Unequal contribution of ALS9 alleles to adhesion between Candida albicans and human vascular endothelial cells

Abstract

The Candida albicans ALS (agglutinin-like sequence) family includes eight genes (ALS1 to ALS7, and ALS9) that share a common general organization, consisting of a relatively conserved 5' domain, a central domain of tandemly repeated sequence units, and a 3' domain of relatively variable length and sequence. To test the hypothesis that the cell-surface glycoproteins encoded by the ALS genes mediate contact between the fungal cell and host surfaces, a set of C. albicans mutant strains was systematically constructed, each lacking one of the ALS sequences. Phenotypes of the mutant strains were evaluated, primarily using adhesion assays. ALS9 is unique within the ALS family due to extensive allelic sequence variation within the 5' domain that may result in functional differences between proteins encoded by ALS9-1 and ALS9-2. Deletion of ALS9 significantly reduces C. albicans adhesion to human vascular endothelial cell monolayers. The mutation was complemented by reintegration of a wild-type copy of ALS9-2, but not ALS9-1, suggesting allelic functional differences. Complementation of the mutation with a gene fusion between the 5' domain of ALS9-2 and the tandem repeats and 3' domain of ALS9-1 also restored wild-type adhesion levels. Analysis of the als9Delta/als9Delta mutant phenotype in other assays demonstrated no significant difference from a control strain for adhesion to buccal epithelial cells or laminin-coated plastic plates. The als9Delta/als9Delta mutant did not show significant differences from the control for adhesion to or destruction of cells in the reconstituted human epithelium (RHE) disease model, or for cell-wall defects, germ-tube formation or biofilm formation in a catheter model. Analysis of ALS9 allelic frequency in a collection of geographically diverse clinical isolates showed a distinct preference for ALS9-2 allelic sequences, within both the 5' and the 3' domain of the ALS9 coding region. These data suggest greater selective pressure to maintain the ALS9-2 allele in C. albicans isolates and imply its greater relative importance in host-pathogen interactions.

Fig. 1

Diagram of (a) ALS9-1, (b) ALS9-2, and (c) the ALS9-2/ALS9-1 fusion. The locations of the ALS domains are shown below ALS9-1. The ALS9-2/ALS9-1 fusion encodes amino acids 1 to 420 from Als9-2p fused to amino acids 421 to 1854 from Als9-1p. Capital letters show the approximate location of PCR primers used for allelic genotyping reactions. Primers E and F amplify a 3’ domain region (speckled) that contains Variable Block 1 (VB1). Primers G and H amplify a region (basketweave) that contains Variable Block 2 (VB2). Solid shading in each region denotes the extra nucleotides that comprise VB1 and VB2.

Fig. 2

Adhesion assay data. Histograms showing the adherence of C. albicans control (CAI12), als9Δ/als9Δ (2028), als9Δ/als9Δ∷ALS9-1 (2740), als9Δ/als9Δ∷ALS9-2 (2064) and als9Δ/als9Δ∷ALS9-1/ALS9-2 fusion (2767) strains to (a) vascular endothelial cell monolayers, (b) buccal epithelial cells and (c) laminin-coated plastic plates. Adhesion assays used C. albicans yeast forms that were grown for 16 h in YPD at 37ºC as described in Methods. Asterisks indicate significant differences from wild-type adhesion (P ≤ 0.05).

Fig. 3

Light micrographs of buccal RHE inoculated with either (a) the C. albicans control (CAI12) or (b) the als9Δ/als9Δ mutant strain (2028) and grown for 8 h at 37ºC. The similar appearance of RHE following incubation with either strain suggested that Als9p does not contribute to epithelial damage in this experimental model.