There are several additional neural inputs to the core reward system that may modulate drug reward by modulating DA function within the core system. GABAergic efferents from the nucleus accumbens form a feedback loop to the ventral tegmental area, and nucleus accumbens medium spiny GABAergic neurons also project to other GABAergic neurons synaptically linked to both the accumbens and ventral tegmental area (Alexander and Crutcher, 1990; Kalivas et al., 1993; Van Bockstaele and Pickel, 1995). Endogenous opioid peptidergic neurons also provide synaptic regulation of core mesoaccumbens DA function and of the accumbens-ventral pallidal projection (Alexander and Crutcher, 1990; Heimer and Alheid, 1991; Zahm and Brog, 1992; Kalivas et al., 1993; McGinty, 1999). Both the ventral tegmental area and the nucleus accumbens also receive serotonergic inputs, and manipulation of these serotonergic inputs appears to modulate reward functions. In fact, serotonergic lesions appear to make cocaine more rewarding (Loh and Roberts, 1990). Cocaine is also rendered more rewarding by fluoxetine-induced acute enhancement of forebrain serotonergic levels, which - in turn -may inhibit serotonergic cell firing by stimulation of serotonergic autoreceptors (Chen et al., 1996). Congruent with this is the observation that microinfusion of the serotonin agonist 8-OH-DPAT into the dorsal raphe nucleus - which produces autoreceptor-mediated inhibition of serotonergic cell firing - potentiates the rewarding effects of electrical brain-stimulation reward (Fletcher et al., 1995). It is also now recognized that the core mesoaccumbens reward system receives a substantial modulatory cholinergic input (Oakman et al., 1995), which would appear to have relevance for nicotine-induced reward.
It would be a serious misstatement to convey the impression that these CNS reward substrates are either anatomically or functionally simple. Indeed, a very large body of experimental evidence (for reviews, see Gardner, 1997, 2000)
suggests that these reward-related systems are functionally heterogeneous - with some neurons encoding reward magnitude per se while others encode expectancy of reward, errors in reward-prediction, prioritized reward, and other more complex aspects of reward-driven learning and reward-related incentive motivation. However, it seems equally clear that one of the primary functions of those reward substrates is to compute hedonic tone and neural "payoffs," that this computation takes place in large measure within the circuits delineated above, that the "second-stage" DA component is the common site of action for addictive drugs (and is crucial to their addictive features), that drug reward per se and drug potentiation of electrical brain-stimulation reward have common mechanisms, and that electrical brain-stimulation reward and the pharmacological rewards of addictive drugs are habit-forming because they act in the CNS circuits that subserve more natural, biologically significant rewards (Wise, 1996b; Shizgal, 1997).
Very importantly, the "second-stage" DA component of these CNS reward substrates appears to be the crucial convergence upon which drugs with euphorigenic properties and/or addictive potential (regardless of chemical structure or pharmacological category) act to enhance neural reward functions, subjective experience of reward, and reward-related behaviors.
Was this article helpful?