1. Introduction
Schizophrenia is a debilitating psychiatric disorder affecting 0.45% of adults worldwide, according to the World Health Organisation,
[1]. According to 2019 data, the prevalence of schizophrenia ranges from 0.2 to 0.5% across nations worldwide
[2]. Emerging evidence suggests the role of the immune system and its various components in the risk of schizophrenia, including cytokines, C-reactive protein, chemokines, antibodies, etc.
[3]. Multiple studies show altered protein and mRNA expression of Toll-like receptors (TLRs) in schizophrenia compared to healthy adults
[4][5][6]. Human postmortem studies carried out on individuals diagnosed with schizophrenia, brain samples showed lower
tlr4 mRNA and protein levels, and downregulation of
il6,
il10, and
tnfa mRNA in the prefrontal cortex (PFC) region
[7][8], and elevated TLR4 protein levels in the cerebellum region
[7]. On the contrary, high
tlr4 and
myd88 mRNA expression was observed in the PFC region of the human brain
[9]. On the flip side, studies on peripheral blood show downregulation of
tlr3 and
5 mRNA levels and upregulation of
il6 and
il10 mRNA levels in patients with schizophrenia
[10], which is in contrast to evidence from brain samples. Studies of acute and first-episode psychosis (FEP) patient groups reported higher expression (measured by mean fluorescence intensity (MFI)) of TLR3, 4, and 5, but not TLR2, in monocytes
[11][12]. Although Keri et al. observed higher TLR4 MFI in schizophrenia post-TLR4 stimulation, the MFI expression of IL-6, TNF-α, and IL-1β was lower than controls
[12]. Similar results were also observed by Muller et al. with IL-1β. This was interpreted to represent an increase in TLR expression to compensate for the functional deficit
[11]. In contrast, whole blood stimulation studies in schizophrenia showed elevated concentrations of IL-6, TNF-α, and IL-1β post-TLR2 stimulation, and increased IL-1β alone post-TLR4 and 8 stimulation
[13]. Individual blood cell studies in FEP patients showed lower TLR protein expression in monocytes, B cells, and T cells
[6]. Overall, researchers can infer that dysregulation of TLR expression exists in patients with schizophrenia and varies at various stages of disease progression (
Table 1), but researchers do not have sufficient evidence to implicate their extent of involvement in the immunopathology of schizophrenia.
Characteristics of Human studies illustrating TLRs in various models and cognition. PFC—prefrontal cortex, CB—cerebellum, PBMC—peripheral blood mononuclear cells, LPS—lipopolysaccharide, TLR—Toll-like receptor, RBANS—Repeatable Battery for the Assessment of Neuropsychological Status, MCCB—Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS™) Consensus Cognitive Battery, MyD88—myeloid differentiation primary response 88, NFkB—nuclear factor kappa B, IkBa—, poly (IC)—polyinosinic:polycytidylic acid, LPS—lipopolysaccharide, DSM-IV—Diagnostic and Statistical Manual of Mental Disorders-IV, SCID-CV—Structured Clinical Interview for DSM-Clinical version, ICD-10—International Classification of Diseases-10.
2. Maternal Immune Activation (MIA) and Early Life Exposure
Increased risk for psychosis-like symptoms following MIA (inflammatory stimuli due to TLR activation), conventionally associated with increased production of cytokines and chemokines, in offspring has been theorised for a long time
[20][21]. Infections such as cytomegalovirus, influenza, herpes simplex virus, and coronavirus lead to upregulation of TLR mRNA (2–4 and 7–9). This is also accompanied by an increase in interferon and cytokine production causing complications in pregnancy, miscarriage, or foetal abnormalities
[22][23]. These elevated cytokines and chemokines can migrate from the placenta into the foetus, and such exposure could cause changes to the CNS, leading to abnormal neurodevelopment
[24]. Moreover, conditions such as diabetes mellitus, obesity, and systemic lupus erythematosus (SLE) in pregnancy were linked to high expression of TLRs, cytokines, and chemokines in peripheral blood, and similar levels were reflected in the placenta
[25][26][27]. This suggests that conditions other than viral/bacterial stimuli could lead to an elevation in proinflammatory cytokines that could cross the placenta leading to prenatal immune activation increasing the risk of schizophrenia.
One of the earliest epidemiological studies linking infection with schizophrenia was in 1989, which was based on the data of subjects born during the epidemic period (1918 and 1957 influenza epidemic) reporting schizophrenia. A significant risk (1.4 risk ratio) of schizophrenia was observed in the subjects whose mothers were exposed to infection during 5–7 months of pregnancy compared to other months of pregnancy in the Scottish population. The same was not observed if all United Kingdom subjects were considered
[28]. Similarly, another study conducted on patients diagnosed with schizophrenia showed that the risk of schizophrenia greatly increased in offspring with elevated maternal antibodies against type 2 HSV infection during pregnancy
[16]. Despite the small sample size (
N = 27/50 case/controls), it was reported that elevated maternal serum TNF-α during pregnancy had an association with the risk (8.5 odds ratio (OR)) of schizophrenia later in their offspring
[29]. Animal models show that neural progenitor cells (NPCs) (NPCs undergo multiple changes to develop into neurons) expressed TLR3 and IRF3, and these cells responded to both in vitro and in vivo (dams) stimulation with poly I:C. Although conflicting results were observed from both models, in vivo results showed a significant decrease in cell proliferation post-treatment
[30]. Similarly, irregular differentiation (high cholinergic neuronal differentiation and undifferentiated progenitors) and neurodevelopment were observed when forebrain from rat embryos were cultured with microglial conditioned medium stimulated by LPS
[31]. Such evidence points to the detrimental neurodevelopmental effects on the foetus during pregnancy postexposure to infections or stimuli (TLR ligands), which could later affect the offspring.
For a long time, TLRs have been considered a target for vaccine adjuvants to increase the efficacy of the vaccine, where they act by low-level stimulation of the target TLRs and activate the interferon pathway
[32]. It was observed in animals that vaccination for influenza promoted better neural development (indicated by increased cell proliferation, neural differentiation, and expression of brain-derived neurotrophic factor (BDNF) in pups) and provided protective function in the mother and offspring through TLR4 stimulation
[33]. The aforementioned evidence indicates that the extent of activation of TLRs is determined by the severity of the infection in mothers, further affecting the neurodevelopment in the offspring while appropriate care such as vaccination, where the virus cannot replicate, can promote protective effects through TLRs on the offspring.
Animal models show a more detailed picture of MIA
(Table 2). Intraperitoneal administration of Poly I:C and LPS in dams led to upregulation of TLR3 and 4 mRNA expression in the hippocampus tissue of the offspring. Moreover, this upregulation was associated with sociocognitive deficits and behaviour abnormalities in them
[34]. In another study, post-LPS treatment in dams, mice showed upregulation of TLR2 and 4 mRNA in addition to COX-2 (cyclooxygenase), iNOS (inducible nitric oxide synthase), and CCL2 (chemokine C-C motif ligand) in the foetal stage, and later in adolescence. This was suggested to affect the microglial and astrocyte development in the amygdala region of the brain
[35].
Table 2. Characteristics of animal studies.
Study |
Animal (Type of Cells/Culture) |
Model |
TLR Analysed |
Analysis Results |
Talukdar et al., 2021 [34] |
Rats |
MIA (Poly IC and LPS) |
TLR2, 3, and 4 mRNA |
Elevated TLR genes were highly associated with a decrease in sociability and normal behaviour in the offspring. |
Lee et al., 2016 [36] |
Mice |
MIA (Zymosan) |
TLR2 |
TLR2 stimulation showed decreased cognitive and behavioural changes only in female offspring. |
O’Loughlin et al., 2017 [35] |
Mice |
MIA (LPS) |
TLR2 and 4 mRNA |
Elevated TLR genes in pre/postnatal stages after stimulation. These were associated with morphological changes in microglia, suggesting activation. |
Connolly et al., 2021 [37] |
Mice |
TLR 4 antagonism |
TLR4 |
TLR4 antagonism showed beneficial cognitive effects (enhanced spatial learning and memory) only in young female mice. |
Drouin-Ouellet et al., 2012 [38] |
Mice |
MyD88 KO |
TLR |
The absence of MyD88 in mice led to spatial learning and working memory deficits, and reduced motor activity. |
Ibi et al., 2009 [39] |
Mice |
Poly IC stimulation |
TLR3 |
TLR3 stimulation showed decreased social interaction, cognitive, and behavioural changes. |
Liu et al., 2013 [40] |
Mice |
TLR7 KO |
TLR7 |
TLR 7 activation reduced dendrite growth mediated by IL-6 in mice. TLR 7 KO reduced dendrite growth and reduced behavioural activity. |
Madar et al., 2015 |
Mice |
TLR2 KO and TLR2+ |
TLR2 |
TLR2 KO decreased mice activity and impaired spatial learning. Neonatal TLR2/1 activation led to impaired spatial learning in adulthood. |
Park et al., 2015 |
Mice |
TLR2 KO |
TLR2 |
TLR2 KO mice showed cognitive and social recognition impairment. |
Melnik et al., 2012 |
Mice |
In utero (poly IC stimulation) |
TLR3 mRNA |
TLR3 stimulation leads to decreased proliferation after Poly IC treatment unlike, in vitro studies. |
Mice (DG NPC) |
Poly IC treatment |
TLR3 mRNA |
TLR3 stimulation leads to increased proliferation after 24 and 48 h of poly IC treatment. |
Liu et al., 2013 [40] |
Mice (cortical neurons) |
|
TLR 7 |
TLR7 activation produced cytokines and hindered dendrite growth. TLR7 KO promoted axon and dendrite growth. |
Nakajima et al., 2014 [41] |
Mice (astrocyte cell culture) |
LPS stimulated |
TLR3 and 4 mRNA |
LPS stimulation showed significantly elevated TLR3 expression up to 24 h and dropped until 72 h. No significant change was seen with TLR4 expression, but a trend of upregulation was observed from 3 to 72 h. |
Zhou et al., 2020 [42] |
Mice (BV2 microglial cells) |
LPS treatment |
TLR4 protein |
TLR4 and NF-kB upregulation post-treatment. |
3. Factors Influencing Toll-like Receptors in Schizophrenia
3.1. Diet and Gut Microbiome
There are many diseases and disorders that are associated with improper diet. Studies showed that prenatal malnutrition during famine throughout the world in various eras increased the risk of nervous system-related disorders and mental illnesses such as schizophrenia, antisocial behaviours, other psychiatric disorders, and congenital anomalies such as neural tube defects
[51]. Although it is unclear if this was due to the direct result of malnutrition (diet change) or other underlying causes such as stress, it was observed in Dutch and Chinese populations who experienced the ‘Dutch hunger of 1944–1945’ and ‘Great leap forward of 1958–1962’, respectively, that the risk of acquiring mental illness, especially schizophrenia, was two-fold higher in the populations born during these great famines
[51][52].
Similar to how dietary deficits can be dangerous to mental health, an imbalanced diet lifestyle can affect the same. The nutritional psychiatry division of Harvard reports that the type of food can affect the inflammatory status and oxidative stress in the body. It was observed that the traditional diet from Asian and Mediterranean areas proved to have a 25–30% lower risk of depression when compared to the western diet
[53]. Moreover, studies show that saturated fatty acids (SFAs) are capable of activating microglia and inducing proinflammatory cytokines, which is facilitated by TLR2 and 4
[54]. The mRNA levels of
il6,
il1b, and
tnfa were found to be elevated post-SFA treatment in mouse cell culture, and their upregulated expression was similar to that of LPS stimulation
[55]. Furthermore, in in vitro studies (mouse and human), TLR2- and 4-dependent activations were observed with SFAs. TLR2, 3, 5, and 9 activations by their respective agonists were attenuated by ω-3- and ω-6-polyunsaturated fatty acids (PUFAs), indicating that some fatty acids activate TLRs, and some others inhibit their activity
[54][56][57].
A high-fat diet is closely related to obesity
[58], and it was observed that a high-fat diet elevated proinflammatory cytokine expression via the TLR pathway (NF-kB activation), causing an inflammatory state which correlated with weight gain
[59][60]. This high-fat diet can also dysregulate the gut microbiome
[61], due to bacterial metabolites migrating to the brain and crossing the BBB
[62], and inducing inflammatory responses in the brain. This further affects neurodevelopment (discussed in detail elsewhere
[63][64]). The “leaky gut” MIA model suggests that metabolites from maternal microbiome dysbiosis can act as inflammatory stimuli on the foetus
[64] and, in turn, potentially increase the risk of psychiatric disorders such as schizophrenia.
Evidence also suggests that the presence of unwanted or harmful microbes such as
Saccharomyces cerevisiae (induced gastrointestinal inflammation) is associated with schizophrenia in both FEP and chronic patients
[64][65]. Furthermore, schizophrenia was found to be associated with gastrointestinal dysfunction symptoms. The prevalence rate of such symptoms and conditions such as irritable bowel syndrome is higher (~19%) in these patients over healthy individuals
[66]. This suggests that the microbiome and diet might have a role in the immunopathology of schizophrenia.
3.2. Drugs
Antipsychotic drugs or neuroleptics have both intended (therapeutic) and unintended (adverse/side effects) effects. For instance, the antipsychotic drug clozapine’s main target is dopamine and serotonin receptors, and this interaction leads to the alleviation of positive and negative symptoms in many patients. In addition, it acts partially on muscarinic, adrenergic, and histamine receptors, which leads to side effects such as drowsiness and dizziness
[67]. Similarly, there is a vast literature that implicates antipsychotic medication as having another function, which is to alter the cytokine network in the schizophrenia population
[68][69][70][71]. It is not yet understood how this medication alters cytokine levels, but researchers can hypothesise two possibilities: Firstly, they could affect the cytokine levels directly by unknown means, or they alter the TLR expression/protein levels and regulators in their downstream signalling, further influencing the cytokine expression. It could possibly be the latter, as TLRs regulate cytokine expression. In congruence with this statement, a study showed that paliperidone
[72] and clozapine
[73] reduced the inflammatory response (LPS-induced) associated with TLR4 by partly inhibiting the calcium/calmodulin-dependent Akt (protein kinase B) in a rat microglial cell line. Inhibition of this enzyme prevents Akt-dependent IKKα phosphorylation, further preventing NF-kB nuclear translocation
[73]. Moreover, the MIA model of mice using poly I:C reported that paliperidone prevented the upregulation of TLR3 protein in the frontal cortex region
[74]. On the contrary, another antipsychotic treatment, olanzapine, resulted in a significant rapid elevation of inflammatory cytokines (IL-6, TNF-α, IL-1β) and TLR4 protein expression in rat PFC
[75]. A study on unmedicated patients reported significantly higher expression (MFI) of TLR2 in monocytes post-antipsychotics treatment, whereas TLR4 and 5 MFI were significantly higher before medication, and no difference after
[76]. These studies indicate that drugs alter TLR downstream signalling, which leads to cytokine imbalance. On the other hand, studies reported that antipsychotic-treated patients showed higher expression of
tlr4 mRNA
[9][77] and
myd88 mRNA
[9] in peripheral blood and postmortem brain tissue samples, respectively, over antipsychotic-free patients (schizophrenia). Interestingly, no significant differences were observed in the other downstream signalling protein mRNA levels such as
nf-kb,
ikbα,
il-1β, and
il-6 [9]. On the other hand, 3 months of antipsychotic treatment resulted in higher TLR6 protein expression, and a trend towards higher TLR8 and 9 in monocytes, T cells, and B cells of FEP patients
[6]. Given the evidence, there is a disparity in the current literature regarding the influence and relationship between antipsychotic drugs and TLRs and cytokines which is not yet clearly understood, and studies are ongoing to find the same.
Studies have reported that recreational drugs such as alcohol consumption (binge) and frequent or continuous use of amphetamines and methamphetamines induces psychosis-like symptoms in nearly half the users
[78][79]. It was found that methamphetamine elevated IL-1β and IL-18 in primary mouse astrocyte cultures, and this was mediated by TLR4, NF-kB, and caspase-11 pathways, causing neuroinflammation
[80]. This is the most probable reason why most of these users are at significant risk of schizophrenia where some of them develop psychosis, which eventually transforms into schizophrenia
[81].
3.3. Genetics
Genetic polymorphisms represent DNA variation in a specific population that may/may not lead to a change in protein expression. One of the commonly occurring polymorphisms is a single nucleotide polymorphism (SNP), which has been used as a genetic target to study multiple diseases
[82]. SNPs within multiple immune-related gene networks have now been investigated in terms of their association with disease risk. Among these, it was found that polymorphisms within genes encoding
tlrs and their pathways were associated with either increased or decreased protection against various diseases and infections. Moreover, altered expression of cytokines in response to infections was associated with
tlr gene polymorphisms
[83].
In schizophrenia, gene polymorphisms associated with pathways of inflammation, immune response, neurodevelopment, and cell death have been identified. Alterations of such pathways due to polymorphisms are significantly associated with schizophrenia risk, and the TLR pathway is suggested to be a modulator of these pathways
[84]. Studies also show that TLR polymorphisms are associated with schizophrenia (
Table 3), such as
tlr2 polymorphism in the Tunisian population
[85]. An insertion/deletion polymorphism of an SNP (rs111200466) located in the promoter region of the
tlr2 gene showed to be protective in healthy females, as this polymorphism was not observed in the schizophrenia group
[85]. Missense polymorphisms (rs5743708 and rs121917864) located in the third exon of
tlr2 gene were considered to increase the risk and susceptibility to schizophrenia. The aforementioned three SNPs were previously associated with decreased activation and expression of TLR2 and cytokines
[85]. In another study, two SNPs, rs3804099 and rs3804100 (of unknown function), located in the third exon of
tlr2 gene, were reported to be significantly associated with poor concentration in the Korean schizophrenia population. Although these 2 SNPs were found to be associated with cognition, none were associated with either positive or negative symptoms
[86]. Patients belonging to north Indian ethnicity showed a similar association with
tlr2 SNP (
rs3804099)
[87]. This indicates the broad (multiethnic) implications of TLR2 polymorphism in schizophrenia. TLR4 polymorphism also showed an association with schizophrenia risk; four SNPs of TLR4 were investigated, of which three (rs11536889-3′UTR, rs1927911-intron, and rs1927914-5′UTR) showed an association with schizophrenia. One of these investigated SNPs was previously associated with multiple diseases and disorders
[88]. Changes in the TLR downstream signalling proteins such as
myd88 SNP (rs7744-3′UTR) were reported to be less prominent in patients with schizophrenia than in controls. Moreover, a model utilising 5 SNPs (3 TLR downstream signalling protein genes (
myd88,
irak1, and
nfkb1) and 2 cytokines (
il6 and
il1b)) for a five-way analysis identifies different combinations of these five genes, predicting the risk of schizophrenia in a given population. This model showed that people having a specific combination of these five genes were at higher risk (OR—6.9) of developing schizophrenia
[9]. This evidence indicates the immune–genetic predisposition in schizophrenia, and it can be understood that multiple polymorphisms of various genes associated with the TLR pathway collectively contribute towards the incidence of schizophrenia (
Table 3).
Table 3. Characteristics of genetic studies.
Study |
Population |
Gene (SNPs) |
Location of SNP |
Region of SNP |
SNP Function |
Analysis Results |
Aflouk et al., 2021 [85] |
Tunisian |
tlr2 (rs111200466) |
Chr4. −196 to −174 Insertion/Deletion |
Promoter region |
Decreased activation and expression of TLR2 and cytokines |
Protective action in females, as this polymorphism was not observed in the schizophrenia group (females). |
tlr2 (rs5743708) |
Chr4. G > A Missense variant |
Third exon (functional) |
Increases the risk and susceptibility to schizophrenia. |
tlr2 (rs121917864) |
Chr4. C > T Missense variant |
Kang et al., 2013 [86] |
Korean |
tlr2 (rs3804099) |
Chr4. T > C Synonymous Variant |
Third exon (functional) |
Unknown function |
C’ allele associated with cognition (poor concentration). |
tlr2 (rs3804100) |
Chr4. T > A/T > C Missense variant |
Sharma et al., 2022 [87] |
Indian |
tlr2 (rs3804099) |
Chr4. T > C Synonymous Variant |
Third exon (functional) |
Unknown function |
Increases the risk of schizophrenia. |
Mostafa et al., 2022 [88] |
Egyptian |
tlr4 (rs11536889) |
Chr9. G > A/G > C UTR Variant |
3′UTR |
Interferes with the translation process and increase the expression of TLR4 |
A strong association between these genes and schizophrenia. |
tlr4 (rs1927911) |
Chr9. A > C/G/T |
Intron |
- |
tlr4 (rs1927914) |
Chr9. G > A |
5′UTR |
Regulation of gene expression and protein levels |
García-Bueno et al., 2016 [9] |
Spanish |
myd88 (rs7744) |
Chr3. A > G/A > T Noncoding Transcript Variant |
3′UTR |
Previously associated with autoimmune or inflammatory processes |
These two polymorphisms were found to be crucial in the schizophrenia predictive model for this population. |
il6 (rs1800795) |
Chr7. C > G/C > T |
Promoter region |
Increases IL-6 levels |