Introduction

Over the last decade, the posttranslational processing pathway ofthe Ras oncoprotein has been elucidated and the enzyme that catalyzes the initial and critical step in this pathway, protein farnesyltransferase (FTase), has been characterized (1-6). These advances led to a great deal ofeffort, both in the pharmaceutical industry and in academia, focused on the discovery of selective FTase inhibitors (FTIs). This chapter discusses two aspects of the FTI program at Schering-Plough Research Institute: 1) the discovery and development of the tricyclic series of inhibitors, and 2) studies on the substrate specificity of the prenyl transferases. The latter studies are relevant to our understanding of the in vivo effects of FTIs on prenylation of Ras and other proteins.

From: Farnesyltransferase Inhibitors in Cancer Therapy Edited by: S. M. Sebti and A. D. Hamilton © Humana Press Inc., Totowa, NJ

Fig. 1. Structures ofkey tricyclic FTIs. (Left) SCH 44342 is an early-generation tricyclic FTI that inhibits FTase with an IC50 value of250 nM. Furthermodification—in particular, halogenation of the tricyclic ring system—enhanced potency. (Middle) SCH 56582 inhibits FTase with an IC50 value of 60 nM. (Right) SCH 66336 inhibits FTase with an IC50 value of 1.9 nM. This compound is currently being evaluated in Phase I clinical trials.

Fig. 1. Structures ofkey tricyclic FTIs. (Left) SCH 44342 is an early-generation tricyclic FTI that inhibits FTase with an IC50 value of250 nM. Furthermodification—in particular, halogenation of the tricyclic ring system—enhanced potency. (Middle) SCH 56582 inhibits FTase with an IC50 value of 60 nM. (Right) SCH 66336 inhibits FTase with an IC50 value of 1.9 nM. This compound is currently being evaluated in Phase I clinical trials.

2. DISCOVERY AND INITIAL CHARACTERIZATION OF TRICYCLIC FTIs

A number of inhibitors of FTase have been reported (reviewed in refs. 7,8). Many of these are derived from substrate-mimetic approaches. The design of CAAX peptido-mimetics (9-14) has resulted in potent and selective FTIs capable of blocking H-Ras processing in cells, inhibiting cellular transformation induced by oncogenic Ras proteins, and slowing the growth of Ras-dependent tumors in nude mice (15-17).

SCH 44342 (Fig. 1) was one of the initial lead compounds in the tricyclic FTI series (18). This class of FTIs was discovered in 1991 by random screening of the Schering-Plough library ofcompounds. Unlike many other FTIs, the tricyclic class is entirely non-peptidic and lacks a sulfhydryl function. SCH 44342 inhibits recombinant human FTase with an IC50 value of approx 0.25 ^M. This inhibition is competitive with respect to the protein or peptide CAAX substrate. Tricyclic FTIs are structurally related to compounds possessing histamine H1 and platelet-activating factor antagonist activity (19), however, the FTase inhibitory activity is separable fromthese other activities. Most ofthe tricyclic FTIs are inactive or weakly active as inhibitors of the related protein prenyltransferase, protein geranylgeranyltransferase-1 (GGTase I). For example, SCH 44342 has no inhibitory activity against GGTase I at concentrations up to 100 ^M.

When the biological effects of SCH 44342 and related tricyclics were examined in cellular or animal systems driven by an activated H-Ras oncogene, the results observed were straightforward. H-Ras driven systems allow the use ofapowerful negative control (20,21). Although H-Ras with its native CAAX sequence ofCVLS is only a substrate for FTase, H-Ras in which the CAAX sequence is changed to CVLL is only a GGTase I substrate. Thus, selective effects on CVLS vs CVLL H-Ras transformed cells is a clear indication that compounds are exerting their biological effects by selective inhibition of FTase.

SCH 44342 inhibits the post-translational processing of H-Ras-CVLS in Cos-7 monkey kidney cells with an IC50 value of approx 3.0 ^M (18). In contrast, it does not inhibit H-Ras-CVLL processing. Additionally, SCH 44342 prevents the phenotypic change that occurs in Cos cells after transient expression of an activated [Val12] form of H-Ras-CVLS. SCH 44342 does not prevent this phenotypic change in cells overexpressing [Val12]-H-Ras-CVLL. It is important to note that SCH 44342 and its analogs have no apparent cytotoxic effects on Cos cells at the concentrations that exert this morphological effect.

Fig. 2. Effects ofSCH 44342 on morphology ofnormal and Ras-transformed Rat2 fibroblasts. Control Rat 2 fibroblasts and fibroblasts transformed with Val12-activated forms of H-Ras-CVLS or H-Ras-CVLL were plated at 250 cells/100 mm per dish in the presence or absence of20 p,M SCH 44342 in Dulbecco's modified Eagle's medium containing 10% fetal calf serum. Plates were re-fed after 5 d and photographed on d 14. Drug treatment had no effect on plating efficiency of any of the cell lines.

Fig. 2. Effects ofSCH 44342 on morphology ofnormal and Ras-transformed Rat2 fibroblasts. Control Rat 2 fibroblasts and fibroblasts transformed with Val12-activated forms of H-Ras-CVLS or H-Ras-CVLL were plated at 250 cells/100 mm per dish in the presence or absence of20 p,M SCH 44342 in Dulbecco's modified Eagle's medium containing 10% fetal calf serum. Plates were re-fed after 5 d and photographed on d 14. Drug treatment had no effect on plating efficiency of any of the cell lines.

The effect of SCH 44342 on the growth properties of Rat 2 fibroblasts stably transformed with activated H-Ras was also examined. SCH 44342 does not effect the plating efficiency of parental Rat 2 cells or Rat 2 cells transformed by H-Ras-CVLS or H-Ras-CVLL, confirming its lack of cytotoxicity at concentrations where H-ras processing is fully inhibited. Figure 2 shows the morphology ofclones grown in the presence ofeither DMSO or 20 ^M SCH 44342. The morphology of the normal Rat-2 cells is unaffected by the presence of drug, although the clones are somewhat smaller on average compared to those grown in its absence. Cells transformed by H-Ras-CVLS undergo a pronounced morphological change in response to SCH 44342, displaying a flatter, less refractile, more contact inhibited phenotype than vehicle control-treated cells. Morphological reversion is not observed in cells transformedwith geranylgeranylated H-Ras-CVLL. Consistent with these selective effects on the transformed phenotype, treatment ofH-Ras-CVLS (but not H-Ras-CVLL) transformed cells with SCH 44342 results in the reformation of stress fibers and a decrease in membrane extensions (Fig. 3). In addition, SCH 44342 dose-dependently reduces anchorage-independent soft agar growth of the H-Ras-CVLS transformants, but not of Rat-2 cells transformed with the geranylgeranylated protein.

During the characterization of early generation FTIs, we and others (see refs. 21,22) found that cellular functions that were predicted to be altered by blocking farnesylation were spared. These included signaling pathways thought to involve Ras activation and assembly of the nuclear envelope, despite the fact that both lamin B and prelamin A are FTase substrates (see below). These findings are consistent with the lack ofgeneral cyto-toxicity observed with FTIs. There are several possible explanations for these paradoxical observations. One possibility is that redundant signaling pathways exist that bypass Ras inactivation. A second possibility is that inhibition of Ras farnesylation is incomplete owing either to incomplete inhibition of FTase or to alternative prenylation of Ras by another cellular prenyl transferase. The discussion to follow focuses on the second possibility; however, the contribution ofpotential redundant pathways has not been fully explored and needs to be addressed further experimentally.

Was this article helpful?

0 0

Post a comment