The gut microbiome is a major player in immunomodulation and immune response both locally and in distant tissues (Rhee et al., 2009; Szabo, 2015; Scheffold and Bacher, 2020), up to fecal transplants determining the survival of influenza-infected mice (Zhang et al., 2020). Metabolites like butyrate are heavily implicated in immunomodulation (Chriett et al., 2019), circulate systemically (Ritzhaupt et al., 1998) and influence the severity of lung infections (Haak et al., 2018; Chemudupati et al., 2020).

Besides the bacterial kingdom, the microbiome includes viruses, archaea and fungi. These players exist in a complex ecosystem that is likely greater than the sum of its parts, though little is known about its order as of yet. Interaction and interdependence of microorganisms are commonly found in nature, so much that sociomicrobiology is extensively studied beyond medical approaches (Simonet and McNally, 2021). Both crosstalk and inter-kingdom interactions are relevant in the human microbiome, especially in the gut (Kruger et al., 2019). While bacteria, the predominant kingdom, keep fungal growth under control, fungi modify bacterial growth in return as fungi have the capacity for immunomodulation and can therefore aggravate or ameliorate illnesses and affect host reaction to bacteria (van Tilburg Bernardes et al., 2020).

We and others recently demonstrated that the gut microbiome of SARS-CoV-2 infected patients is characterized by a dysbiosis with a reduced bacterial richness, an increased abundance of Enterobacteriaceae and depletion of immunomodulatory bacteria (Gu et al., 2020; Zuo et al., 2020b; Reinold et al., 2021; Yeoh et al., 2021). Several bacterial producers of butyrate, a short-chain fatty acid known for anti-inflammatory effects (Segain et al., 2000), are demonstrably reduced in COVID-19 illness (Tang et al., 2020; Moreira-Rosario et al., 2021; Reinold et al., 2021).

Considering that severe cases of COVID-19 involve a dysregulated immune response with a lower abundance of T killer cells and a cytokine storm (Coperchini et al., 2021) and fungi are known to shape immunological responses and T cell action (Enaud et al., 2018; Scheffold et al., 2020), it is astounding that the fungal microbiome in COVID-19 has been largely overlooked so far. Just few studies have investigated the fungal gut microbiome in SARS-CoV-2 infected patients to date, one including 30 hospitalized COVID-19 patients (Zuo et al., 2020a) and the other analyzing 67 COVID-19 patients (Lv et al., 2021).

Our study aims to characterize the composition of the fungal microbiome linked to SARS-CoV-2 infection and COVID-19 severity in a cohort of 53 patients in order to assess the impact of fungal dysbiosis in the context of this pandemic. We found major differences in alpha and beta diversity in fungal blooms between mild and severe cases of COVID-19 and distinct differences in the interconnected fungal communities between SARS-CoV-2 positive and SARS-CoV-2 negative patients, raising the question of the role of the fungal microbiome in shaping the course of COVID-19.

Patients with severe/critical COVID-19 illness in our study exhibited a fungal gut microbiome characterized by a reduced Shannon diversity, richness and evenness and by a relative enrichment of Ascomycota compared to patients with non-severe COVID-19 illness. Furthermore, distinct differences in the pronounced interconnected fungal communities characterized the gut microbiome of SARS-CoV-2 positive and negative patients. The interconnected inter-kingdom networks between fungi and bacteria were sparse. Nevertheless, several anti-bacterial metabolites producing fungal genera exhibited different positive and negative correlations with several distinct bacterial genera in both groups, SARS-CoV-2 positive and negative patients.

Major changes of the bacterial gut microbiota have been associated with SARS-CoV-2 infection (Gu et al., 2020; Zuo et al., 2020b; Moreira-Rosario et al., 2021; Reinold et al., 2021; Yeoh et al., 2021; Zuo et al., 2021). In line with our study, the previous studies on the fungal gut microbiome in SARS-CoV-2 infected patients (Zuo et al., 2020a; Lv et al., 2021) also described a dysbiosis of the fungal gut microbiome in SARS-CoV-2 infection. However, the decrease of Shannon diversity and evenness, and the increase of the relative Ascomycota abundance we observed in patients with severe/critical COVID-19 illness, were not reported previously. In line with our study, Lv et al. reported a reduced fungal richness in COVID-19 (Lv et al., 2021). We found a reduced fungal diversity of the gut microbiome in SARS-CoV-2 positive patients, a feature commonly found in patients with inflammatory bowel disease (Chehoud et al., 2015). Lv. et al. (Lv et al., 2021) reported clustering of COVID-19 samples apart from healthy control samples in PCoA, a finding we observed when performing PCoA on phylum level distance matrix. Zuo et al. described a heterogeneous configuration of the fungal gut microbiome during hospitalization of COVID-19 patients (Zuo et al., 2020a), which is in line with our results of the fungal gut microbiome on genus or species level. In our study, severe/critical COVID-19 illness was characterized by depletion of several fungal taxa, a finding reported in COVID-19 patients previously (Lv et al., 2021). The genus Aspergillus of the Ascomycota phylum may cause invasive infectious complications in COVID-19 patients (Fekkar et al., 2021). Conflicting data, a depletion of Aspergillus (Lv et al., 2021) and an increase of this genus was reported (Zuo et al., 2020a) in the gut microbiome of COVID-19 patients. Aspergillus was not significantly associated with COVID-19 severity or SARS-CoV-2 infection in our study. Therefore, fungal genera, which have been reported as producers of antimicrobial metabolites that affect bacterial and fungal growth, including Debaryomyces (Banjara et al., 2016) and Aureobasidium, were overrepresented in SARS-CoV-2 positive patients of our study. Interestingly, Debaryomyces is enriched in the gut microbiome of patients with Crohn´s disease and preferentially inhabits inflamed intestinal tissues (Jain et al., 2021). In a mouse model, Debaryomyces impairs wound healing by blocking the myeloid cell specific type 1 interferon – CCL5 axis (Jain et al., 2021).

Compared with the bacterial gut microbiome, the fungal gut microbiome is low in diversity and highly variable between individuals (Nash et al., 2017), which is consistent with our data. A single species dominated the fungal gut microbiome, most overwhelmingly in patients with severe COVID-19 illness.

We found an increase in the relative abundance of the Ascomycota phylum with a mean relative abundance of >96% in patients with severe/critical COVID-19 illness. Ascomycota are critically involved in a complex interplay of not only inflammatory but also anti-inflammatory effects, affecting both local and distant organs (Simonet and McNally, 2021). Mono-colonization of the intestine with the commensal Ascomycota species Candida albicans or Saccharomyces cervisiae protected mice from severe influenza A virus infection and this effect was mediated by mannans, which are components of the cell walls of these fungi (Jiang et al., 2017).

Few data are available on fungi-fungi interactions in the gut, and to our knowledge, the within-kingdom interactions of the fungal gut microbiome in SARS-CoV-2 infected patients has not been investigated to date. We found distinct pronounced interconnected genus-level communities of the fungal gut microbiome between both groups, suggesting disease-specific correlations in the fungal gut microbiome linked to SARS-CoV-2 infection. The largest positively correlated community, distinctively observed in SARS-CoV-2 positive patients, included several producers of antimicrobial metabolites like Fusarium (Antonissen et al., 2014), Penicillium and Aureobasidium (Takesako et al., 1991). The latter was overrepresented in SARS-CoV-2 positive patients. Fusarium, one of the genera exhibiting the highest number of positive correlations with other genera in this community, produces a variety of mycotoxins that alter different intestinal defense mechanisms and aggravate infections (Antonissen et al., 2014). Gibberella, another genus with a high degree of positive correlations in this interconnected community is associated with chronic intestinal inflammation (Li et al., 2014). Besides competition between microorganisms of the gut microbiome, many observations indicate that interactions may be mutualistic with advantages for both interacting partners or commensalistic, providing an advantage for the other microorganism without advantage for itself (Richard and Sokol, 2019).

Inflammatory bowel disease is a risk factor for hospitalization in COVID-19 (Creemers et al., 2021) and probiotic fungi have positive effects in inflammatory bowel diseases (Guslandi et al., 2003; Jawhara and Poulain, 2007). Moreover, a crosstalk and interaction between fungi and bacteria exists, with fungi promoting shifts of the bacterial gut microbiome (van Tilburg Bernardes et al., 2020). Compared to the pronounced interconnected fungal communities present in SARS-CoV-2 positive and SARS-CoV-2 negative patients, the inter-kingdom co-occurrences were sparse but differed between both groups. However, investigating the effects of individual microorganisms on the entire human microbiome is complex and may be confounded by many factors, so that in-vitro experimental settings or organisms are frequently used as models. Interestingly, exposure to Debaryomyces, overrepresented in SARS-CoV-2 positive patients of our study, had the power to alter the bacterial gut microbiome in zebrafish larvae (Siriyappagouder et al., 2018).

A strength of our study is that patient groups did not differ significantly for several factors associated with alterations of the gut microbiome, as only one patient received antibiotics and no patient received antifungal therapy before sampling was performed. Nevertheless, sequencing of different amplicons, in our case 16S rRNA and ITS2 region, may cause a bias for the inter-kingdom network analyses. Our study is limited by the number of samples included in the study. Although we initially included 212 samples in PCR for ITS2-amplification, only 53 samples exhibited PCR products in agarose gel that were furthermore included in sequencing analysis. This owes to the fact that fungi comprise only a small portion of the microbial biomass and do not readily overgrow in a healthy gastrointestinal tract (Auchtung et al., 2018; Richard and Sokol, 2019). In addition, we cannot exclude, that other factors not matched between the groups, especially age and sex, which have been reported to be associated with alterations of the fungal gut microbiome (Strati et al., 2016), may contribute to the differences linked to SARS-CoV-2 infection and COVID-19 illness, observed in our study. However, the variables age, gender and comorbidity were not associated with significant differences of the fungal diversity, richness, evenness and relative abundance of Ascomycota within the groups of SARS-CoV-2 negative patients, non-severe and severe/critical COVID-19 patients ( Supplementary Figure 3 ). Another limitation is the lacking information about the diet of patients which may also affect the bacterial and fungal gut microbiome (David et al., 2014; Auchtung et al., 2018). An altered in-hospital nutrition and treatment of patients with severe/critical COVID-19 on intensive care units might also have an effect on the fungal gut microbiota. However, rectal samples in patients included in our study were taken shortly after hospital admission and no patient was treated on intensive care unit at the time point of rectal swab sampling.

Fungi have indirect effects on immune development by inducing ecological changes in the bacterial gut microbiome but also affect immune development by promoting infiltration of macrophages into lungs of mice in germ-free mice (van Tilburg Bernardes et al., 2020). They induce strong ecological changes to the gut microbiome and synergistic effects of bacteria and fungi in colonization, as well as increased cytokine levels in animals co-inoculated with both bacteria and fungi. Considering the influence of fungi on the bacterial gut microbiome, on cytokine levels and T cells (Bacher et al., 2019; Shao et al., 2019), the relevance of the fungal gut microbiome needs to be further investigated, especially in diseases like COVID-19, which present with a cytokine storm as a major pathophysiological player.

As our study was cross-sectional, rather than longitudinal, we cannot be certain whether the fungal gut microbiome changes after infection and affects the course of disease, or if the composition of gut mycobiome is a predisposing factor for critical illness.

The distinct compositional differences of the fungal gut microbiome in SARS-CoV-2 infection, especially in severe COVID-19 illness, studied from samples taken unaffected by the use of an antimicrobial therapy, illuminate the necessity of a broader approach to confirm the differences in the fungal gut microbiome in SARS-CoV-2 infection and to investigate whether the fungal microbiome changes are consequences of SARS-CoV-2 infection or a predisposing factor for critical illness.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: SRA NCBI, accession number PRJNA794154.

The studies involving human participants were reviewed and approved by Ethics Committee of the Medical Faculty of the University of Duisburg-Essen (20-9237-BO). The patients/participants provided their written informed consent to participate in this study.

JKe and AW conceived the study and raised the funding. JKe, AW, JR, and OW designed the study. JR, LB, JKo, OW, SD, MK, and JB collected study specimens. CF and JKe performed the experiments. JR analyzed the clinical data. JKe and FF analysed the microbiome data. JKe, A-KS and FF wrote the original manuscript and drafted it with substantial contributions from all other authors. All authors contributed to the article and approved the submitted version.

This work was supported by the Stiftung Universitätsmedizin Essen under Grant number 20204699115.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.