Nicotine in an animal model of attention

Abstract

Studies in smokers have suggested that at least part of the improved psychomotor performance produced by nicotine is the result of an effect on attention. Many animal experiments have assessed the effects of nicotine and its antagonists on diverse types of learning and memory but relatively few have looked at it in tasks designed to assess attention. In a five-choice serial reaction time task (5-CSRTT), rats with restricted access to food were presented with an array of five holes; illumination of a randomly selected hole signalled that a nose-poke into it would be reinforced by food presentation. Initially, signal length and the inter-trial interval (ITI) were varied and the procedure was demonstrated to satisfy some criteria for a vigilance task. The effects of nicotine on deficits in performance induced by varying signal length and ITI were assessed. Under appropriate conditions, small doses of nicotine increased the percentage of correct responses (accuracy), decreased omission errors and reaction time, and increased anticipatory responses. Subsequently, the effects of varying the ITI were examined more extensively in a slightly modified task. Here, nicotine produced small but robust, highly significant dose-related increases in accuracy, as well as decreases in omission errors and reaction times. Nicotine also increased accuracy when light stimuli were presented in an unpredictable manner. The nicotine antagonist mecamylamine produced a modest deficit in reaction time only. It is concluded that appropriate doses of nicotine can produce robust improvements in performance of normal rats in an attentional task. The effect cannot be attributed easily to changes in sensory or motor capability, learning or memory and may provide the measures needed to investigate the neuropharmacological and neuroanatomical bases of the elusive attentional effect of nicotine.

Introduction

The expressed reasons for human smoking have included reports that it aids concentration, leading to suggestions that such possible benefits may constitute a major motive for the use of tobacco (Warburton, 1990). For several decades, the validity of such reports has been supported by findings from laboratory studies in both animal and human subjects suggesting that nicotine can improve aspects of psychomotor performance and cognition. Some studies reported that nicotine could robustly increase rates of operant responding for food or other reinforcers Morrison, 1967, Risner et al., 1985, Goldberg et al., 1989. Other studies reported variously that underlying processes such as attention, learning and memory were all improved by nicotine, but findings of one report were rarely reproduced by another which was perhaps not surprising due to the enormous variations in methodology Garg and Holland, 1968, Battig, 1970, Nelsen and Goldstein, 1972. These early studies are summarised elsewhere Hunter et al., 1977, Stolerman, 1990. This paper describes how continuing persistent and increasingly systematic studies have, over the years, supported the essence of these inconsistent and variable initial results and have progressed towards establishment of methods that yield relatively well-defined and reproducible effects.

A large proportion of the studies in animal subjects have used tasks designed to assess different forms of learning and memory. However, from human research, notably with smokers, it is apparent that nicotine's cognitive benefits are predominantly and most consistently seen in tasks which tax information processing ability and sustained attention (Warburton, 1990; Parrott and Craig, 1992). Therefore, the focus on learning and memory in many studies with animals is a little difficult to understand. Only under rather specific conditions has it been found that nicotine improves memory in humans (Stolerman et al., 1995).

Studies of nicotine in smokers are to varying extents confounded by baseline shifts associated with the nicotine withdrawal syndrome. Tobacco withdrawal is thought to be associated with an expressed difficulty in concentrating upon a task and by actual impairments in task performance. If the baseline condition for assessing the effect of nicotine is subnormal due to withdrawal, then the effect of nicotine will appear larger than it really is. This has led to the suggestion that in the worst case the cognitive enhancing effects of nicotine are nothing more than the result of relief from withdrawal-induced deficits (Hatsukami et al., 1991; Parrott and Roberts, 1991). However, some studies have attempted to identify a ‘normal’ baseline where smokers are not in withdrawal, and have still demonstrated some enhancement of performance. Thus, studies that have attempted to control for the confounding effects of withdrawal have still demonstrated beneficial effects of nicotine (West and Hack, 1991). A small number of studies have also found that nicotine can improve attention in non-smokers using rapid visual information processing tasks Wesnes and Warburton, 1984, Foulds et al., 1996. These findings are paralleled by the many observations of improvements in laboratory animal performance upon exposure to the acute effects of nicotine Stolerman, 1990, Stolerman et al., 1995. Therefore, it seems that performance enhancement need not be attributed exclusively to either the direct effects of nicotine or to the effects of nicotine withdrawal; the two phenomena appear to co-exist, such that improvements in performance are most likely greater in nicotine-dependent individuals than in non-smokers.

In view of the deficits in attention in aging and age-related diseases, it is surprising that few animal paradigms for assessing attentional function have been developed (Lawrence and Sahakian, 1995). Moreover, attentional deficits may occur in the early stages of Alzheimer's disease and neurotransmitter systems that are affected in this disease, such as the cholinergic and noradrenergic systems, are known to be involved in attentional processes Light, 1991, Robbins and Everitt, 1995. Thus, given the effects of nicotine in humans described above and the evidence that nicotine can improve attention in patients with Alzheimer's disease and attention-deficit hyperactivity disorder Jones et al., 1992, Connors et al., 1996, there is a need for valid animal tasks assessing attention. Furthermore, when using animals the confounding effect of withdrawal-induced deficits is not a problem. Although an early study by Nelsen and Goldstein (1972) demonstrated that chronic treatment with 0.4 mg/kg nicotine for 3 weeks improved the performance of rats on a simple reaction time task, this has not been replicated and the validity of this task has been questioned (Turchi et al., 1995). More recently, Muir et al. (1995) showed that forebrain cholinergic lesion-induced deficits in a five-choice serial reaction time task (5-CSRTT) were reversed by nicotine (0.06–0.1 mg/kg) in rats. By contrast, Turchi et al. (1995) demonstrated no effect of nicotine (0.09–0.8 mg/kg) in non-lesioned animals in a vigilance task, where rats were trained to discriminate between signal and non-signal events. However, they did show that the nicotinic antagonist mecamylamine (1.0 mg/kg) impaired performance implying that nicotinic receptors were important in attention. Moreover, mecamylamine (3 mg/kg) can impair the performance of middle-aged rats on the 5-CSRTT (Jones and Higgins, 1995).

The 5-CSRTT used by Muir et al. (1995) was based on paradigms used to study the effects of arousal and environmental factors on human performance (Hockey, 1984). In the adaptation of this task for animals, a rat is required to respond to brief illumination of stimulus lights that occur unpredictably in one of five spatial locations; responding in the correct location is reinforced by presentation of food (Carli et al., 1983). To perform the task a subject must maintain sustained attention over the entire test session. However, due to the spatial nature of this task it is likely to have aspects of divided attention and visual search Carli et al., 1983, Gutnikov et al., 1994. Parasuraman, 1979, Parasuraman, 1985 has established criteria for the construct validity of vigilance tasks, which include sensitivity to (i) the frequency with which the events to be detected are presented, (ii) the strength (intensity) of the events, and (iii) the overall load on the cognitive ability of the organism. In addition, to capture the essence of the concept of vigilance, there should be a decrease in the detectability of events over time (i.e., the vigilance decrement).

In the present experiments, the 5-CSRTT was used to investigate the effects of nicotine in normal adult rats. Initially, various task parameters were manipulated to ascertain whether the task presented to the rats complied with some of the established criteria for a vigilance task in humans. The effects of nicotine were tested under conditions that differed from those present during the intervening daily training sessions; these changes in conditions, involving manipulations of stimulus duration and frequency (inter-trial interval, ITI), were intended to impair performance and, thus, render it more likely that a facilitatory effect of nicotine could be detected. It is important to note that this task appears to have predictive validity since manipulations of the cholinergic and GABA-ergic systems produce changes in performance in this task similar to those seen in humans performing attentional tasks Robbins et al., 1989, Jäkälä et al., 1992, Muir et al., 1992a, Muir et al., 1992b. If so, then the prediction should be that as with non-smoking humans, and smokers not suffering from withdrawal, nicotine should improve performance in drug-naive rats. Conversely, if some tonic level of nicotinic–cholinergic neurotransmission is important for normal cognitive function, then administration of nicotinic antagonists should impair performance. Mirza and Stolerman, 1998, Mirza and Stolerman, 2000 have presented a more detailed account of some of these experiments.