Amid renewed interest in the role of infectious agents in chronic diseases, the parasite Toxoplasma gondii has emerged as a potential contributor in the development of psychiatric disorders. T.gondii is one of the most widely distributed parasites worldwide, infecting about a third of the human population, with a prevalence ranging from below 10% to above 70% depending on geographic location (Pappas et al, 2009). While adult-acquired T.gondii infection was previously believed to be largely asymptomatic in immunocompetent individuals, several lines of converging evidence have challenged this supposition: studies suggest that the parasite manipulates infected rodents' behaviour and T.gondii infection has been associated with altered humans cognition. The parasite could have more serious health implications than subtle cognitive alterations as T.gondii infection has been strongly correlated with the incidence of psychiatric disorders in humans, notably schizophrenia. This multifactorial psychiatric disease that affects up to 1% of the world but its aetiology remains elusive (Avramopoulos et al, 2015). Several scientists have questioned whether the protozoan parasite may play a role in the development of this condition.
In this review, following a brief overview of T.gondii's lifecycle, the literature investigating this parasite's potential manipulation of rodent behaviour shall be assessed. Next, the existing literature regarding the association of T.gondii and human cognitive alterations and psychiatric conditions, particularly schizophrenia, will be reviewed. Studies investigating pathways by which T.gondii may affect its host and potentially be implied in the development of schizophrenia shall be presented since each theory's legitimacy relies on a viable hypothetical mechanism. These hypotheses include the potentially deleterious consequences of neuroinflammation and parasitic modulation of dopaminergic and glutamatergic neurotransmission.
A brief overview of T.gondii's lifecycle
T.gondii is an obligate intracellular apicomplexan parasite with a multistage lifecycle that can infect and asexually replicate in all warm-blooded animals. The parasite exclusively undergoes sexual reproduction within the gut enterocytes of its definitive hosts, felines, culminating in the production of oocysts that are shed into the environment via cat faeces (Fig.1) (Severance et al, 2016).
Infection of hosts can occur following ingestion of oocysts from contaminated water and soil, ingestion of parasitic tissue cysts in undercooked meat from infected animals, and vertical transmission from mother to foetus. Upon oral infection, the rapidly replicating lytic form of the parasite called tachyzoite causes an acute infection beginning in the gut (Severance et al, 2016). The parasites breach the gut and disseminate throughout the body where they infect a broad spectrum of nucleated host cells. During acute infection, the tachyzoites replicate asexually and destroy host cells as they egress from one to infect others (Kannan and Pletnikov, 2012).
Around 10-14 days post-infection, the host's immune system subdues the parasite and the tachyzoites convert into slow-replicating bradyzoites that cluster to form tissue cysts. During latent infection, the quiescent intracellular tissue cysts form mainly in the brain and muscle tissue and persist for the remainder of the host's life (Severance et al, 2016).
Part 1: Rodent manipulation
The manipulation hypothesis posits that certain parasites alter their host's behaviour to their own selective advantage, often to enhance transmission to the next host in the parasite's lifecycle. Studies demonstrating T.gondii-induced increased activity levels, cognitive deficits, and inappropriate response to cat odours in rodents have laid the foundations of the now entrenched dogma that T.gondii manipulates rodent behaviour to increase their predation rates and enhance transmission to its definitive feline host (Webster, 2001). However, there is conflicting evidence and the design of some of these studies puts into question the reliability of their results. Moreover, the assumptions on which this hypothesis is based are debatable, meaning that the behavioural changes observed in infected rodents may be coincidental byproducts of T.gondii infection rather than a deliberate ploy to advance the parasite's agenda.
Alterations in rodent behaviour have been studied by observing the animals and comparing the performance of T.gondii-inoculated rodents to their uninfected counterparts when faced with certain tasks used as proxies to measure competencies.
Experiments have yielded conflicting results regarding the behavioural alterations in T.gondii-infected rodents. Mice are naturally neophilic animals, but their response to novel stimuli appears to change upon T.gondii infection. Compared to their uninfected counterparts, infected mice spend more time in the familiar section of the experimental apparatus rather than the novel, previously blocked, section (Hay et al, 1984; Hutchison et al, 1980). This has been interpreted as T.gondii-induced neophobia (Hay et al, 1984) or deficits in spatial recognition memory leading to a lower ability to discriminate between familiar and novel surroundings (Hutchison et al, 1980). Infected mice also exhibit increased activity and exploratory behaviour (Hodkova et al, 2007; Hay et al, 1984; Kannan et al, 2010) which supports the manipulation hypothesis as cats are more likely to prey upon moving and exposed targets. Contrary to these findings, another study reported T.gondii-infected mice to be less active (Hutchison et al, 1980).
Whilst diminished cognitive abilities such as learning and memory have been reported in both acutely- and chronically-infected mice using labyrinths and mazes (Piekarski et al, 1978; Witting, 1979; Hodkova et al, 2007), these findings have also been contradicted using alternative experimental designs. For example, impaired spatial recognition memory was shown in chronically-infected female mice using a radial arm maze (Hodkova et al, 2007), yet other studies using object recognition and spatial placement tests (Gulinello et al, 2010) or a Y-maze test (Kannan et al, 2010) failed to detect this deficiency in chronically-infected male and female mice, respectively. Moreover, the cognitive deficits observed in infected rats were less severe than those in mice, with lower learning difficulties (Piekarski et al, 1979) and normal memory function (Witting, 1979).
 
The milder cognitive abnormalities observed in rats may result from species-specific differential resistance to T.gondii. Indeed, the parasite not only has more conspicuous effects on mice behaviour but also their health: mice have an increased potential for severe morbidity during acute infection, higher rates of brain infection during latent toxoplasmosis, and have been found to run in circles with their heads bent to one side while infected rats appeared largely unaffected (Witting, 1979). Importantly, the behavioural changes observed in mice are associated with acute symptoms of infection and appear to be transient, suggesting that they may be a non-specific byproduct of parasitic infection (Hrda et al, 2000).
Irregularities in the experimental method may also account for these inconsistent results. For instance, T.gondii's effects may be strain-dependent: the Prugniaud (PRU) strain but not ME49 strain elicited increased locomotor activity in mice, as measured by the number of broken beams (Fig. 1A), while ME49 but not PRU caused deficits in spatial working memory, as measured by number of alternations the mice made in a three-armed Y-maze without entering the same arm twice in a row (Fig. 1B) (Kannan et al, 2010). Furthermore, the parasite may have a differential cognitive effect depending on the time of infection. Indeed, the "neophobia" observed in infected mice was more accentuated in congenitally compared to adult-acquired infected mice (Hay et al, 1984). Impaired associative learning and memory has only been observed in mice congenitally-infected during early and intermediate gestation but not late gestation (Wang et al, 2011).
Webster and Berdoy's team applied careful experimental design to reduce the biases and confounding factors that compromised the consistency and reliability of previous studies. These experiments were carried out on wild-trapped rats with variable naturally occurring parasites and parasitic loads and laboratory/wild hybrid rats with experimentally induced congenital or adult-acquired parasitic infections. The animals were maintained under naturalist habitats and/or social conditions and controls were used that were matched for sex, age, and weight and hybrid rats with induced S.muris infection (which does not require a definitive host to fulfil its lifecycle) to control for a generalised response to parasitism. Increased activity levels and a higher propensity to explore novel stimuli was demonstrated in infected rats. This decreased "neophobia" in innately neophobic rats could be explained as a memory deficiency-induced inability to discriminate between familiar and novel surroundings. Importantly, the altered behavioural patterns observed in T.gondii-infected rodents appeared to be specific to T.gondii infections rather than a generalised consequence of parasitic infection (Webster, 1994; Webster et al, 1994; Berdoy et al, 1995).
"Fatal attraction" to cat odour is the most consistently demonstrated altered behaviour in infected rodents. Cat urine usually elicits strong innate fear in rodents as they associate this smell with predation risk. However, this aversion is diminished and even turned into imprudent attraction following T.gondii infection (Kannan et al, 2010; Vyas et al, 2007; Berdoy et al, 2000; Lamberton et al, 2008). This phenomenon is specific to cat urine and not the odours of other predatory nor non-predatory animals (Berdoy et al, 2000; Lamberton et al, 2008) and is not the result of olfactory dysfunction (Vyas et al, 2007).
Interestingly, T.gondii selectively alters specific rodent behaviours that enhance predation rate while leaving while other comportments such as social status and mating success intact (Berdoy et al, 1995). Together, these findings support the hypothesis that the alterations observed are specific manipulations to facilitate T.gondii transmission to felines. It should be noted however that all the previously mentioned rodent experiments involved sample sizes that fall consistently and sometimes significantly below 50, thereby compromising the statistical robustness of these studies.
The manipulation hypothesis is an elegant narrative to explain the cognitive deficits and behavioural alterations observed in T.gondii-infected rodents. However, it is based on a set of assumptions that have recently been put into question. Indeed, there is no direct evidence that T.gondii infection increases rodent predation rates. In addition, the parasite's fitness may not rely on its sexual reproduction in felines and dissemination of oocysts. If T.gondii propagates sufficiently through its other means of transmission then the selective pressure to drive the parasite to develop such sophisticated manipulative strategies would be insufficient (Worth et al, 2013).
Part 2: Human cognition and mental health
Cognitive alterations comparable to those observed in infected rodents have been reported in latently-infected humans and range from subtle cognitive deficits such as delayed reaction times and poorer IQ, to severe psychiatric conditions, notably schizophrenia. Altering human behaviour has no adaptive significance to the parasite as it does not influence its likelihood of transmission to cats and is probably a 'by-product' pathology of T.gondii infection (Webster et al, 2013). Studies investigating the link between human behaviour and T.gondii determine the presence of infection by the positive detection of serum anti-T.gondii antibodies (seropositivity). The serointensity of anti-T.gondii antibodies has been used as a proxy for the duration of infection. However, the assumption that decreased antibody titres accurately reflect increased infection time has been questioned (Webster, 2001).
A. Cognitive deficits
Latently-infected individuals exhibit poorer reaction times than non-infected people, as demonstrated by their delayed response in an acoustic signal reaction time test (PÅ™Ãplatová et al, 2014) and a computerised three-minute reaction test (HavlÃcek et al, 2001). This deficit in motor performance has been hypothesised to account for the 2.65 higher incidence of traffic road accidents in latently-infected individuals compared to seronegative people (Flegr et al, 2002).
Early investigations have reported T.gondii-infection to be associated with slow learning abilities (Langset, 1975). T.gondii-infected individuals scored lower in verbal intelligence on the Otis test and have a lower probability of achieving secondary education than non-infected subjects (Flegr et al, 2003). However, caution must be taken when interpreting these results as the subjects studied were male military personnel and may not accurately represent the general population. In addition, this experiment failed to consider the area of residence (rural vs. urban) and the corresponding lifestyle of the studied individuals. This confounding factor is significant as it may affect the likelihood of achieving higher education and studies have demonstrated that T.gondii infection is more prevalent in rural areas (Kodym et al, 2000). Indeed, when the association between T.gondii infection and intelligence/education level was adjusted for the place of residence, the association disappeared (Novotna et al, 2005). Yet, other research also suggests an association with these types of deficits, for instance, children from the same social background on average exhibit a lower IQ (93) compared to uninfected controls (110) (Alford et al, 1974). Interestingly, a questionnaire also revealed that infected humans showed an attenuated aversion to domestic cat odour, similar to the "fatal attraction" seen in rodents (Flegr et al, 2011).
It is unclear whether T.gondii infection induces these cognitive deficits or on the contrary these deficits increase the individual's susceptibility to infection. While the causal relationship between host behaviour and T.gondii infection has been experimentally established in rodents, the same cannot be done in humans. However, the similarity of the cognitive deficits observed in both rodents and humans supports the idea that the parasite may also be the perpetrator of human cognitive deficits. Furthermore, a reaction time test reported a negative correlation between the serointensity of anti-T.gondii antibodies and the intensity of the observed deficit, indicating that T.gondii might induce the change as the trait intensified with the length of infection (HavlÃcek et al, 2001).
Alternatively, a third factor could be responsible for both an increased infection risk and exhibiting cognitive deficits. For example, rural living is associated with increased risk of T.gondii infection and a lower average education level. All confounding factors must be eliminated before a truly reliable association between T.gondii and cognitive alterations can be established.
B. Psychiatric disorders
Early investigations into the association of T.gondii infection and psychiatric disease yielded conflicting results. For example, serological tests performed on mental disorder patients from one institution found T.gondii's seroprevalence to be equal to that of the general population (Burkinshaw et al, 1953), whereas a study performed in a different institution reported T.gondii seroprevalence to be 30% higher (Elias and Porsche, 1960). The inconsistency of these results could be accounted for by several significant variables being overlooked in these experiments, such as the type of T.gondii infection (congenital, acute, chronic, or latent) or potential risk factors for infection such as socio-economic status (Webster, 2001).
Later studies controlled for these caveats and more reliably reported T.gondii infection's correlation with OCD (Miman et al, 2010b), Parkinson's disease (Miman et al, 2010a), Alzheimer's disease (Kusbeci et al, 2011), suicide attempts (Arling et al, 2009), bipolar disorder (Pearce et al, 2012), and self-directly violence (Zhang et al, 2012). A meta-analysis taking into account a range of confounding factors confirmed the association between T.gondii and bipolar disorder (OR: 1.52), addiction (OR: 1.94), and OCD (OR: 3.4) (Stutterland et al, 2015). This correlation could reflect an increased risk of T.gondii infection in psychiatric mental facilities or may indicate that T.gondii plays a role in the aetiology of such conditions.
Since 1952, an overwhelming number of studies has associated T.gondii seropositivity with one psychiatric condition in particular: schizophrenia. Over forty studies investigating this association across a range of countries and epidemiological conditions have demonstrated increased T.gondii seroprevalence in schizophrenia patients (Torrey et al, 2007). In 2012, a meta-analysis of 38 papers reported that T.gondii infection increases the risk of schizophrenia roughly 2.7 times (Arias et al, 2012). This relationship is stronger than the association between schizophrenia and any human gene in the genome-wide linkage analysis study (OR ≤ 1.40) (Purcell et al, 2009).
However, it should be noted that some studies failed to reproduce this association (Avramopoulos et al, 2015; Sugden et al, 2016). Many caveats such as variations in the method of antibody measurement, time and form of infection, parasitic strain, and genetic background of the host may all partly account for the heterogeneity of the results. A more recent meta-analysis attempted to more reliably assess the magnitude of the association between T.gondii and schizophrenia by taking into account potential moderators. They adjusted the calculated odds ratio for confounding factors such as study quality, gender, high antibody titres, mean age, disease phase-specific rates, seroprevalence of healthy control population, regions where the study was performed, and publication bias. The study found a more modest but still significant association between T.gondii and schizophrenia (OR: 1.43). However, the combined covariates analysed still only accounted for 56% of the variance of the data studied. This could potentially be explained by the state of the disease (chronic or acute) that could not be assessed (Stutterland et al, 2015).
There has been considerable debate over the nature of the relationship between T.gondii and schizophrenia, and while the correlation between the two entities is strong, causality remains unproven. Serum and CSF sample analyses from first-episode schizophrenia patients showed that they already had increased T.gondii seroprevalence compared to unaffected controls (Yolken et al, 2001) and samples from military personnel revealed a significantly elevated seroprevalence in individuals that developed schizophrenia up to three years prior to disease onset (Niebuhr et al, 2008). However, military personnel are not representative of the general population. The previously discussed meta-analysis also found seropositivity to precede schizophrenia onset, indicating that T.gondii infection may be a factor in the aetiology schizophrenia. However, this study did not conclude this causal relationship with certainty (Stutterland et al, 2015).
Cognitive deficits are a core feature of schizophrenia, with 85% of patients exhibiting some degree of cognitive impairment, including deficits in speed of processing, attention, learning, memory, reasoning, and social cognition (Kannan and Pletnikov, 2012). Although T.gondii infection may cause or exacerbate the cognitive deficits observed in schizophrenia patients, the evidence remains scarce and inconclusive. One study has reported more severe cognitive dysfunction in schizophrenia, as measured by the time taken to complete a Trail Making Test and the number of errors made on the Wisconsin Card Sorting Test (Brown et al, 2009). However, these patients were prenatally exposed to T.gondii so this may not apply to adult-acquired latent infection. Another experiment reported no significantly increased cognitive dysfunction in seropositive schizophrenia patients aged 13-17 (Shirts et al, 2008).
Part 3: Mechanisms by which T.gondii affects host behaviour
The molecular processes underlying the enigmatic association between T.gondii infection and host cognitive alterations and mental disorders remain largely elusive. Multiple hypothetical mechanisms have been proposed and a selection of the most prominent theories shall be covered here. T.gondii's predilection for the brain places it in a privileged position to directly alter its host's cognition. The parasite has also been posited to affect the brain indirectly via the host's immune system. The dopaminergic and glutamatergic systems are fundamental to many of the behaviours and cognitive functions altered in T.gondii-infected hosts such as locomotor activity, learning, and memory (Beninger, 1983; Riedel, 2003) and have been proposed to play a central role in the pathophysiology of schizophrenia (Torrey and Yolken 2003; Coyle, 2006). Most research therefore focuses on unravelling various mechanisms by which T.gondii potentially modulates these neurotransmitter pathways.
A. Parasitic tropism for specific brain areas
T.gondii could cause cognitive deficits and manipulate specific host behaviours by selectively invading regions of the brain associated with the altered behaviour/cognitive process in question. This hypothesis is based on the following observations:
1) T.gondii has been demonstrated to impair neuronal activity and functionally silence infected neurons (Haroon et al, 2012).
2) Several studies report T.gondii tropism for specific brain areas including the olfactory bulbs, amygdala, nucleus accumbens, cerebral cortex, cerebellum, medulla oblongata, basal ganglia, septohippocampal and perihippocampal regions (Gonzalez et al, 2007, Vyas et al, 2007, Di Cristina et al, 2008, Unno et al, 2008).
However, the reliability the T.gondii tropism finding is compromised as only low total parasite burdens were detected in these experiments (2-500 cysts/mice brain) (Parlog et al, 2015). In fact most studies observe a random distribution of T.gondii cysts across a wide variety of brain regions, rendering this hypothesis unlikely as the disparate distribution of cysts cannot be reconciled with the specific behavioural response consistently elicited in hosts (Tedford and McConkey, 2017).
B. Neuroinflammation
Acute T.gondii infection is characterized by neuroinflammation that could result in permanent immune-mediated cerebral damage and/or alterations in synaptic morphology and organisation with neuropsychological consequences.
Neuroinflammation can cause neural damage and reactive repair, leading to tissue damage, demyelination, and increased apoptosis. Histological analysis of infected mice brains has indeed revealed inflammatory damage characteristics such as progressive deposition of necrotic material and subsequent vesicular occlusion and sclerosis in cyst-containing regions (Werner et al, 1981). Furthermore, increased levels of protein biomarkers for these processes such as C-reactive protein, interleukin-1 beta, interferon γ, plasminogen activator inhibitor 1, tissue inhibitor of metalloproteinases 1, and vascular cell adhesion molecule 1 have been measured in the brains of inoculated mice. Importantly, most of the protein biomarkers upregulated in schizophrenia brain tissue were found to also be elevated in the brain of T.gondii-infected mice, potentially suggesting common pathological pathways. It should be noted that this study used CBA/J mice, a strain that is susceptible to high parasite burden and therefore a strong inflammatory response (Tomasik et al, 2016). However, experiments in chronically-infected mice suggest that inflammation-induced neurodegeneration is limited and therefore is probably only a marginal factor (Parlog et al, 2014).
T.gondii-induced neuroinflammation has also been demonstrated to trigger changes in neuronal morphology and synaptic protein expression. For example, T.gondii infection reduces the levels of synapophysin and PSD95 in the somatosensory cortex and hippocampus and modifies the number and morphology of dendritic spines causing impaired structural connectivity in murine models (Parlog et al, 2014). Alterations in synaptic and dendritic organization have been reported in the aetiology of several neuropsychiatric disorders and linked to decreased cortical volume observed in schizophrenia patients (Harrison, 1999).
However, Coxiella burnetti or Leptospira icterohaemorrhagiae also induce neuroinflammation in rodents but do not change their behaviour (Webster, 1994). The neurophysiological alterations observed during T.gondii infection are therefore not a consequence of neuroinflammation alone.
C. Dopaminergic neurotransmission
Dopaminergic dysregulation has emerged as one of the most potent hypothetical mechanisms associating T.gondii infection with cognitive abnormalities and mental disorders. This theory originates from several lines of evidence:
1) Disrupted dopamine levels have been observed in T.gondii-infected humans (Flegr et al, 2003) and T.gondii has been demonstrated to enhance the concentration and release of dopamine from mammalian dopaminergic cells (PC12 cell line) in vitro and in infected mice (Prandovszky et al, 2011; Stibbs, 1985). Although it should be noted that studies have yielded contradictory results, reporting no significant change in dopamine levels in PC12 cells nor in infected mouse brains (Wang et al, 2015; Goodwin et al, 2012), and demonstrating that dopamine dysregulation returns to baseline once the acute T.gondii infection has resolved (Gatkowska et al, 2013).
2) Selective dopamine uptake inhibitors suppressed the exploratory behaviour (Skalloyá et al, 2006) and significantly reduced the suicidal feline attraction observed in T.gondii-infected rodent (Webster et al, 2006).
3) A number drugs used to treat schizophrenia such as haloperidol and valproic acid inhibit the replication of T.gondii in vitro (Jones-Brando et al, 2003). This suggests that these therapies may attenuate psychiatric symptoms in part by exerting anti-parasitic activity.
The parasite possesses two genes that encode tyrosine hydroxylases (AAH1 and AAH2), homologs of the catecholamine biosynthetic enzymes that represent the rate-limiting step in dopamine synthesis (Fig.3.). These enzymes have been hypothesised to be the effectors by which the parasite increases cerebral dopamine by producing excessive quantities of its precursor L-DOPA (Gaskell et al, 2009). Furthermore, antibody staining of chronically-infected mice brain sections found T.gondii-encoded tyrosine hydroxylase (TgTH) present in intracellular tissue cysts (Fig.4.) (Prandovszky et al, 2011).
However, a study reported that AAH enzymes in T.gondii do not cause global alterations of host dopamine. Overexpression of the AAH2 gene surprisingly did not result in a parasite significantly increasing host dopamine levels in vitro nor in vivo. The same study also found that AAH1 and AAH2 expression was negligible in tachyzoites and very low in bradyzoites (Wang et al, 2015).
Alternative pathways have been suggested for the orchestration of T.gondii-mediated dopamine upregulation. For instance, T.gondii-infection consistently upregulates the expression of MiR-132, an miRNA associated with increased dopamine levels (Xiao et al, 2014).
D. Glutamatergic neurotransmission
T.gondii has been hypothesised to induce hypoglutamatergic activity by dysregulating kynurenic acid (KYNA). KYNA is a tryptophan metabolite that antagonizes NMDA and α7nACh GLU receptors. Elevated cerebral KYNA levels are associated with cognitive impairments in rodents such as deficits in contextual learning and memory (Chess and Bucci. 2006; Chess et al, 2007) and have been displayed in the CSF and postmortem prefrontal cortex of schizophrenia patients (Erhardt et al, 2001; Schwarcz et al, 2001). In accordance with this theory, KYNA downregulation by knocking out its synthetic enzyme kynurenine aminotransferase II ameliorated the cognitive deficits observed in mice (Potter et al, 2010).
Astrocytes are key to the synthesis of KYNA and examining the brains of chronically T.gondii-infected mice revealed massive astrocyte activation and a greater than 7-fold increase in cerebral KYNA (Guidetti et al, 2006). This probably results from increased activity of enzymes such as indoleamine 2,3-dioxygenase (IDO) or tryptophan 2,3-dioxygenase (TDO) upstream in the kynurenine pathway that catabolises tryptophan into KYNA and quinolinic acid (Fig.5.), although further research is needed to confirm this.
In addition, fluctuating KYNA levels indirectly modulate dopaminergic activity via glutamatergic pathway dysregulation. Changes in extracellular glutamate influences striatal dopamine release (Sakai et al, 1997) and focal intrastriatal infusion of KYNA in rats indeed reversibly decreases extracellular dopamine concentrations (Rassoulpour et al, 2005).
T.gondii-associated NMDAR autoantibodies have also been posited to underlie the hypoglutamatergic activity observed in infected individuals (Kannan and Pletnikov, 2012). Moreover, the parasite drives systemic inflammation which compromises the impermeability of the blood-brain-barrier, thereby enabling the autoantibodies to access the immune privileged CNS and produce detrimental neuropsychiatric symptoms. A recent study reported the association of NMDAR autoantibodies and markers of endothelial barrier permeability (gluten-IgG and S100B for gut-blood and blood-brain barrier dysfunction, respectively) with T.gondii infection in schizophrenia patients compared with controls. Mouse cohorts were then used to demonstrate that T.gondii infection induced endothelial barrier defects and caused sustained elevations in anti-NMDAR IgG. Furthermore, combined T.gondii and NMDAR antibody seropositivity in schizophrenia were shown to results in higher degrees of cognitive impairment as measured by tests of delayed memory (Kannan et al, 2017).
The inflammatory response to T.gondii infection has been posited to foster an autoimmune-prone environment that promotes the generation of these NMDAR autoantibodies. Alternatively, T.gondii-specific antibodies have been speculated to cross-react with NMDARs and analysis of T.gondii's proteome and NMDAR's subunits found a massive epitope overlap, particularly the NMDA 2D subunit (Lucchese, 2017). However, UV-inactivated parasites failed to elicit an NMDAR autoantibody response, indicating that these antibodies are not the result of cross-reacting anti-T.gondii antibodies (Kannan et al, 2017).
The majority of these findings were based on rodent models. Despite being key to unravelling the molecular mechanism underlying the T.gondii-schizophrenia association, the absence of a true 'schizophrenia rodent model' and the innate differences between rodents and humans means that caution must be taken when extrapolating human conclusions based on these experiments.
Conclusion and future
Despite being one of the most studied links between a pathogen and a psychiatric disorder, it remains unclear whether T.gondii is implicated in the aetiology of schizophrenia. The hypothesis is supported by the evidence for T.gondii-associated cognitive deficits mirrored in rodent and human hosts, by the strong correlation between T.gondii infection and schizophrenia, and the neuropathological commonalities between these two entities. While T.gondii is very unlikely to cause schizophrenia alone, it may have deleterious cognitive effects that augment the likelihood of developing this condition or exacerbates symptoms in already affected individuals. The molecular mechanisms underpinning T.gondii's potential cognitive effects are increasingly proving to be complex and comprising of a range of interacting pathways including neurotransmission and the host's immune response.
Though inconclusive, the research presented establishes a strong rationale for further investigation into T.gondii's association with psychiatric disease. Current integrated hypotheses posit that psychopathologies are triggered when environmental disturbances interact with genetic predispositions (Severance et al, 2016). There is a need to generate more sophisticated/comprehensive animal models of infection-induced cognitive dysfunction, integrating genetic and environmental factors.
Future research to disentangle the heterogeneous aetiology of schizophrenia and substantiate the role of T.gondii in a subset of cases will hopefully facilitate the development of personalised treatment strategies directed towards the disease's causative agents.