Farnesyltransferase inhibitors: a new class of cancer chemotherapeutics. K. S. Koblan*§, N. E. Kohl*, C. A. Omer*, N. J. Anthonyt, M. W. Conned, S. J. deSolmst, ...
Biochemical Society Transactions
Farnesyltransferase inhibitors: a new class of cancer chemotherapeutics K. S. Koblan*§, N. E. Kohl*, C. A. Omer*, N. J. Anthonyt, M. W. Conned, S. J. deSolmst, T. M. Williamst, S. L. Grahamt, G. D. Hartmant, A. Oliff* and J. B. Gibbs* Departments of *Cancer Research, tMedicinal Chemistry and $Safety Assessment, Merck Research Laboratories, West Point, PA 19486, U.S.A.
688
T h e family of Ras proteins plays a central role in the regulation of cell growth and integration of regulatory signals which can govern the cell cycle and proliferation. Mutant m s genes (derived from murine sarcoma viruses) were among the first oncogenes described for their ability to transform cells to a cancerous phenotype [l]. Mutations in one of the three genes (H-, N- and K-rus) encoding Ras proteins have been intimately associated with unregulated cell proliferation and are found in an estimated 30% of all human cancers [2]. T h e frequency of i-us mutations appears to depend upon the specific tumour type analysed. For example, an estimated 90% of pancreatic carcinomas contain a mutated oncogenic Ras protein while i a s mutations are rarely found in mammary carcinomas. T h u s an understanding of the physiology of the Ras proteins and development of inhibitors of Ras function may be important in their applications as chemotherapeutic agents in clinical cancer therapy. T h e function of normal and oncogenic Ras proteins is absolutely dependent on a series of post-translational modifications which promote membrane association. T h e physical association and localization to the plasma membrane is a prerequisite for Ras function and has propelled development of inhibitors to block the oncoprotein's membrane targeting. T h e first and obligatory modification in post-translational processing is S-farnesylation of a cysteine residue near the C-terminus of the protein. T h e precursor to the mature Ras protein contains a cysteine (C) as the fourth residue from the C-terminal amino acid of the protein. This cysteine residue is followed by two amino acids with aliphatic side chains (AA) and ultimately another residue, X. This tetrapeptide motif, referred to as a CAAX box, is the recognition sequence for protein farnesylation by the enzyme protein farnesyltransferase (PFTase). T h e amino acid in the ' X position is important ~
llbbreviations used: FPP, farnesyl diphosphate; PF'l'ase, protein farnesyltrarisferase; P(;( ;'rase-I, protein geranylgerariyltransferase I. §To whom correspondence should be addressed.
Volume 24
in determining which prenyltransferase will recognize the protein substrate. PF'l'ase modifies the CAAX peptide of Ras (in which X is Met, Ser, Ala or Gln) using farnesyl diphosphate (FPP) as co-substrate in the presence of bivalent cations. T h e subsequent processing event is removal of the C-terminal AAX tripeptide by an endoprotease that recognizes the farnesylatcd substrate. Finally, a methyltransferase esterifies the now free carboxylate of farnesylcysteine. Mutation of the critical cysteine in the CAAX box prevents farnesylation, and the resulting cytosolic form of the Ras protein is non-functional and unable to transform cells. While the exact role of the farnesyl group as a simple non-specific hydrophobic anchor or more provocatively in a spccific protein-protein interaction is unclear, it remains certain that farnesylation is necessary for the function of the Ras protein and inhibition of the enzyme PFTase is a potential drug target.
Synthesis of PFTase inhibitors T h e development of inhibitors o f PFTase has progressed through both screening of natural products and defined chemical compounds and more rational design based on the substrates of the enzyme reaction. Analogues based on either the CAAX peptide substrate or FPP could be developed, but, to be used effectively in humans, the compounds must be stable and permeant to the cell membrane to gain access to the cytosolic transferase. A number of farnesylphosptior~icacids have been reported which are purely competitive inhibitors of PFTase rather than substrates in the enzyme reaction [3-51. These compounds lack the pyrophosphate leaving group of FPP, are potent inhibitors of PFTase and are remarkably selective ( > 1000-fold) in their inhibition of PFTase over other isoprerioid-utilizing enzymes 7 \ztiu. T h e ( x -h ydroxyfarne syl) phos phon i c acid derivative partially suppresses H-Ras farnesylation in cell culture at micromolar concentrations and was one of the first compounds to display biological activity [6]. Compounds that contain a lipophilic diacid motif or polycarboxylic acids (e.g. chaetomellic acid and zaragozic acid) are a
Function and Pharmacology of Protein Prenylation
I
recurring theme in PF'Tase inhibitors that compete with FPP [7,8]. T h e natural product manurnycin inhibits PFTase competitively with respect to FPP, inhibits the function of farnesylated proteins in yeast, and is efficacious against both K- and N-ria-transformed fibrosarcomas in mouse models of tumorigenesis [9]. In general all these compounds display some selectivity for PFTase iri 7 i t t - o ; however, iri 73i730 the precise effect on other FPP-utilizing enzymes (e.g. squalene synthetase, FPP synthase) is unclear and therefore so are the cellular effects. Mechanistic and biochemical studies of PFTase have demonstrated that, although the enzyme utilizes polypeptides as substrates in the cell, tetrapeptides contain the essential determinants for enzyme recognition and thus can be farnesylated [lo]. Several laboratories have developed CAAX-based peptidomimetics which are potent PFTase inhibitors [ 1 1 -1.51. Extensive structure-activity studies show that a wide range of amino acids are tolerated within the central dipeptide, while the C-terminal residue, X, is a critical determinant of specificity of the prenylation reaction. Motifs where X is serine, methionine or glutamine become farnesylated, while those ending in leucine are geranylgeranylated by a closely related prenyltransferase, protein geranylgeranyltransferase I (PGGTase-I) [ 16,171. Substitution of aromatic amino acids (phenylalanine, tryptophan, tyrosine) as the penultimate residue in tetrapeptides or modification of the peptide backbone rendered them non-substrates and thus true inhibitors of PFTase [ 18,191. Modification of the backbone of the C M tetrapeptides to obtain metabolically stable compounds has yielded peptidomimetics which are less readily degraded and display antitumour activity against cells transformed by ras in tissue culture models. T h e incorporation of nonhydrolysable isosteric replacements for the amide bonds of CAAX (reduced amides, olefin or ether linkages) can lead to increases in PFTase inhibition (in vitro potency against the enzyme). However, there appears to be a delicate balance between stabilization of the peptide backbone and selective inhibition of PFTase over the related enzyme PGGTase-I. Recent work in our laboratory has focused on a peptidomimetic, 1,-739,750, based on the tetrapeptide Cys-IlePhe-Met (Figure 1). T h e backbone of L-739,750 contains an oxymethylene peptide bond replacement and substitutes a methioninesulphone in the X position. T h e compound is a potent
Figure I Structure of L-739,749 [2(S)-{2(S)-[2(R)-amino-3mercapto]propylamino-3(S)-rnethyl}pentyloxy- 3 - phenylpropionylmethioninesulfphone methyl ester], L-739,750 [2(S)-{ 2(S)-[2(R)-amino-3-mercapto] propylamino-3(S)methyl}pentyloxy-3-phenylpropionylmethioninesulphone] and L-744,832 [2(S)-{2(S)-[2(R)-amino-3me r c ap t 01 p r o py Iam in o - 3(S) - m e t hy I } penty Ioxy - 3 phenylpropionylmethioninsulphoneisopropyl ester] In L-739.749 R = CH,. in L-739,750 R = H and in L-744.832 R = CH(CH3);.
I
"i kO2CH3
PFTase inhibitor (ICs0= 1.8 nM), and a prodrug form (L-739,749) inhibits Ras farnesylation and membrane association in cells at low micromolar concentrations (half-maximal effect 0.1-1 pM). T h e terminal carboxylate is masked by esterification to enhance cell penetration; in the presence of intracellular esterases the labile ester is converted into the free acid form L-739,750. Importantly, the geranylgeranylation of proteins in these cells was unaffected by this compound and thus remained PFTase-specific. T h e ester L-739,749 also reduces the colony formation of rus-transformed cells in soft agar at concentrations in the range 1-2 pM and blocks the growth of Rasdependent tumour xenografts in nude mice. T h e tumour growth suppression by L-739,749 appears to result from inhibition of the Ras pathway and not from non-specific cytotoxicity, as the growth of cells transformed by either v-Mos or v-Raf in mice, proteins which function downstream of Ras, was unaffected. T h e CAAX tetrapeptide motif has served as a template for replacement of the central hydrophobic residue with aminobenzoic acid or aminobenzodiazepine scaffold with nanomolar IC,,, values and specificity for PFTase over PGGTaseI [ 14,201. Hamilton and co-workers (University of Pittsburgh) [14] have utilized the aminobenzoic acid scaffold to develop potent inhibitors which block H-Ras farnesylation in transformed cells at 1 pM and reduce the growth of Ras tumours in nude mouse xenografts. T h e benzodiazepine analogue BZA-2B (ICs0= 0.3 nM) is a
I996
689
Biochemical Society Transactions
Figure 2
Stereo view of the NMR-derived best-fit structure of the PFTase-bound peptide CVWM The peptide inhibitor most closely resembles a type-Ill /]-turn Adapted from
690
[2I ] with the permission of Cambridge University Press
Function and Pharmacology of Protein Prenylation
potcnt and selective PFTase inhibitor, and an ester prodrug slows the growth of H-Ras-transformed cells in inonolayer and increases the life span of nude mice after intraperitoneal implantation of HT1080 cells (a human fibrosarcoma line) [ 121. Development and refinement of existing non-peptide PFTase analogues may also benefit from the recent elucidation of the PF'l'ase-bound conformation of the tetrapeptide inhibitor, CVWM, and a novel pseudopeptide inhibitor by NMR spectroscopy [21,22]. Both inhibitors were shown to adopt non-ideal reverse-turn conformations most closely approximating a type-I11 p-turn (Figure 2). T h e detailed backbone arid side chain information described may aid in the development of new classes of potent non-peptide PFTase inhibitors. While the demonstration that inhibition of PFTase reduces the rate of growth of transformed cell lines in tissue culture experiments (cytostatic agents), the dramatic successes obtained in whole animal models has provided an incentive to pursue PFTase inhibitors vigorously. I I-Ras tumours implanted in immunocompromised mice are sensitive, while the PFTase inhibitors fail to affect cells transformed by Raf, demonstrating that the effects are due to inhibition of the Ras pathway. Similarly the inhibitors are successful in preventing tumour growth in syngeneic mice using human lung cells [23] or human pancreatic adenocarcinoma cells in culture which have mutations in the iityc or pS3 genes in addition to K-ins [24]. T h e response of the human tumour cell lines suggests that PF'l'ase inhibitors may be effective against tuniours with multiple genetic changes. To date, cell culture studies indicate that PFTase inhibitors would act primarily as cytostatic agents; however, recent studies conducted in transgenic mice that overexpress the v-H-ins oncogene demonstrate that tumour regression can be induced by inhibiting protein farnesylation [25]. Complete regression of 100% of small established tumours and partial regression (partial with respect to the number of respondents and degree of response) of large mammary and salivary tumours was achieved with the isopropyl ester of L-739,750. At a dose of 40 mg of L-744,832/day per kg of body weight, subcutaneously, tumours ranging in size from 50 to 2000 m m j were observed to undergo nearly complete regression within a few weeks of initiation of treatment. Remarkably, mice treated for up to 11 weeks with this compound showed no apparent
signs of toxicity to the gastrointestinal tract, bone marrow or retina, all tissues known to contain farnesylated proteins or be dependent upon rapid growth and differentiation of cells. Thus in rodent systems, as in the tissue culture models, the potential for a high therapeutic index is implied. In order to understand the success of PFTase inhibitors to date and their potential clinical utility, several questions remain to be answered. What mechanisms may normal or transformed cells utilize to compensate for the loss of farnesylation? 'The mechanism by which the related PGGTase-I can recognize the variety of substrates, including K-Ras, has been demonstrated in 7itro [26]. Cross-prenylation may have significant implications and may contribute to the apparent safety of the current inhibitors. T h e long-term consequence of inhibiting PFTase in humans is a prime concern and the rodent models to date are limited in duration and scope. Although the development of resistance is an important unresolved concern, it is promising to note that tumours that re-appear after treatment is stopped (two of three), or fail more traditional doxorubicin therapy (three of five) in the v-H-Ras transgenic mice remain sensitive to the PFTase inhibitor. Future experiments will reveal the answers to these questions and pose more perplexing inquiries while the utility of targeting PFTase for the treatment of cancer is tested.
Barbacid, M. (1987) Annu. Rev. Riochem. 56, 779-827 Rodenhuis, S. (1992) Semin. Cancer Riol. 3, 241247
Pompliano, D. Id., Rands, I. I,., Ikrgstrom, J. I)., Rew, D., Polishook, J. D., Meinz, M., IIuang, I,. Y . , Silverman, K. C. et al. (1993) J. Nat. Prod. 56, 1923-1929 9 IIara, M., Akasaka, K., Akinaga, S., Okabe, M., Nakano, I I . , (iomez, K., Wood, D., Uh, M. and ‘I’amanoi, F. (1993) Proc. Natl. ,\cad. Sci. U.S.A. 90, 2281 -2285 10 Keiss, Y., Goldstein, J. I,., Seabra, M. C., Casey, P. J. and Brown, M. S. (1990) Cell 62, 81-88 11 (iarcia, A. M., Rowell, C., Ackermann, K., Kowalczyk, J. J. and Lewis, M. D. (1993) J. Biol. Chem. 268, 18415-18418 12 James, G. L., Goldstein, J. I,., Brown, M. S., Kawson, ‘1’. E., Somers, ‘I‘. C., Medowell, R. S., Crowley, C. W., Lucas, B. K., I,evinson, A. D. and Marsters, J. C. (1993) Science 260, 1937-1942 13 Kohl, N. E., Mosser, S. D., deSolms, S. J., Giuliani, E. A., Pompliano, D. I,., Graham, S. I,., Smith, K. I,., Scolnick, E. M., Oliff, A. and Gibbs, J. B. (1993) Science 260, 1934-1937 14 Nigam, M., Seong, C. M., Qian, Y., Hamilton, A . I>. and Sebti, S. M. (1993) J. Biol. Chem. 268, 20695-20698 15 Graham, S. I,., deSolms, S. J., Giuliani, E. A., Kohl, N. E., Mosser, S. D., Oliff, A. I., Pompliano, I>. I,., Rands, E., Breslin, M. J., Ileana, A . A. et al. (1994) J. Med. Chem. 37, 725-732 16 Moores, S. I,., Schaber, M. D., Mosser, S. D., Rands, E., O’Hara, M. B., Garsky, V. M., Marshall, M. S., Pompliano, D. L. and Gibbs, J. €3. (1991) J. Biol. Chem. 266, 14603-14610
17 Yokoyama, K., Goodwin, (;. W.,(;hornashchi, I:., Glomset, J. :I and . (ielb, M. 11. (1991) Proc. Natl. Acad. Sci. U.S.A. 88, 5302-5300 18 Goldstein, J. I,., I3rown, M. S., Stradley, S. J., Keiss, Y . and Gierasch, I,. M. (1991) J. I3iol. Chem. 266, 15575-15578 19 Pompliano, I>. I,., Rands, IS., Schaber, M. I)., Mosser, S. I>., :Znthony, N. J. and (;ibhs, J. 13. (1992) Biochemistry 31, 3800-3807 20 Qian, Y . , Blaskovich, M. ’I., Saleern, M., Seong, C.-M., Wathcn, S. P., Iiamilton, 11. I>. and Sebti, S. M. (1994) J. Riol. Chem. 269, 12410-12413 21 Koblan, K. S., Culberson, J. C., deSolrns, S. J . , Giuliani, E. A, Mosser, S. I>., Omer, C. :I., Pitzenberger, S. M. and Bogusky, M. J. (1995) Protein Sci. 4, 681 -088 22 Stradley, S. J., Rizzo, J. and (iierasch, I,. M. (1993) Biochemistry 32, 12586-12590 23 Sun, J., Qian, Y., Hamilton, A. D. and Sebti, S. M. (1995) Cancer Res. 55, 4243-4247 24 Kohl, N. E., Wilson, 1:. K., Mosser, S. I>., (iiuliani, E. A., deSolms, S. J., Conncr, M. W., Anthony, N. J., IIoltz, W. J., Gomez, R. P., I,ec, ‘1’. J. et al. (1994) Proc. Natl. Acad. Sci. U.S.A. 91, 9141-9145 25 Kohl, N. E., Omer, C. A, Conner, M. W., Anthony, N. J., Davide, J. P., deSolms, S. J . , (;iuliani, I
