Thrombospondin-1 Mimetic Peptide Inhibitors of ... - ACS Publications

Mar 22, 2005 - Stephen F. Badylak,§ Richard R. Lesniewski, and Jack Henkin ... Abbott Park, Illinois 60064, and Department of Biomedical Engineering,...
0 downloads 0 Views 200KB Size
Download PDF
2838

J. Med. Chem. 2005, 48, 2838-2846

Thrombospondin-1 Mimetic Peptide Inhibitors of Angiogenesis and Tumor Growth: Design, Synthesis, and Optimization of Pharmacokinetics and Biological Activities† Fortuna Haviv,* Michael F. Bradley, Douglas M. Kalvin, Andrew J. Schneider, Donald J. Davidson, Sandra M. Majest, Laura M. McKay, Catherine J. Haskell, Randy L. Bell, Bach Nguyen, Kennan C. Marsh, Bruce W. Surber, John T. Uchic, James Ferrero, Yi-Chun Wang, Juan Leal, Rae D. Record,§ Jason Hodde,§ Stephen F. Badylak,§ Richard R. Lesniewski, and Jack Henkin Global Pharmaceutical Research and Development Organization of Abbott Laboratories, Abbott Laboratories, Abbott Park, Illinois 60064, and Department of Biomedical Engineering, Purdue University, West Lafayette, Indiana 47907 Received August 13, 2004

The heptapeptide 1, NAc-Gly-Val-DIle-Thr-Arg-Ile-ArgNHEt, a structurally modified fragment derived from the second type-1 repeat of thrombospondin-1 (TSP-1), is known to possess antiangiogenic activity. However, therapeutic utility could not be demonstrated because this peptide has a very short half-life in rodents. To optimize the PD/PK profile of 1, we initiated a systematic SAR study. The initial structural modifications were performed at positions 5 and 7 of peptide 1 and at the N- and C-termini. Out of several hundred peptides synthesized, the nonapeptide 5 (ABT-526) emerged as a promising lead. ABT-526 inhibited VEGF-induced HMVEC cell migration and tube formation in the nanomolar range and increased apoptosis of HUAEC cells. ABT-526 showed acceptable PK in rodents, dog, and monkey. ABT-526, when incorporated in an angiogenic pellet implanted in the rat cornea at 10 µM, reduced neovascularization by 92%. Substitution of DalloIle in place of DIle in ABT-526 provided nonapeptide 6 (ABT-510), which was 30-fold less active than ABT-526 in the EC migration but 20-fold more active in the tube formation assay. In comparison to ABT-526, ABT-510 has increased water solubility and slower clearance in dog and monkey. Radiolabeled ABT-510 demonstrated saturable binding to HMVEC cells at 0.02-20 nM concentrations and was displaceable by TSP1. ABT-510 and ABT-526 were shown to significantly increase apoptosis of HUAEC cells. ABT510 was effective in blocking neovascularization in the mouse Matrigel plug model and inhibited tumor growth in the mouse Lewis lung carcinoma model. Previous studies had shown that ABT-510 was effective in inhibiting the outgrowth of murine melanoma metastases in syngeneic mice and in blocking the growth of human bladder carcinoma implanted in nude mice. It had been also shown that ABT-510 could regress tumor lesions in pet dogs or cause unexpected stabilization of the disease in advanced canine cancer. ABT-526 and ABT-510 are the first compounds in the class of potent inhibitors of angiogenesis that mimic the antiangiogenic function of TSP-1. ABT-510 is currently in phase II clinical studies. Introduction Progressive growth, invasion, and metastases of solid tumors are critically dependent on the process of new blood vessel formation, known as angiogenesis.1,2 Tumors secrete a variety of inducers, such as VEGF, bFGF, and PDGF, that activate microvascular endothelial cells (EC) causing them to proliferate, migrate, and organize into capillary structures.3,4 Activated endothelial cells also enhance malignant progression by producing cytokines that inhibit programmed cell death (apoptosis).5 Thrombospondin-1 (TSP-1), a naturally occurring inhibitor of angiogenesis, is a large multifunctional glycoprotein secreted by most epithelial cells and is involved in the organization of the perivascular matrix.6 TSP-1 has been shown to block all the functions of activated EC and strongly mitigates tumor growth and † Part of this work was presented as a poster at the 93rd Annual Meeting of the American Association for Cancer Research, San Francisco, CA, April 6-10, 2002. Abstr. No. 904 and No. 3665. * To whom correspondence should be addressed. Phone: 847-9374829; fax: 847-938-3766; e-mail: [email protected]. § Purdue University.

metastases, while its absence enhances these effects.6,7 Expression of TSP-1 correlates inversely with malignant progression in melanoma, lung, and breast carcinoma.7 The antiangiogenic effect of TSP-1 could be very useful in cancer therapy.7,8 However, the large size, scarcity, and multiple biological activities of TSP-1 make its direct use impractical as a cancer therapeutic. Small peptide mimetics of TSP-1 could be an excellent alternative if they are potent and specific. It was previously reported by Dawson9 that the heptapeptide 1, NAc-GlyVal-DIle-Thr-Arg-Ile-ArgNHEt, a structurally modified proteolytic fragment derived from the second type-1 repeat in the stalk region of TSP-1, inhibited VEGFinduced EC migration in vitro, and decreased bFGFinduced neovascularization in a rat cornea model. However, since heptapeptide 1 undergoes rapid clearance in rodents (t1/2 < 1 min), it became critical for testing the usefulness in animal models and the development of a drug, to optimize pharmacodynamic/pharmacokinetic (PD/PK) profile of this TSP peptide. All compounds were first tested in vitro for inhibition of VEGF-induced migration of human microvascular cells

10.1021/jm0401560 CCC: $30.25 © 2005 American Chemical Society Published on Web 03/22/2005

Thrombospondin-1 Mimetic Peptide Inhibitors

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2839

Table 1. Structure Characterization and in Vitro Inhibition of Endothelial Cell Migration and Tube Formationa

compd

structure

MH+

1 2 3 4 5 (ABT-526) 6 (ABT-510) 7 8 9 10 11 TSP-1

NAc-Gly-Val-DIle-Thr-Arg-Ile-ArgNHEt NAc-Gly-Val-DIle-Thr-Arg-Ile-NvaNHEt NAc-Gly-Val-DIle-Thr-Nva-Ile-ArgNHEt NAc-Gly-Val-DIle-Thr-Nva-Ile-Arg-ProNHEt NAc-Sar-Gly-Val-DIle-Thr-Nva-Ile-Arg-ProNHEt NAc-Sar-Gly-Val-DalloIle-Thr-Nva-Ile-Arg-ProNHEt NAc-Sar-Gly-Val-DIle-Thr-Nva-Ile-Arg-ProNHIsp NSuccinyl-Sar-Gly-Val-DIle-Thr-Nva-Ile-Arg-ProNHEt NAc-Sar-Gly-Val-DIle-Thr-NMeNva-Ile-Arg-ProNHEt NSuccinyl-Sar-Gly-Val-DIle-Thr-NMeNva-Ile-Arg-ProNHEt NSuccinyl-Sar-Gly-Val-DIle-Thr-MNeNva-Ile-Arg-ProNHIsp

883.1 826.4 826.4 923.1 994.2 994.2 1008.3 1052.3 1008.2 1066.3 1080.3

tR,b

min

9.142 10.165 11.455 11.097 11.029 10.845 9.298 10.974 11.412 11.687 12.378

tR,c

min

18.341 22.385 23.154 22.505 22.565 22.393 22.930 22.017 22.452 22.843 23.291

inhibition of HMVEC migration IC50 (nM)

inhibition of HMVEC tube formation (nM)

>100 1.45 (( 0.380) 4.22 (( 0.500) 0.07 (( 0.004) 0.03 (( 0.015) 0.89 (( 0.029) 29 (( 0.003) 0.26 (( 0.110) 0.54 (( 0.150) 0.45 (( 0.110) 0.06 (( 0.030) 0.00035 (( 0.00027)

200-100 10-5.0 >1000 10-5.0 50-25 10-5.0 0.25-0.10 0.021-0.017

a Reported IC b 50 values are an average of two or more measurements. 5% to 100% over 55 min of acetonitrile/water containing 0.1% TFA. c 5% to 100% over 55 min of methanol/water containing 0.1% TFA.

(HMVEC). Selected compounds were then tested for inhibition of HMVEC tube formation and for induction of apoptosis. The most active compounds were then evaluated for rat clearance following iv administration, and the best were further tested for clearance in the mouse, dog, and monkey. Compounds with the best overall activity profile were then tested for inhibition of angiogenesis in vivo using a rat cornea model or a Matrigel plug mouse model. For inhibition of tumor growth in vivo, a syngeneic mouse model of Lewis lung carcinoma was used. Chemistry. All the peptides were synthesized from Fmoc-protected amino acids utilizing standard solid phase peptide synthesis (SPPS) methods. Fmoc-Proethylamide and Fmoc-Pro-isopropylamide resins, prepared by the method described in the Experimental Section, were used. The appropriate protected amino acids were sequentially coupled using HBTU or HATU as an activator. For the syntheses of peptides containing N-MeNva, HATU was used. The peptides were cleaved from the resin, along with the protecting groups, with (95:2.5:2.5) TFA/water/anisole. The crude products were purified using RP-HPLC. The final products were characterized by analytical RP-HPLC, mass spectrometry (MS), and amino acid analysis (AAA). All the tested compounds were trifluoroacetate salts and were at least 95% pure. A detailed NMR study of peptide 5 was performed and is described in the Experimental Section. Biological Assays. Inhibition of Endothelial Cell Migration. Studies of the inhibition of endothelial cell chemotatic migration have been broadly used to determine the activity of compounds as potential inhibitors of angiogenesis in vitro. Because of the many roles that cell migration plays in the pathological processes, a more rapid method that uses fluorescence to detect migrating HMVEC was developed.10,11 The assay details are described in the Experimental Section. Inhibition of Capillary Sprout Formation. To form functional vessels, activated endothelial cells must not only proliferate and migrate along a gradient of growth factors, but they also have to organize threedimensionally into vessels with true lumens. To examine the test peptide effect on this angiogenic function, we adopted an assay similar to that described by Nehls.12 The assay uses VEGF-induced capillary sprout formation from HMVEC cells grown to confluence on gelatin-coated microcarrier beads. The latter were suspended in fibrin gels containing VEGF with and without

Figure 1. Effect of ABT-510 (6) on apoptosis of HUAEC cells.

the test compound. After 72 h the cells proliferate and assemble radially into tubes, sometimes connecting nearby beads to each other. Secondary sprouting leads to a typical branched appearance. After 96 h the beads are observed microscopically and scored visually for cell proliferation, tube formation and branching. The inhibition of tube formation is reported as the lowest concentration or range that abrogates sprouting from 90% of the examined beads (Table 1). Induction of Microvascular Endothelial Cell Apoptosis. Compound 6 was tested for its ability to induce apoptosis in primary human umbilical arterial endothelial cells (HUAEC). Its effect on the apoptotic status of the cells was measured by the method described by Bonfoco.13 Compound concentrations ranging from 0.1 to 100 nM produced apoptotic indices (ratio of treated over control) ranging from 1.35 to 1.75 (Figure 1). These results are in line with previous data reported by Reiher14 that ABT-510, in cultured bovine capillary endothelial cells, increased apoptosis to a level similar to that induced by TSP-1 in the same cells. In Vivo Angiogenesis Models. We used a rat cornea model to test ABT-526 in vivo. This model consists of generating neovascularization from the limbus of the rat eye by cutting a micropocket in the corneal stroma and implanting a pellet containing bFGF and the test compound mixed with sucralfate (to stabilize the growth factor) and Hydron (to slow the release of the growth factor) as described by Fournier.15 After 7 days the neovascularization density was measured. The data

2840

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8

Haviv et al.

Table 2. Pharmacokinetics in the Rat, Mouse, Dog, and Monkey Following 4.5 mg/kg Dose iv compd (species) 4 5

6

7 8 9 10 11 a

(rat) (ABT-526) (rat) (mouse)a (dog) (monkey) (ABT-510) (rat) mouse) (dog) (monkey) (rat) (rat) (rat) (rat) (rat)

t1/2, h

AUC, µg‚h/mL

Cl, L/h‚kg

0.3

4.63

0.97

0.20 1.20 0.5 0.5

3.56 1.99 10.66 3.20

1.27 2.51 0.42 1.41

0.20 0.15 0.80 1.20 0.90 0.20 1.00 2.10 0.3

5.30 1.47 19.27 14.77 9.40 8.37 6.69 8.38 9.84

0.86 3.06 0.23 0.35 0.49 0.54 0.68 0.54 0.46

5 mg/kg dose.

presented in Figure 1 are the average of readings from six corneas. ABT-510 was tested in a mouse Matrigel plug model, another model for testing angiogenesis in vivo. This test was run according to Passaniti16 and is described in the Experimental Section. Inhibition of Tumor Growth in Vivo. ABT-510 was tested in a mouse syngeneic tumor model using the Lewis lung carcinoma cell line. A fixed number of tumor cells was implanted in the flank of C57 mice. Three days post-inoculation, mice with palpable tumors were divided in four groups, consisting of a control and three treatment groups for the doses of 30, 60, and 90 mg/ kg/day. These doses were continuously administered subcutaneously (sc) by osmotic minipumps (omp). Dosing was continued for two weeks and tumor volume was measured every 2-3 days. Sixteen days post-inoculation, before the tumors approached gram size, mice were sacrificed and the final tumor volume was measured.17 Results and Discussion To improve the PD/PK profile of the heptapeptide 1, NAcGly-Val-DIle-Thr-Arg-Ile-ArgNHEt,9 and to eliminate a possible histamine release propensity due to the close proximity of the two arginines,18 we attempted to replace one of these two basic residues with a neutral one. Substitution of norvaline, and other neutral amino acids, for the C-terminal arginine produced peptides that inhibited the VEGF-induced migration of endothelial cells at concentrations in the single digit nanomolar range (2, Table 1). Likewise, substitution of norvaline for the internal arginine (3) maintained the migration inhibition potency in the same range. Insertion of a proline at the C-terminus, a residue common in all GnRH analogues which are known to be metabolically stable,19 caused a 60-fold increase in inhibition of migration, placing octapeptide 4 in the mid picomolar range. In the rat octapeptide 4, administered at 4.5 mg/ kg intravenously (iv), showed a sustained PK with a half-life of 0.3 h, much improved over heptapeptide 1 (t1/2 < 1 min) (Table 2). To stabilize the N-terminus of 4, we inserted a sarcosine in front of the glycine generating nonapeptide 5 (ABT-526).14,20 This insertion caused a 2-fold improvement in inhibition of migration and showed a moderate inhibition of EC tube formation at 200-100 nM (Table 1). ABT-526 showed acceptable PK in rat, mouse, dog, and monkey with a half-life of

Figure 2. TSP-1 competes with [3H] ABT-510 (0.2 nM) for binding sites on HMVEC cells.

0.2, 2.0, 0.5, and 0.5 h, respectively (Table 2). To ensure the safety of ABT-526, nonapeptide 5 was tested in the rat peritoneal mast cell histamine release assay along with leuprolide, a GnRH agonist in use for prostate cancer,19 and with Nal-Glu, a known histamine releaser GnRH antagonist (ED50 1 µg/mL).18 In this assay ABT526 has ED50 > 3.3 mg/mL, compared to leuprolide with ED50 100-300 µg/mL. Since the DIle residue, which is present at position 4 of ABT-526, has the configurations of the R- and β-carbons inverted relative to the natural amino acid Ile ([R]R, [R]β versus [S]R, [S]β), we attempted to minimize the structural diversity by allowing inversion of configuration only at the R-carbon. This led us to substitute DalloIle in place of DIle ([R]R, [S]β versus [R]R, [R]β) providing nonapeptide 6 (ABT-510),14,20 which was 30-fold less active in migration, but was 20fold more active in the inhibition of tube formation (Table 1). The basis for the lack of correlation between the activities in the migration and tube formation assays is unclear. However, it is important to realize that the migration test is run for 4 h, while the tube formation test is run for 4 days. The large difference in the duration of the tests may possibly affect metabolic stability and/or permeability resulting in activity differences. ABT-510 increased apoptosis of endothelial cells by an apparent maximum of 75% at 1, 10 and 100 nM (Figure 1).21 To demonstrate that ABT-510 is a true mimetic of TSP-1 and occupies the same binding site, we tested this peptide for competition binding on endothelial cells.22 As shown in Figure 2, TSP-1 at 0.220 nM concentrations displaced radiolabeled [3H]-Nacetyl-ABT-510 from its binding sites on HMVEC in a dose-dependent fashion, consistent with our initial hypothesis that it shares a receptor in common with TSP-1. For formulation purposes, the solubility of ABT-526 and ABT-510 was studied. It was found that ABT-510 is more soluble than ABT-526. The solubility of ABT510 in D5W at pH 5.43 is 140 mg/mL compared to 45.4 mg/mL for ABT-526. It is quite remarkable and striking that inversion of configuration of just one chiral center in the side chain of one amino acid of a nonapeptide has such a pronounced effect on solubility. We also found that the PK of ABT-510 in the dog and monkey were significantly improved over ABT-526 (Table 2, Figure 3), as reflected in an over 2-fold increase in halflife and two to three-fold decrease in clearance parameter. Since the allometric scaling analysis of the likely

Thrombospondin-1 Mimetic Peptide Inhibitors

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2841

Scheme 1. Key Structures in the Design Process of ABT-510

PK in human weighed the PK of primate more heavily than that of dog or rodents, ABT-510 was advanced to further preclinical studies and then to human clinical development. Scheme 1 illustrates the milestones in the structure optimization starting from TSP-1 fragment 1 (IC50 > 100 nM) and ending with ABT-510 (IC50 0.9 nM), a clinical candidate. As a part of our attempt to optimize the PK/PD profile of our peptides, we continued focusing on structural modifications at the N- and C-termini. Substitution of isopropyl for the ethyl residue at the C-terminus of ABT-

Figure 3. Mean (( SEM, n ) 3) plasma concentrations of ABT-526 (5) and ABT-510 (6) after a 4.5 mg/kg iv dose in monkey.

526 caused a large loss in activity in both the migration and tube formation assays. However, it remarkably improved the rat PK (7, Tables 1 and 2), as reflected by an over 4-fold increase in half-life and 2-fold decrease in clearance parameter. Substitution of succinyl for acetyl at the N-terminus restored most of the activity in migration and tube formation assays and maintained the improved clearance (8, Tables 1 and 2). Metabolism studies of ABT-526 in the rat, 1 h after iv administration of radiolabled [14C] N-acetyl-ABT-526, showed the presence of unchanged ABT-526 (70.2%) along with one major metabolite (M - 1) (16.1%) in the plasma. The (M - 1) metabolite was identified as NAcSar-Gly-Val-DIle-Thr-OH, presumably deriving from a hydrolytic cleavage of the Thr-Nva bond. To stabilize this scissile bond we substituted N-methyl-norvaline (NMeNva) at position 6 (9), resulting in a 18-fold reduction in inhibition of migration and a 4-fold increase in inhibition of tube formation, compared to ABT-526. Compound 9 had a half-life and clearance modestly improved over ABT-526 (Table 2). Combination of the last two structural modifications generated compound 10 which, in comparison to ABT-526, was 15-fold less potent in the inhibition of migration, but was 20-fold more potent in the tube formation assay. In the rat, compound 10 showed improved PK with more than 10fold increase in half-life and 2-fold decrease in clearance parameter (Table 2). Combination of the last thee modifications (11, Table 1) decreased the migration inhibition potency by 2-fold but boosted the inhibition potency of tube formation by 800-fold and decreased the

2842 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8

Haviv et al.

Figure 4. Effect of ABT-526 on bFGF-induced neovascularization in rat cornea model.

Figure 6. Effect of ABT-510 on tumor growth in the Lewis lung carcinoma model following administration by omp.

types of histologically confirmed and refractory cancers and having measurable lesions, received 0.5 mg/kg of ABT-510 sc bid, with no other concurrent cancer therapy. Unexpected stabilization of disease was observed in approximately 10% of the animals and objective responses were observed in tumor lesions in approximately another 10% of the dogs. Conclusion Figure 5. Effect of ABT-510 on hemoglobin content in the mouse Matrigel plug model following administration by omp.

clearance parameter by 2-fold compared to that of ABT526 (Table 2). To demonstrate inhibition of angiogenesis in vivo and to probe the potential use of ABT-526 and ABT-510 as antiangiogenic therapeutics, we utilized the rat corneal neovascularization assay15 and the Matrigel plug mouse model.16 Local administration of 10 µM of ABT-526, contained within an angiogenic pellet implanted in the rat cornea, reduced bFGF-induced neovascularization by 92% (Figure 4). In the Matrigel plug mouse model a dose of 30 mg/kg/day delivered continuously by an osmotic minipump (omp), reduced neovascularization by 53% (Figure 5). To demonstrate inhibition of tumor growth, ABT-510 was tested in the Lewis lung carcinoma in the mouse at 30, 60, and 90 mg/kg/day, by omp.17 Statistically significant inhibition of tumor volumes was observed with all the three doses as shown in Figure 6. It was previously shown by Reiher14 that systemic delivery of ABT-526 inhibited the growth of murine melanoma metastases in syngeneic mice. He also demonstrated14 that ABT-510 inhibited the growth of primary human bladder cancer cells implanted orthotopically in the bladder of immune-deficient mice. Viloria-Petit23 reported that sc administration of ABT526, 30 mg/kg bid, to 528ras1 fibrosarcoma SCID mice caused over 70% inhibition of tumor volume. To examine the safety and efficacy of these TSP-1 mimetics as antiangiogenic therapy, a prospective preclinical trial in pet dogs with naturally occurring cancers was carried out and reported by Khanna.24,25 Pet dogs, with various

Systematic structural variations of heptapeptide 1, a modified fragment of TSP-1, led to optimization of the PK/PD profile resulting in the discovery of two nonapeptides ABT-526 and ABT-510. ABT-510 has excellent water solubility and acceptable PK in rat, mouse, dog, and monkey. ABT-526 and ABT-510 are potent inhibitors of angiogenesis in vitro. Competition binding studies showed that TSP-1 and ABT-510 share common binding sites on HMVEC cells. Both compounds were also shown to inhibit angiogenesis in vivo. ABT-510 significantly inhibited tumor growth in the Lewis lung mouse carcinoma model. It was previously demonstrated14 that ABT-526 inhibited the growth of murine melanoma metastases in syngeneic mice and that ABT510 blocked the orthotopic growth of a human bladder carcinoma implanted in nude mice. Additionally, ABT510 was reported to block progression or cause regression of tumor growth in pet dogs.24,25 ABT-526 and ABT510 are the first in the class of potent inhibitors of angiogenesis that mimic the antiangiogenic function of TSP-1 while avoiding the potentially deleterious side effects associated with the other functional domains of this glycoprotein. ABT-510 is currently in phase II clinical studies.26,27 Experimental Section All the peptides were synthesized using a Symphony automated peptide synthesizer (Protein Technology Inc., Woburn, MA). The peptides purification was performed using a Gilson HPLC system equipped with an automated liquid handler. Mass spectra were recorded using either a Finnigan SSQ7000 (ESI) or JEOL JMS-SX102A-Hybrid (FAB) mass spectrometers. 1H NMR spectra were recorded at 500 MHz using a Varian UNITY 500 NMR spectrometer. 13C NMR

Thrombospondin-1 Mimetic Peptide Inhibitors

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2843

spectra were recorded at 125 MHz using a Varian UNITY 500 NMR spectrometer. Amino acid analyses (AAA) were performed on Beckman Model 6300 Amino Acid Analyzer using ninhydrin derivatization. The peptides were hydrolyzed with 6 N HCl containing 0.5% phenol at 150 °C for 2 h. The data handling system was PE Nelson ACCESS CHOM. For calibration Beckman standards were used. The values for Thr were generally low because of partial decomposition. The content of other amino acids was generally within (10%. Table of AAA is available in the Supporting Information. The Fmoc-protected amino acids and resins were purchased either from Calbiochem-Novabiochem Corp. (San Diego, CA) or from Bachem Inc. (Torrance, CA). HBTU and HATU were purchased from Aldrich Co. (Milwaukee, WI). Solvents and reagents were used as supplied by commercial sources. Ethylamine Resin. 4-Formyl-3-methoxy-phenoxybutyramide polystyrene resin (5 g, 0.54 mmol/g) was placed into a manual SPPS reaction vessel and suspended in a 90/10 (v:v) DMA/acetic acid solution (200 mL). The resin was gently shaken for 5 min at room temperature. The vessel was drained and the resin was washed again with 90/10 (v:v) DMA/acetic acid (2 × 5 min). The resin was drained again and suspended in a minimal amount of 90/10 (v:v) DMA/acetic acid. To the suspension was added one spoon of preactivated 4 Å molecular sieves, followed by a 2 M solution of ethylamine/THF (6.75 mL, 135 mmol). The suspension was shaken for 1 h at room temperature, and to the slurry was added sodium triacetoxyboronhydride (2.86 g, 135 mmol) in a solution of 90/10 (v:v) DMA/acetic acid. The resin was shaken for 2 h, drained, and washed several times as follows: with DMA (×3, 5 min each), methanol (×5, 5 min each), DMA (×3, 5 min each), methanol (×3, 5 min each), dichloromethane (×3, 5 min each), methanol (×3, 5 min each), and diethyl ether (×3, 5 min each). The resin was drained and dried in vacuo at room temperature. The molecular sieves were manually separated. A sample of the resin was subjected to gel-filled 13C NMR analysis. The results of the gel-filled 13C NMR analysis indicated the complete disappearance of the carbon peak at 188.1 ppm, corresponding to the aldehyde carbon, along with the appearance of new peaks at 48.6 ppm (CH2-phenyl), at 43.2 ppm (CH2), and at 15.3 ppm (CH3). An additional peak was observed at 65.8 ppm, which corresponds to a methylene of benzylic alcohol, which had resulted from the reduction of the starting aldehyde. Approximately 10-15% of the resin contained this alcohol byproduct. However, subsequent peptides prepared from this resin did not result in a significant amount of the acid-peptide. Generally, upon purification by HPLC, any small amount of acid-peptide, if present, was readily separated from the ethylamide-peptide. Isopropylamine Resin. A procedure similar to the one described for ethylamine resin was adopted except isopropylamine was substituted for ethylamine/THF. The results of gelfilled 13C NMR analysis indicated the complete disappearance of the carbon peak of the aldehyde at 188.1 ppm and the absence of the benzylic alcohol peak at 65.8 ppm. The predominant carbon peaks, present at 22.6 ppm (CH3), 47.2 ppm (CH), and 46.1 ppm (CH2-phenyl), corresponded to the desired amine product. Fmoc-Pro-ethylamide Resin. Ethylamine resin (3.203 g, 0.54 mmol/g) was placed in a manual peptide synthesis flask and suspended in DMA and shaken for 5 min at room temperature using a manual or automated shaker. The resin was drained, washed again with DMA (×2, 5 min each), and resuspended in the minimal amount of DMA. To the resinsuspension was added diisopropylethylamine (0.5 mL, 1.7 mmol). Separately, a DMA solution (15 mL) containing FmocPro (1.75 g, 5.19 mmol), HATU (1.97 g, 5.19 mmol), and diisopropylethylamine (1.81 mL, 10.4 mmol) was prepared and kept at room temperature for 5 min for preactivation. The solution of the activated amino acid was then added to the resin suspension and shaken overnight at room temperature. The resin was drained and washed as follows: DMA (×3, 5 min each), methanol (×5, 5 min each), DMA (×3, 5 min each), methanol (×3, 5 min each), dichloromethane (×3, 5 min each),

methanol (×3, 5 min each), and diethyl ether (×3, 5 min each). The resin was drained and dried in vacuo. A sample of the resin was subjected to quantitative Fmoc-chromophore analysis to determine the substitution. The method used for determining this was described by Meienhofer.28 The analysis indicated a resin substitution 0.375 mmol/g. The dried resin was then resuspended in DMA (5 min), shaken, drained, washed with DMA (×2, 5 min each) and treated for 1 h at with a 80/10/10 (v:v:v) solution of DMA/pyridine/acetic anhydride. The resin was drained and washed again as follows: DMA (×3, 5 min each); methanol (×5, 5 min each); dichloromethane (×3, 5 min each); methanol (×3, 5 min each); and diethyl ether (×3, 5 min each). The resin was dried in-vacuo and stored for subsequent peptide synthesis. Fmoc-Pro-isopropylamide Resin. A procedure similar to the one described for Fmoc-Pro-ethylamine resin was used except substituting isopropylamine resin for ethylamine resin. General Synthetic Method and Purification of 1-11. The synthesis of 5, NAc-Sar-Gly-Val-DIle-Thr-Nva-Ile-ArgProNHEt, is described as a general example and was started from the peptide C-terminus with Fmoc-Pro (0.1 mmol, substitution varied from 0.3 to 0.6 mmol/g) already attached to the ethylamine resin, prepared as previously described. The following protected amino acids were used and sequentially coupled: Fmoc-Arg(Pmc) or Fmoc-Arg(Pbf), Fmoc-Ile, FmocNva, Fmoc-Thr(t-Bu), Fmoc-DIle, Fmoc-Val, Fmoc-Gly, FmocSar, and acetic acid for the capping of the N-terminus. The resin was washed with DMF at the beginning of the synthesis and after each step (×3, 2.5 min each) and then treated with a solution of 20% piperidine in DMF for 50 min, to remove the Fmoc-protecting group, and washed again. A solution of 0.3 M DMF solution of the next Fmoc-amino acid was added, followed by 0.3 M DMF solution of HBTU or HATU containing 0.4 M of NMM, mixed for 40 min, and washed. The peptide resin was deprotected, washed, and coupled to the next amino acid as described previously. The final peptide resin was washed and treated with (95:2.5:2.5) TFA/H2O/anisole for 3 h to cleave the peptide from the resin and the side chain protecting groups. The peptide solution was separated from the resin and concentrated in vacuo. The residue was purified by HPLC using a Platinum EPS C18 column (2.2 × 25 cm) (100 Å, 5 µm) and a solvent mixture varying in a gradient of 10% to 50% acetonitrile-water (containing 0.1% TFA) over 20 min. The pure fractions were lyophilized, and then the product was characterized by MS, analytical HPLC, and AAA. The data for MS and HPLC are listed in Table 1 and AAA data are included in the Supporting Information. As an example the 1H and 13C NMR spectra of 6 was studied. The 1H and 13C resonances were assigned from the DQCOSY, ROESY, TOCSY, HMQC, and HMBC spectra. The assignments are listed in Table 3 and position numbers are as marked in Scheme 1. The 13C resonances were assigned relative to DMSO-d6 and the 1H resonances were assigned relative to TMS. There are additional peaks present in both the 1H and 13C spectra due to the presence of rotamers. In the ROSEY spectrum exchange cross-peaks are observed indicating the presence of rotamers. Additional evidence comes from the spectra taken at the higher temperatures, 60-120 °C, showing the peaks in the process are coalescing. Synthesis of Radiolabeled [14C] N-Ac-Sar-Gly-Val-DIleThr-Nva-Ile-Arg-ProNHEt (5). An aqueous solution (1.25 mL, 4.4 µmol, 250 µCi) of radiolabeled [14C]-sodium acetate (Amersham Pharmacia Biotech, Piscataway, NJ) was lyophilized. To the dry lyophilized sample was added DMF (0.3 mL) followed by 4 N HCl/dioxane solution (1 µL) and glacial acetic acid (20 µmol). To this solution was added another DMF solution (2 mL), previously prepared and containing HBTU (25 µmol), HOBt (25 µmol) and diisopropylamine (50 µmol). The obtained solution was added to the peptide-resin H-SarGly-Val-DIle-Thr(OtBu)-Nva-Ile-Arg(Pmc)-Pro-NHEt-resin (20 µmol), previously prepared using the SPPS described above. The mixture was shaken for 16 h at room temperature. The resin was filtered and washed ×3 with DMF (20 mL) and ×3 with THF (10 mL). The resin was dried under nitrogen. A

2844

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8

Table 3. 1H and positiona 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 a

13C

NMR Spectral Data of Peptide 5 13C

(ppm)

21.3, 21.2 170.5 36.9, 34.0 52.8, 50.1 168.5 168.7 41.9 168.8, 168.7 57.8, 57.7 30.7, 30.6 19.2, 17.9 170.9 57.3 36.2 15.4 24.2 10.9 171.1 58.1 66.5 19.3 169.7 52.5 33.7 18.4 13.6 171.4 56.8 36.5 15.2 24.3 10.8 170.7 50.0 27.9 24.5 40.4 157.2 169.5 59.8 29.3 24.4 46.8 171.1 33.3 14.6

1H

(ppm)

2.03, 1.94 3.20, 2.78 4.01, 3.97, 3.92 8.34, 8.16 3.84, 3.79 7.90, 7.79 4.33, 4.30 2.00, 1.99 0.84 8.24, 8.20 4.28, 4.27 1.82 0.84 1.43, 1.14 0.80 0.80 7.94 4.26 4.00 1.03 7.97 4.29 1.64, 1.56 1.32, 1.25 0.84 7.81 4.18 1.71 0.80 1.41, 1.08 0.79 8.03 4.47 1.78, 1.61 1.53 3.12 8.00 7.48, 7.19 4.20 2.01, 1.71 1.93, 1.82 3.65 7.84 3.06 0.99

Atom positions as marked in Scheme 1.

solution containing thioanisole (50 µL), water (25 µL), ethanedithiol (25 µL), and TFA (0.9 mL) was added to the dry peptide-resin, and the mixture was shaken for 2.5 h at room temperature. The resin was separated, and to the solution was added cold diethyl ether (15 mL). The desired product precipitated. The mixture was centrifuged, the supernatant was manually removed, and the solid peptide was washed ×3 with cold diethyl ether. The air-dried peptide was dissolved in water (5 mL) and lyophilized. The crude peptide was purified by HPLC using a Waters Symmetry C18-column (7.8 × 300 mm, 7 µm) and a solvent mixture varying in a gradient of 10% to 40% acetonitrile/water (containing 0.01% TFA) over 20 min. The pure fractions were collected and lyophilized to yield [14C] N-Ac-Sar-Gly-Val-DIle-Thr-Nva-Ile-Arg-ProNHEt (8 mg, 80 µCi). The obtained [14C]-labeled product was co-injected on HPLC (Waters C-18 analytical column 4.6 × 150 mm) with a

Haviv et al. sample of the cold peptide 5 to give one single peak with tR ) 22 min. Synthesis of Radiolabeled [3H] N-Ac-Sar-Gly-Val-DalloIle-Thr-Nva-Ile-Arg-ProNHEt (6). To a solution of N-AcSar-Gly-Val-DalloIle-Thr-Allylgly-Ile-Arg-ProNHEt (2 mg, 0.002 mmol), prepared by the procedure described above, was added 10% Pd/C (2 mg), and the mixture was vigorously stirred under an atmosphere of tritium gas (1.35 Ci, 0.0233 mmol) for 17 h. The catalyst was filtered, the filtrate was concentrated in vacuo, and the residue was dissolved in methanol (×2, 5 mL) and again concentrated in vacuo to remove residues of tritium. The crude product was purified by HPLC using a Phenomenex Luna C-18 column (250 × 4.6 mm, 5 µm) and an isocratic mobile phase of 27% acetonitrile in water containing 0.1% TFA. The fraction containing the pure tritiated product, [3H] N-Ac-Sar-Gly-Val-DalloIle-Thr-Nva-Ile-Arg-ProNHEt, was collected, dried, and redissolved in ethanol (2 mL). Co-injection of the radiolabeled peptide with a sample of cold peptide 6 on the HPLC column (Waters C18 analytical column 4.6 × 150 mm, solvent gradient as described above) gave one single peak. The specific activity was 57 Ci/mmol as determined by measuring the radioactivity concentration with liquid scintillation counting and the mass concentration by comparison of the HPLC peak area to a standard curve. In Vitro Biological Assays. Inhibition of Endothelial Cell Migration. A method similar to Frevert was adopted.11 A primary line of human microvascular endothelial cells (HMVEC) were grown in EBM-2-MV media (BioWhittaker). Upon reaching 80-90% confluency the cells were washed with PBS (GIBCO) and placed into EBM media (no supplements except 5% FCS) containing 1% penicillin-streptomycin solution and left for 18 h. The cells were then washed with PBS, trypsinized, and resuspended in Hanks (no Ca2+) containing 0.1% BSA at a concentration of (2-5) × 106 cells/mL. The cells were incubated with 50 ng of calcein AM (Molecular Probes) for 30 min at room temperature in the dark. The cells were then centrifuged and resuspended in EBM-2 media containing 1% penicillin-streptomycin and 0.1% BSA. A 96-well filter plate (NeuroProbe # 101-8) was used. VEGF (10 ng, R&D Systems) was prepared in the media and added to the bottom wells. The filter was put in place, according to manufacture instructions, and cells with and without the test compound were added to the top. All compounds were tested as solutions in D5W. The plates were incubated at 37 °C in a 5% CO2 incubator. After 4 h, the plates were removed, the free cells (unmigrated) were wiped from the top filter and the plate read in a FLx800 BioTek fluoroscan at 485 excitation and 530 emission. The percent of inhibition was calculated according to the following equation:

100 -

(fluorescence reading of treated cells × 100) ) fluorescence reading of positive control % inhibition

Competition Binding of Tritium-Labeled ABT-510 and TSP-1 to HMVEC Cells. Tritium-labeled ABT-510 (1 nM) and various concentrations of TSP-1 (0-20 nM) were added to eppendorf tubes each containing 100 000 HMVEC cells that had been grown in full media and scraped from T175 flasks at about 80% confluency. The tubes were mixed and incubated for 2 h on ice. The number of counts remaining bound to the cells after extensive washing determined total amount of ABT510 bound.22 The Kd of [3H]-ABT-510 is 0.2 nM as determined from saturation binding experiments (data not shown). A competitive binding experiment with TSP-1 and [3H]-ABT-510 shows similar Kd values for both compounds. HUAEC Apoptosis. The effect of ABT-526 and ABT-510 on HUAEC cells apoptosis was determined using a histone ELISA apoptosis assay (Roche, Indianapolis, IN). Five thousand cells per well were plated in 96-well CoStar tissue culture plates. Cells were allowed to adhere, and the test peptides, as D5W solutions, were added to the wells and incubated overnight. Apoptosis was determined from triplicate samples, and the apoptotic index was determined as a ratio of absorbance of treated cells over absorbance of untreated cells.21

Thrombospondin-1 Mimetic Peptide Inhibitors

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8 2845

Biological Assays in Vivo. Rat Cornea Model. Briefly, both corneas of Sprague Dawley rats were implanted on day 0 with a pellet containing 50 ng of bFGF mixed with sucralfate and Hydron.15 The implanted pellet contained also 0 (for control), 1, 5, and 10 µM of compound 5 (ABT-526) in D5W solution. Five corneas were used for each dose. At day 7 the cornea neovascularization of the rat eyes was observed through a slit lamp. An image analysis system was used to record the image and to measure the density of the neovascularization. The results are expressed as a mean ( SEM of vessel density and are average of the readings from six corneas (Figure 4). Matrigel Plug Model. This assay used a procedure similar to that described by Passaniti.16 High protein Matrigel (0.20 mg/mL) (BD Biosciences, Bedford MA) was thawed overnight at 4 °C on ice. The next day, 500 ng/mL of bFGF and 50 ng/ mL of VEGF (R&D Systems, Minneapolis MN) were added and allowed to mix on ice for 2 h. Six to eight week old female SV129 mice (Charles River Laboratories, Wilmington MA) of 25 g average weight were anesthetized with a ketamine/ xylazine mixture and injected with 0.5 mL of cold Matrigel in the mid-abdominal region using a 251/2 G syringe. Immediately following implantation of the plugs, seven-day osmotic minipumps (Alzet Corporation, Cupertino CA), with a flow rate of 1.06 ( 0.03 µL/h, were implanted subcutaneously. The pumps were preloaded with drugs and allowed to equilibrate for 4 h at 37 °C before implantation. ABT-510 was solubilized in D5W at various concentrations: 30 mg/mL for the 30 mg/kg/day omp dose; 10 mg/mL for the 10 mg/kg/day, omp dose; 3 mg/mL for the 3 mg/kg/day, omp dose, and 0 mg/ mL for the vehicle-treated group (control). Seven days after, the plugs were excised and assayed for hemoglobin content using the Total Hemoglobin Kit (Sigma Chemical Co., St. Louis, MO). The plugs were homogenized in Drabkin’s reagent, spun at 1000 rpm, transferred into 1.5 mL eppendorf tubes, and respun at 14 000 rpm. The supernatant (200 µL) was transferred into a 96-well plate, and its absorbance was read at 540 nm. The absorbance for each sample was compared against a hemoglobin standard (Sigma Chemical Co., St. Louis, MO) to determine heme content (g/ dL) in each sample. Lewis Lung Model. This assay was run similar to that described by McDonnell.17 Six to eight weeks old female C57BL/6 mice were inoculated sc into the right flank with 0.5 × 106/0.1 mL LLC cells in a (1:1) mixture of Matrigel (Collaborative Biomedical Products, Bedford MA) and 1 × PBS. Three days later, the mice were randomized into two groups of ten and anesthetized with xylazine:ketamine mixture, and osmotic mini-pumps (Alzet Co., Cupertino CA) were implanted sc. The test compound, ABT-510, was dissolved in D5W and loaded into the pumps to deliver 30, 60, and 90 mg/kg/day for 14-consecutive days following implantation. Tumor measurements were taken with electronic calipers every 2-3 days starting on day 7. The study was terminated 16-days postinoculation. Tumor volumes were calculated for each group (N ) 10) according to the following formula: V ) L × W2/2 (wherein V ) volume; L ) length; W ) width).

tetramethyluronium hexafluorophosphate; HMVEC, human microvascular endothelial cells; HPLC, highpressure liquid chomatography; HUAEC, human umbilicus arterial endothelial cells; LLC, Lewis lung carcinoma; mkd, mg/kg/day; MS, mass spectrum; NMM, N-methylpiperidine; NMR, nuclear magnetic resonance spectrum; omp, osmotic mini-pump; PD, pharmacodynamics; PDGF, placenta derived growth factor; PK, pharmacokinetics; qd, once a day; rt, room temperature; sc, subcutaneous; SPPS, solid-phase peptide synthesis; VEGF, vascular endothelial growth factor.

Acknowledgment. The authors acknowledge the technical support given by Thomas Pagano, David Whittern, Jan Waters, and Darlene Hepp from Abbott Structural Chemistry Department. The authors also acknowledge Sarah Dorwin, from Abbott Structural Biology Department, for performing the amino acid analyses. Appendix Abbreviations: AAA, amino acid analysis; bFGF, basic fibroblast growth factor; bid, twice a day; BSA, bovine serum albumin; D5W, 5% dextrose in water; EC, microvascular endothelial cell; HATU, O-(7-azabenzotrialzol-1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate; HBTU, 2-(1H-benzotriazole-1-yl)-1,3,3-

Supporting Information Available: Amino acid analysis data. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Folkman, J. Tumor angiogenesis. In The Molecular Basis of Cancer; Mendelson, J., Holwy, P. M., Israel, M. A., Liotta, L. A., Eds.; Saunders: W. B. Co.: Philadelphia, 1995; pp 206-232. (2) Folkman, J. Clinical application of research on angiogenesis. N. Engl. J. Med. 1995, 333, 1757-1763. (3) Bouck, N.; Stellmach, V.; Hsu, S. How tumors become angiogenic. Adv. Cancer Res. 1996, 56, 135-174. (4) Zetter, B. Angiogenesis and tumor metastasis. Annu. Rev. Med. 1998, 4, 209-217. (5) Holmgren L.; O’Reilly M. S.; Folkman, J. Dormancy of micrometastases: balanced proliferation and apoptosis in the presence of angiogenic suppression. Nature Med. 1995, 1, 149-153. (6) DiPietro, L. A. Thrombospondin as a regulator of angiogenesis. In Regulation of angiogenesis; Goldberg, T. D., Rosen, E. M., Eds.; Birkhauser Verlag Co.: Basel, 1997; pp 295-313. (7) Tolsma, S. S.; Volpert, O. V.; Good, D. J.; Frazier, W. A.; Polverini. P. J.; Bouck, N. Peptides derived from two separate domains of the matrix protein thrombospondin-1 have antiantiangiogenic activity. J. Cell Biol. 1993, 122, 497-511. (8) Campbell, S. C.; Volpert, O. V.; Ivanovich, M.; Bouck, N. P. Molecular mediators of angiogenesis in bladder cancer. Cancer Res. 1998, 58, 1298-1304. (9) Dawson, D. W.; Volpert, O. V.; Pearce. F. S.; Schneider, A. J.; Silverstein, R., L.; Henkin, J.; Bouck, N. P. Three distinct D-amino acid substitutions confer potent antiangiogenic activity on an inactive peptide derived from Thrombospondin-1 repeat. Mol. Pharmacol. 1999, 55, 332-338. (10) Majest, S.; Bell, R. Development of an endothelial cell chemotaxis assay. Proc. Am. Assoc. Cancer Res. 92rd Annual Meeting, New Orleans, LA, 2001, 42, 484. (11) Frevert, C. W.; Wong. V. A.; Goodman, R. B.; Goodwin, R.; Martin, T. R. Rapid fluorescence-based measurement of neutrophil migration in vitro. J. Immunol. Methodol. 1998, 213, 4152. (12) Nehls, V.; Drenckhahn, D. A microcarrier-based cocultivation system for the investigation of factors and cells involved in angiogenesis in three-dimensional fibrin matrices in vitro. Histochem Cell Biol. 1995, 104, 459-466. (13) Bonfoco, E.; Krainc, D.; Ankarcrona, M.; Nicotera, P.; Lipton, S. Apoptosis and necrosis: two distinct events induced, respectively, by mild and intense insults with N-methyl-D-aspartate or nitric oxide/superoxide in cortical cell culture. Proc. Natl. Acad. Sci. 1995, 92, 7162-7166. (14) Reiher, F. K.; Volpert, O. V.; Jimenez, B.; Crawford, S. E.; Dinney, C. P.; Henkin, J.; Haviv, F.; Bouck, N. P.; Campbell, S. C. Inhibition of tumor growth by systemic treatment with Thrombospondin-1 peptide mimetics. Int. J. Cancer. 2002, 98, 682-689. (15) Fournier, G. A.; Lutty, G. A.; Fenselau, A.; Patz, A. A corneal micropocket assay for angiogenesis in the rat eye. Invest. Ophthalmol. Vis. Sci. 1981, 21, 351-354. (16) Passaniti, A.; Taylor, R. M.; Pili, R.; Guo, Y.; Long, P. V.; Haney, J. A.; Pauly, R. R.; Grant, D. S.; Martin, G. R. A simple, quantitative method for assessing angiogenesis and antiangiogenic agents using reconstituted basement membrane, heparin, and fibroblast growth factor. Lab. Invest. 1992, 67, 519-528. (17) McDonnell, C. O.; Holden, G.; Sheridan, M. E.; Foley, D.; Moriarty, M.; Walsh, T. N.; Bouchier-Hayes, D. J. Improvement in efficacy of chemotherapy by addition of an antiangiogenic agent in a murine tumor model. J. Surg. Res. 2003, 116, 1923. (18) Karten, M. J.; Hook, W. A.; Siraganian, R. P.; Coy, D. H.; Folkers, K.; Rivier, J. E.; Roeske, R. W. In vitro histamine release with LHH analogs. In LHH and its Analogs: Contraception and Therapeutic Applications; Vickery, B. H., Nestor, J. J., Jr., Eds.; MTP Press Ltd.: Lancaster, UK, 1987; pp 179-190.

2846

Journal of Medicinal Chemistry, 2005, Vol. 48, No. 8

(19) Haviv, F.; Bush, E. N.; Knittle, J.; Greer, J. LHH Antagonists. In Integration of Pharmaceutical Discovery and Development: Case Histories; Borchardt, R. T., Freidinger, R. M., Sawyer, T. K., Eds.; Plenum Publishing Co., New York, 1998, pp 131-149. (20) Several biological activities of compounds 5 (ABT-526) and 6 (ABT-510) were previously reported by F. K. Reheir et al. (see ref 14). In that publication, ABT-526 was denoted as DI-TSP and ABT-510 as DI-TSPa. The structure of DI-TSP or ABT-526, described in the Reiher’s publication on page 683 is missing Sar at the N-terminus. (21) Mayr, M.; Li, C.; Zou, Y.; Huemer, U.; Hu, Y.; Xu, Q. Biochemical stress-induced apoptosis in vein grafts involves p38 mitogenactivated protein kinases. FASEB J. 2000, 15, 261-270. (22) Dudani, A. K.; Ganz, P. R. Endothelial cell surface actin serves as a binding site for plasminogen, tissue plasminogen activator and lipoprotein(a). Br. J. Haematol. 1996, 95, 168-178. (23) Viloria-Petit, A.; Miquerol, L.; Yu, J. L.; Gertsenstein, M.; Sheehan, C.; May, L.; Henkin, J.; Lobe, C.; Nagy, A.; Kerbel, R. S.; Rak, J. Contrasting effects of VEGF gene disruption in embryonic stem cell-derived versus oncogene-induced tumors. EMBO J. 2003, 22, 4091-4102. (24) Khanna, C.; Rusk, T.; Haviv, F.; Henkin, J. Antiangiogenic Thrombospondin-1 peptides result in regression of naturally occurring cancers in pet dogs. Presented at the 38th Annual Meeting of the ASCO, Chicago, IL, 2002; Abstr. No. 85.

Haviv et al. (25) Khanna, C.; Cozzi, E.; Sharpee, R.; Vail, D.; Graham, J.; Kitchell, B.; Rusk, T. A randomized placebo-controlled pre-clinical trial of the anti-angiogenic thrombospondin-mimetic peptide ABT526 plus Lomustine chemotherapy versus Lomustine chemotherapy alone in pet dogs with relapsed non-Hodgkin’s lymphoma. Presented at the 40th Annual Meeting of the ASCO, New Orleans, LA, 2004; Abstr. No. 3088. (26) Karyekar, C. S.; Carr, R. A.; Andre, A.; Wang, Q.; Facey, I.; Marsh, K.; Daszkowski, D.; Humerickhouse, R. Pharmacokinetics (PK) of the angiogenesis inhibitor ABT-510 in healthy subjects. Presented at the 40th Annual Meeting of the ASCO, New Orleans, LA, 2004; Abstr. No. 3080. (27) De Vos, F. Y.; Hoekstra, R.; Eskens, F. A. L. M.; De Vries, E. G.; Van der Gaast, A.; Groen, H. J. M.; Knight, R.; Hummerickhouse, R. A.; Gietema, J. A.; Verweij, J. Dose-finding and pharmacokinetic study of ABT-510 with gemcitabine and cisplatin in patients with advanced cancer. Presented at the 40th Annual Meeting of the ASCO, New Orleans, LA, 2004; Abstr. No. 3077. (28) Meienhofer, J.; Waki, M.; Heimer, E. P.; Lambros, T. J.; Makofske, R. C.; Chang, C. D. Solid-phase synthesis without repetitive acidolysis. Int. J. Pept. Protein Res. 1979, 13, 35-42.

JM0401560