The results of our study show a lower expression of TLR5 and TLR8 in patients regarding healthy volunteers at baseline and after 3-month of treatment. In addition, most of patient TLRs have shown lower functionality at baseline (remarkably in monocytes) and following the medication (significantly in B cells). It is important to highlight that we have not found differences in the intra-subject analyses after 3-month of treatment.
Regarding TLR expression, the most outstanding results were the lower expression of TLR5 and TLR8 on the three PBMCs subsets in patients than in the healthy volunteers at baseline (see Fig. 1) that remained lower after 3 months of treatment (see Fig. 2). Despite TLRs are especially expressed on innate immune cells, they can be also expressed on T and B cells of the adaptive response and some authors have suggested they could have immunotherapeutic consequences (Oberg et al., 2011).
With regard to functionality, most of TLRs but TLR5 showed lower functionality (especially reduced intracellular levels of TNF-α) in patients than in healthy volunteers, mainly on monocytes at baseline and in B cells after medication (see Figs. 3, 4a-4c). Both cell types, but not T cells, are antigen-presenting cells in which the altered expression and function of TLRs could have impact in the inflammatory response of FEP.
Considering our findings, we focused in the possible pathophysiological implications of impaired expression and functionality of TLR5 and TLR8. TLR5 is expressed in lung and intestinal epithelial cells, human endometrium, bladder, granulosa cells leukocytes, adipocytes, and some cancer cells, among others. Recent researches have described that the correct expression of TLR5 during neonatal period influences long-term gut microbiota composition (Fulde et al., 2018) and the alteration of microbiota has been related with psychosis in at least a subgroup of patients (Carvalho et al., 2012; Kelly et al., 2017; Schwarz et al., 2018). Moreover, the alteration of the TLR5 expression has been described in individuals with insulin resistance, that is a very prevalent condition in subjects with schizophrenia (Perry et al., 2016; Pillinger et al., 2017) and that has been recently related to schizophrenia polygenic risk score and a good clinical response to antipsychotic drugs (Tomasik et al., 2019).
TLR8, an endosomal sensor of RNA degradation products in human phagocytes, is involved in the recognition of viral and bacterial pathogens by the modulation of the suppressive activity of regulatory T cells (Sutmuller et al., 2007). Several investigations showed that murine TLR8 plays a pivotal role in the regulation of myeloid cells and prevention of autoimmunity by controlling TLR7 expression (Demaria et al., 2010; Tran et al., 2015). In this sense, in this study we have also found an altered functionality response in B cells after 3-month of treatment (see Figs. 4a-4c). Altered B cell signaling may also increase the risk of autoimmunity (Rawlings et al., 2017), and as it is known, autoimmunity has been related with the etiology of schizophrenia (Benros et al., 2011; Jeppesen et al., 2019).
Thus, the alteration in the functioning of these two TLRs has been related with microbiota composition, insulin resistance and autoimmunity, that are abnormalities that have been previously related to psychosis.
In previous studies based on chronic schizophrenia, Chang and colleges (2011) also described a lower expression of TLR5 in patients regarding healthy volunteers. A recent study by Kozłowska et al. (2019) found that TLR1, TLR2, TLR4, TLR6, and TLR9 expression were lower and that TLR3 and TLR7 manifested higher expression in patients with schizophrenia. Regarding FEP patients, Kéri et al. (2017) indicated that treatment-naïve individuals with psychosis displayed an increased expression of TLR4 on monocytes as well as TLR 5 on monocytes and T cells compared to healthy controls. Furthermore, they found that after 2-month of treatment the initial high expression of TLR4 on monocytes was normalized and they also observed an up-regulation of TLR2 on both monocytes and T cells.
Some of these differences may be explained if we consider the design of the studies: in all of previous studies but Kéri and collages (2017) subjects were on medication, and several studies suggest that antipsychotics may influence TLR expression (Kéri et al., 2017; Benros and Mortensen, 2019). In addition, most of previous researchers were not controlled by some of the most important confounding variables, such as BMI (Chang et al., 2011), as well as cannabis (McKernan et al., 2011; Chang et al., 2011; Müller et al., 2012; Kéri et al., 2017; Kozłowska et al., 2019) and smoking (McKernan et al., 2011; Chang et al., 2011) consumption. Finally, none previous publication has taken into account the use of concomitant medication that also may alter the immune response (Hannestad et al., 2011).
Regarding the correlation between the TLR expression and functionality and the total score of clinical scales at baseline we found that those individuals with more severe symptoms measured with BPRS showed higher expression of TLR2 on monocytes, as well as higher expression of TLR9 on both monocytes and B cells (see Supplementary Figs. 1a-1c). However, we found a negative correlation between the total score of BPRS and SAPS and the production of intracellular cytokines after TLR2, TLR3 and TLR9 stimulation (see Supplementary Figs. 2a-2d).
Taking into account both previous evidence and our findings, we hypotheses that decreased TLR expression and functionality may be a consequence of an excessive cell stimulation via exogenous and/or endogenous ligands in patients in such a way that PBMCs could be exhausted and show a low level of expression and activation in the in vitro analyses (Alvarez-Rodriguez et al., 2012). Namely, if TLRs are already occupied by exogenous ligands (such as pathogens) or endogenous ligands (such as autoimmunity), they could not be occupied at the same time by in vitro substrates and therefore they would show lower level of both TLR expression and production of intracellular cytokines after TLR stimulation in the in vitro analyses, although in fact PBMCs would be being overstimulated in the context of a proinflammatory response that has been previously reported in psychosis individuals (Kinney et al., 2010; Miller et al., 2011). This hypothesis would fit with previous studies that has described an increase of serum concentrations instead of intracellular concentrations of cytokines in patients regarding healthy volunteers (Capuzzi et al., 2017; Juncal-Ruiz et al., 2018).
These findings may be consistent with the neurodevelopmental model of schizophrenia, in which a prenatal infection or inflammation may produce change in both brain areas and immune system that increase the future risk to develop both psychosis disorders and immune disturbances, as autoimmunity (Murray and Lewis, 1987; Kumar, 2019; Benros and Mortensen, 2019). In fact, it has been proposed that certain prenatal infections, such as influenza, may increase the risk of adult psychotic illness through induction of autoimmunity (Khandaker et al., 2013; Jeppesen et al., 2019). In this line, in a study conducted by Benros (2011), autoimmune disease and the number of infections requiring hospitalization were described as risk factors for schizophrenia.
This study has several limitations that should be considered: first, stress level, which is known to influence inflammatory status, was not measured; second, we did not consider the consumption of classical analgesic or anti-inflammatory drugs as possible confounding factors (however, it is important to remark that the subjects from this study are young, healthy individuals without chronic diseases or chronic treatments) and third, analysis was performed on peripheral blood. We assume that these findings are an indicator of what occurs in the CNS and therefore, we would expect to find a greater disturbance of TLR expression and intracellular cytokines production in the CNS.