The current studies examined the effects of the VMAT-2 inhibitor and catecholamine depleting agent TBZ on performance of the PROG/chow feeding choice task in rats. TBZ significantly decreased total lever presses, highest ratio achieved and active lever time. In addition, TBZ did not produce an overall significant effect on chow intake, indicating that appetite for chow consumption was not suppressed at the doses tested. These effects of TBZ are consistent with those recently demonstrated in rats assessed on a concurrent fixed ratio 5 (FR5)/chow feeding choice task (i.e., decreased lever pressing and increased chow intake after TBZ administration; [43]). In addition, Nunes et al. [43] reported that doses of TBZ up to 1.0 mg/kg had no effect on consumption of operant pellets or lab chow, and did not alter preference of pellets over chow, in free feeding preference tests. Thus, it seems unlikely that the present results with TBZ occurred because of a change in appetite or food preference. Importantly, these effects were more pronounced in high performers. High performers showed decreases in total lever presses at 0.5 mg/kg TBZ whereas low performers required 0.75 mg/kg in order to show decreases in total lever presses. Moreover, high performers showed decreases in highest ratio achieved and active lever time at all doses of TBZ tested, whereas low performers showed no changes in highest ratio achieved or active lever time under any dose of TBZ. Finally, high performers showed increases in chow consumption at 0.5, 0.75 and 1.0 mg/kg TBZ, where low performers showed no change in chow consumption, presumably because they were already consuming chow at or near ceiling levels.

Recent research has shown that 0.75 mg/kg TBZ significantly decreased extracellular DA in nucleus accumbens as measured by microdialysis [43]. In addition, TBZ affected expression of phosphorylated DARPP-32 in a manner consistent with reduction of DA transmission at both D1 and D2 family receptors (43). Thus, the present experiments also studied the effects of highly selective D1 and D2 family antagonists. Selective antagonism of D1 or D2 family receptors (via administration of ecopipam or eticlopride, respectively) produced similar effects to TBZ on the PROG/chow feeding task, decreasing total lever presses, highest ratio achieved and active lever time. Neither ecopipam nor eticlopride reduced chow intake in the dose range tested, and in fact, ecopipam significantly increased chow intake, which illustrates that there was a strong shift in effort-related choice from lever pressing to chow intake. Thus, it appears that reductions in DA transmission at both D1 and D2 receptors contribute to the effects induced by TBZ. Taking the present results together with previous studies, it appears that blockade of either D1 or D2 receptors, or pharmacological depletion of DA, can reduce PROG lever pressing and lower break points in a manner that does resemble the effects of appetite suppression or reinforcer devaluation [32], [43]. Instead of providing a simple measure of “reward”, PROG breakpoints represent the outcome of a cost/benefit analysis based partially upon characteristics of the reinforcer itself, but importantly, also on the work-related response costs and time constraints imposed by the increasing ratio requirements [32]. Together with earlier studies, the results of experiments 1–3 demonstrate that it is exceedingly unlikely that TBZ, ecopipam, eticlopride and haloperidol are decreasing PROG lever pressing because of a reduction in primary food motivation, appetite, or the unconditioned reinforcing properties of food [26], [32], [43]. Furthermore, these results are consistent with studies showing that increasing DA transmission by knockdown of the DA transporter [11] or overexpression of DA D2 receptors in nucleus accumbens [59] can increase instrumental response output in mice responding on fixed ratio/chow feeding choice procedures or a PROG schedule, in a manner that is inconsistent with an alteration of primary food motivation or the representation of the value of the food reinforcer.

The effects observed with TBZ or DA receptor blockade differed substantially from those induced by appetite manipulation, as demonstrated by the effects of the cannabinoid CB1 neutral antagonist and putative appetite suppressant AM4113 [35]. The effects of AM4113 described above were similar to those observed for the CB1 inverse agonist AM251 [32]; CB1 antagonism decreased chow intake as well as total lever presses and highest ratio achieved, consistent with previous studies demonstrating the appetite suppressant effects of these drugs [32], [35]. In the present studies, AM4113 produced greater effects on lever pressing in high performers compared to low performers. Specifically, high performers showed decreases in total lever presses and highest ratio achieved, whereas low performers showed no change in total lever presses. This pattern of effects in low performers is likely due to the fact that the low performing animals get the vast majority of their food from chow intake. The overall similarity between the actions of a CB1 neutral antagonist (AM4113) and a CB1 inverse agonist (AM251) suggests that there is endogenous cannabinoid tone regulating food motivated behavior, which is attenuated by occupation of CB1 receptors by the neutral antagonist [35]. Moreover, the present results highly the utility of choice procedures as compared to conventional PROG schedules, because having chow concurrently available enables one to differentiate between the effects of low doses of DA antagonists or TBZ, which leave chow consumption intact, vs. the appetite suppression induced by interference with CB1 transmission, which reduces both food reinforced lever pressing and chow intake.

In recent years, there has been increasing interest in the use of adenosine A_2A antagonists for their potential antidepressant effects [43], [59]. This is supported by studies using traditional animal models of depression including the forced swim and the tail suspension tests, in which adenosine A_2A antagonists have been shown to increase swim and struggle time [60], [61], [62]. Moreover, several studies have demonstrated the ability of A_2A antagonists to reverse the effects of selective DA antagonism on a variety of effort-related tasks including the FR5/chow feeding choice task and the T-maze barrier task [27], [25], [28], [39]. In the present studies, the adenosine A_2A antagonist MSX-3 was capable of reversing the effects of TBZ, increasing total lever presses, highest ratio achieved and active lever time. These effects are consistent with TBZ/MSX-3 reversal studies conducted in rats tested on the FR5/chow feeding choice procedure [43]. Moreover, these results are consistent with studies demonstrating the minor-stimulant properties of adenosine A_2A antagonists such as MSX-3, which increases operant response rates on some schedules of reinforcement [32], [57]. In the present studies, the reversal effects produced by MSX-3 were more pronounced in high performers; in these animals MSX-3 plus TBZ increased total lever presses, highest ratio achieved and active lever time relative to TBZ alone. Low performers on the other hand showed no effect of drug treatments. It is likely that this pattern of effects is due to the fact that the TBZ effects on PROG performance are greater in high performers, which could render those animals more susceptible to reversal.

In addition to testing novel compounds that are putative antidepressants, it is critical for the development and validation of the present model that the effects of well characterized antidepressants on the PROG/chow feeding choice paradigm should be assessed. Because TBZ depletes DA, a logical point of focus was on an antidepressant that works on DA, such as the catecholamine uptake inhibitor bupropion (Wellbutrin). Bupropion is one of the most widely prescribed antidepressants [63], and it is one of the few antidepressants that has shown efficacy in treating the effort-related symptoms such as fatigue, anergia and psychomotor retardation in patients with depression and related disorders [44], [46], [64]–[66]. In the current experiments, bupropion was capable of reversing the effects of TBZ on the PROG/chow procedure, increasing total lever presses, highest ratio achieved and active lever time. Moreover, at the highest dose (15.0 mg/kg) it decreased chow consumption compared to TBZ treated animals. For both high performers and low performers, bupropion increased total lever presses, highest ratio achieved and active lever time compared to TBZ/vehicle, and there also was a decrease in chow consumption in high performers. These results are consistent with previous findings on the FR5/chow choice procedure [43] and the T-maze barrier task (Yohn et al., in preparation), which also demonstrated the ability of bupropion to reverse the effects of TBZ administration.

In addition to compounds that block DA uptake such as bupropion, there has been interest in compounds that block the enzymatic breakdown of DA for their antidepressant effects. MAO inhibitors are one group of compounds that have been assessed for their antidepressant effects. Originally developed as an antiparkinsonian drug [67], the MAO-B inhibitor deprenyl has been shown to have antidepressant effects in humans [68], [69], as well as rodents tested on traditional animal models of depression, including the forced swim test and inescapable shock paradigm [70], [71]. In view of the fact that nonselective MAO inhibitors have been used as antidepressants, and deprenyl and related drugs are recommended for treating akinesia and psychomotor slowing seen in Parkinson's disease, it is reasonable to hypothesize that an MAO-B inhibitor could be useful for the treatment of psychomotor/motivational symptoms. The current studies assessed the ability of deprenyl to reverse the effects of TBZ on the PROG/chow feeding choice procedure. Deprenyl was capable of reversing these effects, increasing total lever presses, highest ratio achieved and active lever time. In addition, deprenyl decreased chow intake at the highest dose (10.0 mg/kg) compared to TBZ alone. In analyzing these results by performance level, it was revealed that these effects were greater in high performers, who demonstrated significant deprenyl-induced reversal of the effects of TBZ on total lever presses, highest ratio achieved and active lever time; low performers showed little to no effect of deprenyl. As discussed above, this could be due in part to the fact that TBZ produces greater effects in high performers, and therefore effects in these animals are easier to reverse.

COMT inhibitors are another group of drugs that act to block the enzymatic breakdown of DA, and thus could have effects on effort-related decision making. There is some clinical evidence of increased COMT levels in depressed patients [72]. Moreover, there is evidence for genetic variations in the COMT gene that could lead to abnormal COMT function [73]. The COMT inhibitor tolcapone has been used clinically in Parkinson's patients to effectively treat their effort-related depressive symptoms [45]. The current studies sought to assess tolcapone for its ability to reverse the effort-related impairments induced by TBZ on the PROG/chow task. As described above, tolcapone failed to reverse the effects of TBZ on any measure of PROG/chow performance in the dose range tested. One explanation for this lack of effect comes from clinical studies demonstrating that improvements in energy-related symptoms following tolcapone treatment in Parkinson's patients were in the presence of L-DOPA/carbidopa treatment [74]. This suggests that tolcapone may be more effective at reversing the effects of TBZ if it is co-administered with other drugs that stimulate DA transmission, such as L-DOPA or bupropion. In addition, it has been suggested that COMT plays a more important role in prefrontal cortex, where the DA transporter is sparse, compared to striatum, where the DA transporter is abundant; such a pattern has been interpreted to mean that COMT exerts a greater effect on prefrontal cortex DA metabolism, while MAO exerts a greater effects on striatal DA metabolism [75], [76]. Furthermore, there is evidence from COMT knockout mice showing that COMT has little effect, if any, on striatal DA levels [77]. When you consider these findings in the context of clinical studies demonstrating that entacapone, a peripherally restricted COMT inhibitor, also has been shown to be effective at treating depressive symptoms when co-administered with L-DOPA [78], [79], it seems possible that the antidepressant effects of COMT inhibition are due to blockade of peripheral breakdown of L-DOPA, increasing the amount that reaches the CNS. In fact, L-DOPA has been shown to improve effort-related symptoms in depressed patients [80].

In conclusion, TBZ alters effort-related choice behavior, reducing various indices of PROG lever pressing at doses that did not suppress chow intake. In fact, TBZ increased chow intake in rats with high levels of lever pressing performance. The ability of TBZ to affect effort-based decision making is consistent with research showing that other manipulations associated with depression, including stress [81], and administration of pro-inflammatory cytokines [42], which can induce fatigue-related symptoms in animals and humans [42], [82], [83], also can alter effort-based choice. The effects of TBZ on PROG output were reversed by co-administration of the adenosine A_2A antagonist MSX-3, as well as two drugs that have been used as antidepressants in humans (i.e., bupropion and deprenyl). Additional studies should investigate the effort-related effects of antidepressant drugs with different pharmacological profiles. However, it also should be emphasized that tests of effort-related decision making in rodents are not designed to serve as animal models of depression, in the broadest sense of the term. Rather, they are being investigated as potential models of a group of symptoms (i.e., effort-related psychomotor/motivational symptoms) that is characteristic of depression [44]–[47], [84], [85], but also is seen across multiple disorders, including conditions associated with high levels of pro-inflammatory cytokines [42], [82], [83], schizophrenia [86]–[90], and Parkinson's disease [91]. This suggestion is consistent with recent approaches in mental health research and theory that place less emphasis on traditional diagnostic categories or disorders, and instead focus on the neural circuits mediating specific pathological symptoms (i.e., the Research Domain Criteria or RDoC approach; [42], [92]).