Introduction

Abstract Introduction Methods Results Discussion References

Both humans and laboratory animals demonstrate a long-lasting vulnerability to AMPH. Repeated use in humans can lead to a psychotic state that clinically resembles paranoid schizophrenia (Davis and Schlemmer, 1980; Kramer et al., 1967; Sato, 1986). A behavioral sensitization to AMPH develops in animals that is characterized by a more rapid onset of stereotyped behavior and more intense stereotyped movements than in controls and a marked increase in AMPH-induced rotational behavior (Robinson, 1991; Robinson and Becker, 1986; Segal and Kuczenski, 1994). The behavioral manifestations of sensitization can persist for 1 year after withdrawal from AMPH (Paulson et al., 1991). Although the neurochemical underpinnings for the expression of behavioral sensitization to AMPH are not entirely known, studies have shown that behavioral sensitization to AMPH is accompanied by an enhanced ability of DA to be released by stimuli, such as AMPH, K⁺ depolarization and electrical stimulation (Castañeda et al., 1988; Kolta et al., 1985; Robinson and Becker, 1982; Vezina, 1993; Wolf et al., 1993; Yamada et al., 1988;). The enhanced release of DA has been demonstrated in rat nucleus accumbens and striatum using both slice preparations and in vivo microdialysis. Enhanced release of DA in response to AMPH- and Ca⁺⁺-dependent stimuli such as potassium depolarization (Castañeda et al., 1988) develops after withdrawal from AMPH and is persistent (Paulson and Robinson, 1995).

We have reported increases in content of the Ca⁺⁺ binding protein, CaM, and phosphorylation of two CaM-binding proteins, synapsin I and neuromodulin, in rat striatum after several regimens of repeated AMPH treatment (Gnegy et al., 1991; Iwata et al., 1996; Roberts-Lewis et al., 1986). Neuromodulin and synapsin I, as well as CaM, have been postulated to play a role in exocytosis and neurotransmitter release from synaptosomes (Dekker et al., 1989b; Greengard et al., 1993; Hens et al., 1995, 1996; Llinas et al., 1991; Nichols et al., 1992). Synapsin I is a CaM-binding protein (Hayes et al., 1991) that binds to the cytosolic surface of synaptic vesicles and to various cytoskeletal proteins such as F-actin, microtubules, neurofilaments and spectrin (for review, see Dekker et al., 1989a; Greengard et al., 1993; Valtorta et al., 1992). Phosphorylation of synapsin I, especially at sites 2 and 3, by CaM kinase II leads to a decrease in its affinity for synaptic vesicles as well as for the cytoskeleton. Relieving the vesicles of cytoskeletal constraints could increase mobility of synaptic vesicles in the terminal. Neuromodulin (B-50, GAP-43, F1, pp46) is a neural-specific protein that binds CaM, actin and G_o (Coggins and Zwiers, 1991; Skene, 1989). PKC-mediated phosphorylation of neuromodulin at Ser⁴¹, which adjoins the CaM-binding region, results in the dissociation of bound CaM (Alexander et al., 1987). Phosphorylation at this site appears to be important for neurotransmitter release (Dekker et al., 1989b; Hens et al., 1995). We found that pretreatment of rats with repeated, intermittent AMPH resulted in an increase in the stoichiometry of synapsin phosphorylation at site 3 and of levels of neuromodulin phosphorylated at Ser⁴¹. These results suggested that there could be an increase in phosphorylating activity or a decrease in phosphatase activity in striatal terminals resulting from the repeated, intermittent AMPH.

In this study, we examined whether the enhancement in AMPH-induced DA release could be detected in striatal synaptosomes from rats pretreated with repeated, intermittent AMPH. We also examined the content of site 3-phospho-synapsin I and Ser⁴¹-neuromodulin in striatal synaptosomes using site-specific antibodies. In addition to using the phosphorylation state of the proteins to assess phosphorylating activity, we directly measured PKC and CaM kinase II activity in the synaptosomes. We found significant increases in the content of site 3-phospho-synapsin I and Ser⁴¹-neuromodulin in striatal synaptosomes from AMPH-treated rats after incubation at 37°C. The increase in site 3-phospho-synapsin I could be due to the enhanced CaM kinase II activity that was detected in the synaptosomes from AMPH-treated rats.

	Methods

Abstract Introduction Methods Results Discussion References

Repeated AMPH regimens. Female Holtzman rats (Harlan Sprague-Dawley, Indianapolis, IN) were used in these studies. Two regimens of repeated, intermittent AMPH were used, both of which have been shown to result in behavioral sensitization and enhanced AMPH-induced DA release on the basis of the use of either microdialysis or striatal slices (Robinson and Becker, 1982; Robinson and Camp, 1987). In repeated AMPH regimen 1, rats were treated using an escalating dose regimen (Gnegy et al., 1991; Robinson and Camp, 1987). Briefly, animals were housed in groups of 6 to 10, and injections were given twice a day for 5 days with 10 to 12 hr separating the two injections; this was followed by 2 drug-free days. This schedule was repeated for 4 weeks. The rats received a total of 40 injections (20 injection days) according to the following schedule: injection days 1 to 3 (1.0 mg/kg), 4 to 5 (2 mg/kg), 6 to 7 (3 mg/kg), 8 to 9 (4 mg/kg), 10 to 11 (5 mg/kg), 12 to 14 (6 mg/kg), 15 to 17 (7 mg/kg) and 18 to 20 (8 mg/kg). The control group received an equivalent number of saline injections. The animals were killed 4 weeks after the last injection. In repeated AMPH regimen 2, rats were given intraperitoneal injections of 2.5 mg/kg AMPH once daily for 5 days. Rats were killed 10 days after the last dose of AMPH.

Dopamine release assay. For measurement of DA release, the striatum was homogenized in 10 volumes of a homogenization solution containing 0.32 M sucrose, 0.25 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 10 µM pepstatin A and 10 µM leupeptin, pH 7.4. Homogenate fractions were centrifuged at 1000 × g for 10 min. The pellet was washed, and the combined supernatants were centrifuged at 15,000 × g for 15 min. The P2 fraction was resuspended in 550 µl of KRB, which was composed of 142 mM NaCl, 5.6 mM KCl, 1.0 mM MgCl₂, 1.7 mM CaCl₂, 24.9 mM NaHCO₃, 10 mM glucose, 1.14 mM ascorbic acid and 30 mM HEPES, pH 7.4, and oxygenated with 95% O₂/5% CO₂ for 20 min. P2 fractions (200 µl) were placed on a glass-fiber filter in a chamber of a Brandel superfusion apparatus and then perfused with KRB at 100 µl/min. Samples were collected at 5-min intervals. After a stabilization period (25 min), a 2.5-min bolus of 10 µM AMPH was perfused through the sample. The stimulation was terminated by replacing AMPH-containing buffer with fresh KRB. Collection was continued for an additional 40 min. The DA content in the superfusate was measured by high performance liquid chromatography with electrochemical detection using dihydroxybenzylamine as an internal standard.

Tissue preparation for phosphorylation and enzyme assays. The rats were killed by decapitation, and striatum was dissected on ice using a brain-cutting block as described by Heffner et al. (1980). In phosphorylation studies, either a crude S1 fraction, prepared as described below, or Percoll-purified synaptosomes (Dunkley et al., 1988a; Gnegy et al., 1993) were used. Briefly, striatum was homogenized in a glass-Teflon homogenizer in 10 volumes of homogenization solution containing 0.32 M sucrose, 2 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 10 µM pepstatin A and 10 µM leupeptin, pH 7.4. The homogenate fraction was centrifuged at 1000 × g for 10 min. The supernatant (S1) was layered on the Percoll (Pharmacia LKB Biotechnology, Piscataway, NJ) gradients, which were composed of 2 ml each of 23%, 15%, 10% and 3% Percoll (v/v) in sucrose solution. The gradient was centrifuged without a brake at 32,500 × g for 5 min. Fraction 4 was collected by aspiration, mixed with sucrose solution and then centrifuged at 15,000 × g for 15 min. The final pellet was resuspended in KRB.

Phosphorylation of neuromodulin and synapsin I. Phosphorylation was initiated by adding 45 µl of KRB to either 40 µg (15 µl) of the S1 fraction or 5 µg of Percoll-purified synaptosomes and then incubated. Incubations measuring neuromodulin phosphorylation were for 2 min at 37°C, and those measuring synapsin I phosphorylation were for 30 sec at 37°C. Reactions were terminated with 20 µl of SDS-stop solution (final concentration; 62.5 mM Tris·HCl, pH 6.8, 2% SDS, 10% glycerol, 5% 2-mercaptoethanol and 0.002% bromophenol blue) and boiled at 100°C for 5 min. Neuromodulin and synapsin I were separated by 10% SDS-PAGE and transferred onto nitrocellulose paper with a BioRad (Hercules, CA) minitransfer apparatus at 100 V. Neuromodulin required a 36-hr transfer for quantification, whereas synapsin I was transferred for 14 hr at 4°C. After transfer, blots for neuromodulin were immersed in blocking buffer (10 mM Tris, 150 mM NaCl, pH 7.4, 0.1% Tween 20, 1% bovine serum albumin) for 1 hr and then incubated with 2G12/c7 (to Ser⁴¹-phosphorylated neuromodulin) or 10E8/E7 (to total neuromodulin) (Meiri et al., 1991) for 1 hr. Immunoblots for total neuromodulin contained 20 µg of protein/lane. Blots were then incubated with anti-mouse IgG antibody at a 1:1000 dilution for 1 hr. Blots containing synapsin I were incubated with specific antibody to site 3- phosphorylated synapsin I (RU-19) (Czernik et al., 1995) at a 1:50 dilution for 3 hr. Immunoreactivity was visualized with ¹²⁵I-protein A (2 µCi). Total radioactivity was quantified by a PhosphorImager (Molecular Dynamics, Sunnyvale, CA), and densitometry values were obtained. Immunoreactivity was also visualized using an anti-rabbit IgG coupled to alkaline phosphatase (GIBCO, Gaithersburg, MD). Statistical significance was determined by Student's t test.

Measurement of CaM kinase II activity. Percoll-purified synaptosomes were prepared as described above, but the final synaptosomal pellet was lysed by resuspension for 15 min in a buffer containing 20 mM HEPES, pH 7.4, 0.5 mM EGTA, 1 mM EDTA, 10 mM sodium pyrophosphate, 0.4 mM ammonium molybdate, 0.25 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 10 µM pepstatin A and 10 µM leupeptin and sonication for 10 sec (Shearman et al., 1991). The lysed synaptosomes were centrifuged at 20,000 × g for 1 hr to obtain lysed supernatant and pellet. CaM kinase II activity was measured in the lysed supernatant and pellets. The activity in supernatant from lysed synaptosomes is synaptosomal and not due to possible adherent postsynaptic densities (Dunkley et al., 1988b). The assay, in a final volume of 50 µl, contained 25 mM HEPES buffer, pH 7.4, 10 mM MgCl₂, 50 µM ATP, 5 µCi of [³²P]ATP (specific activity, 4500 Ci/mmol), 20 µM autocamtide-3 (a specific substrate for CaM kinase II; Hanson et al., 1989), 1.5 mM EGTA and 1 µg of protein in the presence or absence of 1.7 mM CaCl₂ and 2 µg CaM. The assay was conducted for 30 sec at 37°C and stopped by pipetting of 30 µl onto P-81 filters (Whatman, Maidstone, UK), which were placed immediately in 75 mM H₃PO₄. Filters were washed three times in 75 mM H₃PO₄, dried and counted in ScintiVerse BD in a Beckman (Columbia, MD) LS 5800 Scintillation Counter.

PKC assay. PKC activity was measured in nonlysed synaptosomes, and supernatant and pellet of lysed synaptosomes prepared as described above. Because PKC activity could be attached to synaptosomes on postsynaptic membranes, PKC activity was also measured in nonlysed synaptosomes. The pellet obtained after lysing of the synaptosomes, and centrifuging at 100,000 × g was resuspended in 20 mM Tris, pH 7.4, containing 1 mM EDTA, 0.25 mM DTT and 0.1% Triton X-100. The pellet was sonicated twice for 5 sec (Shearman et al., 1991) and diluted in the above buffer without Triton X-100 so the protein was diluted 10-fold into the assay. PKC was assayed as described by Yasuda et al. (1990) in a total volume of 50 µl containing 20 mM Tris·HCl, pH 7.4, 5 mM magnesium acetate, 25 µM specific PKC substrate MBP_4-14 (Upstate Biotechnology, Lake Placid, NY), 10 µM ATP, 0.5 µCi of [³²P]ATP (specific activity, 4500 Ci/mmol), 0.5 µg phosphatidylserine, 0.1 µg diolein and either 0.5 mM CaCl₂ or 1 mM EGTA. Protein amounts per assay were 1 to 2 µg for lysed supernatant, 8 to10 µg for lysed pellet and 2 to 5 µg for nonlysed synaptosomes. The reaction was conducted for 6 min at 30°C and stopped by pipetting of 30 µl onto P-81 paper and immediate placement in 75 mM H₃PO₄. Filters were washed three times in 75 mM H₃PO₄ and counted as described above. To calculate enzyme activity belonging to synaptosomal membranes, activity from the nonlysed synaptosomes, which would contain adherent postsynaptic densities, was subtracted from the total activity in the pellet fraction.

Immunoblotting of PKC and CaM kinase II in fractions from Percoll-purified synaptosomes. PKC and CaM kinase II immunoreactivity were measured in lysed synaptosomal cytosol and membranes using specific antibody to the  isozyme of PKC (generously donated by Dr. Karen Leach, Pharmacia & Upjohn Pharmaceutical, Kalamazoo, MI) and antibody to CaM kinase II  subunit (Boehringer-Mannheim, Indianapolis, IN). The predominant PKC isozyme in dopaminergic cells is the  isozyme (Tanaka and Saito, 1992; Yoshihara et al., 1991). Synaptosomal fractions (10 µg of synaptosomal supernatant/lane and 2.5 µg of synaptosomal pellet/lane) were subjected to SDS-PAGE (7.5% acrylamide). Proteins were electrophoretically transferred onto polyvinylidene difluoride membrane (Immobilon P; Millipore, Bedford, MA) for 1 hr at 100 V at 4°C in a Transphor Transfer Unit (Hoefer Scientific Instrument, San Francisco, CA). Blots were immersed in the blocking buffer for 1 hr and then incubated with the primary antibody diluted 1:500 for PKC and 1: 500 for CaM kinase II in blocking buffer for 1 hr. In regimen 1, immunoblots were detected with ¹²⁵I-goat antimouse. Immunoreactivity was visualized with ¹²⁵I-protein A (2 µCi). Total radioactivity was quantified by a PhosphorImager (Molecular Dynamics), and densitometry values were obtained. In regimen 2, ECL using horseradish peroxidase-conjugated goat anti-mouse (1:20,000) was used for quantification. Blots were scanned using a ScanJet IIC (Hewlett Packard) and analyzed using the IMAGE QUANT program (Molecular Dynamics). Volume integration was performed to obtain the total optical density within the rectangular area delineated. Background values were determined for each blot and were subtracted from the total density. Quantification was performed by the PhosphorImager. Statistical significance was determined by Student's t test. Protein concentration was determined by Bradford protein assay.

Materials. AMPH was purchased from The University of Michigan Laboratory of Animal Medicine. CaM was purified from bovine testes by the method of Dedman et al. (1977). Rice starch, polyethylene glycol 8000, rabbit Serum, Triton X-100, Lubrol PX and Tween 20 were obtained from Sigma Chemical (St. Louis, MO). Immobilon paper was from Millipore (Bedford, MA). The state-dependent antibody for phospho-Ser⁴¹-neuromodulin (2G12/c7) was the generous gift of Dr. Karina Meiri, Department of Pharmacology, SUNY Health Science Center, Syracuse, NY. The state-dependent antibody for site 3-phospho-synapsin I (RU19) was generously donated by Dr. Andrew Czernik, Laboratory of Molecular and Cellular Neuroscience, Rockefeller University (New York, NY). Antibody to the  isozyme of PKC was the generous gift of Dr. Karen Leach (Pharmacia & Upjohn).

	Results

Abstract Introduction Methods Results Discussion References

AMPH-induced DA release in striatal synaptosomes from rats treated with saline and repeated AMPH. AMPH-induced DA release was measured in striatal P2 synaptosomal fractions prepared from rats that had been pretreated with saline or escalating doses of AMPH (repeated AMPH regimen 1). After 2.5 min of perfusion with 10 µM AMPH, a significant increase in endogenous DA release was apparent in a striatal synaptosome preparation from rats treated with repeated AMPH compared with saline controls (fig. 1). The amount of maximum DA release in P2 synaptosomes from repeated saline- treated and repeated AMPH-treated rats was 46.1 ± 2.7 and 64.4 ± 5.6 pmol/mg of protein, respectively (P < .017, individual unpaired two-tailed t test). There was no significant difference in basal DA release.

View larger version (15K): [in this window] [in a new window]   Fig. 1.   AMPH-induced DA release in a P2 fraction from rats treated with escalating doses of AMPH (AMPH regimen 1) () or saline (). Rats were treated with either escalating dose of AMPH (n = 7) or saline (n = 6) for 4 weeks and then withdrawn 30 days. Endogenous DA release was measured as described in Methods. *, P < .05 compared with saline-pretreated control (Student's t test).

Measurement of site 3-phospho-synapsin I and Ser⁴¹-phosphoneuromodulin. Site 3-phospho-synapsin I and phosphoSer⁴¹-neuromodulin were measured in incubated S1 fractions from rats treated with repeated AMPH and saline-treated controls. After a 30-sec incubation in KRB, there was a significant 42% increase in immunoreactivity for site 3-phospho-synapsin I measured in the S1 fraction from repeated AMPH-treated rats compared with saline-treated controls (fig. 2A and table 1). The content of total synapsin I was not changed after repeated AMPH treatment (fig. 2B, table 1). We measured the content of site 3-phospho-synapsin I in nonincubated fractions and found almost no immunoreactivity (data not shown). This demonstrated that rephosphorylation of synapsin I occurred during incubation of the synaptosomes. An increase in the immunoreactivity for phospho-Ser⁴¹-neuromodulin was also detected in the S1 fractions from rats treated with repeated AMPH compared with saline controls (fig. 3A and table 2). Total neuromodulin content was not changed after repeated AMPH treatment (fig. 3B and table 2).

View larger version (28K): [in this window] [in a new window]   Fig. 2.   The effect of repeated, intermittent treatment AMPH on the levels of (A) site 3-phospho-synapsin I and (B) total synapsin. Rats were treated with saline (S) or escalating doses of AMPH (A) (AMPH regimen 1). Immunoblots were performed as described in Methods. There was 40 µg of protein in each lane measuring site 3-phospho-synapsin I and 20 µg of protein in each lane measuring total synapsin I.

View larger version (35K): [in this window] [in a new window]   Fig. 3.   The effect of repeated, intermittent treatment AMPH on the levels of (A) phospho-Ser41-neuromodulin and (B) total neuromodulin. Rats were treated with saline (S) or escalating doses of AMPH (A) (AMPH regimen 1). Immunoblots were performed as described in Methods. There was 20 µg of protein in each lane measuring site 3-phospho-synapsin I and total synapsin I.

AMPH-induced DA release and immunoreactivity for site 3-phospho-synapsin I in striatal synaptosomes using repeated AMPH regimen 2. The enhanced phosphorylation of synapsin I and neuromodulin in striatal fractions from repeated AMPH-treated rats could be due to a greater activity or content of CaM kinase II and PKC in nerve terminals. We examined directly the activities of CaM kinase II and PKC in synaptosomal preparations from rats treated with repeated AMPH. To ensure that our results were generalized to other AMPH treatment regimens, we also used a different, shorter AMPH pretreatment regimen (regimen 2), which results in behavioral sensitization and enhanced AMPH-induced DA release (Robinson and Becker, 1982). To ensure that the enhanced DA release was also detectable after this pretreatment, we measured AMPH-induced DA release in a P2 synaptosomal fraction from saline-pretreated rats and rats pretreated with AMPH regimen 2 (5 days of 2.5 mg/kg AMPH intraperitoneal and 10 days of withdrawal). As shown in figure 4, there was an increase in DA release in response to 1 µM AMPH from striatal P2 fractions from the AMPH-pretreated rats compared with saline-pretreated rats. In addition, the enhanced component of the release was Ca⁺⁺ dependent. Removal of Ca⁺⁺ from the medium abolished the enhancement of AMPH-induced release, but the amount of control AMPH-induced release was not altered. There was a significant increase in immunoreactivity for site 3-phospho-synapsin I in the Percoll-purified synaptosomes from rats pretreated with repeated AMPH regimen 2 (26.2 ± 0.8 O.D. units × 10³) compared with saline-pretreated rats (18.2 ± 1.9 O.D. units × 10³, P < .01, n = 4).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   AMPH-induced DA release from a striatal P2 fraction prepared from rats pretreated with AMPH (, ) or saline (, ). Female Holtzman rats were injected with saline or AMPH (2.5 mg/kg) once daily for 5 days and killed 10 days after the last injection. DA release in response to a 2.5-min bolus of 1 µM AMPH was measured in the presence (, ) or absence (, ) of 1.2 mM CaCl2 in the KRB. Arrow indicates addition of 1 µM AMPH (n = 2).

CaM kinase II activity and immunoreactivity after repeated AMPH. CaM kinase II activity was measured in supernatant fractions from lysed synaptosomes to ensure that synaptosomal CaM kinase II activity and not CaM kinase II on adherent postsynaptic membranes was being measured (Dunkley et al., 1988b). The lysed supernatant was centrifuged at 20,000 × g so synaptic vesicles would be retained in the supernatant. In the presence of Ca⁺⁺ and CaM, there was a significant 25% to 30% increase in CaM kinase II activity in synaptosomal supernatant from AMPH-pretreated rats after both treatment regimens (table 3). There was, however, no significant change in CaM kinase II immunoreactivity in synaptosomal supernatant when samples were analyzed by immunoblotting (table 3). When measured, there was no change in CaM kinase II activity in pellet fractions and no change in immunoreactivity in those fractions (data not shown). As shown in table 3, there was no change in immunoreactivity for CaM kinase II in the total synaptosomal fraction measured after AMPH treatment regimen 1. 

PKC activity and immunoreactivity. PKC activity was measured in the supernatant and pellet fractions of striatal synaptosomes from repeated AMPH- and SAL-treated rats. There was no significant change in PKC activity in either supernatant or pellet fractions from lysed synaptosomes after either treatment regimen 1 or regimen 2 (table 4). Similarly, there was no change in immunoreactivity of the  isozyme of PKC when samples were analyzed by immunoblotting in either whole synaptosomes (regimen 1) or subcellular fractions (regimen 2) (table 4).

	Discussion

Abstract Introduction Methods Results Discussion References

An increase in AMPH-induced DA release has been demonstrated in rat striatum and nucleus accumbens after a variety of repeated AMPH and metamphetamine regimens that result in behavioral sensitization (Castañeda et al., 1988; Kolta et al., 1985; Robinson and Becker, 1982; Vezina, 1993; Wolf et al., 1993; Yamada et al., 1988). The enhanced AMPH-induced DA release has been determined using in vitro slice preparations (Kolta et al., 1985; Robinson and Becker, 1982) or in vivo microdialysis (Paulson and Robinson, 1995; Wolf et al., 1993). In this study, we demonstrated that enhanced AMPH-induced DA release can also be demonstrated in perfused P2 striatal synaptosomes from AMPH-pretreated rats. This suggests that the molecular mechanism of the enhanced DA release remains viable in the nerve terminal on tissue homogenization and is not strongly dependent on intact neuroanatomical connections. In addition, our preliminary data suggest that the enhanced component of release is Ca⁺⁺ sensitive, unlike AMPH-induced DA release in control rats. Recent evidence has suggested that the enhanced AMPH-induced DA release in nucleus accumbens from AMPH- and cocaine-pretreated rats is Ca⁺⁺ sensitive (Hens et al., 1995; Pierce and Kalivas, 1997).

We previously found an increase in stoichiometry of site 3-phospho-synapsin I and in phospho-Ser⁴¹-neuromodulin (Iwata et al., 1996) in striatum from rats treated with regimens of repeated, intermittent AMPH, including the escalating dose regimen used in this study. We now demonstrate that an increase in site 3-phospho-synapsin I and phospho-Ser⁴¹-neuromodulin immunoreactivity is detected in incubated broken cell preparations and synaptosomes prepared from striata of AMPH-pretreated rats compared with saline-treated controls. In vitro phosphorylation of synapsin I at site 3 was increased by 40% in striatal fractions from rats treated with either regimen of repeated AMPH compared with saline-treated controls. There was no change in total synapsin I. These results correlate well with our in vivo study in which there was a 33% increase in the stoichiometry of site 3-phospho-synapsin I in AMPH-pretreated rats measured in whole striatum with no change in total synapsin I (Iwata et al., 1996). During homogenization of the striatum, synapsin I becomes dephosphorylated and is subsequently rephosphorylated on incubation by enzymatic activity in the synaptosome. We found that the same is true for neuromodulin (Gnegy et al., 1993). The demonstration of an increase in CaM kinase II activity in striatal synaptosomes from AMPH-sensitized rats provides an explanation for the concomitant increase in site 3-phospho-synapsin I in synaptosomes from AMPH-sensitized rats compared with saline controls. CaM kinase II activity was measured in the supernatant from lysed synaptosomes because it has been demonstrated that the major proportion of CaM-dependent protein kinase in synaptosomes is soluble (Dunkley et al., 1988b). A significant proportion of CaM kinase II in synaptosomes is on the outside of the synaptosome, probably on associated postsynaptic densities. The  subunit of CaM kinase II is also localized on synaptic vesicles, which would be contained in the supernatant from lysed synaptosomes because centrifugation was at 20,000 × g. Sites 2 and 3 of synapsin I are phosphorylated by the  subunit located on synaptic vesicles (Benfenati et al., 1992). The reason for the increase in CaM kinase II activity is not known, but there was no measurable increase in content of the CaM kinase II  subunit. There could be a change in amount of  subunit because there was an increase in  but not  subunit of CaM kinase II in rat hippocampal homogenates after induction of long-term potentiation (Fukunaga et al., 1995). However, we were unable to detect the  subunit of CaM kinase II in striatal synaptosomes using a specific antibody. Neither the 20,000 × g supernatant, pellet or whole synaptosomal fractions showed any change in content of CaM kinase II  subunit after AMPH pretreatment. The reason for the increase in activity is unclear because exogenous Ca⁺⁺ and CaM were added in the assay. There may be a change in localization or binding of CaM kinase II that allows more activity or a change in dephosphorylation of the enzyme.

Similarly, a significant increase in phospho-Ser⁴¹-neuromodulin was detected in incubated striatal fractions prepared from rats pretreated with AMPH compared with saline controls. As with synapsin I, this could be due to an enhanced phosphorylation or decreased dephosphorylation. We did not find, however, any increase in PKC activity in either supernatant or pellet fractions from AMPH-sensitized rats. We previously reported that there was no change in striatal PKC activity in any fraction in rats that had been treated twice weekly with 2.5 mg/kg AMPH intraperitoneally for 5 weeks and withdrawn 7 days compared with controls (Gnegy et al., 1993), nor was there any significant change in immunoreactivity for the  isozyme for PKC in either the supernatant, pellet or whole synaptosomal fractions. We tested for the  isozyme for PKC because this isozyme has been specifically localized in dopaminergic cells in the rat nigrostriatal system (Tanaka and Saito, 1992; Yoshihara et al., 1991). It is possible that another isozyme was increased in nerve terminals after repeated AMPH treatment or that there was a subtle change in PKC activity in a subset of synaptosomes that we did not detect. Giambalvo (1992) found, using a thiophosphorylation assay with ATPS, that AMPH increased PKC activity in vitro in synaptoneurosomes by increasing the affinity for Ca⁺⁺. Our assay contained higher concentrations of Ca⁺⁺, and we used ATP as a substrate in our assays instead of ATPS. Another possibility is that dephosphorylating activity is decreased as a result of repeated AMPH treatment. Neuromodulin is dephosphorylated by calcineurin (Liu and Storm, 1989) and by phosphatases 1 and 2A (Han et al., 1992). The activity of one of these enzymes could be significantly decreased.

Both CaM kinase II-mediated phosphorylation of synapsin I (Greengard et al., 1993; Llinas et al., 1991; Nichols et al., 1992) and PKC-mediated phosphorylation of neuromodulin (Dekker et al., 1989b; Hens et al., 1995) are postulated to enhance Ca⁺⁺-dependent neurotransmitter release. Whether these phosphorylated proteins contributed to the enhanced AMPH-mediated DA release in striatal synaptosomes from AMPH-pretreated rats is unknown. The fact that the enhanced component of AMPH-induced DA release is Ca⁺⁺ sensitive in AMPH-pretreated rats, however, suggests that these proteins could contribute to the enhanced AMPH-mediated release. In vitro, synapsin I binds reversibly to synaptic vesicles, promotes G-actin nucleation and polymerization and induces the formation of thick bundles of actin filaments (Greengard et al., 1993; Valtorta et al., 1992). Phosphorylation at sites 2 and 3 abolishes the interactions of synapsin I with synaptic vesicles and actin filaments. Recent studies support the hypothesis that synapsin I mediates the clustering of vesicles in the presynaptic terminal, which is required to sustain neurotransmitter release (Pieribone et al., 1995; Rosahl et al., 1995). Although AMPH increases DA release through an exchange diffusion that takes place at the uptake carrier (Fischer and Cho, 1979; Seiden et al., 1993), at certain concentrations AMPH can block uptake into the vesicle, and vesicles have been shown to play a role in AMPH-mediated DA release (Floor and Meng, 1996). This suggests that increased vesicle traffic near the transporter could provide enhanced AMPH-induced release by increasing an available cytoplasmic and vesicular pool of DA (Robinson, 1991). In support of this concept, AMPH-sensitized animals also showed enhanced DA release in response to high K⁺ and electrical stimulation, both of which evoke vesicular release (Castañeda et al., 1988). In addition, some Ca⁺⁺-dependency of DA uptake through the transporter has been demonstrated (Uchikawa et al., 1995).

Similarly, an increase in phospho-Ser⁴¹-neuromodulin could have contributed to enhanced DA release. A correlation between PKC-mediated phosphorylation of neuromodulin and neurotransmitter release has been reported (Dekker et al., 1989b, 1990). Although it is unclear which property of neuromodulin is essential for neurotransmitter release, the phosphorylation state of neuromodulin (at Ser⁴¹) rather than PKC activity per se may be important for Ca⁺⁺-dependent release (Hens et al., 1993, 1995). Use of an antibody specific for the amino-terminal residues 39 to 43 of neuromodulin have shown that these residues play an important role in the release process, perhaps by serving as a local CaM store regulated by Ca⁺⁺ and phosphorylation (Hens et al., 1995). PKC-mediated phosphorylation of neuromodulin has been demonstrated to enhance Ca⁺⁺-dependent release of catecholamines in brain and of DA in PC12 cells (Ivins et al., 1993), but its effect on AMPH-induced DA release is unknown. We found that AMPH enhances phosphorylation of neuromodulin at the Ser⁴¹ site in striatum when given acutely in the animal (Iwata et al., 1996) and in vitro in incubated striatal synaptosomes (Iwata et al., 1997). Giambalvo (1992) reported that AMPH, at concentrations of >0.1 µM, enhanced PKC activity in synaptosomes by lowering the concentration of Ca⁺⁺ required for activation. Because PKC-mediated phosphorylation of neuromodulin leads to a dissociation of CaM, there could be more CaM available in the synaptosomal cytosol to contribute to Ca⁺⁺-dependent processes in AMPH-pretreated rats.

It is not known, however, in what population of synaptosomes the enhanced phosphorylation of neuromodulin or synapsin I is occurring. In the striatum, the predominant population of nerve terminals are glutamatergic, but there also are terminals for -aminobutyric acid, acetylcholine, serotonin and DA (see references in Walaas et al., 1988). It is not known whether these phosphorylation changes are occurring in dopaminergic synaptosomes or whether the release of DA is affecting phosphorylation in another population of synaptosomes. Both synapsin I and neuromodulin are widely distributed in the brain, and virtually all terminals have both proteins (Skene, 1989; Walaas et al., 1988).

In conclusion, we were able to demonstrate an enhanced AMPH-induced release of DA in striatal broken cell preparations and synaptosomes from rats treated with repeated AMPH. The results demonstrate that factors responsible for the enhanced AMPH-mediated release of DA in animals repeatedly treated with AMPH are contained within a synaptosomal preparation. In the incubated synaptosomes from AMPH-pretreated rats, there also were increased levels of phospho-Ser⁴¹-neuromodulin and site 3-phospho-synapsin I. The increased phosphorylation of synapsin I could be due to an enhanced activity of CaM kinase II. The increase in phosphorylation of neuromodulin could be due to a discrete activation of PKC that we were unable to measure or a decrease in dephosphorylation. The fact that the enhanced release of DA appears to be Ca⁺⁺ dependent suggests that either enhanced Ca⁺⁺-dependent phosphorylation in general or these proteins in particular could contribute to the enhanced AMPH-mediated DA release.