Introduction

Top Abstract Introduction Methods and Materials Results Discussion References

Most low molecular weight G proteins and the  subunits of trimeric G proteins possess a specific C-terminal sequence (referred to as the CAAX motif), which makes them suitable candidates for a series of post-translational modifications (Clarke, 1992; Takai et al., 1992; Casey, 1995; Wedegaertner et al., 1995; Kowluru et al., 2000). The first of these modifications involves attachment of either a 15-carbon (referred to as farnesylation) or a 20-carbon (referred to as geranylgeranylation) derivative of MVA to the cysteine via a thioether linkage. These reactions are catalyzed by protein farnesyl- or geranylgeranyl-transferases, respectively (Casey et al., 1989; Kato et al., 1992; James et al., 1993; Kohl et al., 1993; Kowluru et al., 2000). Subsequent modifications include, cleavage of the three terminal amino acids after the isoprenylated cysteine by a protease of microsomal origin, resulting in the exposure of the carboxylate anion of the prenylated cysteine residue (Hancock et al., 1991; Clarke 1992; Kowluru et al., 1996a). This site is further subjected to methylation by a prenyl cysteine methyltransferase, which neutralizes the carboxylate anion, thus making the candidate G protein more hydrophobic, resulting in its translocation to the membrane sites for interaction with their respective effector proteins (Clarke, 1992). Besides the aforementioned modifications, certain G proteins (e.g., H- and N-Ras and the  subunits of trimeric G proteins) undergo acylation (typically, incorporation of palmitate) at a cysteine residue, which is frequently upstream to the prenylated cysteine (Hancock et al., 1989; Wedegaertner et al., 1995).

By using inhibitors of each of these modifications, previous studies from our laboratory have demonstrated involvement of these proteins in physiological insulin secretion (Li et al., 1993; Metz et al., 1993). For example, using lovastatin (LOVA), an inhibitor of MVA biosynthesis from 3-hydroxy-3-methylglutaryl coenzyme A, which is a precursor of farnesyl and geranylgeranyl pyrophosphates, we have reported inhibition of glucose-induced insulin secretion from normal rat islets (Metz et al., 1993). Studies by Li et al. (1993) have demonstrated similar inhibitory effects by LOVA on bombesin- and vasopressin-induced insulin secretion from HIT-T15 cells. In each case, inhibition of isoprenylation of -cell G proteins also resulted in altered subcellular distribution of these proteins in these cells as evidenced by accumulation of nonprenylated proteins in the cytosolic fraction (Li et al., 1993; Metz et al., 1993; Kowluru and Metz, 1996). Furthermore, using acetyl farnesyl cysteine, a specific inhibitor of prenyl cysteine methyltransferases and/or cerulenin, a specific inhibitor of fatty acylation, we (Metz et al., 1993; Kowluru et al., 2000) have provided evidence for the regulatory roles of these post-translational modification steps in physiological insulin secretion; these findings have been subsequently confirmed by other investigators (Deeney et al., 2000; Yajima et al., 2000; Straub et al., 2002).

The current study further explores the roles of protein isoprenylation steps in physiological insulin secretion from insulin-secreting clonal  (TC3) cells. We felt a critical need for these studies because LOVA, which was used in previous studies, also inhibits both sterol and nonsterol limbs of the cholesterol pathways; affecting many compounds that could have specific functional roles in the  cell (Li et al., 1993; Metz et al., 1993; Kowluru and Metz, 1996). Specifically, LOVA inhibits the biosynthesis of both farnesyl and geranylgeranyl pyrophosphates and, thus, it is difficult to determine precisely which of the two signaling pathways (e.g., farnesylation or geranylgeranylation) is critical for the regulation of insulin secretion from the islet  cell. With these facts in mind, we undertook the present study to evaluate specific inhibitors of farnesyl and geranylgeranyl transferases and examine their effects on glucose- and calcium-mediated insulin secretion from TC3 cells to further examine the roles of these modification steps in glucose- and calcium-induced insulin secretion. To provide corroboration for these results, we compared the effects of other commercially available inhibitors of protein farnesylation and geranylgeranylation on glucose- and KCl-induced secretion from these cells. Our data clearly support the original formulation that both farnesylation and geranylgeranylation of proteins are necessary for glucose- and calcium-mediated exocytotic secretion of insulin.

	Methods and Materials

Top Abstract Introduction Methods and Materials Results Discussion References

Materials. Rat Insulin ELISA kit was purchased from American Laboratory Products Company (Windham, NH). [-³²P]GTP was purchased from Amersham Biosciences (Piscataway, NJ). Enhanced chemiluminescence reagents and Hyper film were from Amersham Biosciences. GGTI-2147 was purchased from Calbiochem (San Diego, CA). Manumycin A was obtained from Sigma-Aldrich (St. Louis, MO). MTT assay reagents were purchased from Roche Applied Science (Indianapolis, IN). All other reagents used in the present study were of analytical grade and were of highest purity available.

Cell Culture. TC3 cells were kindly provided by Dr. Shimon Efrat (Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel). Cells were cultured in Dulbecco's modified Eagle's medium containing 25 mM glucose supplemented with 10% fetal bovine serum, 100 IU/ml penicillin, 100 IU/ml streptomycin, and 2 mM L-glutamine under 95% O₂/5% CO₂ (Efrat et al., 1990) The medium was changed twice weekly and cells were trypsinized and subcloned weekly. Cells were used between passages 20 and 65.

Synthesis of 3-Allyl- and 3-Vinylfarnesols and Geranylgeraniols. These compounds were synthesized (Figs. 1 and 2) via our previously described vinyltriflate to isoprenoid analogs (Gibbs et al., 1999). Details of the specificity of these inhibitors, in terms of their ability to inhibit prenyl transferases, were described in detail in our previous publications (Gibbs et al., 1999; see Discussion).

View larger version (15K): [in this window] [in a new window]   Fig. 1.   Scheme for the synthesis of 3vFOH and 3alFOH. a, nBuLi; THF, 0°C; geranyl bromide, 0°C to room temperature. b, (Me3Si)2NK, THF, 78°C; AvN (SO2CF3)2, 78°C and brought to room temperature. c, vinylSnBu3, Pd (AsPh3)2, CuI, NMP, room temperature. d, allylSnBu3, Pd(AsPh3)2, CuI, NMP, 80°C. e, DIBAL-H, PhMe, 78°C.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Scheme for the synthesis of 3VGGOH and 3alGGOH. a, nBuLi; THF, 0°C; farnesyl bromide, 0°C to room temperature. b, (Me3Si)2NK, THF, 78°C; AvN (SO2CF3)2, 78°C and brought to room temperature. c, vinylSnBu3, Pd (AsPh3)2, CuI, NMP, room temperature. d, allylSnBu3, Pd(AsPh3)2, CuI, NMP, 80°C. e, DIBAL-H, PhMe, 78°C.

Insulin Release Studies. TC3 cells were cultured overnight in Dulbecco's modified Eagle's medium containing 5 mM glucose and 10% fetal calf serum in the presence of diluent alone or various concentrations of inhibitors as indicated in the text. After preincubation in the presence of 3.3 mM glucose, cells were then incubated in the presence of either 3 or 20 mM glucose for 45 min at 37°C. The supernatant was then removed, centrifuged at 300g for 10 min, and assayed for insulin. To assess insulin content, cells were extracted overnight in acid/ethanol mixture as described previously (Metz et al., 1993). The amount of insulin was quantitated by ELISA (American Laboratory Products Company; Kowluru et al., 2001).

Cell Viability Assay. TC3 cells were seeded at a density of 1 × 10⁶ cells/ml in round-bottomed 96-well plates and then treated with inhibitors (as described above). Cell viability was determined by a colorimetric assay (at 550-690 nm), using MTT, which measures the reduction of 3-(4,5-dimethylthiazolyl-2) 2,5-diphenyltetrazolium bromide into the blue formazan product by metabolically active cells.

Subcellular Distribution of G Proteins: GTP Overlay Assay. To identify the effects of our inhibitors on the hydrophobicity of G proteins, we used the [-³²P]GTP overlay assay. Experimental details for this method, which determine the relative abundance of the G proteins in total membrane and soluble fractions, have been described in Li et al. (1993), Kowluru et al. (1994), and Kowluru and Metz (1996). Briefly, homogenates of control and inhibitor-treated cells were centrifuged at 105,000g for 90 min in a TL-100 ultracentrifuge (Beckman Coutler, Inc., Fullerton, CA). The supernatant and pellets were designated as the cytosol and membrane fractions, respectively. They were separated by SDS-PAGE (12.5%) and transferred to a nitrocellulose membrane using a transblot apparatus (Bio-Rad, Hercules, CA). The blots were then washed in a medium containing 50 mM Tris-HCl pH 7.4, 5 mM MgCl₂, 1 mM EGTA, and 0.3% Tween 20 for 10 min. The membranes were incubated in the above-described medium containing [-³²P]GTP (1 µCi/ml) for 1 h at room temperature followed by extensive washing in the same buffer to remove unbound label from the membranes. Labeling of proteins was determined by autoradiography using an X-OMAT film (Eastman Kodak, Rochester, NY). Molecular weights of labeled proteins were determined using prestained, authentic molecular weight standards (Bio-Rad). Gel scanning was performed using UN-SCAN-IT software and Adobe Photoshop (Adobe Systems, Mountain View, CA).

	Results

Top Abstract Introduction Methods and Materials Results Discussion References

Insulin release elicited in the presence of either a stimulatory concentration (20 mM) of glucose or a depolarizing concentration (40 mM) of KCl was measured in TC3 cells in the absence or presence (0-20 µM) of 3-vinylfarnesol (3vFOH). Data in Fig. 3 indicate minimal effects of this compound on glucose-stimulated insulin secretion up to 10 µM. However, at 20 µM, glucose-induced insulin secretion was completely abolished by 3vFOH. Potassium-induced insulin release was inhibited by 80% in the presence of 10 µM 3vFOH, and a comparable degree of inhibition was demonstrable even at 20 µM of this compound. Data in Fig. 4 demonstrate inhibition by 3-allylfarnesol (3aFOH) of glucose- or KCl-induced insulin secretion from TC3 cells. At 10 µM 3aFOH, the glucose- and KCl-induced insulin secretion was inhibited by 83 and 94%, respectively. These data indicate potential regulation by protein farnesylation of glucose- and calcium-induced insulin secretion from isolated  cells.

View larger version (32K): [in this window] [in a new window]   Fig. 3.   Inhibition by 3vFOH of glucose- or KCl-induced insulin secretion from TC3 cells. Cells were incubated (for 18 h) in the presence of varying concentrations (0-20 µM) of 3vFOH as indicated in the figure. After incubation in low glucose (3 mM for 45 min), cells were then incubated in Krebs-Ringer medium for 45 min in the presence of either 20 mM glucose or 40 mM KCl in the continuous presence of either the inhibitor or the diluent as indicated in the figure. Insulin was quantitated by ELISA (see Materials and Methods). Data are expressed as mean ± S.E.M. from three different experiments carried out in triplicate. , p < 0.05 versus the control value in the absence of inhibitor.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   Inhibition by 3aFOH of glucose- or KCl-induced insulin secretion from TC3 cells. Cells were incubated (for 18 h) in the presence of varying concentrations (0-10 µM) of 3aFOH as indicated in the figure. After incubation in low glucose (3 mM for 45 min), cells were then incubated in Krebs-Ringer medium for 45 min in the presence of either 20 mM glucose or 40 mM KCl in the continuous presence of inhibitor or diluents as indicated in the figure. Insulin was quantitated by ELISA (see Materials and Methods). Data are expressed as mean ± S.E.M. from three different experiments carried out in triplicate. , p < 0.05 versus the control value in the absence of inhibitor.

We next examined the effects of inhibitors of protein geranylgeranylation on glucose- and KCl-mediated insulin secretion. Data in Fig. 5 demonstrate significant attenuation in glucose- and calcium-induced insulin secretion by 3-vinyl-geranylgeraniol (3-vGGOH). We observed complete inhibition by 3-vGGOH (20 µM) on glucose-induced insulin secretion, whereas KCl-induced secretion was inhibited by >90% under similar experimental conditions. Comparable degrees of inhibition of KCl-induced insulin secretion were demonstrable in the presence of 20 µM 3-allyl geranylgeraniol (additional data not shown). These data indicate a requirement for protein farnesylation and geranylgeranylation modification steps in glucose- as well as calcium-induced insulin secretion from the TC3 cells.

View larger version (29K): [in this window] [in a new window]   Fig. 5.   Inhibition by 3vGGOH of glucose- or KCl-induced insulin secretion from TC3 cells. Cells were incubated (for 18 h) in the presence of varying concentrations (0-20 µM) of 3vGGOH as indicated in the figure. After incubation in low glucose (3 mM for 45 min), cells were then incubated in Krebs-Ringer medium for 45 min in the presence of either 20 mM glucose or 40 mM KCl in the continuous presence of either the inhibitor or diluent as indicated in the figure. Insulin was quantitated by ELISA (see Materials and Methods). Data are expressed as mean ± S.E.M. from three different experiments carried out in triplicate. *, p < 0.05 versus the control value in the absence of inhibitor.

To further solidify our observations that protein farnesylation and geranylation are critical to insulin secretion, we next studied the effects of two commercially available inhibitors of protein farnesylation and geranylgeranylation on glucose or KCl-induced insulin secretion. Manumycin, a natural substance isolated from Streptomyces parvulus, is a selective farnesyl pyrophosphate competitive inhibitor of farnesyl transferase (Tannous et al., 2001). At 10 µM concentration, manumycin A markedly reduced both glucose- and KCl-induced insulin release (Fig. 6), further confirming data described in Figs. 3 and 4. Likewise, GGTI-2147, a peptidomimetic inhibitor of geranylgeranyl transferases, at 20 µM inhibited glucose- and KCl-induced secretion by 95 and 65%, respectively (Fig. 7). Together, data in Figs. 3 to 7 clearly suggest that both farnesylation and geranylation of proteins are critical to glucose as well as calcium-induced insulin secretion TC3 cells.

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Inhibition by manumycin A of glucose- or KCl-induced insulin secretion from TC3 cells. Cells were incubated (for 18 h) in the absence or presence of manumycin (10 µM) as indicated in the figure. After incubation in low glucose (3 mM for 45 min), cells were then exposed in Krebs-Ringer medium for 45 min in the presence of either 20 mM glucose or 40 mM KCl in the continuous presence of either manumycin or diluent as indicated in the figure. Insulin was quantitated by ELISA (see Materials and Methods). Data are expressed as incremental response to either 20 mM glucose or 40 mM KCl and are expressed as mean ± S.E.M. from three different experiments carried out in triplicate. , p < 0.05 versus the control value in the absence of inhibitor.

View larger version (34K): [in this window] [in a new window]   Fig. 7.   Inhibition by GGTI-2147 of glucose- and KCl-induced secretion from TC3 cells. Cells were incubated (for 18 h) in the presence of varying concentrations (0-20 µM) of GGTI-2147 as indicated in the figure. After incubation in low glucose (3 mM for 45 min), cells were then incubated in Krebs-Ringer medium for 45 min in the presence of either 20 mM glucose or 40 mM KCl in the continuous presence of either GGTI-2147 or diluent as indicated in the figure. Insulin was quantitated by ELISA (see Materials and Methods). Data are expressed as mean ± S.E.M. from three different experiments carried out in triplicate. , p < 0.05 versus the control value in the absence of inhibitor.

To assess whether the effects of these inhibitors on insulin secretion are not due to their effects on insulin synthesis, we measured the total insulin content in control and inhibitor-treated (18 h; 10 µM) cells. Data from these studies indicated no major differences in insulin content between the control and inhibitor-treated TC3 cells. For example, these values represented 96 ± 5 and 109 ± 3% of control cells when they were exposed to 3aFOH and 3vFOH, respectively (mean ± variance from two experiments). However, a modest reduction (78 ± 3% of control; mean ± variance from two experiments) in insulin content was demonstrable in cells treated with 3vGGOH. These data indicate a possible modulatory role for geranylgeranylated proteins in insulin biosynthesis as well.

We also examined the effects of our inhibitors on the metabolic viability of  TC3 cells under the conditions they inhibited glucose- and KCl-stimulated insulin secretion. Data in Fig. 8 indicated no significant effects of either the allyl- or the vinylfarnesyl transferase inhibitors (0-20 µM) on the viability of these cells. A modest (16%), but insignificant (p = 0.33) loss in the viability of cells was demonstrable at 20 µM GGTI-2147. These data suggest that the inhibition of glucose- and KCl-induced insulin secretion by these inhibitors is largely due to their ability to inhibit protein prenylation, and not due to their nonspecific cytotoxic effects.

View larger version (20K): [in this window] [in a new window]   Fig. 8.   Lack of effects of farnesyl- and geranylgeranyl transferase inhibitors on metabolic cell viability of TC3 cells. Metabolic cell viability was determined using an MTT assay (see Materials and Methods for additional details) in TC3 cultured in the presence of various inhibitors for 18 h as indicated in the figure. Data are mean ± variance from two different experiments carried out in triplicate.

Previous studies (Li et al., 1993; Metz et al., 1993; Kowluru and Metz, 1996) have demonstrated that inhibition of protein isoprenylation by LOVA, an inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase, resulted in accumulation of unprenylated G proteins in the cytosolic fraction in normal rat islets and clonal  cells. Based on those findings, we proposed that inhibition of glucose-induced insulin secretion by LOVA might, in part, be due to the accumulation of critical G proteins in the cytosolic fraction, thereby impeding the interaction of the candidate G proteins with their membrane-associated effector proteins. To further confirm our hypothesis and to determine the putative modes of action of these inhibitors, we examined the subcellular distribution (i.e., membrane versus cytosolic) of the -cell G proteins in control and inhibitor-treated cells. To address this, isolated TC3 cells were exposed to these inhibitors overnight, and the relative abundance of GTP-binding proteins in the total soluble and membranous fractions was determined by the [-³²P]GTP overlay assay (see Materials and Methods for additional details). Data in Fig. 9 indicate that exposure of TC3 cells either to farnesols or geranylgeraniols results in significant increase in the abundance of soluble G proteins, as evidenced by increase in the labeling in the soluble fraction. A corresponding decrease in the abundance of G proteins in the membrane fraction was also demonstrable (Fig. 9). These data clearly indicate that treatment of isolated  cells with either the farnesols or geranylgeraniols results in abnormal subcellular distribution of the candidate G proteins, presumably interfering with interaction of these signaling proteins with their effector proteins in the membrane fraction. These data also imply that the abundance of isoprenylated proteins within the  cell is significantly decreased in cells treated with inhibitors of protein prenylation as evidenced by the accumulation of nonprenylated proteins in the cytosolic fraction (Fig. 9). We have provided evidence in our earlier studies that such an interaction of specific proteins (e.g., Cdc42) with their effector proteins (e.g., phospholipase C) is critical for physiological insulin secretion (Kowluru et al., 1996a).

View larger version (27K): [in this window] [in a new window]   Fig. 9.   Altered subcellular distribution of GTP-binding proteins in TC3 cells after exposure to 3vFOH or 3vGGOH; GTP overlay assay. Cytosolic and membrane proteins (30 µg/lane) from control or inhibitor-treated cells (as indicated in figure) were separated by SDS-PAGE and transferred to a nitrocellulose membrane. After transfer, the membrane was incubated in a binding buffer containing [-32P]GTP. After incubation, membranes were washed extensively and labeled bands were identified by autoradiography (see Materials and Methods for additional details). Data are representative of two experiments with similar results. CON, control; FTI, 3vFOH; GGTI, 3vGGOH; C, cytosol; and M, membrane.

	Discussion

Top Abstract Introduction Methods and Materials Results Discussion References

One of the main objectives of this study is to further examine the contributory roles of protein farnesylation and geranylgeranylation in glucose- and calcium-induced insulin secretion from isolated  cells. Our data clearly support original observations from our laboratory (Metz et al., 1993) and those of others (Li et al., 1993) suggesting that isoprenylation steps are critical to physiological insulin secretion. To the best of our knowledge, there have been only two studies that examined the contributory roles of protein isoprenylation in insulin secretion. Using LOVA, Li et al. (1993) first reported that inhibition of isoprenylation resulted in significant attenuation in bombesin and vasopressin potentiation of nutrient-induced insulin secretion from HIT-T15 cells. These investigators have also provided evidence to indicate abnormalities in subcellular distribution (i.e., significant accumulation of unprenylated proteins in the cytosolic fraction) of small G proteins in LOVA-treated cells. Based on these data, they proposed that protein isoprenylation plays a critical regulatory role in bombesin and vasopressin-induced insulin secretion from HIT-T15 cells. At the same time, independent studies from our laboratory have provided evidence to suggest that protein isoprenylation, carboxyl methylation, and fatty acylation steps play important regulatory roles in physiological insulin secretion from isolated rat islets (Metz et al., 1993). We have shown that pretreatment of isolated rat islets with LOVA resulted in significant inhibition of glucose-induced insulin secretion; such an inhibition was reversed by exogenous provision of MVA in the culture medium indicating that LOVA-induced inhibition of insulin secretion is due to its inhibition of MVA (and subsequent isoprenoid) biosynthesis. Under conditions in which LOVA inhibited insulin secretion, we have also demonstrated significant accumulation of unprenylated G proteins in the cytosolic fraction. Together, these studies provided initial evidence for the contributory roles of isoprenylation in insulin secretion.

The current studies form a logical extension to our original studies in the sense that they use more specific inhibitors of these pathways (i.e., farnesylation and geranylgeranylation) to study their individual roles in both glucose- and KCl-mediated secretion. We have been able to further establish the roles of these modification steps in insulin secretion by using commercially available inhibitors as well. It may also be mentioned that the purpose of the current studies was not to identify the -cell endogenous farnesylated and geranylgeranylated G proteins that are required for physiological insulin secretion. Several previous studies have identified the candidate G proteins that may be required for insulin secretion (Kowluru and Metz, 1994; Kowluru et al., 1994, 1996a,b,c, 1997a,b, 2000; Robertson et al., 1991; Regazzi et al., 1992; Leiser et al., 1995; Sharp, 1996; Lang, 1999; Kowluru and Amin, 2002). Instead, the purpose of the present studies was to synthesize more specific inhibitors to probe for the roles of protein farnesylation and geranylgeranylation in insulin secretion. In addition to their utility in studies that examined putative contributory roles of these modification steps in physiological insulin secretion (this study), we recently used these inhibitors to demonstrate the roles of farnesylated proteins (e.g., H-Ras) in the signaling cascade leading to interleukin-induced nitric oxide release from normal rat islets and clonal  (HIT-T15, RIN5F, and INS-1) cells (Tannous et al., 2001; Kowluru and Amin, 2002; Kowluru and Morgan, 2002). Therefore, these inhibitors might subserve the roles as valuable tools to study the regulatory function(s) of specific G proteins in various aspects of -cell function. It should be noted that the 3-allyl- and 3-vinylfarnesols and geranylgeraniol analogs have similar effects in the present study on insulin secretion in TC3 cells.

The key issue of the selectivity of the farnesylation and geranylgeranylation inhibitors has been addressed in some detail. Previous studies have demonstrated that 3-allyl-FPP, the active form of the prodrug 3-allylfarnesol, exhibits ~1600-fold selectivity for the inhibition of farnesyltransferase (FTase) versus geranylgeranyl transferase I (GGTase I) (Gibbs et al., 1999). In a similar manner, 3-vinyl-FPP, the active form of 3-vinylfarnesol, is an effective and tight-binding substrate for FTase, and exhibits no productive interaction with GGTase I (selectivity for FTase/GGTase I, ~600-fold). The in vitro selectivity of 3-allyl-FPP is reflected in the in vivo selectivity of 3-allylfarnesol, which blocks the growth of NIH3T3 cells transformed with the oncogenic variant of farnesylated H-Ras, but not the growth of NIH3T3 cells transformed with the oncogenic variant of geranylgeranylated H-Ras, consistent with a lack of intracellular effect on protein geranylgeranylation. Evaluation of 3-vinylfarnesol versus these two NIH3T3 cell lines demonstrated only modest selectivity of this compound for the farnesylated H-Ras cell line, which may be consistent with a potentially different mechanism of action for this compound (vide infra). A similar in vivo pattern was observed with the geranylgeraniol analogs and the two NIH3T3 cell lines; 3-allyl-geranylgeraniol exhibits excellent selectivity for the geranylgeranylated H-Ras cell line versus the farnesylated variant, whereas 3-vinylgeranylgeraniol exhibits only modest selectivity between the two cell lines. However, 3-allyl-GGPP and 3-vinyl-GGPP exhibit no selectivity for GGTase I in vitro. We have previously suggested that the observed cellular selectivity for the geranylgeraniol analogs is due to a low intracellular concentration of the natural GGTase I substrate GGPP, relative to the FTase substrate FPP. However, this has not been conclusively demonstrated, and thus we have also used the alternative GGTase I inhibitor GGTI-2147, which exhibits a >60-fold in vivo selectivity for the inhibition of protein geranylgeranylation (Vasudevan et al., 1999). It should be noted that the 3-allyl- and 3-vinylfarnesol and geranylgeraniol analogs have similar effects in the present study on insulin secretion. This may seem surprising in view of the fact that the vinyl analogs are alternative substrates. These differences may, in part, be due to differences in the cell types used (i.e., NIH3T3 cells versus the pancreatic  cells used in this study). However, recent preliminary unpublished data in our laboratory have suggested that 3-vinyl-FPP may be an alternative substrate with some proteins (such as H-Ras), but an inhibitor with other proteins (such as RhoB). These in vitro data suggest that the in vivo mechanism of action of the vinyl analogs is complex. However, the combined use of the allyl and vinyl analogs may provide more information on the significance of various prenylated proteins in signaling pathways.

Interestingly, in our previous studies, we have observed only a modest, but significant (30%) inhibition by LOVA elicited by depolarizing concentrations of potassium (Metz et al., 1993). In the present study, we observed a significant inhibition in KCl-induced insulin secretion by both FTI and GGTIs in TC3 cells. Even though these findings (i.e., inhibition of KCl-induced insulin secretion) are directionally similar, reasons underlying such pronounced differences (i.e., about 30% inhibition by LOVA in contrast to >80-90% inhibition by farnesols and geranylgeraniols) remain unclear at this time. It may be due to the specificity of the inhibitors used in the present study as well as differences in cell types used in the two studies (i.e., rat islets in LOVA experiments and TC3 cells in this study). Our data indicate relatively a greater degree of inhibition of KCl-induced insulin secretion by these inhibitors compared with their effects on glucose-induced insulin secretion. Based on these data, we speculate that isoprenylated proteins might exert differential contributory roles in insulin secretory responses elicited by glucose or KCl.

Several previous studies have demonstrated regulatory roles for both trimeric and small molecular weight G proteins in insulin secretion (Robertson et al., 1991; Sharp, 1996; Kowluru et al., 2000). Most of these studies used pharmacological inhibitors or select bacterial toxins to identify these two classes of G proteins. For example, using acetyl farnesylcysteine, an inhibitor of prenyl cysteine methyl transferases, it has been shown that carboxyl methylation of  subunits of trimeric G proteins (Kowluru et al., 1997a) as well as small G proteins, such as Cdc42, Rac, and Rap, may be involved in glucose- and calcium-mediated insulin secretion (Leiser et al., 1995; Kowluru et al., 1996a; Kowluru et al., 1997b). Likewise, using cerulenin, a specific inhibitor of protein palmitoylation, several previous studies have demonstrated that acylation of proteins is relevant to insulin secretion (Metz et al., 1993; Yajima et al., 2000; Straub et al., 2002). Some of the candidate proteins for palmitoylation include, but are not limited to, the Ras subfamily of small G proteins as well as the  subunits of trimeric G proteins. Furthermore, bacterial toxins have been used in previous studies to assign roles for these proteins in modulation of insulin secretion. For example, using clostridial toxins that specifically monoglucosylate and inhibit functions of Rho subfamily of small G proteins (Kowluru and Metz, 1994; Kowluru et al., 1997b), we demonstrated inhibition of both glucose- and KCl-induced insulin secretion from normal rat islets and clonal  cells. Some of the farnesylated proteins identified in the  cell include Ras G proteins (Kowluru et al., 2000; Tannous et al., 2001; Kowluru and Amin, 2002),  subunits of trimeric G proteins (Kowluru et al., 1996b), and the nuclear lamin B (Kowluru, 2000). Examples of geranyl geranylated proteins include Cdc42, Rac, and Rap (Regazzi et al., 1992; Leiser et al., 1995; Kowluru et al., 1997b, 2000).

In conclusion, our current studies further confirm and reinforce the postulation that both farnesylation and geranylgeranylation of proteins play important regulatory roles in physiological insulin secretion. These data, together with previous findings, suggest that inhibition of post-translational prenylation impedes the maturation, membrane association, and biological effects of at least a subgroup of these proteins, leading to inhibition of physiological insulin secretion.