Progress in Neuro-Psychopharmacology and Biological Psychiatry
Review articleThe amphetamine-induced sensitized state as a model of schizophrenia
Introduction
Amphetamine-induced sensitization is the process whereby repeated exposure to amphetamine results in increased behavioural and neurochemical responses to the drug. For example, the ability of amphetamine to stimulate locomotor activity, stereotypy, and striatal dopamine release is markedly enhanced in animals previously exposed to a sensitizing regimen of amphetamine (Paulson and Robinson, 1995, Paulson and Robinson, 1996, Robinson and Becker, 1986) Amphetamine-induced sensitization has long been viewed as a potential model of psychosis (Robinson and Becker, 1986, Snyder, 1973), based in part on the awareness that sensitization can induce changes in dopamine function similar to those hypothesized to be associated with schizophrenia. Additionally, it is known that long-term use of amphetamine, and other stimulants, can induce psychosis in humans (Angrist and Gershon, 1970, Bell, 1965, Connell, 1958, Curran et al., 2004, Guttman and Sargant, 1937, Janowsky and Risch, 1979, Lieberman et al., 1987, Young and Scoville, 1938). While it was originally thought that the similarities between the symptoms of long-term amphetamine abuse and schizophrenia were limited to positive symptoms, a recent large-scale naturalistic study of psychostimulant-induced psychosis showed that patients display both positive and negative symptoms (Srisurapanont et al., 2003). Further, recent work has shown that repeated exposure to amphetamine induces a variety of behavioural, molecular, cellular and neurochemical adaptations in animals that extend beyond the striatal dopaminergic system (Castner et al., 2005, Fletcher et al., 2005, Kolb et al., 2003, Selemon et al., 2007, Wolf, 2003). These findings raise the possibility that amphetamine sensitization can be used to model more than just psychosis.
Historically, research into schizophrenia has been dominated by a focus on understanding and treating psychosis and positive symptoms. Although cognitive deficits have long been known to be present in schizophrenia, it is only relatively recently that serious attention has been paid to this aspect of the disease. This change in focus has occurred for several reasons. First, it is increasingly being recognized that the presence of cognitive dysfunction is negatively associated with measures of functional outcome, while psychosis is only weakly, if at all, related, to such measures (Green, 1996). Secondly, while current antipsychotic treatments are able to successfully manage positive symptoms, such treatments show a much more limited effect in ameliorating cognitive dysfunction (Meltzer and McGurk, 1999, Velligan and Miller, 1999). Thus, even after psychosis is managed successfully by antipsychotic drugs, cognitive deficits typically remain. As such, the need to develop more effective treatments for cognitive dysfunction has emerged as a major concern for schizophrenia researchers.
It is likely that animal models will play an important role in this process. Animal models can be used to develop and test hypotheses about the mechanisms responsible for producing different aspects of the disease state (Sarter and Bruno, 2002). For example, the notion that mesocortical dopamine hypofunction could be important in the etiology of schizophrenia (Davis et al., 1991) was derived in large part from studies that demonstrated excessive mesolimbic dopamine activity following cortical dopamine depletion (Carter and Pycock, 1980, Pycock et al., 1980). Animal models can also be used to screen novel drugs and to help understand the mechanisms of action of those drugs. For example, the conditioned avoidance response test has played an important role in the development of antipsychotic drugs (Wadenberg and Hicks, 1999). It is likely that an animal model which reproduces both the cognitive deficits and psychotic symptoms of schizophrenia will be a crucial step towards the development of more effective pharmaceutical treatments. The purpose of this article is to review the present state of research into the sensitization model, emphasizing particularly behavioural and cognitive changes, and to evaluate how the amphetamine-induced sensitized state models various different aspects of schizophrenia.
The neurotransmitter most closely linked to schizophrenia is dopamine. The original dopamine hypothesis of schizophrenia proposed that schizophrenia is the result of overactive striatal dopamine systems (van Rossum, 1967). Further evidence for a hyperdopaminergic basis for schizophrenia was provided by the finding that antipsychotic drugs function by blocking dopamine D2 receptors (Seeman et al., 1976), and that chronic use of psychomotor stimulants can induce psychotic symptoms (Curran et al., 2004). Strong evidence for enhanced dopaminergic activity in schizophrenia has emerged from imaging studies showing that the binding of radiolabelled dopamine D2 receptor ligands to D2 receptors is displaced by amphetamine-induced dopamine release, and that this effect is enhanced in schizophrenia (Abi-Dargham et al., 1998, Laruelle et al., 1999, Laruelle et al., 1996). In one of these studies a low dose of amphetamine worsened psychosis in patients with schizophrenia, and the severity of this response was highly correlated with the estimate of dopamine release (Laruelle et al., 1999). Complementary studies involving dopamine depletion have revealed an increased baseline occupancy of striatal D2 receptors in schizophrenia (Abi-Dargham et al., 2000).
Prefrontal cortex function is also altered in schizophrenia. For example, reduced regional cerebral blood flow in the dorsolateral prefrontal cortex has been noted when schizophrenics perform poorly on tasks such as the Wisconsin Card Sort Task (Daniel et al., 1991). This hypofrontality correlated with low levels of the dopamine metabolite homovanillic acid (Weinberger et al., 1988), and was alleviated by low doses of amphetamine, which also improved task performance (Daniel et al., 1991). The awareness that schizophrenia is associated with changes in prefrontal dopamine function led to a revised dopamine hypothesis that saw schizophrenia as being due both to excess striatal dopamine as well as reduced prefrontal dopamine, with this latter feature giving rise to both cognitive deficits and negative symptoms (Davis et al., 1991). A prefrontal hypodopaminergic state is not incompatible with a subcortical hyperdopaminergic state since animal studies show that a hypoactive cortical dopamine system can co-exist with, and in fact may be a contributor to, a hyperactive subcortical dopamine system. Thus, selective dopaminergic lesions of prefrontal cortex lead to increased levels of dopamine and its metabolites in the nucleus accumbens and enhance behaviours that are dependent on increased mesolimbic dopamine function (Carter and Pycock, 1980, Leccese and Lyness, 1987).
Traditionally, evidence for the existence of amphetamine sensitization in humans has been derived indirectly from observing behavioural and physiological changes in chronic amphetamine abusers. This approach is limited in that it is primarily correlational and cannot rule out the possibility that the observed behavioural changes preceded the start of amphetamine abuse. Recent studies have provided direct evidence for amphetamine sensitization in drug-naïve human subjects (Sax and Strakowski, 2001, Strakowski and Sax, 1998, Strakowski et al., 1996, Strakowski et al., 1997, Strakowski et al., 2001). In one study, subjects were given exposure to a single dose of amphetamine at three separate time points, with a number of pre-selected behaviours being recorded during the first and third amphetamine exposure. Subjects given the full amphetamine treatment showed increased rates of eye-blink responses and increased motor activity following the third amphetamine exposure, as compared to their response following the first or second exposures (Strakowski and Sax, 1998). A later study conducted by a different group provided additional evidence for amphetamine-induced behavioural changes and, further, showed that amphetamine exposure was accompanied by a decrease in D2 receptor radioligand binding ([11C] raclopride) in the ventral striatum following re-exposure to amphetamine, indicative of increased mesolimbic dopamine activity (Boileau et al., 2006). A similar change was also observed in the caudate–putamen, but this only emerged 1 year after amphetamine exposure (Boileau et al., 2006). These studies demonstrate that both behavioural and neurochemical sensitization can be produced in humans under procedures similar to those used in animal studies.
The notion that the amphetamine-induced sensitized state can be used to model schizophrenia is based on a number of lines of evidence. First, the induction of sensitization appears to depend, at least in part, on the development of a hyperdopaminergic state (Paulson and Robinson, 1996, Robinson and Camp, 1987). Second, like schizophrenia, sensitization is a long-lasting phenomenon; a sensitized response has been detected one year after amphetamine treatment in the rat (Paulson et al., 1991). Third, cross-sensitization has been shown to occur between amphetamine and stressors in rodents (Antelman et al., 1980) and it has been suggested that stress can trigger psychosis in susceptible individuals through a sensitization-like process (Myin-Germeys et al., 2005). Further, prior exposure to either amphetamine or stressors can produce elevations in hypothalamic-pituitary-adrenal (HPA) axis response to subsequent stressors, as assessed by glucocorticoid release, suggesting that the HPA axis itself can be sensitized as a result of amphetamine exposure (Barr et al., 2002, Vanderschuren et al., 1999). Fourth, psychotic symptoms have long been known to occur following chronic amphetamine abuse (Bell, 1965, Connell, 1958, Guttman and Sargant, 1937, Young and Scoville, 1938). In many cases these symptoms are quite advanced, and can include complex delusions and hallucinations (Janowsky and Risch, 1979). Moreover, acute amphetamine administration can produce or enhance a psychotic reaction in patients with schizophrenia at doses that are ineffective in healthy controls (Curran et al., 2004, Lieberman et al., 1987), although increased psychotic reaction to repeated amphetamine in schizophrenic patients has not always been observed (Strakowski et al., 1997). The ability for amphetamine to induce psychosis in chronic amphetamine abusers is maintained long after the drug abuse has been stopped (Sato et al., 1983), suggesting that amphetamine abusers undergo a long lasting change in physiology that produces an increased susceptibility to relapse/symptom re-emergence, as appears to be the case in schizophrenia. Fifth, studies of amphetamine abusers have shown that amphetamine can induce psychosis in non-psychotic individuals (Angrist and Gershon, 1970, Curran et al., 2004, Janowsky and Risch, 1979). Finally, amphetamine-induced dopamine release is enhanced in some patients with schizophrenia (Abi-Dargham et al., 1998, Laruelle et al., 1996, Laruelle et al., 1999). The findings of increased behavioural and neurochemical responses to amphetamine in patients with schizophrenia are reminiscent of the increased responses to amphetamine observed after sensitizing regimens of amphetamine in animals. On the basis of this body of evidence some investigators have proposed that a sensitization-like process, targeting dopaminergic systems, may contribute to the pathophysiology of schizophrenia (Howes et al., 2004, Kapur, 2003, Laruelle, 2000, Lieberman et al., 1997, Ujike, 2002).
Given the extensive overlap between the neurochemical changes observed in the amphetamine-sensitized animal and those observed in the schizophrenic patient, it is likely that this model can be used to both further understand how excessive dopaminergic activity may contribute to the broad changes in behaviour and cognition found in schizophrenia as well as to guide the development of novel therapeutic approaches to the treatment of schizophrenia. Further, recent work has suggested that the sensitization model may be useful in identifying the molecular pathways involved in schizophrenia and/or anti-psychotic response (Ko et al., 2006). The following sections of the article review evidence that the amphetamine-induced sensitized state impacts behaviours, other than simple locomotion or stereotypy, which more accurately reflect the cognitive and behavioural disruptions seen in schizophrenia.
Section snippets
The amphetamine-induced sensitized state as a model of positive symptoms
The positive symptoms of schizophrenia include delusions, hallucinations and disorganized thinking (Crow, 1980). There are two general methods of addressing the issue of these types of symptoms in animal models. First, it is possible to induce behaviours that are assumed to reflect an underlying psychotic state. Although this approach is arguably unsuitable for rodent studies, it has met with some success in studies using non-human primates (Castner and Goldman-Rakic, 1999, Castner and
The amphetamine-induced sensitized state as a model of cognitive deficits
Recently the focus of research and treatment in schizophrenia has expanded to include cognition. One development in this area has been the attempt to identify the cognitive domains that are impacted in schizophrenia and to create a battery of cognitive tasks that reliably measure competency in each of these domains, a process that has been named the MATRICS initiative (Green et al., 2004). At present, seven cognitive domains have been identified that are impacted in schizophrenia, including
The amphetamine-induced sensitized state as a model of negative symptoms
Negative symptoms of schizophrenia have rarely been addressed in animal models of schizophrenia. Ellenbroek and Cools (2000) suggested that anhedonia and social withdrawal are the only two negative symptoms of schizophrenia amenable to study in animals. While the amphetamine-induced sensitized state typically results in animals that are hyper-responsive to reward (Vezina, 2004, Wyvell and Berridge, 2001), under certain conditions, rats withdrawn from amphetamine may show a behavioural profile
Sensitization, schizophrenia and drug abuse
In addition to changes in behaviours such as PPI and LI, schizophrenia is associated with increased rates of drug abuse relative to the general population (Fowler et al., 1998, Hambrecht and Hafner, 1996, Regier et al., 1990). While in many cases substance abuse emerges prior to or along with the onset of schizophrenia, a sizeable proportion of schizophrenic patients develop substance abuse problems following the onset of schizophrenia (Hambrecht and Hafner, 1996), suggesting that schizophrenia
Brain changes associated with the amphetamine-induced sensitized state
Repeated exposure to amphetamine, and the resulting sensitized state, are associated with changes in multiple aspects of brain physiology. Perhaps the most frequently described change is at the neurochemical level, where the sensitized state is associated with an enhanced ability of amphetamine, or other challenges, to release dopamine from terminals in the nucleus accumbens and dorsal striatum (Paulson and Robinson, 1995). Other adaptations that have been described for the nucleus accumbens
How complete is the model?
In rats the amphetamine-induced sensitized state is associated with impaired sustained attention and executive function as measured by the 5-CSRTT and attentional set-shifting tasks. Neither working nor long-term memory appears to be affected in rodents, although working memory deficits have been reported in non-human primates (Castner et al., 2005). This suggests that the failure to find impaired working memory in sensitized rats may relate not to problems with the amphetamine-induced
Conclusion
Induction of an amphetamine-sensitized state reproduces many of the deficits associated with schizophrenia, especially those most closely associated with positive symptoms. This conclusion is most strongly supported by those studies that have demonstrated amphetamine sensitization-induced deficits in behaviours such as LI and PPI that are quite similar in form to those observed in patients with schizophrenia. Additionally, induction of an amphetamine-sensitized state appears to reproduce many
Acknowledgements
REF was supported by a research fellowship from the Ontario Mental Health Foundation. SK is supported by a Canada Research Chair and by a Special Initiative grant from the Ontario Mental Health Foundation.
References (233)
- et al.
Influence of novel versus home environments on sensitization to the psychomotor stimulant effects of cocaine and amphetamine
Brain Res
(1995) - et al.
Increased successive negative contrast in rats withdrawn from an escalating-dose schedule of d-amphetamine
Pharmacol Biochem Behav
(2002) - et al.
Effects of withdrawal from an escalating dose schedule of d-amphetamine on sexual behavior in the male rat
Pharmacol Biochem Behav
(1999) - et al.
Exposure to repeated, intermittent d-amphetamine induces sensitization of HPA axis to a subsequent stressor
Neuropsychopharmacology
(2002) - et al.
Reward system depression following chronic amphetamine: antagonism by haloperidol
Pharmacol Biochem Behav
(1980) - et al.
Behavioural and biochemical effects of dopamine and noradrenaline depletion within the medial prefrontal cortex of the rat
Brain Res
(1980) - et al.
Long-lasting psychotomimetic consequences of repeated low-dose amphetamine exposure in rhesus monkeys
Neuropsychopharmacology
(1999) - et al.
Amphetamine sensitization of hallucinatory-like behaviors is dependent on prefrontal cortex in nonhuman primates
Biol Psychiatry
(2003) - et al.
Amphetamine sensitization impairs cognition and reduces dopamine turnover in primate prefrontal cortex
Biol Psychiatry
(2005) - et al.
A neurobiological basis for substance abuse comorbidity in schizophrenia
Biol Psychiatry
(2001)