Genome engineering in the yeast pathogen Candida glabrata using the CRISPR-Cas9 system

Abstract

Among Candida species, the opportunistic fungal pathogen Candida glabrata has become the second most common causative agent of candidiasis in the world and a major public health concern. Yet, few molecular tools and resources are available to explore the biology of C. glabrata and to better understand its virulence during infection. In this study, we describe a robust experimental strategy to generate loss-of-function mutants in C. glabrata. The procedure is based on the development of three main tools: (i) a recombinant strain of C. glabrata constitutively expressing the CRISPR-Cas9 system, (ii) an online program facilitating the selection of the most efficient guide RNAs for a given C. glabrata gene, and (iii) the identification of mutant strains by the Surveyor technique and sequencing. As a proof-of-concept, we have tested the virulence of some mutants in vivo in a Drosophila melanogaster infection model. Our results suggest that yps11 and a previously uncharacterized serine/threonine kinase are involved, directly or indirectly, in the ability of the pathogenic yeast to infect this model host organism.

Figure 1. Design of CRISPR-Cas9 in C. glabrata and schematic of ADE2 disruption.

(A) CAS9 expression was done either with the p414 vector engineered by DiCarlo and colleagues (39) under the S. cerevisiae TEF1 promoter (pTEF1) or with the pRS314 under the C. glabrata CYC1 promoter (pCYC1). sgRNAs were cloned in a pRS315 and expressed either under the RNA Pol III promoters of S. cerevisiae SNR52 (pSNR52) or C. glabrata RNAH1 (pRNAH1). (B) Simplified scheme of adenine biosynthesis. Cells deprived of ADE2 display a red phenotype due to the accumulation of a red pigment. (C) Two target sequences localized in 5′ of the ADE2 were chosen for gene disruption: sgADE2.1 and sgADE2.3. Arrowheads indicate Cas9 cleavage sites. (D) Workflow of the sequential transformations done for sgRNA and CAS9 expression in C. glabrata and description of the following experiments.

Figure 2. Efficient CRISPR-Cas9 gene disruption of the ADE2 locus.

(A) Drop test of ∆HTL, ∆HTL + sgADE2.1 or sgADE2.3, ∆HTL + CAS9 (TEF1 or CYC1), ∆HTL + sgADE2.1 or sgADE2.3 + CAS9 (TEF1 or CYC1) on SCGlc 2% and SCGlc 2% without leucine (-L) or tryptophan (-W) or both -L-W. Plates were incubated 2 days at 30 °C. (B,C) Indels in ADE2 cells were monitored by Surveyor assay in ∆HTL strain compared to the ∆HTL + CAS9 (CYC1), ∆HTL + sgADE2.1 + CAS9 (CYC1) or ∆HTL + sgADE2.3 + CAS9 (CYC1). (D) Events leading to ADE2 disruption were monitored by sequencing each strain and by comparing to the in silico constructs. Insertions and deletions are shown in green. For each deleted strain, 6 different clones were tested.

Figure 3. Recombination efficiency in C. glabrata using targeted gene disruption with CRISPR-Cas9.

(A) Workflow of C. glabrata expressing sgADE2.1 transformation with CAS9 (CYC1) and either XTAG or HIS3 dsDNA bearing either 20 or 200 bp of homology domain (HD). (B) For homologous recombination, XTAG and HIS3 cassettes were inserted at the sgADE2.1 cut site in ADE2. Sequences of XTAG and HIS3 are listed in Supplementary Fig. 3. For each cassette, we appended either 20 or 200 bp of HD corresponding to the sequence around sgADE2.1 cutting site. (C) ∆HTL strain expressing the sgADE2.1 was transformed with a dsDNA cassette XTAG or HIS3 with or without the pRS314-CAS9 (CYC1). To evaluate efficient recombination at the ADE2 locus, we checked integration of XTAG and HIS3 by PCR after selection of mutants on SCGlc 2% -L-W plates. In each experiment, we tested HR with cassettes bearing 20 or 200 bp-long flanking regions. *p < 0,05, **p < 0,1, nd: not different, each bar represents mean values of n > 200 clones, SEM are shown.

Figure 4. Phenotypic characterization of the ∆HTL, ade2, yps11 and vpk1 strains.

(A) Drop test of each strain engineered on YPD plates at pH5.5 and 7.5 supplemented in NaCl (500 mM), ZnCl₂ (10 mM), CaCl₂ (100 mM) or caffeine (7.5 mM). Plates were stored for 1 day at 30 or 37 °C except for YPD plates supplemented with ZnCl₂ stored for 3 days. (B) Average generation time of ∆HTL, ade2, yps11 and vpk1 strains monitored during growth in SCGlc 2% liquid medium. Numbers are mean generation time for each strain. n = 3, SEM are shown.

Figure 5. Study of C. glabrata infection in D. melanogaster after genome engineering by the CRISPR-Cas9 system.

(A) Scheme of Cas9-directed gene disruption in C. glabrata prior to D. melanogaster infection. After sequential transformation of sgRNA- and CAS9 (CYC1)-expressing vectors, targeted gene disruption was checked by sequencing. Vectors were then lost by growing yeasts overnight in liquid SCGlc 2% media. (B) Survival curves of immuno-competent A5001 flies and immuno-compromised MyD88 flies after infection with clean prick (cp), ∆HTL strain or ∆HTL + Cas9. Survival are shown as percentage of surviving flies. **p < 0,05 when compared to survival curve of MyD88 + cp or A5001 + cp. (C) Colony Forming Units (CFU) of cp and C. glabrata ∆HTL or ∆HTL + Cas9 after infection of D. melanogaster A5001 and MyD88 flies. (D) Infection of A5001 and MyD88 flies with the ∆HTL strain and strains disrupted for YPS11 and VPK1. ****p < 0,0001 when compared to survival cure of MyD88 + ∆HTL; ns, not significant. (E) CFU experiments related to infections performed in (D). Drosophila drawing is the work of B. Nuhanen, and was obtained from https://commons.wikimedia.org/wiki/File:Drosophila-drawing.svg, and is licensed under a CC BY-SA license (https://creativecommons.org/licenses/by-sa/3.0/deed.en). (F) Infection of latex (LTX) beads-injected A5001 and MyD88 flies with the ∆HTL strain or the NHEJ-vpk1 mutant.