Vol. 3, 461-469, July 1992 Cell Growth & Differentiation 461

Research Capsule

Protein Prenylation: Key to Ras Function andCancer Intervention?’

Roya Khosravi-Far,2 Adrienne D. Cox,2 Kiyoko Kato,3and Channing J. Der2 �

La lolla Cancer Research Foundation, La lolla, California 92037

Ihe frequent association of mutated, oncogenic formsof cellular ras proteins with a broad spectrum of humanmalignancies has prompted intensive investigations intoidentifying their role in normal cellular physiology andinto establishing the contribution of aberrant ras functionto human tumorigenesis (1-4). Despite considerableknowledge of the structural and biochemical propertiesof ras proteins, we remain ignorant of the function of rasproteins. The recent discovery that ras transformingactivity is critically dependent on modification(prenylation) by a farnesyl isoprenoid, an essentialintermediate in cholesterol biosynthesis, has unveiledpotentially important clues to ras function and identifiednovel prospects for cancer therapy (5-8). In this researchcapsule, we will summarize our present understandingof the role of protein prenylation in ras biological activity,and we will assess the promises and problems ofpharmacological approaches to antagonizing prenylation.Finally, understanding the role of protein prenylation inras function will provide a foundation for defining thesignificance of this previously obscure proteinmodification for the functions of a very diverse andgrowing number of isoprenylated proteins (9, 10).

CXXX-signaled Processing and Membrane AssociationAre Critical for ras Transforming Activity

The three human ras genes (H-, K-, and N-ras) encodefour structurally related proteins of 188-189 amino acidswhose functions as molecular switches are dictated byinteraction with guanine nucleotides (11) (Fig. 1). Thenormal protein cycles between an active, GTP-com-plexed state, and an inactive, GDP-complexed state.Regulatory factors have been identified which stimulateguanine nucleotide dissociation (GDSs5) to promote ras-GIP formation (12-1 5) or which act as GIPase-activatingproteins (ras GAP, NFl) to negatively mediate ras activity(16-21). Oncogenic ras proteins contain single aminoacid substitutions at residues 12, 13, or 61 (1-4), aredefective in GDP/GTP cycling, persist constitutively in

Received 4/22/92.n Our research studies were supported by grants from the NIH (CA42978,

CA52072. CA55008, and CA08791), the American Cancer Society (FRA57817), and the Tobacco Related Disease Research Program (2FT0039(.2 Present address: Department of Pharmacology. University of North

Carolina, Chapel Hill, NC 27599.3 Present address: Kyushu University, Beppu, Oita, lapan.4 To whom requests for reprints should be addressed, at Department ofPharmacology, University of North Carolina. Chapel Hill, NC 27599.

S The abbreviations used are: GDS, guanine nucleotide dissociation slim-ulator; PDE, phosphodiesterase; HMG, hydroxymethylglutaryl.

the active, GTP-bound form, and chronically stimulatean as yet unidentified growth-stimulatory pathway(s) topromote the uncontrolled growth of the tumor cell (22,23).

In addition to guanine nucleotide interactions, thesecond critical requirement for ras function is associationwith the inner face of the plasma membrane (6-8). rasproteins lack the conventional transmembrane or hydro-phobic sequences associated with other membrane-as-sociated proteins and are initially synthesized as soluble,cytoplasmic proteins (24). Their membrane association(25) is triggered by a series of closely linked posttransla-tional processing steps that are signaled by the conservedCOOH-terminal CXXX motif present in all ras proteins(26-28) (Fig. 2). The first modification is the addition, viaa thioether bond to the cysteine residue of the CXXX, ofa Cn5 farnesyl isoprenoid moiety (29-31). The farnesylmoiety is a product of mevalonate (Fig. 3), the essentialprecursor of all cellular isoprenoids, including cholesterol(32). The other modifications to ras proteins are proteo-lytic removal of the three terminal XXX residues (33, 34)

and, finally, carboxyl methylation of the now terminalfarnesyl-cysteine residue (27, 33) (Fig. 2).

The critical contributions of CXXX-signaled processingto ras function have been demonstrated by two comple-mentary lines of experimental evidence. First, COOH-terminal structural mutants of ras proteins that lack eitherthe cysteine or XXX residues of the CXXX motif do notundergo any of the CXXX-signaled processing steps, donot associate with the plasma membrane, and are com-pletely nontransforming (29, 35, 36). Second, inhibitorsof ras isoprenylation (e.g., lovastatin) result in the accu-mulation of unprocessed, cytosolic proteins and alsoprevent ras biological activity (30, 37). Thus the CXXX-signaled modifications are clearly critical in the traffickingof ras proteins to the plasma membrane, and membraneassociation is apparently essential to trigger rastransformation.

Farnesyl Addition Is the Critical Modification for rasMembrane Association and Transforming Activity

Although the three linked CXXX-signaled modificationsimpart a hydrophobic nature to the protein and arecritical for ras function, the precise contribution andfunctional role of each modification are presently notclear. Isoprenylation provides the greatest contributionto increased protein hydrophobicity; the proteolytic re-moval of the XXX residues may then enhance membraneassociation by allowing a better interaction of the prenylgroup either with the lipid bilayer or with a possiblemembrane receptor(s). Carboxyl methylation may pro-vide further enhancement to the hydrophobicity, and/oralter the conformation or ionic charge, of ras proteins.Whereas farnesylation is apparently a stable addition (31,38), carboxyl methylation is reversible (27, 39, 40) andmay have an additional function in regulating association

I HVI1

/II

H-ras HKL RK LNPPDESGPG�r�MS�j�I

K-ras4A V R L K K I S K E E K T P G�V K I K

K-ras4B H ________N-ras Y R M K K L N S S D D G T 0 G�j M G L

I PLASMA MEMBRANE BINDING DOMAIN I

� = GTP BINDING DOMAINS � = HYPERVARIABLE DOMAIN

L�1 = EFFECTOR DOMAIN/GAP INTERACTION � = CXXX MOTIF

462 Re’se’,mrc h (,ipsiile’

I GTP BINDING/HYDROLYSIS DOMAIN-

r IEI �

= PALMITYLATED CYSTEINE �..‘ = LYSINE-RICH DOMAIN

I ig. 1 - Plasn)a nue’n�lIrane’-bind-ing (Iill))ain) ot hun),mn) r,is Ilriite’in)s.

The’ four hun))an) r,ms Proteins cx-hillil Ihe greatest divergence’ in

the’ C()OH-terninn,mI se’que’n) c’s

(designated the’ hypervari,mble do-m,iin)) directly ,uljace’nt to the

CXXX n)olif. )n Iude’d in) this re-glnin) ,Ire’ the l)aln))itylale’(I cysleinere’siduc’s of H-, N., ,lI)(1 K.r,is4Aan)(l the’ lysinc-ric h cl(inlain) (II K-

ras4B. K-ras4A ,)n)(l K-r,ss4B are’v,lrian)ts c)I hun)1,in) K-ras Protein

lh,it (lilfer on)Iy in) their COOH-te’rninin),lI seque’nc c’s ,ss ,i (onse’-

(lci(’l)( C’ of alte’rnat �e’ lou rI h cod-ing e’xons.

with the plasma niembrane. The observations that thedegree of carboxyl methylation of 21 -29-kilodalton pro-teins varied with cellular growth state (41), exhibitedrapid methylation/demethylation by separate enzymaticsteps (40), and exhibited a guanine nucleotide depend-ence (42) are supportive of this possibility. However,there is presently no evidence for regulated methylation/demethylation of ras proteins in particular (33).

Since blocking ras isoprenylation also prevents thesubsequent two CXXX-signaled modifications (Fig. 2), theexact contribution of the individual CXXX-signaled mod-ifications to membrane association and transforming ac-tivity has been difficult to assess. However, recent in vitroand in vivo studies have provided some preliminary clues.First, reconstitution of each CXXX-signaled modificationin vitro using a reticulocyte lysate assay determined thatthe isoprenoid modification alone promoted a limitedmembrane association of K-ras4B, which was increasedboth by proteolytic removal of the XXX residues and bycarboxyl methylation (43). Second, a mutant K-ras4Bprotein (Tyr-187), which undergoes isoprenylation butnot proteolysis or carboxyl methylation in vivo, neverthe-less exhibited significant membrane association (50%)

and transforming activity (44). Third, a yeast mutant(stel4) that lacked ras methyl transferase activity exhib-ited no impairment of RAS function or cell viability andexpressed a partially membrane-associated RAS protein(45). Since RAS function is essential for yeast viability,these results suggest that a nonmethylated RAS proteinretains at least some degree of normal function. Alto-gether, these results suggest that, although all threeCXXX-signaled modifications are required for optimalmembrane association, isoprenylation alone is sufficientto promote the biological activity of both normal andoncogenic ras proteins.

The suggestion that isoprenylation is the critical mod-

ification for ras function contrasts sharply with studies of

the yeast a-factor mating hormone. a-factor undergoesthe same three CXXX-signaled modifications as ras, and,like unprocessed ras proteins, unprocessed a-factor pep-tides are completely inactive for biological activity (46,47). However, since both nonfarnesylated variants andnonmethylated variants of a-factor are drastically im-paired in biological activity, both modifications are criti-cal for a-factor function (46, 48, 49). Thus, the specificcontribution of each CXXX-signaled modification to pro-tein function may well differ among different isoprenyl-ated proteins. Furthermore, although carboxyl methyla-tion may be dispensable for ras transforming activity, itis possible that this modification may still play an integralrole in the function of normal ras proteins in mammaliancells.

CXXX-signaled Modifications Require Complementa-tion by Other COOH-terminal Lipid Modifications andSequences to Promote Full ras Function

Although the CXXX-signaled modifications are clearlycritical for the plasma membrane association of ras pro-teins, observations from the studies of ras mutants andof other similarly modified proteins suggest that thesemodifications alone can confer neither optimal mem-brane affinity nor specific association of ras proteins withthe plasma membrane. First, the lamin B protein under-goes the same three CXXX-signaled modifications as ras(50), yet it localizes to the nuclear envelope (51, 52).Second, a significant number of prenylated proteins arecytosolic rather than membrane associated (38, 53, 54).

Instead, observations from several studies (29, 55, 56)suggest that sequences directly upstream of the CXXXmotif (cysteines or lysine-rich sequences) provide sec-ondary membrane-targeting signals that complement theCXXX-signaled modifications to Promote both the avidityand specificity of the plasma membrane interactions

I Cys Cys-X�X�X

Endoprotease 4�....x.x-x

I Cys Cys]

S - CH2 fs,##P4�,%#4�y#Pss4,rni��

Carboxyl Methyl Transferase ?

Palmityl Transferase ?

S - CH2 �

Cys Cysj.OCH3

S. CH2fy\/�ys�fy

Membrane-bound I y Cys j. OCH3Mature,

0

S - CH2 �

Cell Growth A Differentiation 463

Precursor,

Cytoplasmic I Cys CyS.X.X-X

I ig 2. Posttranslational proc e’ssingelf ras l)rote’Il)s. HMG CoA redu tase

inhibitors (leivastat ill, ( on)pactin(block the’ farnesylation stel) and Ire’-ve’nt all subsequent proc e’ssing steps.

Farnesyl Transferase � Compactin, Lovastatin Inhibition

required for ras function.In the case of the H-, N-, and K-ras4A proteins, a

subsequent modification by the fatty acid palmitate issignaled by cysteine residue(s) upstream of the CXXXmotif (Fig. 1). Mutant ras proteins that lack these cysteineresidues retain CXXX-signaled modifications but do notundergo palmitylation, no longer display a plasma mem-brane localization, and are predominantly cytosolic pro-teins (29, 55). Thus, palmitate complements the CXXX-signaled modifications to promote full association withthe plasma membrane. Additionally, since palmitate hasa very high turnover rate (>70-fold higher than the half-life of ras proteins), palmitylation may potentially play anadditional regulatory role in ras biological activity (57).

The K-ras4B protein, on the other hand, lacks anyupstream cysteines; instead, it contains stretches of lysineresidues (Fig. 1). When the positively charged lysineresidues were replaced with neutral glutamines, the pro-tein was rendered cytosolic (55). Replacement of thelysines with similarly charged arginine residues retainedmembrane association, suggesting that the lysine resi-dues provide a positively charged domain that may pro-mote interaction with negatively charged phospholipidhead groups (56). Thus, these lysine residues serve afunction analogous to that of palmitylation. However,since these residues, in contrast to palmitate, are stableelements of K-ras4B, the K-ras4B protein may associatewith the plasma membrane in a manner distinct from itspalmitylated ras counterparts.

The importance of the palmitate- and lysine-rich com-ponents for the plasma membrane localization of rasproteins has also been demonstrated by characterizingthe consequences of introducing different ras COOH-terminal sequences to heterologous, cytosolic proteins.For example, addition of tetrapeptide sequences repre-senting the ras CXXX sequences to the COOH-terminusof Protein A promoted the isoprenylation of these chi-meric proteins, but localization to the plasma membrane

was not observed (29). However, the addition of theCOOH-terminal 10 residues of H-ras (including bothpalmitylated cysteines) or 1 7 residues of K-ras4B (includ-ing the lysine-rich sequences) promoted efficient asso-ciation of Protein A with the plasma membrane (56).Thus, the plasma membrane targeting information is ap-parently found in the ras sequences directly adjacent tothe CXXX motif. An analogous result was observed instudies of the ras-related rab/YPT proteins, in which theirlocalization to specific membranes ofthe endocytic path-way was dictated by COOH-terminal sequences imme-diately upstream of the isoprenylated terminal cysteineresidues (58).

An interesting, and somewhat unexpected, observa-tion has been that, although the presence of palmitylationor lysine-rich sequences is critical for ras association withthe plasma membrane, oncogenic ras proteins that lackthese elements but retain the CXXX-signaled modifica-tions also retain strong transforming activity in NIH 313cells (29, 55). These results suggest that an associationwith the plasma membrane may not be a critical require-ment for ras transforming function. They also suggest thatthe critical requirement for CXXX-signaled modificationsin transforming activity may involve a role unrelated topromoting plasma membrane interactions. One possiblerole may be to facilitate the interaction of ras proteinswith critical regulatory proteins. For example, GDS stim-ulation of GDP/GTP exchange on K-ras4B is dependenton the presence ofthe CXXX-signaled modifications (1 5).Alternatively, since these transformation analyses havebeen performed in the rather permissive NIH 313 cellline, it is still possible that membrane association may berequired for the biological activity of oncogenic ras intumor cells. Finally, since oncogenic ras proteins arechronically activated, and therefore presumably inde-pendent of stimulation via upstream mitogenic signals, aplasma membrane association may no longer be impor-tant for oncogenic ras function but may remain essential

464 Research Capsule

I, A. D. Cox and C. I. Der, unpublished observations.

Hydroxymethyiglutaryl C0A

HMG C0A reductase � �- - - - - Lovastatin, Compactin Inhibition

� MEVALONATE

Mevalonate . P

Mevalonate . PP Isopentenyl . tRNA

Isopentenyl . PP (C5 ) � Dimethylallyl . PP

r UGeranyl . PP (C10)

�-(_#{149}‘� Squalene - - -0#{176} � CHOLESTEROLj

F-’��Farnesyl . PP (C15) -�---� Heme A

I--�

� FARNESYLATEDPROTEINS

e.g., ras, lamin B1Geranylgeranyl . PP (C�) GGTase � GERANYLGERANYLATED

e.g., rap, rab, rac, rho

Fig. 3. Mevalonate and isoprenoid biosynthetic pathway. Mevalonate isthe essential precursor for all isoprenoids, including cholesterol and the

Cns farnesyl and C20 geranylgeranyl moieties that are added to proteinsby farnesyl (FTase) or geranylgeranyl (GCTase( transferase. HMG CoAreductase inhibitors (e.g., lovastatin( block mevalonale biosynthesis and

prevent all isoprenoid biosynthesis.

for normal ras function. Thus, the significance of mem-brane association for ras function remains to be clarified.

The ras proteins display their greatest divergence inthe COOH-terminal 19-20 residues directly adjacent tothe CXXX motif (designated the hypervariable domain;see Fig. 1). Thus, in addition to cysteine and lysineresidues, other sequences unique to each ras proteinmay also contribute to the avidity and specificity ofplasma membrane interactions, and possibly to otherroles important for ras function. Since the membraneassociation of ras proteins is presumably required forpromoting interaction with membrane-associated up-stream and downstream components oftheir signal trans-duction pathways, the fact that each of the four humanras proteins possesses distinct COOH-terminal modifi-cations and sequences (Fig. 1 ) may therefore reflect theirassociation with distinct pathways.

Normal and Oncogenic ras Proteins May HaveDifferent Membrane Association Requirementsfor Biological ActivityAlthough the majority of studies have concentrated onthe role of posttranslational modifications on ras trans-forming activity, several observations suggest that themembrane association requirements of normal versusoncogenic ras proteins may be distinct. First, studies haveshown that ras modification by the fatty acid myristate

can functionally replace the CXXX-signaled modificationsto promote the membrane association required for on-cogenic ras transforming activity (59, 60). These resultssuggested that the CXXX-signaled modifications are notrequired for the intrinsic biological activity of ras proteinsand that these modifications merely provide a nonspe-cific membrane targeting signal that can be replaced withmyristate. However, the unexpected activation of trans-forming activity when normal ras is modified by myristate(60) suggests that myristate cannot correctly facilitatenormal ras function, which instead may require the pre-cise contributions of CXXX-signaled modifications.

A second distinction between the membrane associa-tion requirements of normal and oncogenic ras proteinswas observed with studies addressing the importance ofspecific isoprenoid modification in ras function (56, 61).Whereas ras proteins are modified by a Cns farnesylisoprenoid moiety, the majority of mammalian proteinsare modified instead by a C20 geranylgeranyl isoprenoidgroup (54, 62, 63). In particular, nearly all of the othermembers of the ras superfamily are modified by geran-ylgeranyl, and to date, none of these related proteins hasdemonstrated oncogenic activity (8). Therefore, two re-cent studies have addressed whether addition of eitherisoprenoid can promote ras biological activity. Interest-ingly, CXXX mutants of oncogenic ras proteins which arespecifically modified by a geranylgeranyl, rather thanfarnesyl, isoprenoid retain full transforming activity (56,61). In contrast, geranylgeranyl-modified normal ras pro-teins were observed to possess growth-inhibitory activityin transfected NIH 313 mouse fibroblasts (61). Similarly,studies in yeast suggested that only farnesylated, but notgeranylgeranylated, RAS protein is capable of stimulatingadenylate cyclase activation (64). These observations in-dicate that oncogenic ras protein requires only a gener-alized membrane association which can be equally facil-itated by different lipid additions, whereas normal rasmay require specific modification by a farnesyl isopren-oid for its function.

The basis for specific isoprenoid modification for nor-mal ras function is presently unclear. One possibility isthat there are specific receptors that interact with eitherfarnesylated or geranylgeranylated proteins. Althoughsuch prenylation-specific receptors have not been iden-tified, the recent observation that geranylgeranyl-modi-fied forms of K-ras4B (56) or H-ras” proteins exhibitdifferent intracellular localizations from their authenticfarnesylated counterparts is consistent with this possibil-ity. Furthermore, there are indications from other studiesthat specific receptors for lipid-modified proteins (e.g.,lamin B p6Ov-src) do exist (65, 66). Thus, the identificationof a membrane component that specifically interacts withfarnesylated ras proteins may provide a clue for identi-fying the elusive biochemical function of these proteins.

Isoprenylation Is a Common Modification ofOther Proteins

Although modification of proteins by prenylation hasbeen known for some time (67), and was first associatedwith yeast mating factors (46), only recently have theidentity of isoprenylated proteins in mammalian cells

Cell Growth & Differentiation 465

been established (6-10). Although the nuclear laminsand ras proteins represent the first characterized isopren-ylated mammalian proteins, it is becoming apparent thatother proteins that undergo this lipid modification rep-resent critical components involved in the regulation ofa range of diverse cellular functions.

The majority of recently identified isoprenylated pro-teins are additional members of the ras protein superfam-ily, whose specific members have been implicated in theregulation of intracellular vesicular transport (rab/YPT1)(68), cytoskeletal organization (rho) (69), the oxidativeburst of phagocytic cells (rac) (70, 71), and negativegrowth (rap/Krev-1) (72). In those cases that have beenstudied, isoprenoid modification has been determinedto be important for the involvement of ras-related pro-teins in these activities. For example, rab/YPT proteinisoprenylation is apparently important for the associationof these proteins with specific intracellular compartments(73) and in their regulation of intracellular transport proc-esses (74, 75). rab/YPT proteins are believed to cyclebetween distinct membrane subcompartments and be-tween membrane-associated and cytosolic states to fa-cilitate transport processes (68, 76), and isoprenylation istherefore expected to play a key role in rab function.

A number of isoprenylated proteins other than ras alsorepresent critical elements in signal transduction path-ways. Isoprenylation is a critical modification for the ysubunits of the heterotrimeric G proteins, and this mod-ification is important for their function in receptor-me-diated signal transduction pathways in mammalian (77)and yeast (78) cells. In addition to the y subunit oftransducin, several other components of the retinal signaltransduction pathway are also modified by isoprenyla-tion. Rhodopsin kinase is modified by a farnesyl moiety,and its serine/threonine phosphorylation activity is mark-edly reduced in the absence of isoprenylation (79). Athird component ofthis pathway, retinal cyclic GMP PDE,is composed of four subunits, of which two are modifiedby farnesyl (a subunit) or geranylgeranyl (�3 subunit) moie-ties (80). Prenylation of PDE is believed to be importantfor its membrane attachment. Finally, several additionalsignal transduction pathways are blocked by inhibitors ofprotein prenylation, implicating the involvement of iso-prenylated proteins in these processes (81, 82).

In summary, determining the role of isoprenylation inras function will provide insights into the contributionsof isoprenoid modification to the function of other iso-prenylated proteins that are involved in signal transduc-tion, intracellular vesicular transport, and other diversecellular processes. Present evidence suggests that, likeras proteins, the biological activities of these proteins arealso dependent on isoprenoid addition. Generally, iso-prenoid addition is believed to influence membrane andprotein-protein interactions, to contribute to specific in-tracellular localizations, and to regulate biological activ-ity. However, it is likely that the precise role for proteinprenylation will be distinct and unique for each isopren-

ylated protein.

Protein Prenyl Transferases: Targets for CancerTreatment?The discovery that the function of oncogenic ras proteinis critically dependent on its modification by an essentialintermediate in cholesterol biosynthesis has attractedsignificant interest in blocking ras isoprenylation as a

means of cancer chemotherapy (5). This possibilityseemed promising when early studies demonstrated theability of drugs such as lovastatin and compactin (32) toinduce cellular growth arrest of cell lines in vitro (83-86)and oftumors in vivo (87). Since these fungal compoundsinhibit mevalonate synthesis (Fig. 3), and consequentlyall isoprenoid biosynthesis, one possible basis for growthinhibition was thought to be via prevention of ras iso-prenylation. However, it has been shown that this growtharrest is not a consequence ofantagonizing ras-transform-ing activity by blocking ras processing (88) but is presum-ably due to perturbations of other mevalonate-depend-ent processes.

Although general inhibitors of mevalonate biosynthesis(e.g., lovastatin and compactin) have been useful clini-cally for lowering cholesterol levels, their use in cancerchemotherapy has several significant drawbacks. First,since mevalonate is the precursor for all isoprenoid bio-synthesis, including cholesterol (Fig. 3), lovastatin willclearly affect other isoprenylation pathways. Second, ithas been shown that protein prenylation is 100-fold lesssensitive than cholesterol biosynthesis to the effects oflovastatin (89). Finally, ras proteins represent but a smallfraction of isoprenylated proteins. Thus, the relativelynonspecific action of these mevalonate inhibitors maylimit their usefulness for blocking oncogenic ras functionin tumors.

A more effective approach to pharmacological inter-vention in oncogenic ras activity is therefore the designof drugs that block ras isoprenylation farther down themevalonate pathway (Fig. 3). The identification and iso-lation of a protein:prenyl transferase that catalyzes theisoprenoid addition to ras proteins (90-92) has provideda promising target for such intervention. Furthermore,although protein prenylation is a relatively common mod-ification, recent studies demonstrate that there is a familyof protein:prenyl transferases, each of which may modifyonly a subset of isoprenylated proteins (Table 1). Forexample, isoprenylated proteins that terminate in CXXXmotifs are specifically modified by either a farnesyl or ageranylgeranyl transferase (61, 93-96), and present evi-dence indicates that geranylgeranyl-modified proteinsthat terminate in either CC, CXC, or CCXX motifs areprocessed by different geranylgeranyl transferases (73,

95, 97, 98). Therefore, it is possible that specific blockageof ras farnesyl transferase will not perturb the processing

and function of the majority of other isoprenylated pro-teins.

However, despite the promise of blocking oncogenicras function via antagonism of ras farnesyl transferase

activity, there are still several potential areas of nonspe-cific complications to this approach. First, other farnesyl-

ated cellular proteins (e.g., nuclear lamins, transducin �y)

apparently are substrates for the same farnesyl transfer-ase as ras proteins (90), so perturbation of their functionsin cell division and visual signal transduction might benonspecific consequences of such attempts to interferewith the function of oncogenic ras. Second, present

evidence suggests that isoprenylation is critical for thebiological activity of normal, as well as oncogenic, ras

proteins. Since ras proteins are expressed in virtually alltissues (99, 100), and since normal ras function is criticalfor cellular proliferation (101), antagonizing normal rasfunction may result in serious consequences for theviability of normal cells. Nevertheless, if these compli-

466 Research Capsule

Table 1 Summary o f isopren ylated proteins and prote in prenyl transferases

SequenceModifications

Enzyme Subunits Yeast genes ExampleF’ GGb �xxx’ Meth”

cxxsM

CAQ

+ - + + FTase is

/3,RAM2DPR1/RAM1

H-ras, N-rasK-ras4B, lamin BRAS1a-factorPDEa�

CXXLF

- + + + GGTase I a/32

RAM2CAL1/CDC4

rapla, raplb, PDE-JJ’G25K (brain(

Cc - + - - GGTase II ?

?

?

BET2/ORF2

rabla, rablb

rab2

cxc - + - + GGTase Ill? ? ? rab3a, rab6

ccxx - + ? ? GGTase IV? ? ? rab5

Farnesylation.b Geranylgeranylation.

CProteolytic removal of the three terminal residues.d Carboxyl methylation of the cysteine residue.

#{176}2’,3’-Cyclic nucleotide phosphodiesterase.

cations can be overcome, inhibitors of ras farnesyl trans-ferase activity may one day provide effective and selec-tive inhibition of tumor cells harboring oncogenic rasproteins.

Synthetic peptides corresponding to the CXXX se-quence of ras proteins have been demonstrated to bepotent inhibitors of ras prenylation in vitro (90-92) andras biological activity in vivo (37). Although peptidesprovide the necessary specificity to target ras isoprenyla-tion, principal limitations for the use of such peptides incancer treatment will be in their introduction into cellsand in delivery of these inhibitors to their intracellulartargets. Nevertheless, such peptide inhibitors have beenvery useful for characterizing ras farnesyl transferasespecificity, and improvements in the design and deliveryof peptides may hold promise for this approach to cancerchemotherapeutics.

The enzymes that catalyze the remaining two CXXX-signaled modifications and palmitylation might provideadditional targets for blocking ras function. Membrane-associated enzymatic activities for proteolysis (43, 102),carboxyl methylation (43, 45, 103-105), and palmitateaddition (106) are only now beginning to be isolated andcharacterized. However, these enzymatic activities areapparently less specific in their substrate utilization thanis the ras farnesyl transferase. For example, the carboxylmethyl transferase responsible for modification of rasproteins apparently also methylates geranylgeranyl-mod-ified CXXX proteins (39, 40, 45, 103, 104). Additionally,since farnesylation alone is sufficient for ras transformingactivity (44), there may be only a limited benefit inblocking these modification steps to prevent oncogenicras function in tumors.

Future DirectionsThe finding that ras proteins are isoprenylated by afarnesyl isoprenoid moiety has stimulated studies ad-dressing several important questions. First, it is clear thatisoprenoid modification is the key addition that triggersras biological activity, and it is likely that isoprenylationpromotes the interaction of ras proteins with targets

necessary for transformation. Therefore, isoprenylationmay provide a convenient handle on identifying the keycomponents for ras biochemical function, and an impor-tant aim for future studies will be to determine whetherreceptors for farnesylated proteins exist. Second, sinceprotein prenylation is emerging as a modification impor-tant for the function of a diverse group of proteins,understanding the contribution of isoprenylation to rasbiological activity may establish important principles forunderstanding the contributions of prenylation to otherproteins. Finally, further studies on the enzymology andsubstrate specificities of protein:prenyl transferases willprovide important information for the identification anddesign of pharmacological agents that preferentiallyblock ras isoprenylation. The identification of such drugswill then allow determination of whether specificallyblocking ras function will be useful for cancer chemo-therapy and whether oncogenic ras contributes clinicallyto the aberrant growth properties of the malignant cell.

Acknowledgments

We thank Cecilia Ruggiero for excellent preparation of the illustrations.

References

1. Barbacid, M. ram genes. Annu. Rev. Biochem., 96: 779-827, 1987.

2. Bos, I. L. The ras gene family and human carcinogenesis. Mulal. Rem.,195: 255-271, 1988.

3. Der, C. I. Cellular oncogenes and human carcinogenesis. Clin. Cheni.,

33: 641-646, 1988.

4. Bos, I. L. ras oncogenes in human cancer: a review. Cancer Res., 49:

4682-4689, 1989.

5, Gibbs, I. B. Ras C-terminal processing enzymes- .new drcig large’ls?Cell, 65: 1-4, 1991.

6. Rine, I.. and Kim, S-H. A role for isopre’noid lipids in the localizationand function (If an oncoprotein. New Biologist. 2: 219226, 1990.

7. Der, C. I.. and Cox, A. D. Isoprenoid mc)dification and plasma mem-brane association: critical factors for ram oncogenicity. Cancer Cells, 3:331-340, 1991.

8. Cox, A. D., and Der, C. I. The ras/cholesterol connection: implications

for ras oncogenicity. Cril. Rev. Oncogenesis, in press, 1992.

9. Maltese, W. A. Posttranslational modification of Prote’ins by isopre’n-

oids in mammalian cells. FASEB I.. 4: 3319-3328, 1990.

Cell Growth & Differentiation 467

10. Glomset, I. A., GeIb, M. H., and Farnsworth, C. C. Prenyl proteins in

eukaryotic cells: a new type of membrane anchor. Trends Biochem. Sci.,15: 139-142, 1990.

11. Haubruck, H., and McCormick, F. Ras p21: effects and regulation.

Biochim. Biophys. Ada, 1072: 215-229, 1991.

12. West, M., Kung, H-F., and Kamata, T. A novel membrane factor

stimulates guanine nucleotide exchange reaction of ras proteins. FEBSLetl., 259: 245-248, 1990.

13. Wolfman, A., and Macara, I. G. A cytosolic protein catalyzes the

release ofGDP from p21’S’. Science (Washington DC(, 248: 67-69, 1990.

14. Downward, I., Riehl, R., Wu, L., and Weinberg, R. A. Identification

of nucleotide exchange-promoting activity for p2lras. Proc. NatI. Acad.Sci. USA, 87: 5998-6002, 1990.

15. Mizuno, T., Kaibuchi, K., Yamamoto, T., Kawamura, M., Sakoda, T.,

Fujioka, H., Matsuura, Y., and Takai, Y. A stimulalory GDP/GTP exchange

protein for smg p21 is active on the post-translationally processed formofc-Ki-ras p21 and rhoA p21. Proc. NatI. Acad. Sci. USA, 88: 6442-6446,

1991.

16. McCormick, F. ras GTPase activating protein: signal transmitter andsignal terminator. Cell, 56: 5-8, 1989.

17. Hall, A. ras and GAP-who’s controlling whom? Cell, 61: 921-923,

1990.

18. Xu, G., O’Connell, P., Viskochil, D., Cawthon, R., Robertson, M.,

Culver, M., Dunn, D., Stevens, I., Gesteland, R., White, R., and Weiss, R.The neurofibromatosis type 1 gene encodes a protein related to GAP.

Cell, 62: 599-608, 1990.

19. xu, G., Lin, B., Tanaka, K., Dunn, D., Wood, D., Gesteland, R.,

White, R., Weiss, R., and Tamanoi, F. The catalytic domain of theneurofibromatosis type 1 gene product stimulates ras GTPase and com-plements ira muants of S. cerevisiae. Cell, 63: 835-841, 1990.

20. Martin, G. A., viskochil, D., Bollag, G., McCabe, P. C., Crosier, W.

I.. Haubruck, H., Conroy, L., Clark, R., O’Connell, P., Cawthon, R. M.,Innis, M. A., and McCormick, F. The GAP-related domain of the neuro-fibromatosis type 1 gene product interacts with ras p21. Cell, 63: 843-

849, 1990.

21. Ballester, R., Marchuk, D., Boguski, M., Saulino, A., Letcher, R.,Wigler, M., and Collins, F. The NFl locus encodes a protein functionallyrelated to mammalian GAP and yeast IRA proteins. Cell, 63: 851-859,

1990.

22. Bourne, H. R., Sanders, D. A., and McCormick, F. The GTPase

superfamily: conserved structure and molecular mechanism. Nature(Lond.(,349: 117-126, 1990.

23. Bourne, H. R., Sanders, D. A., and McCormick, F. The GTPasesuperfamily: a conserved switch for diverse cell functions. Nature’ (Lond.(.

348: 125-132, 1990.

24. Shih, T. Y., Weeks, M. 0., Gruss, P., Dhar, R., Oroszlan, S., andScolnick, E. M. Identification of a precursor in the biosynthe’sis of the P2 1

transforming protein of Harvey murine sarcoma virus. I. virOl., 42: 253-261, 1982.

25. Willingham, M. C., Pastan, I., Shih, T. Y., and Scolnick, E. M. Local-ization (If the src gene product of the Harvey strain of MS� IC) the plasmamembrane of transformed cells by electron microscopic immunocyto-

chemistry. Cell, 19: 1005-1014, 1980.

26. Powers, S., Michaelis, S., Broek, D., Santa Anna-A., S., Field, I..Herskowitz, I., and Wigler, M. RAM, a gene of yeast required for afunctional modification of RAS proteins and for production of matingpheromone a-factor. Cell, 47: 413-422, 1986.

27. Clarke, S., vogel, I. P., Deschenes, R. I.. and Stock, I. Posttranslationalmodification of the Ha-ras oncogene protein: evidence for a third class

of protein carboxyl methyltransferases. Proc. NaIl. Acad. Sci. USA, 85:

4643-4647, 1988.

28. Magee, T., and Hanley, M. Sticky fingers and CAAX boxes. Nature(Lond.(, 335: 114, 1988.

29. Hancock, I. F., Magee, A. I., Childs, I. E., and Marshall, C. I. All ram

proteins are polyisoprenylated but only some are palmitoylated. Cell, 57:1167-1177, 1989.

30. Schafer, W. R., Kim, R., Sterne, R., Thorner, I.. Kim, S-H., and Rine,

I- Genetic and pharmacological suppression of oncogenic mutations in

RAS genes of yeast and humans. Science (Washington DC). 245: 379-

385, 1989.

31. Casey, P. 1.. Solski, P. A., Der, C. I., and Buss, I. E. p2lras is modified

by the isoprenoid farnesol. Proc. NaIl. Acad. Sci. USA, 86: 8323-8327,1989.

32. Goldstein, I. L., and Brown, M. S. Regulation of the mevalonate

pathway. Nature (Lond.(, .343: 425-430, 1990.

33. Gutierrez, L., Magee, A. I., Marshall, C. I.. and Hancock, I. F. Post-

translational processing of p2lras is two.step and involves carboxyl-methylation and carboxyl-terminal proteolysis. EMBO I.. 8: 1093-1098,

1989.

34, Fujiyama, A., and Tamanoi, F. RAS2 protein of Saccharomyces cere-

visiae undergoes removal of methionine at N terminus and removal ofthree amino acids at C terminus. I. Biol. Chem., 265: 3362-3368, 1990.

35. Willumsen, B. M., Norris, K., Papageorge, A. G., Hubbert, N. L., and

Lowy, D. R. Harvey murine sarcoma virus p21 ras protein: biological andbiochemical significance of the cysteine nearest the carboxy terminus.EMBOJ., 3:2581-2585, 1984.

36. Iackson, I. H., Cochrane, C. G., Bourne, I. R., Solski, P. A., Buss, J.

E., and Der, C. I. Farnesol modification of Kirsten-ras exon 4B protein isessential for transformation. Proc. NaIl. Acad. Sci. USA, 87: 3042-3046,1990.

37, Kim, R., Rine, I.. and Kim, S-H. Prenylation of mammalian ras proteinin Xenopus oocyles. Mol. Cell. Biol., 10: 5945-5949, 1990.

38. Repko, E. M.. and Maltese, W. A. Post-translational isoprenylation ofcellular proteins is altered in response to mevalonate availability. I. Biol,

Chem., 264:9945-9952, 1989.

39, Volker, C., Lane, P., Kwee, C., lohnson, M., and Stock, I. A singleactivity carhoxyl methylales both farnesyl and geranylgeranyl cysteine

residues. FEBS. LeIt., 295: 189-194, 1991.

40. Perez-SaIa, D., Tan, E. W., Canada, F. I.. and Rando, R. R. Methylationand demethylation reactions of guanine nucleotide-hinding proteins ofretinal rod outer segments. Proc. NatI. Acad. Sci. USA, 88: 3043-3046,

1991.

41. Haklai, R., and Kloog, Y. Carboxyl methylation of 21-23 kDa mem-

brane proteins in intact neuroblastoma cells is increased with differentia-lion. FEBS Lett., 259: 233-236, 1990.

42. Backlund, P. 5., and Aksamit, R. R. Guanine nucleotide-dependent

carboxyl methylation of mammalian membrane proteins. I. Biol. Chem.,263: 15864-15867, 1988.

43. Hancock, I. F., Cadwallader, K., and Marshall, C. I. Methylation andproteolysis are essential for efficient membrane’ binding of prenylated

p21K��O. EMBO I.. 10: 641-646, 1991.

44. Kato, K., Cox, A. D., Hisaka, M. M., Graham, S. M., Buss, I. E., andDer, C. I. Isoprenoid addition to ram protein is the critical modificationfor its men)l)rane association and transforming activity. Proc. NaIl. Acad.

Sci. USA, 89: in press, 1992.

45. Hrycyna, C. A., Sapperstein, S. K., Clarke, S., and Michaelis, S. The

Saccharomyce’s cerevisiae STE14 gene encodes a methyltransferase thatmediates C-te’rminal methylation of a-ta tor and RAS proteins. EMBO I..10: 1699- 1709, 1991.

46. Ande’regg, R. I.. Betz, R., Carr, S. A., Crahh, I. W., ,lnd Duntze, W.Structure of Saccharomyces cerevisiae’ mating hormone a-factor. I. Biol.

Chem., 263: 18236-18240, 1988.

47. Marcus, S., Caldwell, C. A., Xue, C.B., Naider, F., and Be’cker, I. M.

Total in vitro maturation of the Saccharornyces cerevimi,ie a-factor lipopep-

tide’ mating Phe’romOne. Biochem. Biophys. Res. Commun., 172: 1310-

1316, 1990.

48. Marr, R. S., Blair, L. C., and Thorner, I. Saccharomyce’c cerevisiae5TL14 gene’ is required for C-terminal methylation of a-tactor mating

pheromone’. I. Biol. Chem., 265: 20057-20060, 1990.

49. Marcus, S., CaIdwell, G. A., Miller, D., xue, C-B., Naider, F., andBecker, I. M. Significance of C-terminal cysteine modifications to the

biological activity of the Saccharnmyce’s ere’visiae a-factor niating pher-omone. Mol. Cell. Biol., 1 1: 3603-3612, 1991.

50. Farnsworth, C. C., Wolda, S. L., GeIb, M. H., and Glomset, I. A.

Human lamin B contains a farnesylated cysteine residue. I. Biol. Che’m.,

264: 20422-20429, 1989.

51. Krohne, G., Waizenegger, I. an(l Hoger. T. H. The conserved car-boxy-terminal cysteine of nuclear lamins is essential for lamin association

with the nuclear envelope. I. Cell Biol., 109: 2003-201 1, 1989.

52. Kitten, G. T., and Nigg, E. A. The CaaX motif is require’d for isopren-ylalion, carhoxyl methylation, and nuclear membrane association of laminB2. I. Cell Biol., 1 13: 13-23, 1991.

53. Sepp-Lorenzino, L., Azrolan, N., and Coleman, P. S. Cellular distri-bution of cholesterogenesis-linked. phosphoisoprenylated proteins inproliferating cells. FEBS Lett., 245: 1 10- 1 16, 1989.

54. Epstein, W. W., Lever, D., Leining, L. M., Brue’nger. E., and Rilling,H. C. Quantitation of prenylcysteines by a selective cleavage reaction.Proc. NaIl. Acad. Sci. USA, 88: 9668-9670, 1991.

55. Hancock, I. F., Paterson, H., and Marshall, C. I. A polybasie domain

01’ palmitoylation is required in addition to the CAAX motif to localizep2l’�’ to the’ plasma membrane. Cell, 63: 133-139, 1990.

56. Hancock, I. F., Cadwallader, K., Paterson, H., and Marshall, C. I. A

468 Research Capsule

CAAX or a CAAL motif and a second signal are sufficient for plasma

membrane targeting of ras proteins. EMBO I.. 10: 4033-4039, 1991.

57. Magee, A. I., Gutierrez, L., McKay, I. A., Marshall, C. I.. and Hall, A.Dynamic fatty acylation of p21N��. EMBO I., 6: 3353-3357, 1987.

58. Chavrier, P., Gorvel, I-P., Steizer, E., Simons, K., Gruenberg, I.. andZerial, M. Hypervariable C-terminal domain of rab proteins acts as atargeting signal. Nature (Lond.(, 353: 769-772, 1991.

59. Lacal, P. M., Pennington, C. Y., and Lacal, l. C. Transforming activityof ras proteins translocated to the plasma membrane by a myristoylation

sequence from the src gene product. Oncogene, 2: 533-538, 1988.

60. Buss, I. E., Solski, P. A., Schaeffer, I. P., MacDonald, M. I.. and Der,C. I. Activation ofthe cellular proto-oncogene product p2 1 Ras by additionof a myristylation signal. Science (Washington DC). 243: 1600-1603,1989.

61 . Cox, A. D., Hisaka, M. M., Buss, I. E., and Der, C. I. Specific isoprenoidmodification is required br function of normal, but not oncogenic, Rasprotein. Mol. Cell. Biol., 12: 2606-2615, 1992.

62. Farnsworth, C. C., GeIb, M. H., and Glomset, I. A. Identification ofgeranylgeranyl-modified proteins in HeLa cells. Science (WashingtonDC). 247: 320-322, 1990.

63. Rilling, H. C., Breunger, E., Epstein, W. W., and Cram, P. F. Prenylatedproteins: the structure of the isoprenoid group. Science (Washington,

DC), 247: 318-320, 1990.

64. Marshall, M. S., Davis, L. I.. Keys, R. D., Mosser, S. D., Hill, W. S.,

Scolnick, E. M., and Gibbs, I. B. Identification of amino acid residuesrequired for ras p21 target activation. Mol. Cell. Biol., 1 1: 3997-4004,

1991.

65. Worman, H. I.. Yuan, I.. Blobel, G., and Georgatos, S. D. A lamin Breceptor in the nuclear envelope. Proc. NatI. Acad. Sci. USA, 85: 8531-

8534, 1988.

66. Resh, M. D., and Ling, H.P. Identification of a 32K plasma membraneprotein that binds to the myristylated amino-terminal sequence of p6Ov-src. Nature (Lond.), 346: 84-86, 1990.

67. Schmidt, R. A., Schneider, C. I., and Glomset, I. A. Evidence for post-translational incorporation of a product of mevalonic acid into Swiss 3T3cell proteins. J. Biol. Chem., 259: 10175-10180, 1984.

68. Balch, W. E. Low molecular GTP-binding proteins (LMGPs( involved

in vesicular transport: binary switches or biological transducers? TrendsBiochem. Sci., 15: 469-472, 1990.

69. Paterson, H. F., Self, A. l.� Garrett, M. D., lust, I., Aktories, K., and

Hall, A. Microinjection of recombinant p21’�’ induces rapid changes incell morphology. I. Cell Biol., 1 1 1: 1001-1007, 1990.

70. Knaus, U. G., Heyworth, P. G., Evans, T., Curnutte, I. T., and Bokoch,

G. M. Regulation of phagocyte oxygen radical production by the GTP-binding protein Rac 2. Science (Washington DC). 254: 1512-1515, 1991.

71. Abo, A., Pick, E., Hall, A., Totty, N., Teahan, C. G., and Segal, A. W.Activation of the NADPH oxidase involves the small GTP-binding protein

p21’�#{176}.Nature (Lond.), 353: 668-670, 1991.

72. Kitayama, H., Sugimolo, Y., Matsuzaki, T., Ikawa, Y., and Noda, M.A ras-related gene with transformation suppressor activity. Cell. 56: 77-

84, 1989.

73. Khosravi-Far, R., Lutz, R. J., Cox, A. D., Conroy, L., Bourne, J. R.,Sinensky, M., Balch, W. E., Buss, J. E., and Der, C. I. Isoprenoid modifi-

cation of rab proteins terminating in cc,cxc motifs. Proc. NaIl. Acad.Sci. USA, 88: 6264-6268, 1991.

74. Molenaar, C. M. T., Prange, R., and Gallwitz, D. A carboxyl-terminal

cysteine residue is required for palmitic acid binding and biologicalactivity of the ras-related yeast YPT1 protein. EMBO I., 7: 971-976, 1988.

75. Walworth, N. C., Goud, B., Kabcenell, A. K., and Novick, P. 1.Mutational analysis of SEC4 suggests a cyclical mechanism for the regu-

lation of vesicular traffic. EMBO J., 8: 1685-1693, 1989.

76. Bourne, H. R. Do GTPases direct membrane traffic in secretion? Cell,

53:669-671, 1988.

77. Fukada, Y., Takao, T., Ohguro, H., Yoshizawa, T., Akino, T., and

Shimonishi, Y. Farnesylated �y subunit of photoreceptor G protein indis-pensible for GTP-binding. Nature (Lond.), 346: 658-660, 1990.

78. Finegold, A. A., Schafer, W. R., Rine, J., Whiteway, M., and Tamanoi,F. Common modifications of trimeric G proteins and as protein: involve-ment of polyisoprenylation. Science (Washington DC). 249: 165-168,

1990.

79. Inglese, J., Clickman, J. F., Lorenz, W., Caron, M. G., and Lefkowilz,R. I. Isoprenylation of a protein kinase. J. Biol. Chem., 267: 1422-1425,1992.

80. Anant, I. S., Ong, 0. C., xie, H., Clarke, S., O’Brien, P. J., and Fung,B. K-K. In vivo differential prenylation of retinal cyclic GMP phosphodi-

esterase catalytic subunits. J. Biol. Chem., 267: 687-690, 1992.

81. Bokoch, G. M., and Prossnitz, v. Isoprenoid metabolism is requiredfor stimulation of the respiratory burst oxidase of HL-60 cells. I. Clin.lnvest.,89:402-408, 1992.

82. Law, R. E., Stimmel, I. B., Damore, M. A., Carter, C., Clarke, S., andWall, R. Lipopolysaccharide-induced NF-KB activation in mouse 70Z/3pre-B lymphocytes is inhibited by mevinolin and S’methylthioadenosine:roles of protein prenylation and carboxyl methylation reactions. Mol.

Cell. Biol., 12: 103-111, 1992.

83. Quesney-Huneeus, V., Wiley, M. H., and Siperstein, M. D. Essential

role for mevalonate synthesis in DNA replication. Proc. NatI. Acad. Sci.

USA, 76: 5056-5060, 1979.

84. Habenicht, A. I. R., Glomset, I. A., and Ross, R. Relation of cholesteroland mevalonic acid to the cell cycle in smooth muscle and Swiss 3T3cells stimulated to divide by platelet-derived growth factor. I. Biol. Chem.,

255: 5134-5140, 1980.

85. Fairbanks, K. P., Witte, L. D., and Goodman, D. S. Relationship

between mevalonale and mitogenesis in human fibrclblasts stimulatedwith platelet.derived growth factor. I. Biol. Chem., 259: 1 546- 1 551 , 1984.

86. Fairbanks, K. P., Barbu, v. D., Witte, L. D., Weinstein, I. B., and

Goodman, D. S. Effects of mevinolin and mevalonate on cell growth inseveral transformed cell lines. I. Cell, Physiol., 127: 216-222, 1986.

87. Maltese, W. A., Defendini, R., Green, R. A., Sheridan, K. M., andDonley, D. K. Suppression of murine neuroblastoma growth in vivo bymevinolin, a competitive inhibitor of 3-hydroxy-3-methylglutaryl-coen-zyme A reductase. I. Cell Biol., 76: 1748-1754, 1985.

88. DeClue, I. E., \Jass, W. C. Papageorge, A. G., Lowy, D. R., and

Willumsen, B. M. Inhibition of cell growth by lovastatin is independent

ofras function. Cancer Res., 51: 712-717, 1991.

89. Sinensky, M., Beck, L. A., Leonard, S., and Evans, R. Differentialinhibitory effects of lovastatin on protein isoprenylation and sterol syn-

thesis. I. Biol. Chem., 265: 19937-19941, 1990.

90. Reiss, Y., Goldstein, I. L., Seabra, M. C., Casey, P. 1.. and Brown, M.S. Inhibition of purified p21’�’ farnesyl:protein transferase by Cys-AAxtetrapeplides. Cell, 62: 81-88, 1990.

91. Schaber, M. D., O’Hara, M. B., Garsky, v. M., Mosser, S. D., Bergs-

lrom, I, D., Moores, S. L., Marshall, M. S., Friedman, P. A., Dixon, R. A.

F., and Gibbs, I. B. Polyisoprenylation of ras in vitro by a farnesyl-proteintransferase. I. Biol. Chem., 265: 14701-14704, 1990.

92. Manne, v., Roberts, D., Tobin, A., O’Rourke, E., De Virgilio, M.,Meyers, C., Ahmed, N., Kurz, B., Resh, M., Kung, H-F., and Barbacid, M.

Identification and preliminary characterization of prolein-cysteine fame-syltransferase. Proc. NatI. Acad. Sci. USA, 87: 7541-7545, 1990.

93. Seabra, M. C., Reiss, Y., Casey, P. J., Brown, M. S., and Goldstein, I.L. Protein farnesyltransferase and geranylgeranyltransferase share a com-mon cm subunit. Cell, 65: 429-434, 1991.

94, xue, C-B., Ewenson, A., Becker, I. M., and Naider, F. Solution phase

synthesis of Saccharomyces cerevisiae a-mating factor and its analogs. Int.

I. Pept. Protein Res., 36: 362-373, 1990.

95. Moores, 5, L., Schaber, M. D., Mosser, S. D., Rands, E., O’Hara, M.

B., Garsky, v. M., Marshall, M. S., Pompliano, D. L., and Gibbs, J. B.Sequence dependence of protein isoprenylation. I. Biol. Chem., 266:14603-14610, 1991.

96. Kinsella, B. T., Erdman, R. A., and Maltese, W. A. Postlranslationalmodification of Ha-ram p21 famnesyl versus geranylgeranyl isoprenoids isdetermined by the COOH-terminal amino acid. Proc. NaIl. Acad. Sci.

USA, 88: 8934-8938, 1991.

97. Horiuchi, H., Kawata, M., Katayama, M., Yoshida, Y., Musha, T.,Ando, S., and Takai, Y. A novel prenyltransferase for a small GTP-binding

protein having a C-terminal Cys-Ala-Cys structure. J. Biol. Chem., 266:16981-16984, 1991.

98. Kinsella, B. T., and Maltese, W. A. rab GTP-binding proteins withthree different carboxyl-terminal cysteine motifs are modified in vivo by20-carbon isoprenoids. I. Biol. Chem., 267: 3940-3945, 1992.

99. Furth, M. E., Aldrich, T. H., and Cordon-Cardo, C. Expression of rasproto-oncogene proteins in normal human tissues. Oncogene, 1: 47-58,

1987.

100. Fiorucchi, G., and Hall, A. All three human ras genes are expressedin a wide range of human tissues. Biochem. Biophys. Ada., 950: 81-83,1988.

101. Mulcahy, L. S., Smith, M. R., and Stacey, D. W. Requirement for ras

proto-oncogene function during serum-stimulated growth of NIH 3T3cells. Nature (Lond.), 313: 241-243, 1985.

102. Ashby, M. N., King, D. S., and Rine, I. Endoproteolytic processing

of a famnesylated peptide in vitro. Proc. NaIl. Acad. Sci. USA, 89: 4613-

4617, 1992.

103. Tan, E. W., Perez-Sala, D., Canada, F. J., and Rando, R. R. Identifying

Cell Growth & Differentiation 469

the recognition unit for G protein methylation. I. Biol. Chem., 266: 10719-

10722, 1991.

104. volker, C., Miller, R. A., McCIe’ary, W. R., Rao, A., Pclenie, M.,

Backer, I. M., and Stock, I. B. Effects of farnesylcysteine analogs onprotein carboxyl methylation and signal transduction, I. Biol, Chem., 266:

21515-21522, 1991.

105. Stephenson, R. C.. and Clarke’. S. lde’ntification of a C-terminal

protein carboxyl methyltransferase in rat liver membranes utilizing a

synthetic famnesyl cysteine-containing peptide substrate. I. Biol. Chem.,265: 16248-16254, 1990.

106. Gutierrez, L., and Magee, A. I. Characterization ofan acyltransferaseacting on p2 1 N-ras protein in a cell-free system. Biochim. Biophys. Ada,1078: 147-154, 1991.