The NDR/LATS kinase Cbk1 controls the activity of the transcriptional regulator Bcr1 during biofilm formation in Candida albicans

Abstract

In nature, many microorganisms form specialized complex, multicellular, surface-attached communities called biofilms. These communities play critical roles in microbial pathogenesis. The fungal pathogen Candida albicans is associated with catheter-based infections due to its ability to establish biofilms. The transcription factor Bcr1 is a master regulator of C. albicans biofilm development, although the full extent of its regulation remains unknown. Here, we report that Bcr1 is a phosphoprotein that physically interacts with the NDR kinase Cbk1 and undergoes Cbk1-dependent phosphorylation. Mutating the two putative Cbk1 phosphoacceptor residues in Bcr1 to alanine markedly impaired Bcr1 function during biofilm formation and virulence in a mouse model of disseminated candidiasis. Cells lacking Cbk1, or any of its upstream activators, also had reduced biofilm development. Notably, mutating the two putative Cbk1 phosphoacceptor residues in Bcr1 to glutamate in cbk1Δ cells upregulated the transcription of Bcr1-dependent genes and partially rescued the biofilm defects of a cbk1Δ strain. Therefore, our data uncovered a novel role of the NDR/LATS kinase Cbk1 in the regulation of biofilm development through the control of Bcr1.

Figure 1. The RAM pathway is required for biofilm formation in vitro.

(A) The wild-type (CEC369) and bcr1Δ (JC1081) reference strains, cbk1Δ (JC1080), mob2Δ (JC525), tao3Δ (JC848), kic1Δ (JC798) and ace2Δ (JC343) mutants were incubated during 40 hours in biofilm-inducing conditions on Thermanox plastic slides in microfermentors. (B) Determination of biofilm dry mass collected from microfermentors shown in (A). Average results of two independent experiments done in duplicate are shown. Error bars represent the standard deviation of the data throughout the paper, unless otherwise indicated. (C) Adherence to plastic slides of WT, bcr1Δ, RAM mutants and ace2Δ, as detailed in Materials and Methods. The results shown are the mean of 3 independent experiments counting 30 fields per strain in each one. (D) Biofilms induced on silicon squares for 60 hours were stained with calcofluor white and concanavaline A for CSLM visualization.

Figure 2. Bcr1 is a phosphoprotein that interacts in vivo with Cbk1.

(A) Schematic representation of Cbk1 consensus phosphorylation sites in Bcr1. Gray rectangles: Zn finger domains. (B) Cell extracts from a BCR1-HA strain (JC1144) grown at 37°C in Spider medium were analyzed by Western blot using anti-HA antibodies. A fraction from the same lysates was treated with λ-phosphatase (λPP). (C) Cell lysates from wild-type (WT, JC1144) and cbk1Δ (JC1159) strains grown in Spider medium at 37°C were subjected to 2D-WB and probed with anti-HA antibodies. Half of each lysate was treated with λ-phosphatase. Histograms obtained using ImageJ show the quantification of the relative intensity of each spot of the blot. (D) Protein extracts from a CBK1-myc BCR1-HA strain (JC1151) were immunoprecipitated using anti-HA antibodies. A strain carrying CBK1-myc (JC305) was used as negative control of the immunoprecipitation. Samples were separated by SDS-PAGE and probed with anti-myc or anti-HA antibodies.

Figure 3. Characterization of the phenotype of BCR1-phosphorylation mutants in biofilm formation in vitro.

BCR1 phosphomimetic (bcr1 ^T191E/bcr1Δ, JC1094; bcr1 ^S556E/bcr1Δ, JC1092; bcr1EE/bcr1Δ, JC1180) and phosphodefective (bcr1 ^T191A/bcr1Δ, JC1093; bcr1 ^S556A/bcr1Δ, JC1088; bcr1AA/bcr1Δ, JC1178) mutants were grown in biofilm-inducing conditions for 40 hours using microfermentors. BCR1/bcr1Δ (WT, JC1089) and bcr1Δ (JC1081) strains were used as controls. (A) Quantification of adherent cells. (B) Biofilm dry mass determination. (C) Biofilms grown using silicon squares models were stained with calcofluor white for CSLM visualization.

Figure 4. Phosphorylation of Cbk1 consensus sites is required for Bcr1 transcriptional activity.

(A) Expression of the Bcr1 target genes ALS3 and ALS1 measured by quantitative RT-PCR in the wild-type (JC1089), bcr1Δ (JC1081), bcr1EE (JC1180) and bcr1AA (JC1178) strains normalized using ADE2. The data are the mean of two independent experiments. (B) Cells from the wild-type BCR1-HA (JC1144), bcr1EE-HA (JC1177) and bcr1AA-HA (JC1176) strains were grown at 37°C in Spider medium. Protein extracts were probed with anti-HA antibodies. Anti-PSTAIRE antibodies were used as loading control.

Figure 5. Characterization of the in vitro biofilm formation phenotype of the cbk1Δ bcr1EE double mutant.

(A) Quantification of adherence ability to plastic slides of wild-type (JC1089), cbk1Δ BCR1 (JC1082), cbk1Δ bcr1AA (JC1096) and cbk1Δ bcr1EE (JC1099) cells. (B) The same strains were incubated during 40 hours in microfermentors and biofilm dry mass was determined. (C) CSLM images of biofilms induced using silicone squares after calcofluor white staining. (D) Quantitative RT-PCR measurements of ALS3 and ALS1 transcription levels in the wild-type, cbk1Δ BCR1 and cbk1Δ bcr1EE strains normalized using ADE2.

Figure 6. Cbk1-dependent phosphorylation of Bcr1 is required for virulence during disseminated candidiasis.

Mice were infected intravenously with 5×10⁵ yeast phase cells of the indicated C. albicans strains. The graphs represent the survival over a 14-day period (n = 10 mice per strain). * P<0.005.