Chemically programmed antibodies: Endothelin receptor targeting CovX-Bodies

Article abstract of DOI:10.1016/j.bmcl.2006.10.009

Aryl sulfonamide-based endothelin antagonists were synthesized and covalently linked to the reactive lysine of the m38C2 antibody to create a series of CovX-Bodies. These chemically programmed antibodies behaved as potent endothelin receptor antagonists in vitro and had antitumor efficacy in a prostate cancer xenograft model which, on a molar basis, far exceeded the activity of the parent small molecule.

Full text of DOI:10.1016/j.bmcl.2006.10.009

Bioorganic & Medicinal Chemistry Letters 17 (2007) 501–506
Chemically programmed antibodies:
Endothelin receptor targeting CovX-Bodies
TM
Venkata R. Doppalapudi,* Nancy Tryder, Lingna Li, Teresa Aja,
David Griffith, Francesca-Fang Liao, Giovanni Roxas,
Mysore P. Ramprasad, Curt Bradshaw and Carlos F. Barbas, III
CovX Research LLC, 9381 Judicial Drive, Suite 200, San Diego, CA 92130, USA
Received 22 September 2006; accepted 5 October 2006
Available online 7 October 2006
Abstract—Aryl sulfonamide-based endothelin antagonists were synthesized and covalently linked to the reactive lysine of the m38C2
antibody to create a series of CovX-Bodies. These chemically programmed antibodies behaved as potent endothelin receptor antag-
onists in vitro and had antitumor efficacy in a prostate cancer xenograft model which, on a molar basis, far exceeded the activity of
the parent small molecule.
Ó 2006 Elsevier Ltd. All rights reserved.
Monoclonal antibodies (mAbs) have proven effective as
therapeutics across a spectrum of diseases, including
cancer, heart disease, infection, and immune disorders,
as shown by the growing list of mAbs that have been ap-
proved by the US Food and Drug Administration and
by the European Medicines Agency.1 The therapeutic
potential of antibodies can be expanded by arming them
with drugs, toxins, and radionuclides.2 Although anti-
bodies offer high target specificity, long serum half-life,
and Fc-mediated effector functions, development re-
quires years of research and is costly. On the other hand,
small molecule or peptide-based leads with excellent
activity can be straightforward to find but are often pla-
gued with poor drug-like properties.
(Fig. 1). This approach effectively combines the function
of certain small molecules and peptides with the long
serum half-life of an antibody into one unique therapeu-
tic. Barbas discovered a CovX-Body prototype that is
based on an integrin targeting pharmacophore.3,5 Here,
we further expand the technology to include endothelin
antagonists, and describe the design and synthesis of
endothelin-A (ETA) antagonist CovX-Bodies. In addi-
tion, we also show that ETA-targeting CovX-Bodies
bind ETA and ETB receptors and have anti-tumor effica-
cy in tumor xenograft models.
Endothelins (ET-1, ET-2, and ET-3) constitute a family
of vasoconstrictor peptides, produced by the endotheli-
um of blood vessels and many other tissues.6 Endothe-
lins exert their physiological effects via two specific G-
protein coupled receptors termed ETA and ETB. Both
receptor subtypes are found on smooth muscle cells
and mediate vasoconstriction and/or proliferative disor-
ders. The endothelins and their receptors ETA and ETB
play a major role in tumor growth, proliferation, apop-
tosis, angiogenesis, and bone metastasis.7 ET-1 pro-
motes cell growth and suppresses apoptosis of cancer
cells. Extensive preclinical and clinical studies with
ETA receptor antagonists (and to a lesser degree ETB
antagonists) have shown potential therapeutic benefits
in disease states such as heart failure, pulmonary hyper-
tension atherosclerosis, restenosis, systemic hyperten-
sion, various models of cancer, and chronic renal
failure.6–9
In recent publications, Barbas demonstrated that the
strengths of certain small synthetic molecules and anti-
bodies could be uniquely combined using an aldolase
antibody.3–5 In this approach, a reversible, covalent
enaminone bond forms in a well-controlled fashion be-
tween a b-diketone and active site lysine of the aldolase
antibody m38C2. When the b-diketone is incorporated
into an extended pharmacophore, the result is a well-
characterized, bivalent antibody termed a CovX-BodyTM
Keywords: Monoclonal antibodies; Aldolase antibody; CovX-BodyTM
Endothelin-A; Endothelin-A (ETA) antagonists; Antitumor efficacy.
;
*
Corresponding author. Tel.: +1 858 964 2026; fax: +1 858 964
2090; e-mail: dvramana@covx.com
0960-894X/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.10.009

502
V. R. Doppalapudi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 501–506
Figure 1. Formation of a CovX-Body.
ETA is up-regulated in several tumor types and ETA
antagonists are efficacious in prostate tumor models.
AtrasentanTM, an ETA selective antagonist, recently com-
pleted Phase 3 studies for the treatment of prostate can-
cer.10 Of the known ETA antagonists, the sulfonamide
class is among the most potent and selective.11 One such
sulfonamide developed by Bristol-Myers Squibb, Edon-
entanTM (1), has completed Phase 3 clinical studies for
hypertension. Despite their excellent ETA receptor bind-
ing, the clinical anti-tumor efficacy of sulfonamide-
based ETA antagonists has not been reported. In an at-
tempt to define our CovX-Body technology further, we
selected ETA antagonist 1 as a prototype for an ETA-
targeting CovX-Body. We sought to demonstrate com-
petitive binding to the target receptor and efficacy in a
tumor xenograft model utilizing a CovX-Body equipped
with an ETA targeting pharmacophore.
Synthetic strategy and results. Based upon the published
structure–activity relationship11 of compound 1, the 20-
position amide side chain and the oxazole ring attach-
ment site were chosen as sites for attachment of the b-di-
ketone group (Fig. 2). In order to incorporate a diketone
functionality at the 20-position that would allow anti-
body docking, but not interfere with ETA receptor bind-
ing, a polyethylene glycol (PEG) spacer was included as
shown in compound 2. Alternatively, the upper oxazole
ring was replaced with a PEG spacer-b-diketone func-
tionality resulting in compound 3. The synthetic strategy
for target compound 2 is illustrated in Scheme 1. The
Figure 2. Design of b-diketone containing, ETA binding pharmacophores.

V. R. Doppalapudi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 501–506
503
Scheme 1.
convergent synthesis required three key intermediates 4,
5, and 6.
was confirmed by monitoring the newly generated
enaminone bond spectral absorbance (316 nm) normal-
ized to protein concentration.12 Protein concentration
The masked diketone segment 6 was synthesized from 4-
nitrobenzyl bromide as shown in Scheme 2. Treatment
of 4-nitrobenzylbromide with the dianion of 2,4-pentan-
edione, generated with LDA at ꢀ78 °C, provided the b-
diketone 7 in 40% yield.3 The b-diketone in compound 7
was converted to bis-ketal derivative 8 before the reduc-
tion of the nitro group (Pd/C/H2) to provide aniline 6.
The PEG acid 5 was synthesized by the Jones oxidation
of the corresponding alcohol.
was estimated from the 280 nm absorbance
(A
0:1%
280
¼ 1:4).
Target compound 3, with a tether replacing the oxazole
group, was synthesized as shown in Scheme 4. Mitsun-
obu reaction of 2-iodo, 5-hydroxy benzaldehyde with
PEG alcohol provided compound 13 in 55% yield. Car-
bon–carbon bond formation between the boronic acid
derivative11 and compound 13 proceeded smoothly un-
der Suzuki coupling conditions to provide compound
14. The MEM and tert-butyl protecting groups of 14
were removed with 6 N HCl and the subsequent cou-
pling with the diketone segment 6 resulted in 15.
After synthesis of the biaryl sulfonamide segment 4
according to a reported procedure,11 the condensation
of the PEG acid 5 and compound 4 in the presence of
water soluble carbodiimide provided the biaryl-sulfon-
amide 9 (Scheme 3). Treatment of 9 with 6 N HCl under
reflux conditions removed both the methoxyethoxym-
ethyl ether (MEM) and tertiary butyl protecting groups
to give 10 in 68% yield. Compound 6 was then coupled
to 10 in the presence of (30-dimethylaminopropyl)carbo-
diimide (EDCI) to provide compound 11. Removal of
the ketal-protecting group with a catalytic amount of
p-toluenesulfonic acid (p-TSA) in acetone completed
the synthesis of 2. Compound 2 (2 equiv) was combined
with antibody m38C2 (1 equiv) in phosphate-buffered
saline, pH 7.4, at room temperature to provide CovX-
Body 12 as represented in Figure 1. Reaction completion
Reductive amination of aldehyde 15 with methylamine
and NaBH(OAc)3 and the subsequent treatment with
tert-butyl acetyl bromide at ꢀ5 °C in the presence of
diisopropylethylamine provided 16. Finally, the b-dike-
tone was unmasked by treating 16 with p-TSA in ace-
tone to complete the synthesis of 3. The diketone
compound was transformed into CovX-Body 17 by
reacting with antibody m38C2 as described for synthesis
of CovX-Body 12 above.
The length of the spacer, the attachment point to the
pharmacophore, and the chemical compatibility be-
Scheme 2. Synthesis of protected b-diketone (6) and PEG diacid (5).

504
V. R. Doppalapudi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 501–506
Scheme 3.
Scheme 4.
tween pharmacophore functional groups and the linker
all play a role in tether selection. An ideal tether will
allow reaction of the b-diketone moiety with the anti-
body m38C2 and will also allow the pharmacophore
to interact effectively with the targeted receptor. In or-
der to explore the importance of the spacer length in
compounds tethered at the 20-position, we designed
spacer length variants 19 and 22 (Scheme 5). For the
synthesis of compound 19, a commercially available
PEG diacid (a 32 atom spacer length) was coupled with
compound 6 in the presence of EDCI to provide com-
pound 18. The sulfonamide 4 was then coupled to 18
and the subsequent diketone unmasking gave 19. Com-
pound 19 was reacted with antibody m38C2 to provide
CovX-Body 20.
Compound 22 was synthesized in a similar fashion and
transformed into CovX-Body 23 as shown in Scheme 6.

V. R. Doppalapudi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 501–506
505
Scheme 5.
Scheme 6.
Based on the published structure activity relationship of
compound 1, compounds 27 and 28 were designed to
investigate the effect of a phenyl group near the 20-posi-
tion on ETA/ETB selectivity.11 A Mitsunobu reaction
was employed as shown in Scheme 7 to incorporate
the phenyl group. Addition of masked diketone 6 to
PEG acids 23 and 24 followed by ester hydrolysis pro-
vided compounds 25 and 26. The biphenyl segment 4
was then added and ketals were removed to give 27
and 28. Diketone derivatives 27 and 28 were then con-
verted into the CovX-Body 29 and CovX-Body 30,
respectively.
In vitro ETA and ETB receptor binding. The ETA and
ETB receptor binding affinity was measured using I125
labeled ET-1 and CHO cells that express human ETA
Scheme 7.

506
V. R. Doppalapudi et al. / Bioorg. Med. Chem. Lett. 17 (2007) 501–506
Table 1. ETA and ETB binding assaya results
xenograft model. Endothelin antagonists inhibit the
PC-3 and OVCAR-5 tumor xenograft growth, but
there have been no reports on the anti-tumor effects
of the biphenyl sulfonamides. In a prostate cancer
xenograft (PC-3) study, a 10 mg/kg once weekly dose
of CovX-Body 12 inhibited the tumor volume growth
by >45% by day 31, compared to the vehicle control.
In comparison, a 2 mg/kg daily dose of compound 1
showed 50% tumor growth inhibition by day 31
(Fig. 3). The individual components of a 20 mg/kg
dose of CovX-Body 12, antibody m38C2 alone
(20 mg/kg/once a week), and compound 2 alone
(0.25 mg/kg/once a week) failed to inhibit the tumor
growth. Thus, CovX-Body 12 is highly active at a
low concentration while compound 2 alone was inef-
fective even at dosages ꢁ40-fold higher, on a molar
basis, than those used for CovX-Body 12.
b
ETA IC50 (nM) ETB IC50 (nM) ETB/ETA
Compound
1
1.0
1.6
6003
1110
8978
1121
507
2
CovX-Body 12
8.6
4360
3
19
44.9
8.0
>10,000
4690
9370
564
1372
564
1372
420
95
CovX-Body 20
22
CovX-Body 23 14
7.2
7.7
4510
10,000
1910
570
27
4.6
6.0
CovX-Body 29
28
CovX-Body 30 15.6
21.6
4920
5170
228
331
a The inhibition of endothelin binding to ETA and ETB receptors stably
expressed on CHO cells was measured using 125I-labeled ET-1
competitive assays.
b IC50 values were calculated using means of at least three measure-
ments for six concentrations from 1 lM to 10 pM.
In conclusion, these studies establish the potential of
chemically programmed monoclonal antibodies like
CovX-Body 12 as a novel and effective class of immuno-
therapeutics that combine the merits of traditional small
molecule drug design with immunotherapy.
References and notes
1. Brekke, O. H.; Sandlie, I. Nat. Rev. Drug Discov. 2003, 2,
52.
2. DeNardo, S. J.; Kroger, L. A.; DeNardo, G. L. Curr.
Opin. Immunol. 1999, 11, 563.
3. Rader, C.; Sinha, S. C.; Popkov, M.; Lerner, R. A.;
Barbas, C. F., 3rd Proc. Natl. Acad. Sci. U.S.A. 2003, 100,
5396.
4. Li, L. S.; Rader, C.; Matsushita, M.; Das, S.; Barbas, C.
F., 3rd.; Lerner, R. A.; Sinha, S. C. J. Med. Chem. 2004,
47, 5630.
Figure 3. Anti-tumor efficacy of CovX-Body 12 in PC-3 human
xenograft. *P < 0.05, one-way ANOVA versus post Dunnett’s multiple
comparison test to vehicle.
5. Popkov, M.; Rader, C.; Gonzalez, B.; Sinha, S. C.;
Barbas, C. F., 3rd Int. J. Cancer 2006, 119, 1194.
6. Nelson, J.; Bagnato, A.; Battistini, B.; Nisen, P. Nat. Rev.
Cancer 2003, 3, 110.
7. Bagnato, A.; Natali, P. G. Cancer Treat Res. 2004, 119,
293.
8. Attina, T.; Camidge, R.; Newby, D. E.; Webb, D. J. Heart
2005, 91, 825.
9. Remuzzi, G.; Perico, N.; Benigni, A. Nat. Rev. Drug Disc.
2002, 1, 986.
or ETB receptors13,14 (Table 1). In all binding assays, the
ETA targeting CovX-Bodies, the b-diketone analog, and
the parent small molecule compound 1 were evaluated.
Compound 2 showed an affinity for ETA similar to that
of the parent compound 1 (IC50 1.6 nM vs 1.0 nM). This
result is remarkable considering the presence of a long
PEG tether in compound 2 compared to compound 1.
The corresponding CovX-Body 12, with an IC50 of 8.6
nM, showed a marginal loss in the ETA binding affinity.
On the other hand, the ETB affinity of compounds 2 and
CovX-Body 12 increased slightly relative to compound 1,
with a corresponding alteration to the overall selectivity.
In contrast, compound 3 had an ETA binding affinity al-
most 50-fold lower than that of compound 1, indicating
the importance of the oxazole ring. Increased tether
length (in CovX-Bodies 20 and 23) or addition of a
phenyl group (in CovX-Bodies 29 and 30) did not en-
hance ETA binding affinity relative to CovX-Body 12.
10. Jimeno, A.; Carducci, M. Expert Opin. Investig. Drugs
2004, 13, 1631.
11. Murugesan, N.; Gu, Z.; Spergel, S.; Young, M.; Chen, P.;
Mathur, A.; Leith, L.; Hermsmeier, M.; Liu, E. C.; Zhang,
R.; Bird, E.; Waldron, T.; Marino, A.; Koplowitz, B.;
Humphreys, W. G.; Chong, S.; Morrison, R. A.; Webb,
M. L.; Moreland, S.; Trippodo, N.; Barrish, J. C. J. Med.
Chem. 2003, 46, 125.
12. Wagner, J.; Lerner, R. A.; Barbas, C. F., 3rd. Science
1995, 270, 1797.
13. Mihara, S.; Nakajima, S.; Matumura, S.; Kohnoike, T.;
Fujimoto, M. J. Pharmacol. Exp. Ther. 1994, 268,
1122.
14. Cain, M. J.; Garlick, R. K.; Sweetman, P. M. J.
Cardiovasc. Pharmacol. 1991, 17(Suppl. 7), S150-1.
Xenograft studies. We then evaluated in vivo anti-tu-
mor efficacy of CovX-Body 12 in a nude mouse/human