Catalase produced by Candida albicans protects Streptococcus mutans from H2O2 stress—one more piece in the cross-kingdom synergism puzzle

ABSTRACT Co-infection with Streptococcus mutans and Candida albicans is associated with dental caries, and their co-cultivation results in enhanced biofilm matrix production that contributes to increased virulence and caries risk. Moreover, the catalase-negative S. mutans demonstrates increased oxidative stress tolerance when co-cultivated in biofilms with C. albicans, a catalase-producing yeast. Here, we sought to obtain mechanistic insights into the increased H2O2 tolerance of S. mutans when co-cultivated with clinical isolates of Candida glabrata, Candida tropicalis, and C. albicans. Additionally, the C. albicans SC5314 laboratory strain, its catalase mutant (SC5314Δcat1), and S. mutans UA159 and its glucosyltransferase B/C mutant (UA159ΔgtfB/C) were grown as single- and dual-species biofilms. Time-kill assays revealed that upon acute H2O2 challenge, the survival rates of S. mutans in dual-species biofilms with the clinical isolates and C. albicans SC5314 were greater than when paired with SC5314Δcat1 or as a single-species biofilm. Importantly, this protection was independent of glucan production through S. mutans GtfB/C. Transwell assays and treatment with H2O2-pre-stimulated C. albicans SC5314 supernatant revealed that this protection is contact-dependent. Biofilm stability assays with sublethal H2O2 or peroxigenic Streptococcus A12 challenge resulted in biomass reduction of single-species S. mutans UA159 and dual-species with SC5314Δcat1 biofilms compared to UA159 biofilms co-cultured with C. albicans SC5314. S. mutans oxidative stress genes were upregulated in single-species biofilms when exposed to H2O2, but not when S. mutans was co-cultivated with C. albicans SC5314. Here, we uncovered a novel, contact-dependent, synergistic interaction in which the catalase of C. albicans protects S. mutans against H2O2. IMPORTANCE It is well established that co-infection with the gram-positive caries-associated bacterium Streptococcus mutans and the yeast pathobiont Candida albicans results in aggressive forms of caries in humans and animal models. Together, these microorganisms form robust biofilms through enhanced production of extracellular polysaccharide matrix. Further, co-habitation in a biofilm community appears to enhance these microbes’ tolerance to environmental stressors. Here, we show that catalase produced by C. albicans protects S. mutans from H2O2 stress in a biofilm matrix-independent manner. Our findings uncovered a novel synergistic trait between these two microorganisms that could be further exploited for dental caries prevention and control.

. The development of dental caries is intimately associated with the high consump tion of fermentable carbohydrates like sucrose (table sugar) that favor the growth of acid-producing (acidogenic) and acid-tolerant (aciduric) microorganisms.Among them, Streptococcus mutans is considered an important pathogen in dental caries due to its high acidogenicity, aciduricity, and ability to form robust biofilms when sucrose is present (3).Candida albicans, a fungal pathobiont, is often co-isolated with S. mutans from carious sites in patients with early childhood caries (ECC), dentinal caries, and root caries (4)(5)(6)(7).Moreover, co-infection with S. mutans and C. albicans has been associated with recalcitrant oral infections that often result in rampant caries and increased caries reoccurrence in children after clinical interventions (4,8).In recurring ECC, 80% of children were co-infected with S. mutans or C. albicans, whereas 68% of the children were co-infected with both species.Importantly, less than 10% of the caries-free children were co-infected with S. mutans and C. albicans (6).Furthermore, in rats fed a cariogenic diet, co-infection with S. mutans and C. albicans resulted in more aggressive caries formation than single-species infections (9,10).This increased virulence was attributed to the upregulation of the glucosyltransferases of S. mutans by C. albicans and the ability of glucosyltransferase B (GtfB) to bind to C. albicans surface, de facto enhancing both species' glucan production (11,12).Importantly, C. albicans possesses several traits that make it a desirable partner for S. mutans, including its highly aciduric nature, ability to utilize lactate as a source of energy, secretion of bacterial growth factors, and ability to reduce the oxygen tension in the biofilm (13,14).
In the oral cavity, hydrogen peroxide (H 2 O 2 ) is an important form of reactive oxygen species (ROS) generated as a by-product of microbial metabolism or delivered through oral hygiene products.In fact, H 2 O 2 has been used intraorally for its antimicrobial properties for over a century (15)(16)(17).Importantly, net H 2 O 2 production by commensal oral streptococci is associated with health (18)(19)(20), as S. mutans is susceptible to peroxide stress (3,21,22).Previous studies have demonstrated that the ability of commensal streptococci to effectively compete with S. mutans is compromised when their H 2 O 2 -pro ducing pathways are disrupted, or following catalase addition, which degrades H 2 O 2 (18,23).Moreover, a clinical study showed low salivary H 2 O 2 levels correlate with increased ECC risk (24).Of note, C. albicans is highly H 2 O 2 tolerant, and while co-cultivation with S. mutans improves bacterial survival of H 2 O 2 , fungal survival does not increase (12,25).C. albicans H 2 O 2 tolerance is primarily dependent on the production of a heme-con taining catalase (Cat1), predicted to be localized to multiple intracellular compartments, including the cytosol, mitochondria, and peroxisomes, which is conserved across the genus and whose expression is upregulated in mixed biofilms with S. mutans (26)(27)(28).Contrastingly, S. mutans lacks catalase and instead relies on other ROS scavenging systems to cope with fluctuations in ROS levels (29).Thus, we hypothesized that C. albicans' catalase confers a competitive advantage to S. mutans against peroxigenic streptococci, favoring a dysbiotic cariogenic biofilm.Here, we investigated the role that C. albicans' catalase plays in the increased tolerance of S. mutans against oxidative stress in dual-species biofilms.We showed that in dual-species biofilms, the catalase produced by C. albicans protects S. mutans from H 2 O 2 stress in a contact-dependent manner.

Clinical strains of C. albicans, C. glabrata, and C. tropicalis and the laboratory strain of C. albicans SC5314 protect clinical co-isolates of S. mutans and the laboratory strain UA159 against acute oxidative stress in dual-species biofilms
Here, we explored whether the reported oxidative stress protection of S. mutans by C. albicans is shared among clinical strains of human-associated Candida species and S. mutans (12).For this, we utilized a panel of paired S. mutans and Candida co-isolates from plaque samples of ECC patients, including three co-isolates of S. mutans and C. albicans (SMP1 & CAP1, SMP2 & CAP2, and SMP3 & CAP3), one co-isolate of S. mutans and C. tropicalis (SMP4 & CTP4), and one co-isolate of S. mutans and C. glabrata (SM P5 & CG5) (Fig. 1).Biofilms were exposed to 0.25% H 2 O 2 (73.5 mM) for up to 90 min and plated for viable CFU determination.In all cases, S. mutans clinical isolates demonstrated dramatically enhanced survival in dual-species biofilms with their Candida clinical co-isolates compared to single-species biofilms (P < 0.001) (Fig. 1A).Also, for Candida clinical isolates, compared to single-species biofilms, no changes in survival were observed when they were in dual-species biofilms with S. mutans (Fig. 1B).When the lab strains of C. albicans SC5314 and S. mutans UA159 were co-cultivated in biofilms, we confirmed the previous finding that UA159 had enhanced survival when grown as dual-species biofilms with SC5314 as compared to single-species biofilms (P < 0.001) (12,25).However, SC5314 survival does not appear to benefit from UA159 (Fig. 1).

C. albicans' catalase protects S. mutans against acute oxidative stress in dual-species biofilms
After verifying that Candida's protection of S. mutans is a common trait among humanassociated species and is strain-independent, we continued our studies with the lab strains of S. mutans (UA159) and C. albicans (SC5314), as mutants of our genes of interest were available.First, we utilized a C. albicans catalase mutant (SC5314Δcat1) and an S. mutans glucosyltransferase B/C-deficient strain (UA159∆gtfB/C) to determine whether inactivation of these genes impacts their ability to form biofilms.Dual-species biofilms of UA159 with either C. albicans strain (SC5314 or SC5314Δcat1) showed increased biomass in relation to UA159 single-species biofilms (P < 0.01) (Fig. S1).As expected, the UA159∆gtfB/C mutant, a strain with impaired ability to produce extracellular polysacchar ide (EPS) matrix, formed poor biofilms when compared to the parent strain UA159 in single-species biofilms (P < 0.001).Also, no enhancement of biofilm biomass was observed when UA159∆gtfB/C was co-cultivated with SC5314 or SC5314Δcat1 (P > 0.05).Of note, when compared to C. albicans SC5314, the C. albicans catalase-mutant strain SC5314Δcat1 showed no impairment in biofilm formation (Fig. S1).
After determining that inactivation of C. albicans' catalase did not affect biofilm biomass in either single-species or dual-species biofilms with S. mutans, mature singleand dual-species biofilms were exposed to 0.25% H 2 O 2 for up to 90 min and plated for CFU estimation.We found that the UA159ΔgtfB/C strain behaved similarly to UA159 in mono-species biofilms when challenged with H 2 O 2 (P > 0.95) (Fig. 2).S. mutans strains (UA159 and UA159ΔgtfB/C) were afforded significant protection when grown as dual-species biofilms with SC5314 as compared to single-species biofilms (P < 0.0001).Further, UA159-SC5314 and UA159ΔgtfB/C-SC5314 biofilms did not differ from each other upon H 2 O 2 challenge (P > 0.95) (Fig. 2A), indicating that robust EPS matrix production does not protect S. mutans from lethal doses of H 2 O 2 .However, C. albicans SC5314Δcat1 failed to provide this same protective effect, indicating that C. albicans' catalase is a major player in the protection of S. mutans against H 2 O 2 .While C. albicans SC5314 displayed a high tolerance to H 2 O 2 that was not impacted by co-cultivation with S. mutans (Fig. 2B), the SC5314Δcat1 mutant was always highly sensitive to H 2 O 2 when compared to the parent SC5314 strain (Fig. 2B) (P < 0.01), as previously reported (30,31).To confirm that the C. albicans catalase mutant did not have other mutations that could affect its tolerance to H 2 O 2 stress, a second catalase mutant was independently constructed (SC5314Δcat1#2) and used to demonstrate that the phenotypes observed are solely due to the lack of a functional catalase (Fig. S2).

S. mutans oxidative stress tolerance is contact-dependent
Though C. albicans' catalase is predicted to be an intracellular protein, it was experimen tally identified on the cell wall and thus is likely secreted (32).We then examined whether the ability of C. albicans to protect S. mutans is contact-dependent by perform ing H 2 O 2 survival assays using a Transwell system that allows diffusion of nutrients and metabolic products while keeping the organisms physically separated.Since 48 h of biofilm development allowed C. albicans to penetrate the microporous membrane, a 12-h time point was utilized, and a group of unseparated biofilms was used as a control.Our findings revealed that when physically separated, C. albicans cannot protect S. mutans from the H 2 O 2 challenge (Fig. 3A).Further, since it is known that C. albicans' catalase expression is stimulated by H 2 O 2 (27), S. mutans survival following exposure to acute oxidative stress was monitored in supernatants obtained from C. albicans pre-stimulated with H 2 O 2 .No enhanced survival of S. mutans was observed when it was challenged with H 2 O 2 in the presence of C. albicans H 2 O 2 pre-stimulated superna tant (Fig. 3B).Longitudinally, all groups in the Transwell apparatus and pre-exposed C. albicans supernatant presented statistical differences in surviving S. mutans CFUs after 60 and 90 min compared to 0 min (P < 0.05).All S. mutans groups in the Transwell demonstrated reduced survival rates compared to UA159-SC5314 dual-species control biofilms after 90 min (P < 0.0001), suggesting that any extracellular catalase is likely bound to the fungal cell wall and not available to protect S. mutans at a distance.

Confocal laser scanning microscopy (CLSM) illustrates the importance of C. albicans' catalase in the survival of S. mutans during exposure to H 2 O 2
To gain insight into the potential contribution of C. albicans' catalase to biofilm architecture, SC5314 and SC5314∆cat1 were grown as 48 h single-and dual-species biofilms alongside UA159 and analyzed by CLSM.The biofilms were labeled with bacterial live-dead staining (SYTO 9 and propidium iodide-PI) to allow differentiation of live and dead S. mutans cells and calcofluor to visualize total C. albicans cells.Fig. 4A shows a representative image of a dual-species biofilm with a homogenous distribution of total C. albicans cells (live and dead: purple) surrounded by S. mutans (live: green, and dead: red).In the untreated group, there was no difference in biofilm architecture between UA159-SC5314 and UA159-SC5314Δcat1 dual-species biofilms (Fig. 4A).Also, SC5314 and SC5314Δcat1 single-species biofilms labeled with FUN1 (live cells) and Calcofluor (all cells) were comparable (Fig. S2), indicating that catalase production does not contribute to the inherent biofilm architecture of C. albicans.Of note, contrary to previous studies showing differences in hyphal formation (33,34), we did not observe any difference in hyphal abundance between SC5314 and SC5314∆cat1 (Fig. S3A).
Next, the dual-species biofilms were treated with 0.25% H 2 O 2 for 60 min and stained.While the untreated UA159-SC5314 and UA159-SC5314Δcat1 biofilms displayed no significant differences in the quantification of live or dead S. mutans or C. albicans cells (P > 0.1) (Fig. 4), the UA159-SC5314∆cat1 biofilms treated with H 2 O 2 displayed a higher abundance of dead S. mutans cells throughout the biofilm when compared to UA159-SC5314 biofilms (P < 0.01) (Fig. 4B).Representative images of H 2 O 2 -treated and untreated single-species S. mutans and C. albicans (SC5314 and SC5314∆cat1) biofilms are shown in Fig. S3.

C. albicans restores S. mutans biofilm formation in the presence of a sublethal dose of H 2 O 2 or the peroxigenic commensal Streptococcus A12
After investigating the ability of C. albicans to protect S. mutans from a lethal dose of H 2 O 2 in mature biofilms, we examined whether exposure to sublethal oxidative stress interferes with the growth and accumulation (stability) of early biofilms of UA159 and C. albicans (SC5314 and SC5314Δcat1).Here, we utilized either 0.005% H 2 O 2 (1.446 mM; Fig. 5A), a concentration that is estimated to be produced by peroxigenic streptococci (18), or direct inoculation with Streptococcus A12, a peroxigenic oral Streptococcus associated with health (Fig. 5B) (22).While 0.005% H 2 O 2 does not affect S. mutans viability (data not shown), oxidative stress can impact S. mutans behavior even below inhibitory concentrations (35).Upon exposure to 0.005% H 2 O 2 or Streptococcus A12, UA159 single-species biofilms and UA159-SC5314∆cat1 dual-species biofilms were less stable than UA159-SC5314 dual-species biofilms (P > 0.0001).As for C. albicans (Fig. 5), SC5314Δcat1 single-species biofilm stability was reduced compared to SC5314 singlespecies biofilms a following 0.005% H 2 O 2 or Streptococcus A12 challenge.To rule out other antagonistic mechanisms employed by Streptococcus A12 that were not interfering with the biofilms, we also exogenously added catalase at the time of Streptococcus A12 inoculations (Fig. S4).We found that the addition of exogenous catalase protected these biofilms against the inhibitory effect of Streptococcus A12 (P < 0.05).Our findings suggest that upon exposure to sublethal H 2 O 2, including H 2 O 2 produced by Streptococcus A12, C. albicans' catalase confers stability to both C. albicans mono-species biofilms and to dual-species biofilms with S. mutans.

C. albicans protects S. mutans from sensing oxidative stress
When experiencing oxidative stress, S. mutans upregulates stress response pathways to cope with the damaging effects of ROS (29,36,37).To gain insight into the S. mutans' oxidative response during growth in dual-species biofilms with C. albicans, we quantified the expression of selected oxidative stress genes of UA159 grown in single-or dual-spe cies biofilms with either SC5314 or SC5314∆cat1, with or without exposure to 0.005% H 2 O 2 .The S. mutans oxidative stress genes investigated were superoxide dismutase (sodA), thiol peroxidase (tpx), alkyl hydroperoxide (ahpCF), glutathione reductase (gor), and H 2 O-forming NADH oxidase (nox), which are shown to be upregulated upon peroxide stress (35).To ensure our S. mutans primers would not amplify C. albicans genes, we blasted our primers against the genome of C. albicans SC5314 and no alignment was observed (data not shown).When compared to untreated UA159 single-species biofilms, the expression of all genes was upregulated at least threefold in H 2 O 2 -treated singlespecies UA159 biofilms, but no induction of these genes was seen in UA159-SC5314 biofilms (P < 0.01) (Fig. 6).All five genes were found to be upregulated in dual-species biofilms of UA159-SC5314∆cat1 as compared to those with UA159-SC5314 (P < 0.01).
When comparing dual-species biofilms of UA159-SC5314∆cat1 to UA159 single-species biofilms, the induction of the tpx gene was smaller in UA159-SC5314∆cat1 than that observed for UA159 only (P < 0.01).Although not statistically significant, we observed a similar trend of decreased induction of ahpCF, sodA, gor, and nox in UA159-SC5314∆cat1 compared to UA159 single-species biofilms (Fig. 6).

DISCUSSION
Clinical, in vivo, and in vitro studies have shown that the cross-kingdom synergistic relationship between S. mutans and C. albicans increases the persistence of these organisms in the oral cavity and results in enhanced cariogenicity (4,5,8,10,12,38,39).However, most studies evaluating synergistic mechanisms between these microor ganisms have centered on the increased production of EPS matrix by S. mutans in dual-species biofilms (12,(40)(41)(42)(43).Moreover, since S. mutans is highly susceptible to ROS, it benefits from the presence of C. albicans in dual-species biofilms by displaying increased hydrogen peroxide stress tolerance (12).Here, by using clinical co-isolates of S. mutans and C. albicans, C. glabrata, and C. tropicalis, we showed that the protection of S. mutans against H 2 O 2 by human-associated Candida species is not restricted to laboratory strains or C. albicans.Then, using an S. mutans strain with impaired ability to produce EPS matrix and a catalase-deficient C. albicans mutant strain, we showed that the increased oxidative stress tolerance of S. mutans in dual-species biofilms with C. albicans is independent of EPS matrix production but dependent upon the catalase produced by C. albicans.Thus, the present study expanded, beyond the EPS matrix, the mutualistic repertoire of S. mutans and C. albicans.In addition, our findings point to S. mutans also benefiting from the presence of other human-associated Candida species.Biofilm matrices of cross-kingdom biofilms are often more robust, as seen with dual-species biofilms of C. albicans and multiple bacteria, including Enterococcus faecalis, Staphylococcus epidermidis, S. aureus, and S. mutans (11,(44)(45)(46).These robust matrices create diffusion barriers against environmental stressors.For example, C. albicans has increased resistance to fluconazole in biofilms with S. epidermis or S. aureus (44,45).Moreover, the increased matrix production in S. mutans-C.albicans biofilms appears to serve as a diffusion barrier, protecting both species from the antimicrobial activity of chlorhexidine (12).In S. mutans-C.albicans biofilms, the increased EPS-matrix production results from C. albicans presence, causing upregulation of S. mutans' glucosyltransferase genes (gtfB, C, and D) (10).GtfB produces insoluble glucans, and GtfC produces both soluble and insoluble glucans, contributing to the bulk of the EPS matrix in the presence of sucrose.Of note, S. mutans GtfD produces soluble glucans that are thought to serve mostly as a nutrient reservoir.Hence, while our S. mutans UA159ΔgtfB/C strain has intact gtfD, the soluble glucans produced by GtfD represent a small fraction of the EPS produced in S. mutans biofilms (47).Importantly, in vitro studies revealed that planktonic cultures and mature biofilms of S. mutans are equally susceptible to H 2 O 2 challenge mutans mutant strain showed that this survival enhancement is independent of the EPS matrix.Importantly, C. albicans protection of bacterial species from oxidative stress is not limited to S. mutans.When in biofilms, C. albicans protected the anaerobic species Clostridium perfringens and Bacteroides fragilis from the stress of an aerobic environment (50).Additionally, C. albicans has been shown to modulate the early oral bacterial microbiome by increasing the abundance of facultative and strict anaerobes (51).The mechanism involved in this protection conferred by C. albicans to less aerotolerant species is uncharacterized, but we speculate that consumption of O 2 by respiration and the production of ROS detoxifying enzymes like catalase and the extracellular superoxide dismutases (Sod4,5,6) by C. albicans promote the survival of more oxygen-sensitive species.Here, by using a catalase-deficient strain of C. albicans SC5314, we demonstra ted that, indeed, catalase activity of C. albicans is critical for increased survival of S. mutans against an exogenous H 2 O 2 challenge when both species are in close proximity.Interestingly, in C. albicans, catalase is an ROS-inducible intracellular enzyme that has also been found extracellularly in the cell wall (32).Thus, it appears that C. albicans attempts to detoxify ROS derived from the host or other sources before it can damage intracellular components (52)(53)(54), and thus neighboring ROS-sensitive cells can benefit from these potent detoxification mechanisms.
Clinically, this increased H 2 O 2 tolerance seen for S. mutans when co-cultivated with C. albicans can be alarming not only from the perspective that H 2 O 2 is used as an oral antiseptic but also because H 2 O 2 produced by peroxigenic bacteria plays an important role in maintaining oral biofilm homeostasis (18).Similar protection of the strictly anaerobic pathogen Porphyromonas gingivalis by the catalase-positive Aggregati bacter actinomycetemcomitans showed that A. actinomycetemcomitans' catalase protects P. gingivalis from the oxidative stress of aerobic conditions as well as H 2 O 2 produced by the peroxigenic commensal S. sanguinis (52).Similarly, we showed that C. albicans' catalase protects early S. mutans-C.albicans biofilms from interference by sublethal levels of added H 2 O 2 or the addition of peroxigenic commensal Streptococcus A12.The addition of exogenous catalase to single species UA159 and UA159-SC5314Δcat1 biofilms alleviated the reduction in biomass following treatment with Streptococcus A12, suggesting that catalase at least in part plays a direct role in S. mutans oxidative stress protection.Furthermore, we showed that C. albicans' catalase prevents S. mutans from mounting an oxidative stress response when exposed to sublethal oxidative stress, promoting a favorable environment for S. mutans growth and biofilm development.We also observed that when S. mutans was in biofilms with SC5314∆cat1, an oxidative stress response was still mounted to a lesser extent compared to S. mutans single-species biofilms.Since C. albicans possesses six superoxide dismutases(SODs -three extracellu lar), we speculate that these enzymes could aid in ROS detoxification and provide partial ROS protection to S. mutans when co-cultivated with SC5314∆cat1 and challenged with H 2 O 2 (53).Studies are underway to determine whether C. albicans extracellular SODs also aid in the protection of S. mutans against oxidative stress.
While oral infection with either S. mutans or C. albicans is known as a caries risk factor (4, 54), co-infection with both species is also a risk factor for caries development and reoccurrence (4,39,55).Thus, uncovering the major mechanism responsible for the increased oxidative stress tolerance of S. mutans in dual-species biofilms with C. albicans provides another piece of the puzzle to understand this complex, often synergistic, cross-kingdom relationship.The knowledge that C. albicans' catalase protects S. mutans in close contact with H 2 O 2 stress highlights the importance of developing strategies that target the disassociation of these partners in grime for the successful prevention and treatment of dental caries.Recently, Aljaffari and colleagues showed that the use of nystatin, an antifungal rinse, reduced the burden of S. mutans in saliva and plaque in patients with thrush, suggesting that approaches limiting colonization by C. albicans might help in the prevention and management of caries (56).Thus, when crafting new strategies to effectively combat caries, a multipronged approach including molecules that disrupt the synergism between caries-associated microorganisms, prebiotics that favor antagonistic properties of health-associated microorganisms, and the introduction of beneficial commensals possessing distinct pathways that inhibit different cariogenic traits of caries-associated microorganisms should be considered.

Construction of the catalase mutant in C. albicans
The catalase (CAT1) gene in C. albicans was deleted using a CRISPR-based approach, as described (58,59).Briefly, three separate sets of PCR reactions created linear DNA constructs for (a) codon-optimized Cas9 with the ENO1 promoter, (b) a guide RNA targeting the CAT1 ORF fused to the Pol III promoter of the SNR52 snoRNA gene, and (c) a repair construct in which the SAT1-FLP construct is amplified using primers that also contain homology to the 5′ and 3′ UTR of CAT1 (Primers in Table S1) (60).All three PCR products were transformed into C. albicans strain SC5314 and selected on media containing nourseothricin.Homozygous deletions were identified by PCR genotyping, and the strains were restored to nourseothricin sensitivity by growth on maltose, which activates the FLP recombinase in the repair construct, leaving the CAT1 ORF replaced with the 34-bp FLP Recognition Target sequence.We generated two independent strains, SC5314∆cat1 and SC5314∆cat1#2 that showed the expected peroxide-sensitivity phenotype.

Contact-dependence assay
To determine whether the protection conferred by C. albicans to S. mutans against H 2 O 2 is contact-dependent, S. mutans and C. albicans biofilms were grown physically separated in transwell plates (24 mm Costar Transwell plate with 0.4 um pore size).Starter cultures of S. mutans UA159 and C. albicans (SC5314 and SC5314Δcat1) were grown as described in the growth conditions above, then diluted 1:10 in TYEG to inoculate UA159 in the bottom well.SC5314, SC5314Δcat1, or TYEG (control) were placed in the insert, and cultures were incubated (12 h, 37°C, 5% CO 2 ).Then, the top insert containing C. albicans was removed, and the UA159 cultures were treated with 0.2% H 2 O 2 for 0, 30, 60, and 90 min.An aliquot of the culture was removed at each time, and serial dilutions were plated on BHI agar (48 h, 37°C, 5% CO 2 ) for CFU estimation.Control biofilms consisted of co-inoculation with S. mutans and C. albicans into the bottom well (12 h, 37°C, 5% CO 2 ).
To determine whether the supernatant of C. albicans cultures pre-exposed for 18 h to 0.01% H 2 O 2 to induce catalase expression (62) could protect S. mutans from H 2 O 2 killing, 48 h UA159 biofilms were transferred to filter-sterilized supernatants of C. albicans SC5314 or fresh TYES, followed by treatment with 0.25% H 2 O 2 for 0, 30, 60, and 90 min and CFU enumeration.

Confocal laser scanning microscopy (CLSM)
To gain insight into C. albicans' catalase influence on biofilm composition and architec ture, UA159, SC5314, and SC5314Δcat1 cultures were prepared as described above.Briefly, 48 h single-and dual-species biofilms were cultured in TYES on saliva-coated ibidi µ-slide eight chamber coverslips (Cat.No. 80826), with media replacement after 24 h.Biofilms were treated with hydrogen peroxide (0.25% H 2 O 2 in 0.89% saline solution, 60 min).H 2 O 2 was neutralized with 5 µL of 2,000 U/mL catalase and rinsed with 0.89% saline solution.Prior to imaging, the biofilms were stained for 30 min in the dark using the Live/Dead bacterial viability kit (Molecular Probes Inc., Eugene, OR, USA) to determine the S. mutans viability, whereas C. albicans cells were stained with calco fluor white M2R blue-fluorescence fungal surface labeling reagent (Molecular Probes Inc., Eugene, OR, USA).The dyes were excited and detected per the manufacturer's instructions.The confocal images of biofilms were acquired using a Nikon Ti2 confocal microscope and a Nikon C2plus camera equipped with a Plan Apo λ 60× oil objective with sequential illumination, a z-step of 1 µm, and a 1-s scanning speed.Data represent the average of results obtained from at least four biological replicates (with three images per replicate).The total and individual component biovolume was quantified using the Nis Elements 5.0 Imaging Software (63).

Oxidative stress gene expression in S. mutans
To determine if C. albicans protects S. mutans from sensing oxidative stress when exposed to a sublethal dose of H 2 O 2 , single-species of S. mutans and dual-species (S. mutans-C.albicans SC5314, or S. mutans-C.albicans SC5314Δcat1) biofilms were exposed to 0.005% H 2 O 2 and expression of S. mutans' oxidative stress genes was quantified through qRT-PCR.Briefly, biofilms were grown in six-well plates in TYES (24 h at 37°C, 5% CO 2 ) and treated with 0.005% H 2 O 2 for 5 min, followed by neutralization with 5 µL catalase (2,000 U/mL).After a wash with PBS, biofilms were harvested in RNApro tect (Qiagen), and RNA was isolated by acid-phenol-chloroform extraction, followed by treatment with DNase I (Ambion) (64).The RNA was precipitated and purified with RNeasy Mini Kit (Qiagen) followed by a second on-column DNase treatment (Rnase Free-Dnase, Qiagen) as described elsewhere (35,65).The NanoDrop One Spectropho tometer (Thermo Fisher Scientific, Waltham, MA) was used to determine RNA concentra tions, and samples were run on agarose gels to verify RNA integrity.A high-capacity cDNA reverse transcription kit containing random primers (Applied Biosystems) was used to synthesize cDNA from 1 µg of RNA from each of three independent RNA samples.Quantitative real-time PCRs (qRT-PCRs) were carried out using gene-specific primers (35) and performed in an C1000 Thermal Cycler apparatus (Bio-Rad) (66).Data were normalized by time 0 (untreated) and represented the average and standard deviation of fold change.

Statistical analysis
All experiments were performed with at least three independent biological replicates.The GraphPad Prism 9.2.0 software (GraphPad Software, La Jolla, CA, USA) was used to analyze the data.For H 2 O 2 killing, transwell assays, CLSM, and biofilm stability assays, ANOVA followed by Dunn's or Tukey's multiple comparison tests were used.For all tests, a P-value of <0.05 was considered the threshold for significance.

ADDITIONAL FILES
The following material is available online.

FIG 1
FIG 1 Candida albicans, C. glabrata, and C. tropicalis protect S. mutans against acute oxidative stress in dual-species biofilms.Peroxide tolerance of (A) clinical and laboratory strains of S. mutans and (B) clinical and laboratory strains of Candida in 48-h-old single-or dual-species biofilms after exposure to 0.25% H 2 O 2 .The percentage of survival among P1 through P5 represents co-isolated S. mutans (SM) and Candida (CA-C.albicans, CT-C.tropicalis, and CG-C.glabrata) strains from five distinct ECC patients.P1, P2, and P3 were co-infected with S. mutans and C. albicans.P4 was co-infected with S. mutans and C. tropicalis.P5 was co-infected with S. mutans and C. glabrata.Lab strains UA159: S. mutans UA159 and SC5314: C. albicans SC5314.Data represent the average and standard deviation of survival percentage compared to CFUs at time 0. Statistical difference between S. mutans survival in dual-species biofilm compared to single-species biofilm.**P < 0.01 and ***P < 0.001 (t-tests for each single-species vs. dual-species of the same patient).

FIG 2 C 5 FIG 3
FIG 2 C. albicans protects S. mutans against acute oxidative stress in dual-species biofilms in a catalase-dependent manner.Peroxide tolerance of (A) S. mutans and (B) C. albicans in 48-h-old single-or dual-species biofilms after exposure to 0.25% H 2 O 2 .Data represent the average and standard deviation of survival percentage compared to CFUs at time 0. Statistical difference between survival in dual-species biofilm with C. albicans SC5314 versus single-species biofilm or in dual-species biofilm with SC5314Δcat1 at a specific exposure time is noted with *P < 0.05, ***P < 0.001, and ****P < 0.0001 (one-way ANOVA, followed by Dunn's multiple comparison tests).Statistical difference between survival in biofilm with C. albicans SC5314 versus biofilm with SC5314Δcat1 at a specific exposure time is noted with **P < 0.01 (one-way ANOVA, followed by Dunn's multiple comparison tests).UA159: S. mutans UA159, UA159ΔgtfB/C: S. mutans UA159 with glycosyltransferase B/C deletion mutant, SC5314: C. albicans SC5314, and SC5314Δcat1: C. albicans SC5314 with catalase deletion mutant.

FIG 4
FIG 4 Representative confocal microscopy images and quantification of 48-h-old S. mutans and C. albicans dual-species biofilms illustrate the role of C. albicans' catalase in the protection of S. mutans against H 2 O 2 .The green, red, and purple colors represent live cells of S. mutans (SYTO9-: green), dead cells of S. mutans (propidium iodide (PI)-red), and all cells of C. albicans (Calcofluor-: purple), respectively.The merged 3D reconstruction is presented in the first column, while (Continued on next page)

FIG 4 ( 9 FIG 5
FIG 4 (Continued) the other columns comprise a representative Z-layer for merged and individual channels.Biovolume was quantified for each individual component and then normalized by the total components in the sample.(B) The proportion of the biovolume from each component for treated and untreated biofilms is shown.Statistical difference between the same component shown: a versus b and c versus d, P < 0.01 (one-way ANOVA, followed by Dunn's multiple comparison tests).