Introduction

Schizophrenia (SZ) is a severe and complex psychiatric disorder characterized by cognitive impairment and behavioral dysfunction, which affects a large population worldwide and causes significant physical and social morbidity [1]. Compared to the general population, SZ is associated with a 2- to 2.5-fold higher risk of premature mortality and a reduced life expectancy of 15–20 years [2], making it a significant public health concern.

Family and twin studies have provided compelling evidence for the heritability of SZ, indicating that genetic variations contribute to an individual’s susceptibility to this disorder [3, 4]. Genome-wide association studies and large-scale sequencing efforts have identified numerous genetic loci implicated in synapse function, immune system function, and calcium signaling [5], shedding light on the molecular pathways and biological systems implicated in SZ. However, genetics alone cannot fully explain its etiology, necessitating exploration of other factors, particularly the environmental factors [6]. Among these, human microbiota has emerged as an important player in the pathophysiology of SZ, with potential therapeutic implications. Over the past two decades, next-generation sequencing techniques have revealed that the human microbiota, including bacteria, fungi, virus, and archaea, plays a crucial role in various aspects of health, such as digestion, metabolism, immune function, and even mood regulation. The bidirectional communication between the gut and the brain, known as the gut-brain axis, has been extensively studied. Increasing evidence suggests that the alterations of human microbiota, especially gut bacteria, can influence brain function and contribute to the development of SZ [7] through mechanisms involving the gut-brain axis, immune system modulation, and production of neuroactive compounds.

While the gut microbiota has been extensively studied, the oral microbiota is increasingly recognized as a potential contributor to neurological disorders. The oral cavity, as the origin of the gastrointestinal tract, harbors a diverse community of microorganisms that influence both oral and systemic health. Oral dysbiosis has been associated with the development of SZ [8]. Evidence suggests that the oral microbiota may influence SZ through two main pathways: the direct spread of oral pathogens or their byproducts to the brain, and the indirect modulation of the immune system and systemic inflammation. However, the role of oral fungi, which are an integral part of the oral ecosystem, remains largely unknown. Investigating the fungal component of the oral microbiota can provide a more comprehensive understanding of the complex interactions within the oral microbiome and their impact on brain health.

In our current study, we enrolled 215 elderly participants, including 118 SZ patients and 97 healthy controls (HCs), to explore the characteristics of oral fungal microbiota profiles by fungal-specific internal transcribed spacer 1 (ITS1) amplicon sequencing. Furthermore, we investigated correlations between key SZ-associated oral functional fungi and inflammatory cytokines. Through these investigations, we aim to gain a better understanding of the potential roles of oral microbiota in the pathogenesis of SZ. Our study could provide valuable insights into the diagnosis and treatment of SZ by identifying potential biomarkers and therapeutic targets related to the oral fungal microbiota.

Materials and methods

Participants’ enrollment and sample collection

118 well-controlled Chinese elderly SZ patients (age >62 years), and 97 age- and gender-matched cognitively normal HCs were recruited from Lishui, Zhejiang province (China) from April 2021 to August 2021. The study was approved by the Ethics Committee of Lishui Second People’s Hospital (reference no.: 20180705-1), and written informed consent was obtained from all participants or their caregivers prior to enrollment. The inclusion and exclusion criteria were outlined in our previous study [7]. The detailed characteristics of these participants were the same as our previous study [8].

For oral tongue coat sampling, three sterile swabs were used to collect samples from the middle portion of the tongue, with 20 swabs taken from each participant. The swabs were collected in sterile plastic cups [8]. For serum sample collection, we collected 0.5 mL of serum from each participant’s fasting morning blood. All samples were processed within 15 min and stored at −80 °C until use.

Fungal amplicon sequencing and bioinformatic analysis

Fungal genomic DNA was extracted from the oral swabs using a QIAamp® DNA Stool Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions, with additional glass-bead beating performed on a Mini-beadbeater (FastPrep; Thermo Electron Corporation, Boston, MA, USA). Fungal amplicon libraries targeting the ITS1 region (ITS1-F: 5ʹ- CTTGGTCATTTAGAGGAAGTAA-3ʹ; ITS1-R: 5ʹ- GCTGCGTTCTTCATCGATGC-3ʹ) were constructed by technical staff at Hangzhou KaiTai Bio-lab, followed by sequencing on the Illumina NovaSeq 6000 PE250 platform. After sequencing, QIIME2 (v2020.11) was used for subsequent bioinformatic analysis with default parameters. Reads from each sample were normalized to ensure consistent sampling depths before analysis. Data quality control, annotation, fungal α- and β-diversity analysis, fungal community composition analysis, and correlation analysis among key functional fungi were conducted as previously described for bacterial microbiota [7,8,9]. Additionally, enzyme classification (EC) numbers and MetaCyc pathway predictions were generated based on the amplicon sequence variants (ASVs) using PiCRUSt2 [10, 11]. Receiver operating characteristic (ROC) analysis was performed to evaluate the ability of key functional fungi to discriminate SZ from HCs.

Cytokine analysis

The systemic immune function of participants was evaluated using the Bio-Plex Pro™ Human Cytokine 27-plex assay kit (Bio-Rad, CA, USA; cat. no. M500KCAF0Y) [7,8,9, 12, 13]. Measurements of 16 cytokines, 6 chemokines, and 5 growth factors were conducted using an automatic immunoassay analyzer (Bio-Plex® 200 System; Bio-Rad Laboratories, Inc.) following the manufacturer’s instructions. Prior to analysis, serum samples were diluted fourfold with the sample diluent buffer. Results were presented as mean fluorescence intensity, and the cytokine concentrations were determined from the standard curve.

Statistical analysis

Differences were assessed using various statistical tests, including White’s nonparametric t test, independent t test, or Mann–Whitney U-test, depending on data distribution and assumptions. Correlation analyses were conducted using Spearman’s rank correlation test. Statistical analysis was performed utilizing SPSS v24.0 (SPSS Inc., Chicago, IL) and Statistical Analysis of Metagenomic Profiles (STAMP) v2.1.3. Graphical representations were generated using R packages and GraphPad Prism v6.0. All significance tests were two-sided, and p values were adjusted for multiple testing using the Benjamini-Hochberg method to control the False Discovery Rate (FDR). A threshold of FDR < 0.05 was considered statistically significant.

Accession number

The sequence data from this study are deposited in the GenBank Sequence Read Archive (PRJNA1136659).

Results

Altered structure of the oral fungal microbiota in patients with SZ

In our study, we obtained a total of 17,666,047 high-quality sequence reads from an initial pool of 19,304,576 raw sequence reads. On average, each sample yielded approximately 82,167 reads. To ensure comparability, we normalized the read count to 40,750 per sample for subsequent microbiota analysis. Across the entire cohort, we identified a total of 2473 operational taxonomic units (OTUs), with 2,012 OTUs in the HC group and 1251 OTUs in the SZ group. Fungal α-diversity indices were compared between SZ patients and HCs based on the relative OTU table (Fig. 1A–F). While overall fungal diversity did not show significant alterations, there was a notable decrease in the number of observed species in SZ patients (p < 0.05). Using principal coordinates analysis (PCoA) with Bray-Curtis, unweighted UniFrac, and weighted UniFrac algorithms, we observed significant differences in oral fungal β-diversity between SZ patients and HCs (ADONIS test: p = 0.001, p = 0.001, and p = 0.031, respectively; Fig. 1G–I), despite considerable interindividual variations. The rank abundance curve further illustrated the species richness and evenness of microbial communities across the groups, revealing that the SZ group exhibited lower species richness and evenness compared to the control group (Fig. 1J). Similarly, the Venn diagram showed fewer unique fungal phylotypes in SZ patients compared to HCs (Fig. 1K). Collectively, these findings suggest an altered fungal microbiota structure in individuals with SZ.

Fig. 1: Comparison of the overall structure of the oral fungal microbiota between SZ patients and healthy controls.
figure 1

A–F α-diversity indices (Shannon, Simpson and invsimpson) and richness indices (ACE, Chao1, and observed species) were utilized to assess the overall structure of oral fungal microbiota. Data are presented as mean ± standard deviation. Unpaired two-tailed t tests were employed for inter-group comparisons. G–I Principal coordinate analysis (PCoA) plots illustrate β-diversity of individual fungal microbiota based on Bray–Curtis, unweighted UniFrac, and weighted UniFrac distances. Each symbol represents a sample. J The rank-abundance curve of fungal OTUs derived from the two groups indicates that there were more OTUs present at low abundance in the oral fungal microbiota of healthy controls compared to SZ patients. K Venn diagram depicts overlap of Operational Taxonomic Units (OTUs) in microbiota associated with SZ and healthy controls.

Full size image

Oral fungal dysbiosis in patients with SZ

After taxonomic classification of the fungal community, we identified 9 phyla, 226 families, 436 genera, and 644 species from both cohorts. Figure 2 illustrates the oral fungal microbiota landscape at the phylum (A), family (B), genus (C), and species (D) levels, highlighting differences in composition between SZ patients and HCs. Using bacterial enterotype discovery analysis, we stratified the genus-level fungal compositions of the 215 samples into two distinct groups: a Malassezia-dominated mycotype (Cluster 1) and a Candida-dominated mycotype (Cluster 2) (Fig. 2E).

Fig. 2: Oral fungal microbiota compositions of in SZ patients and healthy controls.
figure 2

A Mean relative abundance of major phyla. B Mean relative abundance of major families. C Mean relative abundance of major genera. D Mean relative abundance of major species. E Oral mycotypes identification: a Malassezia-dominated mycotype (Cluster 1) and a Candida-dominated mycotype (Cluster 2).

Full size image

To identify SZ-associated oral fungi, we compared the abundance of individual fungi with a relative abundance greater than 0.01% against control samples. Employing the linear discriminant analysis (LDA) effect size (LEfSe) method, we identified key functional taxa at multiple taxonomic levels (LDA score >2, p < 0.05, Fig. 3). Figure 3A displayed cladograms that highlight differentiating fungal biomarkers from the phylum to genus levels, illustrating the representation of fungal taxa between SZ patients and HCs. Figure 3B showed potential SZ-associated fungal biomarkers at various taxonomic levels, with the most distinctive biomarkers at the genus level including Candida, Tausonia, Alternaria, Malassezia, Saccharomyces, Trichosporon, and Aspergillus.

Fig. 3: Differential oral fungal taxa between the SZ patients and healthy controls.
figure 3

A LEfSe cladograms illustrating fungal taxa significantly associated with SZ patients (red) or healthy controls (green). The size of each circle in the cladogram corresponds to the relative abundance of the fungal taxon. Circles represent taxonomic levels from inner to outer: phylum, class, order, family, and genus. Statistical significance was determined using the Wilcoxon rank-sum test with p < 0.05. B Histogram depicting the distribution of Linear Discriminant Analysis (LDA) scores (>2) for fungal taxa that show the greatest differences in abundance between SZ patients and healthy controls (p < 0.05).

Full size image

Using MetaStats 2.0, we compared the oral fungal microbiota at specific taxonomic levels. At the phylum level, Basidiomycota was enriched in SZ patients, while Ascomycota was decreased (Fig. 4A). At the family level, ten families including Malasseziaceae, Aspergillaceae, Phaeosphaeriaceae, Trichosporonaceae, Irpicaceae, and Peniophoraceae were elevated in SZ patients, and other five families such as Saccharomycetales_Incertae_sedis, Bolbitiaceae, Chaetomiaceae, Mortierellaceae, and Typhulaceae were significantly reduced (Fig. 4B). At the genus level, nine genera including Malassezia, Trichosporon, Leptospora, Cutaneotrichosporon, Purpureocillium, Peniophora, Irpex, Neosetophoma, and Tilletiopsis were more abundant in SZ patients, while other 12 genera such as Candida, Debaryomyces, Tausonia, Gibberella, Mortierella, Typhula, Gymnostellatospora, and Starmerella were notably decreased (Fig. 4C). At the species level, we observed enrichment of Malassezia restricta, Irpex hydnoides, Peniophora incarnata, Purpureocillium lilacinum, Trichosporon asahii, Periconia epilithographicola, and Dentocorticium bicolor in SZ patients. Meanwhile, species such as Candida albicans, Starmerella etchellsii, Debaryomyces udenii, Tausonia pullulans, and Coprinellus flocculosus were significantly reduced (Fig. 4D).

Fig. 4: Differential oral fungal taxa between SZ patients and healthy controls.
figure 4

Comparisons of the relative abundance of the abundant fungal taxa at the level of bacterial phylum (A), family (B), genus (C), and species (D) in the oral microbiota. The data are presented as the mean ± standard deviation. Mann–Whitney U-tests were used to analyze variation between the SZ patients and controls. * p < 0.05 compared with the control group.

Full size image

Oral fungi exist within intricate interactions rather than in isolation. To analyze the co-occurrence network of these genera and compute their topological properties, we employed the SparCC (Sparse Correlations for Compositional Data) algorithm to delve deeper into their interactions. The co-occurrence pattern of key functional fungi in SZ was more complex than that observed in HCs, as illustrated in Fig. 5. Overall, our findings strongly suggest the presence of dysbiosis in the oral fungal microbiota of SZ patients.

Fig. 5: Strengths of the correlation between abundant oral fungi in the SZ patients and controls.
figure 5

Co-occurrence network inferred from the SparCC algorithm applied to relative abundance genera data. Cytoscape version 3.6.1 was used for network construction. The red and blue lines represent positive and negative correlations, respectively.

Full size image

Using the identified key functional fungal taxa, we evaluated their potential roles in distinguishing SZ patients from HCs through ROC analysis. ROC curves were generated to assess the diagnostic performance of these fungi at species level, with area under the curve (AUC) values indicating accuracy. Among the identified differential species, Candida albicans (AUC = 0.703), Malassezia restricta (AUC = 0.633), Trichosporon asahii (AUC = 0.615), and Peniophora incarnata (AUC = 0.606) demonstrated the ability to distinguish SZ patients from HCs (Fig. 6). However, the overall diagnostic potential is limited due to relatively low AUC values for various fungal species, indicating that the utility of oral differential fungi in distinguishing SZ patients from HCs is constrained.

Fig. 6: The differential oral fungi as SZ diagnostic biomarkers.
figure 6

Receiver operating characteristic (ROC) curves were used to evaluate the discriminatory ability of oral differential fungi in distinguishing SZ patients from healthy controls. The area under the curve (AUC) represents the predictive accuracy of each ROC curve. A Candida albicans; (B) Malassezia restricta; (C) Trichosporon asahii; (D) Peniophora incarnata.

Full size image

Changed oral fungal metabolic pathways in SZ patients

PICRUSt2 was employed to infer enzyme (EC) abundances and MetaCyc pathway profiles from ITS1 sequencing data. The analysis identified a total of 992 ECs, which were categorized into 81 MetaCyc pathways. Significant decreases were observed in pathways related to urate biosynthesis/inosine 5’-phosphate degradation, guanosine nucleotides degradation III, adenine and adenosine salvage III, 1,3-propanediol biosynthesis (engineered), pyrimidine deoxyribonucleotides de novo biosynthesis I, superpathway of pyrimidine deoxyribonucleotides de novo biosynthesis (E. coli), pyrimidine deoxyribonucleotide phosphorylation, and GDP-mannose biosynthesis urate biosynthesis, guanosine nucleotide degradation in SZ patients, while L-arginine biosynthesis II (acetyl cycle) was enriched (p < 0.05; Fig. 7). Using Spearman’s correlation, we observed that SZ-enriched species, such as Malassezia restricta and Malasseziaceae sp., were positively correlated with L-arginine biosynthesis II (acetyl cycle) but negatively correlated with the decreased MetaCyc pathways. In contrast, Candida albicans exhibited a positive correlation with 1,3-propanediol biosynthesis (engineered), adenine and adenosine salvage III, and urate biosynthesis/inosine 5’-phosphate degradation, and a negative correlation with L-arginine biosynthesis II (acetyl cycle) (Fig. 8). These results suggest that complex metabolic dysregulations related to oral fungal taxa may influence the pathogenesis and progression of SZ (Fig. 9).

Fig. 7: Altered oral fungal-related metabolic pathways in SZ patients compared to healthy controls.
figure 7

The oral fungal-related MetaCyc pathways predicted by PiCRUSt2 were evaluated using two-sided Welch’s t test. Percentage comparisons between the groups for each MetaCyc pathway are displayed. The Benjamini–Hochberg method was applied for multiple testing correction based on the false discovery rate (FDR) using STAMP.

Full size image
Fig. 8: Correlations between oral differential fungi and MetaCyc pathways.
figure 8

The heatmap depicts Spearman’s correlation between oral differential fungal taxa and MetaCyc pathways in SZ patients. The color was according to the Spearman’s correlation coefficient distribution. Red represented significant positive correlation (p < 0.05); blue represented significantly negative correlation (p < 0.05), and white represented that the correlation was not significant (p > 0.05).

Full size image
Fig. 9: Correlations between oral differential fungi and systemic immune dysfunction in SZ patients.
figure 9

The heatmap illustrates Spearman’s rank correlations (r) and associated probabilities (p values) between key functional differential genera of gut microbiota and circulating inflammatory cytokines, chemokines, and growth factors in SZ patients. Significant correlations (*p < 0.05) are indicated.

Full size image

SZ-related immunological dysfunction correlated with fungal species

In our study, SZ patients exhibited elevated levels of inflammatory cytokines (IL-2, IL-4, IL-6, IL-8, TNF-α) and increased chemokines (Eotaxin, MIP-1α). Conversely, there were reduced levels of cytokines IL-9 and IL-13, as well as decreased chemokines MCP-1, MIP-1β, and RANTES (Fig. S1, p < 0.05 for each), indicating significant immune dysfunction in SZ patients. Using Spearman’s correlation, we found that SZ-enriched species such as Trichosporon asahii, Malasseziaceae sp., and Purpureocillium lilacinum showed positive correlations with elevated cytokine levels, including IL-2, IL-4, IL-6, IL-8, TNF-α, Eotaxin, and MIP-1α. Conversely, Candida albicans exhibited negative correlations with these cytokines. Additionally, Malasseziaceae sp. and Purpureocillium lilacinum were negatively correlated with IL-13, MCP-1, MIP-1β, and RANTES, while Candida albicans showed positively correlated with IL-13, MIP-1β, and RANTES. These correlation analyses suggest that the key functional differential fungi associated with SZ contribute to the host immune dysfunction, potentially influencing the pathogenesis of SZ through their interactions with host immune responses.

Discussion

The oral cavity is one of the important habitats in the human body, consisting of a diverse microbial community that includes bacteria and fungi, both of which play crucial roles in maintaining oral health and influencing systemic health through various interactions and metabolic activities [14]. Historically, research has predominantly focused on the bacterial component of the oral microbiota, often overlooking the importance of fungi. Despite representing a small fraction of the total microbiome, fungi are indispensable for a balanced oral microbiota and exert significant effects on various physiological and pathological processes in both oral and systemic contexts [15,16,17,18,19]. The advent of next-generation sequencing techniques has revolutionized our understanding of the fungal community within the oral microbiota [20], revealing their roles in microbial balance, immune modulation, and metabolic pathways. This advanced molecular approach has also elucidated fungal contributions to oral diseases and their potential impact on systemic inflammatory disorders. Understanding the fungal contributions is pivotal for advancing our knowledge of oral health and its broader implications for overall well-being. Elucidating the intricate roles of fungi within the oral microbiota will help reveal new strategies for managing both oral and systemic diseases.

Recently, numerous studies have emphasized the possible role of bacterial community in psychiatric disorders such as SZ [7, 8, 21]. The altered bacterial community might contribute to neuroinflammation, affect neurotransmitter levels, or influence immune responses, potentially exacerbating symptoms and influencing the course of SZ. While the role of bacteria in the oral microbiome and SZ is well established, the contribution of oral fungi remains relatively unexplored. Evidence suggests that oral fungal microbiota actively participates in various oral diseases, such as oral lichen planus, periodontitis, dental caries, and oral cancer [16,17,18,19]. However, its involvement in SZ pathophysiology is not well understood. In our previous work, we examined the oral bacterial community in the same cohort using tongue coat sampling, revealing SZ-associated oral dysbiosis characterized by increased levels of Streptococcus and Fusobacterium, as well as decreased levels of Prevotella and Veillonella [8]. This oral bacterial dysbiosis is believed to contribute to chronic low-grade inflammation and may potentially influence SZ symptoms. In our current study, for the first time, we explored the altered fungal community in tongue coat samples and systemic immune dysfunction among Chinese SZ patients. Similar to previous findings on SZ-associated gut fungal communities in a Chinese cohort [22], we found decreased richness (observed species) and altered β-diversity in the oral mycobiota of SZ patients. A higher number of fungal species were identified in healthy subjects, suggesting that the reduced diversity observed in SZ represented a dysbiotic state. However, contrasting with decreased α-diversity observed in the oral bacterial microbiota of the same SZ cohort [8], these findings suggest distinct patterns of fungal dysbiosis compared to bacterial dysbiosis in individuals with SZ. The rank abundance curve also revealed a shift towards lower diversity, with fewer dominant fungal species and a less even distribution of abundance across taxa compared to HCs. This suggests a parallel pattern of fungal dysbiosis across different body sites in individuals with SZ, highlighting the systemic nature of these microbial alterations in SZ.

Based on our current study, we noted significant differences in the composition of the oral fungal community between individuals with SZ and HCs. Similar to previous findings in the gut, our investigation revealed that the predominant phyla in the oral fungal microbiota were Ascomycota and Basidiomycota. Notably, SZ patients displayed a distinct fungal profile characterized by an enrichment of Basidiomycota and a reduction in Ascomycota compared to HCs. At the genus level, the SZ-associated fungal microbiota exhibited higher abundances of genera such as Malassezia, Trichosporon, and Leptospora, which have been associated with high levels of pro-inflammatory cytokines such as IL-6 and TNF-α. Conversely, genera like Candida, Debaryomyces, and Tausonia were found to be less abundant in SZ patients compared to HCs. Interestingly, our findings suggest the presence of two ecologically distinct oral mycotypes (mycobiome profiles) among SZ patients: one characterized by the dominance of Malassezia (belonging to Basidiomycota phylum) and the other by Candida (belonging to Ascomycota phylum). Specifically, Candida albicans was notably abundant and distinctive within the Candida mycotype, whereas Malassezia restricta predominated in the Malassezia mycotype, consistent with prior studies on saliva and oral mucosa samples [23, 24]. Similar findings have been also observed in the gut fungal communities of patients with multiple sclerosis, although each mycotype contains different fungi [25]. Of course, the oral mycotypes exhibited variability across different environmental niches, with distinct compositions of fungal communities observed between dental plaque and saliva, characterized by a higher abundance of Candida in plaque and Malassezia in saliva [26]. These distinct mycotypes in specific niches appear to correlate with different functional roles [27]. Both mycotypes were associated with distinct bacterial groups, wherein, the fungal genus Malassezia was found to be associated with inflammophilic bacteria [24]. Hong et al. noted a higher prevalence of the Candida mycotype in individuals with cancer, which positively correlated with steroid use, removable prosthesis usage, the number of teeth with visible caries, smoking habits, and plaque index [24]. In our oral fungal communities, the Candida mycotype was negatively associated with pro-inflammatory cytokines, while the Malassezia mycotype was positively correlated with them. Their distinct functions of the two oral mycotypes could indicate their interaction with the host immune system and their potential roles in SZ pathogenesis. Using ROC analysis, these specific oral fungi demonstrate potential as effective diagnostic biomarkers for SZ. However, our cohort only suggests the potential diagnostic value of these markers for SZ; further validation in larger, multi-center cohorts is essential to confirm their diagnostic utility. Nevertheless, this represents the first identification of different oral fungal microbiota compositions in SZ patients, which offer novel insights and avenues for exploring the roles and mechanisms of the specific oral fungi in the pathophysiology of SZ.

Our current study has identified Candida and Malassezia as two prominent fungi driving oral fungal dysbiosis, which may contribute to the development and progression of SZ. These fungi, typically associated with oral and skin environments, exhibit notable differences in abundance and diversity within the gut microbiota of individuals diagnosed with SZ compared to HCs [22]. However, their low abundance in these specific habitats has posed challenges for isolation using traditional culture-dependent methods. Instead, the detection of specific IgG antibodies targeting these fungi in serum samples often reflects an individual’s historical exposure to these microorganisms. Candida albicans, a key species within the Candida genus, has been highlighted in previous research for its elevated antibody levels in various psychiatric disorders, including depression, anxiety disorders, and SZ [28,29,30]. This suggests a potential pathogenic role in SZ development. Administration of probiotics, including Lactobacillus rhamnosus and Bifidobacterium animalis, may help normalize Candida albicans antibody levels and alleviate Candida albicans-associated gut discomfort in SZ individuals [30]. However, our ITS sequencing data revealed decreased levels of Candida, particularly Candida albicans, in the oral microbiota of SZ patients. Additionally, we found a negative correlation between Candida albicans levels and pro-inflammatory cytokines IL-6 and TNF-α. These results suggest that oral Candida albicans may play a role in immune modulation among SZ patients, potentially indicating a protective function against the onset or progression of SZ by influencing inflammatory pathways. Previous DNA-based studies have established Candida species, especially Candida albicans, as normal components of the human microbiota [31,32,33]. It commonly colonizes the oral mucosal surface of 30-70% of healthy individuals, coexisting harmlessly with hosts under normal conditions [34, 35]. Evidence indicates that Candida albicans, which has coevolved extensively with humans and cohabiting microbes, plays a critical role in shaping microbial community structure, metabolic function, and immune priming [36, 37]. Moreover, it can induce antigen-specific adaptive immune responses that are essential for maintaining immune homeostasis and preventing fungal infections [38]. However, under certain predisposing conditions, Candida albicans can rapidly transition to a pathogenic state [39], causing localized infections in the mouth, throat, and reproductive tract, or leading to systemic invasive candidiasis that affects the bloodstream, bones, and even the brain. Such infections are commonly observed in immune-compromised individuals, including those with HIV/AIDS, due to local overgrowth of these organisms [40]. In our SZ cohort, Candida albicans generally does not provoke active oral infections; instead, it has exhibited potential anti-inflammatory effects, suggesting it can be considered a commensal fungus within the oral microbiota of SZ patients. Despite the absence of active oral infection, SZ is often associated with chronic low-grade inflammation. Evidence has suggested that chronic low-grade inflammation can alter cognitive performance in SZ and other psychiatric disorders [41, 42]. The release of peripheral pro-inflammatory cytokines such as IL-6 and TNF-α affects the brain via neural (mainly vagal) pathways, interaction with cytokine receptors on cerebral endothelial cells and/or microglial activation [41]. However, the specific roles of Candida albicans in influencing chronic low-grade inflammation, interacting with oral bacterial communities, and participating in the oral fungi-gut-brain axis are still not well defined. Further research is warranted to comprehensively elucidate how Candida albicans contributes to the oral microbiome dynamics in SZ patients and its potential implications for disease progression and management.

In addition to Candida, several other oral fungi, including Malassezia, have been identified using ITS sequencing in oral samples from SZ patients. Malassezia, the second most abundant fungal genus in the oral microbiota of SZ patients, has been highlighted as a representative fungus of the oral mycotype [24]. Malassezia, lipid-dependent basidiomycetous yeast of the normal skin microbiome [43], have been detected in various anatomical sites beyond the skin due to recent advancements in high-throughput sequencing techniques. Dupuy et al. observed a notable prevalence of Malassezia spp. in the oral cavity, with these yeasts comprising 38% of sequences detected using ITS sequencing [23, 44]. However, the specific role of Malassezia within the oral ecosystem remains unclear. Baraniya et al. found that Malassezia is one of the most prevalent/abundant taxa in the children’s supragingival mycobiome, while Malassezia globosa was significantly enriched in caries-free subjects [45]. In the salivary microbiome, Mohamed et al. observed that higher levels of Malassezia were significantly associated with improved overall survival in patients with oral squamous cell carcinoma [46]. Most evidence regarding its presence and potential functions stems from studies in habitats such as the gut and brain, rather than the oral cavity. Malassezia has been more prevalently observed in conditions such as Crohn’s disease, colorectal cancer, chronic rhinosinusitis, cystic fibrosis pulmonary disease, and major neurological diseases like multiple sclerosis (MS), amyotrophic lateral sclerosis (ALS), Alzheimer’s disease (AD), and Parkinson’s disease (PD) [47,48,49,50,51,52]. Unexpectedly, Malassezia DNA has been confirmed in the brains of patients with AD, MS, and ALS through ITS sequencing, suggesting the possible colonization of Malassezia in human brain [52,53,54]. Our present study firstly revealed a notable enrichment of oral Malassezia, especially Malassezia restricta, in SZ patients. These findings indicated a potential systemic influence of Malassezia beyond its traditional association with skin microbiota. Notably, the SZ-associated Malassezia mycotype was positively associated with pro-inflammatory cytokines such as IL-6 and TNF-α. Consistent with previous studies in both mouse models and keratinocytes, Malassezia can stimulate the expression of chemokine CXCL8, which plays a crucial role in neutrophilic inflammation [55]. The correlations with systemic inflammation in SZ suggest a possible role for Malassezia in modulating immune responses in psychiatric disorders, which may influence the pathogenesis and progression of SZ. Evidence from studies on skin-associated Malassezia indicated that Malassezia species possess various virulence factors including the production of indoles, generation of reactive oxygen species, secretion of azelaic acid, and formation of biofilms, collectively contributing to their pathogenicity [56]. However, there was currently no more conclusive evidence to establish the functional role of oral Malassezia, as these abundant Malassezia such as Malassezia restricta could not be cultivated from the sequenced oral samples [24]. This limitation hinders the direct exploration of Malassezia’s exact involvement in the mechanisms underlying the pathogenesis of SZ. While the precise role of Malassezia in these disorders remains unclear, its ability to infect human cells, travel through the bloodstream, cross the blood-brain barrier, reside among lipid-rich neuronal cells, and evade or resist host immune responses are potential mechanisms that could contribute to SZ pathogenesis [57,58,59]. A previous study demonstrated that Malassezia restricta can induce neuroinflammation by activating helper T-cell (Th) 1 and Th17 immune responses in AD mice [48]. As far as the inflammation was concerned, Malassezia-related extracellular vesicles may be one of the crucial metabolites that induce inflammatory cytokines such as IL-4 and TNF-α which have been proven to participate in central nervous system disorders using in vitro studies [60, 61]. Additionally, Malassezia can produce metabolites such as Indirubin, which activate the NF-κB signaling pathway [49], confirming its potential involvement in triggering inflammatory responses in brain. While our understanding of the interactions between Malassezia and the brain has grown significantly in recent years, the specific involvement of oral Malassezia species in SZ remains an emerging and intriguing area of investigation, particularly given the challenges associated with culturing this fungus.

Furthermore, significant alterations were observed in other oral fungi, such as Peniophora incarnata, Purpureocillium lilacinum, and Trichosporon asahii, in the oral cavity of SZ patients. These species can be classified within the Malassezia mycotype, which is associated with systemic inflammation. Similar to Malassezia, increased abundance of Purpureocillium lilacinum and Trichosporon asahii also correlated positively with the pro-inflammatory cytokines and chemokines. Yuan et al. found that an increased abundance of gut Purpureocillium was associated with more severe psychiatric symptoms and poorer cognitive function in SZ patients, marking it as a key SZ-associated marker [62]. Gao et al. demonstrated that the expansion of opportunistic gut fungi Trichosporon may promote colorectal cancer progression [63]. Unlike the well-studied fungi Candia and Malassezia in the oral mycobiota, these less common species have not been thoroughly investigated for their roles in oral and systemic diseases such as SZ. This gap in research is primarily due to the challenges involved in successfully culturing them from samples. These fungi exist in low abundance and varying frequencies, posing challenges for their detection and isolation in the oral cavity. Additionally, each species may have specific growth requirements that are hard to replicate in laboratory settings. Despite these obstacles, understanding the contributions of these lesser-known fungi is crucial, as they may play significant roles in both oral and systemic health. Recent advancements in culturomics offer promise in overcoming the limitations of conventional cultivation methods, facilitating more thorough research into difficult-to-culture oral fungi. These innovations enhance our understanding of microbial ecosystems within the oral cavity. Integrating omics techniques, such as metagenomics and metabolomics, allows for a more comprehensive exploration of these species. Such approaches hold great potential for future studies aimed at uncovering the full spectrum of oral fungal communities and their effects on human health, including possible implications for conditions like SZ.

However, our study has several limitations. First, as a cross-sectional case-control study, it only identifies potential correlations between oral fungi and SZ. Larger-scale prospective longitudinal cohort studies or intervention studies are needed to establish causality for these key functional fungi in SZ. Second, we did not investigate the interactions between oral fungi and bacteria, which could provide valuable insights into the role of fungi within the oral ecosystem of individuals with SZ. Thirdly, our study did not consider potential confounding factors such as diet, oral hygiene practices, or medications, which may influence fungal composition in the oral cavity and contribute to the observed correlations with SZ. Lastly, our focus on Chinese participants limits the generalizability of our findings to other demographics and geographic regions. Future research should address these limitations to enhance our understanding of the relationship between oral fungi and SZ.

Taken together, we have demonstrated for the first time the presence of altered oral fungal communities and systemic immune dysfunction in Chinese SZ patients compared to HCs. We identified two distinct mycotypes: Candida mycotype and Malassezia mycotype, which could serve as biomarkers for SZ. Notably, Malassezia mycotype was positively correlated with peripheral pro-inflammatory cytokines, whereas Candida mycotype showed a negative correlation with these cytokines, suggesting that oral fungi may play a significant role in the pathogenesis of SZ. Further study is warranted to investigate to explore the specific roles and mechanisms of these key functional oral fungi in SZ, their interactions with oral bacteria in the oral ecosystem, and their potential therapeutic implications for SZ treatment.