Discovery of highly potent Src SH2 binders: Structure-activity studies and X-ray structures

Article abstract of DOI:10.1016/S0960-894X(02)00139-7

Optimization of the hydrophobic moiety of caprolactam/thiazepinone based compounds led to the identification of potent Src SH₂ binders in two different series incorporating a phosphotyrosine group (RU 81843) or a phosphobenzoic group (RU 79181). The X-ray co-structures with the Src SH₂ domain revealed different binding modes for RU 81843 and RU 79181, and an excellent fit between RU81843 and the Src SH₂ protein thus explaining its high potency (9 nM, 15-fold more potent than pYEEI reference peptide).

Full text of DOI:10.1016/S0960-894X(02)00139-7

Bioorganic & Medicinal Chemistry Letters 12 (2002) 1291–1294
Discovery of Highly Potent Src SH2 Binders: Structure–Activity
Studies and X-ray Structures
Pierre Deprez,* Isabelle Baholet, Stephane Burlet, Gudrun Lange, Remi Amengual,
Bernard Schoot, Annie Vermond, Eliane Mandine and Dominique Lesuisse
Aventis Pharma, Paris Research Center, Medicinal Chemistry, 102 route de Noisy, 93235 Romainville Cedex, France
Received 23 October 2001; accepted 19 February 2002
Abstract—Optimization of the hydrophobic moiety of caprolactam/thiazepinone based compounds led to the identification of
potent Src SH2 binders in two different series incorporating a phosphotyrosine group (RU 81843) or a phosphobenzoic group (RU
79181). The X-ray co-structures with the Src SH2 domain revealed different binding modes for RU 81843 and RU 79181, and an
excellent fit between RU81843 and the Src SH2 protein thus explaining its high potency (9 nM, 15-fold more potent than pYEEI
reference peptide). # 2002 Elsevier Science Ltd. All rights reserved.
The pp60 Src protein tyrosine kinase1 has become an
attractive target following knock out experiments2
which indicated impaired osteoclast resorption resulting
to osteopetrosis. The Src protein contains several
domains, including SH2 (approximately 100 amino
acids) and SH3 domains involved in protein–protein
interactions.3 With the aim of finding a drug candidate
for the treatment of osteoporosis, we have attempted to
inhibit the Src protein via the binding of non-peptidic
ligands to the SH2 domain of the Src protein.
seven-membered-ring scaffolds: the thioazepinone
(X=S) and caprolactam (X=CH2) scaffolds. These
ligands were obtained by optimization of the substituant
interacting with the hydrophobic pocket. Moreover, to
be as far as possible from a peptidic structure and tak-
ing into account some published Src SH2 ligands incor-
porating the benzoic moiety,6c we decided to explore the
replacement of the phosphotyrosine with a phospho-
benzoic group (Scheme 1).
Examination of the X-ray structures4 of the complex
between the known pYEEI tetrapeptide ligand and the
Src SH2 domain revealed the presence of two major
binding pockets, one interacting with the pY and the
other with pY+3 Ile residue. Between these two pock-
ets, the two glutamate EE residues do not make strong
interactions with the peptide backbone. From rational
drug design, new ligands less peptidic than pYEEI but
Chemistry
The final compounds 4 were prepared using solution-
phase or solid-phase parallel synthesis, as shown in
Scheme 2. The introduction of the hydrophophic group
with a similar binding potency have been recently iden-
5ꢀ8
tified
and we have also described the identification
of the seven-membered ring thioazepinone scaffold as
a promising EE surrogate that proved able to deliver
its two substituants in the respective pTyr and pY+3
pockets.9
In this paper, we would like to report how we
discovered nM Src SH2 binders from two similar
*Corresponding author. Tel.: +33-1-4991-3069; fax: +33-1-4991-
3089; e-mail: pierre.deprez@aventis.com
Scheme 1.
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00139-7

1292
P. Deprez et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1291–1294
Scheme 2. Solution- and solid-phase synthesis of phosphate Src SH2 ligands 4. Reagents and conditions: (a) NaH (1.1 equiv) DMF 1 h at 0 ꢁC, then
add RX (1.1 equiv), overnight at rt or 4 h at 60 ꢁC (35–65% yield); (b) Ar–CH2–Br (1 equiv), KOH (1.1 equiv), Bu4NI (0.1 equiv), THF, overnight at
rt (70–90% yield); (c) CH2Cl2/TFA 2:1 2 h at rt; (d) Ar–A–COOH, BOP, DIEA, CH2Cl2, 3 h at rt; (e) H2, Pd/C, MeOH, overnight; (f) see ref 12;
(g) CCl4, DMAP, HO–Ph–A–COOAll, 1 h at rt; (h) Pd(P(Ph)3)4, DMF/AcOH/NMM (9:1:0.15) 1.5 h; (i) 2; DIC, HOBt, CH2Cl2/DMF overnight;
(j) CH2Cl2, 10% TFA, 30 min.
was achieved through an alkylation of the amide func-
tion of the starting N-Boc protected amino thioazepi-
none10 1a and caprolactam 1b, respectively.10 Two
procedures were used depending on the nature of the
alkylating agent: benzyl halide or alkyl halide. For the
more reactive benzyl halides, an efficient heterogeneous
alkylation (KOH in THF)11 proved to be very con-
venient for parallel synthesis and led to the desired
alkylated amide in good yield (>70% after purification
on silica cartridges) under mild conditions. Unfortu-
nately, this procedure was unsuccessful for alkyl halides
(including alkyl iodides) so we used the standard NaH
method, albeit less convenient for parallel synthesis.
Due to the various reactivities of the substituted alkyl
halides, we decided to convert all the selected alkyl
halides to alkyl iodides and thus the reaction of amide
alkylation with NaH was performed in parallel fashion
with reasonable yields. After deprotection of the N-Boc
protecting group with trifluoroacetic acid, the crude
amine 2 was coupled directly with benzoic acid diben-
zylphosphate or with NAc dibenzyl phosphotyrosine
using the standard BOP coupling procedure. The
resulting protected phosphate 3, was finally treated
under hydrogenolysis conditions to give the desired free
phosphate 4.
Results and Discussion
The solution- and solid-phase syntheses led to the pre-
paration of dozens of compounds resulting from the
combination of the three fragments: caprolactam or
thioazepinone scaffold, phosphobenzoic or phosphotyro-
sine head group and various hydrophobic residues. A
selection of representative compounds is listed in Table 1
and all the purified phosphates were tested in a compe-
tition assay (scintillation proximity assay, SPA),15 with
the pYEEI tetrapeptide as a reference (150 nM).
The lactam substituants have been selected with the help
of molecular modeling: knowing that the hydrophobic
pocket was rather deep, it appeared that a spacer was
needed to link the scaffold with a hydrophobic fragment
located at the bottom of the pocket. Different spacers
(alkyl/benzyl/allyl) bearing aromatic or aliphatic
hydrophobic substituents have been chosen.
In the tyrosine series [with A=CH(NHAc)CH2],
(entries 1–18), significant modifications of the binding
affinity (from mM to nM) were observed depending on
the hydrophobic tail, indicating the crucial role of this
specific interaction for a potent binding affinity. Thus, a
three-carbon atom spacer (entries 1 and 2) appeared to
be the right length with a phenyl hydrophobic residue.
An alternative solid-phase procedure was also devel-
oped based on the DIC coupling between the amino
scaffold 2 and the carboxylic acid supported phosphate
6. The resulting compound 7 was finally cleaved off the
resin with a rapid TFA treatment to generate the
desired phosphate 4, which was further purified using
preparative HPLC (reverse phase). The method of pre-
paration of the phosphate bound compound 6 was per-
formed through the reaction between the phosphite 512
and the phenol moiety of NAc Tyrosine O-allyl (or 1-
carboxy-4-hydroxy phenyl allyl ester) in the presence of
CCl4, as developed recently in our laboratory,13
followed by an allyl ester deprotection with palladium.
The benzyl group was also evaluated and substitution at
the para position of the benzyl group proved to be cri-
tical to the binding affinity, with a 1000-fold difference
depending on the substituent: a low potency with too
short (3, 4) or too bulky substituents (5, 6 and 7) and a
higher potency with a para propyl group (8). A dramatic
improvement was observed with a phenyl substituent
on the benzyl group (10 for RU 81843, 9 nMol, 15-fold
more potent than the parent pYEEI) and we thus
focussed on various biaryl compounds (11–18). It
turned out that the substitution on the second phenyl

P. Deprez et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1291–1294
1293
ring was also very sensitive, indicating its close contact
with the protein surface (as also shown by the X-ray
structure of the complex, Fig. 1). Thus, most of the
substituted phenyl groups appeared to be detrimental
(dramatically in the para position with 14 and 15 at mM
level), except when replaced with a thiophene ring (18;
3 nM).
SH2/citrate crystals, using soaking methodology.16
Among them, the structures of 10 (RU 81843, tyrosine
series) and 21 (RU79181 benzoic series) have been
solved (Figs 1 and 2).
With 10, multiple and very close interactions can be
observed which clearly explained its nM potency. The
two most striking features in the X-ray structure are
that: (1) The second phenyl ring of the biphenyl is going
very deeply in the pocket (much more than the Ileu
residue of pYEEI) and fits perfectly with the hydro-
phobic surface of the protein (in yellow). There are 15
distances shorter than 4 A between the biphenyl moiety
and the protein surface. (2) The caprolactam scaffold
itself has a great influence on the binding affinity, lying
very nicely on the surface generated by the tyrosine 61
residue (hydrophobic contact).
As a general rule, we observed that the compounds with
the caprolactam scaffold (X=CH2) are 10-fold more
potent than those obtained with the thioazepinone
scaffold (X=S), (10 vs 11 and 12 vs 13), indicating sig-
nificant contribution of the scaffold itself to the binding.
In the benzoic series also (with A=no atom), a series of
compounds has been prepared. However, in this case,
we were unsuccessful at improving mM activity. It is
interesting to note some SAR similarities between the
two series. The best substituent is the biphenyl moiety in
both cases (entry 10 and 27), but with a 100-fold differ-
ence in potency. It seems that in the benzoic series, the
potency levels off at the mM level.
It also interacts with a structural water molecule via the
carbonyl of the lactam group and finally H-bonding
with a carbonyl of the protein backbone (H60) with the
initial amino group of the aminocaprolactam. Besides
these interactions, the contribution of the phosphotyro-
sine itself with the pY binding pocket is very similar to
that of pY of pYEEI (not shown).
In the course of this project, we obtained several X-ray
structures of Src SH2/ligand complexes from Src
Table 1. Binding affinity of Src SH2 ligands
Also information rich is the X-ray structure of benzoic
compound 21. When superimposed with 10, the two
Entry
Aa
X=
CH2R
IC50 (nM)b
1
2
3
4
5
6
7
8
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
t
C
C
C
C
C
C
C
C
C
C
S
(CH2)3Ph
CH2–HC¼CH–Ph(trans)
CH2Ph–3OMe
290
280
1700
1800
2300
2040
7800
150
58
9
87
26
315
2160
1300
334
143
3
CH2Ph–3,5CF3
CH2Ph–4tBu
CH2Ph–4c hexyl
CH2Ph–4Bu
CH2Ph–4Pr
9
CH2–Ph–4COOMe
CH2–Ph–4Ph
Figure 1. X-Ray structure of 10 (RU 81843) in Src SH2 (QXP soft-
ware:14 yellow: hydrophobic, blue and red: H bond acceptor and
donor).
10
11
12
13
14
15
16
17
18
CH2–Ph–4Ph
C
S
CH2Ph Ph 20 CN
CH2Ph Ph 20 CN
CH2Ph Ph 40 Me
CH2–Ph–4Ph40 Cl
CH2–Ph–4Ph20 40 diF
CH2–Ph–4 (1naphtyl)
CH2–Ph–4 (2thienyl)
C
C
C
C
C
19
20
21
22
23
24
25
26
27
28
29
30
b
b
b
b
b
b
b
b
b
b
b
b
C
C
S
C
S
S
C
S
C
S
C
C
(CH2)3Ph
CH2–HC¼CH–Ph(trans)
CH2–HC¼CH–Ph(trans)
CH2–Ph–4CO Ph
CH2–Ph–4CO Ph
CH2–Ph–3,5 CF3
CH2–Ph–3,5 CF3
CH2–Ph–4Ph
4500
2300
2800
3000
18,500
9800
10,400
11,900
1500
CH2–Ph–4Ph
CH2–Ph–4Ph 20 CN
CH2–Ph Ph 20 CN
CH2–Ph–4Ph 40 Me
2460
2110
15,200
ab, for benzoic series; t, for tyrosine series.
bSPA binding assay:15 reference peptide pYEEI: 150 nM.
Figure 2. Superimposition of X-ray structures of 10 (RU 81843 pink)
and 21 (RU 79181, white) within Src SH2.

1294
P. Deprez et al. / Bioorg. Med. Chem. Lett. 12 (2002) 1291–1294
major Src SH2 binding pockets are filled similarly, with
a good superposition between the phosphate groups on
one side and the phenyl groups at the bottom of the
hydrophobic pocket on the other side (Fig. 2). As a
consequence, with two carbon less, 21 adopts a ‘short-
cut’ binding mode. However, here again, the seven-
membered ring scaffold appears to be well designed
since it is still lying on the surface generated by the tyr-
osine 61 residue.The structural water molecule (which
was interacting with the lactam ring of 10) is now dis-
placed and replaced directly by the carbonyl of the lac-
tam ring of 21, interacting with the amide nitrogen of
K62. This should be entropically favorable.
5. Charifson, P. S.; Shewchuk, L. M.; Rocque, W.; Hummel,
C. W.; Jordan, S. R.; Mohr, C.; Pacofsky, G. J.; Peel, M. R.;
Rodriguez, M.; Sternbach, D. D.; Consler, T. G. Biochemistry
1997, 36, 6283.
6. (a) Plummer, M. S.; Holland, D. R.; Shahripour, A.; Lun-
ney, E. A.; Fergus, J. H.; Marks, J. S.; McConnel, P.; Mueller,
W. T.; Sawyer, T. K. J. Med. Chem. 1997, 40, 3719. (b) Lun-
ney, E. A.; Para, K. S.; Rubin, J. R.; Humblet, C.; Fergus,
J. H.; Marks, J. S.; Sawyer, T. K. J. Am. Chem. Soc. 1997,
119, 12471. (c) Lunney, E. A.; Para, K. S.; Plummer, M. S.;
Prasad, J. V. N. V.; Saltiel, A. R.; Sawyer, T. K; Shahripour,
A.; Singh, J.; Stankovic, C. J. Patent WO 9712903-A.
7. Beaulieu, P. L.; Cameron, D. R.; Ferland, J.-M.; Gauthier,
J.; Ghiro, E.; Gillard, J.; Gorys, V.; Poirier, M.; Rancourt, J.;
Wernic, D.; Llinas-Brunet, M. J. Med. Chem. 1999, 42, 1757.
8. (a) Buchanan, J. L.; Bohacek, R. S.; Luke, G. P.; Hatada,
M.; Lu, X.; Dalgarno, D. S.; Narula, S. S.; Yuan, R. Y.; Holt,
D. A. Bioorg. Med. Chem. Lett. 1999, 9, 2353. (b) Violette,
S. M.; Shakespeare, W. C.; Bartlett, C.; Guan, W.; Smith,
J. A.; Rickles, R. J.; Bohacek, R. S.; Holt, D. A.; Baron, R.;
Sawyer, T. Chem. Biol. 2000, 7, 225. (c) Shakespeare, W.;
Yang, M.; Bohacek, R.; Cesaroli, F.; Stebbins, K.; Sundar-
amoorthi, R.; Azimioara, M.; Vu, C.; Pradeepan, S.; Metcalf,
C., III; Haraldson, C.; Merry, T.; Dalgarno, D.; Narula, S.;
Hatada, M.; Lu, X.; Van Schravendijk, M. R.; Adams, S.;
Violette, S.; Smith, J.; Guan, W.; Barlett, C.; Herson, J.; Iul-
liucci, J.; Weigele, M.; Sawyer, T. PNAS 2000, 97, 9373.
9. Lesuisse, D.; Deprez, P.; Albert, E.; Duc, T. T.; Sortais, B.;
Gofflo, D.; Jean-Baptiste, V.; Marquette, J.-P.; Schoot, B.;
Sarubbi, E.; Lange, G.; Broto, P.; Mandine, E. Bioorg. Med.
Chem. Lett. 2001, 11, 2127.
However, despite all the positive interactions, benzoic
21 is 300-fold less active than tyrosine 10. One possible
explanation could be that the additional NAc inter-
action present with 10 contributes significantly to the
binding (the carbonyl of Nac interacts with Arg 14). In
this case, the benzoic group is not a potent surrogate of
the tyrosine group for the Src SH2 protein.
In conclusion, we identified compound 10 RU 81843 as
one of the most potent SH2 binder known to date. Its
caprolactam scaffold is able to efficiently deliver the
biphenyl and pY substituents in the respective bind-
ing pockets, whilst also favorably interacting with the
protein.
10. Caprolactam: commercially available (Neosystem);
Thioazepinone: Nobuyoshi, E. FEBS Lett. 1984, 174, 76.
11. Bisarya, S. C.; Roa, R. Synth. Comm. 1992, 22, 3305.
12. (a) Cao, X.; Mjalli, A. M. M. Tetrahedron Lett. 1996, 37,
6073. (b) Zhang, C.; Mjalli, A. M. M. Tetrahedron Lett. 1996,
37, 5457.
13. Deprez, P. Tetrahedron Lett. To be submitted.
14. Mc Martin, C.; Bohacek, R. S. J. Comput.-Aided Mol.
Des. 1997, 11, 333.
References and Notes
1. (a) Brown, M. T.; Cooper, J. A. Biochim. Biophys. Acta
1996, 1287, 121. (b) Liu, D.; Wang, L.-H. J. Biomed. Sci. 1994,
1, 65. (c) Thomas, S. M.; Brugge, J. S. Annu. Rev. Cell Dev.
Biol. 1997, 13513.
2. Soriano, P.; Montgomery, C.; Geske, R.; Bradley, A. Cell
1991, 64, 693.
15. Mandine, E.; Gofflo, D.; Jean-Baptiste, V.; Sarubbi, E.;
Touyer, G.; Deprez, P.; Lesuisse, D. J. Mol. Recog. 2001, 14,
254.
3. (a) Sawyer, T. K. Biopolymers 1998, 47, 243. (b) Pawson, T.
Nature 1994, 373, 573.
4. (a) Waksman, G.; Shoelson, S. E.; Pant, N.; Cowburn, D.;
Kuriyan, J. Cell 1993, 72, 779. (b) Eck, M. J.; Shoelson, S. E.;
Harrison, S. C. Nature 1993, 362, 87. (c) Eck, M. J.; Atwell,
S. K.; Shoelson, S. E.; Harrison, S. C. Nature 1994, 368, 764.
16. Lesuisse, D.; Lange, G.; Deprez, P.; Schoot, B.; Benard,
D.; Delettre, G.; Marquette, J. P.; Broto, P.; Jean-Baptiste, V.;
Bichet, P.; Sarubbi, E.; Mandine, E. J. Med. Chem., in
press.