All animal studies were conducted at the Comparative Medicine Group at the college of Veterinary Medicine, Kansas State University. Experiments had received prior approval from the Institutional Animal Care and Use Committee (IACUC) of Kansas State University (IACUC Protocol #4248.1). Animals were housed with unlimited access to water and mouse chow and permitted to move without restraints within their cages. All efforts were made to minimize any discomfort, distress, and pain to animals.

C. albicans SC5314, a genome sequenced laboratory strain originally isolated from a patient with bloodstream infection, has been widely used in oral biofilm mouse models (Solis and Filler, 2012; Bertolini et al., 2019). C. albicans was routinely maintained in yeast extract (1%) peptone (2%) dextrose (2%) (YPD) agar (1.5%) plate. For animal experiments, a single C. albicans colony, initiated from a glycerol stock, was suspended in YPD broth and grown overnight at 30°C with shaking. Growing C. albicans at 30°C allows it to produce mostly yeast cells, which is convenient for determining the infectious dose. C. albicans is known to produce multicellular hyphae at 37°C. Cells were harvested and washed twice with PBS, and final cell density was adjusted as required.

Cortisone acetate (Sigma-Aldrich, St. Louis, MO, USA) was suspended (225 mg/kg body weight) in sterile PBS without calcium and magnesium (Gibco) containing 0.05% (v/v) Tween 80 (Sigma-Aldrich, St. Louis, MO, USA). The prepared cortisone acetate solution was made as suspension by sonication in a water bath for 1 min using Branson sonicator. Before injection, the preparation was vortexed and injected immediately (Solis and Filler, 2012). To avoid lot-to-lot variation, a single batch of cortisone acetate was used.

Our OPC model was designed with some modifications from previously established oral mouse model protocols (Solis and Filler, 2012; Xu et al., 2014). Eight-week-old female C57BL/6 mice (Charles River Laboratories, Wilmington, MA, USA) were purchased and acclimated one-week to housing before initiating the experiment. Animals were divided into two groups with n=6/group: (1) Cortisone acetate alone control, (2) Cortisone acetate + C. albicans infection. Two independent experiments were performed. Briefly, mice were immunosuppressed on day -1 and on alternate days (day 1 and 3) with cortisone acetate (subcutaneously) dissolved in 0.05% Tween 80/PBS. On day 0, mice were anesthetized by an intraperitoneal injection of ketamine: xylazine (100 and 10 mg/kg of body weight, respectively). Then the oral cavity was swabbed with a sterile cotton swab soaked with 100 µl of C. albicans (6 × 10⁸ cells/mL) or PBS which was left sublingually for 1 h. Fresh suspensions of C. albicans (6 × 10⁶ yeast cells/mL) were added to the drinking water for 3 days (day 0 to 2), to maintain high oral carriage loads (Bertolini et al., 2019). Microbes in drinking water were changed every 24 h. After 2 days, normal drinking water was supplied until the end of the experiment. To ensure the viability of the yeast in the drinking water, the water was sampled and cultured on a YPD plate after 24 h, and the results confirmed the survival of yeast (data not shown). On day 5, all mice were sacrificed and the organs, including tongue, intestine, kidney, and liver were excised and processed for analysis.

All tongues were excised and digitally photographed for biofilm analysis. Image analysis was done by ImageJ software (National Institute of Health, Bethesda, MD, USA). Tongue scoring (0-4) was done based on the percentage of the surface area affected as published elsewhere (Dwivedi et al., 2011; Xu et al., 2014) with slight modifications. Tongues with no lesions = 0, unmeasurable lesions = 1, measurable lesions covering < 25% surface area = 2, between 25–75% area = 3, and > 75% surface area = 4.

For CFU determinations, tongues, kidneys, and livers from control and experimental animals were excised, weighed, and homogenized. To calculate the fungal burden, homogenized tissue aliquots were plated on YPD agar containing chloramphenicol (10 μg/mL, Sigma) and kanamycin (50 μg/mL, Sigma). For Enterococcus sp. CFU determination, samples were plated on bile esculin (BE; Difco Bile Esculin Agar, Becton, Dickinson and Company (BD), Franklin Lakes, NJ, USA) agar and black colored colonies were counted. When necessary, colonies were plated or re-streaked on BHI (brain heart infusion agar, BD) containing 6.5% NaCl to further confirm the identification of Enterococci. On BHI salt-agar, the Enterococci, but not the viridian-D Streptococci (e.g., S. bovis) are expected to grow. Fresh stool samples were collected from each mouse, weighed, and homogenized in PBS. CFU for homogenized stool samples were determined on BHI, BHI + NaCl or BE agar plates. For oral microbe determination after immunosuppression, oral swabs were taken before and after the cortisone injection and plated on BHI or BE agar plates. Oral sampling was done for 30 s as previously described (Abusleme et al., 2017). Briefly, a sterile cotton swab was used to swab the oral cavity, starting from the tongue, then buccal areas, gingiva, palate, and ending with the gingiva on the lower incisors.

To differentiate between E. faecalis and E. faecium from mice oral samples, the 16S rRNA gene was amplified by PCR using genomic DNA and species-specific 16S rRNA primers (Ryu et al., 2013). The PCR primers for E. faecalis (forward 5-CATAACAGTTTATGCCGCATGGCATAAGAG-3, reverse 5-GGGGACGTTCAGTTACTAACG-3, accession number DQ239694) and E. faecium 16S rRNA (forward 5-TATAACAATCGAAACCGCATGGTTTTGATT-3, reverse 5- AGGGATGAACAGTTACTCTCA-3, accession number AY675247.1) genes were employed, and the amplified PCR products were sequenced commercially by the Sanger DNA sequencing method. The DNA sequences were analyzed using the NCBI nucleotide BLAST program under default mode. Sequences with the highest similarities (>70%) were considered matches. The 16S rRNA DNA sequence obtained in this study was deposited in the GenBank with accession number MZ868593.

To determine the changes in the gut bacterial and fungal communities of the immunosuppressed C. albicans infected and uninfected mice, we took the targeted sequencing of the 16S rRNA gene and ITS2 amplicons for bacteria and fungi, respectively. Fecal samples from normal non-immunosuppressed and uninfected mice were also included as additional control. Fresh fecal samples collected from these three mice groups were stored at -80 ° C immediately after collection. Genomic DNAs were extracted using QIAmp PowerFecal isolation kit (Qiagen, Germantown, MD, USA), purified using Qiagen DNA purification kit, analyzed for quantity and quality by Qubit/Nanodrop and kept at -20°C until sequencing. The bacterial 16S rRNA gene V4 region was amplified during library preparation via Illumina’s Nextera XT Index Kit v2 (Illumina, Inc., San Diego, CA) (primers: 515F, 5-GTGCCAGCMGCCGCGGTAA-3 and 806R, 5-GGACTACHVGGGTWTCTAAT-3) (Caporaso et al., 2011). Similarly, the internal transcribed spacer region ITS2 was amplified using the primers fITS7/ITS4 combination (White et al., 1990; Ihrmark et al., 2012) (primers: 5-GTGARTCATCGAATCTTTG-3 & 5-TCCTCCGCTTATTGATATGC-3) (Sarkar et al., 2022). The quality of the libraries was confirmed by a bioanalyzer. Library preparation and subsequent sequencing also included a no-template negative control. Sequencing was performed on an Illumina MiSeq, which generated paired-end 250 bp reads at the Kansas State University’s Integrated Genomics Facility. DNA sequences were analyzed as described (Sarkar et al., 2022), and the taxonomic assignments of the microbes were visualized using QIIME 2 (Bolyen et al., 2019). The mice gut amplicon DNA sequence obtained in this study was deposited in the NCBI (National Center for Biotechnology Information) repository with accession number PRJNA1142050.

Fresh tissues samples were homogenized and smeared over the slide followed by methanol fixation for 5 min. The slide was rinsed with distilled water and stained with calcofluor white solution (0.1%) in PBS for 5 min. After two washes with distilled water, dehydration was done with 95% ethanol, absolute ethanol, and then xylene. Coverslips were placed and slides were imaged using a fluorescence microscope (Leica DM-6B).

Portions of tongue, liver, kidney, and cecum were collected and immediately placed in 10% buffered formalin for fixation. Mouse intestines were collected and processed with a gut bundling technique as published elsewhere (Williams et al., 2016). Fixed tissues were processed on a Sakura Tissue-TEK VIP 6 processor prior to paraffin embedding. Four micrometer thick serial sections were stained with hematoxylin and eosin (H & E), Grocott’s methenamine silver (GMS), or periodic acid Schiff’s (PAS) stains on a Leica Autostainer XL ST5010. An Olympus U-TV0.5XC-3 scope mounted camera using CellSens software acquired the slide images.

Data were analyzed for statistical differences using the Graph-Pad Prism 9.0 software. Statistical significance was determined by unpaired two-tailed t-test, assuming equal variances, or the Mann Whitney test when data were not normally distributed. Statistical significance for all tests was set at P < 0.05. P-values for each analysis are indicated in the representative graphs, where “ns” denotes non-significant. Each symbol represents one mouse or group.

To examine Candida infection under the immunosuppressed condition we used a previously established OPC model with some modifications. To mimic the immunosuppressed condition and to form a visible biofilm, mice were treated subcutaneously with cortisone acetate every two days starting 1 day before infection ( Figure 1A ). We treated the mice sublingually with C. albicans and supplied the fungus in drinking water for 3 days after infection to form a strong biofilm over the mouse tongue ( Figures 1A , 2A ). An oral swab was taken on day -1 (before giving cortisone acetate injection) and day 0 (before infection) to determine the bacterial load. As expected, there was a heavy bacterial load after cortisone acetate injection ( Figure 1B ) when compared to before immunosuppression (on day -1). The overgrowth of bacteria in the oral cavity of immunosuppressed mice indicates the dysbiosis of oral environment and can lead to the appearance of opportunistic pathogens. Our OPC infection model showed a significant reduction in body weight of mice 3 days after C. albicans infection in comparison to the uninfected cortisone acetate group ( Figure 1C ). This reduction in body weight correlated with the oral biofilm that was formed over the tongue and its dissemination to the other organs of the mouse.

Immunosuppressed C57BL/6 mice with C. albicans infection developed robust biofilms over the tongue that were visible as white plaques in contrast to the cortisone acetate alone control group mice where clear tongue surface was seen ( Figure 2A ). The tongue pathology score was as follows: Tongues with no lesions = 0, unmeasurable lesions = 1, measurable lesions covering < 25% surface area = 2, between 25–75% area = 3, and > 75% surface area = 4. Oral thrush lesions (biofilms) were scored based on the surface area affected with an average of ~68% of tongue surface affected and the pathology score was found to be between 3 to 4 in the infected mice group ( Figure 2B ). CFU analysis of the Candida from infected mice tongue shows high fungal burden ( Figure 2C ) in agreement with the tongue pictures showing biofilm and pathological scoring ( Figures 2A, B ). In addition to assessing the fungal burden, tongue tissue harvested from mice was processed for histopathology to visualize the fungal hyphae and epithelial integrity. Microscopic analysis of GMS-stained tissue sections revealed extensive fungal adherence and hyphal penetration of the epithelial tissue ( Figure 2D ). The H & E staining showed the epithelial hyperplasia, hyperkeratosis, and infiltration of inflammatory cells compared to the cortisone acetate only (control) group. In addition, epithelial invasion by C. albicans with epithelial damage was seen ( Figure 2E ).

To determine the dissemination of C. albicans after immunosuppression, we performed the histopathological analysis of GI tissues (small intestine and cecum) and CFU of stool samples. The H & E staining shows severe changes to the intestinal mucosa, with the marked attenuation of the intestinal villi in the Candida infected group ( Figure 3A ). To further evaluate if C. albicans is directly associated with intestinal villi damage, small intestine tissue sections of control and Candida infected groups were stained with periodic acid Schiff’s (PAS) reagent. With the present method, no fungal organisms were observed at the damaged villi ( Supplementary Figure S1 ), suggesting that C. albicans may cause mucosal attenuation indirectly or via other unknown mechanisms, including dysbiosis in the gut microbiome. In the infected cecum sample, there were increased numbers of inflammatory cells in the mucosa ( Figure 3B ). Further, we observed a significant amount of C. albicans CFU in the fecal sample of the infected group (p < 0.0027) ( Figure 3C ).

To assess whether the loss of GI tract integrity leads to dissemination of C. albicans, we performed CFU and histology analysis of the liver and kidney. We found a detectable fungal burden in the liver by determining CFU ( Figure 4A ) but not in the kidney except in one mouse ( Figure 5A ). Even though some mice did not show a fungal burden in the liver, most showed the presence of C. albicans with significantly higher values than controls ( Figure 4A ). These data suggest that immunosuppression and additional fungal load in the drinking water could create a favorable environment for bacterial dysbiosis and dissemination of C. albicans. Fungal dissemination in the liver was further assessed by H & E staining of the tissue sections and calcofluor staining of the liver homogenate. While there was no major change in histology sections of the infected liver except increased lipid vacuolation, fungal hyphae and yeasts were observed by calcofluor staining ( Figures 4B, C ). In agreement with the CFU results, histology staining showed that C. albicans did not disseminate to the kidney ( Figure 5B ).

Enterococcus sp. is one of the members of the Enterococcaceae family found in the oral cavity and GI tracts of animals and humans, and it is known to produce a plethora of virulence factors and cause disease under immunosuppressed conditions (Jett et al., 1994). Since we found an increased load of bacterial colonies from the oral swabs of the immunosuppressed mice ( Figure 1B ), we performed CFU analysis to study the dynamics of oral and intestinal enterococcal populations in response to C. albicans infection in immunosuppressed mice. We used BHI, BHI + NaCl, & BE agar plates. We used BE & salt agar media for confirmation ( Supplementary Figure S2 ). We calculated the % CFU between BHI vs BHI + NaCl agar plates to determine Enterococcus sp. precisely. After C. albicans infection, bacterial CFU counts of BHI vs BHI + NaCl agar plates found about 25% of Enterococcus sp. in the tongues of C. albicans infected and 35% in the uninfected groups. CFU counts of the stool samples found 46% Enterococcus sp. in Candida infected and 41% in uninfected mice stools. Interestingly, Enterococcus sp. was absent in the kidney tissues of the fungus-infected group, while it was present in the cortisone acetate alone control group ( Figure 6A ). The absence of Enterococcus sp. and C. albicans (except one mouse) in the infection group agrees with the normal kidney histology observed ( Figure 5B ), which may suggest that synergistic fungus-bacterial interaction or high fungal burden is required for kidney tissue damage. However, further evidence is needed to support this statement. To determine the identification of Enterococcus species, we performed species-specific 16S rRNA PCR for Enterococcus species from the tongue, liver, kidney (one mouse), and stool samples ( Supplementary Figure S3 ). The PCR and sequencing results from the tongue sample identified E. faecalis, not E. faecium ( Figure 6B , Supplementary Figure S4 ).

The GI tract harbors diverse microbes with a wide array of functions for hosts, such as protecting from pathogen invasion, maintaining a healthy gut epithelial barrier, and aiding in nutrient availability (Barko et al., 2018). The observed damages to the gut epithelial barrier and the spread of Enterococcus sp. and C. albicans to various organs of the mice in this study led us to investigate the alterations in the gut microbial communities. We used fecal samples from three groups of mice, at least three samples for each group, for metagenomic analysis. The bacterial 16S rRNA and fungal ITS2 genes were amplified from the fecal microbial DNA and sequenced on an Illumina platform. The results shown in Figure 7 reveal major changes in bacterial communities. Lactobacillus is found to be high in all the groups with variable percentages. In non-immunosuppressed and uninfected control (OC group) mice, a high number of bacterial members are found to be Lactobacillus and Muribaculaceae, followed by Dubosiella, Turicibacter, Alistipes, Lachnospirales, Bifidobacterium, Enterorhabdus etc. Similar to the OC group, the immunosuppressed and uninfected control (O group) also shows abundant presence of Lactobacillus and Muribaculaceae followed by Candidatus arthromitus, Alistipes, Bacteroides, Lachnospirales, Clostridia and Enterobacteriaceae (Escherichia_Shigella, etc.). Remarkably, the taxonomic order of bacterial members was changed in immunosuppressed and C. albicans infected (OT group) and as follows: Lactobacillus, Odoribacter, Enterobacteriaceae, Candidatus arthromitus, Alistipes, Tannerellaceae, Lachnospirales, and Clostridia, etc. Overall, the members of the Muribaculaceae, Erysipelotrichaceae (Dubosiella), and Bifidobacteriaceae, and Turicibacter are considered beneficial (Liu et al., 2023; Lynch et al., 2023), and they are high numbers in non-immunosuppressed and uninfected control (OC group) mice ( Figure 7 ). These bacterial members are absent in the OT or less in the O (immunosuppressed) groups, and instead, the pathogenic bacteria, including bacteria from Enterobacteriaceae, Clostridiaceae, Odoribacter, Tannerellaceae, and Enterococcaceae (Enterococcus) dominate ( Figure 7 ).

The role of fungal communities in gut-microbial activity is poorly known. To determine the fungal taxa in the GI tracts of the above three groups of mice, we used ITS2-targeted amplicon sequencing, and the results are shown in Figure 8 . While the non-immunosuppressed control mice (OC) group shows saprophytic fungi (Rhizopus, Mucor, Penicillium) abundantly, the immunosuppressed mice (O) group shows a mixture of fungi, including opportunistic pathogens (Cladosporium, Alternaria, Epicoccum, etc.). A small amount of Saccharomycetales yeasts, which are not C. albicans but likely non-albicans sp., such as C. tropicalis (Iliev et al., 2012), are present in these mice. Observing the OT group, we see an overwhelming presence of Candida, making it difficult to detect other fungal species ( Figure 8 ). Although C. albicans is not a natural commensal or pathogen of mice (Savage and Dubos, 1967), the fungus can colonize and cause candidiasis in antibiotic-treated, neonatal, or immunosuppressed mice (Koh, 2013), suggesting bacterial communities play an important role in C. albicans disease.