Powder diet exacerbates oropharyngeal candidiasis in a mouse model

ABSTRACT This study reports on the influence of a powder diet in a mouse model of oropharyngeal candidiasis (OPC), a significant health concern caused primarily by Candida albicans. Despite identical nutritional composition, we found that a powdered diet significantly increased Candida burdens and oral lesions, and aggravated weight loss compared to a standard pelleted diet. High fungal burdens and severe oral lesions were accomplished within 48 hours after infection with only one dose of cortisone. Moreover, mice on a powder diet recovered a week after infection. Using a powder diet, we thus modified the cortisone OPC murine model in a way that simplifies the infection process, enhances reproducibility, and facilitates studies investigating both pathogenesis and recovery processes. Our findings also underscore the pivotal role of the physical form of the diet in the progression and severity of oral Candida infection in this model. Future research should investigate this relationship further to broaden our understanding of the underlying mechanisms, potentially leading to novel prevention strategies and improved disease management. IMPORTANCE Oropharyngeal candidiasis (OPC) is a multifactorial disease and a significant health concern. We found that the physical form of the diet plays a critical role in the severity and progression of OPC. We developed a modified cortisone OPC murine model that facilitates studies investigating pathogenesis and recovery processes.

nutritional composition to include texture.For example, C57BL/6 mice fed with either pelleted or powdered isocaloric diets had different body composition and insulin resistance outcomes (16).Other studies showed that compared to a pellet diet, a powdered diet leads to a higher glycemic response (17) and substantially decreases the alpha diversity of the gastrointestinal microbiome (18).
In our previous study on the effect of a powder-based prebiotic diet in a murine OPC model, we noticed reproducibly high fungal burdens in all mice of the control powder diet groups (15).To follow up on this interesting observation, we tested the hypothesis that the physical form of diet may alter the course of oral Candida infection in mice.Published OPC mouse models use pellet-based diets and most involve administering cortisone twice or more frequently (19)(20)(21)(22)(23), and performing two Candida inoculations under anesthesia (20,23).Some existing models also require the use of antibiotics (21,22) but yield a relatively low Candida biomass.In this work, we compared the most common pellet diet, a two-dose cortisone immunosuppression murine OPC model, with a powdered diet and the same or lower immunosuppression regimens.Our results indicate that a powdered diet increases weight loss, oral Candida biomass, biofilm, and tongue lesion development and requires lower levels of immunosuppression for reproducibly measurable infection outcomes.Importantly, despite developing severe oral lesions and weight loss, mice on a powder diet show signs of recovery over time and clear the infection within 7 days post-infection.Therefore, this model can also serve as a recovery model, allowing the identification of host and microbial factors that play an important role in clearing this infection.

Study animals
Female C57BL/6 mice, aged between 3 and 5 weeks and weighing 17-19 g, were obtained from Jackson Laboratories.Only female mice were used because sex is not a significant biological risk factor in OPC (24,25).Mice were acclimated for 1-2 weeks before initiating the experiments.Before the interventions, mice were randomized into control and experimental groups, relocated to appropriate individually ventilated cages (five or six mice per cage), tagged, and weighed.All animal experiments adhered to the guidelines stipulated by the National Institutes of Health, the Animal Welfare Act, and the University of Connecticut Institutional Animal Use and Care Committee (IACUC).The UConn Health IACUC committee approved all protocols (IACUC protocol #102251-0323).

Diet
Mice were assigned to either a pelleted or a powdered diet (T.2918 diet, Envigo, USA, different lots), starting 2 days prior to infection.The diets have identical nutritional compositions, differing only in their physical form.The diet's macronutrient profile is as follows: total protein 18.4%, total fat 13%, total carbohydrates 65.7%, and others 2.9% (details on the diet's composition can be found at: https://insights.inotivco.com/hubfs/resources/data-sheets/2918X.pdf).The pelleted diet was placed on metallic lids at the top of the cage and was unchanged during the duration of the experiment, whereas the powdered diet was distributed in 4-oz food follower containers with a lid (DYETS Inc., USA, DYET #910016 and #910024, respectively) placed on the bottom of the cage and was replaced daily to reduce soiling and spilling.Mice were provided sterile drinking water, and both food and water were available ad libitum.Diets are provided irradiatedsterile (<50 kGy) by the vendor and were further plated on Sabouraud dextrose (SD) agar (BD-Difco, #210950) to rule out fungal contamination prior to feeding.

Candida oral infection
Frozen stocks of C. albicans SC5314 were streaked onto SD agar, incubated aerobically at 30°C for 48 hours, and then subcultured in Yeast extract-Peptone-Dextrose (YPD) broth and incubated aerobically at 30°C for 24 hours.Overnight cultures of C. albicans were washed with Phosphate-buffered Saline (PBS) solution and adjusted to 1 × 10 8 blastospores/mL in PBS.Two sterile cotton balls were soaked with 100 μL of this yeast suspension.Mice were immunosuppressed either once with a single subcutaneous injection of cortisone acetate (225 mg/kg) a day before infection or twice, 1 day before and 1 day after infection.A non-immunosuppressed infected group was also included.For the oral fungal infection, mice were anesthetized with ketamine-xylazine (100 mg/kg of body weight and 10 mg/kg of body weight, respectively), and then two Candidasoaked cotton balls were placed sublingually for 2 hours.Non-immunosuppressed non-infected under a pelleted diet, and non-immunosuppressed non-infected under a powdered diet mice were used as controls.Changes in body weight were monitored every other day to assess animal morbidity with results expressed as a percentage of initial weight.A timeline of experimental interventions is depicted in Fig. 1.

Histological staining
Tissues were fixed with 4% paraformaldehyde at 4°C, dehydrated, and processed for paraffin embedding.For immunofluorescence staining, sections were deparaffinized and stained with an Fluorescein isothiocyanate (FITC)-conjugated anti-Candida antibody (Meridian Life Science, #B65411F) and anti-E-cadherin (BD Transduction, #610182) used as primary antibodies.A secondary antibody conjugated with Alexa Fluor 568 (Invitro gen, #A11031) was used for assessing the integrity of the mucosal barrier (via E-cad herin), as described previously (27).Finally, Hoechst 33258 (Invitrogen, H3569) was used as a counter stain following a previously published method (20).Tissues were also stained with hematoxylin-eosin (H&E) to assess the structural integrity.Images were obtained using a Zeiss Axio Imager M1 microscope.Whole tongue images were assembled using 200× image tiles from the FITC-conjugated anti-Candida antibody FIG 1 Timeline of infection.To compare powder and pellet diets, mice were infected with Candida 1 day after a cortisone injection, received a second dose of cortisone a day later, and were euthanized 48 hours after infection.In addition, Candida-infected mice from the single cortisone group under the powdered diet were sacrificed at days 3, 5, and 7 to follow the progression of the infection.
immunofluorescence staining, processed in Zen v3.3 Blue edition (Zeiss), stitched in Imaris Stitcher v10.0.0 (Oxford Instruments), and analyzed in Imaris v10.0.1 (Oxford Instruments) to visualize the fungal signal intensity and quantify the biofilm area.The biofilm area was quantified by creating a surface layer over the fluorescence signal of the FITC-conjugated anti-Candida antibody and then generating the total area values.Prism v10.0.2 (GraphPad) was used to perform the statistical analyses and generate the corresponding graphs.Images of the whole tongue and regions of interest from the H&E staining were assembled using 200× image tiles from the H&E stained sections obtained in Zen v3.3 Blue edition (Zeiss), stitched in Imaris Stitcher v10.0.0 (Oxford Instruments), and post-processed in Imaris v10.0.1 (Oxford Instruments).The final images generated by Zen and Imaris were saved in .tiffformat.There may be visible lines from the software stitching the images together.

Powder diet aggravates oropharyngeal candidiasis in the cortisone immuno suppression mouse model
We first compared the standard pellet diet versus the powder diet in a commonly used cortisone immunosuppression oral infection protocol (20,23).In these experiments, mice received two cortisone doses, 1 day prior and 1 day after oral infection, and were sacrificed 48 hours post-infection (Fig. 1).The physical form of the diet had no significant effect on the body weight of uninfected mice from both the non-immunosuppressed and immunosuppressed groups (Fig. 2A).In Candida-infected mice, the powder-fed cohort displayed a significantly increased body weight loss compared to their pellet diet counterparts, suggesting a more severe infection (Fig. 2B).
In uninfected mice, the physical form of the diet did not significantly affect the indigenous fungal biomass, as assessed by qPCR with primers for a broad range of fungi (Fig. 3A).A comparison of Candida-infected mice receiving two doses of cortisone under the two different diets showed significantly higher oral fungal burdens (as assessed by both qPCR and Candida CFU counts) in the powdered diet than the pelleted diet group (Fig. 3B).These results were supported by the finding that the powdered diet cohort manifested a significantly larger tongue lesion area compared to their pellet-fed counterparts (63% ± 1% versus 36% ± 25% of the lesion area, respectively) (Fig. 4A).In addition, fungal biofilms as seen by immunofluorescence staining appeared as a thin or interrupted layer in the pelleted diet group, whereas the biofilms in the powdered diet group appeared to be uniformly thick, covering the entire dorsal tongue surface.Quantification of the biofilm fluorescent area showed that it was significantly higher in the powdered diet group than in the pellet diet group (Fig. 4B).
At the histologic level, infected mice on the powder diet exhibited greater destruc tion of the tongue stratified epithelium.In this group, we observed areas of significant surface erosion and desquamation with only two to three remaining intact epithelial layers and reduced E-cadherin immunoreactivity suggesting a compromised mucosal barrier (27,33,34).Consistent with this finding, we observed fungal organisms invading the superficial layers of the oral mucosa (Fig. 5).In comparison, mice on the pellet diet exhibited more superficial mucosal destruction with minor epithelial atrophy and pronounced desquamation.In these mice, epithelial barrier integrity was maintained as shown by the E-cadherin staining pattern and by the finding of fungal organisms located mostly superficially (Fig. 5).Collectively, our data show that the powder diet aggravates all Candida infection outcomes in the cortisone model of infection.While the mechanism that promotes Candida infection under a powdered diet needs to be further explored, it is possible that the texture and/or consistency of this diet might provide a structural scaffold that facilitates fungal biofilm formation.Alternatively, a powder diet may increase dietary fungal nutrient bioavailability by slower food clearance from the oral cavity due to mucosal dryness and retention.

Powder diet OPC model requires lower immunosuppression levels
The "gold standard" murine OPC model requires three cortisone injections, to obtain reproducible oral fungal burdens, with CFUs typically in the range of 10 5 -10 6 /g of tissue, 5 days post-infection (23).Because we observed the same range of CFU outcomes with the powder diet and only two cortisone injections, we next questioned whether we could reduce the cortisone doses even further and continue to achieve reproducibly high fungal burdens with this diet.In uninfected mice on a powdered diet, increasing the cortisone doses from 0 to 2 was inversely associated with the indigenous fungal biomass (Fig. 3A), indicating that immunosuppression alone cannot promote the growth of oral commensal fungi in mice.
In infected, non-immunosuppressed mice on a powder diet, we observed some weight loss associated with Candida infection, although this did not reach statistical Mice on the powdered diet had significantly higher total fungal biomass (qPCR, hollow squares) and Candida viable counts (CFU, black diamonds) compared to the pellet diet.No significant differences in fungal biomass between one and two cortisone doses in mice on a powder diet.(C) Immunosuppressed Candida-infected mice showed a significantly higher fungal load than the immunocompetent and non-infected cohorts.One-way ANOVA with Dunnett's T3 test.****P < 0.0001.significance (Fig. 2B).This observation is consistent with the mild transient effect of oral C. albicans SC5314 infection on the weight of immunocompetent C57/BL6 mice on a pellet diet (35).Importantly, non-immunosuppressed Candida-infected mice on the powdered diet displayed similar fungal burdens and tongue lesion areas as the Candida-infected mice under the pelleted diet receiving two cortisone injections (Fig. 3B and C).However, the biofilm area was significantly lower (Fig. 4A and B), and the tongue epithelium was mostly intact with few areas of desquamation in immunocom petent mice on a powder diet (Fig. 5).The discrepancy between fungal burdens and biofilm surface area in immunocompetent mice could be explained by the removal of loosely adhering organisms during staining, suggesting that the intact immune response mounted prevented fungal organisms from adhering, superficially invading, and forming robust biofilms on the tongue surface.
In the Candida-infected immunosuppressed groups, regardless of the immunosup pression level (one or two doses of cortisone), mice on the powdered diet exhibited a similar loss of body weight (Fig. 2B), suggesting similar severity of infection.Consistent with the weight loss data, there were no significant differences in the fungal biomass between one and two cortisone injections within the Candida-infected, powdered diet cohorts.This indicated that a single injection is sufficient to achieve high oral fungal burdens when mice receive a powder diet, although reproducibility was somewhat compromised, as shown by the mouse-mouse variability in the qPCR and CFU data within this group (Fig. 3B).The extent of tongue lesions and Candida biofilm areas were also similar in the two cortisone groups (Fig. 4A and B).At the histologic level, mice on one cortisone dose exhibited areas of surface erosion and desquamation, although they had more remaining intact epithelial layers and higher E-cadherin protein expression than mice on two doses.This finding suggested the maintenance of a more intact mucosal barrier with lower levels of immunosuppression.Despite this, we observed fungal invasion of the epithelial layer in the single cortisone dose group, and mucosal destruction was significantly higher compared to the pellet diet group which received two cortisone doses (Fig. 5).
Collectively, these findings support the notion that Candida oral mucosal virulence with the powder diet can be accomplished with lower levels of immunosuppression.This is contrary to the established effects of cortisone dosing on fungal burdens with the pellet diet (23,36,37).However, our results also support the well-established relationship between the extent of immunosuppression and fungal mucosal invasion regardless of the physical form of diet.

The powder diet allows mice to recover from infection in the cortisone OPC model
In the multiple-dose cortisone model of OPC, the mice experience significant weight loss and morbidity after 5 days, which necessitates sacrifice for humane reasons (23).Because the powder diet promoted a severe oral infection with lower levels of immuno suppression, we hypothesized that mice would be able to clear this infection if followed for longer periods of time.To follow the progression of oral infection with the powder diet, mice receiving a single cortisone dose were monitored for up to 7 days.By this time, mice regained their body weight (Fig. 2C) suggesting full recovery.In agreement with the weight recovery data, these mice cleared oral Candida burdens and biofilm lesions within the same time period (Fig. 6).Histopathologic analysis of mucosal tissue changes over time showed that while some papillary atrophy persisted after 5 days of infection, indicative of residual tissue damage, the absence of epithelial desquamation signaled a transition toward tissue restitution and repair, which is critical due to the role of the epithelium as a barrier against microbial invasion (38).By day 7, the filiform papillae and overall stratified epithelial structure had almost completely regenerated, providing further evidence of the ability to recover from Candida infection in this model (Fig. 7).This may be a result of administering a single dose of cortisone which allowed the immune response to rebound earlier in the course of infection.Thus, this infection protocol offers the opportunity to study how homeostasis is re-established, along with pathogen clearance, which is a critical gap in knowledge pertaining to OPC.

Powder diet affects indigenous bacterial communities in infected mice
Different processing of commercial laboratory mouse diets can lead to differences in the gastrointestinal microbiome of C57/BL6 mice [reviewed in references (39,40)].For example, pellet-fed mice contained larger numbers of fecal Lactobacillus species compared to extruded diet (39).Other studies showed that a powder diet substantially decreases the richness of the gut microbiome compared to a pellet diet (18).Observed changes in the composition of the powder diet-fed mouse microbiota included a marked decrease in the abundance of certain Firmicutes accompanied by a bloom of sequen ces assigned to the genus Akkermansia (18).The mechanisms involved in the effects of diet texture on mouse gut bacterial diversity are not clear, although it has been suggested that decreased intestinal transit time, starch gelatinization, and their impact on digestibility and bacterial fermentation may affect gut microbial communities.
The effect of the physical form of diets with identical nutritional profiles on oral bacterial communities has not been studied.To begin to ask this question, we first analyzed the effect of diet texture on total bacterial biomass, as measured by quantifying the 16S rRNA gene amplicons.As seen in Fig. 8A, neither the form of diet nor the amount of immunosuppression had an impact on the tongue bacterial biomass in uninfected mice.Consistent with our previous reports in different models of immuno suppression-associated OPC (14,27), immunosuppressed Candida-infected mice showed an increase in bacterial biomass (Fig. 8B).Candida infection has been associated with oral bacterial dysbiosis, defined as a reduction of bacterial diversity with an overgrowth of bacterial species associated with disease severity.Previous studies from our lab showed that in Candida-infected mice, oral Enterococcus biomass is increased and Lactobacillus biomass is decreased (15,27).To ascertain whether the differences in the severity of oral infection between the powder and pellet diets were related to differences in changes in enterococcal or lactobacillus biomass during infection, we quantified these bacteria with genus-level qPCR.As shown in Fig. 8C and D, Candida-infected mice on a powder diet exhibited a significant increase in Enterococcus biomass and a trend for a reduction in lactobacilli, compared to the pellet diet.Because uninfected mice on the two diets did not differ in these bacterial outcomes, it is more likely that the differences observed in infected mice were a result of the differences in fungal disease severity and not due to the effect of diet texture.As mice recovered from infection, there was a commensurate decline in the bacterial biomass, as shown by both qPCR and viable counts (Fig. 9A).
In agreement with this bacterial biomass decline, we observed a decrease in the oral enterococci biomass over time (Fig. 9B).Collectively these data support the notion that bacterial changes during infection and recovery result from Candida-induced dysbiosis.
In conclusion, our findings highlight the profound influence of diet form in the context of Candida virulence.We also conclude that a powder diet improves the murine OPC cortisone model because it requires lower levels of immunosuppression and opens new avenues for studies investigating not only the pathogenesis but also the recovery processes associated with Candida infection.The mechanisms underlying the increase in oral fungal biomass and OPC severity with a powder diet remain unclear.This highlights the need for a deeper exploration into the multifaceted role of diet in shaping hostmicrobe interactions in the oral cavity, which could pave the way for novel therapeutic strategies and improved disease management.

ACKNOWLEDGMENTS
This study was funded by NIH/NIDCR Grant RO1 DE013986.The funder played no role in the study design, data collection, analysis, and interpretation of data, or writing of this manuscript.R.V.M. contributed to experimental design, conducted experimental work, analyzed, and interpreted the data and was a major contributor to writing the manuscript.A.T. conducted experimental work and analyzed the data.T.S. conducted experimental work and interpreted data.A.D.B. was responsible for the experimental design, data analysis, and interpretation, and was a major contributor to writing the manuscript, and obtained the research funding.All authors read and approved the final manuscript.

FIG 2
FIG 2 Body weight changes in response to diet and Candida infection.(A) Uninfected mice displayed no significant weight variation based on diet physical form.(B) Candida-infected mice on the powdered diet exhibited more significant weight loss compared to the pellet diet.There was no significant difference between one versus two cortisone doses in mice on a powder diet.(C) Candida-infected mice with a single cortisone dose under the powder diet demonstrated a weight recovery trajectory over time.Note: one-way analysis of variance (ANOVA) with Dunnett's T3 test.

FIG 3
FIG 3 Influence of diet and immunosuppression on oral fungal burdens.(A) Uninfected mice.The physical form of the diet did not significantly affect indigenous fungal biomass, although within the powdered diet groups, increasing cortisone doses was inversely associated with fungal biomass.(B) Candida-infected mice.

FIG 4
FIG 4 Candida oral biofilms in relation to diet.(A) The physical form of the diet modulated the ability to form oral biofilms.Left panel: macroscopic tongue images; right panel: Image J quantification of biofilm (white patch) surface area in n = 5 mice per group.(B) Fluorescence staining of biofilm lesions using an FITC-conjugated anti-Candida polyclonal antibody (green, upper panel).Hematoxylin-eosin staining of the same tissue sections (lower panel).The FITC-stained area was quantified in the same groups (right panel).Results show a powder diet to promote Candida mucosal biofilms.Scale bar = 500 µm.Statistical analysis: one-way ANOVA with Dunnett's T3 test.

FIG 6 FIG 7 FIG 8
FIG 6 Candida burdens and oral biofilms over time.(A) Viable counts (black diamonds) and qPCR data (hollow squares) indicated a decline in Candida colonization over time in mice on a powder diet with one cortisone dose, with mice exhibiting recovery from OPC by day 7 post-infection.(B) A temporal decline in fungal biofilm coverage was observed, indicative of recovery from oral Candida colonization.By the seventh day, the biofilms were reduced to 10% of the tongue surface.(C) Tissue sections depicting Candida biofilms with FITC-conjugated anti-Candida antibody (green) confirm that C. albicans presence is rapidly reduced over time.Scale bar = 500 µm.Statistical analysis: one-way ANOVA with Dunnett's T3 test.*P < 0.05 and ****P < 0.0001

FIG 9
FIG 9 Temporal effect of diet and Candida infection on bacterial biomass.(A) Bacterial biomass decreased over time (confirmed by viable counts and qPCR), as mice recovered from Candida infection.By day 7 post-infection, the bacterial biomass returned to the levels of the uninfected mice.(B) Enterococcus biomass (qPCR) declined over time, as mice recovered from the fungal infection.One-way ANOVA with Dunnett's T3 test.*P < 0.05 and **P < 0.01.