Organization and expression of reward in the rodent orbitofrontal cortex. Focus on "orbitofrontal ensemble activity monitors licking and distinguishes among natural rewards".

In this issue of the Journal of Neurophysiology, Gutierrez et al. (p. 119–133) show that neurons in the orbitofrontal cortex play a critical role in organizing the consummatory response of licking (Gutierrez et al. 2005). Using both muscimol-induced reversible lesions and single unit ensemble recording, they show that orbitofrontal cortex neurons code the palatability or rewarding properties of a stimulus, i.e., differentiate water from sucrose as well as predict the onset and offset of voluntary licking. One of the unique features of this study is that neurons in the orbitofrontal cortex predicted the onset of licking based on an “internal” cue and not one provided by the investigator. The rat was free to lick, pause, and resume licking based on its own contingencies. Because neuron ensembles could accurately predict when licking was about to commence, they effectively translated the abstraction of “reward” into overt motor behavior. Licking occurs in clusters that last on average 2–3 s. Within a cluster, licking occurs in a stereotyped fashion and at a fairly constant frequency. Licking clusters are separated by intervals of 0.5–3 s (Davis 1996), but clusters become longer with fewer, shorter pauses as the molarity of a sucrose solution is increased (Davis and Perez 1993). As an animal licks to satiety, clusters of licking get shorter and pauses get longer. Thus the consummatory behavior of licking reflects a balance between positive orosensory cues such as a sweet taste and negative viscerosensory cues that signal satiety (Davis and Perez 1993). When negative feedback cues are eliminated, an animal will lick nearly continuously to a preferred or needed stimulus (Frankmann et al. 1996). This “microstructure” of licking, i.e., clusters of licking of variable duration, separated by pauses of variable length, is thus highly sensitive to oral and visceral inputs driving reward and satiety (Davis 1996; Spector et al. 1998). However, where is the microstructure of licking organized? Although the hindbrain contains the sufficient neural apparatus to lick (Miller and Sherrington 1916), lick longer to higher concentrations of sucrose (Grill and Norgren 1978), and even to integrate viscerosensory feedback (Seeley et al. 1994), the hindbrain is insufficient to initiate voluntary licking. The cluster organization of licking reflects the voluntary nature of this appetitive response and the forebrain is required for the initiation and termination of cluster bouts. Gutierrez et al. show that reversible lesions with muscimol infusions into the orbitofrontal cortex did not abolish licking or appreciably slow the lick rate to a preferred (0.3 M) solution of sucrose. Rather, the microstructure of licking was modulated such that licking became less continuous, with clusters of licking more frequently interrupted by pauses compared with controls. More pauses are a sign of reduced palatability. Thus, one interpretation is that inactivating orbitofrontal cortex reduced the reward value of sucrose. Alternatively, the 0.3 M solution may be just as palatable; after all, the animal continued to return to the spout to lick, but the translation of palatability to the motor pattern for licking was altered. The motor pattern generator for licking, however, is not organized in the forebrain, begging the question of what role orbitofrontal cortex neurons play in organizing licking behavior. Neurophysiological recording provided some provocative insights. Ensembles of up to 14 neurons were simultaneously recorded from the orbitofrontal cortex during voluntary licking of sucrose and water. Some neurons responded before licking, effectively predicting when a licking cluster was about to commence. Others fired during the licking cluster, with a small subset entrained to the lick rate, suggesting continuous gustatory and somatosensory feedback. Other neurons fired after the end of a licking cluster, perhaps contributing to a code for cluster termination. Significantly, ensembles of simultaneously recorded neurons performed better at predicting lick cluster initiation than did single neurons. It is still unclear through what pathways ensembles of orbitofrontal cortex neurons influence and interact with other ingestionand reward-related substrates. The orbitofrontal cortex is intimately and reciprocally connected with the nucleus accumbens, as well as other ventral forebrain structures involved with coding the rewarding properties of stimuli (Berthoud 2002; Kelley 2004). In addition, there are projections to the lateral hypothalamus where internal signals of energy balance are also integrated with orosensory signals (Van Eden and Buijs 2000). Reciprocal interconnections between these forebrain structures constitute parallel “loops” that seem to preclude any reduction of the circuit into a simple input/output organization (Alexander et al. 1986; Berthoud 2002). However, electrical stimulation of the orbitofrontal cortex does induce rhythmic oral movements (Neafsey et al. 1986), and anatomical studies (Van Eden and Buijs 2000) suggest that orbitofrontal cortex may have direct projections to the lower brainstem reticular formation, conceivably to sites containing preoromotor neurons that are part of a central pattern generator for licking (Chen et al. 2001). Establishing the functional relationship of orbitofrontal cortex with other sites critical to the consummatory response of licking remains unexplored. Although a number of studies have established that neurons in several regions of the brain are entrained to the lick cycle, only this study determined a relationship to the microstructure of licking. Thus we do not yet know if neurons Address for correspondence: Section of Oral Biology, Ohio State University, Columbus OH 43218-2357 (E-mail: [email protected]). J Neurophysiol 95: 14–15, 2006. doi:10.1152/jn.00968.2005.