Investigating the role of Candida albicans as a universal substrate for oral bacteria using a transcriptomic approach: implications for interkingdom biofilm control?

Candida albicans is frequently identified as a colonizer of the oral cavity in health and has recently been termed a “keystone” commensal due to its role on the bacterial communities. However, the role that C. albicans plays in such interactions is not fully understood. Therefore, this study aimed to identify the relationship between C. albicans and bacteria associated with oral symbiosis and dysbiosis. To do this, we evaluated the ability of C. albicans to support the growth of the aerobic commensal Streptococcus gordonii and the anaerobic pathogens Fusobacterium nucleatum and Porphyromonas gingivalis in the biofilm environment. RNA‐Sequencing with the Illumina platform was then utilized to identify C. albicans gene expression and functional pathways involved during such interactions in dual‐species and a 4‐species biofilm model. Results indicated that C. albicans was capable of supporting growth of all three bacteria, with a significant increase in colony counts of each bacteria in the dual‐species biofilm (p < 0.05). We identified specific functional enrichment of pathways in our 4‐species community as well as transcriptional profiles unique to the F. nucleatum and S. gordonii dual‐species biofilms, indicating a species‐specific effect on C. albicans. Candida‐related hemin acquisition and heat shock protein mediated processes were unique to the organism following co‐culture with anaerobic and aerobic bacteria, respectively, suggestive that such pathways may be feasible options for therapeutic targeting to interfere with these fungal‐bacterial interactions. Targeted antifungal therapy may be considered as an option for biofilm destabilization and treatment of complex communities. Moving forward, we propose that further studies must continue to investigate the role of this fungal organism in the context of the interkingdom nature of oral diseases.

Candida genus has readily been detected in cases of oral candidiasis, with the incidence increasing in patients undergoing cancer therapy [10,11].
Most infections of the oral cavity seldom arise from mono-species colonization.As such, C. albicans has been co-isolated alongside oral microbial communities in biofilm-related diseases such as denture stomatitis [12], dental caries [13], periodontitis [14], and angular chelitis [15,16].Such interactions are largely beneficial for a variety of commensal and pathogenic bacterial species, with the fungi providing a mechanical support structure for adhesion and colonization: a fungal yeast cell biovolume is almost 150-fold larger (~70 lM) than an average bacterial cell (~0.5 lM), thereby providing an ideal substrata for bacterial attachment [17].Furthermore, it is proposed that the surface area of a hyphal cell is approximately 2-fold greater than that of a yeast cell, offering an even greater landscape for bacterial colonization [18].As such, cellcell interactions between C. albicans and bacterial species associated with the above diseases such as Streptococcus mutans, Porphyromonas gingivalis and Staphlyococcus aureus have all been reported in the literature [19][20][21].For example there is evidence to suggest that S. mutans and P. gingivalis all actively bind to C. albicans through expression of adhesions such as GtfB [22,23] and InlJ [24], respectively.Physical interactions between C. albicans and S. aureus also exist, with the bacteria adhering to the fungal hyphae to facilitate tissue invasion and dissemination during biofilm infection [25][26][27].
Beyond physical adhesion mechanisms, Candidabacterial interactions also result in important mutualistic relationships that are critical for key biological processes in the microbial community at a transcriptional, metabolite and protein level, as well as driving physiological changes in the immediate microenvironment of the cells.For example, GftB secretion by S. mutans drives carbohydrate breakdown providing a carbon source for C. albicans [22], while also upregulating expression of key virulence and adhesion genes such as HWP1, ALS1 and ALS3 in the fungus [28].Conversely, C. albicans can further support S. mutans growth aside from its adhesive capabilities by upregulating gene pathways related to sugar metabolism [29] and driving increased bacterial growth independent of cell-cell contact through metabolite production [30].Other bacterial species of the oral cavity also possess similar symbiotic associations with C. albicans beyond direct cellular interactions.For example, Streptococcus gordonii, an oral commensal bacteria, and C. albicans develop biofilms together through a combination of both physical interactions and production of chemical or protein "signals" such as quorum sensing molecules and secretory proteases [31][32][33].Similar pathways are present in interactions between P. gingivalis and C. albicans: Direct contact between these two microorganisms resulted in P. gingivalis survival in oxic conditions that would be otherwise lethal to the bacterium, likely arising from a dense hyphal network that resulted in O 2 depletion by C. albicans [34].Such protective "shielding" in these dualspecies communities can also promote P. gingivalis invasion of the host via immune evasion [35].
Due to all these interactions discussed above, it has recently been proposed that C. albicans acts as a "keystone" commensal [17].This can be explained by the organism being a minority with respect to the overall bioburden (~0.1% of the total microbial load), but still providing an important physical scaffold and influencing the local microenvironment within the biofilm bringing changes to micronutrient availability, oxygen consumption, and pH buffering [17].We have confirmed that such Candida-bacterial interactions are present in more complex biofilm communities, demonstrating that C. albicans enhances bacterial colonization in a 6-and 11-species soft and hard oral tissue models, while buffering the pH of the spent media [36].However, the effects that such mixed communities have on Candida at a cellular level are largely unknown.
Taken together, the aforementioned interplay between fungi and bacterial species in the oral cavity presents such interactions as a possible avenue for therapeutic target.Given the potent role that fungal microorganisms such as C. albicans play in providing a substrata for bacterial attachment, it could be postulated that destruction of this microbial scaffold could have huge implications in the pathogenesis of various oral diseases.This could be achieved through directing treatment against the fungal element of these biofilms, using antifungal therapies or targeting specific adhesion molecules [37].In recent years, the development of molecular techniques such as CRISPR-Cas9 technologies and increased availability of omics approaches for indepth analyses has made targeting specific genetic pathways and/or biological processes a feasible, yet exploitable, treatment strategy [38,39].However, the role that bacteria play in regulating such pathways in the fungal species is understudied.
The purpose of this study was to investigate the transcriptional response in C. albicans following coculture with three orally relevant bacterial species when grown as a dual-species or a 4-species biofilm model using a transcriptomic approach.We envisage that such analysis may highlight gene pathways at work arising from such interactions, with the hope of identifying key "genes" that could be targeted to remove this "keystone" commensal, and thus preventing such challenging fungal-bacterial interkingdom interactions.

Microbial culture
Yeast and bacterial cultures were prepared according to our previously standardized methods [40].Briefly, C. albicans SC5314 type strain was first cultured on solid agar (Sabouraud; SAB) for 24-48 h, then propagated overnight in Yeast Peptone Dextrose liquid culture at 30 °C.Cells were then retrieved by centrifugation for 5 min and washed twice with phosphate buffered saline (PBS) before being standardized to 1 9 10 7 cells/mL by counting via hemocytometer.Bacterial cultures for P. gingivalis ATCC 33277 and F. nucleatum ATCC 10953 were prepared anaerobically firstly in Fastidious Anaerobic Agar (FAA) for 48-72 h and then sub-cultured in Schaedlers broth for 24 h.S. gordonii ATCC 35105 was initially grown on Columbia Blood Agar (CBA) then propagated in Tryptic Soy Broth in 5% CO 2 overnight.All bacterial cultures were centrifuged for 5 min and washed twice with PBS before being standardized by spectrophotometry at 550 nm.Bacteria were standardized to 1 9 10 8 cells/ml at an OD of 0.2 for P. gingivalis and F. nucleatum and 0.5 for S. gordonii, respectively.Yeast and bacteria were used for biofilm growth as described below following preparation of standardized cells.

Biofilm preparation
All biofilms were prepared in 1:1 media comprised of Roswell Park Memorial Institute (RPMI-1640) and Todd Hewitt Broth (THB) supplemented with 0.01 mg/mL hemin and 2 lg/mL menadione to support the growth of bacterial and fungal biofilms as previously described [40,41].For colony counting and compositional analysis, all biofilms were prepared in 24-well polystyrene plates at an initial concentration of 1 9 10 7 cells/mL for bacteria and 1 9 10 6 cells/mL for yeast.For the dual-species or 4species model, bacteria were added following 4 h of C. albicans biofilm growth.Biofilms that were grown with a preformed 4-h Candida biofilm followed by the introduction of either single or all three bacteria were also replenished with 1:1 media containing the standardized bacteria before being grown for a further 20 h.For RNAsequencing experiments, biofilms were grown in 75 cm 2 cell culture flasks to provide enough bioburden for RNA extraction.All biofilm models were grown for a total of 24 h in 5% CO 2 at 37 °C.

Colony counting and composition analysis
Colony counts and/or biofilm compositional analysis for single, dual, or multi-species biofilms prepared by initial growth of Candida for 4 h followed by 20 h with bacteria as described above was performed using validated methods for Colony Forming Equivalents (CFE) and Colony Forming Units (CFU) [42].Conditions remained the same as previously stated with the amendment of the growth of biofilms on Thermanox TM coverslips allowing for the removal of coverslips and sonication for 15 min in 1 mL of PBS at 35 hz.For CFU counting, the sonicate was serially diluted in 1/10 dilutions before each dilution was plated on SAB (for Candida counts) and FAA (P.gingivalis or F. nucleatum) or CBA (S. gordonii) according to the Miles and Misra method [43].
For CFE analysis, propidium monoazide (PMA) was added to 500 lL of the sonicate in PBS and incubated in the dark for 10 min prior to treatment of the sample with a 650 W halogen light.DNA extraction was performed by QIAamp DNA mini kit according the manufacturers protocol.Real-time qPCR was then used to determine the live and total cell number.Each reaction mixture contained 10 lL of Fast SYBRÒ Green Master Mix, 1 lL of each forward and reverse species primers at 10 lM, 1 lL of either PMA treated or untreated DNA, and 7 lL nuclease free water.PCR cycles were performed on the Step-One plus machine using the following stages 50 °C for 2 min, 95 °C for 2 min, followed by 40 cycles of 95 °C for 3 s and 60 °C for 30 s. CFEs were then calculated using standard curves for each species using serially diluted DNA.All primers used were as previously described [40].

RNA-sequencing
Following biofilm growth, the supernatant was removed from the biofilms before being stored in 1 mL of RNAlater at À80 °C.RNA was extracted from all single species fungal, dual and multi-bacterial species biofilms by the RiboPure TM Purification kit according the manufacturers guidance.RNA quantity was determined using the Nano-Drop spectrophotometer ND-1000 with a total of 2.5 lg deemed adequate for sequencing, and RNA integrity number (RIN) was determined using the Bioanalyser with a RIN of > 7 being necessary for sequencing.Candida transcripts were prepared by PolyA selection and sequenced on the Illumina NovaSeq6000 at the Edinburgh Genomics sequencing facility (http://genomics.ed.ac.uk/).
Paired end 100 bp reads were retrieved from the Illumina sequencing performed by Edinburgh Genomics.Preprocessing was performed on Raw data using our in house pipeline.The reads were first quality controlled by the removal of low quality sequences and removal of adapters using Trimmomatic (v0.38) [44].A haploid version of the Candida genome database (CGD) genome (SC5314_A22) was prepared by removal of the B variants of the chromosomes before the resulting fasta file was indexed using Hisat2 (v2.1.0)[45,46].Hisat2 was then again used to align the remaining trimmed reads to the haploid genome.The aligned SAM files were converted to BAM compressed format using samtools (v1.7) before HTSeq-count (v0.11.0) was implemented to count the occurrence of each transcript [45].HTSeq-count files for each sample were compiled and imported into R using DESeq2 (v1.26) [47].The counts were compiled into one data frame before differential analysis was performed using DESeq2.Subsequently, the differentially expressed genes between our single species fungal biofilm and interkingdom biofilms with a log 2 FC > 1.5 and FDR adjusted p-value < 0.05 were selected for further analysis.Differentially expressed transcripts between each condition were represented in an UpsetPlot using UpsetR (v1.4.0) which displays data in a Venn diagram like manner for larger numbers of samples.

603
PROFILING THE CANDIDA ALBICANS TRANSCRIPTOME Upregulated Candida genes in the dual-or multi-species biofilms compared to Candida only biofilms were compared to find unique and overlapping genes between the dual-and multi-species biofilms.Over representation analysis (ORA) was performed using the hypergeometric analysis in the R package ClusterProfiler (v3.14.3) for the unique and overlapping genes in each comparison based upon the C. albicans Gene Ontological terms.

Data presentation and statistical analysis
RNA-sequencing data were plotted in R using a combination of the packages listed above, base R and ggplot2 (v.3.3.0).Graph generation related to the CFU-and CFEderived data was achieved in GraphPad Prism 9. When it was pertinent a t-test was performed to compare between the CFU counts in the biofilms grown as mono-or dual-species models.Statistical significance was achieved if p < 0.05.

Candida as a scaffold
Initially, we sought to determine the capability of C. albicans in supporting the growth of oral bacterial species in a biofilm.We enumerated the viable Candida and bacterial cells for P. gingivalis, S. gordonii, and F. nucleatum grown alone for 24 h in a single species biofilm and also grown in the presence of a pre-established C. albicans biofilm.We hypothesized that the preformed Candida biofilm would behave as a scaffold and provide an environment that promotes the growth of the oral bacterial species.Candida counts remained unchanged following co-culture with the bacterial species (Fig. 1A).Conversely, there was an increase in bacterial CFUs for all three bacteria in the biofilm model with C. albicans (Fig. 1B).This is indicative that C. albicans supports or promotes the growth of oral pathogens in the biofilm; however, bacteria do not provide the same impact on C. albicans.There is a particularly pronounced effect on P. gingivalis, as there is little to no biofilm formation, on average, in the P. gingivalis single species biofilms, however, there is consistently greater than 1 9 10 6 viable cells/mL in the dualspecies biofilm.This was found to be a significant increase (p = 0.0068; Fig. 1B).Compositionally for the multi-species biofilms we see that there was an increased competition from the S. gordonii and F. nucleatum resulting in reduced numbers of P. gingivalis by ~1 log compared to the dual species whereas the CFE numbers of S. gordonii and F. nucleatum remained consistent in the different models (Figure S1).Nevertheless, viability of the bacterial strains used remained high when grown in dual-and 4-species biofilms.

Transcriptional response of C. albicans to oral pathogens
From our RNA-Seq data, we discerned the upregulated and downregulated C. albicans genes in response to dual-and multi-species biofilms.When comparing the effect of bacteria on Candida transcription, we observed the most pronounced changes in the dual-species biofilms containing S. gordonii and the multi-species containing all three bacterial species and C. albicans (Fig. 2).It is also noteworthy that the multi-species and S. gordonii dual-species biofilms had the largest overlap in expressed genes.There were 36 upregulated genes shared between the two biofilm models (Fig. 2A).We had already noted that S. gordonii was the most prevalent member of the multi-species models (Figure S1), while here it also stands apart as the dominant driver of the C. albicans transcriptional response.However, in each of our comparisons there was a species-specific response from the C. albicans, with 25, 20, and 4 genes being unique to the S. gordonii, F. nucleatum, and P. gingivalis dual-species biofilm models, respectively (Fig. 2B).It is apparent from the UpSet plot and volvano plots that S. gordonii exhibits the largest response from C. albicans and P. gingivalis the least (Fig. 2).Additionally, there were 20 unique transcripts from the multi-species biofilm indicating an influence on the C. albicans transcript which is unique to the combined impact of all three bacterial species.
Next, we performed Gene Ontology (GO) Over Representation Analysis (ORA) on the unique and overlapping genes in each of our dual-species biofilms compared to the multi-species.All the genes included in the ORA were the upregulated genes in each of the conditions compared to the C. albicans biofilm only control.This approach was chosen to elucidate and separate the species-specific interactions compared to those from the complex biofilm Fig. 2. Differential expression of C. albicans genes in response to co-culture with bacteria.UpSet plot depicting the overlap of upregulated C. albicans transcripts in dual-and multi-species biofilms of P. gingivalis, S. gordonii, F. nucleatum or all three species compared to Candida only biofilms.Biofilms were all mature biofilms grown for a total of 24 h.Within the UpSet plot bars depict the number of overlapping upregulated genes between each of the interkingdom biofilms, depicted by the cross sectional blue lines (A).Volcano plots depict the individual upregulated genes in each of the fungal vs interkingdom biofilms following co-culture (B).Genes with a log 2 FC greater than 1.5 (green) and a p-value < 0.01 (red) are shown within the plot.The inverted log 10 p-value is depicted on the y-axis along with the Log 2 FC on the x-axis.The top most prominent features are also overlayed with gene symbols or CGD identifiers.containing all bacterial consortia.The C. albicans-P.gingivalis biofilms presented only 11 upregulated genes in comparison with the control and 9 of these were not present in the multi-species biofilm.Very few of the upregulated genes were over represented in functional pathways, and those that were contained very few genes within the network (Fig. 3).There was no significantly enriched pathways for the 9 unique P. gingivalis genes and only 2 genes that were significant in the overlapping pathways.The over represented terms included genes involved in heme binding and tetrapyrrole binding networks from C. albicans in response to P. gingivalis.
Fusobacterium nucleatum exhibited a more pronounced response from C. albicans in the dualspecies biofilms with significant enrichment (FDR p < 0.05) in the 3 oxidoreductase pathways.The overlapping genes within the multi-species biofilm included the same two genes involved in iron acquisition that were expressed by C. albicans in response to P. gingivalis (Fig. 4).This is indicative of a shared interaction with C. albicans by the two anaerobes in dual-species and multi-species conditions in relation to potential heme acquisition and/ or iron metabolism.
Finally when comparing the unique and shared genes in the S. gordonii dual species compared to the 4 species model, we observed the largest overlap of upregulated genes.It is apparent that S. gordonii produces a much larger transcriptional response from C. albicans than the two anaerobic pathogens.Within the comparison in Fig. 5, there are significantly over represented pathways that are comprised of 38 overlapping genes.Notably, many of these overlapping pathways are between S. gordonii dual-species and multi-species biofilms, which are largely related to protein processing related processes, heat shock, and toxic substance response pathways (Figs 3 and 4).Taken together, our commensal S. gordonii was the largest driver of differential gene expression in the multi-species biofilm.The unique S. gordonii-derived transcriptome contained 39 genes which were specifically related to heat shock protein binding.There were also 24 genes that were only expressed in the multi-species biofilm when compared directly to the S. gordonii dual-species.These genes were over represented in the GO terms involved in amino acid biosynthesis.These are likely genes that are in response specifically to the combined effects of all three microorganisms when cultured with C. albicans.

DISCUSSION
Increasing evidence in the literature suggests that the opportunistic pathogen, C. albicans, may play important, yet largely undefined roles in Fig. 3. Comparative analysis of the enriched Gene Ontology terms in multi-species compared to P. gingivalis dual-species biofilms.Upregulated genes in dual or multi-species biofilms after 24 h were selected based upon a log 2 FC > 1.5 and a FDR corrected p-value < 0.05.The unique and overlapping genes in the dual-species biofilm (with P. gingivalis) vs the multi (4-species) were then compared and depicted in a Venn Diagram.The unique and overlapping genes then were included in an over representation analysis to determine enriched functional pathways.The dot plots depict significantly over represented pathways from each of the gene subsets and the overall gene ratio of the pathway is shown on the x-axis, the weight of the dots is proportional to the number of genes within the analysis and the color is indicative of the FDR adjusted p-value from the hypergeometric testing.pathogenesis of oral diseases such as dental caries and periodontitis [13,14].Conversely, due to the presence of Candida spp. in the oral cavity of healthy individuals [4,5], C. albicans has been proposed as a "keystone" commensal pertaining to its role as a microbial scaffold for bacterial species and the ability of the organism to alter the local microenvironment to support bacterial growth [17].However, studies investigating the effects that bacteria have on the fungus are limited.Thus to further investigate this "keystone" commensal phenomena, the role of C. albicans as a microbial scaffold to three oral bacterial species was first confirmed.Following this, the transcriptomic response in the fungal organism was assessed when grown as monospecies biofilms, or grown as dual-species and 4-species biofilms with the oral bacteria, S. gordonii, F. nucleatum, and P. gingivalis.Results from these RNA-sequencing experiments are suggestive that the fungal response was species-dependent, with transcriptional regulation of C. albicans dictated by the presence of the bacterial species, and predominantly the oral commensal, S. gordonii.
Initial microbiological results showed that C. albicans provided a substrata for bacterial attachment, increasing cell counts for all three bacterial species when grown as dual-species biofilms compared to grown alone.This is in line with previous observations, emphasizing the idea that C. albicans is a structural scaffold to the oral bacterial species, S. gordonii and P. gingivalis [31,35].Interestingly, results for the C. albicans-F.nucleatum co-incubation showed that numbers of the bacteria increased when cultured with the fungi which is in contrast with observations elsewhere [48].Bor et al. (2016) showed that although C. albicans adhered with F. nucleatum, aggregation was detrimental to growth and hyphal morphogenesis for the fungus.This difference between these observations might be explained by experimental design, with C. albicans added first here for 4 h, allowing the organism to form hyphae before addition of the bacteria for a further 20 h.Alternatively, as Candida biofilms display a degree of heterogeneity, strain variability might explain the differences in results between the studies [49].The presence of C. albicans was also essential to support growth of P. gingivalis in the 5% CO 2 conditions.As an obligate anaerobe, P. gingivalis cannot survive in oxygen levels of above 6% unless grown in excess hemin [50,51].Thus to survive, P. gingivalis must attach to other colonizers of the oral cavity Fig. 4. Comparative analysis of the enriched Gene Ontology terms in multi-species compared to F. nucleatum dual-species biofilms.Upregulated genes in dual or multi-species biofilms after 24 h were selected based upon a log 2 FC > 1.5 and a FDR corrected p-value < 0.05.The unique and overlapping genes in the dual-species biofilm (with F. nucleatum) vs the multi (4-species) were then compared and depicted in a Venn Diagram.The unique and overlapping genes then were included in an over representation analysis to determine enriched functional pathways.The dot plots depict significantly over represented pathways from each of the gene subsets, and the overall gene ratio of the pathway is shown on the xaxis, the weight of the dots is proportional to the number of genes within the analysis and the color is indicative of the FDR adjusted p-value from the hypergeometric testing.There were no significantly overrepresented pathways for the dual species only pathways.
PROFILING THE CANDIDA ALBICANS TRANSCRIPTOME such as Enterococcus faecalis and F. nucleatum to persist in oxic conditions [52][53][54].In line with previous observations, results here show that P. gingivalis can use C. albicans as a substrata to survive oxygenated conditions [34].Although we observed that Candida was able to support the growth of our obligate anaerobes in 5% CO 2 in our dual-species biofilm models, within the 4species model S. gordonii was the most competitive member.This may in part be due to these conditions favouring S. gordonii growth thus highlighting a potential limitation of this study.Future work may warrant investigations into other conditions (e.g., an anaerobic microenvironment) when growing these biofilm models to further support F. nucleatum and P. gingivalis growth.However such investigations could be influenced by the inability of C. albicans to form biofilms under such anaerobiosis, while depleted O 2 conditions can also interfere with hyphal formation [55,56].Careful considerations are required for such future work.
Previous studies have shown that S. gordonii interactions with C. albicans are largely synergistic in nature [31][32][33].One study by Bamford et al (2009) reported that contact-dependent interactions between the bacterial cells and Candida hyphae were necessary for mixed-species biofilm development.As well as modulating C. albicans biofilm formation S. gordonii has also been shown to co-aggregate and interact with P. gingivalis and F. nucleatum within the oral cavity [57,58], and therefore was selected for incorporation into our mixed species model.Whether such transcriptional responses are similar in C. albicans when cultured alongside other oral Streptococcus species such as S. oralis, would be an interesting follow-up study.Here, S. gordonii cell counts were increased in the presence of C. albicans, although assessment of hyphal formation and cellcell interactions were not pursued in this study.Nevertheless, results from our RNA-sequencing study highlighted an increase in expression of protein processing machinery (e.g., chaperone binding) and heat shock pathways (e.g., HSP70 and HSP78 in dual species, and HSP78 in the 4-species model) in C. albicans when grown alongside bacterial species.Heat shock proteins have previously been linked to biofilm formation and hyphal development in the organism [59,60].Additionally, the chaperone network involving heat shock proteins have been shown to play a role in drug resistance, morphogenesis, and virulence [61].Here, increases in Candida counts were not observed between monoand dual-species or 4-species biofilms, suggestive that bacterial presence may simply be driving Fig. 5. Comparative analysis of the enriched Gene Ontology terms in multi-species compared to S. gordonii dual-species biofilms.Upregulated genes in dual-or multi-species biofilms after 24 h were selected based upon a log 2 FC > 1.5 and a FDR corrected p-value < 0.05.The unique and overlapping genes in the dual-species biofilm (with S. gordonii) vs the multi (4-species) were then compared and depicted in a Venn Diagram.The unique and overlapping genes then were included in an over representation analysis to determine enriched functional pathways.The dot plots depict significantly over represented pathways from each of the gene subsets and the overall gene ratio of the pathway is shown on the x-axis, the weight of the dots is proportional to the number of genes within the analysis and the color is indicative of the FDR adjusted pvalue from the hypergeometric testing.
hyphal elongation or other biological processes which was not investigated.Nevertheless, these results highlight that heat shock proteins may be a feasible target not only for C. albicans virulence pathways (reviewed in [62]) but also key cell-cell interactions with bacterial communities of the oral cavity.
When investigating the transcriptome of C. albicans grown in mixed-species biofilms with F. nucleatum and P. gingivalis, unique pathways relating to heme binding were identified in the fungal organism.For example, there was an upregulation of the gene responsible for the extracellular heme binding protein CSA2 in dual-species biofilms and the complex 4-species model.This hemophore can sequester heme and subsequently iron from human hemoglobin [63].The ability of C. albicans to utilize heme as an iron source is documented in the literature [64,65], with hemin uptake linked with hyphal formation in the fungi [66].Further work would need to be performed to elucidate the heme-related interaction between Candida and the anaerobes.P. gingivalis has the ability to accumulate heme on the surface of the cell, giving the organism its black pigmentation [67]; thus, the upregulation of the candidal hemophore may relate to an increased heme availability in adjacent or adhered cells.Alternatively, it could be postulated that this is a stress based response to anaerobic bacteria to negate the "competition" for micronutrients.To this end, P. gingivalis has been shown to upregulate expression of HmuR, HmuY, and HusA, three key proteins responsible for heme acquisition systems, when grown alongside C. albicans [68].Nevertheless, the mutualistic relationship of C. albicans-bacterial interactions merits much further consideration.
A number of genes relating to oxidative stress pathways were upregulated in C. albicans grown alongside F. nucleatum.The ability of F. nucleatum to produce reactive oxygen species (ROS) is well known [69,70] and has been linked to the progression of oral cancers through this mechanism.Within the dual-species biofilm of Candida and F. nucleatum, we observed a marked upregulation of the oxidoreductase based processes including genes, OYE2, GST1, and EBP1.Interestingly, similar genes have previously been shown to be upregulated in C. albicans during DNA damage response [71].Given that ROS are potent inducers of DNA damage in yeast [72], Candida may be undergoing protective measures to prevent F. nucleatum-mediated damage.Such mechanisms might explain previous antagonistic relationships between the two organisms, even in the presence of cell-cell attachment [48].Also upregulated were the NAD family of genes related to mitochondrial respiration.Superoxide scavengers such as NADH are known to reduce oxidative stress in eukaryotes such as C. albicans with the mitochondria being central to this antioxidant response [73].This oxidoreductase response is only apparent in the dual species and not the multispecies model suggestive this phenomena is potentially being buffered or otherwise controlled by the other bacterial communities.Alternatively, this could be a specific C. albicans-F.nucleatum interaction, the dynamics of which are unclear.It is evident that further interrogation is necessary to decipher this interaction.A further limitation of our results is that it is unknown whether there is a conserved response with other Candida species, or if it is specific to C. albicans; to this end, there is evidence that there are shared pathways between the species, such as the oxidoreductase and iron acquisition systems (e.g., CSA2) in other Candida spp.such as Candida glabrata [74].

CONCLUSIONS
In conclusion, we have documented an in-depth transcriptomic study looking at the interactions between C. albicans and three orally-relevant bacterial species.Results are indicative that although Candida provides a structural and supportive substrata for bacteria to attach, critical for anaerobes such as P. gingivalis to survive in oxic conditions, the transcriptional profile of the fungal organism can vary in a bacterial species-dependent manner.Admittedly, these results may raise more questions than they answer, but highlights that the ideology of "one size fits all" is not applicable to such fungal-bacterial interactions.
We deem it pertinent for future studies to continue to consider the role that this fungus plays in the dynamics of the oral microbiome.Given the reported presence of C. albicans in the oral cavity of healthy and diseased individuals, and its potential role as a "keystone" commensal, we postulate that the organism could be a potent target for therapeutic treatment to control interkingdom oral diseases.

Ó
2023 The Authors.APMIS published by John Wiley & Sons Ltd on behalf of Scandinavian Societies for Pathology, Medical Microbiology and Immunology.

Fig. 1 .
Fig. 1.Microbial analysis from dual-species biofilms.Colony forming units per ml of C. albicans (A) or P. gingivalis, S. gordonii, and F. nucleatum (B).Comparison is made from growth of mono-species biofilms of C. albicans only or bacterial only biofilms compared to dual-species biofilms formed by initial growth of C. albicans for 4 h followed then by 20 h with each bacterial species.T-tests were performed between dual and single species biofilms for each comparison (**p < 0.01, ****p < 0.0001).All individual points representative of biofilms generated from three technical replicates and three biological replicates (n = 9).A total of 27 points are visible for C. albicans only biofilms as these have been collated from individual controls (e.g., 9) for each bacterial co-culture experiment.No significant differences were observed between the C. albicans mono-species biofilms compared to each dual-species biofilms with P. gingivalis, F. nucleatum, and S. gordonii.