Intracranial self-stimulation as a positive reinforcer to study impulsivity in a probability discounting paradigm

Abstract

Probability discounting is used to study risky decision-making in humans and rodents. In these paradigms, the subject chooses between a small reward that is always delivered and a large reward that is delivered with varying probabilities. Risk-taking is defined as a preference for the large, uncertain reward. The aversive consequence associated with this task involves choosing the large reward and not obtaining it. To study this form of impulsivity in rodents, food reinforcement is commonly used. Using this reinforcer, and the need to food-deprive rodents to enhance task performance, may be problematic in rodent models that exhibit eating disorders, in pharmacological assessments that alter feeding, and for assessments of the neurocircuitry that is engaged by both feeding and risk-taking. We reveal here that electrical intracranial self-stimulation (ICSS) can be used as the positive reinforcer in risk assessments (i.e., probability discounting). ICSS was selected as it is rapidly acquired, the operant procedures are retained for months, and no tolerance or satiety develops to the reinforcer; thus, ICSS can be used in multiple test sessions in a repeated measures design. We developed an efficient, standardized, six phase ICSS-mediated protocol that allowed for the assessments of risk-taking in a probability discounting task. We demonstrated that the discounting behavior remained stable for several weeks. The value of this protocol is discussed in terms of practical as well as theoretical advantages of using ICSS-mediated reinforcement.

Introduction

Impulsivity can be regarded as “actions that appear poorly conceived, prematurely expressed, unduly risky, or inappropriate to the situation” (Daruna and Barnes, 1993). While some beneficial aspects of impulsivity are known (Dickman, 1990), it is generally recognized as a dysfunctional trait that is frequently associated with numerous neurological and psychiatric disorders including frontal lobe damage, schizophrenia, attention deficit-hyperactive disorder and substance abuse disorders. According to the American Psychiatric Association, impulse control disorders (ICDs) are a form of psychiatric disorder (American Psychiatric Association, 2000). ICDs include trichotillomania, intermittent explosive disorder, pathological gambling, kleptomania, pyromania, hypersexuality, compulsive shopping and others.

To understand impulsivity and ICDs and to subsequently develop therapies targeted to particular aspects of the disorder, laboratory protocols that model attributes of impulsivity are required. Risky decision-making is one facet of impulsivity. A common method used to study risky choice in both humans and laboratory rodents is the probability discounting paradigm (Mobini et al., 2000, Rachlin et al., 1991, Richards et al., 1999). In this task, the subject can choose between a small reward that is always delivered and a large reward that is delivered with varying probabilities. Risky behavior is defined as a preference for the large uncertain reward. The aversive consequence associated with this task involves choosing the large reward and not obtaining it (Cardinal and Howes, 2005). In rodent testing of probability discounting, food is often used as the positive reinforcer and to motivate the animal, salience of the food is enhanced by food-deprivation. This approach presents several disadvantages which can potentially confound outcomes. First, internal factors, such as hunger or thirst, can themselves lead to a change in impulsive behavior in animals (Minamimoto et al., 2009, Schuck-Paim et al., 2004). Second, chronic food restriction can lead to adaptations in dopaminergic (Carlson et al., 1988, Carr et al., 2003, Carr et al., 2009, Collins et al., 2008) and serotonergic signaling (Haleem and Haider, 1996, Huether et al., 1997, Kohsaka et al., 1980). These neurotransmitters also play a role in impulsivity (Adriani et al., 2009, Mehlman et al., 1994, Mobini et al., 2000, Soubrié, 1986, Winstanley et al., 2005). Moreover, this reward option may not be possible for assessments of risky choice in rat models of human neuropathologies that present eating disorders or for testing pharmacologics that alter feeding behaviors. Thus, we sought to design a probability discounting paradigm that utilized a positive reinforcer that avoided such shortcomings. To be broadly applicable to a range of laboratory assessments, we determined that criteria for this reinforcer should include the following: (i) It should more directly engage brain “reward centers” than is possible with food reward. (ii) It should be conducive to robust operant task testing. (iii) It should demonstrate a range of reward values that can be discriminated by the rat. (iv) Finally, it should support stable responding for several weeks. We reveal here that intracranial self-stimulation (ICSS) meets these criteria. In ICSS procedures, rats perform an operant task to obtain a positive reinforcing current delivered via an electrode implanted in reward regions of the brain (Olds and Milner, 1954). For the current study, we selected the medial forebrain bundle (MFB) at the level of the lateral hypothalamus (LH) as the stimulation target. This structure is well known to readily support ICSS with a large range of stimulation parameters. We detail how this reward parameter can be successfully implemented for probability discounting paradigms, and we verify performance stability and persistence.