Protein prenylation is a posttranslational modification of proteins involving the addition of isoprenyl lipids (Tamanoi and Sigman 2001). It is a post-translational modification that occurs in the cytosol and is essential for the proper localization and functions of many proteins. There are three kinds of prenyltransferases: protein farnesyltransferase (FTase), protein geranylgeranyltransferase-I (GGTase-I), and Rab geranylgeranyltransferase (RabGGTase). No additional prenyltransferases have been described to date. Farnesyltransferase (FTase) catalyzes the addition of a 15-carbon farnesyl group to proteins such as Ras proteins, Rheb proteins, nuclear lamins, and Hdj2 that end with the Cys-A1-A2-X (CA 1 A 2 X) motif, where A is an aliphatic amino acid and X is usually serine, methionine, glutamine or alanine. The lipid is covalently attached to the Cys via a thioether linkage between C1 of the farnesyl group and the sulfur of Cys. Some CA 1 A 2 X-ending proteins such as K-Ras and N-Ras undergo geranylgeranylation catalyzed by GGTase-I when farnesylation is inhibited, in a process known as alternative prenylation (Whyte et al. 1997). Interestingly, RhoB protein can be either farnesylated or geranylgeranylated (Lebowitz et al. 1997). See chapters on Ras and Rho for the function of these proteins. A recent study identified novel peptide substrates for FTase (Hougland et al. 2009). In their study, Hougland J.L. et al. selected and screened a library of small peptides representing the C-termini of more than 200 human proteins for reactivity with FTase. They identified two peptide substrate classes with distinct sequence selectivities – one class that exhibits multiple-turnover (MTO) reactivity, and another class that exhibits single-turnover (STO) reactivity, presumably due to slow dissociation of the prenylated peptide substrate under steady-state reaction conditions. In addition, they examined the amino acid composition of the two peptide substrate classes. While the STO peptides vary significantly at both the A2 and X residues, the MTO peptides are enriched in a non-polar amino acid at the A2 position and a Phe, Met, or Gln at the terminal X residues (Hougland et al. 2009). Some of the STO peptide substrates identified include CVLL (R-Ras), CAKS (Rab 38), CYSN (Ubiquitin-conjugating enzyme E2 variant 1), and CPLG (Ras association domain family 1) (Hougland et al. 2009).
FTase is a heterodimeric, zinc metalloenzyme that consists of the α-subunit (48 kDa) and the β-subunit (46 kDa). The α-subunit is composed of fewer amino acids than the β-subunit (377 vs 437 residues) (Tamanoi and Sigman 2001; Zhang and Casey 1996). FTase and GGTase-I share the same α-subunit; the cDNA cloning of FTase and GGTase-I confirmed that the α-subunits of both prenyltransferases are encoded by the same gene. The β-subunit of FTase is distinct from that of GGTase-I, but they share an overall amino acid similarity of about 30%. Crystal structures have revealed that both α and β subunits are composed primarily of α helices (Park et al. 1997). They are arranged in a crescent-shaped superhelix (α-subunit) that wraps around an α-α barrel (β-subunit) which has a deep central cleft that forms the active site of FTase (Park et al. 1997). Although the majority of the active site residues are derived from the β-subunit, both subunits are important for substrate binding and catalysis.
Crystal structures of FTase indicate the presence of a single Zn2+ ion bound to the β subunit near the α/β subunit interface (Taylor et al. 2003). The Zn2+ ion is required for catalytic activity and coordinates the Cys thiol of the CA1A2X substrate. Mg2+ ions have also been found to bind to the active site of FTase and can accelerate the protein farnesylation reaction by up to 700-fold (Pickett et al. 2003). Recently, Yang Y. et al. reported a computational study regarding the Mg2+ binding site in FTase. Their calculations support the idea that D352β plays a critical role in Mg2+ binding and Mg2+ is important for the conformational transition step of the reaction (Yang et al. 2010).
In an earlier study, crystallographic analysis of FTase and GGTase-I complexed with substrate peptides, including those that were derived from the C termini of K-Ras, H-Ras, and TC21, was performed to define rules of protein substrate selectivity for both prenyltransferases (Reid et al. 2004). They showed that residues Trp102β, Trp106β, and Tyr361β of FTase bind to the A2 residue of the CA 1 A 2 X. Interestingly, their findings suggested that the X residue of the CA 1 A 2 X of FTase substrates can bind in one of two different sites, one found by residues Tyr131α, Ala98β, Ser99β, Trp102β, His149β, Ala151β, and Pro152β of FTase, the other formed by residues Leu96β, Ser99β, Trp102β, Trp106β, and Ala151β of FTase (Reid et al. 2004).
Mutagenesis studies and crystal structures have provided an insight into the structure of FTase in complex with the isoprenoid of farnesyl diphosphate (FPP), the prenyl donor (Reid et al. 2004; Bowers and Fierke 2004). FPP reaches to the bottom of the hydrophobic, deep central cleft of the β-subunit. The diphosphate moiety of FPP interacts with residues Lys164α, Arg291β, His248β, Lys294β, and Tyr300β at the top of the α-α barrel. Residues Lys164α, Arg291β, His248β, Lys294β, and Tyr300β form the diphosphate binding pocket. Crystal structures of inactive FTase•FPP•peptide complexes illustrate that Lys164α interacts with the α-phosphate of FPP, while His248β and Tyr300β form hydrogen bonds with the β-phosphate of FPP. Consistent with previous reports, molecular dynamic simulations of the ternary structure of FTase•FPP•acetyl-capped tetrapeptide (acetyl-CVIM) showed that residues Arg291β, Lys164α, Lys294β, and His248β contributed to the binding of the diphosphate group (Cui et al. 2005). In addition, the first isoprene unit (C1-C5) of FPP binds in an aromatic pocket consisting of residues Tyr251β, Tyr166α, Tyr200α, His248β, and His201α.
Regulation of Farnesyltransferase Activity
In one line of investigation, it was shown that insulin stimulates phosphorylation of the α-subunit by fourfold and enhances FTase activity in 3 T3-L1 fibroblasts and adipocytes. In another study, phosphorylation of the α-subunit was shown to affect FTase activity. This conclusion was supported by experiments showing that insulin-stimulated vascular smooth muscle cells (VSMC) expressing a non-phosphorylatable mutant FTase-α (S60A, S62A) exhibited reduced phosphorylation and decreased FTase activity (Solomon and Goalstone 2001). Furthermore, expression of the non-phosphorylatable mutant FTase-α (S60A, S62A) in MCF-7 cells blocked IGF-1 and insulin-stimulated BrdU incorporation and cell count. Interestingly, phosphorylation of the α- and β-subunits of FTase was detected in rat adrenal medulla pheochromocytoma (PC12) cells (Kumar and Mehta 1997).
The potential regulation of prenyltransferase activity by glucose, which regulates insulin secretion in the pancreatic β-cell, was investigated (Goalstone et al. 2010). They showed that an insulinotropic concentration of glucose [20 mM] stimulated the expression of the α-subunit of FTase/GGTase-I by about threefold in insulin-secreting INS 832/13 cells and by about fourfold in isolated rate pancreatic islets, but not the β-subunit of FTase or GGTase-I. Moreover, an insulinotropic concentration of glucose stimulated FTase activity in INS 832/13 cells and rat islets by about 2.75- and 3.5-fold, respectively. Likewise, GGTase-I activity was increased by about 3.5 fold in INS 832/13 cells and by about fourfold in rat islets following exposure to high glucose [20 mM] (Goalstone et al. 2010).
In addition, cleavage of the α-subunit of FTase by caspase-3 during apoptosis was reported (Kim et al. 2001). Serial N-terminal deletions and site-directed mutagenesis showed that residue Asp59 of the α-subunit was cleaved by caspase-3. The cleavage resulted in the inactivation of FTase and GGTase-I. In another line of investigation, it was reported that JNK is involved in the C-terminal processing of Ras proteins (Mouri et al. 2008). Inhibition of JNK was shown to prevent C-terminal processing of H-Ras and its subsequent plasma membrane localization. Interestingly, the C-terminal processing of H- and N-Ras but not K-Ras was sensitive to JNK inhibition. Further study is needed to elucidate the biological significance of the involvement of JNK in C-terminal processing of Ras proteins.
In a recent study, Zhou et al. (2009) have shown that FTase forms a protein complex with microtubules and a histone deacetylase HDAC6 in vitro and in cells. FTase was shown to bind microtubules via its α-subunit, and that microtubules are required for the interaction between FTase-HDAC6 (Zhou et al. 2009). Furthermore, treatment with an FTI removes FTase from the protein complex and abrogates the deacetylase activity of HDAC6, suggesting that FTase regulates the function of HDAC6 in a microtubule-dependent manner (Zhou et al. 2009).
Biological Significance of FTase in Tumor Development
Mijimolle et al. (2005) addressed the biological significance of FTase by generating mice with knockout alleles for the gene encoding the β-subunit of FTase. They showed that FTase is essential for embryonic development, but dispensable for adult homeostasis. They reported that mouse embryonic fibroblasts (MEFs) deficient in FTase-β displayed a flat morphology, and reduced motility and proliferation rates. Surprisingly, H-Ras remained associated with the membrane fraction of FTase-β knockout cells, and the development of K-Ras induced tumors was not affected by FTase-β deficiency (Mijimolle et al. 2005). In a more recent study, it was shown that the FTase-β knockout allele generated by Mijimolle et al. yielded a transcript with an in-frame deletion, raising the possibility that this mutant transcript still yielded a protein with some residual FTase activity (Yang et al. 2009).
Conditional knockout of the β-subunit of GGTase-I has been shown to reduce tumor formation and increase survival in mice expressing K-Ras-induced lung cancer (Sjogren et al. 2007). Knockout of GGTase-I-β resulted in disrupted actin cytoskeleton, reduced cell migration and proliferation in fibroblasts expressing oncogenic K-Ras (Sjogren et al. 2007). More recently, Liu et al. (2010) created a conditional knockout allele for FTase-β and reevaluated the effects of FTase-β deficiency on protein prenylation, cell proliferation, and growth of K-Ras induced tumors. Furthermore, they assessed the effect of simultaneous inactivation of both FTase and GGTase-I on the development of K-Ras-induced lung cancer by breeding mice homozygous for conditional knockout alleles in both FTase-β and GGTase-I-β. They showed that FTase-β deficiency blocked proliferation of primary and oncogenic K-Ras-expressing fibroblasts, and that inactivation of FTase-β in mice inhibited K-Ras-induced lung cancer growth and improved survival (Liu et al. 2010). Moreover, simultaneous inactivation of both FTase-β and GGTase-I-β was shown to markedly reduce tumors and improve survival without apparent pulmonary toxicity (Liu et al. 2010). These findings suggested that targeting both prenyltransferases could be useful in cancer therapeutics.
Inhibitors of Farnesyltransferase (FTIs)
There has been considerable clinical interest in developing inhibitors of farnesyltransferase (FTIs) as anti-cancer agents because FTase catalyzes the processing of the Ras family members. In previous preclinical studies, FTIs were shown to block the growth of numerous tumors, including H-, K-, and N-Ras transgenic mouse models (Kohl et al. 1995). However, K- and N-Ras both undergo alternative prenylation when farnesylation is inhibited (Whyte et al. 1997), suggesting that the effects of FTIs on tumor growth might be due to FTI targets other than Ras. Clinical studies of FTI have been reported. For example, a phase I study to evaluate the tolerance and benefice of a combination of FTI tipifarnib (R115777/Zarnestra®, Janssen Research Foundation) and sorafenib (a multi-kinase inhibitor) in patients with advanced cancer has been reported (Chintala et al. 2008).
Recently, a nonpeptidic FTI, LB42708, has been shown to suppress vascular endothelial growth factor-induced angiogenesis by inhibiting Ras-dependent MAPK and PI3K/Akt signaling pathways (Kim et al. 2010). FTIs have also been shown to inhibit mammalian target of rapamycin complex 1 (mTORC1) signaling (Gau et al. 2005). See chapters on MAP kinases, Akt(PKB) and mTOR for signaling pathways. Moreover, it has been shown that FTIs preferentially inhibit mTORC1 signaling in non-small cell lung cancer cells (Zheng et al. 2010).
In addition, FTIs were evaluated for the treatment of Hutchinson Gilford Progeria Syndrome (HGPS), a rare condition that arises from accumulation of farnesylated prelamin A (Fong et al. 2006). Costello syndrome (CS) is a congenital disorder that is characterized by mental retardation, distinctive facial appearance, cardiovascular abnormalities, tumor predisposition, and skin and musculoskeletal abnormalities (Lin et al. 2005). Interestingly, about 80% of CS patients have H-Ras mutations, primarily H-Ras G12S (34G → A), suggesting that the development of FTIs as therapeutics for the treatment of CS could be promising as H-Ras is a farnesylated protein.
To elucidate the mechanism of FTI selectivity, Reid T.S. and Beese L.S. examined the crystal structures of FTIs, R115777 (tipifarnib/Zarnestra®) and BMS-214662 (Bristol-Myers Squibb), complexed with mammalian FTase (Reid and Beese 2004). Both FTIs are selective toward FTase, with almost no activity against GGTase-I. It was shown that both FTIs bind to the active site of FTase, and that drug binding does not induce a change in the structure of the active site of the enzyme. Furthermore, they showed that both FTIs bind as a ternary complex with farnesyl diphosphate (FPP) and coordinate the catalytic zinc ion (Reid and Beese 2004).
Protein farnesyltransferase is a heterodimeric enzyme catalyzing farnesylation of proteins ending with the CA 1 A 2 X motif found at the C-termini of proteins such as Ras, Rheb, nuclear lamins and the γ-subunit of some heterotrimeric G-proteins. The enzyme consists of the α- and β-subunits and its crystal structure has been determined. Regulation of the enzyme activities has been reported. Finally, a variety of small molecule inhibitors have been developed against the enzyme with the intention to inhibit membrane association of signaling proteins such as Ras.
- Gau CL, Kato-Stankiewicz J, Jiang C, Miyamoto S, Guo L, Tamanoi F. Farnesyltransferase inhibitors reverse altered growth and distribution of actin filaments in Tsc-deficient cells via inhibition of both rapamycin-sensitive and -insensitive pathways. Mol Cancer Ther. 2005;4:918–26.PubMedCrossRefGoogle Scholar
- Kim KW, Chung HH, Chung CW, Kim IK, Miura M, Wang S, Zhu H, Moon KD, Rha GB, Park JH, Jo DG, Woo HN, Song YH, Kim BJ, Yuan J, Jung YK. Inactivation of farnesyltransferase and geranylgeranyltransferase I by caspase-3: cleavage of the common alpha subunit during apoptosis. Oncogene. 2001;20:358–66.PubMedCrossRefGoogle Scholar
- Kim CK, Choi YK, Lee H, Ha K-S, Won M-H, Kwon Y-G, Kim Y-M. The farnesyltransferase inhibitor LB42708 suppresses vascular endothelial growth factor-induced angiogenesis by inhibiting ras-dependent mitogen-activated protein kinase and phosphatidylinositol 3-kinase/Akt signal pathways. Mol Pharmacol. 2010;78:142–50.PubMedCrossRefGoogle Scholar
- Lin AE, Gripp KG, Kerr BK. Management of genetic syndromes. 2nd ed. Hoboken: Wiley; 2005.Google Scholar
- Liu M, Sjogren AK, Karlsson C, Ibrahim MX, Andersson KM, Olofsson FJ, Wahlstrom AM, Dalin M, Yu H, Chen Z, Yang SH, Young SG, Bergo MO. Targeting the protein prenyltransferases efficiently reduces tumor development in mice with K-RAS-induced lung cancer. Proc Natl Acad Sci U S A. 2010;107:6471–6.PubMedPubMedCentralCrossRefGoogle Scholar
- Sjogren AK, Andersson KM, Liu M, Cutts BA, Karlsson C, Wahlstrom AM, Dalin M, Weinbaum C, Casey PJ, Tarkowski A, Swolin B, Young SG, Bergo MO. GGTase-I deficiency reduces tumor formation and improves survival in mice with K-RAS-induced lung cancer. J Clin Invest. 2007;117:1294–304.PubMedPubMedCentralCrossRefGoogle Scholar
- Tamanoi F, Sigman DS (Eds). The enzymes. vol. 21. San Diego: Academic Press; 2001.Google Scholar