Identification of N-acyl 4-(3-pyridonyl)phenylalanine derivatives and their orally active prodrug esters as dual acting α4β1 and α4β7 receptor antagonists

Article abstract of DOI:10.1016/j.bmcl.2012.12.019

From a series of N-acyl 4-(3-pyridonyl)phenylalanine derivatives of 4, the trifluoromethyl derivative 28 was identified as a potent, dual acting alpha4 integrin antagonist with activity in primate models of allergic asthma. Investigation of a series of prodrug esters led to the discovery of the morpholinopropyl derivative 48 that demonstrated good intestinal fluid stability, solubility and permeability. Compound 48 gave high blood levels of 28 when dosed orally in cynomolgus monkeys. Surprisingly, hydrolysis of 48 was rapid in liver microsomes from the pharmacological species, mouse, rat and monkey, but slow in dog and human; in vivo studies also indicated there was prolonged exposure to unchanged prodrug in dogs.

Full text of DOI:10.1016/j.bmcl.2012.12.019

Bioorganic & Medicinal Chemistry Letters 23 (2013) 1036–1040
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Identification of N-acyl 4-(3-pyridonyl)phenylalanine derivatives and
their orally active prodrug esters as dual acting
receptor antagonists
a4b1 and a4b7
a
a
a
b
b
Jefferson W. Tilley a, , Achyutharao Sidduri , Jianping Lou , Gerry Kaplan , Nadine Tare , Gary Cavallo ,
⇑
Karl Frank c, Anjula Pamidimukkala c, Duk Soon Choi d, Louise Gerber d, Aruna Railkar d, Louis Renzetti b
a Discovery Chemistry, Pharmaceutical Research and Early Drug Development, Hoffmann-La Roche Inc., 340 Kingsland Street, Nutley, NJ 07110, USA
b Discovery Inflammation and Respiratory Diseases, Pharmaceutical Research and Early Drug Development, Hoffmann-La Roche Inc., 340 Kingsland Street, Nutley, NJ 07110, USA
c Drug Metabolism and Pharmacokinetics, Pharmaceutical Research and Early Drug Development, Hoffmann-La Roche Inc., 340 Kingsland Street, Nutley, NJ 07110, USA
d PARD, Pharmaceutical Research and Early Drug Development, Hoffmann-La Roche Inc., 340 Kingsland Street, Nutley, NJ 07110, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
From a series of N-acyl 4-(3-pyridonyl)phenylalanine derivatives of 4, the trifluoromethyl derivative 28
was identified as a potent, dual acting alpha4 integrin antagonist with activity in primate models of aller-
gic asthma. Investigation of a series of prodrug esters led to the discovery of the morpholinopropyl deriv-
ative 48 that demonstrated good intestinal fluid stability, solubility and permeability. Compound 48 gave
high blood levels of 28 when dosed orally in cynomolgus monkeys. Surprisingly, hydrolysis of 48 was
rapid in liver microsomes from the pharmacological species, mouse, rat and monkey, but slow in dog
and human; in vivo studies also indicated there was prolonged exposure to unchanged prodrug in dogs.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 5 October 2012
Revised 30 November 2012
Accepted 10 December 2012
Available online 20 December 2012
Alpha4 integrins are expressed on a variety of leukocytes,
including B-cells, T-cells, basophils and eosinophils, and are in-
volved in the recruitment, activation and survival of these cell
types. Data supporting a role for alpha4 integrins in a number of
inflammatory diseases, including multiple sclerosis, inflammatory
bowel diseases, asthma, rheumatoid arthritis, and atherosclerosis,
have emerged and are summarized in recent reviews.1 A human-
ized mouse anti-alpha4 antibody, Natalizumab, has been approved
for the treatment of multiple sclerosis2 and Crohn’s disease.3 More
recently a small molecule alpha4 integrin antagonist was shown to
be effective for the treatment of MS,4 fully validating alpha4 inte-
grins as targets for human disease and increasing interest in the
discovery of small molecules with superior activities.
The compounds reported in Table 1 were typically prepared
by a palladium-catalyzed coupling reaction of the 4-iodophenyl-
alanine derivative 17 with an in situ generated organozinc
intermediate derived from the appropriate 4,6-disubstituted
3-iodo-pyridone intermediate such as 11 and 16 (Scheme 3).
Cl
OCH3
O
Cl HN
OCH3
OH
Cl HN
OH
Cl HN
O
O
O
We previously reported that the potent, dual-acting antagonist
O
Cl
of
man5 and sought additional compounds that might have lower
clearance and greater selectivity for 4b7 to further assess their
a4b1 and a4b7, 1, is effective for the treatment of asthma in
CH3
1
2
α4β1 IC50 10 nM
α4β7 IC50 43 nM
a
potential as alpha4 integrin antagonists. Inspired by reports of
the alpha4 integrin antagonist activity of members of the biphenyl-
alanine class such as 2,6 as well as the interesting, selective integrin
antagonist activity of members of the pyridizinone family repre-
sented by 3,7 we report here on a new series of potent alpha4 inte-
grin antagonists, N-acyl 4-(3-pyridonyl)-phenylalanine derivatives
4 and their orally active prodrug esters.
R1
N
CH3
R2
O
N
O
N
R3
OCH3
Cl HN
OH
OH
Cl HN
O
O
O
O
R4
⇑
Cl
Corresponding author.
E-mail address: tilleyjk@optonline.net (J.W. Tilley).
4
3
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.12.019

J. W. Tilley et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1036–1040
1037
Table 1
VCAM/VLA-4 binding inhibition of N-acyl 4-(3-pyridonyl)-L-phenylalanine derivatives
R2
R3
N
O
R4
OH
HN
R1
4b1 Ramos cells/VCAM IC50 nM (Rel Potency)a
O
O
No.
23
R1
R2
R3
H
R4
H
a
a
4b7 RPMI cells/MadCAM IC50 nM (Rel Potency)a
H3CO
CH3
419
ND
Br
Cl
Cl
Cl
Cl
24
25
26
CH3
CH3
CH3
CH3
CH3
H
H
H
H
H
582 (0.03)
687 (0.46)
CH3
CH3
Cl
CH3
CH3
226 (0.23)
90 (0.37)
363 (0.68)
179 (1.6)
27
28
29
30
31
32
33
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CF3
CF3
CF3
CF3
CF3
CF3
CF3
48 (0.47)
32 (0.81)
820 (0.08)
112 (0.40)
463 (0.03)
560 (0.12)
225 (0.41)
154 (3.5)
42 (6.5)
CH3
Cl
Cl
Cl
810 (0.65)
61 (1.9)
Cl
Cl
C2H5
Cl
Cl
1230 (0.34)
600 (0.96)
767 (0.70)
Cl
Cl
CH2–C5H6
Cl
CH3
CH3
C2H5
CH3
34
35
36
CH3
CH3
CH3
CH3
CH3
CF3
CF3
CF3
CH3
366 (0.36)
140 (0.75)
269 (0.11)
1,290 (0.23)
671 (0.54)
126 (2.2)
CH(CH3)2
CH3O
Cl
Cl
Cl
37
CH3
CF3
CH3
267 (0.06)
160 (2.6)
CH3
a
Rel potency refers to activity relative to the reference standard RO0270608 run as a positive control in each binding experiment.

1038
J. W. Tilley et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1036–1040
H
CH3
N
O
O
O
N
O
CH3
N
H3C
O
O
H3C
O
OMe
NH2
OMe
N
a
b
H3C
O
I
F
F
c
+
a, b
CO2Me
F
CF3
CF3
I
CF3
5
7
6
8
OMe
BocHN
11
CH3
N
O
H
N
OEt
H
N
BocHN
O
18
H3C
O
I
O
I
H3C
H3C
e
c
d
17
O
CH3
N
CO2H
H3C
O
R
CH3
N
CF3
CF3
CF3
H3C
O
9
10
11
d, e
CF3
Scheme 1. Reagents and conditions: (a) TMSCl, H2O, 0 °C, 15 h; (b) piperidine,
EtOH, reflux, 15 h; (c) LiOH, THF, MeOH, reflux, 48 h; (d) NaHCO3, I2, KI, MeOH, H2O,
65–70 °C, 15 h; (e) MeI, K2CO3, DME, rt, 15 h.
CF3
R
X
OH
HN
OMe
O
O
O
ClH.H2N
O
R1
X = OH or Cl
O
O
R1
F
F
O
a
19
F
F
F
F
F
+
20
O
O
F
F
O
Scheme 3. Reagents and conditions: (a) Zn dust (activated using 10 mol %
dibromoethane and 10 mol % TMSCl in THF), DMAC, 70 °C, 15 h; (b) Pd(dba)2, TFP,
THF, 50C, 15 h; (c) 4.0 N HCl in dioxane, dioxane, rt, 5 h; (d) X = OH, HBTU, DIPEA,
12
ꢀ
CH3
N
H
DMF, rt, 15 h or X = Cl, DIPEA, CH2Cl2, rt, 15 h; (e) 1.0 N NaOH, EtOH, rt, 5 h.
N
O
F3C
O
O
b
F3C
c
OEt
OEt
OEt
O
H2N
O
CH3
CH3
a, b
13
N
I
N
O
O
O
14
O
O
O
CH3
N
CH3
N
21
22
O
O
I
F3C
F3C
d
e
CH3
N
CH3
N
H3C
H3C
O
CH3
CH3
15
16
Scheme 2. Reagents and conditions: (a) pyridine, CH2Cl2, 0 °C to rt, 15 h; (b) NaOEt,
EtOH, reflux, 15 h; (c) MeI, K2CO3, DME, reflux, 15 h; (d) LiCl, DMF, 160C, 19 h; (e)
c. d
ꢀ
CF3
CF3
O
NIS, trifluoroacetic anhydride, TFA, 70–85 °C, 2 h.
OH
O
N
Cl HN
Cl HN
HCl
O
O
O
O
Cl
Cl
The 4,6-disubstituted 3-iodo-pyridone intermediates (11 and
16) were prepared using the process described in Schemes 1 and
2 while others were prepared using similar approach. In both, a
4,6-disubstituted pyridone core was built by a condensation be-
tween the mono ester malonamide and the diketone 7 or the re-
lated methoxy vinylketone 12 to give 8 and 13, respectively.
Treatment of the acid 9 with iodine/KI/NaHCO3 in aqueous metha-
nol at 65 °C led to a nearly quantitative conversion to the iodide 10,
which was then alkylated to give 11.
28
48
Scheme 4. Reagents and conditions: (a) LAH, THF, rt, 2 h; (b) NaI, MeSO3H, CH2Cl2,
reflux, 15 h; (c) NaHCO3, DMF, rt, 48 h; (d) TMSCl, PrOH, rt, 2 h.
i
The in vitro potency of the compounds listed in Table 1 was as-
sessed by determining the ability of serial dilutions to inhibit the
binding of RAMOS cells (
a4b1-specific binding) and RPMI 8866
cells ( 4b7-specific binding) with recombinant human VCAM-1
a
Surprisingly, the carboxylic acid derived from 13 did not provide the
corresponding iodide under similar conditions. Thus, 13 was converted
to the N-methyl intermediate 14 which was then treated with lithium
chloride in DMF at 160 °C to effect hydrolysis and decarboxylation fol-
lowed by treatment with NIS in trifluoroacetic acid in the presence of
trifluoroacetic anhydride to give 16 exclusively in 62% yield.
The iodide 11 was then reacted with an activated zinc dust in
dimethylacetamide at 70 °C to give the corresponding organozinc
intermediate.8 The coupling reaction of this organozinc intermedi-
ate with 17 proceeded smoothly in the presence of Pd(dba)2 and
trifurylphosphine (TFP) in THF at 50 °C to obtain the desired prod-
uct 18 in 77% isolated yield. The remaining steps, BOC removal,
benzoylation, and hydrolysis proceeded in a straightforward man-
ner to obtain the target compounds.
used as the counter ligand for the RAMOS cell assay and human
MAdCAM-1 for the RPMI 8866 cell assay. Cells were labeled with
Calcein AM, a fluorescent dye, and then activated with a binding
buffer containing Mn2+ to achieve maximum activation prior to as-
say. RO0270608 (1) was used as a positive control on each plate.
The data in Table 1 are reported both as the observed IC50 and as
fold-potency relative to 1 since the activity of the cell lines varied
over time with the number of passages.
Comparing the relative potencies of the compounds reported in
Table 1 in the cell based assay with the reference compound 1 indi-
cates that 28 is nearly equally potent to 1 against
a4b1 and is sev-
eral-fold more potent as an antagonist of 4b7. The data indicate
a
that R2 is preferably –CH3 and that R4 is preferentially trifluoro-
methyl group relative to –CH3 or H. These findings are consistent
with our previously reported observation with 4-benzoylcarbonyl-
aminophenylalanines related to 1 that an electron deficient distal
aromatic ring was favorable for potency9 and esptablished that
R1 is preferably a 2,6-dichlorobenzoyl group.10 Since 28 met our
In general, the alkyl ester prodrugs were prepared by treatment
of the carboxylic acids with the appropriate alkyl iodide or alkyl
bromide in the presence of sodium bicarbonate in DMF. For exam-
ple, the preparation of morpholinopropyl iodide and the corre-
sponding ester prodrug and its HCl salt is shown in Scheme 4.
criteria for potency and selectivity as an
a4b7 antagonist, it was

J. W. Tilley et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1036–1040
1039
Table 3
selected for further profiling. Intravenous PK experiments in rat
PK properties of compounds 28 and 48
(Table 3) indicated that 28 had an acceptable PK properties.
Previous experience with related compounds indicated that the
free acids in the acylphenylalanine class have very poor bioavail-
ability; thus, a prodrug strategy was pursued from the outset.
Compd
Rat
Dog
Monkey
28
Dose (iv) (mg/kg)
AUC (ng h/mL)
Cl (mL/min/kg)
Vss (mL/kg)
Dose (po) (mg/kg)
AUC of 28 (ng h/mL)
Cmax (mg/mL)
F (%)
10
10
15
35,100
3.0
67
30
30,150
5.0
46
10,860
15.7
320
50
6100
2.5
10,880
16.5
250
50
6420
2.4
The ethyl ester 38 is extremely insoluble in water (6 lg/mL) and
while bioavailability in rats was a modest 18%, it was only 2% in
dogs. Other candidate prodrugs were assessed for Caco-2 perme-
ability, solubility, and stability in simulated intestinal fluid, and
promising compounds were further profiled in rat PK (Tables 2
and 3). Most of the prodrugs have good stability and reasonable
conversion in plasma to the parent drug. Only the basic prodrugs
had good solubility and of these, the morpholinopropyl ester 48
emerged as the best candidate for further profiling in large animal
PK and efficacy experiments.
PK data for 48 in rat, dog and monkey are shown in Table 3.
Although oral bioavailability was only modest in rat and dog, the
relatively high bioavailability in monkey, combined with its good
solubility and permeability, prompted us to test this compound
further in the pivotal monkey lung inflammation model. This mod-
el was used to profile 1, which was validated clinically, and thus
became our major in vivo hurdle for taking compounds forward.
Thus cynomolgus monkeys were challenged with aerosolized
Ascaris suum extract 2 h after dosing. Inflammatory cell accumula-
tion in the lungs was assessed by examination of bronchiolar la-
vage fluid withdrawn 24 h after the antigen challenge. At oral
doses of 10 and 30 mg/kg po, 48 was efficacious in blocking influx
of eosinophils, lymphocytes, neutrophils and total leukocytes into
48
11
13
their lungs as shown in Figure 1. Based on its in vivo profile, 48
was selected for preclinical safety studies.
Surprisingly, although 48 was readily hydrolyzed in mouse, rat
and monkey liver microsomes, ester hydrolysis was slow in dog
and human. A graphic representation of the microsomal results
are shown in Figure 2. The significance of these findings was high-
lighted by our observation that there was about 10% unchanged
prodrug in dog plasma on days 1 and 6 during subchronic safety
studies. This gave rise to concerns about tissue distribution and
possible safety issues associated with significant exposure to the
unchanged prodrug, particularly as it could be anticipated to occur
in humans as well.
These concerns were validated when, metabolism studies of
radiolabeled 28 and 48 in liver microsomes from rat, dog, monkey
and human revealed that significant covalent protein binding oc-
curred with an unknown metabolite derived from the unchanged
Table 2
Prodrug esters of the lead compound 28
CH3
H3C
N
O
CF3
O R
HN
Cl
O
O
Cl
Compd
R
Caco-2 Papp
10À7 cm/s
Sol. SGFa
mL
lg/
Sol. SIFb
mL
lg/
Stability SIFb t1/2
min
hERG IC20
Bioavail rat F
(%)
Comment
lM
28
38
39
40
41
42
–H (Parent)
–CH2CH3
1.4
—
—
—
—
—
—
271
526
325
499
228
BQL
BQL
BQL
BQL
nd
BQL
BQL
BQL
BQL
nd
Stable
Stable
Stable
Stable
nd
2.7
nd
nd
nd
nd
18
nd
nd
nd
2
Dog F = 2%
–CH2CH2CH3
–CH2CH(CH3)2
–CH2CH2OCH3
–(CH2)3OCH3
O
43
44
45
63
0.8
0.7
Stable
Stable
Stable
nd
nd
nd
nd
nd
6
N(CH3)2
CH2
O
764
18
0.03
190
0.03
140
CH2
OH
Unchg prodrug in
plasma
OH
Unchg prodrug in
plasma
N
46
162
>13,000
200
240
11
15
O
N
47
48
49
96
>5000
3900
Decomp
30
11.7
2.9
12
11
17
O
165
275
330
4
Stable
Stable
N
O
CH3
Mixture of
diastereomers
O
a
SGF is simulated gastric fluid, pH 2.0.
SIF is simulated intestinal fluid, pH 7.5.
b

1040
J. W. Tilley et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1036–1040
the inhibition of a4b7. Compound 28 stood out among those inves-
Allergen-induced Inflammatory Cell Influx in the Atopic Primate
tigated and a series of prodrugs was investigated to find a suitable
derivative for oral delivery of the otherwise poorly absorbed car-
boxylic acid. Most compounds investigated were rejected due to
poor solubility, permeability or oral bioavailability and attention
was focused on the morpholinopropyl ester 48 as having the best
combination of in vitro properties and reasonable bioavailability,
particularly in monkeys. The unanticipated species selective
hydrolysis of 48 led to the abandonment of this compound as a
lead candidate and raises a significant cautionary flag for others
considering a prodrug approach to solving drug delivery issues.
Scientists interested in pursuing prodrug strategies should ver-
ify that any prodrugs they contemplate undergo sufficiently rapid
conversion in all species used in pharmacological and safety stud-
ies. A similar incidence of species-selective prodrug ester hydroly-
sis was observed in a series of structurally similar amino acid
derived beta2 integrin antagonists.11 Even though many esters un-
dergo rapid hydrolysis and are rarely employed as drugs per se for
this reason, their consistent hydrolysis cannot be taken as a given.
It would be interesting to investigate more systematically, how
common similar species effects occur. One could, for example
examine the time course of hydrolysis of a series of known prodrug
esters in liver microsomes and plasma from several species seeking
structural patterns. In a forthcoming paper, we will describe a suc-
cessful prodrug strategy with a series of closely realted pyrimi-
dine-diones in which this phenomena did not occur.12
35
Vehicle
10 mg/kg
30 mg/kg
30
25
20
15
10
5
0
MAC
NEU
EOS
LYM Total Cells
Cell Type
Figure 1. Accumulation of inflammatory cells in atopic non-human primate BAL
fluid 24 h after challenge with Acaris suum 2 h after oral treatment with 48.
100
90
80
70
60
50
40
30
20
10
0
References and notes
1. (a) Tilley, J. W. Expert Opin. Ther. Patents 2008, 18, 841; (b) Davenport, R. J.;
Munday, J. R. Drug Discovery Today 2007, 12, 569; (c) Bosserhoff, A.-K. Expert
Opin. Ther. Patents 2006, 16, 963.
2. (a) Steinman, L. Nat. Rev. Drug Disc. 2005, 4, 510; (b) O’Connor, P. Expert Opin.
Biol. Ther. 2007, 7, 123.
3. Ghosh, S.; Goldin, E.; Gordon, F. H., et al N. Engl. J. Med. 2003, 348, 24.
4. Miller, D. H.; Weber, T.; Grove, R., et al Lancet Neurol. 2012, 11, 131.
5. Tilley, J. W.; Sidduri, A.; Chen, L.; et al. Abstracts of papers, 234th ACS National
Meeting, Aug. 19–23, 2007; MEDI1250.
6. Sircar, I.; Gudmundsson, K. S.; Martin, R., et al Bioorg. Med. Chem. 2002, 10,
2051.
0
10
30
90
Time (min)
7. (a) Gong, Y.; Barbay, J. K.; Kimball, E. S., et al Bioorg. Med. Chem. Lett. 2008, 18,
1331; (b) Gong, Y.; Barbay, J. K.; Dyarkin, A. B., et al J. Med. Chem. 2006, 49,
3402.
Control
Dog
Mouse
Rat
Monkey
Human
8. Knochel, P.; Singer, R. D. Chem. Rev. 1993, 93, 2117.
9. Chen, L.; Tilley, J. W.; Guthrie, R. W.; Mennona, F.; Huang, T.-N.; Kaplan, G.;
Trilles, R.; Miklowski, D.; Huby, N.; Schwinge, V.; Wolitzky, B.; Rowan, K. Bioorg.
Med. Chem. Lett. 2000, 10, 729.
10. Sidduri, A.; Tilley, J. W.; Lou, J. P.; Chen, L.; Kaplan, G.; Mennona, F.; Campbell,
R.; Guthrie, R.; Huang, T.-N.; Rowan, K.; Schwinge, V.; Renzetti, L. M. Bioorg.
Med. Chem. Lett. 2002, 12, 2479.
Figure 2. Percent of ester hydrolysis of 48 versus time in liver microsomal
preparations from various species. Control refers to incubation of 48 in buffer with
no microsomes added.
11. (a) Khojasteh, S. C.; Leipold, D. D.; Lai, F.; La, H.; Baumgardner, M. J.; Desino, K.
E.; Gudmundsson, O. S.; Bloedow, D. C.; Dodary, S. C.; Reynolds, M. E.; Gadek, T.
R.; Kenkare-Mitra, S. Xenobiotica 2008, 38, 340; (b) Van Gelder, J. V.; Shafee, M.;
De Clercq, E.; Penninck, F.; Ven den Mooter, G.; Kinget, R.; Augustijns, P. Int. J.
Pharm. 2000, 205, 93.
prodrug, but not the active pyridone acid 28 in dog and human, but
not rat or monkey. It is likely that exposure to 48 in the latter was
too transient to observe this metabolite or that the metabolic path-
way simply was not operable in these species.
In the work presented in this manuscript, a series of pyridones
was prepared seeking alpha4 integrin antagonists with a bias for
12. Sidduri, A. Bioorg. Med. Chem. Lett. 2012, http://dx.doi.org/10.1016/
j.bmcl.2012.12.026.