A glucan synthase FKS1 homolog in cryptococcus neoformans is single copy and encodes an essential function

Abstract

Cryptococcal meningitis is a fungal infection, caused by Cryptococcus neoformans, which is prevalent in immunocompromised patient populations. Treatment failures of this disease are emerging in the clinic, usually associated with long-term treatment with existing antifungal agents. The fungal cell wall is an attractive target for drug therapy because the syntheses of cell wall glucan and chitin are processes that are absent in mammalian cells. Echinocandins comprise a class of lipopeptide compounds known to inhibit 1,3-beta-glucan synthesis, and at least two compounds belonging to this class are currently in clinical trials as therapy for life-threatening fungal infections. Studies of Saccharomyces cerevisiae and Candida albicans mutants identify the membrane-spanning subunit of glucan synthase, encoded by the FKS genes, as the molecular target of echinocandins. In vitro, the echinocandins show potent antifungal activity against Candida and Aspergillus species but are much less potent against C. neoformans. In order to examine why C. neoformans cells are less susceptible to echinocandin treatment, we have cloned a homolog of S. cerevisiae FKS1 from C. neoformans. We have developed a generalized method to evaluate the essentiality of genes in Cryptococcus and applied it to the FKS1 gene. The method relies on homologous integrative transformation with a plasmid that can integrate in two orientations, only one of which will disrupt the target gene function. The results of this analysis suggest that the C. neoformans FKS1 gene is essential for viability. The C. neoformans FKS1 sequence is closely related to the FKS1 sequences from other fungal species and appears to be single copy in C. neoformans. Furthermore, amino acid residues known to be critical for echinocandin susceptibility in Saccharomyces are conserved in the C. neoformans FKS1 sequence.

FIG. 1

Integration of pWL42 through homologous recombination at the C. neoformans FKS1 (CnFKS1) locus. Two potential homologous integration events are depicted. In both panels, plasmid pWL42 is shown as a circle with the ADE2 gene (stippled box) inserted within a cloned fragment of the CnFKS1 locus (cross-hatched boxes). The thicker cross-hatched regions on the linear maps correspond to the CnFKS1 DNA segments present on the plasmid. Thick lines with arrowheads or vertical bars at the ends show the direction of transcription and the approximate size of the CnFKS1 reading frame. The vertical bars indicate the position of 5′ or 3′ truncation, and broken lines indicate promoterless reading frames. The ADE2 insertion is transcribed in the direction opposite that of CnFKS1. In panel A, pWL42 is shown integrating at the CnFKS1 locus through a single crossover event (×) within the cloned 5′ region upstream of the CnFKS1 ORF. The integrated plasmid (bracket) creates an interrupted copy (CnFKS1 5′, 3′ Δ) of CnFKS1, but an intact copy of the ORF remains downstream of the site of integration. The striped box above the intact ORF illustrates the approximate size and position of a PCR product obtained from A-type integration, wild-type, or ectopic integration but not B-type integration. In panel B, disrupting integration of pWL42 is depicted, in which copies of CnFKS1 truncated at either the 3′ end (CnFKS1 3′ Δ) or the 5′ end (CnFKS1 5′ Δ) are created. The striped box illustrates a PCR product specific to disrupting and wild-type or ectopic pWL42 integrants but not A-type integration.

FIG. 2

Glucan synthase parsimony tree. Abbreviations: SpFKSx1, Schizosaccharomyces pombe FKS chromosome 1 (accession no. Z98601); SpFKSx2, S. pombe FKS chromosome 2 (accession no. AL021839); AfFKS, Aspergillus fumigatus FKS (accession no. U79728); AnFKS, Aspergillus nidulans FKS (accession no. U51272); CaFKS1, Candida albicans FKS1 (accession no. D88815); CaGSL1, C. albicans GSL1 (accession no. D88816); CaGSL2, C. albicans GSL2 (accession no. AB001077); ScFKS1, Saccharomyces cerevisiae FKS1 (accession no. S50235); ScFKS2, S. cerevisiae FKS2 (accession no. S50240); ScFKS3, S. cerevisiae FKS3 (accession no. S53976). (All accession numbers are GenBank except for the last three which are PIR.) The numbers in parentheses are the percent identity with C. neoformans Fks1p based on GCG GAP alignments.

FIG. 3

Cryptococcus FKS Southern blot. Each lane contains 10 μg of restriction enzyme-digested H99 genomic DNA. The enzymes were BamHI (lane 1), PstI (lane 2), NcoI (lane 3), and SmaI (lane 4). Hybridization probe was ³²P-labeled random-primed 4-kb BamHI-PstI H99 genomic fragment derived from pCG2 and encompasses the first 1,221 codons (of 1,643) of the C. neoformans FKS1 sequence. Arrows indicate positions of HindIII-digested lambda standards.

FIG. 4

PCR screening profiles. The left panel depicts the PCR pattern expected from wild-type (WT), A-type (A), and B-type (B) integration of pWL42. The right gels show actual screening examples from a wild-type control (WT) and an A-type integration (A). A B-type integration was not obtained. The sizes of the A- and B-specific PCR products are 3,371 and 2,046 bp, respectively.

FIG. 5

Southern blot analysis of nondisrupting integrants. Genomic DNA isolated from two pWL42 transformants (lane 1, strain 42-1-3; lane 2, strain 42-1-28) and from strain H99 (lane 3) was digested with PstI (panels A, B, and C) or XmaI and ClaI (panel D), and analyzed on four separate Southern blots. The probes for the blots were as follows: for panel A, the 1.5-kb NotI-XmaI CnFKS1 fragment from pCG4; for panel B, the 3.0-kb pGEM5zf vector made linear with NotI digestion; for panel C, the 0.9-kb XhoI-XhoI CnFKS1 fragment from pCG2 which represents the region of FKS1 that was replaced by ADE2 in pWL42; and for panel D, the 3.0-kb KpnI-XmaI CnADE2 fragment from pCnADE2ΔApa. Arrows on the side of panels A to D denote the migration positions of HindIII fragments of lambda (from top to bottom, 23.1, 9.4, 6.6, 4.3, 2.3, and 2.0 kb). Maps predicted for the CnFKS1 locus following ectopic, nondisrupting, or disrupting integration are presented below panels A to D. Black bars at the bottom of the map show the positions of the probe fragments on the disrupted pWL42 integrant map. Features of the maps (ORFs, plasmid DNA, and regions of CnFKS1 contained in pWL42) are as described in the legend to Fig. 1. The brackets above each map indicate the sizes and positions of fragments generated by PstI digestion; fragments produced by digestion with XmaI and ClaI are shown below. A bar indicating scale is shown at the bottom right.

FIG. 6

CnFKS1 Northern analysis. Lane 1, H99 parent strain; lane 2, 42-1-28; lane 3, 42-2-13; lane 4, 42-1-43; lane 5, 42-1-48; lane 6, 42-6-60. Lane 2 and 3 are A-type, nondisrupting integrations. Lanes 4 to 6 contain strains with putative genomic rearrangements. The hybridization probe was ³²P-labeled random-primed 4-kb BamHI-PstI H99 genomic fragment derived from pCG2. The arrow indicates the position of the ∼9-kb FKS mRNA band.