A new class of 5-HT2B antagonists possesses favorable potency, selectivity, and rat pharmacokinetic properties

Article abstract of DOI:10.1016/j.bmcl.2009.02.126

We have been exploring the potential of 5-HT_2B antagonists as a therapy for chronic heart failure. To assess the potential of this therapeutic approach, we sought compounds possessing the following attributes: (a) potent and selective antagonis

Full text of DOI:10.1016/j.bmcl.2009.02.126

Bioorganic & Medicinal Chemistry Letters 19 (2009) 2206–2210
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A new class of 5-HT2B antagonists possesses favorable potency, selectivity,
and rat pharmacokinetic properties
a
a
b
c
b
c
Neil Moss a, , Younggi Choi , Derek Cogan , Adam Flegg , Andreas Kahrs , Pui Loke , Orietta Meyn ,
Raj Nagaraja a, Spencer Napier b, Ashley Parker b, J. Thomas Peterson a, Philip Ramsden a, Christopher Sarko a,
Donna Skow a, Josh Tomlinson b, Heather Tye b, Mark Whitaker b
*
a Departments of Medicinal Chemistry, Cardiovascular Disease, or Drug Discovery Support, Boehringer Ingelheim Pharmaceutical, Inc., 900 Ridgebury Road,
Ridgefield, CT 06877, United States
b Evotec (UK) Ltd 114 Milton Park, Abingdon, Oxfordshire OX14 4SA, UK
c Evotec AG Schnackenburgallee 114, 22525 Hamburg, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
We have been exploring the potential of 5-HT2B antagonists as a therapy for chronic heart failure. To
assess the potential of this therapeutic approach, we sought compounds possessing the following attri-
butes: (a) potent and selective antagonism of the 5-HT2B receptor, (b) low impact of serum proteins on
potency, and (c) desirable pharmacokinetic properties. This Letter describes our investigation of a biphe-
nyl benzimidazole class of compounds that resulted in 5-HT2B antagonists possessing the above attri-
butes. Improving potency in a human serum albumin shift assay proved to be the most significant SAR
discovery.
Received 3 February 2009
Revised 24 February 2009
Accepted 25 February 2009
Available online 9 March 2009
Keywords:
5-HT2B
Antagonist
Ó 2009 Elsevier Ltd. All rights reserved.
Several groups have explored antagonism of the 5-HT2B recep-
tor as a potential therapeutic for various indications.1 While origi-
nally functionally characterized in the rat stomach fundus,2
subsequent studies implicated 5-HT2B in vascular and gastrointes-
tinal contraction and relaxation. Consequently, 5-HT2B antagonism
has attracted attention for indications such as migraine,3 irritable
bowel syndrome,4 and pulmonary hypertension.5 Our interest in
5-HT2B antagonism stemmed from the growing body of data sup-
porting the potential of 5-HT2B antagonism for chronic heart fail-
ure, a disease typified by maladaptive remodeling of the heart
(hypertrophy).
The phenotype of 5-HT2B receptor knock-out mice demon-
strates the importance of this receptor for heart development. Sur-
viving mice possess underdeveloped hearts resulting from
impaired myocyte proliferation.6 Conversely, 5-HT2B over expres-
sion in mice leads to cardiac hypertrophy.7 Antagonists of 5-HT2B
have been reported to exhibit efficacy in models of cardiac hyper-
trophy.8–10 Genomics data from a model of tachypacing-induced
decompensatory heart failure in dogs showed an upregulation of
5-HT2B mRNA.11 An upregulation of 5-HT2B mRNA has also been re-
ported in humans with cardiomyopathies.12 However, while all of
this data supports the hypotheses of 5-HT2B antagonism being ben-
eficial for heart failure, to date no reports of 5-HT2B antagonist
showing efficacy in animal models of heart failure have appeared.
To explore the potential of 5-HT2B antagonism for the treatment
of chronic heart failure, we desired a class of compounds that at a
minimum would demonstrate potent and selective antagonism of
the 5-HT2B receptor and possess good pharmacokinetic properties.
In addition, we wanted to increase the likelihood of success by
favoring antagonists that could maintain potency in the presence
of physiological concentrations of serum proteins.13
We conducted a high throughput screening campaign to iden-
tify new classes of 5-HT2B antagonists. Compound 1, Table 1, rep-
resents one of the compound classes resulting from this effort. To
help gauge the attributes of compound 1, we compared it to se-
lected literature 5-HT2B compounds that had been reported to be
efficacious in pharmacological models. These benchmark com-
pounds included the SmithKline antagonist SB215505,14 the Roche
antagonist RS-127445,15 and a compound detailed in a patent from
Epix Pharmaceuticals referred to as compound A.16
Compound 1 possessed several attractive features as a lead. It
demonstrated potent antagonism of serotonin-induced intracellu-
lar Ca2+ flux in 5-HT2B receptor-transfected CHO-K1 cells measured
by aequorin luminescence. In this assay it compared essentially
equally with the benchmark compounds. Compound 1 proved
selective for the 5-HT2B receptor, showing no antagonism up to
10
It compared equally to the most selective of the three benchmark
l
M of 5HT2A or 5HT2C in analogous FLIPR based cell assays.17
* Corresponding author. Tel.: +1 203 798 5101; fax: +1 203 791 6072.
E-mail address: nmoss@rdg.boehringer-ingelheim.com (N. Moss).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.02.126

N. Moss et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2206–2210
2207
Table 1
Comparison of compound 1 with literature benchmarks
CN
F
Cl
N
N
O
N
HN
N
N
H
F
N
N
N
H
N
N
N
Cl
NH2
S
O
1
SB215505
RS-127445
Epix Comp A
5-HT2B IC50 (nM)a
5-HT2B 4% HSA IC50 (nM)a
5-HT2A Selectivity (fold)
5-HT2C Selectivity (fold)
Rat PK %F
1 mg/kg iv CL (%Q)
10 mg/kg po oral
AUC (ng-h/mL)
2.4
1.3
64
7
0.5
—
0.7
19
3.5
1200
>10,000
>10,000
130b
13
25,000
270
>3000
>3000
17d
82
500
2000
92
10c
29
—
810
oral Cmax (ng/mL)
5000
470
240
a
Values are means of 2–21 experiments, standard deviations for n > 2 experiments 50% of reported value.
Male Sprague Dawley rats dosed 70/30 PEG400/water.
Male Wistar rats dosed 30/70 Cremophor/aq citric acid.
b
c
d
Male Wistar rats dosed 70/30 PEG400/water.
compounds, the Epix compound A. One of the more distinguishing
group had minimal effect on potency (compound 11). Modifica-
tions at the para position (compounds 12 and 13) additionally de-
fined limitations and sensitivity at this region of the molecule.
Though the data is not shown in Table 2, it is noteworthy that
the terminal phenyl ring tolerate the presence of a 2, 3, or 4 fluoro
(IC50 1.5, 2.7; 2.5 nM), chloro (IC50 5.5, 4.2; 4.4 nM), or methoxy
(IC50 15, 15; 3.4 nM) substituent.
While the presence of heteroatoms in the terminal ring of com-
pound 1 lowered potency, the central ring appeared to better tol-
erate the presence of nitrogen atoms (Table 3). Compounds 15–
19 demonstrate some modest diversity in polar functionality.
While the decreased lipophilicity of these compounds resulted in
improved aqueous solubility over compound 1, it did not provide
a significant improvement in potency in the presence of HSA. The
limited examples in Table 3 do not provide a lot more information
on the likely role of the central phenyl ring in binding to the 5-HT2B
receptor. The substantial decrease in potency observed on saturat-
ing this ring (compound 14) at least highlights the importance of
the central phenyl for binding to the receptor and/or appropriately
positioning the other pharmacophores.
positive attributes of compound 1 is its favorable rat pharmacoki-
netic profile. Compound 1 demonstrated superior bioavailability,
clearance, and oral exposure over both the Epix and RS compounds.
To assess the impact of protein binding on potency, we ran the
5-HT2B aequorin assay in the presence 4% human serum albumin
(HSA) and discovered the potency of compound 1 dropped 500-
fold.18 This was clearly the most prominent negative attribute of
compound 1 compared to the three benchmark compounds.19 Con-
sequently, in exploring the SAR and potential of this new class of
5-HT2B antagonists, we desired compounds that maintained the
positive potency, selectivity, and pharmacokinetic properties of
compound 1 but possessed greater potency in the presence of HSA.
Our SAR investigation commenced with the goal of initially
establishing the key pharmacophores for potency. We hoped this
investigation would also identify areas in which modifications
could be introduced to address the high serum protein shift. As in-
creases in lipophilicity of compounds can correlate with increases
in protein binding,20 attention was placed on identifying modifica-
tions that could reduce lipophilicity.
The compounds listed in Tables 1 and 2 provide a general pic-
ture of the pharmacophore component of the biphenyl unit of com-
pound 1 and highlight some opportunities for introducing polar
functionality.21 The majority of compounds in these and subse-
quent tables were prepared in a straightforward manner according
the general reaction sequences depicted in Scheme 1.22
The lack of potency of compounds 2 and 3 in Table 2 demon-
strate the importance of the terminal phenyl group for activity.
Introduction of a nitrogen atom into the terminal phenyl reduces
potency. This reduction becomes more prominent as the nitrogen
moves from the ortho to the meta and para positions (cf. com-
pounds 4–6 to compound 1). Although compound 6 has somewhat
weaker potency than compound 1, the high throughput aqueous
A comparison of compound 1 to compounds 20 and 21 (Table 4)
suggests that the carbonyl group may not be a critical element for
potency. While the reduced amide derivative 21 is less potent, re-
moval of the carbonyl introduces substantial conformational flexi-
bility plus introduces a basic amine functionality. The decrease in
potency observed for the piperazine derivative 22 and the azeti-
dine derivative 23 suggested the need to appropriately position
the benzimidazole functionality relative to the rest of the molecule.
At the initiation of our SAR studies, we hypothesized that the
benzimidazole would turn out to be a key pharmacophore for
binding to the 5-HT2B receptor. The substantial decrease in potency
observed upon methylating the benzimidazole N–H (compound
24) or even more so upon replacing it with oxygen (compound
25) supported this belief. We subsequently initiated synthesis of
derivatives containing modifications to the phenyl portion of the
benzimidazole. However, additions such as a 5-methoxy or 5-CON-
HCH3 (IC50 51 and 4200 nM, respectively) or replacement with a 4
or 5 aza-benzimidazole (IC50 210 and 300 nM, respectively) did not
look encouraging. Eventually we tested the essentiality of the phe-
nyl portion of the benzimidazole by preparing the corresponding
imidazole analog 26. To our surprise not only did 26 retain potency
for 5-HT2B, it improved potency approximately 15-fold in the HSA
solubility of compound 6 at pH 7.4 improves to 63
lg/mL from
3
l
g/mL for compound 1. Unfortunately the introduction of two
ortho nitrogens (compound 7) resulted in a further loss of potency.
The terminal phenyl in compound 1 can be saturated to cyclo-
hexyl (compound 8) with a modest 13-fold reduction of potency.
A comparison of the structurally related piperidine analog 9 to
the morpholine derivative 10 further emphasized the undesirabil-
ity of a heteroatom in the terminal ring. However, the presence of
an oxygen atom at the para position in the form of a methoxy

2208
N. Moss et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2206–2210
Table 2
Terminal phenyl SAR
H2N
H2N
N
N
n
n
CO2H
n
a, b
HN
HN
N
H
BOCN
X
Y
N
n = 0 or 1
R
N
H
c, b
CHO
n
N
N
H
BOCN
O
Compd
1
R
5-HT2B IC50 (nM)a
5-HT2B 4% HSA IC50 (nM)a
X or Y
R
1-6, 8-12, 14,16,18,19, 26, 29, 33, 35
13, 15, 17
CO2H
d
2.4
1200
>10,000
>26,000
>13,000
>14,000
1700
e
3
H
Ar
2
3
>10,000
>13,000
>1300
340
X or Y
20, 30-32, 34
Cl
f
S
Br
N
O
O
Scheme 1. Reagents and conditions: (a) P(OPh)3, pyridine, 125 °C, 18 h, 40–50%; or
4 M HCl, water, 130 °C, 2 days, 38%; (b) 2 M HCl, ether, 100%; (c) 40% glyoxal, water,
NH3(g), MeOH, 47–50%; (d) EDC, DMAP, CH2Cl2 or ClCH2CH2Cl; or EDC, HOBt, Et3 N,
CH2Cl2, 2–96%; (e) ArB(OH)2, Pd(dppf)Cl2, 2 M aq NaHCO3, MeOH–dioxane, 7–80%;
(f) iPr2EtN or Et3N, CH2Cl2.
4
5
N
benzimidazole analog (cf. compounds 1 and 29) although the ser-
um shift potency did improve.
6
15
N
N
Given the similar potency of the carbonyl and sulfonamide ana-
logs 1 and 20, we examined the effect of the new imidazole mod-
ifications on antagonists possessing a sulfonamide group (Table 6).
We saw an equivalent effect between the carbonyl and sulfon-
amide series for replacing a directly attached benzimidazole with
an imidazole (cf. compounds 1 to 26 with compounds 20 to 30).
In contrast, the sulfonamide series lost potency relative to the car-
bonyl series when the imidazole/benzimidazole group was at-
N
N
7
120
>2700
8
31
1100
9
48
12,000
>26,000
3200
tached via
a methylene (cf. compounds 26 to 28 with
compounds 30 to 31; compounds 1 to 29 with compounds 20 to
32). These latter results reemphasized the subtleties influencing
optimal imidazole orientation.
O
10
11
12
13
2300
3.4
N
The imidazole compounds highlighted in Tables 5 and 6 show
improved potency in the presence of HSA into the range of the best
literature benchmark. Table 7 completes the initial profiling of this
new series of 5-HT2B antagonists. All three antagonists possess
favorable rat pharmacokinetic properties. The imidazole analog
33 provides a combination of oral bioavailability, iv clearance
and oral exposure comparable to the original lead compound.
Interestingly, the sulfonamide analog 34 possesses a weaker phar-
macokinetic profile compared to the corresponding amide analog.
Overall, compound 35 appears to have the best combination of po-
tency and rat pharmacokinetic properties.
Part of our SAR strategy involved identifying ways to reduce
lipophilicity with the hope of improving potency in the HSA shift
assay. The greatest impact came from replacing the benzimidazole
with an imidazole. The difference in protein shift between com-
pounds 1 and 26 is approximately 15-fold. The magnitude of this
shift is lower when comparing compounds 28 to 29 and com-
pounds 20 to 30 (both approximately fourfold). However, we ob-
tained data on eight additional benzimidazole imidazole pairs
containing modifications to the biphenyl ring beyond those intro-
duced in Tables 1 and 2 (compounds and data not shown). The
11 pairs in total returned an average shift of 7.4 3.9-fold. It is
noteworthy that the reduction in lipophilicity as calculated by
clogP, AlogP, or xlogP in going from a benzimidazole to an imidaz-
O
O
O
F3C
95
16,000
>24,000
H
2N
>2400
Effect of R group on potency
a
Values are means of typically three experiments, standard deviations for n > 2
experiments typically 50% of reported value.
shift assay. The imidazole appears to contain the critical pharmaco-
phore for binding as judged by the drop in potency of the corre-
sponding amide analog 27.
The SAR in Table 4 suggested that potency was sensitive to posi-
tioning of the imidazole portion of the antagonist. Thus we were
further surprised when we discovered that inserting a methylene
group between the imidazole and the piperidine ring increased po-
tency (cf. compound 26 and 28, Table 5). This increase in potency
also translated to the HSA shift assay. The effect of the extra meth-
ylene group appeared to be less pronounced for the corresponding
ole (
D
logP = 1.45–1.63 depending on method) is slightly less than
logP = 1.68–2.21, i.e.,
going from a benzene to a pyrimidine (
D

N. Moss et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2206–2210
2209
Table 3
Table 5
Central Phenyl SAR
Benzimidazole SAR
N
R
N
H
N
X
N
O
O
Compd
R
5-HT2B IC50 (nM)a
5-HT2B 4% HSA IC50 (nM)a
Compd
X
5-HT2B IC50 (nM)a
5-HT2B 4% HSA IC50 (nM)a
N
N
1
2.4
1200
1
2.4
1200
N
H
24
80
10,000
14
15
16
610
12
>20,000
>2400
710
N
O
N
N
25
26
27
28
29
1,500
2.4
>20,000
76
N
N
2.5
N
H
N
N
O
O
>6500
0.4
—
17
4
1200
NH2
N
N
O
O
14
N
H
18
19
6
910
N
H
N
2.3
300
N
H
22
>3000
Effect of R group on potency.
a
Effect of X group on potency.
See footnote for Table 2.
a
See footnote for Table 2.
Table 4
Piperidyl carbonyl SAR
Table 6
Sulfonamide antagonists
N
N
R
X
H
N
S
Compd
X
5-HT2B IC50 (nM)a
5-HT2B 4% HSA IC50 (nM)a
O
O
Compd
20
R
5-HT2B IC50 (nM)a
5-HT2B 4% HSA IC50 (nM)a
1
2.4
1200
N
O
S
N
N
1.9
840
20
21
22
23
1.9
840
N
O
N
H
O
30
0.3
32
65
7300
N
H
N
N
N
N
N
31
32
6.5
990
6900
150
>25,000
>2800
N
H
O
O
N
1100
>23,000
N
H
Effect of R group on potency.
Effect of X group on potency.
a
a
See footnote for Table 2.
See footnote for Table 2.

2210
N. Moss et al. / Bioorg. Med. Chem. Lett. 19 (2009) 2206–2210
Table 7
Potency and pharmacokinetic data for top compounds
F
N
F
F
N
N
N
H
N
H
N
HN
N
N
S
O
O
O
O
33
34
35
5-HT2B IC50 (nM)a
1.1
35
100
17
14,000
1700
0.6
30
54
60
2300
870
0.2
7
118
29
6500
600
5-HT2B 4% HSA IC50 (nM)a
Rat PKb %F
1 mg/kg iv CL (%Q)
10 mg/kg po oral AUC (ng h/mL)
Oral Cmax (ng/mL)
a
See footnote for Table 2.
Male Sprague Dawley rats dosed 70/30 PEG400/water.
b
6. Nebigil, C. G.; Nickel, P.; Messanddeq, N.; Vonesch, J.-L.; Douchet, M. P.;
Monassier, ; Gyorgy, K.; Matz, R.; Andriantsitohaina, R.; Manivet, P.; Launay, J.-
M.; Maroteaux, L. Circulation 2001, 103, 2973.
7. Nebigil, C. G.; Jaffré, F.; Messanddeq, N.; Nickel, P.; Maroteaux; Monassier, L.;
Launay, J.-M.; Maroteaux, L. Circulation 2003, 107, 3223.
8. Jaffré, F.; Callebert, J.; Sarre, A.; Etienne, N.; Nebigil, C. G.; Launay, J.-M.;
Maroteaux, L.; Monassier, L. Circulation 2004, 110, 969.
9. Liang, Y.-J.; Lai, L.-P.; Wang, B.-W.; Juang, S.-J.; Chang, C.-M.; Leu, J.-G.; Shyu, K.-
G. Cardiovasc. Res. 2006, 72, 303.
10. Monassier, L.; Laplante, M.-A.; Jaffré, F.; Bousquet, P.; Maroteaux, L.; de
Champlain, J. Hypertension 2008, 52, 301.
11. Ojaimi, C.; Qanud, K.; Hintze, T. H.; Recchia, F. A. Physiol. Genomics 2007, 29, 76.
12. Oxford, A. W.; Borman, R. A.; Coleman, R. A.; Clark, K. L.; Hynd, G.; Archer, J. A.;
Aley, A.; Harris, N. V.; Goulter, A. USPTO Application 20050176791.
13. Trainor, G. L. Expert Opin. Drug Discovery 2007, 2, 51.
14. Reavill, C.; Kettle, A.; Holland, V.; Riley, G.; Blackburn, T. P. Br. J. Pharmacol.
1999, 126, 572.
15. Bonhaus, D. W.; Flippen, L. A.; Greenhouse, R. J.; Jaime, S.; Rocha, C.; Dawson,
M.; Van Natta, K.; Chang, L. K.; Pulido-Rios, T.; Webber, A.; Leung, E.; Eglen, R.
M.; Martin, G. R. Br. J. Pharmacol. 1999, 127, 1075.
compound 1 to 16). Similarly, HSA binding calculations (LogKHSA
)
predict the phenyl to pyrimidine change would lead to a slightly
greater reduction in binding to HSA than the benzimidazole to
imidazole change.23 Despite the predictions from calculated phys-
icochemical properties, introducing polar functionality had less of
an impact on potency in our HSA assay than removing a phenyl
ring. We measured equilibrium dialysis plasma protein binding
for compounds 1, 16, and 26 (99.9, 99.1, and 99.9, respectively).
Interestingly, despite all being very high, these values do not match
the trends from either the HSA shift assay or the predictions from
calculated physicochemical properties. There are of course caveats
with interpreting each of these three methods for ranking protein
binding, not the least of which is the role of plasma components
other than HSA. Nonetheless, we hypothesize that the conse-
quences of relative binding to serum protein in a functional cell as-
say will be more predictive of impact in vivo. The veracity of this
hypothesis requires an in-depth analysis of pharmacological effi-
cacy and associated pharmacokinetics which is beyond the scope
of this publication. Nevertheless we felt it instructive to highlight
some results involving structure–protein binding relationships.
In summary we have discovered a new series of 5-HT2B antag-
onists that maintain good potency in the presence of HSA and pos-
sess favorable pharmacokinetic properties in rats. These
compounds provide a better balance of potency, selectivity and
PK than currently published literature benchmarks. Consequently
these compounds should make superior tools for assessing the po-
tential of 5-HT2B antagonists for chronic heart failure.
16. Dhanoa, D. S.; Becker, O.; Noiman, S.; Alla, S. R.; Melendez, R. E.; Sharadendu,
A.; Chen, D.; Marantz, Y.; Shacham, S.; Heifetz, A.; Inbal, B.; Kesavan, V.; Bar-
Haim, S.; Cheruka, S. R. WO Patent Application 2006034511, 2006.
17. Compound 1 returned <65% inhibition at 10
conducted at MDS Pharma Services. Dopamine D4.2 was the one receptor
showing significant inhibition (99% at 10 M).
lM of a panel of 40 receptors
l
18. We chose to use a serum shift assay to estimate the impact of protein binding
because standard equilibrium dialysis experiments to determine protein
binding are not high throughput and are not sensitive enough to distinguish
between highly protein bound compounds. Rusnak, D. W.; Lai, Z.; Lansing, T. J.;
Rhodes, N.; Gilmer, T. M.; Copeland, R. A. Bioorg. Med. Chem. Lett. 2004, 14,
2309. Using 4% HSA proved convenient as an estimate for the effect of serum
components, since the use of serum in the assay stimulated calcium release
presumably due to endogenous serotonin.
19. The plasma protein binding for compound 1, RS-127445, and compound A were
determined by equilibrium dialysis to be 99.9%, 99.9%, and 99%, respectively.
20. Gleeson, M. P. J. Med. Chem. 2007, 50, 101.
21. All compounds in Tables 2–6 did not inhibit either 5-HT2A/C at concentrations
References and notes
up to ꢀ10
lM.
22. Most of the biaryl carboxylic acids are commercially available. So are the amine
pieces in compounds 22, 25 and 27 which were coupled to biphenyl carboxylic
acid by the options in Scheme 1. The azetidine fragment in compound 23 was
prepared in the same manner as compound X.
1. Poissonnet, G.; Parmentier, J. G.; Boutin, J. A.; Goldstein, S. Mini-Rev. Med. Chem.
2004, 4, 325.
2. (a) Clineschmidt, B. V.; Reiss, D. R.; Pettibone, D. J.; Robinson, J. L. J. Pharmacol.
Exp. Ther. 1985, 235, 696; (b) Baez, M.; Yu, L.; Cohen, M. L. Mol. Pharmacol. 1990,
38, 31.
3. Johnson, K. W.; Nelson, D. L.; Dieckman, D. K.; Wainscott, D. B.; Lucaites, V. L.;
Audia, J. E.; Owton, W. M.; Phebus, L. A. Cephalalgia 2003, 2, 117.
4. Borman, R. A.; Tilford, N. S.; Harmer, D. W.; Day, N.; Ellis, E. S.; Sheldrick, R. L.
G.; Carey, J.; Coleman, R. A.; Baxter, G. S. Br. J. Pharmacol. 2002, 135, 1144.
5. Esteve, J. M.; Launay, J.-M.; Kellermann, O.; Maroteaux, L. Cell Biochem. Biophys.
2007, 47, 33.
23. These calculations were based on a QSAR model implemented in the program
QIKPROP that was developed from high-performance affinity chromatography
determinations of binding affinities of 95 diverse drugs and druglike compounds
to HSA. See Colmenarejo, G.
J . Med. Chem. 2001, 44, 4370–4378; also
Colmenarejo, G. Med. Res. Rev. 2003, 23, 275–301. LogKHSA values for the
training set range from À2.69 to +1.37 for low HSA to high HSA binding. The
values for compounds 1, 16, 26 were calculated to be +1.01, +0.62, and +0.66
respectively.