Fecal Microbiota Underlying the Coexistence of Schizophrenia and Multiple Sclerosis in Chinese Patients

Both schizophrenia (SZ) and multiple sclerosis (MS) affect millions of people worldwide and impose a great burden on society. Recent studies indicated that MS elevated the risk of SZ and vice versa, whereas the underlying pathological mechanisms are still obscure. Considering that fecal microbiota played a vital role in regulating brain functions, the fecal microbiota and serum cytokines from 90 SZ patients and 71 age-, gender-, and BMI-matched cognitively normal subjects (referred as SZC), 22 MS patients and 33 age-, gender-, and BMI-matched healthy subjects (referred as MSC) were analyzed. We found that both diseases demonstrated similar microbial diversity and shared three differential genera, including the down-regulated Faecalibacterium, Roseburia, and the up-regulated Streptococcus. Functional analysis indicated that the three genera were involved in pathways such as “carbohydrate metabolism” and “amino acid metabolism.” Moreover, the variation patterns of serum cytokines associated with MS and SZ patients were a bit different. Among the six cytokines perturbed in both diseases, TNF-α increased, while IL-8 and MIP-1α decreased in both diseases. IL-1ra, PDGF-bb, and RANTES were downregulated in MS patients but upregulated in SZ patients. Association analyses showed that Faecalibacterium demonstrated extensive correlations with cytokines in both diseases. Most notably, Faecalibacterium correlated negatively with TNF-α. In other words, fecal microbiota such as Faecalibacterium may contribute to the coexistence of MS and SZ by regulating serum cytokines. Our study revealed the potential roles of fecal microbiota in linking MS and SZ, which paves the way for developing gut microbiota-targeted therapies that can manage two diseases with a single treat.


Introduction
Schizophrenia (SZ), characterized by positive and negative symptoms such as delusions, hallucinations, apathy, thought disorder, apathy, and cognitive and functional impairment, is a complex psychiatric neurodevelopmental disorder, afecting more than 20 million people worldwide and over 7 million people in China [1].Multiple sclerosis (MS), which is associated with myelin loss, varying degrees of axonal pathology, and progressive neurological dysfunction, is a multifactorial disease of the central nervous system (CNS), afecting nearly 2.3 million people worldwide with a prevalence of 50-300/100,000 [2].Both SZ and MS are leading causes of disability worldwide and impose heavy psychosocial burden for caregivers and high costs on society.Recently, some studies have suggested a link between MS and SZ [3].It was proposed that the immune dysregulation that afects CNS function in MS may also play a role in SZ [4].Tere was also a signifcantly elevated risk of SZ in priorrecorded MS and of subsequent MS in people with priorrecorded SZ [5].Moreover, MS and SZ patients show varying degrees of dysfunctions in immune and brain, suggesting possible shared mechanisms [6,7].However, the pathological mechanisms underlying the co-occurring of both diseases remain elusive.
Most recently, Ahangari et al. reanalyzed the largest available genome-wide association analysis (GWAS) datasets for SZ and MS and identifed the shared genetic architecture of both diseases [8].Other than the efects of genetic and antigenic exposures [9], gut microbiota have been found to play a vital role in various brain diseases such as neurodegenerative disease via the gut microbiome-brain axis [10].Te potential roles of gut microbiota in modulating the immune system and the homeostasis of the central nervous system can be mediated by multiple kinds of microbiota-related molecules, such as cytokines, metabolites, and even bacterial extracellular vesicles [11][12][13].Mounting evidences suggested that patients with MS from western developed countries and China as well as patients with various stages of SZ demonstrated gut microbial dysbiosis with both depletion and enrichment of certain bacteria as compared to controls.Jangi et al. reported an increase of Methanobrevibacter, Akkermansia, and decrease of Prevotella, Faecalibacterium in patients with MS as compared to controls [14].We found that several key functional bacteria, primarily Faecalibacterium, decreased remarkably in stable Chines MS patients [15].We also reported that Chinese elderly SZ patients were associated with decreased abundances of Faecalibacterium, Roseburia, and increased abundances of Prevotella, Akkermansia [16].Moreover, transplantation of gut microbiota from drug-free SZ patients led to SZ-like behavioral abnormalities and dysregulated kynurenine metabolism in mice [17].In addition,, both SZ and MS patients demonstrated signifcantly perturbed profles of proinfammatory and antiinfammatory cytokines and chemokines, which demonstrated signifcant correlations with certain fecal microbiota in patients [15,16].Such results indicated that fecal microbiota dysbiosis participated in the development of both diseases and may contribute to the associations between SZ and MS.
Considering the fndings that the microbiome infuences on neuroimmune interactions in neurodegenerative disease [18,19], we reanalyzed the fecal microbial sequencing data and serum cytokines from Chinese patients with SZ and MS published in our previous studies [15,16], with the aim to probe the microbial disorders that link the development of both diseases.Our study may pave the way for developing gut microbiota-targeted intervention strategies capable of treating one disease and reducing the incidence of the other disease at the same time.

Datasets. Datasets deposited in the Sequence Read
Archive with accession numbers PRJNA807473 and SRP258890 were retrieved and reanalyzed in this study [15,16].Both datasets contained raw reads of fecal microbiota targeting the V3-V4 regions of rRNA gene sequenced with Illumina MiSeq instrument [20].Dataset PRJNA807473 was obtained from 90 well-controlled Chinese elderly SZ patients and 71 age-, gender-, and BMImatched cognitively normal subjects (referred as SZC), while dataset SRP258890 was from 22 patients with MS and 33 age-, gender-, and BMI-matched healthy subjects (referred as MSC).All SZ patients were diagnosed according to the criteria of the Diagnostic and Statistical Manual of Mental Disorders Fourth Edition (DSM-IV), and they were not treated with any psychiatric drugs such as antidepressants and mood stabilizers during the last 1 month.SZ patients meeting either of the following criteria were excluded: body mass index higher than 28 kg/m 2 ; family history of dementia; any kind of other neurodegenerative diseases such as Alzheimer's disease or Parkinson's disease; any kind of mental disease such as depression; any kind of tumor; treated with psychiatric drugs such as antidepressants, mood stabilizers, and so on; known active infections such as bacterial, viral, or fungal infection; other diseases such as irritable bowel syndrome, infammatory bowel disease, or other autoimmune diseases.MS patients, diagnosed based on the 2005 McDonald criteria, were stable and not of new onset or active relapse, and they did not receive immunosuppressive medications such as steroids and beta-interferon/glatiramer acetate in the preceding 3 months.MS patients meeting either of the following criteria were excluded: less than 20 years old; body mass index higher than 30 kg/m 2 ; pregnancy; known active infections such as bacterial, viral, or fungal infections; use of antibiotics, probiotics, or prebiotics in the last month; any other diseases such as hypertension, diabetes mellitus, infammatory bowel disease, irritable bowel syndrome, and other autoimmune diseases.Te levels of cytokines quantifed with the Bio-Plex assay (Bio-Rad) in serum samples of corresponding patients (MS, SZ) and controls (MSC, SZC) were also retrieved.Detailed information about the samples and datasets was provided in our previous studies [15,16].

Study Design.
Figure 1(a) demonstrated a schema of the study design.After retrieving the sequencing reads from both datasets, the reads from SZ and MS patients as well as corresponding controls (SZC and MSC) were processed with the same pipeline and parameters and pooled together for ASV (amplicon sequence variants) clustering, taxonomy annotation, and quantifcation.Based on the abundance table of microbiota obtained from the results of ASV quantifcation and taxonomy annotation, microbiota differential between MS patients and MSC, SZ patients and SZC were retrieved, separately.At the same time, the cytokines difered between MS patients and MSC, SZ patients and SZC were also obtained, respectively.Ten, we evaluated 2 Canadian Journal of Infectious Diseases and Medical Microbiology the diferential microbiota and cytokines that were shared by both diseases.Combined with the associations between microbiota and cytokines, as well as the predicted functions of diferential microbiota, the potential roles of gut microbiota in driving the development of both diseases were estimated.

Processing of 16S rRNA Sequences.
Read sequences deposited in both datasets were processed with QIIME2 (version 2022.8).Sequence processing procedures such as quality control, clustering of amplicon sequence variants (ASVs) were performed with default parameters [21].Te ASV abundance table was constructed based on downsized sampling reads.Taxonomy of ASVs was annotated based on the SILVA database (version 138.1) [22], while functions of them were predicted with Tax4Fun2 [23,24].For alpha diversity, parameters Shannon and Simpson were considered.Potential batch efects in the two datasets were evaluated with PERMANOVA (permutational multivariate analysis of variance) and visualized with principal component analysis (PCA).

Statistical Analysis.
Te diferences in the composition of fecal microbiota at diferent taxonomic levels were analyzed using the linear discriminant analysis efect size method (LEfSe) [25].For serum cytokines and clinical indicators, normal distribution was frstly estimated using the Kolmogorov-Smirnov test.and SZ patients, and it also showed no apparent diference with regard to both indices (P > 0.05).

Microbial Features Common in MS and SZ Patients.
Te microbial features common in MS and SZ patients were conducted using the following two steps.Firstly, to preclude the potential impact of confounding factors such as sex and age, microbiota associated with MS and SZ patients were evaluated by comparing MS, MSC, and SZ, SZC, respectively.Ten, the microbial features associated with both kinds of diseases were retrieved by comparing both lists of diferential microbiota.Based on P value less than 0.05 and LDA score (linear discriminant analysis) higher than 2, three genera were found to be signifcantly associated with MS and SZ, simultaneously.Figure 2(a) illustrates the relative abundances of the three genera that were signifcantly perturbed in MS and SZ patients, while the LDA scores of them are shown in Figure 2(b).It was revealed that two genera Faecalibacterium and Roseburia decreased in MS or SZ patients as compared to corresponding controls, while Streptococcus increased in patients with both kinds of diseases.
We further retrieved the ASVs belonging to the three common genera and predicted the functions of them using Tax4Fun2 [23,24].In Figure 2(c), we can see that the most abundant function associated with the three genera was "metabolism," followed by "environmental information processing," "genetic information processing," and "cellular processes."Among the pathways associated with "metabolism," "carbohydrate metabolism," "amino acid metabolism," "metabolism of cofactors and vitamins," "nucleotide metabolism," and "energy metabolism" were the most abundant (Figure 2(d)).Moreover, pathways such as "signal transduction," "membrane transport," and "translation" were also found to be associated with the three genera.

Serum Cytokines Associated with Both Kinds of Diseases.
We also evaluated whether MS and SZ patients shared some patterns of serum cytokines.By retrieving cytokines with P values less than 0.01 and fold changes (FC) larger than 1.5, the cytokines signifcantly diferential between MS patients and MSC, SZ patients and SZC were demonstrated in Figures 3(a) and 3(b), with the relative abundances provided in Figures 4(a) and 4(b).Te Venn diagram (Figure 3(c)) illustrated that six cytokines were perturbed in both kinds of diseases, while three and eight cytokines were dysregulated specifcally in MS and SZ patients as compared to corresponding controls.Te log2 (fold change) values for the six cytokines in MS patients versus MSC (MSvsMSC) and SZ patients versus SZC (SZvsSZC) were also evaluated.As shown in Figure 3(d), IL-8 (interleukin 8) and MIP-1a (macrophage infammatory protein 1-alpha) were downregulated, while TNF-α (tumor necrosis factor-alpha) was upregulated in both kinds of disease.Moreover, IL-1ra (interleukin-1 receptor antagonist), PDGF-bb (plateletderived growth factor BB), and RANTES (regulated upon activation, normal T-cell expressed and presumably secreted) increased in SZ patients but decreased in MS patients as compared to corresponding controls.Such results indicated that although immune responses were present in both kinds of diseases, the patterns of infammatory responses and underlying mechanisms may not be the same for MS and SZ patients.

Discussion
Te coexistence of symptoms associated with MS and SZ patients indicated that these two disorders may have similar etiologies.To reveal the mechanisms underlying the coexistence of both diseases, we retrieved and reanalyzed the 16S rRNA sequencing data and serum cytokines from our 6 Canadian Journal of Infectious Diseases and Medical Microbiology previous studies.It was revealed that the genera Faecalibacterium, Roseburia, and Streptococcus demonstrated similar trends of perturbation in both diseases, while the patterns of serum cytokines were a bit diferent.We found that T17-related responses contributed to MS pathogenesis, as evidenced by the increase of IL-17 in MS patients [26].Whereas, the infammatory disorders in SZ patients may be associated with the increased expression of RANTES and MIP-1b since RANTES was thought to promote leukocyte infltration to sites of infammation and induce the activation of T-cells, followed by diverse efects including T-cell proliferation or apoptosis and the release of proinfammatory cytokines such as MIP-1b [27,28].Moreover, association analysis showed that Faecalibacterium correlated extensively with serum cytokines in both diseases.Infammation in the periphery and the CNS, reported to be connected to the pathogenesis of MS and SZ [29,30], was found to be increased in this study, as evidenced by the elevated levels of infammatory markers such as TNF-α in both kinds of diseases.On one hand, TNF-α was reported to mediate monocyte infltration into the intestinal tissues, resulting in tissue damage, disruption of epithelial barrier, and fnally leaky gut associated with MS and SZ patients [31,32].Bacterial translocation due to the increased intestinal permeability further drove proinfammatory responses in both diseases [33,34].Moreover, TNF-α was found to stimulate the hypothalamic-pituitary-adrenal (HPA) axis and increase microglial activation in the central nervous system (CNS), which would further contribute to infammatory responses.Te infammation mediators may further infuence the levels of neurotransmitters such as dopamine, noradrenaline, and serotonin and fnally induce neuronal damage [35,36].In a word, infammation was an important mediator in the pathogenesis of neurological disorder that may contribute to the coexistence for MS and SZ.
We found in our study that the abundance of Faecalibacterium negatively correlated with TNF-α and was signifcantly reduced in both diseases as compared to corresponding controls.Faecalibacterium is a kind of Grampositive bacteria showing anti-infammatory properties [37].It was found that Faecalibacterium and its supernatant can suppress the expression of TNF-α and then alleviate the infammation in MS [15].Faecalibacterium also produces short chain fatty acids (SCFAs), especially butyrate.Previous studies indicated that butyrate can cross the blood-brain barrier, help improve gut integrity, shape the immune milieu by attenuating proinfammatory cytokine expression in microglia and peripheral, and have a crucial impact on brain plasticity by stimulating the production of BDNF (brain derived neurotrophic factor) in the CNS, which is involved in the survival of neurons [38,39].Lower abundance of Faecalibacterium may lower the production of butyrate in MS and SZ patients [40], which in turn could contribute to lower BDNF levels, and fnally the development of both diseases [41,42].Besides Faecalibacterium, Roseburia was also downregulated in both diseases.Evidences have shown that Roseburia played an important role in maintaining gut health by improving the gut ecosystem and exhibiting antiinfammatory efects by upregulating genes involved in the innate immune responses such as antimicrobial peptides and Toll-like receptors [43,44].A reduction in Roseburia spp.has been found to contribute to an infammatory milieu [45].As another important butyrate-producing bacterium, the reduction of Roseburia would further decrease the butyrate level and exacerbate dysfunctions such as gut barrier impairment, peripheral, and brain neuroinfammation associated with the pathophysiology of both diseases.
On the other hand, Streptococcus was found to increase in both diseases.Some Streptococcus spp.strains were potential pathogens associated with infammation and an increased risk of a broad spectrum of neuropsychiatric conditions such as SZ [46].Streptococcus spp.strains were also found to be potent acetate and serotonin (5-HT) producers [47].Increased plasma acetate has been found to correlate with disability and immune response in multiple sclerosis [48].Although the role of 5-HT in the pathophysiology of SZ and MS is still unclear, it was reported that long-lasting 5-HT overload can disrupt neuronal activity within the cerebral cortex, anterior cingulate cortex, and dorsolateral frontal lobe.Te disturbance of the serotonergic system has been found to play an important role in the development of multiple psychiatric disorders ranging from anxiety to SZ [49].Moreover, a previous study reported that Streptococcus spp.can produce neurotoxins such as streptokinase and streptomycin, which might irreversibly damage neurons, including dopaminergic neurons relevant for the pathogenesis of Parkinson's disease [50].In other words, reduced Faecalibacterium, Roseburia, and increased Streptococcus may together contribute to the brain infammation and neuronal disorders associated with the coexistence of MS and SZ.
Based on the abovementioned information, the increase of opportunistic pathogen Streptococcus and decrease of probiotics (Faecalibacterium, Roseburia) that played a role in anti-infammation, improving gut ecosystem and barrier may contribute to the coexistence of MS and SZ via regulating serum cytokines.Supplementing Faecalibacterium and Roseburia may be helpful in reducing the incidence and co-occurring of both diseases.However, the results of the study were based on statistical analyses, and further studies such as targeted fecal manipulation are still required to confrm the fndings of this study.

Conclusion
In summary, we found that three genera including Faecalibacterium, Roseburia, and Streptococcus were perturbed in a similar way in MS and SZ patients, while the associations of them with serum cytokines were a bit diferent.MS pathogenesis was found to be associated with T17-related responses, while the increased expression of RANTES and MIP-1b may contribute to the infammatory disorders in SZ.Despite the abovementioned discrepancies, increased expression of TNF-α indicated the presence of infammation in both diseases.Te reduced Faecalibacterium and Roseburia and increased Streptococcus in together may contribute to systematic infammation and neuron damages, while Streptococcus might also be associated with the dysregulation of neurotransmitters, another possible driver for the neuronal dysfunction.Although further studies with higher depths and resolutions are still required for fully understanding the pathologies relating to both diseases, our study depicted the potential roles of gut microbiome and paved the way for developing gut microbiota-targeted therapies that are capable of managing two diseases with a single treat.

Figure 1 :
Figure 1: (a) A schema of the study design.(b) PCA score plot of ASV abundance tables.(c) Shannon index for MS patients, SZ patients, and corresponding controls.(d) Simpson index for MS patients, SZ patients, and corresponding controls.

Figure 2 :
Figure 2: (a) Boxplot illustrating the relative abundances of the three common genera signifcantly perturbed in MS and SZ patients.(b) LDA scores of the three common genera in datasets MS and SZ.(c, d) Functional pathways associated with the three common genera in diferent levels.

Figure 3 :
Figure 3: (a) Volcano plot showing the diferential serum cytokines between MS patients and corresponding controls.(b) Volcano plot showing the diferential serum cytokines between SZ patients and corresponding controls.(c) Venn diagram illustrating the diferential serum cytokines shared by both diseases and unique in each disease.(d) Heatmap illustrating the log2 (fold change) values for the six common cytokines in MS and SZ patients versus corresponding controls.

Figure 4 :Figure 5 :
Figure 4: (a) Heatmap illustrating the relative abundances of serum cytokines perturbed in MS patients.(b) Heatmap showing the relative abundances of serum cytokines perturbed in SZ patients.