Deficits in dopamine clearance and locomotion in hypoinsulinemic rats unmask novel modulation of dopamine transporters by amphetamine


Address correspondence and reprint requests to Lynette C. Daws, University of Texas Health Science Center at San Antonio, Department of Physiology, 7703 Floyd Curl Drive, Mail Code 7756, San Antonio, TX 78229–3900, USA. E-mail:


Insulin affects brain reward pathways and there is converging evidence that this occurs through insulin regulation of the dopamine (DA) transporter (DAT). In rats made hypoinsulinemic by fasting, synaptosomal DA uptake is reduced. Interestingly, [3H]DA uptake is increased in hypoinsulinemic rats with a history of amphetamine self-administration. The possibility that amphetamine and insulin act in concert to regulate DAT activity prompted this study. Here we show that [3H]DA uptake, measured in vitro and clearance of exogenously applied DA in vivo, is significantly reduced in rats made hypoinsulinemic by a single injection of streptozotocin. Strikingly, amphetamine (1.78 mg/kg, given every other day for 8 days) restored DA clearance in streptozotocin-treated rats but was without effect on DA clearance in saline-treated rats. Basal locomotor activity of streptozotocin-treated rats was lower compared to control rats; however, in streptozotocin-treated rats, hyperlocomotion induced by amphetamine increased over successive amphetamine injections. In saline-treated rats the locomotor stimulant effect of amphetamine remained stable across the four amphetamine injections. These results provide exciting new evidence that actions of amphetamine on DA neurotransmission are insulin-dependent and further suggest that exposure to amphetamine may cause long-lasting changes in DAT function.

Abbreviations used



dopamine transporter



The dopamine transporter (DAT) is a primary site of action for drugs of abuse such as amphetamine and is critical in regulating dopamine (DA) neurotransmission by high affinity transport of DA into the nerve terminal. Understanding how amphetamine regulates the DAT is therefore of fundamental importance to studies of its abuse. In this regard, there is converging evidence that insulin can produce profound regulatory control of DAT activity. It is known, for example, that food restriction as well as experimentally induced diabetes have marked effects on behavioral responses to amphetamine (Marshall 1978; Carroll and Stotz 1983). The significance of these observations is underlined by the high co-morbidity of eating disorders and drug abuse (Wolfe and Maisto 2000; Grigson 2002). Moreover, there is a rapidly growing body of evidence that insulin plays a considerably greater role in regulating dopaminergic reward pathways in the central nervous system than was previously thought (see Figlewicz 2003).

Evidence that insulin may influence DAT function comes from reports that rates of DA uptake are decreased in rats made hypoinsulinemic by fasting (Patterson et al. 1998) and that insulin applied to cells stably transfected with the DAT increases [3H]DA uptake (Carvelli et al. 2002). Importantly, insulin can interfere with the action of amphetamine at the DAT and can prevent amphetamine-induced cell surface redistribution of the DAT away from the plasma membrane (Carvelli et al. 2002; Garcia et al. 2005). Of particular importance to the present study is a recent report by our group that rats trained to self-administer amphetamine prior to induction of diabetes with streptozotocin (STZ) show increased rather than decreased [3H]DA uptake into striatal synaptosomes (Galici et al. 2003). Although at first glance these data appear anomalous, they led us to hypothesize that insulin and amphetamine might act at a common site to affect DA neurotransmission and that the signaling pathway(s) involved in mediating the action of these compounds may represent a novel target for the development of new treatments for drug abuse.

To date there are no published reports of the role of insulin in controlling DAT activity in vivo or the potential impact of amphetamine in modifying this putative humoral regulation of the DAT. Because of the possible role of insulin in modulating the abuse potential of drugs, this study investigated the effect of repeated amphetamine treatment on DAT function in the presence and absence of insulin in vivo. Streptozotocin-induced diabetes was used as a model for hypoinsulinemia (Shimomura et al. 1990; Cheta 1998; Reagan et al. 1999).

Materials and methods


Male Sprague-Dawley rats (Harlan, IN, USA) weighing 250–300 g at the beginning of the experiments were treated with saline or made hypoinsulinemic by a single injection of STZ (50 mg/kg, i.p., Sigma). STZ destroys insulin-producing beta cells in the pancreas and has been used extensively to induce diabetes in rats (e.g. Carr et al. 2000; Galici et al. 2003). For a detailed description of its mechanism of action on beta cells in rat pancreas see Szkudelski (2001). To confirm the effectiveness of treatment, blood glucose levels were measured before treatment and again at each experimental end point. A rat was considered hypoinsulinemic when blood glucose exceeded 300 mg/dL. All animal procedures were approved by the local institutional animal care and use committee and were in strict accordance with the NIH Guide for the Care and Use of Laboratory Animals. All efforts were made to minimize both the number of animals used and stress or discomfort to the animals during experimental procedures.

All rats were housed individually and provided food and water ad libitum. Three groups of rats were treated with either STZ or saline. A time line for the study is shown in Fig. 1. One group was killed 9 days later for measurement of [3H]DA uptake into striatal synaptosomes. In another group, clearance of exogenously applied DA in striatum was measured in vivo 7 days following the injection of STZ or saline. The third group of rats was used to assess the effect of amphetamine on DA clearance in vivo and on locomotor activity in STZ- and saline-treated rats. Eleven days after STZ or saline treatment this group was divided into two further groups. One group received a short dosing regime of d-amphetamine hydrochloride (1.78 mg/kg, i.p., every other day for 8 days) and the other received time matched injections of saline. Locomotor activity was assessed for 30 min immediately following injection of amphetamine or saline and DA clearance was measured either 1 or 25 days following the final amphetamine injection.

Figure 1.

Timeline for treatment schedules, behavioral and neurochemical endpoints. Three groups of rats were used. Rats were treated with either streptozotocin (STZ) or saline (SAL) and then clearance of dopamine (DA) applied locally into the striatum was measured 7 days later (Group 1) or [3H]DA uptake into striatal synaptosomes measured 9 days later (Group 2). A third group of rats was treated with amphetamine (Amph, 1.78 mg/kg) or vehicle on days 11, 13, 15 and 17 and locomotor activity measured 30 min after injection. Clearance of DA applied locally into the striatum was measured 1 or 25 days after the final amphetamine or saline injection. These time points correspond to 18 and 42 days after the initial treatment with STZ or saline.

The dose and treatment schedule (every other day) for amphetamine administration was based on preliminary studies showing the development of robust conditioned place preference for amphetamine under these conditions. The assessment of DA clearance and [3H]DA uptake at 7-day and 9-day time points following administration of STZ were selected primarily because blood glucose levels are stable and consistently high by this time and therefore the effectiveness of the STZ injection can be confirmed. These time points were also selected because staggering the time line between collecting data for clearance of locally applied DA in vivo and [3H]DA uptake in vitro afforded a ‘replication’ of the effect of STZ to inhibit DAT function. Thus, a robust effect of STZ to inhibit DAT function was established at two different time points by two different methodologies prior to commencement of further pharmacological manipulation.

In vitro[3H]dopamine uptake into synaptosomes

Rats were given either an injection of STZ or saline vehicle and 9 days later striatal synaptosomes from these animals were prepared and [3H]DA uptake measured according to a modification of the method of Fleckenstein et al. (1997). [3H]DA uptake was performed for two rats on each experimental day. Briefly, the brain was rapidly removed from each rat and striata dissected out on platforms embedded in ice and then weighed. The striata from each rat were homogenized in 20 volumes of ice-cold 0.32 m sucrose using 12 up-and-down strokes of a Teflon homogenizer in a thick-walled glass tube. Homogenates were centrifuged at 1000 g for 10 min at 4°C to give a nuclear pellet. Supernatants were removed and then centrifuged at 17 500 g for 30 min at 4°C, after which the supernatants were discarded and the pellets of crude synaptosomes were resuspended in ice-cold modified Kreb's assay buffer (126 mm NaCl, 4.8 mm KCl, 1.25 mm CaCl2, 1.4 mm MgSO4, 16.4 mm NaH2PO4, 11 mm glucose, 1.14 mm ascorbate, and 1 µm pargyline, pH 7.4) to a concentration of 20 mg of original tissue weight per ml.

Assays were conducted in modified Kreb's assay buffer (see above) at a final assay volume of 2 mL. Each assay tube contained 200 µL of synaptosomal tissue homogenate containing about 50 µg protein (approximately 2 mg of original striatal tissue). Protein concentration was determined using the Bio-Rad protein assay. [3H]DA uptake experiments were performed using duplicate samples for total and non-specific uptake at six concentrations of [3H]DA (7.81, 15.6, 31.3, 62.5, 125, and 250 nm). Non-specific uptake was determined in the presence of 1 mm non-radioactive (cold) DA. Tissue and [3H]DA was added to the sample tubes sequentially in a manner to maintain proper incubation times. After preincubation of tissue in the assay tubes for 10 min at 37°C, assays were initiated by addition of [3H]-DA. Samples were incubated at 37°C for 6 min, then the reaction was quenched by addition of 5 mL of ice-cold 0.32 m sucrose wash buffer to each tube, placement of the samples on ice, and then filtration through Whatman GF/B filters soaked previously in 0.05% polyethylenimine using a Brandel Cell Harvester. Filters were then washed rapidly three times with 5 mL ice-cold 0.32 m sucrose. Radioactivity trapped in filters was counted using a liquid scintillation counter and the counts converted to moles of DA. The affinity (Km) and maximal velocity (Vmax) for [3H]DA uptake were determined using Prism 4 software from Graph Pad. Km and Vmax of [3H]DA uptake are expressed as nm and pmol/(min mg protein), respectively. [3H]DA (approximately 40 Ci/mmol) was purchased from New England Nuclear (Boston, MA, USA). All other reagents were purchased from Sigma Chemical Co. (St. Louis, MO, USA).

In vivo chronoamperometric recordings of dopamine clearance

Clearance of exogenously applied DA from the striatum was measured in rats either 7, 18 or 42 days following treatment with STZ or saline. The 18 and 42 day groups correspond to recordings made 1 or 25 days following the final injection of amphetamine (Fig. 1). Clearance of exogenously applied DA was measured by high-speed chronoamperometry using the FAST-12 system (Quanteon, Nicholasville, KY, USA). A detailed description of the methods can be found in Daws et al. (2000). Carbon fiber electrodes were coated with Nafion® (5% solution, Aldrich Chemical Co, Milwaukee, WI, USA) to prevent interference from anionic substances in extracellular fluid and calibrated to DA in vitro. The detection limit for the measurement of DA in these experiments averaged 37 ± 5 nm (n = 67 electrodes). Rats were anesthetized by i.p. injection of chloralose (85 mg/kg) and urethane (850 mg/kg) and prepared for chronoamperometry. The electrode-micropipette recording assembly was lowered into the striatum (AP, +1.2; ML, +2.2; DV, −3.5 to −5.5) (Paxinos and Watson 1986). Barrels were filled with DA (200 µm, Sigma) dissolved in 0.1 m phosphate-buffered saline with 100 µm ascorbic acid added as an antioxidant (pH 7.3–7.4). Different volumes of DA were pressure ejected at 5-min intervals to deliver 5, 10, 20, 40 or 50 pmols DA. The order was randomized between rats. These delivery procedures and pmol amounts of DA were selected to allow determination of the maximal rate of DA clearance and also for the calculation of the apparent affinity of the DAT (KT) for DA in vivo (see Zahniser et al. 1999).

Locomotor activity

Locomotor activity was assessed by placing the rat in a 26 × 61 × 23 cm high plexiglass chamber located within sound-attenuating cubicles. Horizontal activity was measured with four pairs of infrared photobeams positioned 4 cm above the floor of the chamber. Each beam was placed 15 cm away from the next immediate photobeam and the two extreme photobeams were located 8 cm away from the floor sides. Recording began immediately after administration of amphetamine or vehicle and continued for 30 min.

Data analysis

[3H]DA uptake into synaptosomes was analyzed using a t-test for independent samples. For DA clearance, only data from animals in which electrode placement was confirmed to be within the striatum were used. Chronoamperometry data were analyzed using three signal parameters: (i) the maximal amplitude of the signals resulting from local application of DA; (ii) T80, the time (in s) for the signal to decline by 80% of the maximal amplitude; (c) clearance rate (Tc in nM/s), defined as the slope of the decay curve from 20–60% of maximal signal amplitude, i.e. the most linear portion of the decay. Maximal Tc and KT values were determined by fitting a rectangular hyperbola to a plot of Tc (nm/s) vs. DA (µm) (Prism 4.0, GraphPad). Data were analyzed by anova followed by Tukey–Kramer post hoc comparisons between treatment groups.

Repeated measures anova, followed by Tukey–Kramer post hoc test, was used to compare locomotion counts between groups and across trials. Body weight and blood glucose values were analyzed by anova followed by Tukey–Kramer post hoc comparisons between treatment groups. A two-tailed probability level of p < 0.05 was accepted as statistically significant for all the tests.


Blood glucose levels increase and body weight decreases in hypoinsulinemic rats

Blood glucose levels did not differ between any group of rats prior to injection of STZ or saline. Blood glucose levels were significantly elevated in STZ-treated rats compared to control animals regardless of whether amphetamine was administered and regardless of time elapsed after STZ-administration (i.e. 7, 9, 18 or 42 days). Blood glucose levels did not vary as a function of time elapsed after STZ or saline injection and so the data were pooled. In amphetamine naïve rats, blood glucose levels were 163 ± 57 mg/dL and 483 ± 47 mg/dL in control and STZ-treated rats, respectively. In rats that received amphetamine the corresponding values were 152 ± 22 mg/dL and 393 ± 45 mg/dL.

Body weights did not differ between groups prior to treatment with STZ or saline. At the conclusion of the study STZ-treated rats weighed significantly less than their saline-treated counterparts. There was no significant difference between amphetamine naïve and treated rats. In experiments where DA clearance was measured 1 day following behavioral testing, the pooled means for body weight were 350 ± 6 g and 279 ± 8 g for saline and STZ-treated rats, respectively. In experiments where DA clearance was measured 25 days following behavioral testing, the pooled means for body weight were 402 ± 4 g and 273 ± 29 g for saline and STZ-treated rats, respectively.

[3H]Dopamine uptake is decreased in hypoinsulinemic rats

As shown in Fig. 2, specific [3H]DA uptake was significantly reduced by approximately 45% in rats treated 9 days earlier with STZ compared to those given saline [Vmax: 36.2 ± 8.1 (n = 21) vs. 66.3 ± 8.5 (n = 19) pmol/(min mg protein), respectively, p < 0.025]. There was no difference in the Km for [3H]DA uptake between groups (Km: 54.8 ± 7.5 and 49.8 ± 14.0 nm for STZ- and saline-treated rats). Blood glucose values were 75 ± 4 mg/dL and 379 ± 33 mg/dL in saline- and STZ-treated rats, respectively (p < 0.05, t-test for independent samples).

Figure 2.

Specific [3H]dopamine ([3H]DA) uptake into striatal synaptosomes of rats treated 9 days previously with streptozotocin (STZ) (n = 21, solid circles) or saline (n = 19, open circles). Data are expressed as mean ± SEM.

Dopamine clearance in vivo is decreased in hypoinsulinemic rats

Shown in Fig. 3 is the kinetic profile for DA clearance (Tc) in the striatum of anesthetized rats that were treated 7 days earlier with STZ or saline. DA clearance was significantly reduced (58–69%) in STZ-treated rats compared to control rats. At the greatest amount of DA ejected (50 pmol), clearance rates were 110 ± 32 nm/s (n = 6) and 355 ± 44 nm/s (n = 7) for STZ- and saline-treated rats, respectively. The maximal Tc for DA clearance (determined by fitting the data to a rectangular hyperbola) in STZ-treated rats was 245 ± 59 nm/s and in saline-treated rats 583 ± 48 nm/s (p < 0.05). The KT values for DA clearance were 8.2 ± 7.3 µm and 11.8 ± 8.0 µm for STZ- and saline-treated rats, respectively, and were not significantly different.

Figure 3.

Dopamine (DA) clearance in striatum of anesthetized rats treated 7 days previously with streptozotocin (STZ) (n = 6, solid circles) or saline (n = 7, open circles). Data are expressed as mean ± SEM. *p < 0.05, 1-way anova followed by Tukey–Kramer post hoc comparisons.

At this point it is worth underlining key differences between in vitro and in vivo methods for assessing DA ‘uptake’. In intact brain, rapid passage of transmitter is not permitted and diffusion becomes a rate-limiting factor. Indeed, this is reported to be the reason why relatively low affinity (high Km) values are obtained when uptake kinetics are measured in vivo compared with in vitro systems (Near et al. 1988; see also Cragg and Rice 2004). Thus, it is important to keep in mind that chronoamperometry measures clearance of DA and not exclusively active uptake. The values reported here are consistent with the literature for both DA clearance in striatum (e.g. Hebert and Gerhardt 1999; Zahniser et al. 1999) and [3H]DA uptake in striatal synaptosomes (e.g. Fleckenstein et al. 1997; Page et al. 2004). Because the reduced rate of DA ‘uptake’ in STZ-treated rats was of a similar magnitude when determined either by more conventional methods (synaptosomal uptake) or using methods to index DA ‘uptake’in vivo (chronoamperometry), the remainder of our study employed the latter approach so as to determine the consequence of drug treatment in a system where neural circuitry remained fully intact.

Locomotor stimulant effect of amphetamine in hypoinsulinemic rats is restored to control levels over repeated amphetamine administration

Baseline locomotor activity was significantly lower in STZ-treated rats compared to saline-treated control rats (Fig. 4). A first injection of 1.78 mg/kg of amphetamine increased locomotion in control as well as STZ-treated rats. Over repeated every-other-day injections of amphetamine, locomotor-stimulating effects of amphetamine remained stable for control rats; however, the locomotor-stimulating effects in STZ-treated rats increased over successive injections such that there was no significant difference between groups after the fourth amphetamine injection (Fig. 4). Over repeated every-other-day injections of saline, locomotor activity did not differ from that of baseline and remained stable within both STZ and saline treatment groups. Locomotor activity in STZ-treated rats receiving saline injections was significantly lower than control rats across trials.

Figure 4.

Locomotor activity (counts per 30 min) of control (saline-treated, open circles) and hypoinsulimenic (streptozotocin-treated, closed circles) rats upon repeated every-other day treatment (conditioning trails) with 1.78 mg/kg amphetamine (for details see Fig. 1 legend). Data are mean ± SEM of 10 rats per group. *p < 0.05, compared to locomotion of control rats in the corresponding trial. STZ, streptozotocin.

Reduced dopamine clearance in hypoinsulinemic rats is normalized by repeated amphetamine administration

Shown in Fig. 5 are values for clearance rate (Tc) of 50 pmol DA in rats that received their final injection of amphetamine or saline either 1 or 25 days earlier (and correspond to those rats that received a single injection of STZ or saline either 18 or 42 days earlier, respectively). Regardless of days post-treatment (1 or 25), STZ-treated rats displayed significantly reduced rates of DA clearance compared to saline-treated control rats. Tc values in saline-treated rats were similar regardless of days post-treatment and regardless of whether they had prior exposure to amphetamine or not. Remarkably however, in STZ-treated rats that received amphetamine, Tc values returned to those of control rats at both the earliest time point measured after amphetamine (1 day) and persisted through to the longest time point measured after amphetamine (25 days).

Figure 5.

Dopamine (DA) clearance (Tc) in striatum of drug naïve or amphetamine (1.78 mg/kg, four injections every other day) exposed hypoinsulinemic (streptozotocin) rats compared to control (saline) rats. For details of treatment and time course see Fig. 1 legend. Data are Tc values for clearance of 50 pmol DA and expressed as mean ± SEM. The number of rats/group is shown in parentheses. *p < 0.05 from all other groups; two-way anova with Tukey–Kramer post hoc comparisons. amph, amphetamine; SAL, saline; STZ, streptozotocin.

Shown in Fig. 6 are the kinetic profiles for rats that had received their final injection of amphetamine or saline 25 days earlier. These data highlight the profound effect of STZ to decrease the rate of DA clearance in the striatum. Figure 6(a) shows the oxidation current, converted to a micromolar value using a calibration factor determined in vitro, produced by pressure-ejection of DA into the striatum of an amphetamine naïve rat. The decay of the signal is markedly slower in STZ-treated rats than in saline-treated controls. As summarized in Fig. 6(c), the rate of DA clearance is decreased by almost twofold in STZ-treated rats across the range of pmol amounts of DA tested. Figure 6(b and d) show the corresponding data sets for rats that received amphetamine. The most notable feature is the restoration of DA clearance to that recorded in saline-treated rats. The maximal rate for DA clearance and KT values are summarized in Table 1. The maximal rate for DA clearance was lower in amphetamine naïve STZ-treated rats compared to all other groups. The KT value was also greater in these rats than any other group but this difference failed to reach significance.

Figure 6.

Dopamine (DA) clearance in striatum of drug naïve (a and c) or amphetamine exposed (b and d) hypoinsulinemic (streptozotocin) rats compared to controls (saline) (see Fig. 1 legend for details). (a) Shows the oxidation current (converted to a micromolar concentration using a calibration factor determined in vitro), produced by pressure ejection of 40 pmol of DA into the striatum of an amphetamine naïve streptozotocin-treated rat (squares) and saline-treated rat (circles). (b) Shows comparable data from a streptozotocin- and saline-treated rat subsequently given amphetamine (1.78 mg/kg, i.p., every other day for 8 days). (c and d) Show the corresponding summary data where the rate of DA clearance is plotted as a function of increasing DA concentration. All data are from rats that had undergone behavioral testing 25 days earlier. Data are expressed as mean ± SEM. The number of rats per group was 6–8. The curves represent the fit of a rectangular hyperbola to each data set. amph, amphetamine; sal, saline; STZ, streptozotocin.

Table 1.  Kinetic parameters for dopamine clearance in striatum of drug naïve or amphetamine exposed hypoinsulinemic (streptozotocin-treated) rats compared to controls (saline-treated)
 Clearance ratemax (nm/s)KTm)
  1. Tcmax and KT values were derived by fitting a 1-site rectangular hyperbolic function to a plot of clearance rate (nM/s) vs. peak DA signal amplitude (µM) attained for each pmol amount of DA delivered locally into striatum.

  2. AMPH, amphetamine; DA, dopamine; STZ, streptozotocin.


Changes in the Tc values for DA clearance were also reflected in the T80 time course for DA clearance. The T80 value was significantly longer in amphetamine naïve STZ-treated animals than in any other group. T80 values for clearance of the greatest amount of DA ejected (50 pmol) are summarized in Table 2. There was no significant difference in the signal amplitude produced by pressure-ejection of DA into the striatum between treatment groups. Signal amplitudes increased in a linear fashion with increasing amounts of DA. The average signal amplitudes ranged from 0.5 µm after 5 pmol DA to 10.5 µm after 50 pmol DA were pressure-ejected into striatum.

Table 2. T80 time course parameter for dopamine clearance in striatum of drug naïve or amphetamine exposed streptozotocin- and saline-treated rats
Days after final injection of AMPH or salineT80 for DA clearance (s)
SalineSTZSaline +  AMPHSTZ +  AMPH
  1. The number of rats/group is shown in parentheses. Days after final amphetamine or saline injection correspond to 18 and 42 days after receiving a single injection of either STZ or saline. Data are T80 values for clearance of 50 pmol DA and expressed as mean ± SEM.

  2. p < 0.05 from all other groups; two-way anova with Tukey–Kramer post hoc comparisons.

  3. AMPH, amphetamine; DA, dopamine; STZ, streptozotocin.

135 ± 4 (6)52 ± 10* (6)32 ± 3 (6)31 ± 3 (6)
2530 ± 10 (7)50 ± 9* (7)24 ± 4 (7)35 ± 8 (8)


These data provide exciting new support for the hypothesis that insulin-related pathways modulate dopaminergic striatal systems in brain, and that insulin may act at a site common to drugs of abuse such as amphetamine. Here we demonstrate that hypoinsulinemia, produced by a single injection of STZ, causes a marked reduction in the capacity of DA to be cleared from extracellular fluid in striatum and that this deficit can be restored by a modest amphetamine dosing regimen. Moreover, relative to control rats the reduced level of hyperactivity observed in hypoinsulinemic rats after one exposure to amphetamine was restored to levels observed in control rats by the fourth injection of amphetamine. Interestingly, in control rats, neither DA clearance nor amphetamine stimulated locomotion was altered by repeated amphetamine exposure. It appears, therefore, that under these conditions the ability of amphetamine to affect DA clearance and locomotion is related to levels of circulating insulin in brain. That DA clearance was not altered by this amphetamine dosing regimen in rats with normal insulin levels is consistent with reports of unchanged DA uptake after repeated amphetamine administration, even when given at dosing schedules known to cause behavioral sensitization to amphetamine (e.g. Kuczenski and Segal 1988). Importantly, these studies support our contention that an imbalance in insulin signaling affects the ability of amphetamine to produce lasting effects on DA clearance.

The slower rate of DA uptake (in vitro) and clearance (in vivo) in hypoinsulinemic rats is consistent with reports of decreased DA uptake in fasted rats (Patterson et al. 1998). This decrease is not likely attributable to loss of DAT as we (Galici et al. 2003) and others (Patterson et al. 1998) report no change in total DAT binding in brains of hypoinsulinemic compared to control rats. Consistent with the ability of insulin to translocate other proteins (e.g. GLUT4 glucose transporter) into and out of the plasma membrane (Cheatham and Kahn 1995), it is likely that DAT is being redistributed from the plasma membrane to the cytosol. Support for this idea comes from studies in cultured cells stably expressing hDAT. Here blockade of insulin signaling decreases cell surface expression of hDAT (Carvelli et al. 2002; Garcia et al. 2005). Based on the present data we speculate that in vivo, insulin also mediates recruitment of DATs to the plasma membrane. In STZ-treated rats insulin is absent and so the signal to mobilize DATs to the plasma membrane is ‘turned off’, ultimately leading to a reduction in the rate of DA clearance. In general, affinity values determined in vivo (KT) and in vitro (Km) were not altered by STZ-treatment, supporting the notion that the decreased ability to clear DA in hypoinsulinemic rats is a function of reduced DAT expression on the plasma membrane.

As mentioned earlier, the finding that DAT function is reduced in insulin-depleted rats is in agreement with studies in food-deprived rats (e.g. Patterson et al. 1998). However, the present data are seemingly at odds with those of Galici et al. (2003) who reported increased [3H]DA uptake into striatal synaptosomes prepared from STZ-treated rats. However, differences in experimental protocol make comparison of these data sets inappropriate, e.g. rats trained to self-administer amphetamine prior to STZ treatment (Galici et al. 2003) compared to four single injections of amphetamine after STZ-treatment (present study). Rather, the different outcomes on DAT function between these studies underlines the complexity of the potential interaction between insulin status and amphetamine actions. Clearly systematic time course studies investigating the dose and number of amphetamine exposures required to normalize/enhance DA clearance in STZ-treated rats are needed. For example, an alternate interpretation of the present data is that low DA clearance is the important variable in determining amphetamine's effect rather than insulin levels at the time per se. How amphetamine-induced changes in DA clearance are influenced by the presence, removal and replacement of insulin will be pivotal in teasing apart the relationship between insulin and how amphetamine interacts with the DAT.

Of course factors other than insulin and/or insulin signaling pathways may underlie the changes in DAT function reported here. For example, hyperglycemia is also a consequence of STZ-treatment, thus increased glucose might well factor into these findings. However, compelling evidence that glucose does not drive changes in DAT function reported here come from that observation that blood glucose levels did not differ between STZ-treated rats as a function of amphetamine history and yet DA clearance rates were markedly lower in STZ-treated rats with no amphetamine history. Moreover, there is mounting evidence that points to insulin as a regulator of DAT function. For example, Carvelli et al. (2002) reported that insulin increases [3H]DA uptake into HEK cells expressing the human DAT, and in the study of Patterson et al. (1998), the decrease in DA uptake into striatal synaptosomes prepared from fasted rats could be normalized by addition of a physiological concentration of insulin (1 nm) to the suspension. These results provide strong support for the notion that decreased DA uptake is a direct consequence of reduced insulin and not other consequences of hypoinsulinemia such as hyperglycemia.

The possibility that STZ itself may alter DAT function or expression also seems unlikely. For example, STZ does not cross the blood–brain barrier and so a direct action of STZ at the DAT can be ruled out. As mentioned earlier, STZ-treatment does not alter the density of DATs in the striatum (Patterson et al. 1998; Galici et al. 2003) and so it also seems unlikely that STZ-treatment leads to the production of compounds that are toxic to DA terminals or the DAT. However, STZ-induced diabetes does represent a systemic stress above and beyond loss of insulin. Ensuing changes in the HPA axis (e.g. Barber et al. 2003; Chan et al. 2005), production of free radicals (e.g. Chen et al. 2005) and other central neuropathies (e.g. Nitta et al. 2002) may account for the present observations and although do not deny the present interpretation of these data, should not be discounted.

The most striking finding of this study was restoration of DA clearance rates in hypoinsulinemic rats to control rates by a short amphetamine-dosing regime. ‘Normalization’ of DA clearance rates after treatment with amphetamine was apparent 1 day after the last amphetamine injection (the earliest time point tested). Moreover, normalization of DA clearance rates was long lasting, persisting for at least 25 days after amphetamine treatment was stopped (the longest time point tested). Consistent with other reports (Marshall 1978; Merali et al. 1988; Shimomura et al. 1990), hypoinsulinemic rats displayed decreased locomotor activity. Given that DA clearance is also reduced in hypoinsulinemic rats, this finding might at first seem at odds with the widely accepted belief that increased extracellular DA increases activity. However, it is important to note that chemically induced diabetes affects many centrally and peripherally mediated physiological functions. It is therefore very likely that hyperactivity that might result from reduced DA clearance could be masked by other opposing factors or was simply not detected under the paradigm used here. For example, Shimomura and coworkers found that ambulatory activity and DA turnover in STZ-treated rats are dependent on circadian rhythm. During certain periods of the light phase (9.00–12.00 h), ambulation of STZ-treated rats was decreased relative to control rats. However later during the light phase (12.00–15.00 h), ambulation of STZ-treated rats was increased relative to control rats. A factor that should also be considered is that STZ-induced decreases in DA uptake and the resulting initial increase in extracellular DA concentration might impact other components of DA neurotransmission. For example, DA receptor number or sensitivity is likely changed by prolonged stimulation, as might occur under conditions used in this study.

An important new finding from this study is that hyperlocomotion induced by amphetamine administration increased over successive amphetamine injections in STZ-treated rats but remained stable in control rats. That sensitivity to locomotor-stimulating effects of amphetamine increased only in STZ-treated rats suggests that any neuroadaptation occurring under these conditions is not identical to the sensitization that can occur after repeated, infrequent administration of amphetamine or other stimulants. Therefore, one interpretation of these data is that the gradual increase of locomotor response to amphetamine in STZ-treated rats does not reflect ‘sensitization’per se, but rather ‘normalization’ of the locomotor response to amphetamine that appears to correlate with amphetamine-induced ‘normalization’ of DAT activity. These data, together with the finding that DA clearance rates were sensitive to change by exposure to amphetamine only in STZ-treated rats, raise the distinct possibility that amphetamine can produce long lasting changes in DAT activity under certain physiologic conditions (e.g. hypoinsulinemia).

An exciting avenue for future investigations will be to determine how amphetamine acts to reverse the reduction in DA clearance observed after streptozotocin treatment. The actions of amphetamine on DA neurotransmission are very complex and both dose- and brain region-dependent. We speculate that one action of amphetamine is at the same site/pathway through which insulin regulates expression and/or activity of DAT in the plasma membrane. There is a rapidly growing body of literature supporting a role for auto- and hetero-receptors in regulating expression and/or intrinsic activity of DAT in the plasma membrane of neurons (for reviews see Zahniser and Doolen 2001; Torres et al. 2003; Vaughan 2004). The D2 receptor regulates the rate of DA uptake in brain (e.g. Meiergerd et al. 1993; Cass and Gerhardt 1994; Dickinson et al. 1999) although the precise mechanism of regulation is not established. Because both amphetamine (as an indirect agonist) and insulin are known to influence activity and expression of D2 receptors in brain (Lim et al. 1994; Sumiyoshi et al. 1997; Seeman et al. 2002), this receptor seems a likely target for regulation of DA uptake by these compounds. Moreover, an interaction between insulin and D2 receptors in behavioral measures of reward have been reported (Kamei and Ohsawa 1996; Sipols et al. 2000; Figlewicz 2003) adding support to the notion that insulin plays a major role in mediating dopamine reward pathways in brain.

In the clinical setting, these results create exciting new roads for investigation into addiction. For example, it is known that eating disorders often cluster with drug abuse (Wolfe and Maisto 2000; Grigson 2002), and common neural circuits are thought to underlie food and drug rewards (e.g. Schultz 2001). Our data support the proposal of Figlewicz (2003) that long-term changes in the function of those neural systems that occur as a consequence of self-imposed reductions in circulating insulin (e.g. that occur during food deprivation periods preceding eating binges in bulimics) enhance the rewarding stimulus of both food and drugs of abuse.


This work was supported by National Institutes of Health Grants R21 DA018992 (LCD), RO1 DA14684 (AG), KO2 DA00211 (CPF), and a NARSAD Young Investigator Award (LCD).