Targeted changes of cell wall proteome influence Candida albicans ability to form single- and multi-strain biofilms.. Candida albicans

Biofilm formation is an important virulence trait of the pathogenic yeast Candida albicans. We have combined gene overexpression, strain barcoding and microarray profiling to screen a library of 531 C. albicans conditional overexpression strains (~10% of the genome) for genes affecting biofilm development in mixed-population experiments. The overexpression of 16 genes increased strain occupancy within a multi-strain biofilm, whereas overexpression of 4 genes decreased it. The set of 16 genes was significantly enriched for those encoding predicted glycosylphosphatidylinositol (GPI)-modified proteins, namely Ihd1/Pga36, Phr2, Pga15, Pga19, Pga22, Pga32, Pga37, Pga42 and Pga59; eight of which have been classified as pathogen-specific. Validation experiments using either individually- or competitively-grown overexpression strains revealed that the contribution of these genes to biofilm formation was variable and stage-specific. Deeper functional analysis of PGA59 and PGA22 at a single-cell resolution using atomic force microscopy showed that overexpression of either gene increased C. albicans ability to adhere to an abiotic substrate. However, unlike PGA59, PGA22 overexpression led to cell cluster formation that resulted in increased sensitivity to shear forces and decreased ability to form a single-strain biofilm. Within the multi-strain environment provided by the PGA22-non overexpressing cells, PGA22-overexpressing cells were protected from shear forces and fitter for biofilm development. Ultrastructural analysis, genome-wide transcript profiling and phenotypic analyses in a heterologous context suggested that PGA22 affects cell adherence through alteration of cell wall structure and/or function. Taken together, our findings reveal that several novel predicted GPI-modified proteins contribute to the cooperative behaviour between biofilm cells and are important participants during C. albicans biofilm formation. Moreover, they illustrate the power of using signature tagging in conjunction with gene overexpression for the identification of novel genes involved in processes pertaining to C. albicans virulence. Overall design: A total of 8 samples are included in this study. For fitness profiling of planktonic cells, 2 biological replicates were analyzed (samples Pool_Fitness_planktonic_rep1 and Pool_Fitness_planktonic_rep2). For quantification of strain abundance during biofilm formation, 6 biological replicates were analyzed (Pool_Biofilm_rep1, Pool_Biofilm_rep2, Pool_Biofilm_rep3, Pool_Biofilm_rep4, Pool_Biofilm_rep5, Pool_Biofilm_rep6). Genomic DNA was purified and used as template to PCR-amplify barcodes, which were then used as probes for microarray hybridization. This experiment was done twice independently. For quantification of strain abundance during multi-strain biofilm formation, strain pools were grown in minimal GHAUM medium with or without doxycycline, each inoculum was then diluted to an OD600 of 1 in fresh minimal GHAUM medium with or without doxycycline and left at room temperature for 30 min, to allow further overexpression. Plastic slides (Thermanox™; Nunc) were immersed in the inoculum for 30 min at room temperature to allow adhesion of cells to the plastic substrate. The plastic slides were then transferred to the glass vessel of a 40-mL incubation chamber. This vessel has two glass tubes inserted to drive the entry of medium and air, while used medium is evacuated through a third tube. The flow of medium is controlled by a recirculation pump (Ismatec®) set at 0.6 mL.min-1 and pushed by pressured air supplied at 105 Pa, conditions minimizing planktonic phase growth and promoting biofilm formation. The chambers with the plastic substrate were incubated at 37C and biofilms were grown for 40h followed by genomic DNA extraction, barcode amplification and differential labeling (dox-treated samples with Cy5, untreated samples with Cy3) and hybridization to barcode microarrays.