The Role of G Protein Activation in the Toxicity of Amyloidogenic Aβ-(1–40), Aβ-(25–35), and Bovine Calcitonin

More than 16 different proteins have been identified as amyloid in clinical diseases; among these, β-amyloid (Aβ) of Alzheimer's disease is the best characterized. In the present study, we performed experiments with Aβ and calcitonin, another amyloid-forming peptide, to examine the role of G protein activation in amyloid toxicity. We demonstrated that the peptides, when prepared under conditions that promoted β-sheet and amyloid fibril (or protofibril) formation, increased high affinity GTPase activity, but the nonamyloidogenic peptides had no discernible effects on GTP hydrolysis. These increases in GTPase activity were correlated to toxicity. In addition, G protein inhibitors significantly reduced the toxic effects of the amyloidogenic Aβ and calcitonin peptides. Our results further indicated that the amyloidogenic peptides significantly increased GTPase activity of purified Gαo and Gαi subunits and that the effect was not receptor-mediated. Collectively, these results imply that the amyloidogenic structure, regardless of the actual peptide or protein sequence, may be sufficient to cause toxicity and that toxicity is mediated, at least partially, through G protein activation. Our abilities to manipulate G protein activity may lead to novel treatments for Alzheimer's disease and the other amyloidoses.

The amyloidoses are complex, multiform disorders characterized by the polymerization and aggregation of normally innocuous and soluble proteins or peptides into extracellular insoluble fibrils. More than 16 biochemically unique proteins, including transthyretin, ␣-synuclein, calcitonin, ␤ 2 -macroglobulin, gelsolin, amylin, and ␤-amyloid, have been isolated as the fibrillar components of disease-associated amyloid deposits (1)(2)(3)(4)(5). These proteins share no conserved primary structural motives or other structural homologies, but their fibrils all possess some common structural features (3,6,7). All amyloid fibrils contain ␤-sheet structures in which the polypeptide chains are orthogonally aligned in the fibril directions (8 -10). With Congo Red staining, amyloids show a green birefringence under polarized light, and under electron microscopy, their morphology consists of bundles of nonbranching, long filaments about 5-12 nm wide (4,5,11).
The most characterized amyloid-forming peptide is ␤-amyloid (A␤) 1 of Alzheimer's disease. The toxicity of A␤ has been directly linked to structure and amyloid content. In an aggregated state (containing fibrils, protofibrils, and low molecular weight intermediates), A␤ has been consistently shown to be toxic to neurons in culture (12)(13)(14)(15)(16)(17)(18). Although there is some disagreement as to the exact structure of the aggregated species associated with toxicity, whether it be a protofibril (14,19), a diffusible, nonfibrillar ligand (20), or some other low molecular weight intermediate (19), toxicity is associated with peptide structures that are part of the aggregation pathway associated with amyloid formation. In addition, A␤ neurotoxicity has been shown to be attenuated by Congo Red and rifampicin, which bind to and selectively inhibit the formation of A␤ amyloid fibrils (21)(22)(23)(24). Clearly, all of these observations imply a causal link between A␤ fibril formation and neurodegeneration. Various research groups have hypothesized potential molecular mechanisms of ␤-amyloid toxicity, but there is no consensus. Cellular responses to A␤ that have been postulated to result in toxicity encompass destabilization of calcium homeostasis, membrane depolarization, increased vulnerability to excitotoxins, increased membrane permeability due to free radical generation, blockage or functional loss of potassium channels, and direct disruption of membrane integrity (17,18,(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35)(36). The preceding plethora of observed biochemical responses to A␤ suggests that perhaps a more common, fundamental pathway is initially being activated and that this pathway subsequently diverges to produce many unique intracellular responses.
Analogous to A␤, calcitonin is another model amyloid peptide associated with medullary carcinoma of the thyroid (37)(38)(39)(40)(41)(42)(43). The fibrils of human calcitonin have also been shown to be neurotoxic (44 -46), suggesting that the amyloidoses may possess a shared mechanism of toxicity related to their secondary and macromolecular structures.
To explore if a common, fundamental mechanism of toxicity exists in the amyloidoses, we examined the structure-function relationships of several synthetic A␤ sequences, A␤-(1-40), A␤-(25-35), A␤-(1-16), and bovine calcitonin. We were able to manipulate the secondary and macromolecular structures of these peptides to produce stable amyloid and nonamyloid structures. With these model systems, we demonstrated that the peptides in an amyloid state (with high ␤-sheet content and the ability to bind Congo Red) altered G protein activity associated with both cell membrane extracts and purified G␣ subunits. We showed that the abilities of the peptides to induce GTPase activation were correlated with their toxicities, and the neurotoxicities of the peptides were attenuated by specific and nonspecific GTPase inhibitors. In addition, we demonstrated that significant GTPase activities were still induced even when the cell surface receptors were removed with a nonspecific protease. These results suggest that G protein activation, possibly induced via a protein-membrane interaction, plays an important role in the toxicity of A␤ and other amyloid-forming proteins.
A␤ Peptide Preparation-The A␤ peptides were prepared analogously to established methods in the toxicity and structural literature for forming ␤-sheet structures and fibril formations (47)(48)(49)(50). Stock solutions of 10 mg/ml were prepared by dissolving the A␤ peptides in 0.1% (v/v) trifluoroacetic acid in water. After incubating for 1 h at 25°C, the peptide stock solutions were diluted to concentrations of 0.5 mg/ml in sterile phosphate-buffered saline (PBS) (0.01 M NaH 2 PO 4 , 0.15 M NaCl, pH 7.4) with antibiotics. These solutions were rotated on a model RD4524 rotator (Glas-col, Terre Haute, IN) at 60 rpm at 25°C for 24 h. The peptides were then diluted to final concentrations of 20 M in sterile medium and rotated for an additional 24 h prior to being added to the culture wells or plates for the toxicity and GTP studies.
The Structural Characterization and Preparation of Bovine Calcitonin-Bovine calcitonin was directly dissolved in various solvents and buffers at concentrations of 40 and 80 M. CD measurements of the solutions were recorded 2-24 h later on a model 62DS spectrometer (Aviv Instruments, Lakewood, NJ) at 25°C using a bandwidth of 1.0 nm, a step interval of 0.5 nm, and an averaging time of 2 s. A 0.01-cm quartz cell was used for the far-UV (190 -250 nm) measurements. The instrument was calibrated using D(ϩ)-10-camphorsulfonic acid. Three scans each of duplicate samples were measured and averaged. Control buffer and solvent scans were run in duplicate, averaged, and then subtracted from the sample spectra. Spectra were analyzed using the secondary structural parameters reported by Chang (51) to ascertain the sample percentages of ␣-helix, ␤-sheet, ␤-turn, and random-coil.
To assess the presence of amyloid fibrils in the calcitonin solutions, Congo Red binding studies were performed. Congo Red dye was dissolved in PBS to a final concentration of 112 M. Congo Red absorbances of the calcitonin solutions and free dye controls were determined by adding Congo Red to a final concentration of 12 M and acquiring spectral measurements from 300 to 900 nm at 25°C on a model 420 UV-visible spectrophotometer (Spectral Instruments, Tucson, AZ) (52,53). Both the calcitonin solutions and the control solutions were allowed to interact with Congo Red for 1 h prior to recording their spectra. Congo Red difference spectra were calculated by subtracting the free dye absorbance from the calcitonin-dye absorbances.
Finally, bovine calcitonin was prepared in the following manner for the toxicity and GTP studies. Calcitonin was dissolved at a concentration of 1.5 mg/ml in either deionized water or a solution of 5 mM CaCl 2 and 1 mM MgCl 2 in water. These solutions were rotated at 60 rpm at 25°C for 24 h. Then the calcitonin solutions were diluted to 40 and 80 M with sterile medium and rotated for an additional 24 h prior to being added to the culture wells or plates for the toxicity and GTP studies.
Cell Culture-Human neuroblastoma SH-SY5Y cells (a gift of Dr. Evelyn Tiffany-Castiglioni, College of Veterinary Medicine, Texas A & M University, College Station, TX) were cultured in a humidified 5% (v/v) CO 2 /air environment at 37°C in minimum Eagle's medium supplemented with 10% (v/v) fetal bovine serum, 3 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 2.5 g/ml amphotericin B (fungizone). Likewise, rat pheochromocytoma PC12 cells (ATCC, Manassas, VA) were cultured in RPMI medium supplemented with 10% (v/v) horse serum, 5% (v/v) fetal bovine serum, 3 mM L-glutamine, 100 units/ml penicillin, 100 g/ml streptomycin, and 2.5 g/ml fungizone in a 5% (v/v) CO 2 /air environment at 37°C. For the GTP studies, cells were plated at densities ranging from 2.5 to 4 million cells in 35-mm tissue culture dishes. For the viability assays, cells were plated at a density of 1 ϫ 10 5 cells/well in 96-well plates. During both the GTP and viability studies, the peptides were added to the cells 24 h after plating.
Membrane Preparation for the GTPase Assay-After incubation with the peptides or controls for 30 min at 37°C, the PC12 or SH-SY5Y membranes were isolated using the widely accepted method of Seifert and Schultz (54). Cells were harvested with a cell scraper and collected by centrifugation (1600 ϫ g, 4°C, 20 min). Subsequently, they were washed with a buffer consisting of 10 mM triethanolamine (TEA) and 140 mM NaCl (pH 7.4) and disrupted by nitrogen cavitation in a 50 mM KH 2 PO 4 buffer with 100 mM NaCl, 3 mM EDTA, and 15 mM ␤-mercaptoethanol (pH 7.0). The nuclear portion of the cells was removed by a short centrifugation (1000 ϫ g, 4°C, 2 min), and membrane sedimentation was attained with a long centrifugation (15,000 ϫ g, 4°C, 60 min). The resulting membrane pellet was suspended in a 10 mM TEA/ HCl buffer (pH 7.4), and the total membrane protein content was measured with the BCA assay (55).
GTPase Assay-The GTPase activities of the PC12 and SH-SY5Y membranes were assessed similarly to procedures described previously (56 -60). Low affinity or nonspecific GTPase activity was measured by adding excess unlabeled GTP (50 M) to the aforementioned reaction mixture and conducting the reaction as described. Specific high affinity GTPase activity was calculated as the difference between the total GTPase activity in the absence of unlabeled GTP and the low affinity GTPase activity.
Pronase Experiments-Pronase studies were conducted analogously to established procedures (61)(62)(63). Pronase at a concentration of 3 mg/ml in serum-free RPMI was incubated with the plated PC12 cells for 1 h at 4°C. The cells were harvested with a cell scraper and collected by centrifugation (1600 ϫ g, 4°C, 20 min). Then the cells were thoroughly washed with PBS and centrifuged again (1600 ϫ g, 4°C, 20 min) prior to the addition of the peptides or controls for the GTPase assays. For the epinephrine control, 200 M epinephrine in serum-free RPMI was incubated with these Pronase-treated PC12 cells for 30 min prior to the membrane isolation step.
MTT Reduction Assay-SH-SY5Y and PC12 cell viability was measured using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) reduction assay. MTT is reduced by viable cells to form blue formazan crystals, and inhibition of this reaction is indicative of cellular redox alterations that could result in toxicity (64). The peptides were incubated with the SH-SY5Y and PC12 cells for 24 h, after which time MTT reduction was assessed. MTT was added to the culture medium to yield a final concentration of 0.5 mg/ml. The cells were allowed to incubate with the MTT for 4 h in a CO 2 incubator after which time 100 l of a 5:2:3 N,N-dimethylformamide/SDS/water solution (pH 4.7) was added to dissolve the formed formazan crystals. After 18 h of incubation in a humidified CO 2 incubator, the results were read using an Emax Microplate reader at 585 nm (Molecular Devices, Sunnyvale, CA). Viability is reported relative to control cells unexposed to the peptides.
GTPase and Toxicity Inhibition-Pertussis toxin (PT) (100 ng/ml), GDP␤S (600 M), and suramin (20 M) were incubated with the PC12 and SH-SY5Y cells for 24, 3, and 3 h, respectively, at 37°C prior to the peptide additions for the GTPase or toxicity assays. The peptide solutions for these assays also contained the same inhibitors at the same concentrations. Control cells were treated identically except for the presence of peptide. Statistical Analysis-The significance of results was determined using a Student's t test on n independent measurements, where n is specified in the figure legend. Unless otherwise indicated, significance was taken as p Ͻ 0.05.
Because the solution structures of bovine calcitonin are not as well documented, we identified conditions that promoted the formation of ␤-sheet structure and amyloid using CD spectroscopy and Congo Red binding assays. As determined by CD (Fig.  1A), both 40 and 80 M bovine calcitonin in water containing 5 mM CaCl 2 and 1 mM MgCl 2 adopted structures with ϳ55 Ϯ 10% ␤-sheet and only 15 Ϯ 10% ␣-helix. Incubating the peptide in deionized water alone at these same concentrations (Fig. 1A) produced structures devoid of ␤-sheet character with 95 Ϯ 10% ␣-helical contents. As depicted by Congo Red difference spectra (Fig. 1B), the 40 and 80 M water solutions of bovine calcitonin with 5 mM CaCl 2 and 1 mM MgCl 2 significantly bound and shifted the spectral properties of Congo Red, indicating the formation of amyloid. The zero difference spectrum of the peptide in deionized water indicated an absence of Congo Red binding and substantial amyloid fibril formation (spectrum not shown).
GTPase Activity-By using A␤-(1-40), A␤- (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), and bovine calcitonin in water with 5 mM CaCl 2 and 1 mM MgCl 2 as models of peptides with amyloidogenic structures and A␤-(1-16) and bovine calcitonin in deionized water as models of peptides without amyloidogenic structures, we examined the relationship between peptide structure and GTPase activity. We found that the rate of high affinity GTP hydrolysis in PC12 membranes increased by 31 Ϯ 12% on average with exposure to the peptides containing extensive ␤-sheet and amyloid contents relative to the rate of hydrolysis of control cells unexposed to peptides ( Fig. 2A). In all cases, the increases in GTPase activity were significant relative to the control cells (p Ͻ 0.001). The rates of GTP hydrolysis were significantly greater for A␤-(25-35) and 80 M calcitonin (p Ͻ 0.001), the two peptides with the greatest amyloid content, than for A␤-(1-40) and 40 M calcitonin, suggesting that the extent of macromolecular structure influences the process. Similarly, A␤-(1-16) and bovine calcitonin in deionized water, the peptides devoid of amyloid and ␤-sheet structures, did not significantly alter the GTPase activities of the PC12 cells relative to untreated controls ( Fig. 2A) To ensure that the observed phenomena were not isolated to PC12 cells, we examined the GTPase activities of SH-SY5Y membranes exposed to bovine calcitonin. We found similar trends to those observed with the PC12 cells (Fig. 2B). The rate of GTP hydrolysis increased from 16.0 Ϯ 0.5 pmol/mg/min for the control cells to 23.8 Ϯ 0.6 and 18.7 Ϯ 0.6 pmol/mg/min for the cells exposed to 80 and 40 M bovine calcitonin in water with 5 mM CaCl 2 and 1 mM MgCl 2 , respectively (p Ͻ 0.002). The nonamyloidogenic 80 M calcitonin in water did not significantly alter GTPase activity relative to the controls (16.0 Ϯ 0.3 pmol/mg/min, p Ͼ 0.4).
Toxicity-In parallel to the GTPase activity experiments, we examined the relationship between peptide structure and toxicity using our model peptides. As seen in Fig. 3, A and B, analogous to our GTPase results, we found that exposure to the peptides containing extensive ␤-sheet and amyloid contents resulted in significant PC12 (Fig. 3A) and SH-SY5Y (Fig. 3B) cell toxicity (p Ͻ 0.001). Conversely, A␤-(1-16) and bovine calcitonin in deionized water, the peptides devoid of amyloid and ␤-sheet structures, did not significantly alter cell viability relative to untreated controls (p Ͼ 0.2).
GTPase Inhibition-To identify the family or families of G proteins activated by the amyloid-forming peptides, we investigated the effects of GDP␤S and suramin, two nonspecific G determinations are depicted. Untreated PC12 or SH-SY5Y cells were exposed to the amyloids for 30 min at 37°C in a humidified 5% CO 2 environment in the absence of all pharmacological agents. Treated PC12 or SH-SY5Y cells were preincubated with 100 ng/ml PT, 20 M suramin, or 600 M GDP␤S for 24, 3, and 3 h, respectively, at 37°C in a humidified 5% CO 2 environment. Following this preincubation, treated cells were exposed to amyloidogenic peptide solutions containing the same inhibitors for 30 min at 37°C in a humidified 5% CO 2 environment prior to the GTPase membrane isolation step. PT (open bars), suramin (diagonally striped bars), and GDP␤S (cross-hatched bars) significantly reduced membrane GTPase activity relative to the cells exposed to the amyloids in the absence of these compounds (solid bars). * and ** indicate that the decreases in the rates of hydrolysis relative to the untreated cells were significant at p Ͻ 0.005 and p Ͻ 0.001, respectively. protein inhibitors (66 -73), and PT, a specific inhibitor of the G i and G o families of G proteins (74 -78). As illustrated in Fig. 4A, GDP␤S, suramin, and PT were each able to inhibit significantly the increases in GTP hydrolysis observed in PC12 membranes exposed to the amyloidogenic A␤-(1-40), A␤- (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), and bovine calcitonin (p Ͻ 0.005). Analogously, the inhibitors significantly reduced the rate of GTP hydrolysis induced by amyloidogenic calcitonin in membranes from SH-SY5Y cells (Fig. 4B, p Ͻ 0.001).
To demonstrate the effectiveness of Pronase treatment at receptor removal, we incubated cells with 200 M epinephrine and then measured the rate of GTP hydrolysis. Without Pronase treatment, 200 M epinephrine increased the rate of GTP hydrolysis in PC12 cells by 100% (data not shown). However, as seen in Fig. 5, after Pronase removal of cell receptors, incubation with epinephrine did not significantly alter the rate of GTP hydrolysis in the Pronase-treated cells relative to Pronase-treated control cells (p Ͼ 0.3). These results indicate that the cell receptors had been effectively removed by the Pronase.
G␣ o and G␣ i GTPase Assays-To demonstrate more specifically that the peptides were interacting with heterotrimeric G proteins, we performed GTPase assays with lipid vesicles containing purified G␣ o and G␣ i subunits (Fig. 6, A and B). Analogous to the cell membrane GTP results, the amyloidogenic A␤-(1-40), A␤- (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), and 80 M bovine calcitonin significantly increased GTPase activity in both the G␣ o and G␣ i vesicles (p Ͻ 0.001). In contrast, A␤-(1-16) and bovine calcitonin in deionized water, the peptides devoid of amyloid and ␤-sheet structures, did not significantly stimulate GTPase activity in either vesicle system at any of the experimental times. No receptors were included in the G␣ preparations.

DISCUSSION
In previous cell culture studies, the most important predictor of A␤ toxicity was the macromolecular state of the A␤ peptides with only aged or aggregated A␤ peptides (including fibrils, protofibrils, and/or low molecular weight intermediates) consistently eliciting toxic responses (12)(13)(14)(15)(16)(17)(18). Additionally, both the L-and D-enantiomers of A␤ exhibited nearly identical structural characteristics and induced similar levels of toxicity, implying that A␤ neurotoxicity was mediated by A␤ fibril features instead of any stereoisomer-specific interactions (79). Compounds such as Congo Red, rifampicin, and recognition peptides that bind to and/or inhibit the formation of amyloid fibrils have also been shown to attenuate the toxicity of A␤, further establishing the causal link between A␤ structure and function (21-24, 64, 80, 81).
Ample evidence exists that suggests that G protein activation and other signal transduction events such as phospholipase D and adenylate cyclase activation may be associated with the biological activity of A␤ (82)(83)(84). Also, the reported changes in K ϩ and Ca 2ϩ ion channel activity and changes in calcium homeostasis are all consistent with GTPase activation being an early event in the mechanism of action of A␤ (26,30,85,86). However, in very few of these studies has anyone shown the relationship between the biological activity and the structure of the peptide, which is essential in establishing the connection between activity and a toxicity mechanism.
Unlike our results, reports of A␤-induced barium conductances in N1E-115 neuroblastoma cells (87) and calcium fluxes (85,88) did not correlate aggregation state of the peptide with activity. When A␤ blockage of the fast-inactivating K ϩ current was investigated, structure-function relationships were examined, but no structure dependence upon ion channel activity was observed (30). In addition, the data associated with the free radical model of A␤ toxicity such as the ability of A␤ . Untreated PC12 and SH-SY5Y cells were exposed to the peptides for 24 h at 37°C in a humidified 5% CO 2 environment in the absence of all pharmacological agents. Treated PC12 or SH-SY5Y cells were preincubated with 100 ng/ml PT, 20 M suramin, or 600 M GDP␤S for 24, 3, and 3 h, respectively, at 37°C in a humidified 5% CO 2 environment. Following this preincubation, treated cells were then exposed to amyloidogenic peptide solutions containing the same inhibitors for 24 h at 37°C in a humidified 5% CO 2 environment cells. The data are reported as the percentage of the MTT reduced by the cells incubated with the peptides alone or with the peptides and inhibitors relative to the MTT reduced by control cells unexposed to the peptides. The means Ϯ S.D. of 8 determinations are presented. In the absence of pharmacological reagents, exposure to the amyloid-forming peptides led to a significant reduction in cell survival relative to the control cells (p Ͻ 0.0001). Incubation of the PC12 and SH-SY5Y cells with PT (open bars), suramin (diagonally striped bars), and GDP␤S (cross-hatched bars) protected them from amyloid-induced cell death relative to untreated cells exposed to the amyloids (solid bars). Incubation of the cells with inhibitors in the absence of the amyloid peptides had no significant effect on cell viability (Control Cells). * and ** indicate that the increases in cell viability relative to the untreated cells exposed to the amyloids were significant at p Ͻ 0.0005 and p Ͻ 0.0001, respectively. peptides to generate EPR signals were not structure-specific (31).
We suggest that amyloidogenic A␤-induced G protein activation could be an early step in the molecular level mechanism of A␤ toxicity and that activation of alternative G protein pathways could produce many of the observed diverse cellular responses. For example, in our studies, PT had significant inhibitory effects on GTPase activity and toxicity, indicating that the G i and G o families of G proteins were being activated (74 -78). The results of our purified G␣ o and G␣ i subunit GTPase assays confirm this hypothesis. The amyloidogenic A␤-(1-40) and A␤- (25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35) significantly increased the GTPase activity of both the G␣ o and G␣ i subunits. Activation of these particular families of G proteins could account for some of the previous ion channel results because they have been linked to certain K ϩ and Ca 2ϩ channel modulations (89 -93). Our results do not preclude the activation of multiple GTPases, which could result in tremendously diverse intracellular phenomena since G proteins and their effectors have been associated with selective protein phosphorylation, gene transcription, cytoskeletal reorganization, secretion, and membrane depolarization (71, 74, 92, 94 -96).
In both our Pronase experiments and in our G␣ o and G␣ i subunit experiments, we demonstrated that the presence of receptors was not necessary for the observed amyloidogenic peptide-induced increases in GTPase activity. In the Pronasetreated cells, few receptors remained, as evidenced by the absence of an epinephrine-induced increase in GTP hydrolysis. Similarly, no receptors were included in the G␣ o and G␣ i subunit preparations. However, in both systems, significant peptide-induced increases in GTPase activity were observed. Nonreceptor mediated GTPase activation has been documented with mastoparan, ethanol, and shear stress (59,(97)(98)(99). The discovery that our A␤ GTPase activities may be membranemediated is consistent with a number of other previous findings, which imply the importance of A␤-membrane interactions to neurotoxicity. A␤ has been shown to interact with the lipid bilayer of the plasma membrane, forming cation-selective channels and disrupting ion homeostasis (25,26). Interaction with the plasma membrane may also influence aggregation and amyloid formation of the peptide (33,34,100,101).
In conjunction with our A␤ GTPase results, our calcitonin results suggest that the amyloidoses may share some common steps in the mechanism of toxicity. Analogous to our A␤ findings, we found that amyloidogenic bovine calcitonin in water containing 5 mM CaCl 2 and 1 mM MgCl 2 increased cell toxicity and increased GTPase activities associated with PC12 membranes and G␣ o and G␣ i vesicles, whereas the nonamyloidogenic calcitonin in deionized water did not significantly alter cell viability or G protein activities of PC12 membranes or G␣ o and G␣ i vesicles. Amyloidogenic bovine calcitonin increased the GTPase activity in Pronase-treated cells and G␣ o and G␣ i vesicles devoid of receptors, and inhibition of GTPase activation attenuated cell toxicity. These results again suggest the potential importance of protein-membrane interactions to amyloidmediated toxicity and indicate that GTPase activation is an early step in the toxicity of amyloid-forming peptides.
In conclusion, we demonstrated that A␤ and calcitonin peptides altered G protein activity in a structure-specific manner; only the peptides with extensive ␤-sheet and amyloid contents significantly increased GTPase activity. Both G o and G i activation was observed. At least some of the observed amyloidinduced increases in GTPase activity were not receptor-mediated, pointing to the potential importance of peptidemembrane interactions in the biological activity of the peptides. We demonstrated that the observed increases in G protein activity were linked to neurotoxicity by showing that G protein inhibitors significantly reduced the neurotoxic effects of the amyloidogenic A␤ and calcitonin peptides. These results may help to elucidate the mechanism of toxicity and may lead to novel treatments for the 16 or more diseases associated with amyloid proteins.