Synthesis of Novel c(AmpRGD)-Sunitinib Dual Conjugates as Molecular Tools Targeting the αvβ3 Integrin/VEGFR2 Couple and Impairing Tumor-Associated Angiogenesis

Article abstract of DOI:10.1021/acs.jmedchem.6b01266

On the basis of a previously discovered anti-α_Vβ₃ integrin peptidomimetic (c(AmpRGD)) and the clinically approved antiangiogenic kinase inhibitor sunitinib, three novel dual conjugates were synthesized (compounds 1-3), featuring the covalent and robust linkage between these two active modules. In all conjugates, the ligand binding competence toward α_Vβ₃ (using both isolated receptors and α_Vβ₃-overexpressing endothelial progenitor EP cells) and the kinase inhibitory activity (toward both isolated kinases and EPCs) remained almost untouched and comparable to the activity of the single active units. Compounds 1-3 showed interesting antiangiogenesis properties in an in vitro tubulogenic assay; furthermore, dimeric-RGD conjugate 3 strongly inhibited in vivo angiogenesis in Matrigel plug assays in FVB mice. These results offer proof-of-concept of how the covalent conjugation of two angiogenesis-related small modules may result in novel and stable molecules, which impair tumor-related angiogenesis with equal or even superior ability as compared to the single modules or their simple combinations.

Full text of DOI:10.1021/acs.jmedchem.6b01266

Subscriber access provided by NEW YORK UNIV
Article
Synthesis of Novel c(AmpRGD)-Sunitinib Dual Conjugates
V
3
as Molecular Tools Targeting the ## Integrin/VEGFR2
Couple and Impairing Tumor-Associated Angiogenesis
Andrea Sartori, Elisabetta Portioli, Lucia Battistini, Lido Calorini, Alberto Pupi,
Federica Vacondio, Daniela Arosio, Francesca Bianchini, and Franca Zanardi
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01266 • Publication Date (Web): 06 Dec 2016
Downloaded from http://pubs.acs.org on December 6, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Synthesis of Novel c(AmpRGD)ꢀSunitinib Dual
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conjugates as Molecular Tools Targeting the αvβ3
Integrin/VEGFR2 Couple and Impairing Tumorꢀ
Associated Angiogenesis
[a]
[a]
[a]
[b]
[b,c]
Andrea Sartori, Elisabetta Portioli, Lucia Battistini, Lido Calorini, Alberto Pupi,
[
a]
[d]
[b,c]
[a]
Federica Vacondio, Daniela Arosio, Francesca Bianchini,*
and Franca Zanardi*
[
a] Dipartimento di Farmacia, Università degli Studi di Parma, Parco Area delle Scienze 27A,
3124 Parma, Italy
4
[
b] Dipartimento di Scienze Biomediche, Sperimentali e Cliniche “Mario Serio”, Università
degli Studi di Firenze, Viale G. B. Morgagni 50, 50134 Firenze, Italy
[c] Centro Interdipartimentale per lo Sviluppo Preclinico dell’Imaging Molecolare (CISPIM),
Università degli Studi di Firenze, Viale G. B. Morgagni 50, 50134 Firenze, Italy
[
d] Istituto di Scienze e Tecnologie Molecolari, Consiglio Nazionale delle Ricerche, Via Golgi
19, 20133 Milano, Italy
KEYWORDS: Angiogenesis, Dual conjugates, Growth Factor Receptors, Integrins, RGD
Ligands, Peptidomimetics, Sunitinib
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
ABSTRACT
Based on a previously discovered antiꢀα β integrin peptidomimetic – c(AmpRGD) – and the
V
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
clinically approved antiꢀangiogenic kinase inhibitor sunitinib, three novel dual conjugates were
synthesized (compounds 1ꢀ3), featuring the covalent and robust linkage between these two active
modules. In all conjugates, the ligand binding competence toward α β (using both isolated
V
3
receptors and α β ꢀoverexpressing endothelial progenitor EP cells) and the kinase inhibitory
V
3
activity (toward both isolated kinases and EPCs) remained almost untouched and comparable to
the activity of the single active units. Compounds 1ꢀ3 showed interesting antiꢀangiogenesis
properties in an in vitro tubulogenic assay; furthermore, dimericꢀRGD conjugate 3 strongly
inhibited in vivo angiogenesis in Matrigel plug assays in FVB mice. These results offer proofꢀofꢀ
concept of how the covalent conjugation of two angiogenesisꢀrelated small modules may result
in novel and stable molecules, which impair tumorꢀrelated angiogenesis with equal or even
superior ability as compared to the single modules or their simple combinations.
INTRODUCTION
1ꢀ3
Targeted therapy that selectively addresses oncogenic drivers, as well as the use of drugs
4
ꢀ6
concomitantly perturbing multiple molecular targets and signaling pathways are arising as
privileged therapeutic options. Angiogenesis, the process by which new blood vessels arise from
preexisting vasculature, plays crucial roles in both normal and pathological events. In particular,
aberrant angiogenesis – involving constantly activated vasculature – is widely accepted as a key
player in a variety of pathological conditions including cancer growth and metastasis,
ACS Paragon Plus Environment
2

Page 3 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
rheumatoid arthritis and ageꢀrelated macular degeneration. Many diverse endogenous molecular
systems participate in the angiogenesis regulation, including proꢀangiogenic vascular endothelial
growth factors (VEGFs), plateletꢀderived growth factors (PDGFs), their associated tyrosine
kinase receptors (VEGFRs and PDGFRs), matrix metalloproteinases (MMPs), ephrinꢀephrin
receptor complexes, and specific extracellular matrix (ECM)ꢀrecognizing integrin receptors as
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7ꢀ12
α β , α β and α β .
V
3
V
5
5 1
Substantial four decadeꢀlong body of research in this field resulted in worldwide approval by
drug agencies of effective antiꢀangiogenic drugs including the humanized monoclonal antibodies
trastuzumab and bevacizumab, which are VEGF antagonists, and several small molecules such
as sunitinib and sorafenib, which mainly target the highly conserved cytosolic tyrosine kinase
9
,12
domain of VEGFRs.
Sunitinib, in particular (Figure 1), is an alkylidene 2ꢀoxindole agent
which acts as a highly effective multitarget tyrosine kinase inhibitor (TKI, mainly against
1
3ꢀ16
VEGFR2, PDGFRβ, cKit, and Fltꢀ3);
it is indicated as firstꢀline therapy for metastatic renal
cell carcinoma, pancreatic neuroendocrine tumors, secondꢀline therapy for imatinibꢀresistant
1
3ꢀ16
17
gastrointestinal stromal tumors,
and it is still the focus of countless clinical trials.
Among antiangiogenic α β ꢀintegrin inhibitors, the small molecule cilengitide [c(RGD)NMefV]
V
3
is the most widely studied, and several advanced clinical trials are still in progress concerning the
1
8ꢀ21
use of this drug as either single agent or in combination with radiotherapy.
However, the
limited longꢀterm efficacy and the systemic toxicity associated with the clinical use of the
approved antiꢀangiogenic drugs posed serious concerns about the actual benefit and safety of
2
2ꢀ29
these treatments.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. The antiangiogenic drug sunitinib and the 4ꢀaminoprolineꢀbased RGD
a
b
c
cyclotetrapeptide c(AmpRGD)ꢀ(CH ) NH . Ref. 13; Ref. 42; Ref. 43.
2
2
2
22ꢀ29
While the debate about the actual usefulness of antiꢀangiogenesis therapy remains open,
current researches suggest that possible solutions could entail the use of drugs capable of hitting
1
0,12,25,29,30
multiple targets/pathways and cell types involved in the tumor microenvironment,
while possessing supplemental selective targeting moieties.
Among the intricate, often overlapping cell signaling networks regulating angiogenesis, growing
31ꢀ36
evidence emerged for strict crosstalk between the VEGFR2 and α β receptors.
These two
V
3
receptors are physically and functionally connected in common cell populations (e.g. endothelial
cells, ECs, and several cancer cell types) and their interactions are important for both integrin
3
1ꢀ36
activation and mutual regulation of the kinase activity.
Blockage of the α β /VEGFR2
V
3
3
7ꢀ40
couple may thus be of high therapeutic potential.
In fact, the combined use of two small
molecules – cilengitide and the sunitinib analogue SU5416 (Figure 2, A) – showed antiꢀ
angiogenic effect and inhibition of tumor melanoma growth and metastasis during in vivo
3
7
preclinical studies. In this instance, however, the two drugs were independently delivered, with
38
possible differences in localization and pharmacokinetics. In an enlightening study, a dual
ACS Paragon Plus Environment
4

Page 5 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
specific scVEGF protein was engineered, capable of binding the extracellular portions of αVβ3
and VEGFR2 simultaneously, showing promise for effective in vitro and in vivo antiꢀangiogenic
action (Figure 2, B). Though highly promising, this work had the limitation of dealing with
complex engineered 25 kDaꢀweighty proteins.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2. A) α β antagonist cilengitide (outside cell)+VEGFR2 antagonist (inside), two
V
3
discrete small molecules, different localization (Ref. 37); B) one engineered dualꢀspecific protein
(outside), simultaneous α β /VEGFR2 inhibition (Ref. 38); C) and D) one small molecule, the
V
3
c(AmpRGD)ꢀTKI conjugate, recognizing α β (outside), partially entering the cell (possibly via
V
3
α β ), and antagonizing VEGFR2 (inside) (this work).
V
3
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A complementary and conceptually different approach is here proposed, according to which a
selective binder of the extracellular segment of α β is covalently linked to a proven TKI such as
V
3
sunitinib, whose interaction with the cytoplasmic domain of VEGFR2 is widely recognized. As
α β binder, we could rely on a recently discovered series of aminoprolineꢀbased RGD
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
V
3
4
1ꢀ43
cyclotetrapeptides of type c(AmpRGD)ꢀ(CH ) NH (Figure 1)
which showed remarkable and
2
2
2
selective binding capability toward the α β integrin receptor in both cellꢀfree and cell assays.
V
3
The covalent assemblage of an anchorable sunitinibꢀlike moiety to the c(AmpRGD) portion
through a suitable linker would furnish dual conjugates (Figure 2 C and 2 D, schematic
representation) wherein the RGD unit would possibly provide i) ECꢀselective targeting by α β ꢀ
V
3
RGD recognition, ii) α β ꢀdependent antiꢀangiogenic effect, and iii) α β ꢀmediated cell
V
3
V 3
internalization. On the other hand, the sunitinib unit could exert its intracellular TKI effect after
4
4ꢀ46
internalization, while playing a role in overall perturbation of the α β ꢀVEGFR2 crosstalk.
V
3
We herein disclose the design, synthesis, and biological activity evaluation of three novel
covalent prototypes 1ꢀ3 (Figure 3). In particular, the binding properties toward α β integrin,
V
3
kinase inhibitory activity, cell uptake, and antiꢀangiogenesis potential in vitro and in vivo are
reported and discussed, visꢀàꢀvis the behavior of the single modules and their simple
combinations.
RESULTS AND DISCUSSION
Design of cAmpRGD-sunitinib Conjugates. To fulfill the objectives of this work (Figure 2 C
and 2 D), the projected dual conjugates had to embody several stringent requisites. First, the
ACS Paragon Plus Environment
6

Page 7 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
active units should not disturb each other, that is, the sunitinib moiety should not compromise
the RGDꢀbinding capability while the RGD unit should not impede the tyrosine kinase activity of
sunitinib. Second, to exclude premature detachment of the two active units (outside the targeted
cells), the linker between them should be either uncleavable or cleaved within cells exclusively;
and third, the conjugate must enter the targeted cells possibly via α β ꢀmediated endocytosis.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
V
3
13,47
As a background, extensive structureꢀactivity relationship studies on sunitinib analogues
and
Xꢀray analysis of the complex between the drug and the tyrosine kinase domain of the VEGFR2
4
8,49
active site
revealed that the aromatic portion of the molecule is directly involved in the
binding, while the terminal tertiary amine stands outside the pocket and may allow certain
margins of structural modifications. Thus, connection of sunitinib to the linker exploiting this
amine terminal would likely be uninfluential toward the tyrosine kinase activity. Furthermore,
we were aware of the binding capability and selectivity toward the α β integrin of
V
3
4
1ꢀ43,50,51
c(AmpRGD)ꢀbased ligands and related conjugates,
anticipating that conjugation of these
ligands with the ancillary sunitinib moiety would not hardly compromise their α β ꢀintegrin
V
3
binding ability. Lastly, the cell internalization potential of c(AmpRGD)ꢀconjugates was
preliminarily assayed using a c(AmpRGD)ꢀfluorescein conjugate control which demonstrated
complete α β ꢀdependent internalization in A375 melanoma cells within 25 min exposure (see
V
3
details in the Supporting Information).
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Structure of the targeted dual conjugates 1ꢀ3.
With these clues at hand, conjugates 1ꢀ3 were designed, wherein the two active units are
α
ω
positioned 11ꢀtoꢀ22 bonds away (Figure 3, N to N ) and are connected via robust
triazole/ether/amide linkages, as shown in Figure 3. Monomeric compounds 1 and 2 differ from
each other in the linker length and type; compounds 2 and 3 share a common pegylated linker
and maintain the tertiary amine functionality of the parent drug, while compound 1 replaces this
amine with a secondary amide. Finally, compound 3 features a dimeric RGD presentation, which
could be important for enhanced integrin recognition and integrinꢀmediated cell
52ꢀ57
internalization.
ACS Paragon Plus Environment
8

Page 9 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Chemistry. The synthesis began with the preparation of three constitutive modules namely, the
c(AmpRGD)ꢀazide 7 (Scheme 1), the sunitinib analogue 13 (Scheme 2), and the linker moieties
14, 16, and 18 (Scheme 3 and Supporting Information). Thus, commercially available protected
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
cisꢀaminoꢀLꢀproline 4 was selectively deprotected at the N(α)ꢀsite and alkylated via reductive
amination using 4ꢀazidopropanal (Scheme 1).
a
Scheme 1. Synthesis of the c(AmpRGD) Modules 7 and 8
a
Reagents and conditions: (a) (i) TFA, DCM, rt; (ii) 4ꢀazidobutanal, NaBH(OAc) , 1,2ꢀDCE,
3
rt; (b) FmocꢀSPPS: (i) ClꢀcTrtꢀOꢀGlyꢀ(Pmc)ArgꢀNH , 5, HATU, HOAt, collidine, DMF, rt; (ii)
2
piperidine, DMF, rt; (iii) FmocꢀAsp(tBuO)ꢀOH, HATU, HOAt, collidine, DMF, rt; (iv)
piperidine, DMF, rt; (v) AcOH, TFE, DCM, rt; (c) HATU, HOAt, collidine, DCM/DMF, rt; (d)
(
i) H , Pd/C, EtOH, rt; (ii) TFA/TIS/H O (95:2.5:2.5), rt.
2
2
Azidoproline 5 was efficiently obtained (95% yield), and inserted into the projected peptide
chain via conventional Fmocꢀbased solid phase peptide synthesis (FmocꢀSPPS) using chlorotrityl
(cTrt) resin. After detachment from the resin, 6 was recovered in 94% yield and subjected to inꢀ
solution cyclization under diluted conditions (13:1 DCM/DMF solvent mixture, 2.2 mM), using
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the HATU/HOAt reagent couple, giving protected azidoꢀterminating cyclotetrapeptide 7, ready
for the subsequent conjugation step. Overall, the novel azide module 7 was prepared in a eightꢀ
step sequence and rewarding 63% overall yield from proline 4. Azide 7 could be also
conveniently converted to free amine 8 via reduction and acidic deprotection (82%, two steps),
which served as a reference control during the biological assays (vide infra). It is to be noted that
the synthesis of similar c(AmpRGD) azide/amine congeners possessing two carbonꢀlong alkyl
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
2,43
chains instead of four was reported by us in previous works;
in that instances, however,
longer linear sequences were experienced (14ꢀ15 steps) and lower yields were obtained (10ꢀ16%
overall yields).
a
Scheme 2. Synthesis of the Sunitinib-Like Module 13
a
Reagents and conditions: (a) NH NH , 100 °C; (b) KOH, H O, MeOH, reflux; (c) piperidine,
2
2
2
EtOH, 60 °C.
As for the sunitinib portion, carboxylic acid 13 was judged a good precursor. As shown in
Scheme 2, Knoevenagel condensation between oxindole 10, in turn obtained from commercial 5ꢀ
fluoroisatin 9 by WolffꢀKishner reduction, and pyrrole acid 12 (from saponification of the
ACS Paragon Plus Environment
10

Page 11 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
corresponding commercial ethyl ester 11) consigned 3ꢀalkylidene 2ꢀoxindole 13 in a good 80%
1
3
58
yield, as the sole detectable Zꢀconfigured isomer.
Simple chemistry was used to access alkyneꢀterminating amine 14, pegylated counterpart 16, and
bisꢀalkyne amine 18 (structures in Scheme 3), whose straightforward preparation from
commercial starting materials is described in the Supporting Information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
All was ready for preparation of the three dual conjugates 1ꢀ3, where the main modules could be
connected through common synthesis pathways. Thus, as shown in Scheme 3, parallel BOPꢀ
promoted condensation of carboxylic acid 13 with either amines 14, 16 or 18 provided access to
the respective alkyneꢀterminating amides 15, 17, or 19 in good isolated yields (64ꢀ88%). Copperꢀ
catalyzed 1,3ꢀdipolar cycloaddition between these alkynes and the previous c(AmpRGD) azide 7
(twoꢀfold equivalents in case of the dimeric execution in eq. 3) followed by acidic deprotection
gave conjugates 1ꢀ3 in good yields and purity after semipreparative reverseꢀphase HPLC
purification (43ꢀ70% yields; 96ꢀ98% purity, recovered as TFA salts).
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
a
Scheme 3. Modular Synthesis of the Sunitinib-c(AmpRGD) Conjugates 1-3
O
N3
H
N
O
N
3
N
O
H
HN
b
O
(tBu)Asp
Arg(Pmc)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
N
Gly
a
H
15
N
H
F
7
eq. 1) 13
+
+
1
2
NH2
88%
O
69%
1
4
N
H
O
N
O
2
N
H
O
NH2
a
7, b
N
O
eq. 2) 13
N
O
H
17
70%
64%
F
F
2
O
16
N
H
O
H
N
N
O
NH2
+
N
O
H
N
H
2
N
O
O
a
7%
7, b
eq. 3) 13
2
3
N
6
H
43%
O
O
18
19
O
O
O
N
H
a
Reagents and conditions: (a) BOP, DIPEA, DCM/DMF (2:3), rt; (b) (i) azide 7, Cu(OAc) , Na
2
L
ꢀasc, H O/DMF, rt; (ii) TFA/TIS/H O (95:2.5:2.5), rt.
2 2
In Vitro Stability of Conjugates 1-3. The in vitro stability of conjugates 1ꢀ3 in 80% v/v rat and
human plasma was firstly evaluated by HPLCꢀUVꢀVis analysis. The cyclopeptides were
incubated and analyzed up to 8 h, as detailed in Table 1 (see also the Supporting Information).
Invariably and regardless their intimate structure, compounds 1ꢀ3 showed complete resistance to
rat and human plasma degradation during the observed time. This demonstrated that the covalent
connection of the modules resulted in robust conjugates anticipating that, whatever the biological
response, it would be the result of the interaction of the cell environment with the integral,
preserved structure of the conjugates and not the individual detached components. Whether this
would translate into a benefit or disadvantage in biological assays remained to be seen.
ACS Paragon Plus Environment
12

Page 13 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. In Vitro Plasma Stability of Compounds 1-3 vs Sunitinib
Compound
Rat plasma
% compd at 8h)
97.3 (±9.1)
Human plasma
(% compd at 8h)
102.3 (±5.2)
a
a
(
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
sunitinib
1
2
3
88.2 (±12.7)
104.1 (±3.3)
108.8 (±10.5)
99.0 (±10.9)
106.6 (±11.5)
98.6 (±15.4)
a
Percentage of compound remaining after 8 h of incubation in 80% v/v plasma, 37 °C,
protected from light. Reported are Means ± SD.
Lipophilicity and Cellular Uptake of Conjugates 1-3. At physiological pH, compounds 1ꢀ3
proved highly hydrophilic in accordance with the measured negative values of the LogDoct,7.4 (i.e.
the distribution coefficient in nꢀoctanol/buffer at pH 7.4, Table 2). As expected, attachment of
the c(AmpRGD) module to the sunitinibꢀlike portion turned the lipophilic character of the drug
to hydrophilic, which was magnified by the pegylated linker (compounds 2 and 3 vs 1) and the
dipeptide presentation (3 vs 1 and 2).
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. Lipophilicity and Cellular Uptake of Compounds 1-3 and Sunitinib in EPCs
Compound
Log
MW
Intracellular
Content
Cell uptake
(pmol/min
/mg prot)
(1 h)
Log
Cell
Intracellular
Content
a
Doct,7.4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
nmol/mg prot)
Uptake
(1 h)
(nmol/mg
b
(1 h)
prot)
b
(8 h)
sunitinib
2.51 398.5
− 2.03 960.0
− 2.56 1078.2
− 3.02 1788.9
5.72 (±0.71)
0.26 (±0.02)
0.48 (±0.07)
0.94 (±0.02)
95.3
4.3
1.98
0.64
0.90
1.19
2.90 (±0.40)
0.22 (±0.02)
0.43 (±0.09)
0.54 (±0.03)
1
2
8.0
3
15.7
a
Distribution coefficient in the nꢀoctanol/buffer system, pH 7.4. Reported are Means ± SD.
EPCs were incubated with 1ꢀM final concentration of test compound. After 1 h, the medium
b
containing the tested compounds was removed and intracellular content was quantified
immediately and after 8 h. Experiments were conducted in triplicate and data were expressed as
nmol/mg of total cell proteins in each sample.
The capability of endothelial progenitor cells (EPCs) to internalize conjugates 1ꢀ3 as compared
to free sunitinib was next investigated. Total intracellular concentrations of 1ꢀ3 and sunitinib
were measured by HPLCꢀESIꢀMS/MS. EPCs were incubated in standard conditions for 1 h in the
presence of the different compounds at 1 ꢀM final concentration (see also Experimental and
Figure S1 in the Supporting Information). As illustrated in Table 2, all compounds were detected
in the intracellular extracts, with the smallꢀsized sunitinib drug showing maximum levels at 1 h,
while conjugates 1ꢀ3 were found in the cell extract to a much lesser extent. In particular,
compound 1 showed a scarce entrance in cells after 1 h treatment, which remained almost
ACS Paragon Plus Environment
14

Page 15 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
invariable after 8 h. Pegylated counterpart 2, having similar molecular weight, almost doubled its
ability to enter cells as compared to 1 at both 1 h and 8 h treatment. Finally, dimeric RGD
conjugate 3, notwithstanding its higher molecular weight, showed a 4ꢀfold and 2ꢀfold ability to
enter cells as compared to 1 and 2, respectively.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
These data are quite interesting since they emphasize the following points: i) the requisite
delivery of the sunitinibꢀlike moiety inside cells is provided by conjugates 1ꢀ3, even if the
internalization is not as efficient as the free drug, and ii) the amount of each conjugate (expressed
as pmol/min/mg prot) which passes through the EPC membrane in the first hour is, on a log
scale, inversely related to its lipophilicity, expressed by the distribution coefficient at pH 7.4, i.e.
2
Log(cell uptake) = ꢀ0.55(±0.04)log Doct, 7.4 – 0.50(±0.10); n=3; r = 0.995; s=0.03; F=191. The
more hydrophilic and bulkier 3 is more efficiently internalized than 1; the dependence of the
internalized content upon the RGD presentation (monomeric vs dimeric) suggests a direct
involvement of the RGD moiety during the internalization process possibly via α β ꢀmediated
V
3
5
2ꢀ57
endocytosis.
The assay was repeated for compounds 2 and 3 in the presence of excess αVβ3
integrin ligand 8 (100 ꢀM). Significant decrease of cell uptake was witnessed for both
compounds (Table S1 in the Supporting Information) further corroborating the notion of an
active role of this integrin during internalization.
Solid-phase receptor binding assay. The integrin activity and selectivity profile of compounds
1
ꢀ3 were firstly evaluated by measuring their ability to bind to human, isolated α β and α β
V
3
5 1
integrin receptors by competitive displacement assays using either biotinylated vitronectin VN
for α β ) or biotinylated fibronectin FN (for α β ). To better evaluate the impact of the sunitinib
(
V
3
5 1
moiety on binding capability, the results were compared to those obtained for the unconjugated
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
counterpart c(AmpRGD)ꢀ(CH ) NH and commercial ligand c(RGDfV). As shown in Table 3,
2
2
2
compounds 1ꢀ3 exhibited oneꢀdigit nanomolar affinity toward α β integrin, which was even
V
3
superior to the unconjugated AmpRGDꢀbased counterpart and showed in all cases an appreciable
α β /α β selectivity. Compound 3, bearing a 2ꢀfold RGD repeat, showed an increased binding
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
V
3
5 1
affinity as compared to monomer 2, even if it was lower than 1. Overall, the presence of a
sunitinibꢀlinker cargo attached to the integrinꢀrecognizing RGD unit did not compromise the
exquisite binding affinity and selectivity of these conjugates.
Table 3. Inhibition of Biotinylated VN and FN Binding to α β and α β Receptors,
V
3
5 1
a
Respectively
Compound
IC (nM)±SD for α β
IC (nM)±SD for α β
5
0
V
3
50
5 1
1
2
3
1.24±0.01
5.1±0.6
30.7±17.7
101.3±31.3
95.8±46.7
151.6±67.6
166.0±28.0
3.8±0.6
b
b
c
c(AmpRGD)ꢀ(CH ) NH
6.1±1.6
2 2
2
b
c(RGDfV)
3.2±1.3
a
IC values were calculated as the concentration of compound required for 50% inhibition of
5
0
biotinylated VN or FN binding to human, isolated receptors. Each data point represents the
average of triplicate wells; data analysis was carried out by nonlinear regression analysis using
b
c
GraphPad Prism software. Each experiment was repeated in duplicate. Ref. 42. Ref. 43.
Inhibition of EPC adhesion to the α β -ligand vitronectin using conjugates 1-3. The
V
3
synthesized compounds 1ꢀ3 were evaluated for their ability to inhibit the adhesion of natural
ligand VN to α β ꢀoverexpressing cells. Endothelial progenitor cells were chosen due to their
V
3
abundant α β integrin receptor expression (as certified by flow cytometric analysis, Figure S2)
V
3
5
9ꢀ61
and for their recognized role in tumor angiogenesis.
The assay of adhesion inhibition was
ACS Paragon Plus Environment
16

Page 17 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
performed in the presence of 2.0 mmol/L MnCl to switch α β integrin to its activated form
2
v 3
with increasing concentrations of compounds 1, 2 and 3 (1, 10, 100, 1000, 10000 nM); for
comparison purposes, sunitinib alone, unconjugated c(AmpRGD) 8, and a combination of both
were also assayed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Inhibition of EPCs adhesion to VN in the presence of compounds 1ꢀ3, 8, sunitinib, and
the combination sunitinib+8 (ꢀ compound 1; ꢀ compound 2; ꢁ compound 3; ꢂ compound 8;
ꢃ sunitinib; ꢀ compound 8+sunitinib). Top insert indicates the percentage of α β integrin
V
3
expression in EPCs. The inhibitory activity was calculated as percentage of cell adhesion to VN
in untreated cells and was expressed as mean±SD. Experiments were carried out in triplicate.
As shown in Figure 4, conjugated cyclopeptides 1 and 3 strongly inhibited cell adhesion in a
doseꢀrelated manner with IC values nearly approaching 500 nM, while conjugate 2 showed a
5
0
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
less efficient activity (IC ca 10 ꢀM); notably, the binding capability of compounds 1 and 3 was
5
0
even better than the unconjugated counterpart 8 (IC 1.8 ꢀM). As expected, the binding
50
capability of sunitinib alone remained negligible at these concentrations. Overall, the covalent
conjugation of the c(AmpRGD) portion to the sunitinibꢀlike moiety as described in the diverse
topologies of compounds 1ꢀ3 does not significantly alter the ligand binding capability towards
these endothelial α β ꢀoverexpressing cells.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
v
3
Effect of conjugates 1-3 on cell proliferation and cell viability. The effect of the different
compounds was evaluated on EPCs in a proliferation assay performed in the presence of VEGFꢀ
A (20 ng/mL), and the various compounds 1ꢀ3, 8, sunitinib, and 8+sunitinib at 1 ꢀM
concentration every 24 h. The effect was followed after 24 h, 48 h and 72 h exposure. Cell
proliferation was measured by cell count and cell viability was evaluated using trypan blue
exclusion assay (Figure 5). After the first 24 h treatment, conjugates 1ꢀ3 showed inhibition of
VEGFꢀinduced proliferation of 46%, 24%, and 41%, respectively, with respect to untreated cells.
On the other hand, sunitinib inhibited cell proliferation of 33%, and the combination of sunitinib
with compound 8 poorly impacted cell proliferation. After 48 h treatment, conjugates 1ꢀ3 and
sunitinib alone maintained almost the same level of inhibition of proliferation. Interestingly, the
combination of sunitinib and compound 8 revealed a 23% of inhibition. After 72 h, inhibition of
EPCs proliferation found in cells treated with compound 3 was similar to that of cells exposed to
sunitinib, while other treatments did not show a significant inhibition of proliferation.
During the entire experiment, cell viability was monitored and no significant difference was
found in EPCs exposed either to the conjugated compounds or to separate drugs.
ACS Paragon Plus Environment
18

Page 19 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Effect of the different compounds on VEGFꢀmediated EPCs proliferation. EPCs were
grown in a serum and growth factorꢀfree medium containing 20 ng/mL VEGFꢀA. Cells were
exposed to 1ꢀM concentration of different compounds every 24 h (ꢁ compound 8, ꢁ sunitinib;
ꢁ
compound 8+sunitinib, ꢁ compound 1, ꢁ compound 2, ꢁ compound 3). After 24 h, 48 h,
and 72 h incubation, cells were counted and cell viability was assessed. Representative of three
independent experiments.
Inhibition of TKI activity by conjugates 1-3. To evaluate whether conjugation within 1ꢀ3
would affect the TKI activity of the sunitinibꢀlike portion toward its targeted kinases, we firstly
evaluated the inhibitory activity of representative compound 3 against human recombinant
PDGFRβ and VEGFR2. As shown in Table 4, IC values were in the nanomolar range, slightly
5
0
13
superior than those reported for sunitinib, demonstrating that appendage of the two RGD
moieties and linker was not detrimental for TKI activity in vitro.
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 4. Inhibition of TKI Activity for Compound 3 and the Reference Compound
a
Sunitinib against Human Recombinant PDGFRβ and VEGFR2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound
PDGFRβ (nM)
VEGFR2 (nM)
420
3
9
b
b
sunitinib
2
80
a
IC50 values for 3 were calculated as the concentration of compound required for 50%
inhibition of control specific activity (staurosporine). Each data point represents the average of
duplicate wells; data analysis was carried out by nonlinear regression analysis using software
b
developed at Cerep (Hill software). Ref. 13.
The ability of compounds 1ꢀ3 to inhibit VEGFꢀstimulated VEGFR2 phosphorylation was
investigated by Western blotting using EPCs, which were proven to express high levels of
VEGFR2 (besides α β ). Sunitinib alone, c(AmpRGD) 8 alone, and a combination of the two
V
3
were also assayed for comparison purposes. Percent inhibition at 1 ꢀM concentration is reported
in the densitometric analysis histogram (Figure 6). EPCs were treated for 1 h with the different
6
2
compounds and then activated with 50 ng/mL VEGFꢀA for 5min before cell lysis for VEGFR2
phosphorylation detection. Among the different compounds, 1, 2 and 8 showed a weak induction
of VEGFR2 phosphorylation that might be the result of the synergistic intracellular interaction
between α β and VEGFR2 probably through Src domains leading to a mild activation of the
V
3
32,35
VEGF receptor.
As mentioned before, the biological behavior of monomeric compounds 1
and 2 might be influenced more by their RGD moiety rather than the sunitinib moiety, as a
consequence of their weak propensity to enter EP cells. Interestingly, dimeric compound 3
induced a marked reduction of VEGFR2 phosphorylation, comparable to that found in EP cells
ACS Paragon Plus Environment
20

Page 21 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
exposed to sunitinib alone or to the combination of sunitinib + 8; and this would support the
notion that the biological behavior of compound 3 is heavily influenced by its sunitinib moiety,
likely due to the enhanced ability to be delivered to the intracellular compartment through the
double RGD moieties. Overall, the TKI activity of compound 3 is attributable to the direct
interaction with the Y951 domain of VEGFR2 (as sunitinib does) supporting the evidence that 3
acts as genuine VEGFR2 antagonist.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Inhibition of VEGFR2 phosphorylation in EPCs treated for 1 h with 1 ꢀM
concentration of different compounds followed by VEGFꢀA activation (50 ng/mL) for 5 min. A)
densitometric analysis of the inhibition of VEGFꢀinduced VEGFR2 phosphorylation; B) western
blot analysis. Results are reported as the mean±SD of percent of inhibition of VEGFR
phosphorylation compared to VEGFꢀtreated cells. Representative of three independent
experiments.
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Conjugates 1-3 inhibit the angiogenic process in vitro and in vivo. The ability of conjugates
1
ꢀ3 to interfere with EP cells in organizing capillary network in vitro was determined. Cells were
seeded on Matrigel and exposed to a medium containing VEGFꢀA (20 ng/mL). Cells were
incubated for 6 h in the presence of conjugates 1ꢀ3, unconjugated c(AmpRGD) 8, sunitinib, and
the combination 8+sunitinib at 0.01, 0.1, and 1.0 ꢀM concentrations. As shown in Figure 7A, a
significant or even dramatic reduction in the number of newly formed tubules was observed
when EP cells were incubated on Matrigel with the various compounds. The quantification was
performed by measuring the number of loops formed by connecting capillary projections
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(branches) and expressed as percentage of reduction compared to untreated cells, as reported in
Figure 7B.
Unconjugated c(AmpRGD) 8 shows a good and doseꢀrelated antiꢀangiogenic activity which is
likely due to the inhibitory interaction between the RGD moiety and the extracellular binding
domain of α β integrin. Slightly superior antiꢀangiogenic response is witnessed with sunitinib
V
3
alone, which clearly owes this behavior to its interaction with the intracellular domain of
different kinases, including VEGFR2. Treating EPCs with the combination 8+sunitinib results in
a remarkable doseꢀdependent (slightly sloped) antiꢀangiogenic trend, with an exceptional 60%
inhibition at 10 nM, much higher than that observed in 8 (19%) and sunitinib (30%) separately.
Passing to conjugates 1ꢀ3, the inhibitory activity of angiogenesis is more pronounced for 1 and 3
than for 2. In particular, for monomeric compound 1 and dimeric derivative 3, the antiꢀ
angiogenesis activity is similar to that of the single 8 or sunitinib at both 1 ꢀM and 0.1 ꢀM
concentrations, while it is highly improved (52% for 1 and 43% for 3) at 10 nM, somehow
paralleling the behavior of the combination. This demonstrates that for both the combined and
conjugated ingredients, a favorable synergy could exist, given by both the extracellular RGDꢀ
ACS Paragon Plus Environment
22

Page 23 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
integrin interaction (likely provided by the nonꢀinternalized fraction of compounds) and the
sunitinibꢀVEGFR2 kinase interaction (provided by the amount of internalized compound, see
also uptake data). Indeed, the partial internalization of the conjugates may be considered a
benefit allowing the contemporary action both outside and inside cells.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Overall, the association of the two active modules either in the guise of a combination or as a
covalent conjugate is beneficial to gain antiꢀangiogenic effect in vitro. Which of the two options
is better has to be judged after in vivo antiꢀangiogenesis evaluation, which is able to measure the
putative targeting effect within the covalent conjugates.
Figure 7. In vitro inhibition of tubulogenesis in VEGFꢀA activated (20 ng/mL) EPCs seeded on
Matrigel and incubated for 6 h with compounds 1ꢀ3, c(AmpRGD) 8, sunitinib, and 8+sunitinib at
1 ꢀM (black), 0.1 ꢀM (dark grey), and 0.01 ꢀM (grey) concentrations. A) Representative images
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of the different treatments at 1 ꢀM concentration. B) Histograms refer to the inhibition of
branches development as compared to untreated cells and expressed as percentage.
Representative of three independent experiments.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We next measured the ability of conjugate 3 (as compared to 8, sunitinib and their combination)
to block angiogenesis in vivo using a Matrigel plug assay. We chose compound 3 for this study
as it had the strongest binding to EPCs, it was best internalized, it was the most effective in
inhibiting VEGFR2 phosphorylation and capillary tube formation. Despite the very preliminary
character of the data obtained, we found that in vivo angiogenesis is scarcely impaired by
unconjugated compound 8, while it is downregulated by sunitinib treatment (Figure 8). This
inhibition was comparable to that obtained in the coꢀtreatment, while the injection of compound
3
revealed a consistent reduction of in vivo angiogenesis. These results corroborate the
substantial role of conjugate 3 as an antiꢀangiogenic tool in vivo and substantiate the hypothesis
according to which the covalent conjugation of two key angiogenesisꢀrelated players (as in 3)
results in a synergic action, even superior to their simple combination.
Figure 8. Inhibition of in vivo angiogenesis in Matrigel plugs implanted in FVB mice. Matrigel
plugs contained VEGFꢀA/heparin+10 mg/kg sunitinib or equivalent quantity within 8 or 3
ACS Paragon Plus Environment
24

Page 25 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
compound 8, sunitinib, sunitinib+8, compound 3). Plugs were removed from mice and
photographed after 4 days.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
CONCLUSIONS
Three novel molecules, compounds 1ꢀ3, were efficiently synthesized and characterized, which
featured the robust covalent linkage of a sunitinibꢀlike portion to one or two cyclic aminoproline
RGD moieties. Subsequent biological investigations gave important clues about the relative
weight of the active modules within the conjugates. Overall, compound 3 seems to best
summarize the structural characteristics required for optimal biological response, where both the
RGD and the sunitinib modules may exert an active role. The preliminary, relevant antiꢀ
angiogenic effect of compound 3 in mice assesses its potential as an effective tool against tumorꢀ
associated angiogenesis and is even superior than the simple combination of the two discrete
modules. This work stands as a proofꢀofꢀconcept of how antiꢀangiogenic small molecules may be
selectively delivered to cells through simple, superꢀtargeted antiꢀangiogenic conjugated
molecules to be used in tumorꢀrelated or angiogenesisꢀrelated therapy.
EXPERIMENTAL SECTION
Chemistry. General. All chemicals were of the highest commercially available quality and were
used without further purification. Solvents were dried by standard procedures and reactions
requiring anhydrous conditions were performed under nitrogen or argon atmosphere. HꢀGlyꢀ2ꢀ
ClTrt resin (loading 0.63 mmol/g) was purchased from Novabiochem, (2S,4S)ꢀFmocꢀ4ꢀaminoꢀ1ꢀ
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Bocꢀpyrrolidineꢀ2ꢀcarboxylic acid (4) from PolyPeptide and all other reagents from Alfa Aesar,
TCI, or SigmaꢀAldrich. Flash column chromatography was performed using 40ꢀ63 ꢁm silica gel
using the indicated solvent mixtures. Automated flash column chromatography was carried out
with the Biotage Isolera One system using Biotage KPꢀSil cartridges (direct phase) or KPꢀC18ꢀ
HS (reverse phase). Melting points (mp) were measured with an optical Optiphot2ꢀPol thermoꢀ
microscope and are uncorrected. Optical rotations were measured using a PerkinꢀElmer model
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
341 polarimeter at ambient temperature using a 100 mm cell with a 1 mL capacity and are given
ꢀ1
2
ꢀ1
in units of 10 deg cm g . ESIꢀmass spectra were recorded on API 150EX apparatus and are
reported in the form of (m/z). HPLC purifications were performed on a Prostar 210 apparatus
(Varian, UV detection) equipped with C18ꢀ10 µm columns (Discovery BIO Wide Pore 10 × 250
mm or 21.2 × 250 mm). Routine NMR spectra were recorded on Avance 300 or 400 (Bruker)
NMR spectrometers. Chemical shifts (δ) are reported in parts per million (ppm) with TMS
(
CDCl ), CD HOD, and HOD resonance peaks set at 0, 3.31, and 4.80 ppm, respectively.
3 2
Multiplicities are indicated as s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), and b
1
13
(
broad). Coupling constants, J, are reported in Hertz. H and C NMR assignments are
corroborated by 1D and 2D experiments (gCOSY and gHSQC sequences). High resolution mass
analysis (ESI) was performed on LTQ ORBITRAP XL Thermo apparatus. Purity of all tested
compounds was determined by analytical highꢀpressure liquid chromatography (HPLC) and was
in the 96 ꢀ >99% range.
Materials. HꢀGlyꢀ2ꢀClTrt resin (loading 0.63 mmol/g), (2S,4S)ꢀFmocꢀ4ꢀaminoꢀ1ꢀBocꢀ
pyrrolidineꢀ2ꢀcarboxylic acid (4), FmocꢀAsp(tBu)ꢀOH; FmocꢀArg(Pmc)ꢀOH, 2,4,6ꢀcollidine,
piperidine, glacial acetic acid, 4ꢀpentynoic acid, 5ꢀfluoroisatin, ethyl 5ꢀformylꢀ2,4ꢀ
dimethylpyrroleꢀ3ꢀcarboxylate, triethylene glycol, NꢀBocꢀethylenediamine, propargyl bromide,
ACS Paragon Plus Environment
26

Page 27 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
acetaldehyde, 2ꢀ[2ꢀ(2ꢀaminoethoxy)ethoxy]ethanol, 3,5ꢀdihydroxybenzoic acid were
commercially available and were used as such without further purification. Sunitinib malate salt
was purchased by LC Laboratories (USA) with a purity of >99%.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of the c(AmpRGD) Azide 7. To a solution of linear tetrapeptide 6 (137 mg, 0.16
mmol, 1 equiv) in dry DCM (35 mL), 2,4,6ꢀcollidine (52 µL, 0.39 mmol, 2.5 equiv) was
added. The mixture was stirred under argon at room temperature, and added dropwise to a
solution of HATU (119 mg, 0.31 mmol, 2 equiv) and HOAt (43 mg, 0.31 mmol, 2 equiv) in
dry DMF (5mL) and dry DCM (30 mL). The reaction mixture was degassed by argon/vacuum
cycles (3 ×) and left to stir under argon at room temperature for 5 h. After completion, the
solution was concentrated under vacuum, treated with aq NaHCO saturated solution and
3
extracted with EtOAc (4 ×). The combined organic layers were dried with MgSO , filtered
4
and evaporated under reduced pressure, keeping the temperature under 50 °C. The crude was
purified by reverse phase flash chromatography [H O (0.1% TFA)/MeCN: linear gradient
2
8
0:20 to 20:80] furnishing the protected c(AmpRGD)ꢀN 7 (90 mg, yield 67%) as a white
3
2
5
1
solid; mp 118.6 °C; [α]D = + 19.2 (c 0.5; MeOH). H NMR (400 MHz, MeOD) δ 4.64 (dd, J
6.3, 5.6 Hz, 1H, Hα Asp), 4.44 (ddd, J = 6.9, 6.9, 2.0 Hz, 1H, H4), 4.16 (d, J = 13.9 Hz, 1H,
Hα Gly), 4.08 (t, J = 7.5 Hz, 1H, Hα Arg), 3.64 (d, J = 9.2 Hz, 1H, H2), 3.34 (d, J = 13.9 Hz,
=
1
9
2
1
H, Hα Gly), 3.30 (t, J = 6.6 Hz, 2H, Hδ Arg), 3.27ꢀ3.13 (m, 3H, H1′a,b + H5a), 3.04 (bd, J =
.6 Hz, 1H, H5b), 2.76ꢀ2.63 (m, 6H, Hβ Asp + CH Pmc + H4′a,b), 2.59 (s, 3H, CH Pmc),
2
3
.57 (s, 3H, CH Pmc), 2.41 (ddd, J = 13.6, 9.5, 7.1 Hz, 1H, H3a), 2.12 (s, 3H, CH Pmc),
3
3
.99 (d, J = 13.6 Hz ,1H, H3b), 1.86 (t, 2H, CH Pmc), 1.74ꢀ1.69 (m, 2H, Hβ Arg), 1.65ꢀ1.62
2
1
3
(
m, 6H, Hγ Arg + H2′ + H3′), 1.46 (s, 9H, tBu), 1.33 (s, 6H, CH Pmc). C NMR (75 MHz,
3
MeOD) δ 177.3 (Cq), 175.4 (Cq), 170.6 (Cq), 170.1 (Cq), 169.8 (Cq), 156.8 (Cq), 153.5 (Cq),
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 53
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
38.5 (Cq), 135.3 (Cq), 134.9 (Cq), 133.5 (Cq), 123.8 (Cq), 118.2 (Cq), 81.0 (Cq), 73.7
(CH ), 62.7 (CH ), 60.0 (CH ), 55.1 (CH), 52.3 (CH ), 51.0 (CH ), 49.8 (CH), 49.1 (CH),
2 2 2 2 2
44.4 (CH ), 40.1 (CH), 37.3 (CH ), 36.1 (CH ), 32.6 (CH ), 27.1 (3C, CH ), 26.9 (CH ), 26.5
2 2 2 2 3 2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(CH ), 25.8 (2C, CH ), 25.7 (CH ), 21.2 (CH ), 17.8 (CH ), 16.7 (CH ), 11.1 (CH ). HRMS
2 3 2 2 3 3 3
+
+
(ES ) C H N O S calcd for [M+H] 860.4447, found 860.4470.
39
61 11
9
Synthesis of the c(AmpRGD) Amine 8. The cyclic tetrapeptide 7 (29 mg, 0.03 mmol) was
dissolved in EtOH (4 mL) and a catalytic amount of 10% palladium on carbon was added. The
reaction vessel was degassed under vacuum and thoroughly purged with hydrogen (3 ×). The
resulting heterogeneous mixture was stirred overnight under hydrogen at room temperature, then
the catalyst was filtered off and the filtrate was concentrated under vacuum. The protected
intermediate AmpRGDꢀNH (27 mg, 0.03 mmol) was dissolved in 1.6 mL of a TFA/TIS/H O
2
2
(95:2.5:2.5) mixture and stirred at room temperature for 2.5 h. Then, the solvent was evaporated
and the crude residue was thoroughly washed with Et O (4 ×) and petroleum ether (2 ×).
2
Preparative RPꢀHPLC purification was performed [C ꢀ10 µm column, 21.2 × 250 mm; solvent
1
8
A: H O (0.1% TFA) and solvent B: MeCN, flow rate 8 mL/min; detection 220 nm] using a linear
2
gradient from 100% A to 25% B in 25 min. The removal of the solvent under vacuum, keeping
the temperature under 50 °C, furnished c(AmpRGD)ꢀNH 8 (18.4 mg, TFA salt, yield 91%) as a
2
2
5
1
colourless glassy solid; [α]D = –13.3 (c 1.0, H O). H NMR (400 MHz, D O) δ 4.60 (dd, J =
2
2
6.2, 6.2 Hz, 1H, Hα Asp), 4.52 (d, J = 10.8 Hz, 1H, H2), 4.26 (bt, J = 4.9 Hz, 1H, H4), 4.12 (dd,
J = 7.5, 7.5 Hz, 1H, Hα Arg), 3.94 (d, J = 13.9 Hz, 1H, Hα Gly), 3.93 (m, 1H, H5a), 3.37 (bd, J
=
8.8 Hz, 1H, H5b), 3.21 (m, 2H, H4′), 3.09 (t, J = 6.8 Hz, 2H, Hδ Arg), 2.90–2.81 (m, 3H, H1′
+
H3a), 2.80 (d, J = 5.9 Hz, 2H, Hβ Asp), 2.42 (bd, J = 15.2 Hz, 1H, H3b), 1.68–1.46 (m, 8H,
13
Hβ Arg + Hγ Arg + H2′ + H3′). C NMR (100 MHz, D O) δ 176.2 (Cq), 174.5 (Cq), 172.7 (Cq),
2
ACS Paragon Plus Environment
28

Page 29 of 53
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
71.2 (2C, Cq), 156.9 (Cq), 66.0 (CH), 60.2 (CH ), 56.1 (CH), 49.9 (CH), 44.6 (CH ), 40.6
2
2
(CH ), 38.8 (CH ), 35.1 (CH ), 35.0 (CH ), 34.9 (CH ), 26.5 (CH ), 24.6 (CH ), 23.8 (CH ), 22.3
2 2 2 2 2 2 2 2
+
+
(CH ). HRMS (ES ) C H N O calcd for [M+H] 512.2940, found 512.2932.
2
21 37
9
6
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Synthesis of compound 15. To a solution of 13 (94 mg, 0.31 mmol, 1 equiv) in dry DCM (1.5
mL) and dry DMF (2.5 mL), BOP (70 mg, 0.16 mmol, 1.3 equiv) and DIPEA (85 µL, 0.49
mmol, 4 equiv) were added and, after 5 min, a solution of 14 (103 mg, 0.41 mmol, 1.3 equiv) in
dry DCM (1.5 mL) and dry DMF (1 mL) was added. The reaction was stirred under argon at
room temperature, protected from light. A yellow precipitate started to form after 1 h. After 5 h
the reaction was over and Et O (3 mL) and petroleum ether (3 mL) were added to the mixture.
2
The solid was collected by vacuum filtration and washed with H O (3 ×) furnishing compound
2
1
15 (116 mg, yield 88%) as an orange solid; mp> 220 °C. H NMR (400 MHz, DMSOꢀd ) δ 13.69
6
(s, 1H, NH), 10.89 (s, 1H, NH), 8.02 (bt, J = 4.5 Hz, 1H, NH), 7.76 (dd, J = 9.5, 2.0 Hz, 1H,
H4), 7.72 (s, 1H, H1′), 7.60 (bt, J = 5.0 Hz, 1H, NH), 6.92 (ddd, J = 8.8, 8.8, 2.0 Hz, 1H, H6),
6.85 (dd, J = 9.2, 4.4 Hz, 1H, H7), 3.28 (m, 2H, CH ), 3.24 (m, 2H, CH ), 2.76 (t, J = 2.4 Hz,
2
2
≡
CH), 2.44 (s, 3H, CH ), 2.42 (s, 3H, CH ), 2.37 (m, 2H, CH ), 2.29 (t, J = 7.1 Hz, 2H, CH ).
3 3 2 2
1
3
1
C NMR (100 MHz, DMSOꢀd ) δ 171.1 (Cq), 170.0 (Cq), 165.4 (Cq), 158.6 (d, J = 233 Hz,
6
CF
3
Cq), 137.1 (Cq), 135.0 (Cq), 130.8 (Cq), 127.6 (d, J = 10 Hz, Cq), 126.3 (CH), 125.3 (Cq),
CF
2
3
2
1
21.0 (Cq), 115.1 (Cq), 112.8 (d, J = 25 Hz, CH), 110.4 (d, J = 8 Hz, CH), 106.3 (d, J =
CF CF CF
26 Hz, CH), 84.2 (Cq), 71.7 (CH), 46.6 (CH ), 34.8 (CH ), 19.3 (CH ), 14.7 (CH ), 13.8 (CH ),
2
2
2
2
3
+
+
1
1.0 (CH ). HRMS (ES ) C H FN O calcd for [M+H] 423,1827, found 423,1823.
3
23 23
4
3
Synthesis of compound 17. To a stirred suspension of 13 (11mg, 0.04 mmol, 1 equiv) in dry
DMF (500 µL), BOP (21 mg, 0.05 mmol, 1.3 equiv) and DIPEA (26 µL, 0.15 mmol, 4 equiv)
were added and, after 5 min, a solution of 16 (15 mg, 0.04 mmol, 1.1 equiv) in dry DCM (400
ACS Paragon Plus Environment
29