Parasites as causative agents of human affective disorders? The impact of anti-psychotic, mood-stabilizer and anti-parasite medication on Toxoplasma gondii's ability to alter host behaviour

1. Introduction

Certain parasites selectively alter host behaviour to enhance their transmission (Barnard 1990). The protozoan Toxoplasma gondii provides a convincing example of such manipulation. T. gondii has an indirect life cycle, in which members of the cat family are definitive hosts (Hutchison et al. 1969). If oocysts shed with the faeces of an infected cat are ingested by an intermediate host such as a wild rodent, or another secondary host such as a human or domestic livestock, schizogony occurs and small thin-walled cysts form in various tissues and organs, most commonly the brain, where they remain potentially for the host's lifetime (Remington & Krahenbuhl 1982). The parasite completes its life cycle when a cat consumes an infected intermediate host (Hutchison et al. 1969). In addition to that inherent for all indirectly transmitted parasites, since sexual reproduction of T. gondii can be accomplished only in the feline, there are likely to be particularly strong selective pressures on the parasite to evolve mechanisms to enhance the transmission rate from the intermediate to the definitive host. The predilection of the parasite for the brain of its intermediate host places T. gondii in a privileged position to cause such a manipulation (Werner et al. 1981). Accordingly, studies on rats have demonstrated that T. gondii causes an increase in activity (Webster 1994) and a decrease in neophobic (fear of novelty) behaviour (Webster et al. 1994), both of which may facilitate transmission to the feline definitive host. Moreover, while rats have innate defensive reaction to predator odours, in particular feline (Blanchard et al. 1990), T. gondii appears to alter the rats’ cognitive perception of cat predation risk, turning their innate aversion into a ‘suicidal’ feline attraction (Berdoy et al. 2000).

T. gondii's ability to infect all mammals, often at very high prevalence, with for example, 20–80% of humans infected (Desmonts & Couvreur 1974), makes the implications of its behaviour-altering activity of significant theoretical and clinical importance. In humans, while the often severe sequelae of congenitally acquired toxoplasmosis are well known, most infections are acquired postnatally (Webster 2001). Latent toxoplasmosis results when the, usually mild, symptoms of the acute stage disappear after a few weeks (Remington & Krahenbuhl 1982). While cases of psychiatric complications such as disorientation, anxiety, depression and even psychoses with schizophreniform characters are well recognized within the immunosuppressed, due to either acute infection or secondary reactivation of the disease (Arendt et al. 1999), latent toxoplasmosis has been traditionally viewed as asymptomatic within immunocompetent humans and animals (Roberts & Janovy 2000). This view, however, appears to be largely based on a lack of research into the later stages of the disease and, more specifically, possible associated diseases. Indeed, similar psychiatric complications and meningoencephalitis can also occur within T. gondii-infected immunocompetent human hosts (Couverur & Thulliez 1996; Carme et al. 2002; Kaushik et al. 2005), and recent human studies have revealed that latent toxoplasmosis may even cause personality changes (Flegr et al. 1996), decreased IQ (Flegr et al. 2003) and psychomotor performance (Havlícek et al. 2001). T. gondii has also been demonstrated to contribute to the formation of certain types of brain tumours (Ryan et al. 1993), and a potential link between T. gondii and neuropsychiatric disorders, in particular schizophrenia, has been proposed.

While any potential association between toxoplasmosis and the development of schizophrenia is likely to occur only in a small proportion of infected individuals, neuropathological studies have reported that glial cells, especially astrocytes, are selectively affected in both toxoplasmosis and schizophrenia (Halonen et al. 1996; Cotter et al. 2001), as are levels of dopamine, serotonin, norepinephrine and other neurotransmitters (Torrey & Yolken 2003). Likewise, epidemiological studies have demonstrated an association between T. gondii infection with increased incidence of schizophrenia (Torrey et al. 2000; Torrey & Yolken 2003). For example, analyses of serum samples obtained from mothers shortly before or after giving birth revealed a significantly raised proportion of IgM antibodies to T. gondii in those whose children subsequently develop schizophrenia in later life (Torrey & Yolken 2003), and individuals suffering from first-episode schizophrenia have significantly elevated levels of IgG, IgM and/or IgA class antibodies to T. gondii antibodies, within both serum and cerebral spinal fluid (CSF), compared to uninfected control subjects (Yolken et al. 2001). Indeed, meta-analysis here of all reported studies testing for an association between T. gondii antibody titres and first incidence schizophrenia status (reviewed in Torrey & Yolken (2003)) revealed that 18 of the 19 studies showed higher (11 significantly higher) T. gondii prevalence within schizophrenia patients than controls (table 1). Of particular interest here are studies that have demonstrated that T. gondii antibodies of schizophrenia patients treated with anti-psychotic drugs are intermediate between those of patients never treated and those of control groups, with a significant further reduction in those patients undergoing current drug treatment, suggesting that anti-psychotic treatment may affect T. gondii levels (Leweke et al. 2004). This is supported by the observation that many anti-psychotic drugs commonly used in the treatment of schizophrenia inhibit the replication of T. gondii tachyzoites in cell culture (Jones-Brando et al. 2003). One could, therefore, conclude that the anti-psychotic and mood stabilizing activity of some medications may be achieved, or at least augmented, through their inhibition of T. gondii replication and invasion in infected individuals.

The aim of the current study was thus to test the hypothesis that anti-psychotic drugs commonly used in the treatment of human affective disorders such as schizophrenia inhibit T. gondii replication in vivo and thereby help alleviate T. gondii-induced cognitive and behavioural alterations. In particular we predicted, using the epidemiologically and clinically applicable rat-T. gondii model system, and following a previously described and neurologically characterized ‘feline attraction’ protocol (File et al. 1993; Hogg & File 1994; Adamec et al. 1999; Berdoy et al. 2000) that haloperidol (HAL, an anti-psychotic used in the treatment of mental illnesses including schizophrenia) and/or valproic acid (VAL, a mood stabilizer used in the treatment of mental illnesses including schizophrenia), would be, at least as, effective in preventing the development of T. gondii-associated behavioural and cognitive alterations as the standard anti-T. gondii chemotherapeutics pyrimethamine with Dapsone (PD). HAL and VAL were chosen as, in a battery of tests across 12 neuroleptic compounds, both were demonstrated to be most effective in inhibiting T. gondii replication in vitro (Jones-Brando et al. 2003), although no in vivo test has yet been performed. Moreover, HAL's mode of action is thought to involve acting as a dopamine D2 antagonist, with dopamine having been proposed as one of the ‘missing links’ in elucidation of the potential association between schizophrenia and toxoplasmosis (Flegr et al. 2003), particularly since both disorders are characterized by raised levels of this neurotransmitter (Stibbs 1985; Torrey et al. 2000; Torrey & Yolken 2003). The results obtained should have important clinical and theoretical implications.

2. Material and methods

Observations were carried out using adult Lister-hooded laboratory rats. The rat provides a useful model here as the chronic impact of T. gondii within the CNS of rats has been demonstrated to be more applicable to human latent toxoplasmosis, relative to that of the more acute infections in animals such as mice (Hrda et al. 2000; Webster 2001). Moreover, while as yet there are no specific neuropsychiatric disorder rodent models available (Anon. 2003), to test the prediction here that some anti-psychotics can alleviate the impact of T. gondii on host behaviour, through their direct or indirect impact upon the parasite, it is imperative to use a model system with a well characterized T. gondii-altered behavioural repertoire (Crabbe & Morrise 2004).

Experimental rats (n=49) were orally exposed with 20 cysts of the low virulence cyst-forming ME-49 strain. This strain had been maintained by continuous passage of infective brain homogenate in mice from the University of Leeds. Since it has been demonstrated that the route of infection (horizontal versus congenital) has no significant effect on the characteristics of T. gondii-altered behaviour in rats (Webster 1994) and adult-acquired infection is more applicable to human latent infection (because most human congenital infection results in death or severe symptoms), adult-acquired infection was used throughout this study. Uninfected control rats (n=39) were sham orally exposed with an equal quantity of isotonic saline. Rats were divided into eight groups each containing 8–12 rats (four drug treatment groups, each subdivided into an infected and uninfected control group).

The treatments were either: (i) water control; (ii) pyrimethamine (3 mg kg⁻¹ d⁻¹) with Dapsone (PD, 5 mg kg⁻¹ d⁻¹), an anti-T. gondii treatment which has previously been shown to prevent latent toxoplasmosis both in rodents (Derouin et al. 1991; Brun-Pascaud et al. 1996) and humans (Girard et al. 1993); (iii) haloperidol (HAL, 1.5 mg kg⁻¹ d⁻¹), an anti-psychotic used in the treatment of mental illnesses including schizophrenia (Jones-Brando et al. 2003); or (iv) valproic acid (VAL, 40 mg kg⁻¹ d⁻¹) a mood stabilizer used in the treatment of mental illnesses including schizophrenia (Jones-Brando et al. 2003). HAL and VAL were dissolved in distilled water, while PD was dissolved in ethanol. Despite the highly variable doses, which are administered to human patients, and because of the half-life differences of these anti-psychotics between humans and rodents, the doses of each drug administered here were those specifically recommended for rodent models (Kapur et al. 2003). Daily oral drug (or water placebo) administration commenced 14 days after T. gondii exposure, and continued for a further 14 days in all groups. Thus, our aim here was to examine the impact of the drugs primarily on the replicating tachyzoite T. gondii stage, prior to the development of (the relatively more drug-insensitive) bradyzoite cysts.

Drugs were administered via the oral route as this most accurately mimics the most common method by which humans receive the equivalent drug treatments, and also minimizes the amount of stress which drug administration causes to the rats, which was deemed especially important in the current investigation. Likewise for ethical reasons, all drugs (or placebos) were provided contained within fruit-flavoured jelly cubes (Flecknell et al. 1999), to which all rats had been previously trained (habituated) to eat daily. All rats were housed 2–4 per large (0.6 (L)×0.6 (W)×0.3 (H) m) (hence 4–9 times larger that the required Home Office minimum specifications for grouped rats (HMSO 1989), extensively environmentally enriched cage, randomly mixed between experimental groups, with food and water available ad libitum. All rats were monitored closely and none displayed any symptoms of illness or experimentally induced stress at any point during the study. The work was performed under Home office project licenses PPL 30/2032 and PPL 30/1805, and all procedures were classed as ‘mild’.

(a) Feline attraction, activity and behaviour

To test the potential effect of T. gondii and drug treatments on the rat's perception of predation risk, activity levels and behavioural repertoire, we observed the continuous exploratory behaviour of rats in 1×1 m pens (figure 1). Following our previously published protocol (Berdoy et al. 2000), the base was covered with a layer of woodchips to provide a homogeneous surface that could be cleaned between each test. Each corner contained 15 drops of one of four distinct odours deposited within plastic nest-boxes: the rat's own smell (own urine-soiled woodchip bedding), neutral smell (fresh woodchips treated with water), cat odour (fresh woodchips treated with undiluted cat urine) and rabbit odour (fresh woodchips treated with undiluted rabbit urine). Rabbit odour was used as a control for a mammalian non-predator. A food bowl and water bowl were placed on either side of each of the nest-box entrances, each of which was treated to three drops of their respective odour. The position of the four smells was changed between each test to avoid positional biases. The positioning and behaviour displayed within each 50×50 cm area was recorded individually by continuous 10 s interval sampling for a period of four hours per rat, judged to be a sufficient test period based on preliminary trials which revealed that the majority of test subjects, irrespective of infection status or drug treatment, remained in a single nest-box after approximately 2.5 h of a trial. Behavioural traits recorded focused on those likely to increase feline predation rate and hence completion of the parasites life cycle, either through time spent ‘still’ for greater than 3 s while exposed (conspicuousness), or increased activity, either measured as proportion of time spent ‘out’ versus inside a nest-box, or ‘active’ versus inactive, when a rat has already been within a nest-box for greater than 5 min with no nose poking out/movement inside the nest-box); or ‘grooming in exposed areas’ (decreased attention to predator avoidance). Additional behavioural traits recorded included ‘water consumption’ (drinking from water bowl), ‘feed’ (eating from food bowl or spilt pellets); rearing; mount (rat on top of nest-box); or HDB (on the next box with its head down the back). The pen was located in a test room, which did not contain any other rats or potential disturbances during each trial, adjoining the area in which the rats were housed to minimize the stress induced by relocation to the test pen. During each test period, the test room remained fully lit (to facilitate visual clarity of the videos), and the test subject was left undisturbed for the duration of the trial. Prior to infection with T. gondii or administration of drugs, 4 h videotaped trials of a limited number of randomly selected rats (3–4) from each treatment group were recorded (analysis of this pre-infection group (104 rat-hours of observation) revealed as expected, no significant differences between groups). Six to eight weeks after infection and drug administration, each rat underwent a single 4 h videotaped trial (representing the post-infection trials). Six-eight weeks post T. gondii exposure was chosen here as this is when avirulent cyst-forming T. gondii strains tend to form bradyzoites and hence represent the latent state of infection (Hutchison et al. 1969).

• Download figure
• Open in new tab
• Download PowerPoint

At the end of the investigation, all experimental rats were euthanased by rising concentration of carbon dioxide and cervical dislocation. T. gondii antibodies were determined by the IgG indirect latex agglutination test (ILAT: Toxoreagent; Eiken) and IgM Direct agglutination test (BioMerieux, UK). Titres greater than 1 : 16 were considered positive (Webster 1994; Webster et al. 1994), and data from any exposed rat found to be serologically negative in both tests (n=2) were excluded from the analysis. Thus, the final sample size (and median IgG titre among the T. gondii exposed/positive groups) per group was: positive control, n=12 (1 : 32); positive HAL, n=11 (1 : 16); positive PD, n=11 (1 : 32); positive VAL, n=12 (1 : 32); negative control, n=8; negative HAL, n=11; negative PD, n=10; negative VAL, n=10. Notably, those rats within the positive HAL group which failed to seroconvert on the IgG ILAT, were however IgM seropositive on the Direct agglutination test. (Only T. gondii sero-positive rats showed immunohistochemically T. gondii-positive neurons and glial cells; data to be published separately.) Those rats unexposed to T. gondii remained seronegative.

3. Results

(a) Feline attraction

Among untreated rats, those infected spent ‘proportionately’ much more time in the ‘cat’ area than uninfected rats (OR=19.81, 95% CI: 10.49–37.41, p<0.001). This difference was a combination of the untreated infected being both more likely to enter the cat area (OR=1.63, 95% CI: 1.37–1.95, p<0.001 comparing cat with non-cat entries; OR=1.71, 95% CI: 1.38–2.12, p<0.001 comparing cat with rabbit entries) and, having entered, stayed 175% longer on average in the cat area than untreated uninfected rats (95% CI: 110–261% longer, p<0.001, figure 2a).

• Download figure
• Open in new tab
• Download PowerPoint

The effects on the drug treatments differed significantly between infected and uninfected rats (p<0.01 for each measure of feline attraction). Among infected rats, all three drug treatments reduced their likelihood to enter the cat area, although each failed to reach significance (HAL, p=0.12; PD p=0.11; VAL, p=0.74 comparing cat with non-cat; HAL, p=0.22; PD, p=0.13; VAL, p=0.65 comparing cat with rabbit). Both HAL and PD significantly reduced the duration of their stays in the cat area (p=0.02 and 0.01, respectively), while a similar non-significant trend (p=0.11) was observed for VAL (figure 2b). Among uninfected rats, all three drug treatments significantly increased both their likelihood to enter the cat area and the duration of their stays in the cat area (HAL, p=0.004; PD, p<0.001; VAL, p<0.001 comparing cat with non-cat; HAL, p=0.02; PD, p=0.002; VAL, p<0.001 comparing cat with rabbit; HAL, p=0.002; PD, p=0.001; VAL, p<0.001 for duration of stays in cat). However, these increases were in every case less than those associated with untreated infection.

No significant effect of infection was observed on proportion of time spent within the non-predatory control ‘rabbit’ area, indicative of the specificity of the T. gondii feline attraction response (untreated infected versus treated infected, p=0.20). Likewise, there was no overall significant effect of treatment on the proportion of time spent within the rabbit area (effect of drugs among infected: HAL, p=0.39; PD, p=0.64; VAL, p=0.63 and among uninfected: HAL, p=0.93; PD, p=0.20; VAL, p=0.64).

4. Discussion

An apparent suicidal feline attraction and risk behavioural profile among untreated infected individuals was clearly evident here. Under natural conditions, these are all behavioural traits likely to increase predation rate by the definitive host and hence completion of the parasites life cycle (Webster 2001), either through increased risk behaviours (presence in areas with evidence of cat presence), conspicuousness (increased activity or still and exposed), or through decreased attention on predator avoidance (grooming in exposed areas) relative to their untreated uninfected counterparts (figure 2, table 2). Treatment of infected rats with, in order of decreasing efficacy: HAL, PD and VAL, reduced these predator-risk behavioural traits. Therefore, our results lend support to the hypothesis that the anti-psychotic and mood stabilizing activity of some medications used in the treatment of schizophrenia and human affective disorder may be augmented through their inhibitory impact upon T. gondii in infected individuals

In terms of potential mechanistic explanations, the treatments, at least PD and HAL, may function by directly minimizing T. gondii replication and invasion of host brain cells, as has been demonstrated in vitro (Jones-Brando et al. 2003). Indeed, our unpublished observations on T. gondii immunohistochemical staining of tissue sections throughout the brains of these experimental rats indicate that the frequencies of T. gondii exposed animals showing immunohistochemically positive neurons and glial cells was reduced following drug treatment, with a superiority of HAL over PD and VAL (S. Weis & I. C. Llenos 2005, personal communication). Such anti-T. gondii activity may be related, at least in part, to the calcium inhibitory properties of these drugs. T. gondii tachyzoites require calcium in order to invade host cells, and this invasion is inhibited by calcium channel blockers and calmodulin antagonists, including trifluoperazine, another phenothiazine anti-psychotic (Pezzella et al. 1997). Accordingly, HAL and VAL, are each capable of inhibiting calcium transport through cellular ion channels (Itoa et al. 1996; Johannessen 2000).

Another, not mutually exclusive, explanation for the effects of T. gondii and drug treatment on feline avoidance behaviour relates to their potentially neuromodulatory impact, either directly or indirectly (Blanchard et al. 1990; Berdoy et al. 2000; Torrey & Yolken 2003). The reaction by potential prey to cat stimuli is used to study the neurological basis of anxiety and the mechanisms of anxiolytic (anxiety relieving) drugs, and such studies have found, for example that blocking the normally anxiogenic NMDA receptors in the amygdala also causes laboratory rats to ‘fearlessly’ approach areas treated with cat urine (Adamec et al. 1999). Likewise, while exposure of rats to predator odours induces fast wave activity in the dentate gyrus of the hippocampus (File et al. 1993; Hogg & File 1994), such a response can be blocked by serotonin antagonists (Blanchard et al. 1990), or alternatively by the presence in mice of another protozoan, Eimeria vermiformins (Kavaliers & Colwell 1994), which could suggest a similar neuromodulatory action of certain protozoan parasites including T. gondii. Indeed, T. gondii infection is known to be associated with raised dopamine levels in particular (Stibbs 1985; Flegr et al. 2003). Dopamine levels are also often raised within patients with schizophrenia (Torrey & Yolken 2003). Thus, as HAL is a dopamine D2 antagonist (Seeman 1980), one could propose that its superior therapeutic impact here may be through a combination of both its ability to inhibit T. gondii replication and to reduce, directly and indirectly, dopamine levels. This may contrast to the restricted anti-parasitic, rather than neuromodulatory, action of PD. Likewise, while the mechanisms by which VAL may exert mood-stabilizing effects are not yet fully elucidated, recent studies have found little evidence that these are mediated by its effects on serotonin or dopamine receptors (Delva et al. 2002). Thus, one may postulate that the relatively lowered success of VAL in decreasing the parasite-induced behavioural alterations observed here may also be restricted to those achieved through its direct inhibition on parasite replication alone, albeit less efficaciously than PD.

Our results raise several important theoretical and applied implications, from, for example providing further support for the theory of T. gondii as a causative agent in some cases of schizophrenia, to predicting that such drugs may be expected to be particularly effective in individuals with schizophrenia who are also infected with T. gondii. It may thus be worth investigating the T. gondii inhibitory effects of other anti-psychotics, especially the second-generation agents (olanzapine, clozapine, quetiapine, ziprasidone, risperidone and aripiprazole), as well as the effectiveness of alternative anti-T. gondii treatments, such as pyrimethamine with clindamycin, co-trimoxazole, or ponazuril, as adjunct therapies for schizophrenia. These potential treatments could also prove particularly valuable as prophylactic treatments for groups at serious risk of developing psychiatric disorders in later life as a result of T. gondii infection. Moreover, the results here, particularly with HAL, suggest that further research into anti-psychotics may provide potential alternatives for anti-T. gondii drug treatment, where new anti-bradyzoite treatments, in particular, are imperative.

However, the behavioural changes observed among the uninfected, but treated, rats indicate that serious consideration must be given to possible side-effects of such treatments on human behaviour: all three drug treatments caused uninfected rats to behave, albeit to a much reduced extent, in a similar way to infected untreated rats, in terms of feline attraction and activity levels, with further mild behavioural alterations, such as increased water and food consumption. Many of these behavioural alternations may also relate to the neuromodulatory action of such chemotherapeutics, as discussed above. Indeed, any anxiolytic drug effects may well be predicted to result in treated, but uninfected, rats being less averse to the cat smell, and also potentially more active, than their untreated uninfected counterparts, as was observed here (Blanchard et al. 1990). The potential generalizability of such side-effects is certainly worthy of further inter-host specific research. In terms of the current study, the mild drug-induced feline attraction observed here among uninfected rats may, moreover, detract from the drug-induced reductions in the infection-related increases in entrances and duration spent within cat areas.

Our results, as is always the case in research, raise as many questions as they do answers. What will be important to elucidate now is, for example, why any potential effect of T. gondii on host behaviour, in particular in terms of the clinical outcome of human behaviour, may differ between individuals. Potential key factors may relate to inherent differences in individual genetic predisposition, the state of the immune system, the time of exposure (e.g. infections in the first trimester of pregnancy may differ from those in the third trimester; and/or prenatal infection may differ from postnatal), the duration of exposure (e.g. humans live longer than the average rodent intermediate host), and/or the part(s) of the brain affected. From the perspective of the parasite, key factors may relate to the dose, the source of infection (i.e. oocyst or cyst stage consumption), the genotype of the infecting strain, and/or even an interaction with other infectious agents.

In summary, our results to date do demonstrate that the behavioural changes associated with T. gondii can be effectively reduced by those anti-psychotic drug treatments previously demonstrated to inhibit parasite replication in vitro. This may provide further evidence for a potential role of T. gondii in the aetiology of schizophrenia, and specifically here that the actions of anti-psychotics may work in part via parasite inhibition. Clinical trials based on these findings are warranted, including those perhaps of anti-psychotic drugs with patients separated into those with and without additional T. gondii infection. Such trials could lead to improved prognosis and potentially new medication combinations and therapeutic modalities for the treatment of both toxoplasmosis and severe psychiatric disorders.