To study the contribution of Bcr1-regulated genes in mucosal biofilms, a mouse oral infection model was used where C. albicans forms white pseudomembranes (biofilms) on the dorsal surface of the tongue 
. Tongues from animals infected with genetically manipulated strains were excised and examined by macroscopic “clinical” evaluation, assessment of cultivable fungal burden, and histologic analysis to visualize the thickness of biofilms. Consistent with results in the mouse vaginal mucosa model 
, we found that the bcr1/bcr1
strain was deficient in forming a clinically visible mucosal biofilm on the tongues of immunocompromised mice in vivo (). At the histologic level this mutant formed a thin, interrupted biofilm on the dorsal surface of the tongue (, arrows). These results are in agreement with the recently reported attenuated biofilm phenotype of a bcr1/bcr1
mutant in the rat denture biofilm model 
Biofilm formation and histological examination of the tongues of mice infected with the bcr1/bcr1 mutant, DAY185 (reference) and complemented strains.
Surface area estimates of pseudomembranes, examined macroscopically during necropsy, showed approximately 80–100% coverage of the tongue dorsal surface with biofilm formed by the reference and reconstituted strains, while less than 10% of the tongue surface in mice infected with the bcr1/bcr1 mutant was covered by biofilm (). In accordance with this we also found that the tongue fungal burden of mice infected with the bcr1/bcr1 mutant was significantly lower than that of mice infected with either the reference or complemented strains ().
Biofilm surface area and fungal burden of animals infected with the bcr1/bcr1 mutant and related strains.
C. albicans also forms a biofilm when grown on a three-dimensional model of the human oral mucosa 
. When grown on this model, the bcr1/bcr1
mutant was slow in forming a mature biofilm, and after 24 hours of growth formed a biofilm comprised mainly of yeast cells, in contrast to its reference and complemented strains (). Moreover, this mutant was significantly less capable of triggering mucosal tissue damage (). This could be attributed to the dominant yeast morphotype in biofilm cells, which lacks expression of several hyphae-specific epithelial adhesins that may also be involved in oral epithelial cell damage 
Biofilm formed by the bcr1/bcr1 mutant, complemented and reference (DAY185) strains on a three-dimensional organotypic model of the oral mucosa.
Mucosal damage by the indicated strains in the three dimensional model of the oral mucosa.
Adhesion is a fundamental process under Bcr1 control that promotes biofilm formation on catheter surfaces 
. Therefore, we evaluated the contribution of the Bcr1-regulated adhesins Als1, Als3 and Hwp1 in the capacity of C. albicans to form a biofilm on the oral mucosa in vivo. Increased expression of ALS1
in the bcr1/bcr1
mutant background did not significantly affect the surface area covered by biofilm (), increase the tongue fungal burden (), or promote mucosal biofilm formation at the histologic or macroscopic level (). Similarly, expression of ALS3
under the TEF1
promoter did not significantly affect the surface area covered by biofilm (), the tongue fungal burden (), or the biofilm thickness (, arrows); the strain's intermediate phenotype is discussed further below. A clear example of phenotypic rescue from overexpression was observed with HWP1
. Specifically, HWP1
overexpression in this genetic background improved biofilm formation and increased fungal burden in the animals significantly (, ). This finding agrees with observations in the catheter biofilm model where this construct partly but significantly reversed the bcr1/bcr1
. In addition to being important in biofilm development 
, Hwp1 is also a well established oral epithelial cell adhesin 
, which plays a role in the pathogenesis of oral infection 
. Our findings provide additional support for the role of Hwp1 in oral infection, and in addition argue that the reduced expression of HWP1
in the bcr1/bcr1
mutant is a major cause of the mutant's oral epithelial biofilm defect.
Biofilm formation and histological examination of the tongues of mice infected with strains overexpressing adhesins ALS1, ALS3 and HWP1 in the bcr1/bcr1 background.
overexpression in the bcr1/bcr1
background partly reversed the bcr1/bcr1
phenotype, based upon visual inspection (). However, the mean percentage surface area covered by biofilm and the CFU counts from infected tissue with this strain did not reach statistical significance (). These findings were thus somewhat in contrast with data in the venous catheter biofilm model showing that, although Als3 is not absolutely required for biofilm formation, ALS3
overexpression completely reverses the biofilm phenotype of the bcr1/bcr1
. We note that these two biofilm experimental systems are quite different, and that C. albicans
adhesins exhibit high substrate specificity, even when they have highly related sequences in their binding domains 
. In addition, innate host defense mechanisms in the oral cavity, coupled with salivary flow and mechanical cleansing by chewing, may modify the ability of these strains to establish a biofilm in vivo.
However, because regulation of biofilm-associated gene expression may vary significantly in different biofilm model systems 
, we wanted to rule out the possibility that Als3 expression is Bcr1-independent in the oral mucosa, which could explain the disparate results with these strains in different model systems. Therefore, we quantified Als3 gene expression in the bcr1/bcr1
mutant and adhesin-overexpressing strains, when grown on a three dimensional model of the human oral mucosa. As anticipated, when grown on an oral mucosa tissue analogue Als3 expression levels were lower in the bcr1/bcr1
deletion mutant as well as in the ALS1-
-overexpressing strains, compared to the reference strain (). In contrast, Als3 expression levels in the ALS3
-overexpressing strain were three fold higher than the reference strain (). This finding argues against the possibility of differential regulation of Als adhesins in the oral mucosa, and further supports the idea that different experimental systems can reveal tissue-specific functions of adhesins. Thus our results establish that Hwp1, and not Als3, is a critical Brc1 adhesin target relevant to oral thrush. In agreement with our findings, it has been reported that biofilm formation in a subcutaneous rat model requires Bcr1 but not Als3 
Detection of Als3 expression in the bcr1/bcr1 mutant, reference (DAY185) and adhesin-overexpressing strains on the 3D model of the human oral mucosa.
Due to the complexity of the structure of oral mucosal biofilms 
, transcriptional deregulation of a single C. albicans gene is unlikely to be responsible for mediating loss of the mucosal biofilm phenotype in the bcr1/bcr1
mutant. This explains the finding that HWP1
overexpression only partially reversed the bcr1/bcr1
phenotype. We thus sought to identify additional, adhesion-unrelated genes under the transcriptional control of Bcr1, contributing to its inability to form mucosal biofilms.
A prominent feature of infections characterized by soft tissue biofilms is infiltration of infected tissues by neutrophils, which confer innate immune protection 
. We previously showed that neutrophils infiltrate the oral mucosal biofilm mass in this mouse oral infection model 
. We thus hypothesized that the bcr1/bcr1
mutant fails to develop a robust biofilm on the oral mucosa at least partly because it is more efficiently cleared by biofilm-infiltrating neutrophils. To begin to test this hypothesis we examined susceptibility of this mutant to killing on plastic by the HL-60 neutrophil-like cell line using a modification of the XTT assay 
. Indeed, we found that the bcr1/bcr1
mutant was more susceptible to killing than its reference and reconstituted strains, regardless of the effector to target ratio used in killing assays (). In fact the effector to target ratio of the bcr1/bcr1
strain corresponding to the MIC50 in this assay was five times lower than the reference and reconstituted strains (). These findings were confirmed when freshly isolated human neutrophils were used in killing assays (). Finally, we extended these findings to the oral mucosa, by testing susceptibility of the bcr1/bcr1
mutant to leukocyte killing on a three-dimensional model of the human oral mucosa (). As expected, this mutant was also more susceptible to leukocyte-inflicted damage when grown on a three dimensional model of the human oral mucosa (), supporting our hypothesis that it may be more effectively cleared in the oral environment.
Susceptibility of the bcr1/bcr1 mutant, reference (DAY185) and reconstituted strains to leukocyte-inflicted damage.
Susceptibility of the bcr1/bcr1 mutant, hyr1/hyr1 mutant, and HYR1- or HWP1- overexpressing strains in the bcr1/bcr1 background, to human leukocytes.
Overexpression of HYR1
, but not HWP1
, in the bcr1/bcr1
background, significantly rescued the higher susceptibility phenotype of this mutant to killing by leukocytes both on plastic and on the oral tissue surface (). These findings confirmed previous reports showing that Hyr1, but not Hwp1 is involved in susceptibility to neutrophil killing 
. More importantly, these data, combined with the finding that the hyr1/hyr1
mutant was highly susceptible to killing in an oral mucosal tissue model (), suggest that this gene is indirectly contributing to the observed mucosal biofilm phenotype by conferring resistance to neutrophil killing. Thus, this study now strongly implicates Hyr1 in innate immune cell evasion of C.
albicans in the oral mucosa.
Hyr1 is a GPI-anchored cell wall protein, expressed during hyphal development 
and repressed upon neutrophil encounter 
, however, relatively little is known about its exact role in virulence. Because there is no severe biofilm defect on catheter surfaces in the hyr1/hyr1
, we hypothesized that this mutant may have only a moderately attenuated biofilm phenotype on the tongue surface. As expected, this mutant formed biofilms covering part of the tongue surface, and overexpression of HYR1
in the bcr1/bcr1
background did not rescue the oral mucosal biofilm phenotype (, ). However, we also anticipated that due to increased susceptibility to neutrophil killing, preventing deep tissue invasion 
, the tissue fungal burden in mice infected with the hyr1/hyr1
mutant would be severely attenuated. Consistent with this hypothesis, histological assessment showed that, in contrast to the reference strain which reached the granular and prickle epithelial cell layers, the hyr/hyr1
mutant was confined within the keratin layer ( arrows). As a result, the tissue fungal burden was severely attenuated in mice infected with the hyr1/hyr1
mutant (). However, overexpression of HYR1
in the bcr1/bcr1
background was not sufficient to reverse the tissue invasion or fungal burden phenotype of the bcr1/bcr1
mutant ( arrows, and ). This finding is consistent with the fact that, in addition to neutrophil clearance, other virulence mechanisms directly related to epithelial adhesion and invasion, as noted above, are contributing to the severely attenuated phenotype of this mutant.
Biofilm formation and histological examination of the tongues of mice infected with DAY286 (reference), hyr1/hyr1 mutant and HYR-1 overexpressing strains in the bcr1/bcr1 background.
Biofilm surface area and fungal burden in animals infected with the bcr1/bcr1 mutant, hyr1/hyr1 mutant, HYR1- overexpressing and reference strains.
In conclusion, our studies provide valuable new insights which promote the understanding of the pathogenesis of Candida mucosal biofilm infections. We have shown that the transcription factor Bcr1 is a critical regulator of oral mucosal biofilm formation and identified two genes, HWP1 and HYR1, under Bcr1 control that govern the inability of the bcr1/bcr1 strain to form robust oral mucosal biofilms. Although much is known about the function of Hwp1 and its role in oral mucosal infection, this is the first study implicating Hyr1 in the pathogenesis of oral biofilm infection.