Quinoline-carboxylic acids are potent inhibitors that inhibit the binding of insulin-like growth factor (IGF) to IGF-binding proteins

Article abstract of DOI:10.1016/S0960-894X(03)00322-6

4-Benzylquinolines 5, based on a series of isoquinolines 1, were prepared and tested as inhibitors of the IGF/IGFBP-3 complex based on their ability to displace IGF-I from its binding to IGF-binding protein-3. SAR studies on the 6,7-dihydroxy moiety of the quinoline 5a showed that the catecol moiety could be replaced with other functional groups. Computational modeling of the 5a/mini-IGFBP-5 complex revealed the possible binding site of 5a on IGFBP-5.

Full text of DOI:10.1016/S0960-894X(03)00322-6

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1931–1934
Quinoline-Carboxylic Acids are Potent Inhibitors that Inhibit
the Binding of Insulin-Like Growth Factor (IGF)to
IGF-Binding Proteins
Yun-Fei Zhu, Xiao-Chuan Wang, Patrick Connors, Keith Wilcoxen,
Yinghong Gao, Raymond Gross, Nathalie Strack, Timothy Gross,
James R. McCarthy, Qiu Xie, Nicholas Ling and Chen Chen*
Neurocrine Biosciences, Inc., 10555 Science Center Drive, San Diego, CA 92121, USA
Received 29 October 2002; accepted 11 February 2003
Abstract—4-Benzylquinolines 5, based on a series of isoquinolines 1, were prepared and tested as inhibitors of the IGF/IGFBP-3
complex based on their ability to displace IGF-I from its binding to IGF-binding protein-3. SAR studies on the 6,7-dihydroxy
moiety of the quinoline 5a showed that the catecol moiety could be replaced with other functional groups. Computational modeling
of the 5a/mini-IGFBP-5 complex revealed the possible binding site of 5a on IGFBP-5.
# 2003 Elsevier Science Ltd. All rights reserved.
The small polypeptide hormones insulin-like growth
factors (IGF-I and II) regulate cell proliferation, cell
differentiation, cell death, and cell metabolic activities.1
The mitogenic and metabolic actions of the IGFs are
mediated by their binding and activation of the cell
surface IGF-I receptor, a a2b2 heterotetramer closely
related to the insulin receptor.2 However, the concen-
tration of freely circulating IGF in blood and interstitial
fluids, including the cerebrospinal fluid, is exceedingly
low because most of the IGFs are bound to one or more
of six high affinity IGF-binding proteins (the major
binding protein in circulation is IGFBP-3), which inhi-
bit their interaction with the IGF-I receptor.3
Notable discoveries within the isoquinoline-3-carboxlyic
acid class of IGFBP inhibitors included the potency-
enhancing 3-carboxylic group and the 1-(3,4-dihydroxy-
benzoyl) substituents. Methylation of one or more of
the hydroxy groups of compound 1a to replace meta-
bolically labile catechol group resulted in less active
analogues.5 In this paper we report the synthesis, SAR
and computational docking studies on a series of quin-
oline analogues, represented by 5a (Fig. 1), as potent
IGFBP inhibitors.
IGFBPs are proteins of 219–289 residues, with mature
IGFBP-5 consisting of 252 residues and all IGFBPs
share a common domain organization.6 Although the
three-dimensional structures of IGF-I and IGF-II are
known,7 there is little structural information on any
protein of the IGFBP family. The only available infor-
mation on IGFBP-3 is a domain resolution model for
the IGF binding to IGFBP-3.8 Very recently, the only
high-affinity binding site of the entire IGF-BP5 protein
for IGFs was identified to reside between residues 40–92
(mini-IGFBP-5).9 The crystal structure of IGF-I com-
plex with mini-IGF-BP5 has been solved.10
Previous communications from our laboratory reported
the discovery and SAR of isoquinoline-3-carboxylic
acids and 3-hydroxy isoquinolines as potent non-pepti-
dyl IGFBP inhibitors (Fig. 1).4 Isoquinolines 1 exhibit
excellent in vitro potency in both binding and functional
assays.4 This is an alternative approach to develop an
orally active compound that can potentiate the action of
endogenous IGF-I by displacing the bound IGF-I from
the large pool of IGF/IGFBP complexes in the body
fluids.
The IGF-I binding sites of the IGFBP-5 and IGFBP-3
are nearly identical (Fig. 2) with only two residues lining
the binding cavity being different.11 Val49/Ala55 in
IGFBP-5 are equivalent to Ile57/Gly63 in IGFBP-3.
*Corresponding author. Tel.:+1-858-658-7600; fax:+1-858-658-7619;
e-mail: cchen@neurocrine.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00322-6

1932
Y.-F. Zhu et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1931–1934
Figure 1. IGFBP-3 Inhibitors.
Figure 2. The amino acid sequences of the IGF-binding region of hIGFBP-5 and hIGFBP-3. The boxed residues in hIGFBP-5 and the corre-
sponding sites in hIGFBP-3.
Since compound 5a and its close analogues also bind to
IGFBP-5 with high potency (Ki for IGFBP-5: 13 nM,
for IGFBP-3: 8 nM), we selected the crystal structure of
the ligand binding domain of IGFBP-5 to dock com-
pound 5a. The energy-minimized conformation of 5a
was manually docked into the IGF-I binding cavity of
mini-IGFBP-5 after IGF-I was removed from the crystal
complex. As shown in Figure 3 of the final mini-IGFBP-
5–5a complex, which resulted from the computational
study using Docking and Affinity software provided by
Accelrys.12 Four amino acid residues from the IGFBP-5
are identified that may potentially involve in key inter-
actions with 5a. Those residues are Arg59, Arg53,
Leu73, and Val49 (highlighted by colored balls, whereas
5a is shown in the style of colored ball and stick, oxygen
atoms are colored in red, nitrogen in blue, and carbon
in green, respectively).
4-Benzylquinolines bearing a 2-carboxylic acid (5) were
synthesized by cyclization of a substituted aniline 2 with
acetylenedicarboxylate under heating conditions, and
the resulting hydroxylquinolines were then converted
to the corresponding 4-chloroquinolines 3. Replace-
ment of the chlorine in
3 with 3,4-dimethoxy-
phenylacetonitrile promoted by sodium hydride in
THF gave quinolines 4.13 Acidic hydrolysis of 4 with
48% HBr resulted in saponification of the methyl
ester, hydrolysis and decarboxylation of the nitrile,
and deprotection of the methoxy groups to afford
benzylquinolines 5 (Scheme 1).
Figure 3. Compound 5a binding to the mini-IGFBP-5. The four residues (Val49, Arg53, Arg59, Leu73) that contacted with 5a are highlighted.

Y.-F. Zhu et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1931–1934
1933
Scheme 1. Reagents and conditions: (a) (i) MeO2CCꢀCCO2Me/heat; (ii) POCl3/reflux; (b) 3,4-(MeO)2PhCH2CN/NaH/THF; (c) 48% HBr/reflux.
Table 1. SAR of substituted quinolines 5
Compd
R1
R2
Ki (nM)
13
>10,000
110
53
5a
5b
5c
5d
5e
5f
OH
H
COOH
F
Cl
OH
COOH
OH
OH
OH
OH
240
52
5300
NH2
NH2
5g
OMe
There is no obvious interaction of the 2-nitrogen of the
isoquinoline 5a with the IGFBP-5 protein. That may
explain why quinoline 5a had very similar activity as its
isomeric isoquinoline 1b (1b, Ki=7.2 nM, 5a, Ki=13
nM). From this binding model, there is clearly a charge-
charge interaction between the 3-carboxlyic group of
isoquinoline 1 or quinoline 5, with the Arg53 residue of
the IGFBP-5. The 6,7-dihydroxy moiety of the iso-
quinoline 5 is proposed to interact with another basic
residue, Arg59. Arg53 and Arg59 have strong interac-
tion with Asp20 and Glu58 of IGF-I, respectively, in the
IGF-I/mini-IGFBP-5 complex. Attempts to replace this
catechol moiety with a carboxylic acid at the 6-position
of the quinoline core resulted in a totally inactive ana-
logue (5b). It is rather difficult to interpret this result
since the arginine residue is located on the surface of the
IGFBP-5 protein and should be flexible. Replacing the
7-hydroxy group with a carboxylic acid reduced binding
activity (5c), possibly because of a steric clash with
Arg59. Interestingly, the 6-hydroxy-7-fluoro compound
5d was more active than its 7-chloro analogue 5e (Ki 53
nM for 5d; 240 nM for 5e). In these two cases, the
acidity of the 6-hydroxy group seems to play an
important role, because the fluoro analogue should be
more acidic due to high electron-withdrawing effect of
the fluorine atom. Finally, the 7-amino compound 5f
only lost about 4-fold binding affinity (Ki=52 nM), but
its 6-methoxy analogue was much less active (5g,
Ki=5300 nM). This result again emphasizes the
importance of an acidic hydroxy group in this region
as we had previously reported in the isoquinoline series
that replacement of the 6,7-dihydroxy moiety with a
Figure 4. Illustration of the interaction of compound 5a with IGFBP-5.
6-methylsulfonamido-7-hydroxy group resulted in a
much less active compound (Table 1).14
Based on our docking model, the benzyl group of 5a
interacts with several lipophilic residues including Val49
and Leu73 of IGFBP-5 (Fig. 4), depending on the size
and shape of the substituted phenyl group. A bigger
group such as 3-phenoxyphenyl (see previous paper)
may also interact with Pro62 and Leu70 based on this
model. The Thr51 residue may have a hydrogen-bond-
ing interaction with the benzoyl carbonyl group of 1a.
This may explain why some benzyl alcohols were more
potent inhibitors.12 This binding cavity is the host of
Phe16 of IGF-I based on crystal structure.8 One can
then predict that replacement of the benzyl with a sui-
table aliphatic group may also generate potent IGFBP
antagonists.
In conclusion, we have discovered that quinolines 5
bearing a carboxylic acid are potent inhibitors of IGF-I
binding to IGFBP-3 and IGFBP-5. While the catechol
moiety on the quinoline ring seems to be important in
binding, it could be mimicked with other functional
group to generate potent but more stable analogues.
The 3,4-dihydroxybenzyl group should be replaced by

1934
Y.-F. Zhu et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1931–1934
different substituted benzyl moieties, possibly an ali-
phatic group, as we observed in the isoquinone series.5
This discovery may help to design compounds with
better physicochemical properties. Lastly, the binding
model may guide the design of potent inhibitors with
selectivity among the six different IGF-binding proteins.
8. S penser, E. M.; Chan, K.Prog. Growth Factor Res. 1995, 6, 209.
9. Kalus, W.; Zweckstetter, M.; Renner, C.; Grol, M.;
Demuth, D.; Schumacher, R.; Dony, C.; Lang, K.; Holak,
T. A. EMBO J. 1998, 17, 6558.
10. Zestawski, W.; Beisel, H.-G.; Kamionka, M.; Kalus, W.;
Engh, R. A.; Huber, R.; Lang, K.; Holak, T. A. EMBO J.
2001, 20, 3638.
11. Hong, J.; Zhang, G.; Dong, F.; Rechler, M. M. J. Biol.
Chem. 2002, 277, 10489.
References and Notes
12. Docking and Affinity molecular modeling software are
provided by Accelrys.15 In docking, the interaction energy is
computed by summing the energy contributions between all
atoms of the two molecules. The objective of a docking type
calculation is to evaluate the interaction energies of many
orientations of one molecule relative to the other, while
searching for the orientations that result in low interaction
energies. A more advanced simulation module Affinity is
employed that automatically moves the small molecule, eval-
uates energies, and checks if the structure is acceptable by
energy evaluation. In our simulation, the small molecule and
the side chains of mini-IGFBP are flexible during the search,
whereas the backbone of IGFBP is fixed.
1. (a) For recent reviews, see: Roith, D. L.; Bondy, C.; Yakar,
S.; Liu, J.; Butler, A. Endocr. Rev. 2001, 22, 53. (b) Jones, J. I.;
Clemmons, D. R. Endocr. Rev. 1995, 16, 3.
2. Ullrich, A.; Gray, A.; Tam, A. W.; Yang-Fang, T.; Tsubo-
kawa, M.; Collins, C.; Henzel, W.; Le Bon, T.; Kathuria, S.;
Chen, E.; Jacobs, S.; Francke, U.; Ramachandran, J.; Fujita-
Yamaguchi, Y. EMBO J. 1986, 5, 2503.
3. (a) Shimasaki, S.; Ling, N. Progress in Growth Factor Res.
1992, 3, 243. (b) Shimasaki, S.; Shimonaka, M.; Zhang, H.;
Ling, N. J. Biol. Chem. 1991, 266, 10646.
4. Liu, X.-J.; Xie, Q.; Zhu, Y.-F.; Chen, C.; Ling, N. J. Biol.
Chem. 2001, 276, 32419.
5. Chen, C.; Zhu, Y.-F.; Liu, X.-J.; Lu, Z.-X.; Xie, Q.; Ling,
N. J. Med. Chem. 2001, 44, 4001.
6. Rajaram, S.; Baylink, D. J.; Mohan, S. J. Endocrinol. Rev.
1997, 18, 801.
7. (a) Cooke, R. M.; Harvey, T. S.; Campbell, I. D. Biochem-
istry 1991, 30, 5484. (b) Terasawa, H.; Kohda, D.; Hatanaka,
H.; Nagata, K.; Higashihashi, N.; Fujiwara, H.Sakano; Ina-
gaki, F. EMBO J. 1994, 13, 5590.
13. Cutler, M. J. Am. Chem. Soc. 1949, 71, 3375.
14. Zhu, Y. F.; Wilcoxen, K.; Gross, T.; Connors, P.; Strack,
N.; Gross, R.; Huang, C. Q.; McCarthy, J. R.; Xie, Q.; Ling,
N.; Chen, C. Bioorg. Med. Chem. Lett. See previous paper.
doi: 10.1016/50960-894X(03)00321-4.
15. Luty, B. A.; Wasserman, Z. R.; Stouten, P. F. W.; Hodge,
C. N.; Zacharias, M.; McCammon, J. A. J. Comp. Chem.
1995, 16, 454.