Antiangiogenic agents.

A greater understanding of the complex process of tumor-induced angiogenesis, coupled with the notion that tumors require a blood supply to both grow and metastasize, has fueled the search for agents that block or disrupt the angiogenic process. Because normal vascular endothelial cells (ECs) turn over so slowly, conventional wisdom suggests that an antiangiogenic approach to cancer therapy should offer improved efficacy and reduced toxicity, without the potential for drug resistance. Numerous reports have identified small molecules or antibodies that can interfere with one or more key steps in EC signaling, migration or differentiation. Three new compounds, ZD4190, SU6668 and PD 0173073, have been reported during the past year to have significant and selective antiangiogenic activity, as well as antitumor activity.

Synthesis and tyrosine kinase inhibitory activity of a series of 2-amino-8H-pyrido[2,3-d]pyrimidines: identification of potent, selective platelet-derived growth factor receptor tyrosine kinase inhibitors.

Screening of a compound library led to the identification of 2-amino-6-(2,6-dichlorophenyl)-8-methylpyrido[2,3-d]pyrimidine (1) as a inhibitor of the platelet-derived growth factor receptor (PDGFr), fibroblast growth factor receptor (FGFr), and c-src tyrosine kinases (TKs). Replacement of the primary amino group at C-2 of 1 with a 4-(N,N-diethylaminoethoxy)phenylamino group yielded 2a, which had greatly increased activity against all three TKs. In the present work, variation of the aromatic group at C-6 and of the alkyl group at N-8 of the pyrido[2,3-d]pyrimidine core provided several analogues that retained potency, including derivatives that were biased toward inhibition of the TK activity of PDGFr. Analogues of 2a with a 3-thiophene or an unsubstituted phenyl group at C-6 were the most potent inhibitors. Compound 54, which had IC50 values of 31, 88, and 31 nM against PDGFr, FGFr, and c-src TK activity, respectively, was active in a variety of PDGF-dependent cellular assays and blocked the in vivo growth of three PDGF-dependent tumor lines.

2-Substituted aminopyrido[2,3-d]pyrimidin-7(8H)-ones. structure-activity relationships against selected tyrosine kinases and in vitro and in vivo anticancer activity.

While engaged in therapeutic intervention against a number of proliferative diseases, we have discovered the 2-aminopyrido[2, 3-d]pyrimidin-7(8H)-ones as a novel class of potent, broadly active tyrosine kinase (TK) inhibitors. An efficient route was developed that enabled the synthesis of a wide variety of analogues with substitution on several positions of the template. From the lead structure 2, a series of analogues bearing variable substituents at the C-2 position and methyl or ethyl at N-8 was made. Compounds of this series were competitive with ATP and displayed submicromolar to low nanomolar potency against a panel of TKs, including receptor (platelet-derived growth factor, PDGFr; fibroblast growth factor, FGFr; epidermal growth factor, EGFr) and nonreceptor (c-Src) classes. One of the more thoroughly evaluated members was 63 with IC50 values of 0.079 microM (PDGFr), 0.043 microM (bFGFr), 0.044 microM (EGFr), and 0.009 microM (c-Src). In cellular studies, 63 inhibited PDGF-mediated receptor autophosphorylation in a number of cell lines at IC50 values of 0.026-0.002 microM and proliferation of two PDGF-dependent lines at 0.3 microM. It also caused inhibition of soft agar colony formation in three cell lines that overexpress the c-Src TK, with IC50 values of 0.33-1.8 microM. In in vivo studies against a panel of seven xenograft tumor models with known and/or inferred dependence on the EGFr, PDGFr, and c-Src TKs, compound 63 produced a tumor growth delay of 10.6 days against the relatively refractory SK-OV-3 ovarian xenograft and also displayed activity against the HT-29 tumor. In rat oral bioavailability studies, compound 63 plasma concentrations declined in a biexponential manner, and systemic plasma clearance was high relative to liver blood flow. Finally, in rat metabolism studies, HPLC chromatography identified two metabolites of 63, which were proved by mass spectrometry and synthesis to be the primary amine (58) and N-oxide (66). Because of the excellent potency of 63 against selected TKs, in vitro and in vivo studies are underway for this compound in additional tumor models dependent upon PDGFr, FGFr, and c-Src to assess its potential for advancement to clinical trials.

Inhibitors of tyrosine kinase.

This review covers the literature on significant studies of small molecule inhibitors of the epidermal growth factor receptor (EGFR), platelet-derived growth factor receptor (PDGFR), fibroblast growth factor receptor (FGFR), Flk-1, and src family tyrosine kinases from 1996 through mid-1997. During this period, there has been substantial progress in the discovery of new and highly specific tyrosine kinase inhibitors (TKIs), particularly for the EGFR family. The last 18 months saw a focused effort to discover tyrosine kinase inhibitors with increased potency, increased selectivity, better animal pharmacokinetics, and decreased toxicity. Indeed, some EGFR TKIs are now in clinical trials or are about to enter clinical trials as potential anticancer agents. Potent and selective kinase inhibitors have also been described for PDGFR, but none of these compounds have appeared to advance in the developmental process as far as kinase inhibitors for the EGFR family. Surprisingly, potent and selective inhibitors of receptors involved in neovascularization such as FGFR, Flk-1, or Flt-1 are less prevalent in the literature, and the discovery of TKIs that can inhibit angiogenesis remains a fertile area for drug discovery. Tyrosine kinases continue to remain an extremely attractive target for the design of potent and selective inhibitors that will represent an important new class of therapeutic agents for the treatment of a variety of diseases where current therapy is still insufficient.

Lack of effect of corticosteroids and tamoxifen on suramin protein binding and in vitro activity.

Stout and colleagues [Proc Am Assoc Cancer Res 1993, 34, p. 298] previously reported that both hydrocortisone and tamoxifen increased the free fraction of suramin in human plasma. We examined several corticosteroids as well as tamoxifen for their effects on suramin protein binding and also evaluated hydrocortisone for its ability to modulate suramin activity in PC-3 and MCF-7 cells. Greater than 99% of the suramin was protein bound in undiluted human plasma. However, the free fraction of suramin was increased with the reduced plasma protein levels and increased suramin concentrations. At concentrations ranging from 1 to 30 microM, neither tamoxifen, hydrocortisone, prednisone nor dexamethasone had any effect on the binding of suramin to human plasma, regardless of protein concentrations. Similar results were observed with fetal calf serum. Hydrocortisone also had no effect on suramin activity against PC-3 and MCF-7 cell in vitro. We conclude from these studies that neither corticosteroids nor tamoxifen affect suramin protein binding or its cytotoxic activity.

Role of a small molecular weight phosphoprotein in the mechanism of action of CI-994 (N-acetyldinaline).

The mechanism of action of the novel anti-cancer compound CI-994 was studied in C26 murine colon tumor and HCT-8 human colon adenocarcinoma cells. Treatment of either cell line resulted in the specific loss of a 16-kDa phosphoprotein in a time- and concentration-dependent manner. Treatment with salicylanilide, CI-940, mimosine, aphidicolin, quercetin or ciclopirolxalamine, which, like CI-994, block cells in the G1-S phase of the cell cycle, did not affect the production of this protein. Loss of the 16-kDa protein preceded the block in cell proliferation induced by CI-994 treatment, and recovery of this protein was evident prior to the resumption of cell growth. Cellular fractionation studies demonstrated that the 16 kDa phosphoprotein is confined to the nuclear compartment. Our data indicate that loss the 16-kDa nuclear phosphoprotein appears to be a direct effect of CI-994 treatment and that the inhibition of this phosphoprotein may play a critical role in the mechanism of action of CI-994.

The effect of lysosomotropic agents and secretory inhibitors on anthracycline retention and activity in multiple drug-resistant cells.

The effect of lysosomotropic agents and secretory inhibitors were compared with verapamil for their effect on the activity of doxorubicin (DOX) in multiple drug-resistant (MDR) P388 leukemia cells (P388R) and in blocking anthracycline efflux from these cells. Agents known to interact with the plasma membrane did not potentiate DOX activity in P388R cells unless these same agents were also capable of interacting with acidic compartments within the cell. The lysosomotropic detergent Triton WR-1339, for example, potentiated DOX activity in P388R cells and stimulated the net accumulation of daunorubicin (DAU) in P388R cells by inhibiting drug exodus. However, another detergent, deoxycholate, and two membrane active antibiotics, amphotericin B and filipin, had no effect on DOX activity and/or DAU efflux in P388R cells. Lysosomotropic agents such as chloroquine and secretory inhibitors such as monensin, cytochalasin B, and vinblastine all inhibited DAU efflux from P388R cells. In a MDR B16 melanoma cell line, the activity of DOX was potentiated by both verapamil and reserpine. These same two agents also inhibited melanin secretion from this same cell line. Based on these observations, we propose that secretory vesicles derived from the Golgi apparatus might be involved in the MDR phenomenon. We further suggest that drugs such as DOX might be concentrated in these acidic vesicles, where they would be released to the outside of the cell by exocytosis.

Possible link between the intrinsic drug resistance of colon tumors and a detoxification mechanism of intestinal cells.

The insensitivity of colon tumors to various anticancer agents was studied in vitro. The activity of Adriamycin (ADR) in several colon tumor cell lines was potentiated by the calcium channel blocker verapamil (VER). In the HCT-8 human colon adenocarcinoma cell line, VER potentiation of the activities of ADR and the anthrapyrazole CI-937 appeared to be related to its ability to enhance the net accumulation of both drugs and inhibit their efflux. VER, which potentiated ADR activity in HCT-8 cells by 4-fold, caused a 3.5-fold stimulation of ADR accumulation and 3.5-fold inhibition of ADR efflux, when compared to non-VER-treated cells. The low level of VER potentiation of CI-937 activity in HCT-8 cells (1.4-fold) was also reflected in CI-937 transport studies which demonstrated a 1.5-fold enhancement of CI-937 accumulation and a 1.4-fold inhibition of its efflux. VER was also found to stimulate ADR activity and accumulation in a normal small intestinal crypt cell line (IEC-6). The mechanism of drug efflux was examined in HCT-8 cells. Agents known to increase the permeability of the plasma membrane did not alter ADR accumulation or its efflux in HCT-8 cells unless these same agents were also capable of interacting with the lysosome. Tween 80 and the lysosomotropic detergent Triton WR-1339 as well as proton ionophores and lysosomotropic amines all stimulated ADR uptake and/or inhibited its efflux from HCT-8 cells. ADR efflux was also partially blocked by cytochalasin B. Based on these observations, we suggest that at least part of the inherent drug resistance of colon tumor cells results from the retention of an enhanced drug efflux mechanism which is found in normal intestinal epithelium where this property may provide protection from plant alkaloids and other xenobiotic agents ingested in the diet. The mechanism of this drug efflux from HCT-8 cells may involve drug partitioning into acidic vesicles within the cell and their subsequent release from these cells by exocytosis.

Resistance to anthrapyrazoles and anthracyclines in multidrug-resistant P388 murine leukemia cells: reversal by calcium blockers and calmodulin antagonists.

A series of anthrapyrazoles was examined for their cytotoxic effect on P388 cells resistant (P388R) to anthracyclines, N-[4-(9-acridinylamino)-3-methoxyphenyl] methanesulfonamide, trimetrexate, and vinblastine. The degree of resistance of P388R cells to Adriamycin (ADR) and daunomycin was 50-fold and 38-fold, respectively, when compared to the parent cell line (P388S). The Adriamycin-resistant cells were highly cross-resistant to some anthrapyrazoles, but the degree of cross-resistance was not uniform and was less than 3-fold for one member of the series. The lipophilicity of these compounds appeared to correlate to some extent with the level of resistance. The calcium channel blockers verapamil (VER) and diltiazem and the calmodulin antagonist trifluoperazine potentiated the cytotoxicity of the anthrapyrazoles and ADR in P388R. This potentiating effect was concentration dependent with VER being the most efficacious. VER increased ADR cytotoxicity by greater than 10-fold and CI-937 by almost 40-fold. However, VER, diltiazem, and trifluoperazine had no effect on ADR or anthrapyrazole activity in P388S cells. The antiarrhythmic drug, quinidine, and the detergent, Tween 80, also potentiated ADR activity in P388R cells to the same extent as VER. Both the net accumulation and efflux of [3H]daunomycin were altered in P388R cells by nontoxic concentrations of Tween 80 in a fashion virtually identical to that demonstrated for VER. These data suggest that agents which potentiate drug cytotoxicity in P388R cells may do so by their interaction with the lipid domain of the plasma membrane. In addition, these results demonstrate that some members of the new series of DNA binding drugs, the anthrapyrazoles, may be active against anthracycline-resistant tumors and that, where cross-resistance to them occurs, it can be partially reversed by agents such as VER.

Cross resistance of pleiotropically drug resistant P338 leukemia cells to the lipophilic antifolates trimetrexate and BW 301U.

Several antifolate compounds were examined for their cytotoxic activity in a pleiotropically resistant P388 cell line (P388R). The sensitivity of P388R cells to methotrexate (MTX) and the lipophilic antifols, metoprine and methotrexate gamma-mono t-butyl ester (MTX-gamma-t-butyl ester) were comparable with that activity observed in the parental cell line (P388S). P388R cells were, however, resistant to 2 other lipophilic antifols, trimetrexate (TMQ) and BW 301U. The degree of resistance to TMQ and BW 301U was 22-fold and 15-fold, respectively and could be partially overcome by the calcium channel blocker, verapamil (VER) or the detergent Tween 80. Transport studies showed that net accumulation of trimetrexate was markedly reduced in P388R cells resulting in a steady-state level which was 25% of the sensitive line. This impaired uptake was reversed by 5 micrograms/ml VER which increased the steady-state to a level comparable to P388S. P388R also exhibited a 50% reduction in the unindirectional influx rate, however, this defect could not be reversed by VER. The resistance of P388R cells to TMQ and BW 301U and their potentiation by VER extends pleiotropic resistance to yet another class of drugs which have important clinical implications.

Evaluation of galactosyltransferase isoenzyme II in a human colon carcinoma-derived cell line, HCT-8.

Polyacrylamide gel electrophoresis of galactosyltransferase (GT) extracted from a human colon adenocarcinoma cell line, HCT-8, demonstrated the presence of two peaks of activity: a slow-moving peak, referred to as GT-II, and a more anodally migrating peak, designated as GT-I, which was also found for normal human serum. However, if GT solubilized from HCT-8 cells was separated by isoelectric focusing, no unique isoenzymes could be detected. Total GT activity from HCT-8 cells was purified by alpha-lactalbumin-Sepharose affinity chromatography followed by ion exchange chromatography on either DEAE-cellulose or FPLC using a Mono Q anion exchange resin. Three major peaks of activity were resolved from anion exchange chromatography. Electrophoresis of each peak revealed a GT pattern identical with that originally observed for the crude (detergent) solubilized homogenate. No enrichment of either GT-I or GT-II was observed in the three enzyme fractions. The data suggest that GT-II may be an artifactual activity of cancer cells composed of GT-I associated with some contaminating protein.

Collagen-production inhibitors evaluated as antitumor agents.

Proline analogues such as cis-4-hydroxy-L-proline (CHP) and L-azetidine-2-carboxylic acid (A2C) were tested for their antitumor activity in tissue culture and in vivo. In culture, CHP specifically inhibited those tumor cells that synthesized basement-membrane collagen. CHP appeared to selectively inhibit collagen biosynthesis with only a slight effect on protein synthesis. Culturing cells on type IV collagen matrix did not alter the antiproliferative effect of CHP. The inhibition of 450.1 mouse mammary tumor cells was fully reversible when cultures were incubated for 6 or 12 hours with 25 micrograms CHP/ml but was irreversible after 24 hours of exposure. Of the proline analogues tested against 450.1 tumor cells, A2C and CHP were the most potent inhibitors of cell growth. These two compounds were therefore tested in vivo using 3 transplantable tumors, all of which synthesized basement-membrane collagen. CHP and A2C were given twice daily to mice for 7 to 10 days at doses ranging from 50 mg/kg (body wt) to 600 mg/kg (body wt) per injection. Both CHP and A2C were completely inactive against the 450.1 mammary tumor and the EHS sarcoma. Both compounds also caused considerable liver toxicity. Against CD8F1 mammary tumors, treatment with maximum tolerated doses of CHP and A2C resulted in a slight but insignificant inhibition of tumor growth. While our studies confirmed previous findings that CHP specifically inhibited those tumor cells that synthesized basement-membrane collagen, CHP and A2C did not appear to be efficacious antitumor agents.

Decrease by cycloheximide of calcium binding and nonesterified fatty acids in rat-intestinal Golgi-enriched membrane fractions.

Rat intestinal Golgi-enriched membrane fractions bind more Ca2+ than do basolateral and microvillus-enriched membrane fractions, and this uptake is reduced by vitamin D-deficiency. The effect of the protein synthesis inhibitor, cycloheximide, on this Ca2+ binding was determined in rat fed a normal, vitamin D-sufficient diet. Cycloheximide, 1.5 mg/kg, rapidly reduced protein synthesis (measured by [3H]leucine incorporation) to 12% of control values within 15 min, but Ca2+ binding diminished gradually to 50% of control values by 60 min. Ca2+ transport across gut sacs was also decreased. The reduction in Ca2+ binding was not due to an alteration in vesicle morphology or to a direct effect of cycloheximide. Nonesterified (free) fatty acids, the probable binding sites for Ca2+ in these membrane fractions, were reduced by cycloheximide to 48% of control values by 60 min. There was no significant change in total lipid phosphate. Cycloheximide may affect the synthesis of proteins necessary for the presence of nonesterified fatty acids in these Golgi membranes.

Galactosyltransferase activity and cell growth: uridine diphosphate (UDP)galactose inhibition of murine leukemic L1210 cells.

UDPgalactose inhibits the growth of mouse leukemic L1210 cells. In calf serum supplemented Dulbecco's medium (CS-DMEM), 1.2 mM UDPgalactose (UDPgal) inhibited cell growth by 50% (IC50), and 5 mM UDPgalactose inhibited cell growth by 92%. Other nucleotide sugars as well as galactose, glucose, and galactose-1-phosphate had little or no effect on cell growth. Uridine nucleotides, which inhibit galactosyltransferase activity, protected L1210 cells from the growth inhibitory effect of UDPgalactose when both were added simultaneously to culture media. Unlike mouse 3T12 cells, in which no inhibition of cell growth was observed with heat-inactivated calf serum (HICS)-DMEM, 5 mM UDPgalactose inhibited L1210 cell growth in HICS-DMEM to the same degree as that observed in CS-DMEM. In contrast to 3T12 cells, L1210 cells secrete significant galactosyltransferase activity into the media. Complete inhibition of 3T12 cell growth by UDPgal was observed if HICS-DMEM medium was first conditioned by L1210 cells for 48 hours. No difference in cell growth or [3H]thymidine uptake was detected after 6 hours of exposure to UDPgalactose, but both were significantly decreased at 24 and 48 hours. Flow cytometric analysis of UDPgalactose effects on L1210 cells revealed no differences in the distribution of cells in G1, S, or G2-M of the cell cycle after 6 hours of incubation, but after 16 hours of UDPgalactose treatment, L1210 cells were arrested in early S phase. These cells were completely viable and morphologically similar to control L1210 cells. Normal growth was resumed when UDPgal was removed. The data suggest that UDPgalactose inhibition of cell growth requires extracellular galactosyltransferase activity and that the effect is mediated via the cell membrane.

Kinetics of subcellular distribution in rat intestine of 1,25-dihydroxycholecalciferol administered in vivo. Evidence for concentration within 5 min into purified nuclei.

To better understand the initial steps in the induction of intestinal Ca2+ transport by 1,25-dihydroxycholecalciferol [1,25(OH)2D3], we studied the early subcellular localization of 1,25(OH)2D3 in rat intestine. Vitamin D-deficient rats received 300 pmol of 1,25(OH)2[3H]D3 intravenously at 5 min to 4h before being killed. Cells homogenized in buffer of I = 90 mmol/litre were fractionated by centrifugation into a crude nuclear pellet, purified nuclei, Golgi and basal-lateral membranes, cytosol and a post-nuclear pellet. Nuclear purification was established by biochemical and morphological criteria and gave a yield of 32 +/- 2% (mean +/- S.E.M.; n = 21). Although re-establishment of Ca2+ uptake by Golgi is one of the earliest reported intestinal responses to 1,25(OH)2D3, no direct localization of 1,25(OH)2D3 to Golgi was detected. Purified nuclei had the highest specific radioactivity at all times studied, with nuclear localization detectable at 5 min and peak nuclear uptake at 1 h. Relative specific radioactivity of nuclei to cytosol increased from 5 min to 30 min, at which time equilibrium between cytosol and nucleus appeared to be attained. Nuclear uptake occurred in all cells from villus to crypt. Of total nuclear binding 10% was resistant to high ionic strength buffer (I = 365 mmol/litre); peak nuclear uptake was observed at 30 min in this buffer. This tight binding may represent the active fraction of 1,25(OH)2D3. These results indicate that localization of 1,25(OH)2D3 to rat intestinal nuclei precedes the observed Golgi-membrane effects and suggest the existence of high-affinity nuclear 1,25(OH)2D3-binding sites.

Release of glycosyltransferase and glycosidase activities from normal and transformed cell lines.

The release of galactosyltransferase, sialyltransferase, and several glycosidase activities into the growth media from several normal and transformed cell lines was examined. Six of the seven cell lines released galactosyltransferase into their culture media. Only the human leukemia CCRF-CEM cells failed to release demonstrable galactosyltransferase activity. Release of galactosyltransferase activity into the media closely paralleled the growth curves for all but the BHKpy cells. These cells continued to release peak levels of galactosyltransferase activity into the culture media after their growth had plateaued. Media galactosyltransferase activity was unaffected by Triton X-100 treatment had remained in the supernatant fraction of a 100,000 X g, 12-hr centrifugation, suggesting that the cells release galactosyltransferase in a soluble form. In contrast to galactosyltransferase activity, only one of the cell lines (L1210) released sialyltransferase activity in appreciable amounts. Even this level of activity was 20-fold less than that observed for galactosyltransferase in the media from L1210 cells. Of the nine glycosidase activities assayed, only N-acetylglucosaminidase was observed in significant amounts in the media from all but the CCRF-CEM cells. However, N-acetylglucosaminidase release did not correlate closely with cell growth. These findings suggest a relatively specific release of galactosyltransferase and N-acetylglucosaminidase activities by cells in tissue culture. Moreover, the release of galactosyltransferase closely parallels cell growth. The significance of these released enzymes, especially to cell growth, has yet to be determined.

Characterization of nucleoside diphosphatase in the onion root tip. III. Solubilization and activation of membrane-bound enzyme(s).

UNLABELLED: The latency of nucleoside diphosphatase (NDPase) in onions root homogenates has been examined by comparing the activation of NDPase activity resulting from detergent treatment with that due to storage of homogenates for several days in the cold. Both detergent treatment and cold storage activated NDPase approximately two-fold. In both cases this activation was paralleled by the loss of enzyme activity from the membrane fractions and its appearance in the supernatants. Electrophoresis of these supernatants revealed an identifical isoenzyme pattern of 5 NDPase bands for both preparations. Enzyme kinetic studies demonstrated that NDPase from the detergent-treated homogenate and the homogenate stored in the cold as well as NDPase from the membrane and supernatant fractions from each of the homogenates all had the same Km value. These data suggest that latency of NDPase is the result of a breakdown of cellular membranes and subsequent release of NDPase. ABBREVIATIONS: DOC, deoxycholate; NDPase, nucleoside diphosphatase; IDP, inosine 5'-diphosphate; UDP, uridine 5'-diphosphate; GDP, guanosine 5'-diphosphate.