Association for Behavior Analysis International

Traumatic brain injury is a major problem that affects approximately 1.7 million people annually in the United States alone. Brain injuries result in long-term disabilities for a number of people. Animal models of brain injury have been used for the last 25 years to test numerous pharmaceutical agents. However, despite the many years of research, to date, all clinical trials have failed. One potential cause of this is a lack of behavioral characterization in experimental models. In particular, models of frontal brain injury have been underutilized despite a robust literature of operant behavioral tasks that directly assess frontal function. Additionally, these frontally-mediated behaviors have direct relevance to everyday function in human patients. This symposium will examine some of the operant measures that are currently being used to assess animals with frontal traumatic brain injury and highlight the potential for using these behavioral models as a rigorous screen for pharmaceutical agents.

The present study assesses the effects of TBI using a single subject operant conditional discrimination technique as a novel assay. Rats were trained on a conditional discrimination task that involved 200 discrete 10-s trials, separated by a 25 s inter-trial interval (ITI), where either a high pitch or low pitch tone was presented. Milk delivery would reinforce correct responses when a single left lever press occurred in the presence of a low pitch tone or a single right lever press occurred in the presence of a high pitch tone. Incorrect responses would result in transition to the ITI without milk delivery. Following 85% or higher accuracy, subjects received either a frontal TBI or a sham surgery. Following recovery, subjects resumed training on the discrimination until reacquisition occurred and then we reversed the lever-tone contingencies. Relative to the sham subjects, the TBI subjects showed an initial reduction in completed trials, a loss in the discrimination accuracy, and an increase in side bias. Reversals produced reduction in accuracy for both groups, but the sham group showed slightly better accuracy. This experiment demonstrated that single subject operant procedures might provide a sensitive tool to assess the effects of TBI.

We have previously demonstrated large deficits in discrimination performance following traumatic brain injury (TBI) across previously learned discriminations, novel discriminations and reversals. In a frontal model of TBI, the medial prefrontal cortex (mPFC) is severely damaged and has been implicated in many behaviors, including set-shifting and contingency detection. In the current study, rats were given mPFC lesions and then tested on a scent discrimination. Lesioned rats readily acquired the discrimination. However, after acquisition, differing reinforcement schedules were presented across the two choices. Choice 1 (previously reinforced at FR-1) was gradually tapered up on an FR schedule, while Choice 2 (previously unreinforced) was available on an FR-1 schedule. The lesioned rats took longer than controls to switch to Choice 2. This suggests that the damage to the mPFC may contribute strongly to some of the deficits in sensitivity to contingencies after TBI. However, given that the TBI-induced deficits are much larger and more pervasive, contributions from other cortical and subcortical regions still need to be investigated.

Disinhibition is a symptom resulting from TBI and the present study assesses response inhibition using a successive discrimination procedure. Experimental conditions exposed rats to a multiple schedule using LED flash rates as component stimuli correlated with a reinforcement contingency (SD) and extinction (S?). First, multiple schedule acquisition phase was conducted until 85% of responses occurred only in the SD component. Second, a series of generalization test probes were conducted to assess graded stimulus control. Third, a frontal TBI or sham surgery was administered followed by seven days of recovery. Finally, post-injury multiple schedule reacquisition sessions, interspaced with additional generalization probes, were run. Generalization probes involved components with flash rates that vary in resemblance to the SD component flash rates to establish generalization gradients. Injured subjects showed an abrupt reduction in response rates and poor discrimination performances in initial reacquisition. Generalization probes prior to injury showed generalization decrement around the SD. Injured subjects generalization gradients were less steep, however, still showed decrement around the SD. These results suggest that following a TBI incident there are slight but enduring errors in successive discrimination performances in comparison to rats experiencing a sham surgery.