Alterations in inflammasome-related immunometabolites in individuals with severe psychiatric disorders

In this exploratory, cross-sectional study, 39 individuals with severe mental disorder and 39 age and sex-matched healthy controls were included. The inclusion criteria for participants with mental disorder were ages between 16 and 47 years and being diagnosed with SSD, ASD, OCD, and/or nonsuicidal self-injury disorder (NSSID) prior to enrolment by a board-certified psychiatrist. The Clinical Global Impression – Severity scale (CGI-S) was used. The exclusion criteria included neurological autoimmune disorder and/or an ongoing infection at the time of blood sampling. Participants with mental disorders were recruited from psychiatric clinics across Örebro County and surrounding counties. Interviews with participants were held at Örebro University hospital or in the participants´ homes.

Inclusion criteria for the healthy controls were based on age and sex, matched to the participants with psychiatric disorders. Exclusion criteria included being diagnosed with any psychiatric or medical disorder, however, three of the controls had been diagnosed with a psychical condition (psoriasis, asthma, nut allergy) but presented no current symptoms. However, none of the controls received medical treatment for their diseases at the time of the study, and they were regarded as being in clinical remission. Diagnostic interviews were performed with the controls to exclude individuals with current or previous psychiatric disorder (see below). Healthy controls were recruited through flyers or word of mouth. None of the participants was related to one another.

All participants with severe psychiatric disorder were recruited between November 2016 and June 2018, and the controls were recruited between January 2017 and June 2018.

Participants had been diagnosed with a psychiatric disorder by a board-certified psychiatrist prior to enrolment. According to Swedish diagnostic routines, the diagnosis of ASD requires a confirmatory interview with a parent, in addition to a diagnostic interview and cognitive tests performed by board-certified psychiatrist and a psychologist, respectively, while the diagnostic procedure for SSD, NSSID, and OCD only require an assessment by a psychiatrist. For diagnostic validation all participants were interviewed by a psychiatrist for approximately three to four hours on one single appointment, using validated rating scales. In addition, psychiatric and medical records were available to the researchers throughout the study, whereby the prior psychiatric diagnoses were confirmed.

Psychiatric diagnoses and comorbidities were assessed in the current study using the Mini International Neuropsychiatric Interview (M.I.N.I. version 7). Severity was measured with the Clinical Global Impression Severity scale (CGI-S) and ranges from 1 to 7 (1 = not at all ill, 2 = borderline mentally ill, 3 = mildly ill, 4 = moderately ill, 5 = markedly ill, 6 = severely ill and 7 = extremely ill). The level of functioning was evaluated with the Global Assessment of Functioning (GAF) ranging from 1 to 100 with separate ratings for symptoms and disability (1 = severely impaired and 100 = extremely high functioning).

Depressed mood was assessed using the self-administered nine item Patient Health Questionnaire (PHQ-9), and signs suggestive of personality disorder traits were measured with The Personality Inventory for DSM-5, brief form (PID-5-BF). Both PHQ-9 and PID-5-BF are self-rated. Severity of SSD, ASD, and OCD, respectively, were estimated with the Clinician-Rated Dimensions of Psychosis Symptom Severity, Clinician-Rated Severity of Autism Spectrum and Social Communication Disorders, and the Global Obsessive-Compulsive Scale (NIMH-GOCS). For assessment of the NSSID, the Deliberate Self-Harm Inventory (DSHI-9) and Clinician Administered Non Suicidal Self-Injury Disorder Index (CANDI) were used. In order to estimate the participants’ perception of the extent of their disability, a self-rated severity of illness scale, Patient Global Evaluation (PGE), was administrated in addition to the self-administered, 12-item version of the WHO Disability Assessment Schedule (WHODAS 2.0). The complex method of WHODAS scoring was used to calculate a total score indicating the perceived percentage of disability.

After the psychiatric interview and assessment with rating scales, each participant was designated to a primary diagnostic group: SSD, ASD, OCD, or NSSID, and all comorbidities were documented.

The healthy controls were similarly assessed with all the above-described instruments by a registered nurse (UH), trained in the assessment methods.

Peripheral whole blood was collected between 8 and 10 a.m., following an overnight fast. A total of 86 mL was drawn in eight 10 mL ethylenediaminetetraacetic acid (EDTA) tubes, one four mL EDTA tube, and one PAXgene Blood RNA tube. The four mL EDTA tubes were immediately placed on ice after sampling. The tubes were centrifuged for 15 min at 1500 x g, and plasma was then separated from the whole blood, distributed into aliquots of 2 mL, and stored at -80 °C until further analysis. Inflammasome genes and cytokines were analyzed as previously reported [28].

A mass spectrometry panel including nine selected immunometabolites previously shown to play a role in regulation in the NLRP3 inflammasome/IL-1 family axis was established [17, 19, 20, 22‐27]. The analytical method used to determine plasma concentrations of targeted immunometabolites in all samples (n = 78) was successfully validated in terms of recovery, accuracy, and precision. Briefly, 100 µL serum was added to an Ostro 96-well plate well followed by internal standards (D3-Lactic acid, D4-3-hydroxybutyric acid, D3-Serine, D4-Succinic acid, D5-Glutamine, D7-Arginine, D5-Tryptophan) in 450 µL acetonitrile. Itaconate acid was normalized using D5-Tryptophan. The filtrate was evaporated to dryness and reconstituted in 50 µL methanol and water (1:1) followed by the addition of recovery standards. Ultra-high pressure liquid chromatography-tandem mass spectrometry measurements were performed on a Waters Acquity UHPLC TQS MSMS. Two µL of sample extract was injected onto a 1.7 μm, 2.1 mm x 100 mm Atlantis PREMIER BEH C18 AX column in combination with a guard column. The column temperature was 30 °C. The gradient elution buffers were A (10 mM ammonium formate in 100% milliQ, pH adjusted with formic acid to pH 2.9) and B (10 mM ammonium formate in 5% milliQ, 95% acetonitrile). The flow rate was 0.3 mL/min, and the total run time per sample was 8 min. The mass analysis was performed using electrospray ionization operating in negative ion mode. Source temperature was 150 °C and desolvation temperature 350 °C. The cone gas flow was 150 L/h and the desolvation gas flow was 700 L/h. A seven-point calibration curve was used for the quantification of target analytes. The concentration of each point was 0 µg/mL, 0.5 µg/mL, 1 µg/mL, 2 µg/mL, 4 µg/mL, 12 µg/mL, and 16 µg/mL, and R2 ranged between 0.9990 and 0.9999 depending on the analyte. Internal standard recoveries ranged from 88 to 107%. The LOD was defined as the average amount of trace in blank samples plus three standard deviations and ranged between 0.002 and 0.763 µg/mL. The LOQ defined as the average amount of trace in blank samples plus five standard deviations ranged between 0.002-1.21ug/mL.

Before the analysis, the data were transformed to a logarithmic scale to improve the normality, and standard scores were calculated. Mann-Whitney U test was used to compare differences between the individuals with psychiatric disorders and healthy controls. A Spearman’s rank-order correlation test was used to assess the relationship between the immunological markers and the metabolites, as well as correlations between the metabolites and disease severity (GAF-S, GAF-F, CGI, and WHODAS) and diseases specific rating scales (PHQ-9, CRS-PSS, NIMH-GOCS, DSHI-9). Logistic regression was used to control for sex, age, smoking, and BMI as confounding variables. Stratified analyses were used to test if comorbid depression, antipsychotic medication, or medication with Serotonin re-uptake inhibitors (SRI) had any impact on the results. The stratification were made by removing the patients and the matched controls with depression, comorbid depression, antipsychotic medication, or SRI medication one at a time from the analysis. A probability level of p < 0.05 was set as the threshold for statistical significance. To account for multiple comparisons, the Benjamini-Hochberg procedure with a false discovery rate of 0.05 was used [29]. Principal component analysis (PCA) was used to explore the variation and patterns in the data. Statistics were conducted using SPSS version 22, Graph Pad Prism 7, and STATA 16.

Participants with a psychiatric disorder were designated to a primary psychiatric disorder group (SSD, OCD, ASD, or NSSID) regardless of comorbidities. At inclusion, the patients were markedly to severely ill with a low level of functioning (CGI-S median 6, GAF median 42). Twenty-two had a childhood onset (before age 13) with a median age of 11 years (range 16–33). Thirty were on permanent sick leave, seven were students, and only three had a regular job. The psychiatric comorbidity was substantial; 25 patients fulfilled the criteria for major depression. The comorbidity, diagnostic assessment and medication in the cohort has been described more in detail previously [28]. A summary of the patient characteristics is shown in Table S1.

Nine immunometabolites were selected for analysis based on their reported association with inflammasome activity;of which five were considered positive regulators of inflammasome activity (serine, glutamine, succinate, citrate, and lactic acid) and four considered as negative regulators (tryptophan, arginine, itaconic acid, and 3-hydroxybutyric acid). Of these, all, except itaconic acid (0% detection rate), were successfully detected and quantified in 100% in the study participants (the absolute concentrations of the immunometabolites are shown in supplemental Table S2). To test for interrelationship between the immunometabolites, Spearman correlations were performed. The results revealed that all the metabolites moderately to strongly correlated with each other (r = 0.4–0.8), with noted exceptions in lactic acid and 3-hydroxybutyric acid. Differences between patient and controls were primarily observed in the case of lactic acid, which correlated significantly to other immunometabolites in controls (Fig. 1A), but not in patients (Fig. 1B).

To study differences between individuals with psychiatric disorders and healthy controls, all plasma levels of the selected immunometabolites were compared between the two groups. Among the positive regulators of inflammasome activity, serine (p = 0.01), glutamine (p = 0.01), and lactic acid (p < 0.01) were significantly higher in the plasma of patients compared to the controls (Fig. 2). After adjusting for BMI, age, sex, and smoking, the differences between patients and controls regarding lactic acid (p = 0.01), serine (p = 0.02), and glutamine (p = 0.02) remained significant. No differences were observed for the other positive regulators, i.e. succinate and citrate, nor for the negative regulators of inflammasome function, i.e. tryptophan, arginine, and 3-hydroxybutyric acid, respectively.

Overall functioning and disease severity were estimated by the scales GAF-S, GAF-F, CGI, and WHODAS, respectively. To investigate if metabolic perturbations were associated with any of these scales, correlation analyses were performed among the patients. However, no significant correlations were found. Correlations were also made between diagnosis-specific rating scales (PHQ-9, CRS-PSS, NIMH-GOCS, DSHI-9) and the immunometabolites. The results showed a significant positive correlation between the psychosis severity rating scale and lactic acid (p = 0.01 r = 0.41) but the result did not remain significant after adjusting for multiple comparisons.

Subgroup analyses performed in the different diagnostic groups showed that lactic acid remained significant different between patients and controls in the SSD group (p < 0.01), ASD group (p = 0.03), and NSSID group (p = 0.02) but not in the OCD group. However, no significant differences were found for serine or glutamine when diagnostic subgroups were compared to controls. PCA analysis revealed no separation of specific diagnoses regarding immunometabolites nor inflammasome parameters, including inflammasome related genes and cytokines (Fig. 3).

Stratification of patients according to comorbid depression did not affect most of the results, nor did antipsychotics or medication with SRIs, except for serine in analyses without depressed patients and for glutamine in analyses with patients without antipsychotics (Figure S1).