Autoantibodies against acetylcholine receptors are increased in archived serum samples from patients with schizophrenia

Previous studies have demonstrated that the levels of IgG against neurotransmitter receptors are increased in patients with schizophrenia. Genome-wide association (GWA) studies of schizophrenia confirmed that 108 loci harbouring over 300 genes were associated with schizophrenia. Although the functional implications of genetic variants are unclear, theoretical functional alterations of these genes could be replicated by the presence of autoantibodies. This study examined the levels of plasma IgG antibodies against four neurotransmitter receptors, CHRM4, GRM3, CHRNA4 and CHRNA5, using an in-house ELISA in 247 patients with schizophrenia and 344 non-psychiatric controls. Four peptides were designed based on in silico analysis with computational prediction of HLA-DRB1 restricted and B-cell epitopes. The relationship between plasma IgG levels and psychiatric symptoms, as defined by the Operational Criteria Checklist for Psychotic Illness and Affective Illness (OPCRIT), were examined. The results showed that the levels of plasma IgG against peptides derived from CHRM4 and CHRNA4 were significantly increased in patients with schizophrenia compared with control subjects, but there was no significant association of plasma IgG levels with any symptom domain or any specific symptoms. These pre-liminary results suggest that CHRM4 and CHRNA4 may be novel targets for autoantibody responses in schizophrenia, although the pathogenic relationship between increased serum autoantibody levels and schizophrenia symptoms remains unclear.


Introduction
Schizophrenia is a psychiatric condition that is characterised by the presence, in an individual, of hallucinations, delusions and a range of negative symptoms, including depressive symptoms, as well as cognitive deficits (Tandon et al., 2013).The pathological mechanisms of schizophrenia have yet to be conclusively elucidated but alterations in neurotransmission and the structure of the brain have been consistently observed in patients with schizophrenia and, therefore a resultant disruption of cortical connectivity may underlie some of the pathomechanisms of the condition (Baker et al., 2018;Remington, 2008;Shepherd et al., 2012).
All antipsychotic drugs licensed for the treatment of schizophrenia block dopamine D 2 receptors and hyperactivity of the dopaminergic system has been one of the most robust pathophysiological findings in schizophrenia (Abi-Dargham et al., 1998;Bruijnzeel et al., 2014).Functional alterations of some neurotransmitter receptors have been well documented in patients with schizophrenia, including serotonin receptors, glutamate receptors and acetylcholine receptors, with each spawning their own respective hypotheses of the development of schizophrenia (Eggers, 2013;Hu et al., 2015;Raedler et al., 2007).It has been hypothesised that perturbations of specific neurotransmitter systems may result in distinct symptomatic profiles in schizophrenia, with decreases in serotonin 5-HT receptors being implicated in the development of negative symptoms and the impaired function of nicotinic acetylcholine receptors (nAChr) being implicated in the development of delusional thinking (Carson et al., 2008;Caton et al., 2020;Newman-Tancredi et al., 1996).
A link between the immune system and schizophrenia was identified in the earliest characterisations of the condition and there now exists a wealth of literature documenting immune-system alterations in schizophrenia (Noll, 2007).In the central nervous system (CNS), activated microglia are more abundant in the brains of patients with schizophrenia (Bloomfield et al., 2015) and are able to eliminate synapses in a complement-dependent manner (Mastellos, 2014), a physiological process that may be more active in patients with schizophrenia (Sekar et al., 2016;Sellgren et al., 2019).Dysfunction of the immune system has often been observed in patients with schizophrenia and some evidence suggests that a pro-inflammatory immune profile is associated with negative, particularly depressive, symptoms in schizophrenia (Fernandez-Egea et al., 2016;Goldsmith et al., 2019).
The discovery that IgG directed against the n-methyl-D-aspartate (NMDA) receptor induces a type of encephalitis, with symptoms that overlap significantly with schizophrenia, suggests that antibodies directed against neurotransmitter receptors can induce at least some of the psychotic symptoms found in schizophrenia (Dalmau et al., 2011).Subsequently, other studies demonstrated increased levels of IgG antibodies against other receptors in patients with schizophrenia including, but are not limited to, the NR1 subunit of the NMDA receptor, the M1 muscarinic acetylcholine receptor (mAChr), the α7 nicotinic acetylcholine receptor and the gamma-aminobutyric acid (GABA) receptor (Chandley et al., 2009;Haussleiter et al., 2017;Jones et al., 2014;Lennox et al., 2017).The impact of these antibodies on the development of schizophrenia is largely unknown, though studies of anti-NMDAR encephalitis suggest that IgG directed against the NMDA receptor results in the internalisation of the receptor, disruption of glutamatergic signalling in the brain and the subsequent development of the pathology (Dalmau et al., 2011;Masdeu et al., 2016).
A recent large genome-wide association (GWA) study performed by the Schizophrenia Working Group of the Psychiatric Genomics Consortium confirmed that 108 genetic loci were associated with schizophrenia, with these loci harbouring >300 genes (Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014).While the functional implications of these variants have yet to be fully addressed, antibodies have previously been used experimentally to knockout gene products in functional studies (Marschall et al., 2014) and therefore it is possible that speculative functional alterations of a gene could be replicated by the presence of an antibody against the protein of interest.Consequently, schizophrenia-associated genes identified by the GWA study (Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014) have proven promising candidates for autoantibody discovery in schizophrenia in previous studies (McLean et al., 2023;Whelan et al., 2018).Accordingly, this study applied an enzyme-linked immunosorbent assay (ELISA) developed in-house to measure the levels of IgG against neurotransmitter receptors in plasma from patients with schizophrenia and non-psychiatric controls.Additionally, associations between plasma levels of IgG against neurotransmitter receptors and the presence of individual, biologically relevant schizophrenia symptoms in the Operational Criteria Checklist for Psychotic Illness and Affective Illness (OPCRIT), was also examined.

Patient samples
The samples used for this study were collected through the University of Aberdeen and NHS Grampian in the period between 2002 and 2008 and stored at the Grampian Biorepository.This study used plasma from 247 patients with schizophrenia (mean aged 40 ± 13 years) and 344 non-psychiatric controls (aged 44 ± 12 years).The case subjects were identified through psychiatric hospitals or outpatient facilities, and the control subjects, recruited from local communities, were screened for psychiatric disorders or a family history of mental illness by a selfreporting questionnaire.All schizophrenia samples met the DSM-IV criteria for schizophrenia diagnosis at the time of sample collection and all participants ethnically self-identified as British Caucasian.All participants gave written consent for taking their blood samples for genetic and serological analyses.The study was approved by local ethics committees (13/13/NS/0125) and conformed to the Declaration of Helsinki and its amendments.
The available demographic data for the samples used in this study are listed in supplementary Table 1 showing that the case group was more female and older than the control group (p < 0.05).Information on the medication status was not available for all participants in this study but the available medication information is given in supplementary Table 2. Similarly, the patient database contains symptom information in the form of the Operational Criteria Checklist for Psychotic Illness and Affective Illness (OPCRIT), but a full checklist of symptom information was not available for all patients (McGuffin, 1991).
Plasma samples were separated from whole blood by centrifugation using a routine protocol and then aliquoted before storage at − 80 • C until use.Identical storage conditions, and parity of freeze-thaw cycles were maintained between case and control group samples throughout the study to minimise the impact of repeated freeze-thaw cycles on any differences between the groups.Levels of total IgG in the case and control group have been measured in our previous publication, demonstrating no significant difference between the two groups (Whelan et al., 2018).

Peptide design
Four neurotransmitter receptors whose genes were shown to be strongly associated with schizophrenia by the Schizophrenia Working Group of the Psychiatric Genomics Consortium (Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014), were selected as targets in this study and are listed in supplementary Table 3.The peptide antigens used for antibody testing were designed in silico based on the protein sequences retrieved from the NCBI protein database (https://www.ncbi.nlm.nih.gov/).The target peptides were analysed through computational prediction for HLA class II (HLA-II) restricted and B-cell epitopes (http://www.iedb.org).Twelve HLA-DRB1 alleles that represent the most common alleles in a European-derived Caucasian population were used for the HLA-II epitope prediction ((Schizophrenia Working Group of the Psychiatric Genomics Consortium, 2014)).A peptide derived from a maize protein was used as the control antigen to control for non-specific binding signal.In the absence of a native cysteine residue, to link the peptides to maleimide-activated plates (Corning® 96 Well Sulfhydryl-BIND™ 96 Microplate, Corning, New York, USA), a terminal cysteine residue was added to those peptides that do not natively contain a cysteine residue within their sequences.All peptides used were synthesised by solid-phase chemistry with purity of >95 % (Mimetopes, Mulgrave, Australia), and reconstituted in 67 % acetic acid to 5 mg/mL as stock solution.

Detection of circulating IgG against neurotransmitter receptor peptides
An in-house ELISA was developed to measure the levels of plasma IgG against neurotransmitter receptor-derived peptides, as described in our previous study (McLean et al., 2017).Briefly, 5 mg/mL of stock peptide solution was diluted into 10 μg/mL working solution to coat the 96-well microplates with 100 μL/well.The antigen-coated plates were incubated for 1.5 h at room temperature.After washing 3 times, unoccupied sulfhydryl groups were blocked using 10 mg/mL of freshly reconstituted L-cysteine (VWR, Radnor, Pennsylvania, USA), and the plates were then dried and stored at 4 • C until use within six months.Following washing in duplicate, plasma samples were diluted 1:150 in assay buffer and 50 μL was added to each well, followed by incubation at room temperature for 1.5 h.The plates were then washed in triplicate and 50 μL of the goat anti-human IgG-HRP (Abcam, Cambridge, UK), diluted 1:50,000 in assay buffer, was added to each well before incubation at room temperature for 1 h.After an additional triplicate wash, IgG levels were visualised by colour development using 50 μL stabilised TMB solution (ThermoFisher Scientific, Waltham, Massachusetts, USA) with incubation at room temperature for 30 min, followed by the addition of 25 μL STOP solution (ThermoFisher Scientific, Waltham, Massachusetts, USA).The optical density (OD), proportional to the level of IgG present, was measured using the VarioSkan Lux plate reader (ThermoFisher Scientific, Waltham, Massachusetts, USA) at 450 nm, with 620 nm as a reference wavelength.
To reduce the interference on the detected signal from sample background noise due to nonspecific IgG binding to the 96-well microplates, a specific binding index (SBI) was introduced to express the relative levels of circulating IgG antibodies against the target peptides when compared to the control antigen; a peptide derived from maize.The SBI is calculated as follows: SBI = [OD antigen − OD NC ]/[OD maize − OD NC ].
An inter-assay deviation was estimated using quality control (QC) samples, pooled from 150 healthy control samples, and tested on every 96-well plate.The SBI of the QC samples was expressed as a coefficient of variation (CV%) to represent the reproducibility of the in-house ELISA.

Data analysis
Associations between sex and age, and levels of IgG against the target peptides were examined using a linear regression model.Levels of IgG against Drug and alcohol use was recorded for patients with schizophrenia only and thus plasma levels of IgG against the target peptides were examined by excessive drug and alcohol use in the year prior to sample collection in the patient group but not the control group.Due to a skewed distribution (supplementary Table 4), the differences in the levels of plasma IgG against the target peptide antigens between case and control samples were examined using the Mann-Whitney U test.Due to multiple testing of this hypothesis, the resulting p-values were corrected for using the Bonferroni method.
Available symptom data from schizophrenia patients in this study cohort were used in the form of Operational Criteria Checklist for Psychotic Illness and Affective Illness (OPCRIT) categories to examine symptom association with plasma IgG levels by the presence or absence of depressive and delusional symptoms.Individual symptoms were classified under these headings and the Mann-Whitney U Test was applied to compare plasma IgG levels between cases who experienced the specific symptom and those who did not.In order to examine the overall association between plasma IgG levels by depressive and delusional symptoms, the resulting p-values from each Mann-Whitney U test were used to calculate a combined p-value using Fisher's Combined Probability and the combined p values were corrected for multiple antigens tested using the Bonferroni method.For individual symptom associations, the resulting p values were corrected for the number of symptoms tested using the Bonferroni method.
In order to examine the impact of antipsychotic medication on the IgG levels, available information on medication status and dose was converted into Chlorpromazine (CPZ) Equivalent Daily Doses, by an experienced clinical pharmacist according to Bazire's Psychotropic Drug Directory (Bazire, 2021).A linear regression model was then used to examine the impact of CPZ equivalent doses on plasma levels of individual IgG against each peptide, in which IgG levels act as the dependent variable and the CPZ equivalent dose as the independent variable.

Circulating levels of IgG against neurotransmitter receptor peptides
An in-house ELISA was used to examine the levels of plasma IgG against neurotransmitter receptor-derived peptides in non-psychiatric controls and individuals with schizophrenia.All assays in this study had an inter-assay deviation of CV <15 % (supplementary Table 5).
As shown in Table 1, the median levels of plasma IgG were significantly higher in patients with schizophrenia than control subjects in both the CHRM4 IgG assay (Z = − 4.608, p < 0.001) and the CHRNA4 IgG assay (Z = − 7.207, p < 0.001).The median level of plasma IgG against the GRM3-derived peptide was significantly lower in the case group than the control group (Z = − 2.945, p = 0.003), while there was no significant change in the median levels of plasma IgG against CHRNA5 (Z = − 1.776, p = 0.076) (Table 1).
Since both age and sex were significantly different between the case group and the control group (supplementary Table 1), a linear regression model was used to examine the impact of age and sex on the levels of plasma IgG against the peptides of interest (supplementary Table 6).Sex did not significantly affect the levels of plasma IgG against any of target peptides used (p > 0.05), but age was significantly associated with the levels of plasma IgG against CHRM4 (Standardised β = − 0.089, p = 0.036), therefore values of the anti-CHRM4 IgG assay reported in this study (Table 1) have been corrected for age.Linear regression analysis demonstrated that the levels of plasma IgG against target peptides were not significantly associated with CPZ equivalent dose (supplementary Table 7).Plasma IgG levels were not significantly different between patients who had a history of excessive drug or alcohol use and those who did not have a history of excessive drug or alcohol use (supplementary Table 8).

Association of circulating IgG levels with the presence of depressive symptoms
Due to noted associations between immune-system alterations and depressive symptoms, both generally and within schizophrenia, levels of plasma IgG against neurotransmitter receptors in patients with schizophrenia were examined by the presence and absence of depressive symptoms (Table 2).The median levels of plasma IgG against the target peptides were not significantly associated with the presence of depressive symptoms (p > 0.012) (Table 2), nor were levels of plasma IgG significantly different between patients with a specific depressive symptom when compared to patients who did not manifest that specific symptom (p > 0.007) (Supplementary Table 9).

Association of circulating IgG levels with the presence of delusional symptoms
The levels of plasma IgG against neurotransmitter receptor peptides in patients with schizophrenia were examined by the presence of delusional symptoms (Table 3).Levels of IgG against all peptides examined were not significantly associated with delusional symptoms (p > 0.003).Similarly levels of IgG against neurotransmitter derived peptides were not significantly increased in patients who displayed a specific delusional symptom when compared to those who did not (p > 0.013) (Supplementary Table 10).To reduce the possibility of type-I error, the cut-off for statistical significance was set at p < 0.013.Levels of plasma IgG against CHRM4 and CHRNA4 were significantly higher in patients with schizophrenia than controls (p < 0.013).Levels of plasma IgG against GRM3 were significantly lower in patients with schizophrenia than controls (p = 0.003).Levels of plasma IgG against CHRNA5 were not significantly different between case and control samples (p > 0.013).SBI = Specific Binding Index.a Levels of plasma IgG against CHRM4 have been corrected for a significant age effect (supplementary Table 6).

Discussion
This study examined the levels of plasma IgG against 4 neurotransmitter receptor-derived peptides in patients with schizophrenia and nonpsychiatric controls and the results demonstrated that there was an increase in the levels of plasma IgG against CHRM4 and CHRNA4 derived peptides (Table 1).The levels of plasma IgG against the 4 neurotransmitter-receptor-derived peptides were not altered in patients manifesting any of the depressive symptoms, in contrast to previous studies suggesting that pro-inflammatory markers were associated with negative symptoms, though this is consistent with our previous study that suggested levels of IgG against voltage gated calcium channels were also not associated with negative symptoms (Fernandez-Egea et al., 2016;Goldsmith et al., 2019;McLean et al., 2023) (Table 2).Based on the present results, therefore, autoantibodies involved in schizophrenia may function independently of pro-inflammatory processes; in support of this hypothesis, a study of anti-NMDAR IgG encephalitis did not find elevated clinical measures of neuroinflammation in a Brazilian population, and so the nature of neuroinflammation in the context of autoantibodies in schizophrenia requires further investigation (DiSabato et al., 2016;Nóbrega et al., 2019).
Due to their widespread expression in the frontal cortex, striatum and hippocampus, and their ability to regulate dopaminergic signalling, mAChrs have been the focus of intense research into schizophrenia (Jeon et al., 2010).A recent genetic study confirmed that a single nucleotide polymorphism (SNP), rs72910092, proximal to CHRM4, was associated with schizophrenia, although the functional effect of this SNP on M 4 mAChr is unknown (Pozhidaev et al., 2020).A decrease in M 4 mAChr levels has been observed in the hippocampus of patients with schizophrenia (Scarr et al., 2007).Previous studies of anti-mAChr antibodies in schizophrenia demonstrated increased levels of IgG directed against the M 1 mAChr (Ryan et al., 2019).To our knowledge, the present study is the first to identify an association between schizophrenia and antibodies against the M 4 mAChr (Table 1), although replication is required to confirm this initial finding and to examine the precise relationship between the antibodies against mAChrs and schizophrenia.
Despite a common ligand, nicotinic and muscarinic acetylcholine receptors differ in both structure and function.Unlike mAChrs, nAChrs form homomeric or heteromeric dimers, trimers and pentamers from twelve distinct subunits, resulting in incredible structural and functional diversity depending on the subunit composition of the receptor (Albuquerque et al., 2009).The α7 nAChr is most abundant in the CNS, but levels of circulating IgG against the α7 subunit have shown inconsistent results in patients with schizophrenia (Chandley et al., 2009;Hoffmann et al., 2018).In this current study, based on a selection of genetic loci associated with schizophrenia, we examined plasma IgG against peptides derived from the α4 and α5 subunits of nAChr and found that the levels of plasma IgG against CHRNA4 but not CHRNA5, were increased in patients with schizophrenia compared to control subjects.
A GCT insertion mutation, 776ins3, in the CHRNA4 gene has been found to be associated not only with autosomal dominant nocturnal frontal lobe epilepsy but also with a schizophrenia-like psychiatric illness (Magnusson et al., 2003).Apart from the presence of epilepsy, which is not recorded in OPCRIT data, the 776ins3 mutation does not appear to result in a symptom-specific subgroup of schizophrenia and therefore a symptomatic link between the effects of the 776ins3 mutation and increased levels of plasma IgG against CHRNA4-derived peptides was unable to be examined in this study.Cholinergic neurotransmission may play a role in the development of delusional thinking (Caton et al., 2020), though this present study failed to demonstrate a statistically robust association between the delusional symptoms and levels of IgG against CHRNA4-derived peptides (Table 3).Most studies of nAChr disruption in schizophrenia, whether mediated by autoantibodies or otherwise, have focussed on the α7 subunit, while this present study suggests that the α4 subunit may be a potential target of the immune system in schizophrenia.Of note, the 776ins3 mutation results in a gain of α4 subunit function; it is possible that IgG against the α4 nAChr subunit may be agonising as is observed in Grave's hyperthyroidism, though further mechanistic studies are required to examine this hypothesis (Steinlein and Bertrand, 2008;Xia and Kellems, 2011).
It is important to acknowledge the limitations of this study.Foremost, the OPCRIT system was conceived to aid in the diagnosis of psychiatric illness instead of as a scoring system for the symptoms of schizophrenia.In this present study individuals were reorganised into categories denoting the 'presence' or 'absence' of symptoms.Therefore, this study was unable to examine potential linear associations between plasma IgG levels and symptom severity or duration.Similarly, only symptoms with a pre-established connection to the dysfunction of the neurotransmitters of interest were selected for analysis, limiting the breadth of the analysis performed in the interests of reducing unnecessary multiple comparisons.While this study controls for important demographic factors such as age, sex and alcohol and drug use, a number of relevant factors, such as BMI, are not included in the data available for these samples and it is possible that these factors my influence levels of IgG against the peptides examined in this study.Finally, due to the nature of the in-house ELISA test it was not possible to establish cut-off values based on the assay of known positive and negative samples, and so the analysis was limited to examining differences in median reactivity against the peptides of interest rather than differences in frequencies of autoantibodies.Since it is likely that autoantibodies in schizophrenia are pathogenic in a minority of patients it is plausible that differing levels of total IgG may influence levels of IgG against specific proteins and therefore impact the results reported.However, in a previous study, the concentration of total IgG was examined in a subset of samples used in this present study and no significant difference in the levels of total IgG was observed between cases and controls (Whelan et. al, 2018).In samples for which total IgG data is available, Spearman's rho analysis indicted that levels of IgG against peptide antigens used in this study were not significantly correlated with total IgG concentration (Supplementary Table 11).
Furthermore, this study reports levels of IgG against selected neurotransmitter receptors in serum samples and therefore questions surrounding pathogenicity of autoantibodies in Schizophrenia are not advanced by this study and it remains unclear whether increased levels of IgG against peptide antigens correlate with increased pathogenicity.While an assay of Cerebral Spinal Fluid is often considered a more sensitive, and pathogenically relevant, marker of diagnosis in autoimmune encephalitis, the diagnostic relevance of autoantibodies in CSF The p-values of individual comparisons levels of IgG against peptides in patients with or without specific depressive symptoms (supplementary table 9) were combined using Fisher's Combined Probability in order to examine the association between IgG levels for the peptide antigens of interest and depressive symptoms.Plasma IgG levels for the target peptides were not significantly associated with depressive symptoms.The cut-off for statistical significance was p < 0.007.
can vary between different autoantibody targets (Lee and Lee, 2016).It has been suggested that serum markers of immune factors are poor correlates of CSF measurements but it is possible that there may be disease-specific correlations, particularly in autoimmune diseases, and therefore there is an urgent need for further studies examining autoantibodies in the CSF of patients with Schizophrenia (Gigase et al., 2023;Orlovska-Waast et al., 2019).
In conclusion, this study applied an in-house ELISA to examine the levels of plasma IgG against peptides derived from neurotransmitter receptors encoded by the genes associated with schizophrenia in individuals with schizophrenia and non-psychiatric controls.This approach uncovered novel autoantibodies against neurotransmitterrelated targets in schizophrenia, however no associations between levels of IgG against any of the peptides examined and depressive or delusional symptoms were observed.It is possible that PANSS, or other symptom-rating scales may be more appropriate tools for symptom analysis, nonetheless a long-standing criticism of the field of autoantibody research in schizophrenia, expressed most recently in 2018 (Pollak, 2018), is that studies in the field focus too frequently on antibody discovery and not enough on biological significance and so these associations are reported here in the spirit of this valid criticism.

Role of the funding source
The work undertaken in this manuscript was funded entirely by legacy donations from the Schizophrenia Association of Great Britain (SAGB).The SAGB is a defunct charity and no longer accepts new donations.The funding organisation played no role in the design, collection of study materials, conduct of the work nor in any subsequent decision regarding publication.

Declaration of competing interest
All the authors declare that they have no biomedical financial interests or potential conflicts of interest.

Table 3
Association of plasma IgG levels with delusional symptoms.The p-values of individual comparisons levels of IgG against peptides in patients with or without specific delusional symptoms (supplementary table 10) were combined using Fisher's Combined Probability in order to examine the association between IgG levels for the peptide antigens of interest and delusional symptoms.Plasma IgG levels for the target peptides were not significantly associated with delusional symptoms.The cut-off for statistical significance was p < 0.003.

Table 1
Levels of plasma IgG against peptides encoded by schizophrenia-associated genes.

Table 2
Association of plasma IgG levels with depressive symptoms.