extracts were dried
over anhydrous magnesium sulfate, filtered, and evaporated.
The residue was dissolved in a solution of methyl 4-aminoben-
zoate (15.2 g, 0.10 mol), 2 M sodium hydroxide (30 mL), and
methanol (60 mL). The resulting mixture was heated to reflux
for 4 h, cooled in an ice bath, and acidified to pH 1 with 1 M
hydrochloric acid. A white solid formed, was collected by
filtration, and was washed with petroleum ether. Recrystal-
lization from 50% methanol-water (v/v) gave 70 as colorless
1
crystals (22.9 g, 79%): mp 240-242 °C; H NMR (300 MHz,
alkaline D2O) δ 7.7 (d, 2H), 7.3 (d, 2H), 2.1 (M, 2H), 1.4 (m,
7H), 1.0 (m, 6H), 0.7 (m, 2H). Anal. Calcd for C17H23
-
NO3‚1.5H2O: C, 60.34; H, 7.38; N, 4.13. Found: C, 60.37; H,
6.81; N, 4.00.
An im a l Exp er im en ts. All animal experimental proce-
dures and protocols were approved in advance by the Emi-
sphere Institutional Animal Care and Use Committee. Male
Sprague-Dawley rats weighing 225-250 g were used to
demonstrate the delivery of rhGH. Each experiment was
performed on a group of six or seven rats. The rats were
housed under standard conditions with free access to water
for 24 h prior to the study. All of the rats in these studies
were anesthetized with 44 mg/kg ketamine and 1.5 mg/kg
thorazine immediately prior to dosing. The rats were admin-
istered the dosing solutions by oral gavage or intracolonic
instillation. rhGH serum concentrations were measured by
an ELISA assay for recombinant human growth hormone
from Medix Biotech, Inc., Foster City, CA. The data is
reported as mean ( standard error. Histology studies were
performed by Pharmaco LSR, East Millstone, NJ . Recombi-
nant human growth hormone (rhGH) was a gift from Eli Lilly,
Indianapolis, IN.
Gen er a l P r otocol for Ra t Exp er im en ts. The following
protocol is a general description of the rat experiments
performed as part of these studies. The oral dosing of rhGH
with 2 is given as a representative example. An rhGH dosing
solution was prepared by dissolving compound 2 to a concen-
tration of 300 mg/mL in water and adjusting the pH of the
solution to 7.2-7.8 with aqueous sodium hydroxide (1.0 N).
rhGH was added to obtain a final concentration of 3 mg/mL.
Six rats were given 2 mL/kg of this rhGH dosing solution. The
total dose of rhGH was 6 mg/kg and the total dose of delivery
agent was 600 mg/kg for oral gavage studies, and the total
dose of rhGH was 1 mg/kg and the total dose of delivery agent
was 25 mg/kg for intracolonic studies. Control groups, each
containing six rats, were administered a solution of rhGH or
a solution of compound 2. All of the groups were dosed at the
same time. Blood samples were collected serially from the tail
artery immediately prior to dosing and at 0.25, 0.5, 1.0, and
Compounds 15-16, 19, 21-22, 39-40, and 51 were pre-
pared by this process. Compounds 18 and 27 were also
prepared by this process using the appropriate anhydride in
place of an acid chloride.
P r ep a r a tion of 34. A 100 mL round bottom flask equipped
with a magnetic stirrer and a reflux condenser was charged
with a suspension of 4-(4-aminophenyl)butyric acid (5.0 g,
0.028 mol) in absolute methanol (60 mL). Benzaldehyde (2.96

2578 J ournal of Medicinal Chemistry, 1996, Vol. 39, No. 13
Leone-Bay et al.
(9) Strobl, J . S.; Thomas, M. J . Human growth hormone. Pharm.
Rev. 1994, 46, 1-34.
1.5 h after dosing, allowed to clot for 30-60 min, and
centrifuged, and the serum was harvested and frozen until
assay.
(10) Hintz, R. L. Untoward events in patients treated with growth
hormone in the USA. Horm. Res. 1992, 38, 44-49 (suppl).
(11) Kearns, B. S.; Kemp, S. F.; Frindik, L. A.; Single and multiple
dose pharmacokinetics of growth hormone in children with
idiopathic growth hormone deficiency. J . Clin. Endocrinol.
Metab. 1991, 72, 1148-1156.
Gen er a l P r otocol for P r im a te Exp er im en ts. The fol-
lowing protocol is a general description of the primate experi-
ments performed as part of these studies. An rhGH dosing
solution was prepared by dissolving compound 2 to a concen-
tration of 300 mg/mL in water and adjusting the pH of the
solution to 7.2-7.8 with aqueous sodium hydroxide (1.0 N).
rhGH was added to obtain a final concentration of 2.2 mg/
mL. Monkeys were given 2.7 mL/kg of this rhGH dosing
solution. The total dose of rhGH was 6 mg/kg and the total
dose of 2 was 800 mg/kg. Three cynomolgus monkeys weigh-
ing 3-4 kg were used. The monkeys were fasted overnight,
anesthesized with ketamine, placed in chairs, and allowed to
regain consciousness before dosing. Conscious monkeys re-
ceived a solution of rhGH and 2 by nasogastric gavage, and
blood samples were collected from the saphenous vein. Blood
samples were collected at 1 and 0.5 h before dosing and 0.17,
0.33, 0.5, 0.67, 0.83, 1, 1.5, 2, 3, and 6 h after dosing. The
samples were allowed to clot at 25 °C for 30-60 min and
centrifuged, and the serum was harvested and frozen until
assay. All the samples were analyzed by ELISA for the
determination of rhGH serum concentrations.
(12) Ron, E.; Turek, T.; Mathiowitz, E.; Chasin, M.; Hageman, M.;
Langer, R. Controlled release of polypeptides from polyanhy-
drides. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 4176-4180.
(13) Hedin, L.; Olsson, B.; Diczfalusy, M.; Flug, C.; Petersson, A.-S.;
Rosberg, S. Intranasal administration of human growth hormone
(hGH) in combination with a membrane permeation enhancer
in patients with growth hormone deficiency: a pharmacokinetic
study. J . Clin. Endocrinol. Metab. 1993, 76, 962-967.
(14) Steiner, S.; Rosen, R. Delivery system for pharmacological agents
encapsulated with proteinoids. U.S. Patent 4,925,673, 1990.
(15) Leone-Bay, A.; McInnes, C.; Wang, N.-F.; DeMorin, F.; Achan,
D.; Lercara, C.; Sarubbi, D.; Haas, S.; Press, J .; Barantsevich,
E.; O’Broin, B.; Milstein, S.; Paton, D. Microsphere formation
in a series of derivatized R-amino acids: properties, molecular
modeling and oral delivery of salmon calcitonin. J . Med. Chem.
1995, 38, 4257-4262.
(16) Leone-Bay, A.; Santiago, N.; Achan, D.; Chaudhary, K.; DeMorin,
F.; Falzarano, L.; Haas, S.; Kalbag, S.; Kaplan, D.; Leipold, H.;
Lercara, C.; O’Toole, D.; Rivera, T.; Rosado, C.; Sarubbi, D.;
Vuocolo, E.; Wang, N.-F.; Baughman, R. A. N-Acylated R-amino
acids as novel oral delivery agents for proteins. J . Med. Chem.
1995, 38, 4263-4269.
(17) Greenstein, J . P.; Winitz, M. Chemistry of the Amino Acids; J ohn
Wiley & Sons: New York, 1962; Vol. 2, p 1266.
(18) Chen, M. F. F.; Benoiton, N. L. Diisopropylethylamine eliminates
dipeptide formation during the acylation of amino acids using
benzoyl chloride and some alkyl chloroformates. Can. J . Chem.
1987, 65, 1224-7.
Refer en ces
(1) Biotechnology Medicines and Vaccines Approved and Under
Development. Gen. Eng. News 1995, 15, 12-16.
(2) Smith, P. L.; Wall, D. A.; Gochoco, C. H.; Wilson, G. Routes of
delivery: Case studies. Oral absorption of peptides and proteins.
Adv. Drug. Del. Rev. 1992, 8, 253-290.
(3) Lee, V. H. L. Enzymatic barriers to peptide and protein absorp-
tion. CRC Crit. Rev. Ther. Drug Carrier Syst. 1988, 5, 69-97.
(4) Gonella, P. A.; Walker, W. A. Macromolecular absorption in the
gastrointestinal tract. Adv. Drug Del. Rev. 1987, 1, 235-248.
(5) Humphrey, M. J . The oral bioavailability of peptides and related
drugs. In Delivery Systems for Peptide Drugs; Davis, S. S., Illum,
L., Tomlinson, E., Eds.; Plenum Press: New York, 1986; pp 139-
151.
(6) Neely, E. K.; Rosenfeld, R. G. Use and abuse of human growth
hormone. Annu. Rev. Med. 1994, 45, 407-20.
(7) Gilman, A. G.; Goodman, L. S.; Gilman, A. The Pharmacological
Basis of Therapeutics; McMillan Publishing Company: New
York, 1980; pp 1337-1343.
(8) Russo, L.; Moore, W. V. A comparison of subcutaneous and
intramuscular administration of human growth hormone in the
therapy of growth hormone deficiency. J . Clin. Endocrinol.
Metab. 1982, 55, 1003-1006.
(19) Shinkai, H.; Toi, K.; Kumashiro, I.; Seto, Y.; Fukuma, M.; Dan,
K.; Toyoshima, S. N-acylphenylalanines and related compounds.
A new class of oral hypoglycemic agents. J . Med. Chem. 1988,
31, 2092-7.
(20) Pidgeon, C.; Ong, O.; Liu, H.; Qiu, X.; Pidgeon, M.; Dantzig, A.
H.; Munroe, J .; Hornback, W. J .; Kasher, J . S.; Glunz, L.;
Szczerba, T. IAM chromatography: an in vitro screen for
predicting drug membrane permeability. J . Med. Chem. 1995,
38, 590-594 and references cited therein.
(21) Hansch, C.; Leo, A. Exploring QSAR Fundamentals and Ap-
plications in Chemistry and Biology; American Chemical Soci-
ety: Washington, DC, 1995; pp 1-22, 69-92.
J M960038F