Mechanisms of Negative Regulation of Protein Prenylation: Investigation of Murine Guanylate-Binding Protein 1 " (2000)

Iowa State University Capstones, Theses and Retrospective Theses and Dissertations Dissertations

2000 Mechanisms of negative regulation of protein prenylation: investigation of murine guanylate- binding protein 1 John Thomas Stickney Iowa State University

Follow this and additional works at: https://lib.dr.iastate.edu/rtd Part of the Biochemistry Commons, Cell Biology Commons, and the Molecular Biology Commons

Recommended Citation Stickney, John Thomas, "Mechanisms of negative regulation of protein prenylation: investigation of murine guanylate-binding protein 1 " (2000). Retrospective Theses and Dissertations. 13934. https://lib.dr.iastate.edu/rtd/13934

This Dissertation is brought to you for free and open access by the Iowa State University Capstones, Theses and Dissertations at Iowa State University Digital Repository. It has been accepted for inclusion in Retrospective Theses and Dissertations by an authorized administrator of Iowa State University Digital Repository. For more information, please contact [email protected]. INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author dkl not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletkNi.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overiaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6' x 9" t)lack and white photographic prints are availat)le for any photographs or illustratkxis appearing in this copy for an additional charge. Contact UMI directiy to order.

Bell & Howell Information and Learning 300 North Zeeb Road. Ann Arbor, Ml 48106-1346 USA 800-521-0600

Mechanisms of negative regulation of protein prenylation: Investigation of murine guanylate-binding protein 1

John Thomas Stickney

A dissertation submitted to the graduate faculty

in partial fulfillment of the requirements for the degree of

DCXn-OR OF PfflLOSOPHY

Major: Molecular, Cellular, and Developmental Biology

Major Professor; Janice E. Buss

Iowa State University

Ames, Iowa 2000 UMI Number: 9962850

UMI Microform9962850 Copyright 2000 by Bell & Howell Information and Leaming Company. All rights reserved. This microform edition is protected against unauthorized copying under Title 17, United States Code.

Bell & Howell Information and Leaming Company 300 North Zeeb Road P.O. 80x1346 Ann Arbor, Ml 48106-1346 ii

Graduate College Iowa State University

This is to certify that the Doctoral dissertation of

John Thomas Stickney has met the dissertation requirements of Iowa State University

Signature was redacted for privacy.

Major Professor

Signature was redacted for privacy. or the Major Program

Signature was redacted for privacy. For the Graduate College iii

In Memory

This woric is dedicated to the memory of my father Thomas Allen Stickney who, with my mother Joan, always stressed the importance of education. iv

TABLE OF CONTENTS

CHAPTER 1. GENERAL INTRODUCTION 1 Introduction 1 Dissertation Organization 1 Literature Review 2 References 17

CHAPTER 2. MURINE GBP I: INCOMPLETE C20ISOPRENOID 35 MODinCATION OF AN IFNy-INDUCIBLE GTP-BINDING PROTEIN Abstract 35 Introduction 36 Materials and Methods 38 Results 40 Discussion 47 Acknowledgments 51 References 52

CHAPTER 3. INHIBITION OF C20 MODIHCATION OF MGBPl'S CAAX 67 MOTIF IS MEDIATED BY MGBPl C- AND N-TERMINAL PROTEIN ELEMENTS Absu-act 67 Introduction 68 Materials and Methods 71 Results 74 Discussion 82 Acknowledgments 86 References 86

CHAPTER 4. GENERAL CONCLUSIONS 98 Overview 98 Regulation of Prenylation? 99 Biological Significance 100 References 101

ACKNOWLEDGMENTS 102 I

CHAPTER 1. GENERAL INTRODUCTION

Introduction

Since 1989, the study of isoprenoid-modified (prenylated) proteins has emerged as an extremely active area of research due to the great potential for novel therapeutic applications, as many prenylated proteins play prominent roles in cancer biology, viral and microbial infections, and other disease states. The numbers and types of proteins that are modified on their C-termini with 15 or 20 carbon isoprenoids continues to expand, and as the list of prenylated proteins grows, so does the potential for pharmacological intervention. Thus, it is imperative that research continues into both the basic mechanisms governing protein prenylation, as well as the subsequent biological consequences of this modification.

The work presented in this dissertation describes a novel aspect of prenylation - one in which a protein (mGBPl) fails to be prenylated. The mechanism behind this lack of prenylation appears to involve contributions from both N- and C-terminal regions of mGBPl, and as such represents a previously unrecognized means of cellular regulation of prenylation.

A complete understanding of this unique mechanism may lead to the development of new compounds for the treatment of cancer and other diseases.

Dissertation Organization

This first chapter of the dissertation will detail the current status of what is known regarding the mechanisms and processing events associated with protein prenylation as well as the biological effects this class of modiHcation can have on the various proteins to which it is attached. Additionally, the potential benefits and shortcomings of preventing cellular prenylation in the treatment of diseases will also be discussed as will our model system for studying novel aspects of prenylation (Guanylate Binding Proteins).

Chapter 2, which has been submitted for publication in the journal Molecular Biology of the Cell, is the initial characterization of the prenylation of murine Guanylate Binding

Protein-1 (mGBPl). The main fmding of this paper was the surprising defect in isoprenoid 2 attachment to mGBPl in vivo, even though this protein possesses a suitable CaaX motif at its

C-terminus. Our initial mutational analysis of mGBPl showed that amino acids within its C- terminus play a role in preventing efficient modification by geranylgeranyl transferase I

(GGTase I).

The third chapter, to be submitted to the Journal of Biological Chemistry, extends the initial findings of Chapter 2 by demonsu-ating that the observed difficulty in the prenylation of mGBPl seen in cells also occurs in in vitro conditions and therefore does not result from difficulties in cellular metabolism of isoprenoids or prenyl transferase inaccessibility, but instead, must arise from an intrinsic property of mGBPl itself. We also show in Chapter 3 that the C-terminus of mGBPl is an acceptable substrate for prenylation when it is removed from the context of the native protein, thus implicating the body of mGBPl as an inhibitory regulator of prenylation of the C-terminus. Removal of various internal domains of mGBPl revealed that the N-terminal region possessing the GTP-binding domain strongly participates in blocking geranylgeranylation of the C-terminal CaaX motif. This highly unexpected finding demonstrates that mGBPl represents a new example in the field of protein prenylation whereby a GTP-binding domain can negatively influence CaaX motif utilization.

Chapter 4 will present the general conclusions drawn from this work as it applies to the field of protein prenylation.

Literature Review

INTRODUCTION TO PROTEIN PRENYLATION

Naturally occurring covalent modification of peptides by isoprenoids was first described over twenty years ago for a mating factor from the fungus Rhodosporidium toruloides (61). However, because researchers were unaware of cellular signal sequences for prenylation, similar modification of eukaryotic proteins remained largely ignored for over a decade. Not until the Ras family of proteins was shown to be prenylated (48) with famesyl 3 isoprenoids (17) did the scientific community turn their full attention to this novel posttranslational protein modiHcation.

Famesyl (15 carbons) and geranylgeranyl (20 carbons) isoprenoids are found in C- terminal thioether linkages with cysteine residues in approximately 2% of all cellular proteins

(34) (for reviews of prenylation, see (16,42, 83, 138)). The enzymes that attach these lipids are grouped into two classes. The first, diose of the CaaX prenyl transferases, catalyze the prenylation of proteins possessing a CaaX motif as their final four amino acids. These enzymes, and their potential substrates, will form the focus of the review and also the subsequent data chapters. The second class of enzyme, Rab geranylgeranyl transferase

(GGTase 11), geranylgeranylates the Rab family of GTPases and will not be discussed.

The CadX motif refers to the final four amino acids of a protein where C stands for the cysteine to be prenylated, followed by two generally aliphatic (a) amino acids and the final amino acid of the protein, X. The existence of this motif was first hypothesized as researchers began to compare the C-termini of known prenylated proteins (38). However, confirmation of the "CaaX hypothesis" as a cellular signal for prenylation arrived after the first prenyl transferase was purified and shown to be active on peptides of only 4 amino acids in length

(102). Further work showed that although the nature of the a residues can influence the efficiency of prenylation and additional CaaX processing (62, 87), specificity for which isoprenoid will be attached resides in the final amino acid X (67, 87, 104). This is achieved through specific recognition by either famesyl transferase (FTase) or geranylgeranyl transferase I (GGTase I). FTase has been shown to prefer CaaX proteins terminating most commonly in Met, Ser, Ala, or Gin whereas GGTase I preferentially recognizes and prenylates CaaX motifs ending with Leu or Phe. In both cases, efficiency of FTase and

GGTase I-mediated modification is assumed to be near 100% based on mass-spectrometric analysis of CooX-containing proteins (37,78,96). 4

FTase and GGTase I are distinct but related homodimeric enzymes of a and P subunits with the a subunit being conmion to both enzymes (138). Due to the shared subunit, it was hypothesized early on that the p subunit would provide for subsu-ate specificity (103). Indeed, biochemical characterization has shown that the P subunits are primarily responsible for the selection of the appropriate lipid and CaaX substrates (27,73,119). However, mutational and photolabeling experiments have also demonstrated a clear role for the a subunit in catalysis (4,

98, 135), prompting researchers to speculate that both subunits participate in catalyzing the prenylation reaction. These speculations were conHrmed after the crystal structure of the FTase holoenzyme was solved (97). This structure showed the existence of an active site cleft at the o/p subunit interface. Subsequent co-crystal structures of FTase with lipid, or lipid and peptide substrate, bound (79, 117) showed that a hydrophobic pocket within the P subunit is responsible for binding the appropriate isoprenoid whereas protein substrate binding is accomplished by residues from both subunits.

The presence of an isoprenoid-binding pocket has lead to the "molecular ruler" hypothesis for lipid specificity (97). The depth of the isoprenoid-binding pocket of FTase is such that only famesyl pyrophosphate is the correct length for the CI isoprenoid carbon to be in register with the CaaX cysteine during catalysis. This information gives a mechanistic explanation for previous results showing that FTase could bind famesyl and geranylgeranyl pyrophosphate with similar efficiencies, but only famesyl could be utilized for the prenylation reaction (101). The mechanism by which GGTasel preferentially utilizes geranylgeranyl pyrophosphate is currently unknown, but it is clear that GGTase I does have a distinct preference for geranylgeranyl, as the C20 isoprenoid binds 300 times tighter than famesyl

(137).

A structural explanation as to how FTase and GGTase I selectively identify CaaX- containing proteins, and the subsequent discrimination of CIS vs. C20 substrates based on the

CaaX 'X, also remains a mystery. However, it is clear that the choice of an appropriate CaaX 5 substrate by a prenyl transferase is dictated predominantly by the p subunit even though, as previously mentioned, both the a and P subunits of FTase and GGTase I contribute to the binding of protein substrates. Mutational studies of the P subunit of FTase show that single point mutations can be made that allow for binding and famesylation of proteins with GGTase- specific CaaX motifs (27). It is believed that steric factors contribute to this change of CaaX acceptance, as the size of the substituted amino acid within FTase appeared to be the main factor influencing the ability to famesylate a GGTase-specific CaaX. Similar changes in substrate specificity (i.e. famesylation of a C20-type CaaX) are seen when the active site zinc atom of wild type FTase is replaced with a cadmium atom (139). In this example, changing the type of metal ion may also change steric factors within, or perhaps change the local geometry of, the active site. Even though the aforementioned FTase P subunits retained preference for the famesyl isoprenoid, other studies have suggested that the type of isoprenoid (CIS vs. C20) bound to the prenyl transferase may also play a role in CaaX motif selection in the wild type prenyl transferases (15). Therefore, until further biochemical and structural work is performed, the mechanisms governing choice of CaaX substrates by FTase and GGTase I will not be known in complete detail.

ENZYMOLOGY OF PROTEIN PRENYLATION

The reaction pathways governing the binding of substrates for both FTase and

GGTase I have been investigated at great length (29, 40, 116, 141). For both enzymes, substrate binding follows an ordered pathway, with isoprenoid association preceding the binding of CaaX substrate. Release of prenylated protein is the rate-limiting step in the reaction for both enzymes (40, 137). This may have biological significance, as it has been shown that for FTase, famesyl pyrophosphate binding is required for the release of prenyl product (120).

These authors have hypothesized that release of prenylated protein from prenyl transferases (at least for FTase) could be a regulated event based on the availability and cellular location of isoprenoid pyrophosphates. 6

Although there is good agreement as to the reaction order for both FTase and

GGTase I, the precise mechanism behind formation of the thioether linkage between isoprenoid and CaaX cysteine remains a topic of debate. Both electrophilic (18, 30) and nucleophilic (39, 52,54) mechanisms have been suggested. Although a consensus has not yet been reached, the evidence supports a model in which perhaps both electrophilic activation of the isoprenoid (through a carbocation intermediate) as well as nucleophilic activation of the

CaaX cysteine (through zinc-mediated thiolate formation) play a role in formation of the thioether bond. PROCK.SSING OF PRENYLATED PROTEINS

Once a CaaX protein has been prenylated, subsequent enzymatic processing of the C- terminus then occurs (36). Proteolysis of the final three amino acids {aaX), followed by carboxymethylation of the newly exposed prenylcysteine C-terminus, is believed to occur immediately following prenylation of all CaaX proteins (for exceptions, see (51, 108)). The protease and carboxymethyltransferase responsible for catalyzing these reactions are both localized in vivo to the endoplasmic reticulum (26, 95). Because prenyl transferases are soluble enzymes, newly prenylated proteins must be shuttled to the endoplasmic reticulum for further enzymatic processing. This has lead to the speculation that perhaps the prenyl transferases themselves may actually "chaperone" their products to the endoplasmic reticulum

(120).

For the vast majority of mammalian proteins, the functional significance of these processing steps is not fully known. However, transgenic mice deficient in the CaaX protease usually die in late gestation or very soon post-partum (66). Fibroblasts from these mice demonstrate that the presence of this enzyme is necessary for proper association of Ras with the plasma membrane. Thus, the processing of the C-termini of newly prenylated proteins may play a very important role in the localization/biological activity of this class of lipid-modiHed 7 protein. A more comprehensive discussion of the biological significance of protein prenylation will follow in the next section.

Because the thioether linkage between isoprenoid and cysteine residue is not labile, cells must possess a mechanism to dispose of the prenylcysteine that accumulates after the natural degradation of prenylated proteins. Famesyl cysteine has been shown to interact with the P-glycoprotein (Pgp) transporter (142, 143), and ATP-dependent export via Pgp may be a possible means of prenylcysteine clearance for cells expressing the transporter. However, most cells also express an enzyme capable of speciHcally removing isoprenoids from cysteine amino acids. This enzyme, prenylcysteine lyase, is likely to be the final processing step in the life cycle for the majority of an animal's prenylated proteins (121,144).

CONSEQUENCES OF PROTEIN PRENYLATION

When discussing the ultimate biological role covalently attached isoprenoids play in protein function, many researchers commonly assume (quite erroneously) that prenylation acts simply as a tether for membrane binding. However, up to one half of the total cellular prenylated protein is cytosolic (34); obviously, prenylation confers other properties to proteins in addition to membrane association. This section will therefore focus on the known effects prenylation has on protein localization, trafficking, and function by utilizing examples in which prenylation performs different functions depending on the protein to which the isoprenoid is attached.

The most studied family of prenylated proteins is that of the Ras family (80). This group of low molecular weight GTPases functions as signal transduction conduits by coordinating signaling events from numerous growth factor pathways to a constantly growing list of downstream effectors (for a recent review of Ras signaling, see (125)). Ras proteins toggle between GDP-bound inactive, and GTP-bound active, states normally. However, transforming mutations can arise which lock Ras in its active GTP-bound state, thus activating downstream effectors even in the absence of upstream signal. Accordingly, Ras proteins 8 possess tremendous oncogenic potential; 30% of all human tumors harbor an activated Ras allele (7). Therefore, a great deal of effort has gone into understanding Ras biology in an attempt to design specific inhibitors that can block Ras signaling in human cancers.

There are 4 major cellular forms of Ras: H-, N-, and K- (two alternatively spliced forms termed K4A and K4B). They are highly identical at the amino acid level, with especially high sequence conservation in their N-terminal effector binding regions. All possess CaaX motifs that signal for famesylation, and all but the K4B form are additionally palmitoylated on

C-terminal cysteine residues (48). Instead of additional lipidation, K4B possesses a polybasic region in its C-terminus; the presence of a 'second signal' of palmitate or basic residues, in conjunction with the famesyl group, is necessary for sustained plasma membrane interaction of all Ras proteins. Stable membrane association is required for Ras activity, as mutants of Ras that cannot interact with membranes are non-transforming (130).

The initial membrane binding event for Ras proteins is believed to be facilitated by the famesyl group, as mutants lacking isoprenoid are incapable of membrane association and transformation (55, 62). This discovery led to the search for compounds that could prevent

Ras famesylation, and subsequently, its membrane association and oncogenic potential. To this end, potent, selective inhibitors of FTase have been developed; a discussion of their cellular effects will be presented later.

Studies demonstrating the importance of famesyl in Ras membrane association have prompted questions regarding the mechanism of isoprenoid-mediated membrane association.

Peptide studies clearly show that isoprenoids are capable of relatively weak hydrophobic interactions with membrane lipids, but require the presence of basic amino acids or secondary lipid modifications to achieve stable membrane interaction (76, 109, 113). Due to the weak interaction with lipids, some researchers have hypothesized that perhaps the isoprenoid on Ras can mediate protein-protein interactions with a putative 'Ras receptor' at the plasma membrane

(114). However, the famesyl group is not required for Ras activity if the protein is artificially 9 targeted to the plasma membrane through the use of an alternate membrane anchor (11,14) or alternative lipids on the CaaX cysteine (6, 32). These studies suggest that plasma membrane association alone is sufficient for Ras activity. However, other work has shown the isoprenoid of Ras to play an important role in mediating events other than membrane association.

In addition to initiating plasma membrane interaction, the famesyl group on Ras has been shown to participate in a number of other cellular events. Activation of Raf kinase, and other effector proteins, by Ras may be mediated in part by famesyl interaction with these proteins (84, 106). The C-terminus of Ras also plays a role in the trafficking and targeting of processed (proteolyzed and carboxymethylated) Ras from the endoplasmic reticulum to the plasma membrane (23, 131). Additionally, once at the plasma membrane, the combination of palmitates and isoprenoid on H-, N-, and K4A Ras may control the interaction of these proteins with specialized signaling subdomains of the plasma membrane such as detergent- resistant membranes (DRMs) orcaveolae (33, 46, 82, 85, 86,92, 107).

Although the role of prenylation in Ras biology has received a tremendous amount of attention, the physiological consequences of prenoid addition to a number of other proteins has also been investigated. Famesyl modification of the nuclear protein lamin A has proven to be an interesting example where prenylation does NOT confer plasma membrane localization.

Instead, prenylation of lamin A is required for proper assembly into the nuclear lamina (S3).

However, it is not the prenylated form of lamin A that is integrated into the nuclear envelope.

Lamin A is substrate for a nuclear-localized preiamin A endoprotease that specifically recognizes the prenylated form of lamin A (preiamin A) and cleaves its Hnal 18 amino acids to generate mature lamin A (64,65). It is this form of the protein that is integrated into the lamina

(81). Recent work has shown that the isoprenoid of prenylated, but not proteolyzed, preiamin

A is also required for interaction with a newly discovered nuclear protein termed Narf (8).

Therefore, in addition to directing plasma membrane localization, prenylation can also direct the 10 localization of proteins to various cellular locations (i.e. nuclear lamina) via isoprenoid- mediated protein-protein interactions (i.e. recognition by prelamin A endoprotease).

In addition to merely directing a protein's localization, work in the Held of G-protein coupled receptors (GPCR) and G-protein signaling has shown that isoprenoid addition can also directly influence the specificity and efficiency of signal transduction events. The prostacyclin receptor is the first GPCR shown to be prenylated (50). Preventing famesylation of this receptor decreases adenylate cyclase activation in response to ligand without influencing the ability to bind ligand. This suggests that the presence of famesyl on the C-terminus of this receptor is necessary for efficient activation of Gas. The type of isoprenoid attached to certain

Gy subunits has also been shown to influence their ability to interact both with their receptors and also downstream effectors. For the yl subunit, famesyl is more efficient at promoting interaction with the rhodopsin receptor than is geranylgeranyi (68). Conversely, coupling of GPy subunits to the A1 adenosine receptor is favored by geranylgeranylation rather than famesylation of y subunits (134). The ability to activate the downstream effector phospholipase C-P has also been shown to be affected by the type of isoprenoid attached to the

Gy subunit (88). Thus, prenylation not only can promote protein-protein interactions but may also be able to modulate the specificity and strengths of those interactions as well.

Therefore, as our knowledge of the identities of prenylated proteins expands, so will the list of known functions isoprenoids play in protein activity. This information will be extremely important in the analysis of the effects of preventing cellular protein prenylation in treating human disease. The use and effects of prenyl transferase inhibitors will be discussed next.

USE OF PRENYL TRANSFERASE INHIBITORS IN THE TREATMENT OF DISEASE

As mentioned earlier, preventing prenylation of oncogenic proteins such as Ras holds the promise as a new area of cancer chemotherapy (for reviews on prenyl transferase inhibitors, see (13, 69, 93)). However, prenylated proteins play roles in many other human 11 diseases. Many non-cancerous pathologies that result from hypertrophic cellular growth may be responsive to prenyl Uransferase inhibitors. In fact, blocking cellular prenylation can reduce the progression of atherosclerosis (35), promote smooth muscle cell apoptosis (115) which may help in the treatment of hypertension and other vascular diseases, and can slow the growth of hypertrophic mesangial cells, which are associated with many renal disorders (9). In addition to the ever-growing list of naturally arising human disorders, prenylated proteins also may contribute the to the pathologies associated with viral (43,94) or microbial infection (49,

136). Thus, the use of inhibitors of prenylation to treat disease may extend far beyond the scope of cancer biology.

Competitive inhibitors of FTase or GGTase I can be grouped into two classes: compounds that prevent isoprenoid binding and compounds that block CaaX recognition.

Although competitive inhibitors of isoprenoid binding exist (126, 133), nearly all of the inhibitors of FTase and GGTase I that are cunently being developed are designed to prevent protein substrate recognition. Through rational drug design based on the structure of the CaaX motif, and screening of drug libraries, numerous compounds have been identified that are potent and selective inhibitors of the prenyl transferases. Because of the interest in selectively preventing Ras famesylation, interest in the creation and discovery of FTase inhibitors (FTIs) has far outweighed that for GGTase I inhibitors (GGTIs). Therefore, although GGTIs have been shown to possess anti-tumor therapeutic potential (2, 118), the focus of this section will be on inhibitors of famesyl transferase.

Because peptides as short as 4 amino acids can act as competitive inhibitors of FTase

(102), rational drug design of small organic compounds modeled after the CaaX motif has yielded a number of promising inhibitors of FTase. Drug library screening has also yielded non-peptide inhibitors of protein binding to FTase. All of these drugs show great in vitro inhibition of FTase, with IC50 values in the low nanomolar range (10,24,58,70). In addition to being excellent in vitro inhibitors of FTase, these compounds are also biologically active 12 with respect to regressing Ras-mediated transformation both in cell culture (10,24,58,70) and in animal models (71). These drugs also show good activity in vivo against H-, N-, and K-

Ras tumor explants in nude mice (72). That these compounds appear efficacious in Ras-driven transformation models is quite interesting in light of recent work regarding their mode of action.

Initial studies of FTI activity naturally centered on testing their in vivo ability to abrogate Ras signaling potential by preventing Ras prenylation. It was hypothesized that if prenylation of activated Ras proteins was prevented, then their ability to activate downstream effectors may be hindered, perhaps by the accumulation of cytosolic (and hence inactive) signaling complexes. Although early work showed that this did occur (56,75), the effective concentration of FTI that was necessary to inhibit K-Ras signaling was 100 times greater than that for H-Ras (57,75). This presents a signiHcant problem in the use of FTIs in the treatment of cancers, as the K-Ras oncogene is by far the most commonly activated form of Ras in human cancers (7). An explanation for this large difference in the effective concentrations of

FTIs came with the discovery that the K- and N-Ras isoforms are capable of being alternatively

C20-modified by GGTase I both in vitro (140) as well as in vivo in the presence of FTIs (105,

129). Therefore, because of the potential for continued prenylation and membrane association of K-Ras in FTI-trealed cancer cells, the use of FTIs in the treatment of human tumors was questioned.

Good news for researchers working on famesyl transferase inhibitors arrived with the knowledge that many tumor lines appear to be responsive to FTIs regardless of the activation status of their Ras proteins (112). Thus, it appears that prenylated proteins in addition to Ras may play a role in promoting cellular transformation. One non-Ras candidate prenyl protein that may be involved in mediating the cellular effects of FTIs is RhoB. RhoB is a C20-type

CaoX-containing protein of the Ras superfamily that deviates from the CaaX hypothesis, as it naturally exists in famesylated and geranylgeranylated forms (1). RhoB is a substrate for both 13

FTase and GGTase I in vitro, and is exclusively geranylgeranylated in cells treated with FTIs

(74). The C20-modified form has been hypothesized to promote the growth-inhibiting effects of FTIs, as co-expression of a C20-specific mutant of RhoB with activated Ras mimics the phenotype of FTI-treatment of cells expressing activated Ras (31). Thus the overall mechanism behind FTI activity may be the result of a decrement in Ras protein prenylation with accompanied loss or alternative prenylation of other CaaX proteins. However, continued work in this field is required before a complete understanding of FTI activity is reached.

NOVEL ASPECTS OF CELLULAR PRENYLATION

Although FTIs are in clinical trials for use in human cancers (93), the aforementioned studies clearly illustrate that prenyl transferases and their substrates (such as K-Ras and RhoB) do not always follow the CaaX hypothesis in vivo. Thus, more information is needed regarding the in vivo specificity and fidelity of prenyl transferases, as new classes of FTIs and

GGTIs are being developed and used to prevent cellular prenylation.

Presently, very little data exists to suggest that the enzymatic activity of prenyl transferases can be regulated. There are some reports that FTase activity may be influenced by insulin (44,45), but the effects seen in these reports appear minimal at best. The a subunit of

FTase and GGTase I has been shown to associate with, and become phosphorylated by, the

TGF-P receptor (63, 122, 128). However, the functional significance of this association and phosphorylation remains unknown.

As mentioned with K- and N-Ras and RhoB, some CaaX proteins can be modified by the "wrong" prenyl transferase (i.e. the transferase NOT dictated by the CaaX motif). This phenomenon, termed cross-prenylation, demonstrates that N-terminal, non-CaaX regions of substrate proteins can have an effect on which type of prenyl transferase will modify the C- terminus. This effect of N-terminal substrate protein elements on the fidelity of CaaX prenylation was also shown in a study using chimeric Gysubunits (60). Proteins possessing a

C15-^pe CaaX motif could be cross-prenylated by GGTase I, but this cross-recognition could 14 be modulated based on the identity of N-terminal sequences. Clearly, prenyl transferases are influenced by substrate structures in addition to those encoded by the CaaX motif.

In addition to cross-prenylation, limited examples of incomplete prenylation also exist.

Primary cardiac myocytes (41) and 3T3-L1 adipocytes (44,45) possess considerable levels of non-prenylated, cytoplasmic Ras. This is quite unusual, because as mentioned previously, virtually all CaaJf-containing proteins, including Ras, are believed to be stoichiometrically prenylated (25, 78). One well known exception to this rule is that of the heterotrimeric Gai proteins. Gai subunits, terminating in 'CGLF CaaX motifs, should be targets for modification by GGTase I. However, these proteins are not prenylated in vivo, but instead remain free for ADP-ribosylation of the CaaX cysteine (59). Interestingly, if the CaaX of Gtti is mutated to resemble the FTase-type CaaX found on H-Ras, prenylation can occur. Thus, the lack of prenylation of Gai appears to be specific for GGTase I, and appears to be due in part to some non-CaaX structural element of Gai, the CGLF CaaX is at least partially prenylated when placed onto H-Ras (25).

Thus, GGTase I's ability to cross-prenylate proteins such as K-Ras, N-Ras, and

RhoB, along with its inability to recognize some of its own potential substrates (Gai), suggest that this prenyl transferase may have unique substrate requirements. Because it shares a subunit with FTase, and because the known binding site for CaaX substrates resides at the subunit interface, this poses a series of interesting questions as to how GGTase I recognizes and prenylates its substrates in vivo. We have attempted to address these issues using the

Guanylate Binding Protein family as our model system.

THF GIJANYLATE BINDING PROTEINS rGBPs)

The Guanylate Binding Proteins, or GBPs, are a family of proteins that were discovered based on their interferon (IFN)-inducibility and were named due to their affinity for guanine nucleotides (19, 21). The cellular function(s) of this family of GTPases remains a mystery, although one report exists showing antiviral activity for a human GBP (3). Although 15 the role of these proteins in IFN-treated cells is unclear, this family of proteins has been shown to be a potentially useful marker for interferon activity in IFN-treated patients (20).

Additionally, the gene promoters for human and murine GBPs have been instrumental in deciphering the signal transduction pathways associated with interferon-mediated transcription

(12, 77, 91). Therefore, although the physiological function of GBPs remains unclear, this family of proteins has already contributed greatly to the field of interferon research. And as we will see, GBPs are now being shown to be useful models for the study of GTPases as well as novel aspects of protein prenylation.

Presently, 8 members of the GBP family (2 human, 4 mouse, 1 rat, and 1 chicken) have been cloned (5, 22, 47, 110, 124, 132). All are roughly 600 amino acids (-65 kD) in length and possess nonstandard GTP-binding motifs. Although they show great specificity for binding to guanine nucleotides (22, 47), GBPs lack the canonical third element of the GTP- binding motif that confers specificity for guanine (28). Based on their unusual GTP-binding domains, some researchers have speculated that GBPs belong to a new class of G-proteins that include members from diverse species such as arabidopsis and yeast (127).

In addition to their unusual GTP-binding domains, all of the GBPs so far tested have the equally unusual ability to hydrolyze GTP to both GDP and GMP. GMP constitutes only

10-15 % of the final hydrolysis product for human GBP-2 and chicken GBP (90, 110); however, it is the predominant product of GTP hydrolysis (85%) for human GBP-1 (111). If the GBPs interact with nucleotide-dependent effectors, this additional characteristic of the

GBPs to produce GMP may have functional significance as it would allow for 3 separate nucleotide binding states. However, these individual states would most likely be transient if not stabilized by effector binding, as the rates of GTP hydrolysis are quite high. This appears to be due to very fast dissociation of hydrolysis products (99).

Recent structural data on human GBP-1 has emerged showing this protein to adopt a highly peculiar structure as well (100). The N-terminal half of the molecule exists as a folded, 16

globular GTP-binding domain with some structural similarity to Ras. The C-terminal half is an elongated, rod-like domain composed entirely of a-helices. The authors of this work also showed that human GBP-1 exhibits nucleotide-dependent oligomerization and that human

GBP-1 GTPase activity is protein concentration-dependent. This cooperative effect on GTPase activity, along with other biochemical and structural features, prompted the authors to suggest that GBPs are in fact members of the dynamin superfamily of GTPases. Further work is necessary to confirm or reject this hypothesis.

Of the 8 cloned GBP family members, 6 possess CaaX motifs, with only human GBPl

(hGBPl) having a CIS-type CaaX motif. Early investigation into the prenylation status of the

GBPs showed that most behave as predicted by their CaaX motifs. Human GBP-1 has been shown to be famesylated in vitro (111) and in vivo (89). Ras p67 (123) and murine GBP-2

(124) are also both prenylated in vivo. In vivo labeling experiments have shown the identity of isoprenoid attached to murine GBP-2 to be geranylgeranyl, also as predicted. However, an unexpected finding was obtained when we began studies into the prenylation status of murine

GBP-1.

Murine GBP-1 (mGBPl) was cloned independently from two different laboratories using interferon-treated embryo cells (22) or macrophages (132). This protein shares tremendous similarity to the other GBP family members, being 70% identical to human GBP-1 at the amino acid level. Like the other GBPs, mGBPl shows exclusive specificity for binding guanine nucleotides, and it also possesses the same GGTase I-type CaaX motif (CTIL) found to be functional for rat p67 and murine GBP-2. However, unlike the other GBPs tested, our studies into the prenylation status of mGBPl, detailed in this thesis, show this protein to exhibit a startling deficiency in prenylation. This problem in prenylation appears to be an intrinsic property of mGBPl and demonstrates a new paradigm in prenylation whereby a protein with a functional CaaX motif fails to be stoichiometrically prenylated. A discussion of 17 the significance of this finding to the study of prenylated proteins, as well as the biological function of GBPs, follows the data chapters.

References

1. Adamson, P., C. J. Marshall, A. Hall, and P. A. Tilbrook. 1992. Post-

translational modifications of p21rho proteins. J. Biol. Chem. 267:20033-20038.

2. Adnane, J., F. A. Bizouarn, Y. Qian, A. Hamilton, and S. M. Sebti.

1998. p21wafl/cipl is upregulated by the geranylgeranyltransferase I inhibitor GGTI-

298 through a transforming growth factor P- and spl-responsive element: involvement

of the small GTPase RhoA. Mol. Cell. Biol. 18:6962-6970.

3. Anderson, S. L., J. M. Carton, J. Lou, L. Xing, and B. Y. Rubin. 1999.

Interferon-Induced Guanylate Binding Protein-1 (GBP-1) mediates an antiviral effect

against vesicular stomatitis virus and encephalomyocarditis virus. Virology 256:8-14.

4. Andres, D. A., J. L. Goldstein, Y. K. Ho, and M. S. Brown. 1993. Mutational analysis of a-subunit of protein famesyltransferase. J. Biol. Chem.

268:1383-1390.

5. Asundi, V. I., R. C. Stahl, L. Showalter, K. J. Conner, and D. J. Carey.

1994. Molecular cloning and characterization of an isoprenylated 67 kDa protein. BBA

121:257-265.

6. Bader, B., K. Kuhn, D. J. Owen, H. Waldmann, A. Wittinghofer, and J.

Kuhlmann. 2000. Bioorganic synthesis of lipid-modified proteins for the study of

signal transduction. Nature 403:223-226.

7. Barbacid, M. 1987. Ras genes. Ann. Rev. Biochem. 56:779-827.

8. Barton, R. M., and H. J. Worman. 1999. Prenylated prelamin A interacts with

narf, a novel nuclear protein. J. Biol. Chem. 274:30008-30018. 18

9. Bassa, B. V., D. D. Roh, N. D. Vaziri, M. A. Kirshenbaum, and V. S.

Kamanna. 1999. Effect of inhibition of cholesterol synthetic pathway on the activation

of Ras and MAP kinase in mesangial cells. BBA 1449:137-149.

10. Bishop, W. R., R. Bond, J. Petrin, L. Wang, R. Patton, R. Doll, G.

Njoroge, J. Catino, J. Schwartz, W. Windsor, R. Syto, J. Schwartz, D.

Carr, L. James, and P. Kirschmeier. 1995. Novel tricyclic inhibitors of famesyl

protein transferase. J. Biol. Chem. 270:30611-30618.

11. Booden, M. A., T. L. Baker, P. A. Solski, C. J. Der, S. G. Punke, and

J. E. Buss. 1999. A non-famesylated HRas protein can be palmitoylated and trigger

potent differentiation and transformation. J. Biol. Chem. 274:1423-1431.

12. Briken, V., H. Ruffner, U. Schultz, A. Schwarz, L. F. L. Ries, I.

Strehlow, T. Decker, and P. Staeheli. 1995. Interferon regulatory factor 1 is

required for mouse GBP gene activation by gamma interferon. Mol. Cell. Biol. 15:975-

982.

13. Buss, J. E., and J. C. Marsters, Jr. 1995. Farnesyl transferase inhibitors: the

successes and surprises of a new class of potential cancer chemotherapeutics. Chem.

Biol. 2:787-791.

14. Buss, J. E., P. A. Solski, J. P. Schaeffer, M. A. MacDonald, and C. J.

Der. 1989. Activation of the cellular proto-oncogene product p21ras by addition of a

myristylation signal. Science 243:1600-1603.

15. Caplin, B. E., Y. Ohya, and M. S. Marshall. 1998. Amino acid residues that

deHne both the isoprenoid and CAAX preferences of the saccharomyces cerevisiae

protein famesyltransferase. J. Biol. Chem. 273:9472-9479.

16. Casey, P. J., and M. C. Seabra. 1996. Protein prenyltransferases. J. Biol. Chem.

271:5289-5292. 19

17. Casey, P. J., P. A. Solski, C. J. Der, and J. E. Buss. 1989. p21Ras is

modified by a farnesyl isoprenoid. Proc Natl Acad Sci U S A 86:8323-8327.

18. Cassidy, P. B., and C. D. PouUer. 1996. Transition state analogs for protein

famesyltransferase. J. Am. Chem. Soc. 118:8761-8762.

19. Cheng, Y.-S. E., M. F. Becker-Manley, T. P. Chow, and D. C. Horan.

1985. Affinity purification of an interferon-induced human guanylate-binding protein and

its characterization. J. Biol. Chem. 260:15834-15839.

20. Cheng, Y.-S. E., M. F. Becker-Manley, R. G. Rucker, and E. C.

Borden. 1988. Accumulation of guanylate binding proteins in patients treated with

interferons. J. of Interferon Res. 8:385-391.

21. Cheng, Y.-S. E., R. J. Colonno, and F. H. Yin. 1983. Interferon induction of

fibroblast proteins with guanylate binding activity. J. Biol. Chem. 258:7746-7750.

22. Cheng, Y.-S. E., C. E. Patterson, and P. Staeheli. 1991. Interferon-induced

guanylate-binding proteins lack an N(T)KXD consensus motif and bind GMP in addition

to GDP and OTP. Mol. Cell. Biol. 11:4717-4725.

23. Choy, E., V. K. Chiu, J. Silletti, M. Feoktistov, T. Morimoto, D.

Michaelson, I. E. Ivanov, and M. R. Philips. 1999. Endomembrane trafficking

of Ras; the CaaX motif targets proteins to the ER and Golgi. Cell 98:69-80.

24. Cox, A. D., A. M. Garcia, J. K. Westwick, J. J. Kowalczyk, M. D.

Lewis, D. A. Brenner, and C. J. Der. 1994. The CAAX peptidomimetic

compound B581 specifically blocks famesylated, but not geranylgeranylated or

myristylated, oncogenic ras signaling and transformation. J. Biol. Chem. 269:19203-

19206.

25. Cox, A. D., S. M. Graham, P. A. Solski, J. E. Buss, and C. J. Der.

1993. The carboxyl-terminal CXXX sequence of Gi alpha, but not Rab5 or Rabll,

supports Ras processing and transforming activity. J. Biol. Chem. 268:11548-11552. 20

26. Dai, Qm E. Choy, V. Chiu, J. Romano, S. R. Slivka, S. A. Steitz, S.

Michaelis, and M. R. Philips. 1998. Mammalian prenylcysteine carboxyl

methyltransferase is in the endoplasmic reticulum. J. Biol. Chem. 273:15030-15034.

27. Del Villar, K., H. Mitsuzawa, W. Yang, I. Sattler, and F. Tamanoi. 1997.

Amino acid substitutions that convert the protein substrate specificity of

famesyltransferase to that of geranylgeranyltransferase type I. J. Biol. Chem. 272:680-

687.

28. Dever, T. E., M. J. Glynias, and W. C. Merrick. 1987. GTP-binding domain:

three consensus sequence elements with distinct spacing. Proc Natl Acad Sci U S A

84:1814-1818.

29. Dolence, J. M., P. B. Cassidy, J. R. Mathis, and C. D. Poulter. 1995.

Yeast protein famesyltransferase: steady-state kinetic studies of substrate binding.

Biochemistry 34:16687-16694.

30. Dolence, J. M., and C. D. Poulter. 1995. A mechanism for posttranslational

modifications of proteins by yeast protein famesyltransferase. Proc Natl Acad Sci U S A

92:5008-5011.

31. Du, W., P. F. Lebowitz, and G. C. Prendergast. 1999. Cell growth inhibition

by famesyltransferase inhibitors is mediated by gain in geranylgeranylated RhoB. Mol.

Cell. Biol. 19:1831-1840.

32. Dudler, T., and M. H. Gelb. 1997. Replacement of the H-Ras famesyl group by

lipid analogues: implications for downstream processing and effector activation in

Xenopus oocytes. Biochemistry 36:12434-12441.

33. Elad, G., A. Paz, R. Haklai, D. Marciano, A. Cox, and Y. Kloog. 1999.

Targeting of K-Ras 4B by S-trans,trans-famesyl thiosalicylic acid. BBA 1452:228-242. 21

34. Epstein, W. W., D. Lever, L. M. Leining, E. Bruenger, and H. C.

Rilling. 1991. Quantitation of prenylcysteines by a selective cleavage reaction. Proc

Natl Acad Sci U S A 88:9668-9670.

35. Essig, M., G. Nguyen, D. Prie, B. Escoubet, J.-D. Sraer, and G.

Friedlander. 1998. 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors

increase Hbrinolytic activity in rat aortic enothelial cells. Circ. Res. 83:683-690.

36. Farh, L., D. A. Mitchell, and R. J. Deshenes. 1995. Famesylation and

proteolysis are sequential, but distinct steps in the CaaX box modification pathway.

Arch. Biochem. Biophys. 318:113-121.

37. Farnsworth, C. C., P. J. Casey, W. N. Howald, J. A. Glomset, and M.

H. Gelb. 1990. Structural characterization of prenyl groups attached to proteins.

METHODS: A companion to Methods in Enzymology 1:231-240.

38. Farnsworth, C. C., S. L. Wolda, M. H. Gelb, and J. A. Glomset. 1989.

Human iamin B contains a famesylated cysteine residue. J. Biol. Chem. 264:20422-

20429.

39. Fu, H.-W., L. S. Beese, and P. J. Casey. 1998. Kinetic analysis of zinc ligand

mutants of mammalian protein famesyltransferase. Biochemistry 37:4465-4472.

40. Furflne, E. S., J. J. Leban, A. Landavazo, J. F. Moomaw, and P. J.

Casey. 1995. Protein famesyltransferase: kinetics of famesyl pyrophosphate binding

and product release. Biochemistry 34:6857-6862.

41. Gadbut, A. P., L. Wu, D. Tang, A. Papageorge, J. A. Watson, and J. B.

Galper. 1997. Induction of the cholesterol metabolic pathway regulates the famesylation

of RAS in embryonic chick heart cells: a new role for Ras in regulating the expression of

muscarinic receptors and G proteins. EMBO J. 16:7250-7260. 22

42. Gelb, M. H., J. D. Scholten, and J. S. Sebolt-Leopold. 1999. Protein

prenylation: from discovery to prospects for cancer threatment. Curr. Opin. Chem. Biol.

2:40-48.

43. Glenn, J. S., J. Marsters, J.C., and H. B. Greenberg. 1998. Use of a

prenylation inhibitor as a novel antiviral agent. J. Virol, 72:9303-9306.

44. Goaistone, M. L., and B. Draznin. 1996. Effect of insulin on famesyltransferase

activity in 3T3-L1 adipocytes. J. Biol. Chem. 271:27585-27589.

45. Goaistone, M. L., J. W. Leitner, K. Wall, L. Dolgonos, K. I. Rother, D.

Accili, and B. Draznin. 1998. Effect of insulin on famesyltransferase: specificity of

insulin action and potentiation of nuclear effects of insulin-like growth factor-1,

epidermal growth factor, and platelet-derived growth factor. J. Biol. Chem. 273:23892-

23896.

46. Haklai, R., M. G. Weisz, G. Elad, A. Paz, D. Marciano, Y. Egozi, G.

Ben-Baruch, and Y. Kloog. 1998. Dislodgement and accelerated degradation of ras.

Biochemistry 37:1306-1314.

47. Han, B. H., D. J. Park, R. W. Lim, J. H. Im, and H. D. Kim. 1998.

Cloning, expression, and characterization of a novel guanyiate-binding protein, GBP3 in

murine erythroid progenitor cells. BBA 1384:373-386.

48. Hancock, J. F., A. I. Magee, J. E. Childs, and C. J. Marshall. 1989. All

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

49. Hasne, M., and F. Lawrence. 1999. Characterization of prenylated protein

methyltransferase in Leishmania. Biochem. J. 342:513-518.

50. Hayes, J. S., O. A. Lawler, M.-T. Walsh, and B. T. Kinsella. 1999. The

prostacyclin receptor is isoprenylated: isoprenylation is required for efficient recepetor-

effector coupling. J. Biol. Chem. 274:23707-23718. 23

51. Heilmeyer, J., L.M.G., M. Serwe, C. Weber, J. Metzger, E. Hoffmann- Posorske, and H. E. Meyer. 1992. Farnesylcysteine, a constituent of the a and ^

subunits of rabbit skeletal muscle phosphorylase kinase: Localization by conversion to

S-ethylcysteine and by tandem mass spectrometry. Proc Natl Acad Sci U S A 89:9554-

9558.

52. Hightower, K. E., C.-c. Huang, and C. A. Fierke. 1998. H-Ras peptide and

protein subsu-ates bind protein famesyltransferase as an ionized thiolate. Biochemistry

37:15555-15562.

53. Holtz, D., R. A. Tanaka, J. Hartwig, and F. McKeon. 1989. The CaaX motif

of lamin A functions in conjunction with the nuclear localization signal to target assembly

to the nuclear envelope. Cell 59:969-977.

54. Huang, C.-C., P. J. Casey, and C. A. Fierke. 1997. Evidence for a catalytic

role of zinc in protein famesyltransferase. J. Biol. Chem. 272:20-23.

55. Jackson, J. H., C. G. Cochrane, J. R. Bourne, P. A. Solski, J. E. Buss,

and C. J. Der. 1990. Famesol modiHcation of Kirsten-ras exon 4B protein is essential

for transformation. Proc Natl Acad Sci U S A 87:3042-3046.

56. James, G., et al., and Goldstein, J.L. 1994. Benzodiazepine peptidomimetic

BZA-5B interrupts the MAP kinase activation pathway in H-Ras-transformed Rat-1 cells,

but not in untransformed cells. J. Biol. Chem. 269:27705-27714.

57. James, G., J. L. Goldstein, and M. S. Brown. 1996. Resistance of K-

Ras4BV12 proteins to famesyltransferase inhibitors in Rati cells. Proc Natl Acad Sci

USA 93:4454-4458.

58. James, G. L., J. L. Goldstein, M. S. Brown, T. E. Rawson, T. C.

Somers, R. S. McDowell, C. W. Crowley, B. K. Lucas, A. D. Levinson,

and J. C. Marsters, Jr. 1993. Benzodiazepine peptidomimetics: potent inhibitors of

ras famesylation in animal cells. Science 260:1937-1942. 24

59. Jones, T. L. Z., and A. M. Spiegel. 1990. Isoprenylation of an inhibitory G

protein alpha subunit occurs only upon mutagenesis of the carboxyl terminus. J. Biol.

Chem. 265:19389-19392.

60. Kalman, V. K., R. A. Erdman, W. A. Maltese, and J. D. Robishaw. 1995.

Regions outside of the CAAX motif influence the specificity of prenylation of G protein

gamma subunits. J. Biol. Chem. 270:14835-14841.

61. Kamiya, Y., A. Sakurai, S. Tamura, and N. Takahashi. 1978. Structure of

rhodotorucine A, a novel lipopeptide, inducing mating tube formation in rhodosporidium

toruloides. BBRC 83:1077-1083.

62. Kato, K., A. D. Cox, M. M. Hisaka, S. M. Graham, J. E. Buss, and C.

J. Der. 1992. Isoprenoid addition to Ras protein is the critical modification for its

membrane association and transforming activity. Proc Natl Acad Sci U S A 89:6403-

6407.

63. Kawabata, M., T. Imamura, K. Miyazono, M. E. Engel, and H. L. Moses. 1995. Interaction of the transforming growth factor-p type I receptor with

famesyl-protein transferase-a. J. Biol. Chem. 270:29628-29631.

64. Kilic, F., M. B. Dalton, S. K. Burrell, J. P. Mayer, S. D. Patterson, and

M. Sinensky. 1997. In Vitro assay and characterization of the famesylation-dependent

prelamin A endoprotease. J. Biol. Chem. 272:5298-5304.

65. Kilic, F., D. A. Johnson, and M. Sinensky. 1999. Subcellular localization and

partial purification of prelamin A endoprotease: an enzyme which catalyzes the

conversion of famesylated prelamin A to mature lamin A. FEBS Letters 450:61-65.

66. Kim, E., P. Ambroziac, J. C. Otto, B. Taylor, M. Ashby, K. Shannon,

P. J. Casey, and S. G. Young. 1999. Disruption of the mouse Reel gene results in

defective Ras processing and mislocalization of Ras in cells. J. Biol. Chem. 274:8383-

8390. 25

67. Kinsella, B. T., R. A. Erdman, and W. A. Maltese. 1991. Posttranslational

modiHcation of Ha-ras p21 by farnesyl versus geranylgeranyl isoprenoids is determined

by the COOH-terminai amino acid. Proc Natl Acad Sci U S A 88:8934-8938.

68. Kisselev, O., M. Ermolaeva, and N. Gautam. 1995. Efficient interaction with a

receptor requires a specific type of prenyl group on the G protein gamma subunit. J.

Biol. Chem. 270:25356-25358.

69. Kohl, N. E., M. W. Conner, J. B. GIbbs, S. L. Graham, G. D. Hartman,

and A. Oliff. 1995. Development of inhibitors of protein famesylation as potential

chemotherapeutic agents. J. Cellular Biochemistry 22:145-150.

70. Kohl, N. E., S. D. Mosser, S. J. deSolms, E. A. Giuliani, D. L.

Pompliano, S. L. Graham, R. L. Smith, E. M. Scolnick, A. Oliff, and J.

B. Gibbs. 1993. Selective inhibition of ras-dependent transformation by a

famesyltransferase inhibitor. Science 260:1934-1937.

71. Kohl, N. E., C. A. Omer, M. W. Conner, N. J. Anthony, J. P. Davide,

S. J. deSolms, E. A. Giuliani, R. P. Gomez, S. L. Graham, K.

Hamilton, L. K. Hant, G. D. Hartman, K. S. Koblan, A. M. Krai, P. J.

Miller, S. D. Mosser, T. J. O'Neill, E. Rands, M. D. Schaber, J. B.

Gibbs, and A. Oliff. 1995. Inhibition of famesyltransferase induces regression of

mammary and salivary carcinomas in ras U'ansgenic mice. Nature Med. 1:792-797.

72. Kohl, N. E., F. R. Wilson, S. D. Mosser, E. Giuliani, S. J. deSolms,

M. W. Conner, N. J. Anthony, W. J. Holtz, R. P. Gomez, T.-J. Lee, R.

L. Smith, S. L. Graham, G. D. Hartman, J. B. Gibbs, and A. Oliff. 1994.

Protein famesyltransferase inhibitors block the growth of ras-dependent tumors in nude

mice. Proc Natl Acad.Sci USA 91:9141-9145. 26

73. Krai, A. M., R. E. Diehl, S. J. deSolms, T. M. Williams, N. E. Kohl, and C. A. Omer. 1997. Mutational analysis of conserved residues of the p-subunit of

human famesyl:protein transferase. J. Biol. Chem. 272:27319-27323.

74. Lebowitz, P. F., P. J. Casey, G. C. Prendergast, and J. A. Thissen.

1997. Famesyltransferase inhibitors alter the prenylation and growth-stimulating function

of RhoB. J. Biol. Chem. 272:15591-15594.

75. Lerner, E. C., Y. Qian, M. A. Blaskovich, R. D. Possum, A. Vogt, J.

Sun, A. D. Cox, C. J. Der, A. D. Hamilton, and S. M. Sebti. 1995. Ras

CAAX peptidomimetic Fn-277 selectively blocks oncogenic Ras signaling by inducing

cytoplasmic accumulation of inactive Ras-Raf complexes. J. Biol.Chem. 270:26802-

26806.

76. Leventis, R., and J. R. Silvius. 1998. Lipid-binding characteristics of the

polybasic carboxy-terminal sequence of K-Ras4B. Biochemistry 37:7640-7648.

77. Lew, 0. J., T. Decker, I. Strehlow, and J. E. Darnell. 1991. Overlapping

elements in the guanylate-binding protein gene promoter mediate transcriptional induction

by alpha and gamma interferons. Mol. Cell. Biol. 11:182-191.

78. Lindorfer, M. A., N. E. Sherman, K. A. Woodfork, J. E. Fletcher, D. F.

Hunt, and J. C. Garrison. 1996. G protein y subunits with altered prenylation

sequences are properly modified when expressed in Sf9 cells. J. Biol. Chem.

271:18582-18587.

79. Long, S. B., P. J. Casey, and L. S. Beese. 1998. Cocrystal structure of protein

famesyltransferase complexed with a famesyl diphosphate substrate. Biochemistry

37:9612-9618.

80. Lowy, D. R., and B. M. Willumsen. 1993. Function and regulation of Ras. Ann.

Rev. Biochem. 62:851-891. 27

81. Lutz, R. J., M. A. Trujillo, K. S. Denham, L. Wenger, and M. Sinensky.

1992. Nucleoplasmic localization of prelamin A: implications for prenylation-dependent

lamin A assembly into the nuclear lamina. Proc Natl Acad Sci U S A 89:3000-3004.

82. Marom, M., R. Haklai, G. Ben-Baruch, D. Marciano, Y. Egozi, and Y.

Kloog. 1995. Selective inhibition of ras-dependent cell growth by famesylthiosalicylic

acid. J. Biol. Chem. 270:22263-22270.

83. Marshall, C. J. 1993. Protein prenylation: a mediator of protein-protein interactions.

Science 259:1865-1866.

84. McGeady, P., S. Kuroda, K. Shimizu, Y. Takai, and M. H. Gelb. 1995.

The famesyl group of H-Ras facilitates the activation of a soluble upstream activator of

mitogen-activated protein kinase. J. Biol. Chem. 270:26347-26351.

85. Melkonian, K. A., A. G. Ostermeyer, J. Z. Chen, M. G. Roth, and D. A.

Brown. 1999. Role of lipid modifications in targeting proteins to detergent-resistant

membrane rafts: many raft proteins are acylated, while few are prenylated. J. Biol.

Chem. 274:3910-3917.

86. Mineo, C., G. L. James, E. J. Smart, and R. G. W. Anderson. 1996.

Localization of epidermal growth factor-stimulated Ras/Raf-1 interaction to caveolae

membrane. J. Biol. Chem. 271:11930-11935.

87. Moores, S. L., M. D. Schaber, S. D. Mosser, E. Rands, M. B. O'Hara,

V. M. Garsky, M. S. Marshall, D. L. Pompliano, and J. B. Gibbs. 1991.

Sequence dependence of protein isoprenylation. J. Biol. Chem. 266:14603-14610.

88. Myung, C.-S., H. Yasuda, W. W. Liu, T. K. Harden, and J. C.

Garrison. 1999. Role of isoprenoid lipids on the heterotrimeric G protein y subunit in

determining effector activation. J. Biol. Chem. 274:16595-16603. 28

89. Nantais, D. E., M. Schwemmie, J. T. Stickney, D. J. Vestal, and J. E. Buss. 1996. Prenylation of an interferony-induced GTP-binding protein: the human

guanylate binding protein, huGBPl. J. Leukoc. Biol. 60:423-431.

90. Neun, R., M. F. Richter, P. Staeheli, and M. Schwennnile. 1996. GTPase

properties of the interferon-induced human guanylate-binding protein 2. FEBS Letters

390:69-72.

91. Nicolet, C. M., and D. M. Paulnock. 1994. Promoter analysis of an interferon-

inducible gene associated with macrophage activation. J. Immunology 152:153-162.

92. Niv, H., O. Gutman, Y. I. Henis, and Y. Kloog. 1998. Membrane interactions

of constitutively active GFP-Ki-Ras4B and their role in signaling. J. Biol. Chem.

273:1606-1613.

93. OlifT, A. 1999. Famesyltransferase inhibitors: targeting the molecular basis of cancer.

BBA 1423:C19-C30.

94. Otto, J. C., and P. J. Casey. 1996. The hepatitis delta virus large antigen is

famesylated both in vitro and in animal cells. J. Biol. Chem. 271:4569-4572.

95. Otto, J. C., E. Kim, S. G. Yound, and P. J. Casey. 1999. Cloning and

charaterization of a mammalian prenyl protein specific protease. J. Biol. Chem.

274:8379-8382.

96. Page, M. J., A. Aitken, D. J. Cooper, A. I. Magee, and P. N. Lowe.

1990. Use of the insect/baculovirus expression system to produce lipid-modified ras

proteins for electrospray mass spectrometric studies. METHODS: A Companion to

Methods in Enzymology 1:221-230.

97. Park, H.-W., S. R. Boduluri, J. F. Moomaw, P. J. Casey, and L. S.

Beese 1997. Crystal structure of protein famesyltransferase at 2.25 angstrom

resolution. Science 275:1800-1804. 29

98. Pellicena, P., J. D. Scholten, K. Zimmerman, M. Creswell, C. C. Huang, and W. T. Miller. 1996. Involvement of the a subunit of famesyl-protein

transferase in substrate recognition. Biochemistry 35:13494-13500.

99. Praefcke, G. J. K., M. Geyer, and C. Herrmann. 1999. Nucleotide-binding

characteristics of human Guanylate-binding Protein 1 (hGBPl) and identification of the

third GTP-binding motif J. Molec. Biol. 292:321-332.

ICQ. Prakash, B., G. J. K. Praefcke, L. Renault, A. Wittinghofer, and C.

Herrmann. 2000. Structure of human guanylate-binding protein 1 representing a unique

class of GTP-binding proteins. Nature 403:567-571.

101. Reiss, Y., M. S. Brown, and J. L. Goldstein. 1992. Divalent cation and prenyl

pyrophosphate specificities of the protein famesyltransferase from rat brain, a zinc

metalloenzyme. J. Biol. Chem. 267:6403-6408.

102. Reiss, Y., J. L. Goldstein, M. C. Seabra, P. J. Casey, and M. S. Brown.

1990. Inhibition of purified p21ras famesyhprotein transferase by Cys-AAX

tetrapeptides. Cell 62:81-88.

103. Reiss, Y., M. C. Seabra, S. A. Armstrong, C. A. Slaughter, J. L.

Goldstein, and M. S. Brown. 1991. Nonidentical subunits of p21H-ras

famesyltransferase. J. Biol. Chem. 266:10672-10677.

104. Roskoski, J., R., and P. Ritchie. 1998. Role of the carboxyterminal residue in

peptide binding to protein famesyltransferase and protein geranylgeranyltransferase.

Archives of Biochemistry and Biophysics 356:167-176.

105. Rowell, C. A., J. J. Kowalczyk, M. D. Lewis, and A. M. Garcia. 1997.

Direct demonstration of geranylgeranylation and famesylation of Ki-Ras in vivo. J. Biol.

Chem. 272:14093-14097. 30

106. Roy, S., A. Lane, J. Yan, R. McPherson, and J. F. Hancock. 1997.

Activity of plasma membrane-recniited Raf-1 is regulated by Ras via the Raf zinc fmger.

J. Biol. Chem. 272:20139-20145.

107. Roy, S., R. Luetterforst, A. Harding, A. Apollini, M. Etheridge, E.

Stang, B. Rolls, J. F. Hancock, and R. G. Parton. 1999. Dominant-negative

caveolin inhibits H-Ras function by disrupting cholesterol-rich plasma membrane

domains. Nature Cell Biol. 1:97-105.

108. Sakagami, Y., M. Yoshida, A. Isogai, and A. Suzuki. 1981. Peptidal sex

hormones inducing conjugation tube formation in compatible mating-type cells of tremella

mesenterica. Science 212:1525-1527.

109. Schroeder, H., R. Leventis, S. Rex, M. Schelhaas, E. Nagele, H.

Waldmann, and J. R. Silvius. 1997. S-acylation and plasma membrane targeting of

the famesylated carboxy-terminal peptide of N-Ras in mammalian fibroblasts.

Biochemistry 30:13102-13109.

110. Schwemmle, M., B. Kaspers, A. Irion, P. Staeheli, and U. Schultz. 1996.

Chicken guanylate-binding protein. J. Biol. Chem. 271:10304-10308.

111. Schwemmle, M., and P. Staeheli. 1994. The interferon-induced 67-kDa

guanylate-binding protein (hGBPl) is a GTPase that converts GTP to GMP. J. Biol.

Chem. 269:11299-11305.

112. Sepp-Lorenzino, L., S. Ma, E. Rands, N. E. Kohl, J. B. Gibbs, A.

Oliff, and N. Rosen. 1995. A peptidomimetic inhibitor of famesyl:protein transferase

blocks the anchorage-dependent and -independent growth of human tumor cell lines.

Cancer Research 55:5302-5309.

113. Shahinian, S., and J. R. Silvius. 1995. Doubly-lipid-modified protein sequence

motifs exhibit long-lived anchorage to lipid bilayer membranes. Biochemistry 34:3813-

3822. 31

114. Siddiqui, A. A., J. R. Garland, M. B. Dalton, and M. Sinensky. 1998.

Evidence for a high afflnity, saturable, prenyiation-dependent p21Ha-ras binding site in

plasma membranes. J. Biol. Chem. 273:3712-3717. lis. Stark, J., W.W., M. A. Blaskovich, B. A. Johnson, Y. Qian, A.

Vasudevan, B. Pitt, A. D. Hamilton, S. M. Sebti, and P. Davies. 1998.

Inhibiting geranylgeranylation blocks growth and promotes apoptosis in pulmonary

vascular smooth muscle cells. Am. J. Physiol. 275:L55-L63.

116. Stirtan, W. G., and C. D. Poulter. 1997. Yeast protein geranylgeranyltransferase

type-I: steady-state kinetics and substrate binding. Biochemistry 36:4552-4557.

117. Strickland, C. L., W. T. Windsor, R. Syto, L. Wang, R. Bond, Z. Wu,

J. Schwartz, H. V. Le, L. S. Beese, and P. C. Weber. 1998. Crystal

structure of famesyl protein transferase coupled with a CaaX peptide and famesyl

diphosphate analogue. Biochemistry 37:16601-16611.

118. Sun, J., Y. Qian, Z. Chen, J. Marfurt, A. Hamilton, and S. M. Sebti.

1999. The geranylgeranyltransferase I inhibitor GGTI-298 induces hypophosphorylation

of retinoblastoma and partner switching of cyclin-dependent kinase inhibitors. J. Biol.

Chem. 274:6930-6934.

119. Tnieblood, C. E., V. L. Boyartchuck, and J. Rine. 1997. Substrate specificity

determinants in the famesyltransferase P-subunit. Proc Natl Acad Sci U S A 94:10774-

10779.

120. Tschantz, W. R., E. S. Furfine, and P. J. Casey. 1997. Substrate binding is

required for release of product from mammalian protein famesyltransferase. J. Biol.

Chem. 272:9989-9993.

121. Tschantz, W. R., L. Zhang, and P. J. Casey. 1999. Cloning, expression, and

cellular localization of a human prenylcysteine lyase. J. Biol. Chem. 274:35802-35808. 32

122. Ventura, F., F. Liu, J. Doody, and J. Massague. 1996. Interaction of

transforming growth factor-P receptor I with famesyl-protein transferase-a in yeast and

mammalian cells. J. Biol. Chem. 271:13931-13934.

123. Vestal, D. J., J. E. Buss, G. S. Kelner, D. Maciejewski, V. K. Asundi,

and R. A. Maki. 1996. Rat p67 GBP is induced by interferon-gamma and isoprenoid-

modified in macrophages. Biochem. Biophys. Res. Commun. 224:528-534.

124. Vestal, D. J., J. E. Buss, S. R. McKercher, N. A. Jenkins, K. G.

Copeland, G. S. Kelner, V. K. Asundi, and R. A. Maki. 1998. Murine GBP- 2: A new IFN-y-induced member of the GBP family of GTPases from macrophages. J.

Interferon and Cytokine Res. 18:977-985.

125. Vojtek, A. B., and C. J. Der. 1998. Increasing complexity of the Ras signaling

pathway. J. Biol. Chem. 273:19925-19928.

126. Wallace, A., K. S. Koblan, K. Hamilton, D. J. Marquis-Omer, P. J.

Miller, S. D. Mosser, C. A. Omer, M. D. Schaber, R. Cortese, A. OliH',

J. B. Gibbs, and A. Pessi. 1996. Selection of potent inhibitors of famesyl-protein

transferase from a synthetic tetrapeptide combinatorial library. J. Biol. Chem.

271:31306-31311.

127. Wang, H., S. K. Lockwood, M. F. Hoeltzel, and J. W. Schiefelbein.

1997. The root hair defectives gene encodes an evolutionarily conserved protein with

GTP-binding motifs and is required for regulated cell enlargement in arabidopsis. Genes

and Development 11:799-811.

128. Wang, T., P. D. Danielson, B. Li, P. C. Shah, S. D. Kim, and P. K.

Donahoe. 1996. The p2Iras fanesyltransferase a subunit in TGF-P and activin

signaling. Science 271:1120-1122. 33

129. Whyte, D. B., P. Kirschmeier, T. N. Hockenberry, I. Nunez-Oliva, L.

James, J. J. Catino, W. R. Bishop, and J.-K. Pai. 1997. K- and N-Ras are

geranylgeranylated in cells treated with famesyl protein transferase inhibitors. J. Biol.

Chem. 272:14459-14464.

130. Willumsen, B. M., A. Christensen, N. L. Hubbert, A. G. Papageorge,

and D. R. Lowy. 1984. The p21ras C-terminus is required for transformation and

membrane association. Nature 310:583-586.

131. Willumsen, B. M., A. D. Cox, P. A. Solski, C. J. Der., and J. E. Buss.

1996. Novel determinants of H-Ras plasma membrane localization and transformation.

Oncogene 13:1901-1909.

132. Wynn, T. A., C. M. Nicolet, and D. M. Paulnock. 1991. Identification and

characterization of a new gene family induced during macrophage activation. J. Immunol.

147:4384-4392.

133. Yang, W., K. Del Villar, J. Urano, H. Mitsuzawa, and F. Tamanoi. 1997.

Advances in the development of famesyltransferase inhibitors: substrate recognition by

protein famesyltransferase. J. Cellular Biochemistry 27:12-19.

134. Yasuda, H., M. A. Lindorfer, K. A. Woodfork, J. E. Fletcher, and J. C. Garrison. 1996. Roie of the prenyl group on the G protein y subunit in coupling

trimeric G proteins to A1 adenosine receptors. J. Biol. Chem. 271:18588-18595.

135. Ying, W., L. Sepp-Lorenzino, K. Cai, P. Aloise, and P. S. Coleman.

1994. Photoaffinity-labeling peptide substrates for famesyl-protein transferase and the

intersubunit location of the active site. J. Biol. Chem. 269:470-477.

136. Yokoyama, K., P. Trobridge, F. S. Buckner, W. C. Van Voorhis, K. D.

Stuart, and M. H. Gelb. 1998. Protein famesyltransferase from trypanosoma

brucei. J. Biol. Chem. 273:26497-26505. 34

137. Yokoyama, K., K. Zimmerman, J. Scholten, and M. H. Gelb. 1997.

Differential prenyl pyrophosphate binding to mammalian protein

geranylgeranyltransferase-I and protein famesyltransferase and its consequences on the

specificity of protein prenylation. J. Biol. Chem. 1997:3944-3952.

138. Zhang, F. L., and P. J. Casey. 1996. Protein prenylation; molecular mechanisms

and functional consequences. Annu. Rev. Biochem. 65:241-269.

139. Zhang, F. L., H.-W. Fu, P. J. Casey, and W. R. Bishop. 1996. Substitution

of cadmium for zinc in famesyUprotein transferase alters its substrate specificity.

Biochemistry 35:8166-8171.

140. Zhang, F. L., P. Kirschmeier, D. Carr, L. James, R. W. Bond, L. Wang,

R. Patten, W. T. Windsor, R. Syto, R. Zhang, and W. R. Bishop. 1997.

Characterization of Ha-Ras, N-Ras, Ki-Ras4A, and Ki-Ras4B as in vitro substrates for

famesyl protein transferase and geranylgeranyl protein transferase type I. J. Biol. Chem.

272:10232-10239.

141. Zhang, F. L., J. F. Moomaw, and P. J. Casey. 1994. Properties and kinetic

mechanism of recombinant mammalian protein geranylgeranyltransferase type I. J. Biol.

Chem. 269:23465-23470.

142. Zhang, L., C. W. Sachs, R. L. Fine, and P. J. Casey. 1994. Interaction of

prenylcysteine methyl esters with the multidrug resistance transporter. J. Biol. Chem.

269:15973-15976.

143. Zhang, L., C. W. Sachs, H. W. Fu, R. L. Fine, and P. J. Casey. 1995.

Characterization of prenylcysteines that interact with P-glycoprotein and inhibit drug

transport in tumor cells. J. Biol. Chem. 270:22859-22865.

144. Zhang, L., W. R. Tschantz, and P. J. Casey. 1997. Isolation and

characterization of a prenylcysteine lyase from bovine brain. J. Biol. Chem. 272:23354-

23359. 35

CHAPTER 2. MURINE GBPl: INCOMPLETE C20

ISOPRENOID MODIFICATION OF AN IFN7-INDUCIBLE GTP-BINDING PROTEIN

A paper submitted to Molecular Biology of the Cell

John T. Stickney and Janice E. Buss

Abstract

Famesylation of Ras proteins is necessary for U'ansforming activity. Although faroesyl transferase inhibitors show promise as anti-cancer agents, prenylation of the most commonly mutated Ras isoform, K-Ras4B, is difficult to prevent because K-Ras4B can be alternatively modified with geranylgeranyl (C20). Little is known of the mechanisms that produce incomplete or inappropriate prenylation. Among non-Ras proteins with CaaX motifs, murine guanylate binding protein (mGBPI) was conspicuous for its unusually low incorporation of

[^H]mevalonate. Possible problems in cellular isoprenoid metabolism or prenyl transferase activity were investigated, but none that caused this defect were identified, implying that the poor labeling actually represented incomplete prenylation of mGBPI itself. Mutagenesis indicated that the last 18 residues of mGBPI severely limited C20 incorporation, but surprisingly were compatible with famesyl (C15) modification. Features leading to expression of mutant GBPs with partial isoprenoid modification were identified. The results demonstrate that it is possible to alter a protein's prenylation state in a living cell so that graded effects of isoprenoid on function can be studied. The C20-selective impairment in prenylation also identifies mGBPI as an important model for study of substrate:GGTase I interactions. 36

Introduction

Interest in prenylation has stemmed from the discovery that key proteins in multiple signal transduction cascades contain covalently attached isoprenoids (Casey, 1995). Perhaps the most notable examples are the Ras proteins. Mutated forms of Ras proteins are found in

30% of all human tumors (Lowy and Willumsen, 1993). However, these mutant Ras proteins are not oncogenic if they cannot be prenylated (Lowy and Willumsen, 1993). Prevention of

Ras prenylation thus holds promise as a new tactic for cancer chemotherapy (Cox and Der,

1997). To this end, many prenylation inhibitors have been developed, several of which appear to be effective anti-cancer agents in animal studies and are undergoing clinical uials (Buss and

Marsters, 1995; Gelb et al., 1999; Kohl et al., 1995; Liu et al., 1998).

It is currently presumed that in order to be effective these drugs will need to prevent isoprenoid modification of oncogenic Ras entirely. However, forms of Ras which are incompletely modified have received little study (Gadbut et al., 1997; Goalstone and Draznin,

1996), largely because of the assumption, based on direct physical studies, that prenyl proteins are fully and completely modified (Famsworth et al., 1990; Myung et al., 1999; Page et al.,

1990). It is still not known if all functions of oncogenic Ras require prenylation or if some effector pathways may remain active regardless of prenylation state.

The signals which permit isoprenoid attachment are known in some detail. Either a famesyl (15 carbon) or geranylgeranyl (20 carbon) isoprenoid is attached through a thioether linkage to a cysteine residue at the C-terminus of the protein (Gelb et al., 1999; Seabra, 1998;

Zhang and Casey, 1996). Enzymes that catalyze isoprenoid modification of proteins are grouped into two major classes: geranylgeranyl transferase U, which recognizes C-terminal X-

X-Cys-Cys, Cys-Cys-X-X, and Cys-X-Cys motifs, and CaaX motif prenyltransferases. CaaX motifs consist of a cysteine followed by two amino acids that frequently are aliphatic, then the final amino acid of the protein, X. The X residue is currently believed to be the major factor determining which of the two CaaX protein prenyl transferases will modify the CaaX cysteine 37

(Kinsella et al., 1991; Moores et al., 1991). Farnesyl transferase (FTase) modifies proteins with X residues such as Met, Ser, Ala, Gin or Asn. Geranylgeranyl Transferase I (GGTase I) preferentially modifles CaaX proteins with X residues of Leu or Phe. However, in cells treated with an FTase inhibitor, the K-Ras4B protein, whose CaaX motif (CVIM) is usually modified by FTase, can become C20 modiHed (James et al., 1996; Rowell et al., 1997; Sun et al., 1998;

Zhang et al., 1997). This ability of GGTase I to modify particular FTase substrates is a difficult problem for design of drugs aimed at preventing Ras protein famesylation. Current inhibitors are designed to specifically inhibit FTase, and can do little to prevent K-Ras4B cross-prenylation by GGTase I.

Despite the details identified regarding how isoprenoids are attached to proteins, little is known of the mechanisms that GGTase I or FTase use to exclude inappropriate proteins.

FTase and GGTase I are heterodimers with a shared a, but distinct |3 subunits (Zhang and

Casey, 1996). The ^ subunit binds the isoprenoid and is therefore responsible for the lipid specificity of each prenyl transferase. Recent crystallographic studies of FTase indicate that amino acids from both a and p subunits contribute to the site at which the CaaX motif of the protein substrate binds (Park et al., 1997; Strickland et al., 1998). This combination presents a further challenge for the design of CaaX -based inhibitors for FTase, which must bind FTase tightly, yet avoid interactions with the GGTase I enzyme in order to preserve the function of critical proteins that are C20 modified.

Although Ras and other small GTPases are the most well known prenyl proteins, several other classes of CaaX-containing proteins exist. One such group is the family of

Guanylate Binding Proteins (GBPs). GBPs are 65 kD proteins originally characterized based on their interferon-inducibility and ability to bind to guanine nucleotide affinity columns

(Cheng et al., 1991; Wynn et al., 1991). Six of the 8 GBPs identified thus far possess CaaX motifs (Asundi et al., 1994; Cheng et al., 1991; Han et al., 1998; Schwemmle et al., 1996;

Vestal et al., 1998; Wynn et al., 1991). Only one of the GBPs (human GBPl) has a CIS-type 38

CaaX (CTIS), and appears to be famesylated, as predicted (Nantais et al., 1996; Schwemmle and Staeheli, 1994). Other GBP family members have C20-type CaaX boxes. The CaaX sequence of mOBPl (CTIL) was predicted to be a good motif for isoprenoid attachment, based on the successful prenylation of 5 proteins with the same CTIL sequence (rat p67 GBP, murine

GBP2, G74, murine mRpgr protein, and a plant calmodulin (Kalman et al., 1995; Rodriguez-

Concepcion et al., 1999; Vestal et al., 1996; Vestal et al„ 1998; Yan et al., 1998). However, despite having the hallmarks thought to be necessary, even favorable, for prenylation, isoprenoid attachment to mGBPl proved difflcult to detect. Detailed examination of possible reasons for the meager incorporation of [^H]mevalonate suggested that the defect was not the result of problems in cellular isoprenoid metabolism or prenyl transferase activity, but arose from mGBPl itself.

Materials and methods

Construction of DNAs: The 1.8 kb mGBPl coding region in pRC/RSV (Wynn et al., 1991) was amplified by PGR and cloned directionally into the pcDNAS vector (InVitrogen) using

Not I and Xba I sites included on the PGR primers. Additional mutants were generated from this mGBPl gene using reverse primers that harbored the desired mutations. The GBPtRas chimera, Chim. was constructed using a primer complementary to the coding strand of the vector, with 24 mGBPl nucleotides plus 57 nucleotides corresponding to the Hnal 18 codons of K-Ras4B, plus stop codon and an Xba I site. Full length human GBPl was cloned into pcDNA3 after PGR amplification from the pHMG-SVpolyA vector (Cheng et al., 1991) using

5' and 3' primers containing Bam HI or Not I restriction sequences respectively. Sequences of the mutated regions in all DNAs were veriHed before use in transfections.

Cell Culmre. Transfection and IFN treatment: COS-1 cells were maintained in DMEM (Gibco) containing 10% calf serum (HyClone) supplemented with 2 mM L-glutamine, 1 mM sodium pyruvate, and penicillin/streptomycin. Purified plasmid DNA (1 -2 pg) was introduced into COS-1 cells using Lipofectamine (Gibco), and cells harvested 48 hours later. 39

RAW264.7 cells were maintained in RPMI media (Gibco) with similar additions but with 10% fetal bovine serum (HyClone). Endogenous mCBPl was induced by treatment of RAW264.7 cells with IFNy (300 U/ml) and LPS (10 M-g/ml) for 20 hr.

Metabolic labeling. Electrophoresis and Fluorographv: Transfected COS-1 cells were labeled overnight in medium containing 10% serum and 100 nCi/ml [3H]MVA (American

Radiolabeled Chemicals, Inc.) in the presence of compactin (a generous gift from D. Graves) at

SO |iM or as indicated in the Hgure legends. RAW264.7 cells were simuhaneously exposed to

IFN/LPS and labeled with 100 pCi/ml [^HJMVA or 50 pCi/ml [^HJgeranylgeraniol (American Radiolabeled Chemicals, Inc.) in the presence of 50 compactin. Cells were lysed directly in electrophoresis sample buffer and samples separated by SDS-PAGE. Proteins were transferred by electroblotting onto PVDF membrane (NEN Life Science Products), the membrane sprayed with En^Hance (DuPont/NEN) and the [3H]-labeled proteins detected by fluorographic exposure. After fluorography, blots were stripped of fluorographic enhancer by rinsing with methanol and TTBS (Tris-buffered saline containing 0.5% Tween20). GBPs were then detected directly by immunoblotting the same membrane used for fluoragraphy.

Tmrnunoblotting. Immunoprecipitation and Subcellular Fractionation: GBPs were detected using a rabbit polyclonal antibody to recombinant mGBPl (provided by D. Paulnock) with a biotinylated secondary antibody and alkaline phosphatase (Vector Labs). Endogenous H-Ras in

COS-1 cells was detected similarly using the H-Ras-specific mouse monoclonal antibody 146-

03E4 (Quality Biotech, Camden, NJ). Immunoprecipitates were isolated by solubilizing cells as described (Vestal et al., 1998) and incubating the clarified supernatant with anti-GBP-coated

Pansorbin (Calbiochem). Samples were washed and resuspended in electrophoresis sample buffer for analysis by SDS-PAGE. Membranes (PI00) were separated from cytosol (SI00) by centrifugation at 100,000 x g as described (Nantais et al., 1996). Proteins in both fractions were precipitated by the addition of four volumes of cold acetone, collected by centrifugation and resuspended in electrophoresis sample buffer. 40

Results

FNmnr.ENOUS MGBPI IS POORLY LABELED BY flHIMVA IN IFN-TREATED RAW264.7

CELLS

Our early studies on the effects of IFNy/LPS on protein prenylation in murine bone marrow-derived macrophages had identified a 65 kilodalton prenyl protein (p65) that was strongly induced by this cytokine (Vestal et al., 1995). Incorporation of [3H]mevalonate into p65 was sensitive to the famesyl transferase inhibitor BZA-5B, implying that p65 was farnesylated. Clear potential candidates for p65 were the 65 kD IFN-inducible murine GBPs.

However, both identiHed murine GBPs had C20-type CaaX motifs, and the mGBP2 protein appeared to be successfully C20 modified (Vestal et al., 1998). Therefore mGBPI was examined to learn if it would also be C2G modified or might instead be farnesylated.

Unexpectedly, when the endogenous mGBPI protein was induced by treatment of RAW264.7 cells with IFNy (Fig. 1), no prenyl protein was visible on the film at the correct molecular weight (Fig. 1). Just above the position of mGBPI two IFNy-inducible proteins were clearly

[3H]MVA-labeled. These minor proteins were also recognized by the polyclonal anti-GBP serum. This indicated that RAW264.7 cells were capable of attaching isoprenoids to IFNy- induced proteins, and suggested that some of these were potential GBP family members.

Surprisingly, although the immunoblot showed that mGBPI was abundantly expressed, the amount of isoprenoid it contained appeared to be below our level of detection.

Because [^HJMVA is converted into both [^HJfamesyl pyrophosphate and

[^HJgeranylgeranyl pyrophosphate ([^HJGGPP), the possibility was tested that the poor labeling of mGBPI resulted from difficulties in transport of [^HIMVA or production of pHJGGPP in RAW264.7 cells. Examination of the other proteins present in the radiolabeled cell lysates did not encourage this theory, as many 21-28 kD small GTPases of the Rho and

Rab families, most of which are geranylgeranylated (Reese and Maltese, 1991; Zhang and

Casey, 1996), were successfully labeled (Fig. 1). To more directly examine geranylgeranyl 41 utilization, cells were incubated with [3H]geranylgeraniol ([^HJGGOH), an alcoholic isoprenoid that specifically labels C20-modified proteins (Crick et al., 1994). The several proteins between 30-80 kD that had been labeled with [^HJMVA, including the two ~66 kD IFNy-induced proteins, did not incorporate label from [^HJgeranylgeraniol. This suggested, as had been reported for other cell lines, that these larger proteins contained a CIS isoprenoid

(James et al., 1994). The group of 21- 28 kD proteins was labeled effectively (Fig. 1), demonstrating that IFN-treated RAW264.7 cells metabolize [^HJGGOH to [^HJGGPP properly. However, once again, mGBPl labeling was below detectable levels. These results indicated that IFN treatment and the presence of mGBPl did not impair C20 isoprenoid incorporation into the appropriate proteins, and thus could not explain mGBP's poor labeling.

MGBPl ALSO DISPLAYS POOR ISOPRKNOID INCORPORATION IN COS-1 CELLS

Two other possible explanations for the poor labeling of mGBPl were that the poorly labeled, IFNy-induced protein was not actually mGBPl, or that the monocytic RAW264.7 cells specifically had difficulty with mGBPl modification. These possibilities were tested by transiently expressing a molecular clone of authentic mGBPl in COS-1 cells so as to produce an amount of mGBPl similar to that attained naturally through IFNy treatment. An endogenous prenyl protein that migrated just above mGBPl (as well as the labeling of other prenyl proteins in the cell lysates) showed no differences in incorporation between mock transfected cells or cells expressing mGBPl (Fig. 2, upper panel). The positive control human GBPl protein

(Nantais et al., 1996) was clearly labeled (arrowhead). However, isoprenoid incorporation into mGBPl was still below the level of detection. These results indicated that authentic mGBPl also encountered difflculties in labeling in another cell type. Thus the problem in labeling mGBPl was not restricted to monocytic or IFN-treated cells.

To allow us to examine mGBPl more clearly, immunoprecipitation was used to isolate mGBPl from the endogenous prenyl protein of COS-1 cells. With this technique, incorporation of low amounts of label into mGBPl were detected (Fig. 2, lower panel). 42

Using immunoprecipitation to isolate larger amounts of mGBPl from RAW264.7 cells also revealed [3H]MVA incorporation in IFNy-induced mGBPl after long film exposure times (data not shown). This ability to detect mGBPl [3H]MVA incorporation in an immunoprecipitate did not derive from a more favored interaction of the GBP antiserum with prenylated forms of the protein, as the serum could capture uniform amounts of proteins from lysates even when the proteins showed up to 8-fold differences in labeling (see below). Reciprocally, the poor labeling of mGBPl seen directly in cell lysates (Fig. 1) also indicated that immunoprecipitates had not selectively lost a prenylated form of mGBPl. The antiserum thus could recognize equally both prenylated and non-prenylated forms of mGBPl. Therefore neither native, cytokine-induced, nor artifically expressed mGBPl were totally devoid of isoprenoid, but each appeared to contain so little lipid that the small amounts present were difHcult to detect unless the protein was concentrated and purified by immunoprecipitation.

Finally, to determine if the problem with mGBPl labeling might be kinetic, from a slow equilibration of C20 pools or inefficient modiHcation by prenyl transferase, the amount of compactin utilized to inhibit cellular HMG-CoA reductase was decreased (see for example.

Fig. 4, upper) to avoid unintentional depletion of C20 isoprenoids (Rilling et al., 1993), and extended labeling periods (up to 52 h, data not shown) were used. Both approaches failed to improve mGBPl incorporation. These data indicated that mGBPl underwent neither rapid but short-lived, nor slow but eventual prenylation, but was, at all times, poorly labeled.

POOR LABELING OF MGBPl IS NOT DUE TO REMOVAL OF A PRENYLATED C-TERMINUS

All of the above experiments indicated that, despite the expectation that the mGBPl

CaaX motif should make the protein a good target, the difficulties underlying mGBPl's poor labeling did not involve isoprenoid metabolism or changes in prenyl transferase activity.

Therefore reasons for limited incorporation that might arise from special properties of the protein itself were considered. An initial hypothesis was that mGBPl processing might resemble that of prelamin A. The prelamin A protein undergoes famesylation, but subsequently 43 loses the isoprenoid when the C-terminal 18 amino acids are removed (Kilic et al., 1997). It was judged that if a C-terminal domain of similar size were removed from mGBPl the change in length should be detectable, since the 589 amino acid mGBPl and 592 amino acid huGBPl proteins can be distinguished on our gel system (see Fig. 2). When compactin was used to limit isoprenoid synthesis and force accumulation of precursors of prenylated proteins, the endogenous H-Ras protein in the COS-1 cells collected in its unprocessed form, but the mobility of the transfected mGBPl protein did not change (Fig. 3). The 2 forms of mGBPl with slightly different mobilities also persisted after exposure to compactin and both incorporated [^H]isoprenoid (see Figs. 4 and 5) al the same low levels, suggesting that neither protein was a non-prenylated precursor of the other. Presumably some other modiflcation, or internal initiation of translation, generates these two forms. Thus, compactin u-eatment failed to cause accumulation of any specific precursor form of mGBPl and also failed to detect any potential isoprenoid modification that might have been difficult to observe by radiolabeling techniques.

MGBP 1 WITH A C1 5-TYPE CAAX MOTIF IS MODIFIED WELL

A second possibility was that the bulk of newly synthesized mGBPl might be concealed in some location that was inaccessible to GGTase I and (because they share an a subunit) also the FTase. This possibility was tested by creating a chimeric mGBPl (desipated

Chim) with a signal for famesylation derived from K-Ras4B. [^HIMVA incorporation into this chimeric GBP:Ras protein showed a prominent increase above the native, wild type mGBPl

(mGBPlwt) (Fig. 4), showing that FTase had no difficulty gaining access to the Chim protein. Therefore mGBPl was not sequestered.

However, the loss of 18 residues and their replacement with a lysine-rich domain was a rather significant alteration of the mGBPl C-terminus. To address more precisely how well

FTase could modify mGBPl, a less drastic mutant of mGBPl was made, with only the final

"X" amino acid changed from leucine to serine. This mutant, designated CTIS, also 44 incorporated more label than mGBPlwt (Fig. 4). To judge if the increased labeling of the ens protein resulted simply from the change to FTase or might in addition result from an unintended improvement in presentation of the serine at the C-terminus to a prenyl (famesyl) transferase, an additional mutant was constructed. An alanine substitution in the CaaX "X" position (CTIA) was chosen to more closely mimic the leucine present in the natural CaaX motif. The mGBP-CTIA protein also showed improved labeling, to an extent similar to that of the mGBP-CTIS protein (data not shown). As this conservative change was unlikely to alter the C-terminal structure, it appeared that increased [3H]MVA incorporation resulted from the predicted change in the modifying prenyl transferase from GGTase I to FTase.

The size of isoprenoid attached to these proteins was veriHed using [^HJGG-OH labeling. The mGBPl protein incorporated small amounts of [3H]GG-0H (Fig. 4, lower panel), providing evidence that what little isoprenoid was incorporated into mGBPl was of the expected, C20 type. The Chim and the CTIS proteins were not labeled by the [3H]-C20 isoprenoid (Fig. 4, lower panel), indicating that the lipid detected on these constructs after

[3H]MVA labeling was famesyl, as intended by their CaaX motifs. These results implied that mGBPl was acceptable as a substrate for FTase as long as it had the appropriate CaaX motif

In addition, mGBPl was not the still-elusive p65 protein of bone marrow-derived macrophages.

C-TERMINUS OF MGBPl INTERFERES WITH C20 MODINCATION

At this point all the data pointed to the unexpected possibility that the problem with mGBPl prenylation was C20-specific. To determine if it was the C-terminus or a more distant part of mGBPl that caused this difHculty, a new, C20 version of Chim was constructed by remodeling the CIS-type K-Ras4B CVIM motif of Chim back to Cl'lL. This C20-Chim would thus be modified by [^H] C20 isoprenoid of the same specific activity as mGBPlwt. The only differences between mGBPlwt and C20-Chim were in the C-terminal 14 amino acids that mimicked those of K-Ras4B, adjacent to the C20-CaaX motif. As shown in Figure 5, 45

[3H]MVA labeling of C20-Chim was 2-3-fold higher than mGBPlwt. Therefore, mGBPl modification by C20 isoprenoid could be improved by inserting 14 K-Ras4B amino acids directly upstream of the CaaX motif. This result identiHed this region of the mGBPl C- terminus as part of a structure that impaired C20 modification.

Among these C-terminal residues of mGBPlwt were 3 consecutive prolines (Fig. 6).

To determine if these prolines hindered C20-modification of mGBPl, a new mutant was constructed in which these residues were changed to arginines, the residues found in these positions in hGBPl. This "P to R" mutant was labeled -3 fold better than mGBPlwt, to a level similar to that of the more extensively altered C2Q-Chim. Therefore, among the C-terminal residues of mGBPlwt, these prolines were responsible for at least a portion of the difficulty in

C20 modification. TsopRRNoro MODIFICATION OF MGBPI IS INCOMPLETE

The consistent inequalities in labeling observed among these various GBPs suggested that these proteins might not be fully prenylated in mammalian cells (Fig. 6A). Direct measurement of mGBP's isoprenoid content by mass spectrometry was not practicable, as methods have not yet been developed for purification of sufficient amounts of mGBPl from mammalian cells (where the impairment occurs), especially given the small portion of native protein that appeared to be prenylated. Therefore, the fraction of each protein that contained isoprenoid was determined by dividing the amounts of [^H] isoprenoid incorporated into the various proteins by the amount of each protein that was expressed. To increase the accuracy of this analysis, both [^H] and protein measurements for each protein were drawn from a single immunoprecipitate blotted onto a single membrane. The membrane was first exposed to film for detection of [^H], then developed with GBP antiserum for quantification of protein. As an additional control, comparisons of immunoblots from total cell lysates and immunoprecipitates of cells expressing the various GBPs showed that all forms, including those well or poorly labeled, were immunoprecipitated with equal efficiency by the antibody. 46

These calculations indicated that the ClS-Chim protein contained the most

[3H]isoprenoid (Fig. 6B). The [3H]:protein ratio of ClS-Chim was therefore set to 1 (as each protein molecule could contain no more than 1 isoprenoid, although it conceivably might contain less) and the ratios of the other GBP variants expressed relative to this value. This analysis quantised the previous visual results, and indicated that mGBPlwt contained only 13

± 2% as much [^HJisoprenoid as ClS-Chim (Fig. 6B), thus implying that greater than 85% of mGBPl was not prenylated. Although the C20-Chim and "P to R" proteins were labeled more strongly than mGBPl wt, modification of these proteins also appeared to be incomplete

(-30% of ClS-Chim). Thus all of the mGBP variants with C20-type CaaX motifs appeared to be poorly prenylated (Fig. 6B). Even the CIS-type CTIS protein was modified only -40% as well as the ClS-Chim. That the CTIS and ClS-Chim proteins had different levels of [3H]-C1S labeling further conHrmed that differences detected in incorporation of [^H] radioactivity reflected differences in amount of isoprenoid attached. Thus the level of isoprenoid modification of mGBPl that occurred within an intact cell could be manipulated over an 8-fold range by altering the last 18 residues. A more precise replacement of the 3 prolines with arginines could double the amount of C20-modined mGBPl, while simply switching to a CIS- modified form could triple levels of prenylated mGBPl.

Interestingly, hGBPl was also labeled less well than the ClS-Chim. These results suggested that modification of human GBPl, although efficient enough to allow detection, was also incomplete. Finally, a hGBPl with the C20 motif CTIL was labeled as poorly (12± S%) as mGBPl. The hGBP-C20 protein provided a second example where prenylation was far below stoichiometric amounts. The 3-fold difference in labeling of native hGBPl and hGBP-

CTIL illustrated again that in a pair of proteins in which all non-CaoX structures were identical,

CIS and C20 modification could be di^erent. 47

PRF.NYT.ATTQN OF GBPS DOES NOT AFFECT MRMBRANE ASSOCIATION

Because for some proteins isoprenoid modiHcation enhances membrane binding, mutant GBPs with wide variations in prenylation were examined to learn if this might lead to variable extents of membrane binding. The mGBPl wt, CTIS, "P to R", and huGBPl proteins were largely soluble (-70%, Fig. 7A), even though their extents of isoprenoid modification differed 3-fold, from 15% to 45%. Earlier work had already found 75% of (^HJMVA-labeled huGBPl in cytosolic fractions (Nantais et al., 1996), showing that the prenylated form of huGBPl was not restricted to membranes. Prenylation thus had little impact on membrane association of these GBPs.

However, C15-Chim exhibited a distinct distribution, with significantly greater association with the PlOO fraction (-60%). When compactin was used to deplete the amount of famesyl available for prenylation of C15-Chim (Fig. 7B) the portion of C15-Chim present in the P fraction decreased. The compactin sensitivity of C15-Chim membrane interaction provided additional evidence, independent of radioactive labeling, that indicated the C15-Chim contained significant amounts of isoprenoid.

Discussion

mGBPl is the Hrst example of a protein in which a seemingly adequate CaaX motif almost completely escapes modification. The RhoA GTPase has been shown to exist naturally as a mixture of C15- and C20-modified forms (Adamson et al., 1992), but the stoichiometry of isoprenoid attachment has been presumed to be nearly complete. Rather than existing as a fiilly prenylated protein modified by one or the other isoprenoid, mGBPl is one of the few documented cases of incomplete prenylation. The ability of mGBPl to evade prenylation is particularly unusual because it applies selectively to modification of the protein with C20 isoprenoid. Because the CllL CaaX sequence can be C20-modified in other proteins, it is clear that this negative control of isoprenoid modiHcation not only arises from regions of mGBPl 48 outside of its CaaX box, but must predominate over the otherwise acceptable CaaX motifs role.

PRFNYI.ATION OF MGBPl IS INCOMPLETE

Identifying mGBPl as a protein having a severe deficit in [^HJMVA labeling but with an outwardly acceptable CaaX motif was unexpected. Careful examination of cellular isoprenoid utilization determined that difficulty with production of [^HJisoprenoid or IFN- induced alteration of prenyl transferase activity was not the cause of the poor incorporation.

These results provide an important and necessary foundation for the precedent that difficulty in detecting incorporation of [3H] isoprenoid into a protein need not reflect experimental or cellular flaws, but may actually result from incomplete prenylation of the substrate protein.

From comparison of labeling of mGBPl and the C15-Chim protein, we calculate that only

15% of mGBPlwt molecules contain isoprenoid. This estimate is based on an assumption that the specific activity of the [^HJFPP and [^HJGGPP pools used to modify these proteins will be equal after >18 hours of labeling in minimal compactin. Even if C20 and CIS isoprenoids were to continue to harbor differences in specific activities, the much better labeling of mGBPl's

C20-counterpart, C20-Chim. indicates that, at best, only one third of mGBPlwt is likely to be modified. Therefore the bulk of mGBPl, within an intact cell, genuinely appears to lack isoprenoid. Even this simple estimate gives mGBPl the distinction of being the most poorly prenylated protein identiHed thus far (Famsworth et al., 1990; Page et al., 1990). However, there are recent reports of non-famesylated Ras proteins in cholesterol-deprived cardiac cells

(Gadbut et al., 1997) and insulin-starved 3T3-L1 adipocytes (Goalstone and Draznin, 1996).

Our work with mGBPl and human GBPl now suggests that serious deficits in prenylation can occur in other proteins and cell types, and with either class of transferase.

Native murine GBPl thus appears to exist persistently in the cell as a mixture of C20- modified and (the more predominant) non-modified forms. However, the amount of isoprenoid-modiHed mGBPl could be increased, by up to 8-fold, by replacing mGBPl's 18 49

C-terminal residues. Furthermore, mutant proteins with varying degrees of C20 and CI5 modiHcation could be produced by altering either more specific residues of the C-terminal domain or by changing the prenyl transferase responsible for modiHcation to FTase. Thus the extent of mGBP prenylation can be manipulated through relatively modest changes in the C- terminal domain. Mutation of C-terminal residues of other proteins may similarly allow mixed populations of prenyl and non-modified proteins to be produced within a living cell and isoprenoid-sensitive functions studied. Experiments are underway to determine if replacing the

C-terminus of K-Ras4B with that of mGBPl can be used to produce a mixture of lipidated and non-lipidated forms of an oncogenic Ras. This will allow study of responses that may be encountered if treatment with prenyl u-ansferase inhibitors is only partially effective.

MECHANISM LIMITING PRENYLATION OF MGBPl

The incomplete prenylation of mGBPl does not appear to result from Pre-Lamin A-like proteolysis or from sequestration of the protein from cellular prenyltransferases. Simply changing the CaaX motif of mGBPl to a form recognized by FTase significantly improved mGBPl modification. This indicates that the C-terminus of mGBPl is not likely to be obscured by another protein, as such masking would presumably impede interaction with either

FTase or GGTase I.

Occupation of the mGBPl CaaX cysteine by another modifying group remains a possibility, although we have not been able to detect incorporation of [^HJpalmitate into mGBPl (unpublished results). ADP-ribosylation of the mGBPl CaaX is a theoretical possibility, as pertussis toxin modifies the alpha subunits of heterotrimeric Gi and GQ proteins at a cysteine in the position analogous to the CaaX (West et al., 1985). However, in normal circumstances these alpha subunit cysteines are unmodified (Jones and Spiegel, 1990) and available to interact with heptahelical receptors (Blahos n et al., 1998; Yang et al., 1999).

Additional information on the physiological fiinction of mGBPl will be needed to form a 50 clearer picture of whether the CaaX cysteine of mGBPl might undergo an alternative modiHcation.

Notably, the mechanism that interferes with isoprenoid attachment to mGBPl appears to differ from that of the G alpha subunits. For the pertussis toxin-sensitive Gi alpha proteins, the glycine in the second position of the "pseudo-CaaX" sequence (CGLF) interferes with prenyl transferase interaction, making this sequence a very poor substrate (Jones and Spiegel,

1990). In contrast, the CTIL CaaX motif of mGBPl has been shown to be a good substrate in other proteins (Kalman et al., 1995; Rodriguez-Concepcion et al., 1999; Vestal et al., 1998;

Yan et al., 1998). Thus, the mechanism that negatively regulates mGBPl prenylation is sufficiently strong to limit prenylation of an acceptable CaaX motif.

THF DEFECT IN MGBPI PRENYLATION IS SELECTIVE FOR C20 MODIFICATION

The negative influence of C-terminal non-CaaX regions on mGBPl prenylation is intriguing because it appears to selectively impair the ability of GGTase I to modify the protein. This is most clearly seen with the CTIS mutant, which retains all residues of the wild- type save for the final residue that confers FTase recognition, and which is modified 3 times as well as native mGBPl. The major impediment to GGTase I interaction appears to be in the C- terminal region as replacement of 14 C-terminal residues (C20-Chim) or the somewhat unusual triplet of prolines (P to R) increases mGBPl's C20 modiHcation -3 fold. This C20-selective impairment in prenylation identifies mGBPl as an important model for study of substrate.GGTase I interactions. Using mGBPl as a platform, additional model structures that can produce this selective interference can be tested and help guide design of compounds to better prevent C20 modification of K-Ras4B. Including non-CaaX elements modeled on mGBPl may decrease a prenyl transferase inhibitor's affinity for GGTase I while retaining good FTase interaction.

It will be important to determine if more distant residues outside the C-terminal region also contribute to the difHculQr in C20-modification. Work with chimeric G/s of heterotrimeric 51

G-proteins has previously suggested that GGTase I recognizes protein sequences outside the

CaaX box (Kalman et al„ 1995). In vitro prenylation assays and further testing of deletion mutants of mGBPl in living cells will be necessary to clarify the location of these residues and whether they directly interfere with GGTase I interaction or bind another protein that selectively blocks enzyme access.

An explanation of why mGBPl might evade prenylation is as yet unclear. Two somewhat divergent forms of murine GBPs (mag-2 and mGBP3) lack CaaX motifs, and therefore will not be prenylated (Han et al., 1998; Wynn et al., 1991). However, very little is known about the cellular function of any of the GBP proteins. A recent report indicates that human GBPl may, like the more well studied interferon-inducible MxA GTPase proteins, produce antiviral effects (Anderson et al., 1999). Whether mGBPl will show any natural variation in its prenylation or stay largely unmodified remains to be studied. The variety of mGBPl proteins constructed here, expecially the C15-Chim protein with its gain in prenylation and membrane association, may be useful for studies on GBP function.

The exceptionally poor prenylation of mGBPl indicates that our understanding of the mechanisms which regulate interactions of prenyl transferases with their substrates is far from complete. Such information is particularly needed for the newly described prenyl transferases of protozoa (Yokoyama et al., 1998) or fungi (Omer and Gibbs, 1994), and for isoprenoid modified viral proteins such as Hepatitis Delta antigen (Glenn et al., 1998), where differences from mammalian enzymes and substrates may be exploited in the treatment of infectious diseases.

Acknowledgements; We thank Dave Lucas and Donna Paulnock for the original mGBPl plasmid and polyclonal anti-GBPl antiserum and also for helpful suggestions; Peter Staeheli for the huGBPl plasmid, W. Robert Bishop and Deborah Vestal for comments on the manuscript, Molly Foster for technical assistance and Michelle Booden for technical advice and 52 assistance. These studies were funded by NSF Research Training Group award DBI960-2259 to JTS, PHS award CAS189, and the Roy J. Carver Charitable Trust.

References

Adamson, P., Marshall, C. J., Hall, A., and Tilbrook, P. A. (1992). Post-translational

modifications of p21rho proteins. J. Biol. Chem. 267, 20033-20038.

Anderson, S. L., Carton, J. M., Lou, J., Xing, L., and Rubin, B. Y. (1999). Interferon-

Induced Guanylate Binding Protein-1 (GBP-1) mediates an antiviral effect against vesicular

stomatitis virus and encephalomyocarditis virus. Virology 256,8-14.

Blahos n, J., Mary, S., Perroy, J., de Colle, C., Brabet, I., Bockaert, J., and Pin, J.-P.

(1998). Extreme C terminus of G protein a-subunit contains a site that discriminates

between Gl-coupled metabotropic glutamate receptors. J. Biol. Chem. 273, 25765-25769.

Buss, J. E., and Marsters, J. C., Jr. (1995). Famesyl transferase inhibitors: the successes and

surprises of a new class of potential cancer chemotherapeutics. Chem. Biol. 2,787-791.

Casey, P. J. (1995). Protein lipidation in cell signaling. Science 268,221-225.

Cheng, Y.-S. E., Patterson, C. E., and Staeheli, P. (1991). Interferon-induced guanylate-

binding proteins lack an N(T)KXD consensus motif and bind GMP in addition to GDP and

GTP. Mol. Cell. Biol. 11, 4717-4725.

Cox, A. D., and Der, C. J. (1997). Famesyltransferase inhibitors and cancer treatment:

targeting simply Ras? BBA Rev. Cancer 1333, F51-F71.

Crick, D. C., Waechter, C. J., and Andres, D. A. (1994). Utilization of geranylgeraniol for

protein isoprenylation in C6 glial cells. Biochem. Biophys, Res. Comm. 205,955-961.

Famsworth, C. C., Casey, P. J., Howald, W. N., Glomset, J. A., and Gelb, M. H. (1990).

Structural characterization of prenyl groups attached to proteins. METHODS: A companion

to Methods in Enzymology 1,231-240. 53

Gadbut, A. P., Wu, L., Tang, D., Papageorge, A., Watson, J. A., and Galper, J. B. (1997).

Induction of the cholesterol metabolic pathway regulates the famesylation of RAS in

embryonic chick heart cells: a new role for Ras in regulating the expression of muscarinic

receptors and G proteins. EMBO J. 16, 7250-7260.

Gelb, M. H., Scholten, J. D., and Sebolt-Leopold, J. S. (1999). Protein prenylation; from

discovery to prospects for cancer threatment. Curr. Opin. Chem. Biol. 2,40-48.

Glenn, J. S., Marsters, J., J.C., and Greenberg, H. B. (1998). Use of a prenylation inhibitor

as a novel antiviral agent. J. Virol. 72,9303-9306.

Goalstone, M. L., and Draznin, B. (1996). Effect of insulin on famesyltransferase activity in

3T3-L1 adipocytes. J. Biol. Chem. 271,27585-27589.

Han, B. H., Park, D. J., Lim, R. W., Im, J. H., and Kim, H. D. (1998). Cloning,

expression, and characterization of a novel guanylate-binding protein, GBP3 in murine

erythroid progenitor cells. BBA 1384,373-386.

James, G., Goldstein, J. L., and Brown, M. S. (1996). Resistance of K-RasBV12 proteins to

famesyltransferase inhibitors in Rati cells. Proc. Natl. Acad. Sci. U.S.A. 93,4454-4458.

James, G. L., Goldstein, J. L., Pathak, R. K., Anderson, R. G. W., and Brown, M. S.

(1994). PxF, a prenylated protein of peroxisomes. J. Biol. Chem. 269, 14182-14190.

Jones, T. L. Z., and Spiegel, A. M. (1990). Isoprenylation of an inhibitory G protein alpha

subunit occurs only upon mutagenesis of the carboxyl terminus. J. Biol. Chem. 265,

19389-19392.

Kalman, V. K., Erdman, R. A., Maltese, W. A., and Robishaw, J. D. (1995). Regions

outside of the CAAX motif influence the specificity of prenylation of G protein gamma

subunits. J. Biol. Chem. 270, 14835-14841.

Kilic, F., Dalton, M. B., Burrell, S. K., Mayer, J. P., Patterson, S. D., and Sinensky, M.

(1997). In Vitro assay and characterization of the famesylation-dependent prelamin A

endoprotease. J. Biol. Chem. 272, 5298-5304. 54

Kinsella, B. T., Erdman, R. A., and Maltese, W. A. (1991). Posttranslational modification of

Ha-ras p21 by famesyl versus geranylgeranyl isoprenoids is determined by the COOH-

terminal amino acid. Proc. Natl. Acad. Sci. USA 88, 8934-8938.

Kohl, N. E., Omer, C. A., Conner, M. W., Anthony, N. J., Davide. J. P., deSolms, S. J.,

Giuliani, E. A., Gomez, R. P., Graham, S. L., Hamilton, K., Hant, L. K., Hartman, G.

D., Koblan, K. S., Krai, A. M., Miller, P. J., Mosser, S. D., O'Neill, T. J., Rands, E.,

Schaber, M. D., Gibbs, J. B., and Oliff, A. (1995). Inhibition of famesyltransferase

induces regression of mammary and salivary carcinomas in ras transgenic mice. Nature

Med. 7, 792-797.

Liu, M., Bryant, M. S., Chen, J., Lee, S., Yaremko, B., Lipari, P., Malkowski, M., al., e.,

and Bishop, W. R. (1998). Antitumor activity of SCH 66336, an orally bioavailable

tricyclic inhibitor of famesyl protein transferase, in human tumor xenograft models and

Wap-ras transgenic mice. Cancer Research 58,4947-4956.

Lowy, D. R., and Willumsen, B. M. (1993). Function and regulation of Ras. Ann. Rev.

Biochem. 62, 851-891.

Moores, S. 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. (1991). Sequence dependence of

protein isoprenylation. J. Biol. Chem. 266, 14603-14610.

Myung, C.-S., Yasuda, H., Liu, W. W., Harden, T. K., and Garrison, J. C. (1999). Role of isoprenoid lipids on the heterotrimeric G protein y subunit in determining effector

activation. J. Biol. Chem. 274, 16595-16603.

Nantais, D. E., Schwemmle, M., Stickney, J. T., Vestal, D. J., and Buss, J. E. (1996).

Prenylation of an interferony-induced GTP-binding protein: the human guanylate binding

protein, huGBPl. J. Leukoc. Biol. 60, 423-431.

Omer, C. A., and Gibbs, J. B. (1994). Protein prenylation in eukaryotic microorganisms:

genetics, biology and biochemistry. Mol. Microbiol. II, 219-225. 55

Page, M. J., Aitken, A., Cooper, D. J., Magee, A. I., and Lowe, P. N. (1990). Use of the

insect/baculovirus expression system to produce lipid-modified ras proteins for

electrospray mass spectrometric studies. METHODS: A Companion to Methods in

Enzymology 7, 221-230.

Park, H.-W., Boduluri, S. R., Moomaw, J. F., Casey, P. J., and Beese, L. S. (1997).

Crystal structure of protein famesyltransferase at 2.25 angstrom resolution. Science 275,

1800-1804.

Reese, J. H., and Maltese, W. A. (1991). Post-translational modiflcation of proteins by 15-

carbon and 20-carbon isoprenoids in three mammalian cell lines. Mol. Cell. Biochem. 104,

109-116.

Rilling, H. C., Bruenger, E., Leining, L. M., Buss, J. E., and Epstein, W. W. (1993).

Differential prenylation of proteins as a function of mevalonate concentration in CHO cells.

Arch Biochem Biophys 301, 210-215.

Rodriguez-Concepcion, M., Yalovsky, S., Zik, M., Fromm, H., and Gruissem, W. (1999).

The prenylation status of a novel plant calmodulin directs plasma membrane or nuclear

localization of the protein. EMBO J. 18,1996-2007.

Rowell, C. A., Kowalczyk, J. J., Lewis, M. D., and Garcia, A. M. (1997). Direct

demonstration of geranylgeranylation and famesylation of Ki-Ras in vivo. J. Biol. Chem.

272, 14093-14097.

Schwemmle, M., and Staeheli, P. (1994). The interferon-induced 67-kDa guanylate-binding

protein (hGBPl) is a GTPase that converts GTP to GMP. J. Biol. Chem. 269, 11299-

11305.

Seabra, M. C. (1998). Membrane association and targeting of prenylated Ras-Iike GTPases.

Cell. Signal. 10, 167-172. 56

Strickland, C. L., Windsor, W. T., Syto, R., Wang, L., Bond, R., Wu, Z., Schwartz, J., Le,

H. v., Beese, L. S., and Weber, P. C. (1998). Crystal structure of famesyl protein

transferase coupled with a CaaX peptide and famesyl diphosphate analogue. Biochemistry

57, I6601-1661I.

Sun, J., Qian, Y., Hamilton, A. D., and Sebti, S. M. (1998). Both faroesyltansferase and

geranylgeranlytransferase I inhibitors are required for inhibition of oncogenic K-Ras

prenylation but each alone is sufficient to suppress human tumor growth in nude mouse

xenografts. Oncogene 16,1467-1473.

Vestal, D. J., Buss, J. E., Kelner, G. S., Maciejewski, D„ Asundi, V. K., and Maki, R. A.

(1996). Rat p67 GBP is induced by interferon-gamma and isoprenoid-modified in

macrophages. Biochem. Biophys. Res. Commun. 224, 528-534.

Vestal, D. J., Buss, J. E., McKercher, S. R., Jenkins, N. A., Copeland, K. G., Kelner, G. S., Asundi, V. K., and Maki, R. A. (1998). Murine GBP-2: A new IFN-y-induced

member of the GBP family of GTPases from macrophages. J. Interferon and Cytokine

Res. 18, 977-985.

Vestal, D. J., Maki, R. A., and Buss, J. E. (1995). Induction of a prenylated 65-kd protein in

macrophages by interferon or lipopolysaccharide. J. Leukoc. Biol. 58,607-615.

West, J., R.E., Moss, J., Vaughan, M., Liu, T., and Liu., T.-H. (1985). Pertussis toxin-

catalyzed ADP-ribosylation of transducin. J. Biol. Chem. 260,14428-14430.

Wynn, T. A., Nicolet, C. M., and Paulnock, D. M. (1991). Identification and characterization

of a new gene family induced during macrophage activation. J. Immunol. 147,4384-4392.

Yan, D., Swain, P. K., Breuer, D., Tucker, R. M., Wu, W., Fujita, R., Rehemtulla, A.,

Burke, D., and Swaroop, A. (1998). Biochemical characterization and subcellular

localization of the mouse retinitis pigmentosa GTPase regulator (mRpgr). J. Biol. Chem.

273, 19656-19663. 57

Yang, C.-S., Skiba, N. P., Mazzoni, M. R., and Hamm, H. E. (1999). Conformational changes at the carboxyl terminus of Ga occur during G protein activation. J. Biol. Chem.

274, 2319-23^5.

Yokoyama, K., Trobridge, P., Buckner, F. S., Van Voorhis, W. C., Stuart, K. D., and Gelb,

M. H. (1998). Protein famesyltransferase from trypanosoma brucei. J. Biol. Chem. 275,

26497-26505.

Zhang, F. L., and Casey, P. J. (1996). Protein prenylation: molecular mechanisms and

functional consequences. Annu. Rev. Biochem. 65,241-269.

Zhang, F. L., Kirschmeier, P., Carr, D., James, L., Bond, R. W., Wang, L., Patton, R.,

Windsor, W. T., Syto, R., Zhang, R., and Bishop, W. R. (1997). Characterization of Ha-

Ras, N-Ras, Ki-Ras4A, and Ki-Ras4B as in vitro substrates for famesyl protein

transferase and geranylgeranyl protein transferase type I. J. Biol. Chem. 272, 10232-

10239.

Figure Legends Figure 1. Incorporation of [^HJisoprenoid into mGBPl cannot be detected in EFNy treated

RAW264.7 cells. RAW264.7 cells were labeled for 18 hr with either [^HIMVA (lanes I and

2) or [3H]GG-0H (lanes 3 and 4) in the absence (lanes 1 and 3) or presence (lanes 2 and 4) of

IFNy and LPS. Total cell lysates were displayed by SDS-PAGE and mGBPl detected by immunoblotting (Blot). Film is the fluorogram from the immunoblot shown, after 14 days exposure. With either labeled precursor, no radiolabeled band of the correct size for the induced mGBPl protein (dotted line) was detected after IFNy/LPS. However, two new

[3H]MVA-labeled proteins (solid line) of slightly slower mobility appeared after IFNy/LPS treatment. Positions of molecular weight markers are shown on the side.

Figure 2. Poor incorporation of [^H]MVA occurs in authentic mOBPl expressed in COS-1 cells. {Upper) COS-1 cells were labeled for 18.5 hr with pH]MVA in the presence of 25 pM compactin, starting at 30 hr. after transfection with no DNA (U), or DNAs for mGBPl (mG) 58 or huGBPl (hG). Cell lysates were resolved by SDS-PAGE and radiolabeled proteins detected by fluorographic exposure for 13 days. (Lower) After transfection and labeling as above, mGBPl was isolated by immunoprecipitation (IP) and separated by SDS-PAGE. Incorporation of radiolabel into mGBPl was detected after fluorographic exposure for 21 days. HC denotes the heavy chain of the anti-GBP antibody.

Figure 3. Compactin treatment does not alter mGBPl mobility. COS-1 cells were transfected with DNA encoding mGBPl with compactin (50 pM) added to the media 5 hr after transfection and remaining until samples were prepared 42 hr post-transfection. Cell lysates were separated by SDS-PAGE and endogenous H-Ras or the transfected mGBPl detected by immunoblotting. Lane 1 is from cells transfected with empty vector; lanes 2 and 3 are from mGBPl-transfected cells treated either in the absence or presence of compactin. Asterisk (*) denotes unprocessed H-Ras; the arrowhead shows fiilly processed, famesylated H-Ras.

Figure 4. [^HJMVA labeling improves in mGBPl proteins with CI5 type CaaX boxes.

COS-1 cells were transfected with empty vector (V) or DNAs encoding wild type mGBPl

(wt), huGBPl (hG), or mGBPl variants with C15-type CaaX motifs (CTIS and Chim). Cells were labeled with either 1(X) |iCi/ml [3H]MVA plus 10 pM compactin or 50 MCi/ml [3H]GG-

OH and no compactin, and immunoprecipitates (upper) or cell lysates (lower) prepared and analyzed by SDS-PAGE. Incorporation of radiolabel was detected by fluorographic exposure for 22 days (pHJMVA) or 6 1/2 months ([^HJGG-OH), then GBPs on the membranes were visualized by immunoblotting.

Figure 5. The C-terminus of mGBPl selectively hinders C20 modification. COS-1 cells were transfected with vector DNA or DNAs for various GBPs, labeled with [^HJMVA, immunoprecipitates formed and separated by SDS-PAGE. After fluorographic exposure for 22 days, GBPs present in the immunoprecipitates were visualized by immunoblotting.

Figure 6. C-terminal residues of GBPs and [^H]MVA incorporation relative to C15-Chim.

(A) C-terminal amino acids of various GBP constructs are aligned. Changes from wild type 59 mGBPl or hGBPl are indicated in Bold. (B) Using the fluorogram and immunoblot from the same membrane, [3H]MVA incorporation and protein amounts of each immunoprecipitated

GBP were quantified by scanning and values of [^HlMVA-derived label corrected for variations in protein recovery. The amount of [^H] for each GBP was then expressed relative to C15-Chim. Values shown are averages of these relative amounts, ± S.E.M. The number of independent experiments is indicated by n. One-way analysis of variance indicated that all GBP proteins incorporated significantly greater amounts of [^H] than mGBPl (p< 0.05), with the exception of the C20-modified huGBP-CTIL (hG-C20), which was not significantly different than mGBPl.

Figure 7. Membrane association of GBPs which differ in prenylation state. (A) COS-1 cells were transfected with DNAs for the indicated proteins and 48 hr later separated into cytosolic

(S) and membrane (P) fractions. Samples were separated by SDS-PAGE and GBPs detected by immunoblotting. (B) COS-1 cells were transfected with DNAs encoding mGBPlwt or

C15-Chim. Treatment with 50 compactin was started 5 hr later. After 48 hr, lysates were separated into cytosolic (S) and membrane (P) fractions, separated by SDS-PAGE, and GBPs detected by immunoblotting. 60

MVA GG-OH

83 mQBP1

m. 19 12 3 4 12 3 4

Figure 1. 61

Blot Film U mG hG U mG hG

Lysates hGBP1

U mG U mG mGBPI

IP HC

Figure 2. 62

+ COMPACTIN

mGBP1

^ H-Ras 1 2 3

Figure 3. 63

V wt ens Chim hG [^HJMVA mmm^l Film

Blot

V wt CTIS Chim

["'HIGG-OH :

BIM

Figure 4. 64

V wt ens Chitn C20 hG RoR

Film

Blot

Figure 5. 65

Predicted mGBP-l C-terminal Residues Isoprenoid wt LQLRQEIEKIKNMPPPRSCTIL C20 CTIS LQLRQEIEKIKNMPPPRSCTIS C15 ClS-Chim LQLRSXDGKKKKKKSKTKCVIM C15 C20-Chim LQLRSKDOKKKKKKSKTKCTIL C20 P to R LQLRQEIEKIKNMRRRRSCTIL C20

hu-wt IMKNEIQDLQTKMRRRKACTIS CI 5 hu-CTIL IMKNEIQDLQTKMRRRKACTIL C20

wt CTIS C15-Ch C20-Ch RoR hGBP hG.C2Q Mean .13 .41 1.00 .34 .27 .33 .12 SEM .02 .05 .11 .06 .07 .05 n 8 8 8 4 6 8 6

Figure 6. 66

Vector i wt CTIS CHIM C20 PtoR S P S P S P S P S P S P Blot

wt wt CHIM CHIM COMPACTIN SPSP SP SP

Blot

Figure 7. 67

CHAPTER 3. INHIBITION OF C20 MODIFICATION OF

MGBPl'S CAAX MOTIF IS MEDIATED BY MGBPl C- AND

N-TERMINAL PROTEIN ELEMENTS

A paper to be submined to the Journal of Biological Chemistry

John T. Stickney and Janice E. Buss

Abstract

Covalent attachment of isoprenoid lipids occurs on cysteine residues of C-terminal CaaX motifs of proteins. The thioether linkage formed between the isoprenoid and cysteine sulfhydryl is extremely stable and essentially irreversible in vivo. Thus, for isoprenoid attachment to be regulated, this control must occur prior to lipid addition. At the moment, there is little evidence that the enzymatic activity of prenyl transferases can be modulated, suggesting that control of prenylation would have to involve protein substrate preference and/or accessibility. In our efforts to better characterize how substrates are recognized by FTase and

GGTase I, murine Guanylate Binding Protein-1 (mCBPl) was identifled as a protein that was very poorly prenylated in intact cells. Our initial studies showed that impairment occurred in mutant proteins with CIS-type CaaX motifs but was especially severe for C20 modification of the native CaaX motif, and also implicated 18 amino acids within the C-terminal region as part of a structure of mOBPl that limited prenylation. However, in the present studies, the C- terminal 18 amino acids of mOBPl were shown to function well when appended to the K-

Ras4B protein, while expression of deletion mutants in COSl cells showed that proteins retaining as little as only the 300 N-terminal amino acids of mOBPl remained poorly prenylated. These results implicated both the extreme C-terminal structure and the N-terminal

GTP-binding region of mGBPl as contributors to a prenylation inhibitory domain. Additional 68 in vitro studies showed that mGBP's C20-specific defect in prenylation could be replicated in reticulocyte lysates, thus verifying that full-length mGBPl is intrinsically a poor substrate for

GGTase I. These studies provide the first example of a protein in which the native structure negatively regulates prenylation. This regulation appears to function through (selectively) limiting access of the CaaX motif to GGTase I by a combination of C-terminal and N-terminal mGBPl structures.

Introduction

Isoprenoid modification of the Ras family of proteins is critical for their biological activity. These binary molecular switches function normally to coordinate signals from numerous growth factor and signal transduction cascades to downstream effector proteins (for review, see (1)). However, activating mutations occur quite frequently within Ras genes; in fact, 30% of all cancers harbor a constitutively active, oncogenic Ras allele (2). Importantly, both normal and oncogenic Ras molecules rely on isoprenoid-mediated plasma membrane association to perform their biological activities (3,4). Thus, preventing prenylation of Ras proteins holds promise as a new class of cancer chemotherapy.

Proteins that possess a CaaX motif as their final four amino acids are often substrates for prenylation (for a review of prenylation see (5)). Isoprenoid linkage occurs through the formation of a thioether bond between the isoprenoid and the CaaX cysteine (C of CaaX).

Two types of isoprenoids (consisting of IS or 20 carbons) are attached to prenyiated proteins.

The fmal amino acid of the CaaX motif (X) is thought to be the major determinant for which isoprenoid will be used. If X is Leu or Phe, then the CaaX cysteine should be modified with the 20 carbon isoprenoid (geranylgeranyl) by geranylgeranyl transferase I (GGTase I). If the final amino acid is Met or Ser (as is found in Ras proteins) or Gin or Ala, then famesyl transferase (FTase) will attach the IS carbon isoprenoid famesyl. Of the total number of prenyiated proteins found in mammalian cells, only 20% are famesylated (6). Therefore, 69 blocking famesylation in vivo is hoped to provide inhibition of Ras activity while having low toxicity due to the small number of other prenylated cellular proteins that would be affected.

To this end, many inhibitors of famesyl transferase have been developed, and most appear to be very specific inhibitors of this enzyme in vitro and in vivo (7-10). However, in vivo studies show that not all Ras isoforms are equally sensitive to the action of FTIs (11,12).

In the presence of FTIs, both of the K-Ras isoforms (4A and 4B) as well as the N-Ras protein can be alternatively modified with the 20 carbon geranylgeranyl by GGTase I (12,13). This presents a significant problem in the use of FTIs in treating human tumors as the most commonly mutated forms of Ras in human cancers are K-Ras and N-Ras (14).

There is currently only a poor understanding of the mechanisms through which GGTase I can recognize and prenylate inappropriate substrates such as K-Ras. In vitro studies using peptide substrates and in vivo work using chimeric Ras proteins have shown that GGTase I can modify certain CIS-type CaaX motifs that terminate in Met and that possess multiple basic amino acids (a polybasic domain) N-terminal to the CaaX motif (12,15). Additional work using chimeric heterotrimeric G-protein y subunits has suggested that GGTase I is also sensitive to more distal N-terminal regions of its substrates (16). All of these studies have focused on GGTase I's ability to recognize CIS-type CaaX substrates that can be successfully modified by FTase. Information regarding proteins that should be prenylated by GGTase I but aren't would be extremely helpful in designing inhibitors to prevent unwanted geranylgeranylation of substrates such as K-Ras in cancer cells. We have recently identiHed murine Guanylate Binding Protein 1 (mGBPl) as a protein in which geranylgeranylation of its

CaaX in vivo is extremely poor, and suggested that this difficulty might arise from a GGTase- specific defect in mGBPl's prenylation.

The Guanylate Binding Proteins (GBPs) are a family of proteins that were discovered based on their interferon-inducibility and afHnity for guanine nucleotides (17,18). Little is known regarding the physiological function of the GBPs, but they are among the RNAs that 70 are most highly induced by interferon stimulation (17). Recently, the 3-dimensional structure of human GBPl was solved, and the authors of that report suggest that the GBPs may be distantly related to the dynamin family of GTPases (19). Other work suggests human GBPl may also exhibit antiviral activity (20).

Six of the eight identified members of the GBP family possess CaaX motifs (21-24), and as such, should be targets for prenylation. Human GBPl, rat p67, and murine GBP2 all appear to be modified as predicted by their CaaX motifs (23-25). However, in our initial studies of the prenylation of murine GBPl, we discovered that unlike other GBP family members, mGBPl failed to be effectively prenylated in vivo, even though it had the same

CaaX motif found to be suitable for prenylation of both rat p67 and murine GBP2 (Chapter 2).

No anomalies in cellular metabolism or utilization of isoprenoids, or alterations in cellular protein prenylation, were found that could explain the observed absence of modification of mGBPl. Analysis of mGBPl mutants revealed that residues within the C-terminus played a role in limiting prenylation in vivo. Replacement of the fmal 18 amino acids of mGBPl with the final 18 from K-Ras4B improved [^HJMVA incorporation 8-fold in COS-1 cells.

Additionally, a single point mutation in the Hnal amino acid of the protein (Leu to Ser) that changed the prenyl transferase specificity from GGTase I to FTase also greatly improved prenylation of mGBPl. These results suggested that the C-terminus of mGBPl played some role in preventing prenylation of the CaaX motif in vivo, and that the deficiency was more severe with regard to modification by GGTase I.

In order to better understand the mechanism behind mGBPl's GGTase I-specific defect in prenylation, we have studied the prenylation of this protein in vitro. The work reported here demonstrates that the C20-specific defect in mGBPl prenylation seen in vivo extends to in vitro conditions as well. Although previous studies showed the C-terminus of mGBPl to play a role in limiting C20 modification, analysis of deletion mutants presented in this paper indicates that the mechanism inhibiting geranylgeranylation of the mGBPl CaaX also includes 71 contributions from the N-tenninal GTP-binding domain. This suggests the possibility that prenylation of mGBPl may be a GTP-regulated event in vivo.

Materials and Methods

Construction of DNAs: Hemagglutinin (HA) epitope-tagged K-Ras4B constructs were created by PCR amplification of the 0.6 kb coding region of human K-Ras4B followed by cloning into the eukaryotic expression vector pcDNA3 (Invitrogen). PCR primer pairs containing Bam I and Xho I sites for directional cloning were used to create constructs possessing an N-terminal

HA epitope (protein sequence MYPYDVPDYA) and the indicated C-terminus. HA-tagged

GBPs were created in a similar fashion except that Not I and Xho I were used for cloning and the original initiator methionines in the GBP proteins were replaced by Leu. For DEL290/571, mGBPlwt DNA was digested with Tai I to remove nucleotides 871-1713, leaving in-frame Tai

I ends that were re-ligated to create the internal deletion. For bacterial expression, all constructs were subcloned in-frame (using the restriction sites listed above) from pcDNAS into the pPROEX prokaryotic vector system (GIBCO BRL) which introduces 6 histidine residues at the amino-terminus of the protein. Sequences of the mutated regions of DNA constructs in pcDNAB were verified before use in transfections or subcloning into bacterial expression vectors.

Cell Culmre and Transfection: COS-1 cells were maintained in DMEM (Gibco) containing 10% calf serum (HyClone) supplemented with 2 mM L-glutamine, 1 mM sodium pyruvate, and penicillin/streptomycin. Purified plasmid DNA (1-2 ^g) was introduced into COS-1 cells using Lipofectamine (Gibco) according to the manufacturer's protocol, and cells harvested 48 hours later.

In Vivo r3H1MVA labeling and Immunoprecipitation: Transfected COS-1 cells were labeled overnight in medium containing 10% serum and [^HJMVA (American Radiolabeled Chemicals,

Inc.) at 67 ^iCi/ml (for K-Ras4B proteins) or 200 nCi/ml (for HA-GBP deletions) in the presence of 10 pM compactin (a generous gift from D. Graves). Cells were either lysed 72 directly in electrophoresis sample buffer, or proteins immunoprecipitated, and samples separated by SDS-PAGE. Immunoprecipitates were isolated by solubilizing cells as described

(24) and incubating the clarified supernatant with anti-GBP or anti-HA-coated Pansorbin

(Calbiochem). Samples were washed and resuspended in electrophoresis sample buffer for analysis by SDS-PAGE.

Production of Recombinant Proteins: 100 ml cultures of transformed bacteria were grown at

'iTC in LB containing ampicillin (50 |ig/ml) to a density of OD600=0.6. IPTG was then added to 0.6 mM and cultures were grown an additional 4 hours then collected by centrifiigation and stored at -70°C. For lysis, pellets were thawed, resuspended in 1ml of storage buffer containing protease inhibitors (20 mM Tris, pH 7.7, 1 mM DTT, 3 mM MgCh, 1% aprotinin,

1 mM Pefabloc, 1 pg/ml leupeptin, and 1 ^ pepstatin) and sonicated 4 times at 4W for 30 seconds, with cooling on ice. Insoluble material was removed by centrifiigation at 25,000 rpm at 4°C and the supernatant containing the soluble proteins was stored in aliquots at -1Q°C.

Protein concentration was estimated by comparison of Coomassie Blue staining of the soluble

GBPs to that of BSA standards.

In Vitro T^^SIMethionine and T^HIMVA labeling: For in vitro transcription and translation of

HA-KRas4B or mGBPI constructs, pcDNA3 vectors containing these genes were transcribed and translated in vitro using TNT T7 Quick rabbit reticulocyte lysate (Promega) per the manufacturer's protocol. Protein synthesis was detected by adding 10 pCi Redivue L-[3^S] methionine (Amersham) to the assay. For [3H]MVA labeling of in vitro transcribed/translated proteins, or of recombinant His-tagged proteins, 35 nCi [^HIMVA per reaction was added to individual assay tubes and dried as described (26) before addition of the reticulocyte mix containing DNA or recombinant protein. Reactions were allowed to proceed for 2 hours at

30°C and were stopped by the addition of an equal volume of 2X SDS-PAGE sample buffer and 3 min. of boiling. 73

Fliinrngraphv and Immunoblotting: After SDS-PAGE, radiolabeled proteins were transferred by electroblotting onto PVDF membrane (NEN Life Science Products), the membrane sprayed with En^Hance (DuPont/NEN), and [^H] or [35S]-labeled proteins detected by fluorographic exposure to film. After fluorography, blots were stripped of fluorographic enhancer by rinsing with methanol and TTBS (Tris-buffered saline containing 0.5% Tween20). Proteins were then detected directly by immunoblotting the same membrane used for fluorography. GBPs were detected using a rabbit polyclonal antibody to recombinant mGBPl (provided by D. Paulnock) with a biotinylated secondary antibody and alkaline phosphatase (Vector Labs). Proteins possessing the HA epitope were detected using the hemagglutinin (HA)-specific mouse monoclonal antibody HA. 11 (Covance/BAbCO) with a horseradish peroxidase-conjugated secondary antibody (Pierce) and enhanced chemiluminescence (Pierce).

Scanning and Quantification of Bands: Fluorograms and immunoblots were scanned, and band intensities quantified using the ImageQuant (Molecular Dynamics) program. Stoichiometry of prenylation was determined as the ratio of ([^HJMVA incorporated):(protein amount). Protein levels were based on immunoblot analysis or [35S] incorporation (normalized for the number of methionines). Potential differences in the specific activity of the C20 and CIS isoprenoid pools in [3H]MVA labeling experiments were not considered.

Subcellular fractionation and TX-114 Phase Separation: Membranes (PI00) were separated from cytosol (S100) by centrifugation at 100,000 x g as described (25). After centrifugation, the SI00 supernatant was removed to a separate tube and the PI00 pellet resuspended in a volume of buffer equal to the S100. For TX-114 extraction, Triton X-114 (Sigma) was added to all fractions and phase separation performed by warming at 37°C for 2 min. and cenuifiigation at room temperature for 2 min. as described (27). Proteins in all fractions were precipitated by the addition of four volumes of cold acetone, collected by centrifugation, then resuspended and boiled in electrophoresis sample buffer. 74

GMP agarose: GMP agarose affinity experiments were performed essentially as described

(21). 30 pi aliquots of GMP agarose slurry (Sigma) were added to 1.5 ml microcentrifuge tubes and were equilibrated on ice for IS min. with 1 ml of binding buffer (20 mM Tris, pH

7.2, 150 mM NaCI, 5 mM MgCh, 0.1% Triton X-100). The slurry was sedimented and supernatant buffer was aspirated. GBP-containing cell lysates were added to the agarose bed and the final volume brought to 100 pi. After 30 min. on ice, the agarose was again sedimented, then washed 3 times with ice-cold binding buffer. To release proteins bound to

GMP agarose, 30pl of IX SDS-PAGE sample buffer was added, samples boiled 5 min., then spun 5 min. at 14K to pellet agarose. The entire sample was used for SDS-PAGE and immunoblot analysis.

Results

THF. SOURCE OF THE C20-SPECIFIC DEFECT IN PRENYLATION IS MGBPL ITSELF

Our previous work (Chapter 2) showing poor incorporation of [3H]MVA into mGBPl provided the first clear example of incomplete modification of a CaaX protein within intact cells. It also suggested that the difficulty in prenylation of mGBPl was more pronounced for

C20 modification by GGTase I. To develop a more detailed understanding of the mechanism(s) underlying this impairment of isoprenoid modification, in vitro studies of prenylation of mGBPl were undertaken. Rabbit reticulocyte lysate was employed as a source of enzyme for synthesis of radiolabeled isoprenoid diphosphates from [^Hlmevalonate and also as a source of prenyl transferases (FTase and GGTase I). To examine the specific activities of the famesyl and geranylgeranyl diphosphate generated from [^HIMVA, K-Ras4B proteins with

C15 and C20-type CaaX motifs were expressed in the reticulocyte lysate. K-Ras proteins possessing either FTase (CVIM) or GGTase-specific (Cl'lL) CaaX motifs were labeled very well, and showed equivalent [^H]MVA incorporation (Fig. 2). Thus, the reticulocyte lysates contained ample amounts of both the FTase and GGTase I enzymes and also efficiently produced both prenyl phosphates. This indicated that the reticulocyte lysate system could be 75 used to detect differences in stoichiometry of prenylation of GBP mutants by observing differences in [^HJMVA labeling.

mGBPlwt (C20-type CaaX [CTIL]) and an FTase-specific mutant (CTIS) were expressed in bacteria, lysates prepared by sonication, insoluble proteins removed, and soluble proteins used in the in vitro assay. The concentration of GBP substrate proteins in the clarified bacterial lysates was estimated by Coomassie Blue staining and the volume of each lysate used was adjusted to provide equivalent amounts of each protein in the assay (compare lanes 1 and

3, lower panel. Fig. 1).

Comparison of the in vitro prenylation of the two GBPs that had been expressed in bacteria showed that the mGBPlwt protein possessing a C20-type motif incorporated much less label from [^HjMVA than did its counterpart with a CIS-type CaaX (CTIS) (Fig. 1, top panel). Because the previous results with K-Ras had shown that the specific activities of

[^HJFPP and [^HJGGPP pools used for protein prenylation in the reticulocyte lysate were similar, this difference in labeling appeared to result from a selective inability of the C- terminus of mGBPl to be modified by GGTase 1.

To test if the poorer labeling of mGBPlwt was the result of a difference in the amount of

C20-type protein that was folded correctly, both protein preparations were compared for the amount of mGBP that could bind to GMP-coupled agarose. Both proteins bound the GMP agarose, and the proportion of the C20-version of mGBPl that bound was equivalent to that of the CIS-version (Fig. 1, lower panel). The association of these GBPs with GMP-agarose was shown to be appropriately guanine nucleotide specific, as binding could be competed by the addition of GTP, but not ATP (data not shown). Thus improper folding did not explain the poor prenylation of the mGBPl protein. Therefore, as had been seen in previous in vivo experiments, mGBPl appeared to be an inefHcient substrate for prenylation. Moreover, mGBPl displayed a C20-speciflc defect in prenylation under cell-free in vitro conditions as well as in intact cells. 76

THF r-TF.RMINUS OF MGBPI CAN BE FULI-Y PRRNYLATED WHEN ATTACHED TO K-RAS

Earlier results in COS-1 cells had shown that mutations within the final 18 amino acids could significantly improve [^HJMVA incorporation into mGBPl (Chapter 2). Those previous in vivo results together with results gained in vitro (Fig. 1) suggested that either the C-terminus of mGBPl directly impeded interaction and modification by GGTase I, or that the C-terminal domain of mGBPl formed part of a larger structure that inhibited GGTase I interaction. To test if the C-terminus of mGBPl was itself sufHcient to impair GGTase I interaction, the final

18 amino acids of K-Ras4B were replaced by those of mGBPl, and the labeling of this chimera tested in COS-1 cells (Fig. 2A, K-Ras:GBP). A second mutant of K-Ras4B was also made, in which the wild type CIS-type K-Ras4B CaaX (CVIM) was replaced with the C20- type CTIL CaaX of mGBPl. This pair of proteins thus allowed direct comparison of prenylation of proteins with identical CaaX motifs that would be modified by GGTase I using

C20 isoprenoid from a common cellular pool. All K-Ras proteins were constructed to possess an N-terminal hemagglutinin (HA) epitope tag to allow the proteins to be immunoprecipitated and quantified (by immunoblot analysis) using a single antibody.

Fig. 2B shows the results of one experiment in which these variant K-Ras4B proteins were expressed in COS-1 cells and labeled with [^HIMVA. Both the CIS and C20 versions of

K-Ras4B incorporated radioactivity very well. The K-Ras:GBP protein incorporated less label than the other K-Ras4B proteins, but immunoblots showed that the total amount of chimeric protein that was expressed was also lower (data not shown). When the incorporation of label was normalized for the amount of protein immunoprecipitated, all three K-Ras4B proteins showed similar levels of prenylation (Fig. 2B, bar graph). Notably, the C20-version (CTIL) of the K-Ras4B protein had the same labeling as the K-Ras:GBP chimera with the same C20- type CTIL motif (bar graph, Fig. 2B). These results indicated that, when removed from the context of the entire mGBPl protein, the C-terminus of mGBPl was a good substrate for prenylation and was capable of being fully modified by GGTase I in vivo. 77

Similarly robust incorporation of label into the chimeric K-Ras:GBP protein was also observed when the K-Ras variants were transcribed, translated, and labeled in vitro with

[3H]MVA in a rabbit reticulocyte lysate (Fig. 2C). Equivalent amounts of [^H] isoprenoid were incorporated into K-Ras:GBP and K-Ras-CTIL. This again suggested that the C- terminus of mGBPl could function well as a substrate for prenylation when removed from the structure of the mOBFl protein. These results with the K-Ras:GBP chimera are quite different from those obtained previously with full-length mGBPl proteins expressed in COS-1 cells, in which the presence of the wild-type mGBPl C-terminus was associated with very poor labeling. Although mutations in this C-terminal domain could improve [^HjMVA incorporation into wild type mGBPl, from this new data it appears that the C-terminal domain of mGBPl does not by itself impair interaction with GGTase I, but instead is part of a larger structure that inhibits GGTase I prenylation.

THF C-TERMINUS OF MGBPl IS POORLY EXPOSED

The ease of prenylation of the C-terminal domain of mGBPl when placed onto K-Ras suggested that in the wild type mGBPl the CaaX motif might be hidden by structures within mGBP itself or concealed by a GBP-interacting protein. To test if the C-terminus of mGBPl was exposed to solvent or hidden, the Triton X-114 extractability of a series of K-Ras chimeras and mGBPl variants was compared. Proteins that possess a hydrophobic group such as an isoprenoid can partition into the detergent ("D") phase of TX-114 if the lipid is exposed sufficiently to allow partitioning into the detergent micelles (28). The mGBPl and K-

Ras proteins have no hydrophobic domains within their primary structure, and thus must both have an isoprenoid and display it on the protein surface in order for the protein to be extracted into the "D" phase.

To allow comparisons of proteins from a single cell lysate, COS-1 cells were co- transfected with GBP and K-Ras proteins that possessed C-termini with the same amino acid residues. Cells were then separated into cytosolic and membrane-containing subcellular 78 fractions, and proteins in these fractions further divided (based on their hydrophobic character) by extraction into a Triton X-114 detergent phase. As previously shown, the native, fully prenylated mature K-Ras4B protein was entirely membrane-bound, and all protein in the membrane fraction was extractable into a TX-114 detergent phase (Fig. 3, panel 3). Upon longer exposure of the immunoblot, small amounts of the precursor form of K-Ras4B could be detected in the soluble (SlOO) phase, and that non-prenylated protein remained in the aqueous

("A") phase after TX-l 14 extraction (data not shown). In contrast, the K-Ras:GBP protein, which has the same final 18 amino acids as mGBPl, was significantly more soluble (Fig. 3, panel 1). This partially cytosolic distribution results from replacement of the C-terminal polybasic domain of K-Ras4B with non-basic residues, and has been observed previously

(29). Importantly, the cytosolic form of K-Ras:GBP partitioned completely into the detergent phase, indicating that the cytosolic protein was fully prenylated and that the isoprenoid was available for extraction. In addition, the 50% of the K-Ras:GBP that was found in the PlOO fraction also partitioned into the detergent phase. This indicated that the GBP tail was compatible with membrane binding, in addition to being suitable for prenylation. These results showed that when the GBP tail was placed onto another protein and modified with isoprenoid, the lipid was accessible and capable of directing that protein into the TX-114 detergent phase.

In contrast, native mGBPl with the same C-terminal domain was only 30% in the PlOO fraction (Fig. 3, panel 2), and, of more importance, none of the mGBPl in the PlOO fraction partitioned into the "D" phase. This suggested that for the small amount of protein that did contain C20 isoprenoid, the lipid was not exposed.

To address if the isoprenoid on a more fully prenylated GBP would exist in a similarly concealed state the partitioning of mGBPl :Ras, a chimeric form of mGBPl in which the C- terminal 18 amino acids were replaced with the final 18 from K-Ras4B, was examined. The mGBPl:Ras protein had previously been shown to be prenylated quite well, as it incorporated

8-fold more [^H]MVA than native mGBPl in COS-1 cells (Chapter 2). As had also been seen 79 previously, the mGBPl:Ras protein was found predominantly in the membrane-containing

PlOO fraction (panel 4). But, during TX-114 extraction, the mGBPlrRas protein partitioned entirely into the aqueous ("A") phase. Therefore, although the 18 K-Ras amino acids

(including the multiple basic residues that have been implicated in membrane binding of K-

Ras4B) increased membrane binding of mCBPl :Ras, the isoprenoid that was attached to the final C-terminal residue was inaccessible to TX-114 detergent extraction. These detergent extraction studies indicated that the body of mOBPl, or proteins that associated with mOBPl, could obscure even a prenylated C-terminus. The data of both the in vitro assays and that from intact cells further suggested that the C-terminus of mCBPl, although competent for prenylation, was a poor substrate when present in the context of mOBPl. This suggested that even without isoprenoid, the C-terminus of native mOBPl might assume a conformation that at least partially concealed the CaaX motif and restricted its access to GGTase 1.

RFMOVAL OF C-TERMINAL AND INTERNAL DOMAINS DOES NOT IMPROVE MGBPI

PRFNYLATION IN VIVO

To identify region(s) of mGBPI that might act in conjunction with the C-terminus to impair prenylation by GGTase I, a series of deletion mutants was constructed (Fig. 4A).

These were designed to remove amino acid segments that might interact with the C-terminus to conceal the CaaX motif, or bind another protein that masked this region. Each of these truncated GBPs was engineered to retain the mGBPI wt CaaX motif (CI'IL) and to contain an

N-terminal HA epitope tag. The first mutant, DEL290/571 (DEL), had an internal deletion from amino acid positions 290 to 571 that placed the final 18 amino acids of mGBPI in-frame with the first 290 amino acids of mGBPI. This protein was used to ascertain whether removal of residues between position 290 and 571 would now allow the 18 amino acid mGBPI C- terminus to be modified as efficiently as when it was present on the K-Ras:GBP protein.

Mutant /^71 had the 14 amino acids directly N-terminal to the CaaX deleted. This mutant was constructed to learn if removal of these amino acids had the same positive effect on prenylation 80 as had been seen in the C20-Chim protein (Chapter 2) in which these amino acids had been replaced with the analogous sequence from K-Ras4B. Other mutants contained successive 100 amino acid truncations of mGBPl with a CTIL CaaX motif appended to the residue number listed. A200 was chosen as the smallest potentially viable protein as this mutant included only the GTP-binding domain, which for Ras-family members is crucial for overall protein folding and stability (30).

Expression and subcellular distribution of each truncated protein in COS-1 ceils was examined in order to determine the expression levels and stability of the proteins, and whether any gained the ability to bind to membranes (a crude indicator of possible improvements in prenylation). The AS71, ASOO, and A400 proteins were expressed at levels equivalent to mGBPlwt (Fig. 4B). DEL, and to a lesser degree A300, showed lower amounts of expression, suggesting that these deletions might cause some loss in protein stability. All of the mutants except A200 retained a mostly cytosolic distribution similar to that of mGBPlwt

(Fig. 4B). The ^(X) protein was expressed very poorly, indicating that with nearly 4(X) amino acids removed, the remainder of the protein showed poor stability. The mechanism through which this small amount of protein became associated with the PKX) fraction was not further investigated. Thus for all but the smallest deletion mutant, these proteins showed no change in membrane binding or indication of increased prenyladon.

To determine how well these constructs were prenylated in vivo, COS-i cells were transfected with the various GBP deletions and were labeled overnight with [^HIMVA. K-

Ras4B possessing the same C20-type CaaX motif as mGBPl (CTIL, Fig. 2A) served to establish the ratio of [^HIMVA incorporation/protein expression for a fiilly prenylated protein. mGBPlwt with the same HA epitope tag served as the baseline for poor prenylation (-10% of

K-RasCTIL). If mGBPl possessed an N-terminal protein element that prevented prenylation, we predicted that its removal would result in a dramatic improvement in the MVA/protein ratio to a value similar to the fully prenylated K-RasCTIL protein. Unexpectedly, initial experiments 81 resulted in barely detectable labeling of any of the truncated GBP proteins (data not shown).

Only after increasing the concentration of radiolabel 3-fold (2(X)nCi/ml) could the low level of

MVA incorporation into these GBPs be readily observed (Fig. 5A). Comparison of [^HIMVA incorporation per amount of protein showed that none of the deletion mutants incorporated significantly more [^H] label from MVA than mGBPlwt (Fig. 5B). In fact, the A300 form, whose expression was roughly similar to the other mutants, showed a marked decrease in prenylation to barely detectable levels. Additionally, the DEL protein in which the 18 amino acid C-terminus of mGBPI was appended to position 290 was a very poor substrate for prenylation, even though the C-terminus had been shown earlier to be a suitable substrate for modification when placed onto K-Ras (Fig. 2B,C). This poor MVA incorporation into the truncated GBP proteins could not be explained by protein sequestration or insolubility, as the majority of the proteins were found to be cytosolic (Fig. 4B) and thus should have been available for modification by the soluble GGTase I enzyme. Therefore, although the C- terminal 14 residues of mGBPI were shown previously to have a negative impact on isoprenoid modification of the CaaX motif, their removal, or the removal of up to nearly 300

C-terminal amino acids, did not appear to alleviate the difHculty of mGBPI prenylation in vivo.

These results make it clear that the regions of mGBPI that impede prenylation of the CaaX do not reside solely in the C-terminus, but also include the N-terminal third of mGBPI that encompasses the GTP-binding domain.

PRRNYLATION OF MGBPI DELETION MlfTANTS IS NOT IMPROVED IN VLTRO

To address whether the in vivo results obtained with the mutant GBP proteins would extend to cell-free conditions as well, we again employed the rabbit reticulocyte in vitro prenylation assay. Recombinant GBP deletion mutant proteins possessing the HA epitope were expressed in bacteria, and GBPs in the soluble fraction were added to the reticulocyte lysate in the presence of [3H]MVA. The results, similar to those that had been seen in vivo, showed that none of the GBP deletion mutants incorporated substantially more [^HIMVA than 82 mGBPlwt (Fig. 6). In fact, the relative levels of MVA incorporation closely mimicked those seen in COS-1 cells, with the A300 protein being an extremely poor target for geranylgeranylation. Even the A200 mutant, which could be stably produced in bacteria, remained largely unmodified. Control reactions showed that the K-Ras4B protein produced in bacteria was a much better substrate (- 6-fold, data not shown) for prenylation than the GBP deletion mutants, indicating that the reticulocyte lysate could prenylate good substrates easily.

Thus, the lack of improvement in prenylation of these mutants observed in vivo was not due to their presence in a mammalian cell or to problems in access to GGTase I. Instead, the N- terminus of mGBPl, including the GTP-binding domain, appeared to prevent prenylation of the C-terminal CaaX motif.

Discussion

The experimental results presented here have confirmed mGBPl's status as the only known protein possessing a largely non-prenylated CaaX motif. The mechanism underlying this deficiency in prenylation appears to be complex, consisting of contributions from both the extreme C-terminus and N-terminal GTP-binding domain. Thus, mGBPl provides a novel model for the study of cellular mechanisms governing in vivo selection of substrates by FTase and GGTase I. This identification of a new inhibitory domain contributes to the understanding of prenyl transferases and the continued development of improved inhibitors of these enzymes.

THF C20-SPECIF1C DEFECT IN MGBPl PRENYLATIQN IS AN INTRINSIC PROPERTY OF ITS

STRUCTURE

Our previous in vivo results showed that the endogenous form of mGBPl authentically induced by IFN treatment of macrophages, and also forms transiently expressed in COS-1 cells, exhibit a C20-specif1c defect in prenylation. Changing the transferase speciHcity at the

CaaX motif from GGTase I to FTase (via a Leu to Ser CaaX substitution) resulted in a 4-fold increase in [^HJMVA incorporation (Chapter 2). Our current data shows that this C20-specific defect does not arise from in vivo problems in GGPP production, GGTase I activity, or 83 substraterenzyme inaccessibility, as limited in vitro C20-modification also occurs using rabbit reticulocyte lysate and recombinant mGBPlwt (Fig. 1). Therefore, the poor GGTase I- associated prenylation of mGBPl is not due to its presence in the environment of a living cell, but instead reflects an intrinsic inability of the protein to be effectively modifled by GGTase 1. Studies using chimeric Gy subunits from heterotrimeric G proteins showed that N- terminal regions of CaaX substrate proteins can play a large role in determining which prenyl transferase will recognize and prenylate the CaaX cysteine (16). However, in contrast to the work presented here, the chimeric Gy proteins in the aforementioned study continued to be recognized and prenylated by either FTase or GGTase I. Our work with mGBPl shows that non-CaaX regions of substrate proteins can selectively, and negatively, influence CaaX modification by impeding prenylation by one prenyl transferase over another. This finding is quite interesting in light of the fact that the CaaX substrate binding site of both FTase and GGTase I resides in a cleft between a and P subunits, of which the a subunit is common to both enzymes (5). Our work with mGBPl provides new information that improves our understanding of the mechanisms behind prenyl transferase substrate recognition.

THF C-TERMINUS OF MGBPl IS A GOOD SUBSTRATE FOR PRENYLATION

Earlier work by James et al showed that a polybasic domain within the C-terminal 18 amino acids of K-Ras promoted cross-prenylation of the CaaX motif by GGTase 1(15). Our previous results implicated a similar 18 amino acid C-terminal region of mGBPl in blocking efficient prenylation by GGTase (Chapter 2). We therefore reasoned that amino acids within this region of mGBPl might play a key role in limiting GGTase I interaction. However, chimeric K-Ras proteins possessing the C-terminus of mGBPl were prenylated well, both in vivo in COS-1 cells and in vitro in a rabbit reticulocyte lysate (Fig. 2). Thus, in the case of mGBPl, the final 18 amino acids of mGBPl appear to require N-terminal parts of the native protein to hinder prenylation of the CaaX. 84

That the C-terminus of mGBPlwt appears to be poorly recognized by GGTase I may be due to the conformation of mCBPl itself or perhaps concealment of the C-terminal domain by a GBP partner protein. The poor TX-114 extractability (Fig. 3) of both prenylated

(mGBPChim) and non-prenylated (mGBPlwt) forms of mGBPl suggests that the C-terminus of mGBPl is concealed from GGTase I, thus limiting its prenylation.

THF N-TERMINUS OF MGBPI ACTS TO LIMIT CAAX PRENYLATION

Because the final 18 amino acids of mGBPl were compatible with GGTase I modification, it appeared that other regions of mGBPl were required to produce the prenylation deficiency. Using a series of GBP deletion mutants, the N-terminal third of mGBPl, containing the GTP-binding domain, was identiHed as a region that impaired CaaX modification (Fig. 5). To our knowledge, this is the first case in which a nucleotide-binding domain has been shown to affect prenylation of a CaaX motif. For the Rab GTPases, protein prenylation by the Rab prenyltransferase (GGTase 11) is influenced by the Rab protein nucleotide state. This effect arises because, unlike CaaX prenyltransferases, GGTase n has a third subunit termed Rab Escort Protein (REP). REP preferentially associates with newly synthesized GDP-bound Rab proteins and presents them to the GGTase Q transferase machinery for prenylation (5). However, REPs are quite selective for GGTase n and Rab proteins, and are unlikely to provide the explanation for the lack of prenylation of mGBPl.

However, it does remain a possibility that mGBPl associates with a partner protein(s) that could hinder access to the CaaX motif.

SRI.F ASSOCIATIVE VS. PARTNER PROTEIN OCCLUSION MODELS FOR PRENYI ATTON

Mimnm In an effort to construct models for how the N-terminus of mGBPl could be hindering access to its own CaaX motif, we have turned to the recently solved three-dimensional structure of human GBP-1 (hGBPl) (19) as a source of GBP structural information. mGBPl and hGBPl are 70% identical at the amino acid level, and secondary structure predictions for 85 mGBPl show close resemblance to the determined secondary structure of hOBPl

(unpublished data). Thus, it is very likely that the overall folding of mGBPl mimics that of its human homologue.

The N-terminal 300 amino acids of hGBPl (the GTP-binding domain) are folded into a globular domain loosely resembling that of Ras. The C-terminal half of the molecule is comprised entirely of a-helices and folds into what appears to be a very rigid, rod-like domain.

This rod-like structure is composed of two three-helix bundles that extend away from the globular domain and a single -80 amino acid helix that is folded back, making contact with both the helix bundles and the GTP-binding domain. The long helix terminates with a turn and another short helix that is anti-parallel to the long helix. Although the extreme C-terminal region containing the CaaX is disordered, it is unlikely that the main part of the C-terminal domain of the protein exhibits much flexibility due to the apparent rigidity of the structure and number of intramolecular contacts between helices.

Based on this structure, and other data suggesting that hGBPl can undergo GTP- dependent oligomerization, the authors of the structural paper speculate that hGBPl is a member of the dynamin family of GTPases (19). Dynamins, as well as the Mx family of interferon GTPases, have been shown to exhibit the type of C-terminal molecular

"backfolding" present in hGBPl and are also capable of oligomerization (31-34). Thus, if mGBPl were to behave similarly (i.e. oligomerize), then perhaps multimerization of mGBPl could bury the CaaX motif such that it would not be available for modification.

At the present time, our additional preliminary data do not strongly support an oligomerization model for three main reasons. First, we do not observe a shift in SDS-PAGE mobility of mGBPl after chemical cross-linking COS-1 lysates containing mGBPl. Secondly, we have not seen any evidence of mGBPl self association using yeast two-hybrid analysis.

Finally, and most tellingly, dynamin and Mx protein self-associations require the interaction of

C- and N-terminal elements (31-34). Thus, it is difficult to envision how truncated GBP 86 proteins such as A300, which lack the entire C-terminus and are comprised of only the globular

GTP-binding domain, could retain the ability to multimerize in a fashion similar to the dynamin or Mx proteins.

Currently, our data favor a model in which a partner protein associates with the N- terminal GTP-binding domain, and the nearby CaaX motif, of mGBPl and blocks access of the CaaX to GGTase I. This model is based loosely on the example from RhoGDI binding to the members of the Rho family of small GTPases. In this case, RhoGDI is a partner protein of the Rho/Rac/Cdc42 GTPases and makes contacts with both the GTP-binding domain and the prenylated C-terminus of the GTPase to extract the protein from the membrane to which its attached (35). The effect of GDI binding is to sequester the GTPase protein in the cytosol where it remains inactive until released from the GDI. Our model proposes that mGBPl has a similar binding partner that covers the CaaX motif until a signal for its release arrives, at which time the CaaX would be uncovered and available for prenylation.

This model presents a testable hypothesis for further experimentation into the novel mechanism mGBPl uses to block its own prenylation. Future study of this protein should allow for both a better understanding of the cellular mechanisms governing substrate choices of prenyl transferases as well as new insight into the biology of the GBPs.

Acknowledgments

We would like to thank Janina Brandt and Molly Foster for their excellent technical assistance and Robert Bellin and Pierig Lepont for their technical advice and useful conmients.

References

1. Vojtek, A. B., and Der, C. J. (1998) /. Biol. Chem. 273, 19925-19928

2. Lowy, D. R., and Willumsen, B. M. (1993) Ann. Rev. Biochem. 62, 851-891

3. Jackson, J. H., Cochrane, C. G., Bourne, J. R., Solski, P. A., Buss, J. E., and Der,

C. J. (1990) Proc. Natl Acad. Sci. USA 87, 3042-3046 87

4. Kato, K., Cox, A. D., Hisaka, M. M., Graham, S. M., Buss, J. E., and Der, C. J.

(1992) Proc. Natl. Acad. Sci. USA 89, 6403-6407

5. Zhang, F. L., and Casey, P. J. (1996) Annu. Rev. Biochem. 65, 241-269

6. Epstein, W. W., Lever, D., Leining, L. M., Bruenger, E., and Rilling, H. C. (1991)

Proc. Natl .Acad .Sci .USA 88, 9668-9670

7. Bishop, W. R., Bond, R., Petrin, J., Wang, L., Patton, R., Doll, R., Njoroge, G.,

Catino, J., Schwartz, J., Windsor, W., Syto, R., Schwartz, J., Carr, D., James, L.,

and Kirschmeier, P. (1995) J. Biol. Chem. 270, 30611-30618

8. Cox, A. D., Garcia, A. M., Westwick, J. K., Kowalczyk, J. J., Lewis, M. D.,

Brenner, D. A., and Der, C. J. (1994) J. Biol. Chem. 269, 19203-19206

9. James, G. L., Goldstein, J. L., Brown, M. S., Rawson, T. E., Somers, T. C.,

McDowell, R. S., Crowley, C. W., Lucas, B. K., Levinson, A. D., and Marsters, J.

C., Jr. (1993) Science 260, 1937-1942

10. Kohl, N. E., Mosser, S. D., deSolms, S. J., Giuliani, E. A., Pompliano, D. L.,

Graham, S. L., Smith, R. L., Scolnick, E. M., Oliff, A., and Gibbs, J. B. (1993)

Science 260, 1934-1937

11. Lemer, E. C., Qian, Y., Blaskovich, M. A., Possum, R. D., Vogt, A., Sun, J., Cox,

A. D., Der, C. J., Hamilton, A. D., and Sebti, S. M. (1995) J. Biol.Chem. 270, 26802-26806

12. Whyte, D. B., Kirschmeier, P., Hockenberry, T. N., Nunez-Oliva, L, James, L.,

Catino, J. J., Bishop, W. R., and Pai, J.-K. (1997) J. Biol. Chem. 272, 14459-14464

13. Rowell, C. A., Kowalczyk, J. J., Lewis, M. D., and Garcia, A. M. (1997) J. Biol.

Chem. 272, 14093-14097

14. Bos, J. L. (1989) Cancer Res. 49, 4682-4689

15. James, G. L., Goldstein, J. L., and Brown, M. S. (1995) /. Biol. Chem. 270, 6221-

6226 88

16. Kalman, V. K., Erdman, R. A., Maltese, W. A., and Robishaw, J. D. (1995) J. Biol.

Chem. 270, 14835-14841

17. Cheng, Y.-S. E., Colonno, R. J., and Yin, F. H. (1983) J. Biol. Chem. 258, 7746-

7750

18. Cheng, Y.-S. E., Becker-Manley, M. F., Chow, T. P., and Horan, D. C. (1985) J.

Biol. Chem. 260, 15834-15839

19. Prakash, B., Praefcke, G. J. K., Renault, L., Wittinghofer, A., and Herrmann, C.

(2000) Nature 403, 567-571

20. Anderson, S. L., Carton, J. M., Lou, J., Xing, L., and Rubin, B. Y. (1999) Virology

256, 8-14

21. Cheng, Y.-S. E., Patterson, C. E., and Staeheli, P. (1991) Mol. Cell. Biol. 11,4717-

4725

22. Wynn, T. A., Nicolet, C. M., and Paulnock, D. M. (1991) J. Immunol. 147, 4384-

4392

23. Vestal, D. J., Buss, J. E., Kelner, G. S., Maciejewski, D., Asundi, V. K., and Maki,

R. A. (1996) Biochem. Biophys. Res. Commun. 224, 528-534

24. Vestal, D. J., Buss, J. E., McKercher, S. R., Jenkins, N. A., Copeland, K. G., Kelner,

G. S., Asundi, V. K., and Maki, R. A. (1998) J. Interferon and Cytokine Res. 18,977-

985

25. Nantais, D. E., Schwemmle, M., Stickney, J. T., Vestal, D. J., and Buss, J. E. (1996)

J. Leukoc. Biol. 60, 423-431

26. Barton, R. M., and Worman, H. J. (1999) /. Biol. Chem. 274, 30008-30018

27. Hancock, J. F., Paterson, H., and Marshall, C. J. (1990) Cell 63, 133-139

28. Hancock, J. F., Magee, A. I., Childs, J. E., and Marshall, C. J. (1989) Cell 57, 1167-

1177 89

29. Jackson, J. H., Li, J. W., Buss, J. E., Der, C. J., and Cochrane, C. G. (1994) Proc.

Natl. Acad. Sci. USA 91, 12730-12734

30. Willumsen, B. M., Papageorge, A. G., Hubbert, N., Bekesi, E., Kung, H. F., and

Lowy, D. R. (1985) Embo J 4, 2893-2896

31. Schumacher, B., and Staeheli, P. (1998) J. Biol. Chem. 273, 28365-28370

32. Di Paolo, C., Hefti, H. P., Meli, M., Landis, H., and Pavlovic, J. (1999) J. Biol.

Chem. 274, 32071-32078

33. Muhlberg, A. B., Wamock, D. E., and Schmid, S. L. (1997) EMBO J16, 6676-

6683

34. Smimova, E., Shurland, D.-L., Newman-Smith, E. D., Pishvaee, B., and van der

Bliek, A. M. (1999) J. Biol. Chem. 274, 14942-14947

35. Hoffman, G. R., Nassar, N., and Cerione, R. A. (2000) Cell 100, 345-356

Figure Legends

Figure 1. mGBPl displays a C20-specific defect in prenylaiion when labeled with [^HIMVA in vitro. {MVA) Bacterially produced His-tagged mGBPlwt and an FTase-specific point mutant (CTIS) were added in equal amounts to separate tubes containing rabbit reticulocyte lysate with 35 fiCi [3H]MVA and incubated at 30°C for 2 hours. Aliquots were displayed by

SDS-PAGE, transferred to PVDF, and radiolabeled proteins detected by fluorography after 4 days exposure to film. Protein amounts of GBPs in the bacterial lysates were adjusted so that equal amounts were used in the in vitro assay. Amounts were confirmed by intensity of bands after immunoblot analysis (lower panel, L). (Blot) Amounts of bacterial lysates identical to those used in the in vitro prenylation assay were loaded (L) onto GMP-agarose and after washing, the amount of protein remaining bound to the agarose (B) was released by the addition of SDS-PAGE sample bufler followed by boiling.

Figure 2. The C-terminus of mGBPl can be fully modified when placed onto K-Ras4B. A)

Alignment of the C-termini of HA-tagged K-Ras4B proteins with differences from wild type 90 residues shown in bold. B) COS-1 cells were transfected with the indicated constructs or empty vector (V) and labeled overnight with 67 pCi/ml [^HJMVA in the presence of 10 pM compactin. upper panel-MWK Proteins were immunoprecipitated using a monoclonal anti-HA antibody and labeled proteins visualized by fluorography as in Fig. 1 after 28 days exposure. lower panel-E\ot After fluorography, the membrane was stripped of enhancer and used for immunoblot detection using the anti-HA antibody. LC denotes the light chain of the antibody used for immunoprecipitation. Bar graph -Ratios of [3H]MVA incorporation per amount of protein present in the immunoblot were calculated and expressed relative to K-Ras4B possessing a wild type, CIS-type CaaX (wt). Number of experiments is indicated over columns and bars represent S.E.M. Q Plasmid constructs used in transfection experiments

(B) were transcribed and translated in a rabbit reticulocyte lysate containing 35 fiCi [^H]MVA to detect prenylation or lOpCi [3^S]methionine to visualize protein levels. Radiolabeled proteins were visualized by fluorography after 4 days for M\'A or 12 hours for Bar graph

- Ratios of [3H]MVA incorporation per amount of protein were calculated as in B) with error bars representing the range of values from two separate experiments.

Figure 3. The exposure of the C-terminus of mGBPl differs greatly from that of K-Ras4B. mGBP and K-Ras4B proteins with equivalent C-termini (mGBPl or K-Ras4B) were co- expressed in COS-1 cells and subjected to SlOO/PlOO fractionation followed by TX-114 phase separation. Calculated Fraction with Isoprenoid refers to the observed stoichiometry of prenylation as determined by MVA/protein ratios in COS-1. Comparisons of the ratio of

[3H]MVA/protein for mGBPl wt and a CIS-specific chimeric mutant (mGBPl:Ras) in COS-1 cells showed that mGBPl wt incorporates approximately 12% the amount of MVA label/protein as does mGBPl;Ras (Chapter 2). These values were used to assign a stoichiometry of prenylation of 12% for mGBPlwt assuming 100% prenylation for mGBPlrRas.

Figure 4. Deletion mutants of mGBPl are stable and soluble when expressed in COS-1. A)

Cartoon diagram of various mGBPl deletion mutants with the GTP-binding domain (G- 91 domain) depicted by shading. B) COS-l cells expressing proteins described in A) were separated into SlOO and PlOO fractions as described above. After acetone precipitation, proteins were resuspended in 100 (il of IX SDS-PAGE sample buffer and 20 |il of each sample loaded per lane. After SDS-PAGE and transfer to PVDF, proteins were detected using an anti-HA antibody as described in Materials and Methods. Because of limited expression of

DEL290/571 (Del) and A200, these proteins were visualized by longer exposure times during development of the ECL immunoblot (inset panel, bottom).

Figure 5. Removal of the C-terminal half of mGBPl does not improve MVA incorporation in

COS-1. A) COS-1 cells were transfected with K-Ras (arrow) or GBP proteins possessing N- terminal HA epitope tags and labeled overnight in the presence of 200 pCi/ml [^HIMVA.

Following immunoprecipitation, captured proteins were released from immunoconjugates by boiling in the presence of IX SDS-PAGE sample buffer. After SDS-PAGE, transfer to

PVDF, and fluorographic exposure, proteins possessing the HA epitope were detected on the same membrane using immunoblot analysis as described above. B) The bar graph summarizes MVA:protein ratios of the various GBP mutants relative to mGBPl wt. Percent of mGBPlwt prenylation was set to 12% of stoichiometric prenylation (i.e. Ras) as in Fig. 3.

Number of experiments is shown as N and error bars represent S.E.M. Due to the poor expression of A200, analysis of prenylation of this mutant was not possible.

Figure 6. MVA incorporation into GBP deletion mutants is not improved in reticulocyte lysates. A) Recombinant GBP deletion mutant proteins, expressed in bacteria, were added to rabbit reticulocyte lysates and their MVA incorporation tested as in Fig. 1. Blot To quantify the amount of recombinant protein that was added to the reticulocyte lysate, representative amounts of protein were analyzed by anti-HA immunoblot analysis. B) Bar graph represents mean

[^HJMVA/protein ratios for the experiment in Aj which was performed in duplicate. All ratios were normalized to mGBPlwt. 92

wt CTIS

L B L B 12 3 4

Figure 1. 93

A) Predicted C-terminal Residues isoprenoid KRas KEKMSKDGKKKKKKSKTKCVIM CI 5 KRasCTIL KEKMSKDGKKKKKKSKTKCTIL C20 KRas:GBP KEKMQEIEKIKNMPPPRSCTIL C2 0

300 wt CTILRaszGBP

MVA

35S

CTIL Ras:GBP

Figure 2. 94

C-term 18 a.a. = mGBP1

S P Calculated Fraction AD AD; Protein wttti Isoprenold

K-Ras:GBP 100%

mGBPIwt 12%

C-term 18 a.a. = K-Ra84B S P Calculated Fraction ~A D A D Protein with Isoprenoid

K-Raswt 100%

mGBP1:Ra8 100%

Figure 3. 95

wt i Del I 5711 5001400 3001200 SP SP SP SP SP SP S P

Figure 4. 96

Ras wt Del 571 500 400 300 Ras wt Del 571 500 400 300

MVA Blot B)

NsS 2 6 4 5 4

Figure 5. 97

A) wt 571 500 400 300 200 wt 571 500 400 300 200

MVA Blot

B) 150-1

a n o E 100- X o >a

e o

>e s &

571 500 400 300 200

Figure 6. 98

CHAPTER 4. GENERAL CONCLUSIONS

Overview

The work presented in this dissertation describes the initial characterization of the prenylation of murine Guanylate Binding Protein-1 (mGBPl). Based on the amino acid sequence of mGBPl and our current understanding of cellular prenylation, this protein should be recognized by geranylgeranyl transferase I (GGTase I) and each molecule of mGBPl should have a geranylgeranyl isoprenoid covalently attached to its C-terminus. However, our experiments with mGBPl both in vivo and in vitro suggest that this protein possesses the intrinsic ability to block its own prenylation, and naturally exists in a largely unmodified state.

This discovery of incomplete prenylation is perhaps the most important finding of the dissertation, as it challenges the widely-held belief that prenylation is both automatic and stoichiometric for proteins bearing the canonical CaaX motif. That the N-terminal, OTP- binding region of mGBPl plays a large role in this inhibitory event is also quite interesting, as it suggests that this block in modification may be a regulated event in vivo.

Initial efforts to characterize the basis for the observed poor incorporation of [3H]MVA into mGBPl focused on thoroughly exploring, and ultimately excluding, cellular factors other than mGBPl itself. Based on the experiments presented in Chapter 2, we tested and discarded hypotheses of metabolic and cell type-specific explanations for the deficiency in mGBPl prenylation. Expression of mGBPl in cell types other than IFN-treated macrophages did not improve its [^H]MVA incorporation. Comparison to other C20-modined proteins in other cell types showed that mGBPl itself, and not metabolic issues related to isoprenoid metabolism, was the culprit in blocking its own prenylation. Interestingly, work from Chapter 2 led us to believe that this deficiency was more severe with respect to GGTase I-mediated prenylation than FTase. This transferase specificity provides a foundation for further studies into how

FTase and GGTase I choose their substrates in vivo. Perhaps a definable region of mGBPl will eventually be identified that will explain this transferase-specific defect in prenylation. If 99 so, the inclusion of such elements into new generations of prenyl transferase inhibitors could lead to the development of highly specific inhibitors of FTase.

As mentioned before, a failure in prenylation has been described for some members of the Ga family of heterotrimeric signaling proteins. However, these G-proteins possess

"pseudo CaaXs" which are simply poor substrates for prenylation. In the case of mGBPl, its

CaaX motif (CTIL) has been shown to be functional on numerous other proteins, including two other GBP family members. Thus, the lack of efficient modification of mGBPl is not due to a non-functional CaaX.

We demonstrated in Chapter 2 that non-CaaX amino acids in the C-terminal region of mGBPl contribute to mGBPl's prenylation deficiency. Non-CaoX amino acids have been shown to promote GGTase I-mediated cross-prenylation of both Gysubunits and Ras proteins possessing CIS-type CaaX motifs. However, in each of these cases, prenylation by GGTase I was promoted, not impaired, by the CaaX-extrinsic amino acids. Therefore, mGBPl represents the first example whereby non-CaaX amino acids actually diminish the capacity for prenylation and override the positive signal from the "good" CaaX.

However, as was shown in Chapter 3, the ability of these non-CaaX amino acids to participate in vitiating prenylation by GGTase I relies on participation from N-terminal portions of mGBPl. Thus, the unique mechanism that mGBPl uses to hinder its own prenylation results from contributions from both N- and C-terminal protein elements.

Regulation of Prenylation?

That the region of mGBPl that impairs prenylation includes the GTP-binding domain suggests that GTP binding/hydrolysis of mGBPl may potentially regulate prenylation of its C- terminus. Prenylation of the Rab family of proteins is sensitive to their guanine nucleotide state. This is due to a GDP dependence for binding of an accessory factor (REP) necessary for prenylation by GGTase II (Rab geranylgeranyl transferase). However, both CaaX 100 prenyltransferases identified so far are heterodimers and give no indication that a third subunit is required for activity, making a Rab-like scenario for mGBPl unlikely.

GTP-dependent oligomerization of mGBPl may also be the cause of mGBPl's lack of prenylation. Human GBP-1 (hGBPl) as well as the distantly related Mx and dynamin families of proteins are known to undergo multimerization. However, we have not been able to decisively detect the presence of any higher molecular weight forms of mGBPl that would suggest its multimerization. Additionally, we have been unable to detect any mGBPl protein- protein interactions with itself using the yeast two-hybrid assay. Therefore, while oligomerization of mGBPl may be responsible for its block on prenylation in vivo, we currently have no evidence to support this.

Guanine dissociation inhibitors (GDIs) of the Rho family sequester these small

GTPases into the cytoplasm through simultaneous interaction with the GTP-binding domain and prenylated C-terminus. If a similar situation exists with mGBPl, then perhaps a putative mGBPl binding partner binds to both the GTP-binding domain as well as the unmodified mGBPl C-terminus to preclude efficient geranylgeranylation. Such a mechanism would allow for regulated access of the C-terminus for prenylation and could possibly be an integral component of mGBPl's biology.

Biological Signiflcance

Presently, very little is known regarding the physiological function of the guanylate- binding proteins. Early studies have shown that in IFN-treated cells, mRNA for GBPs make up the most abundant transcripts in interferon-treated cells. Thus, it seems quite likely that

GBPs play an important role in mediating the cellular effects of IFN. Only one report exists regarding possible anti-viral activity for human GBP-1. In that report, no mention is made whether the famesyl isoprenoid attached to hGBPl is essential for activity of the protein.

It is tempting to speculate reasons why mGBPl prevents its own modification.

Because this virtual absence of isoprenoid appears to be an essential characteristic of mGBPl, 101 it appears likely that selective pressures have created this protein with a fully functional, yet grossly-unmodified, CaaX motif.

Perhaps mGBPl in its unmodified state is actually a suitable partner for the aforementioned RhoGDI class of regulatory molecules. In this case, the high levels of GBP proteins that accumulate in IFN-treated cells could titrate out interactions of RhoGDI with their natural substrates. This would free the Rho/Rac/CDC42 proteins for plasma membrane interactions and signaling.

Alternatively, mGBPl may exist in a non-prenylated state until the proper cellular cue

(i.e. presence of virus) triggers geranylgeranylation of the mGBPl CaaX. The recoverin and

ARF proteins are myristoylated proteins who extrude their myristoyl groups using Ca2+- or

GTP-dependent switch mechanisms respectively (1,2). Perhaps mGBPl undergoes a similar switch from a non-prenylated form to a prenylated version required for putative anti-viral activity.

Finally, the lack of mGBPl prenylation could be an adaptive strategy. Perhaps at one time, viruses or other pathogens utilized prenylated mGBPl, or another prenylated GBP, to their advantage. In this case, the lack of isoprenoid observed on mGBPl could be an attempt to thwart pathogenesis through sequestration of pathogen binding to non-prenylated mGBPl.

Titration of the binding of pathogenic proteins from "good", more efficiendy prenylated GBPs necessary for host defense to the non-prenylated mGBPl may be a means by which prenylation (or lack thereof) could be playing a crucial role in prevention of disease.

References

1. Ames, J. B., R. Ishima, T. Tanaka, J. I. Gordon, L. Stryer, and M.

Ikura. 1997. Molecular mechanics of calcium-myristoyl switches. Nature 389:198-

202.

2. Goldberg, J. 1998. Structural basis for activation of ARF GTPase: mechanisms of

guanine nucleotide exchange and GTP-myristoyI switching. Cell 95:237-248. 102

ACKNOWLEDGMENTS

I would first and foremost like to heartily thank my major professor, Jan Buss, for her many years of encouragement, wisdom, guidance, and (most importantly) patience. Though my foibles were numerous and frequent, Jan never wavered in her display of faith in my abilities. She was also a pillar of strength for me during my father's long, unsuccessful battle with cancer. For this, I owe her a debt of gratitude I most certainly will never be able to repay.

Thus, she has been not only an excellent mentor but also a cherished friend - one whose daily company I will miss greatly after I leave Iowa State.

My wife, Janese O'Brien Stickney, has also been a constant source of support for me during my time in graduate school. Not only has she taken a very active part in directly assisting me with many of my academic endeavors, but she has served as my ultimate standard for how to conduct myself as a person as well as a graduate student. She has earned my utmost respect for both her academic abilities as well as her caring and compassionate personality, and I am ever grateful for her love and the opportunity to spend my life with her.

If there are two people responsible for the successes I may experience in life, it is most certainly my parents Tom and Joan Stickney. As an only child of more "mature" parents, I was quite naturally the focal point of their lives (although they went to great lengths to disguise this from me). Through their faith in me and the countless hours they invested in my rearing, I was prepared to face the world as an adult with the self-confidence that comes from belief in oneself. Their indelible mark on my character will shine brightly for my entire life.

I must also thank the many students in the Department of Biochemistry, Biophysics, and Molecular Biology for their years of companionship. Specifically, Robert Bellin and

Pierig Lepont have become close friends whose opinions on life and science are invaluable to me. I thank you for your many years of friendship and advice.

I'd also like to thank my many friends outside of the department who have given me limitless moral support. Michael Belshan and Aaron Cook and I were members of a rock band 103 for a number of years and they will always be cherished friends. Likewise, Michael Sears and

Shawn Lyons are longtime friends who have shown me that only children can still have brothers. Their support and friendship through good times and bad has enriched my life and has been a priceless gift to me.