Targeted gene deletion in Candida parapsilosis demonstrates the role of secreted lipase in virulence

Abstract

Candida parapsilosis is a major cause of human disease, yet little is known about the pathogen's virulence. We have developed an efficient gene deletion system for C. parapsilosis based on the repeated use of the dominant nourseothricin resistance marker (caSAT1) and its subsequent deletion by FLP-mediated, site-specific recombination. Using this technique, we deleted the lipase locus in the C. parapsilosis genome consisting of adjacent genes CpLIP1 and CpLIP2. Additionally we reconstructed the CpLIP2 gene, which restored lipase activity. Lipolytic activity was absent in the null mutants, whereas the WT, heterozygous, and reconstructed mutants showed similar lipase production. Biofilm formation was inhibited with lipase-negative mutants and their growth was significantly reduced in lipid-rich media. The knockout mutants were more efficiently ingested and killed by J774.16 and RAW 264.7 macrophage-like cells. Additionally, the lipase-negative mutants were significantly less virulent in infection models that involve inoculation of reconstituted human oral epithelium or murine intraperitoneal challenge. These studies represent what we believe to be the first targeted disruption of a gene in C. parapsilosis and show that C. parapsilosis-secreted lipase is involved in disease pathogenesis. This efficient system for targeted gene deletion holds great promise for rapidly enhancing our knowledge of the biology and virulence of this increasingly common invasive fungal pathogen.

Figure 1. Gene deletion in C. parapsilosis.

(A) Design of gene targeting for the C. parapsilosis lipase locus showing the SAT1 flipper cassette with the homologous C. parapsilosis lipase fragments 5′ LIP and 3′ LIP as well as the FLP recombination target sequences (FRT). The probe used to verify correct integration and deletion of the SAT1 flipper by Southern blot hybridization is represented by a black bar. (B) Southern blot hybridization analysis of genomic DNA (PacI, AvrII, double digested) isolated from the WT GA1 (lane 1), the heterozygous mutant CpLIP1-2/Δcplip1-2:SAT1-FLIP (HebFLP) before FLP activation (lane 2), the heterozygous mutant CpLIP1-2/Δcplip1-2:FRT (HE) after excision of SAT1 flipper cassette (lane 3), the homozygous mutant Δcplip1-2/Δcplip1-2:SAT1-FLIP (KobFLP) before FLP activation (lane 4), the homozygous mutant Δcplip1-2/Δcplip1-2:FRT (KO) after excision of SAT1 flipper cassette (lane 5), the reintegration mutant Δcplip1-2/Δcplip1-2/CpLIP2:SAT1-FLIP (RebFLP) before FLP activation (lane 6), and the reconstituted mutant Δcplip1-2/Δcplip1-2/CpLIP2:FRT (RE) after FLP activation (lane 7). Diagrams of the structures and size of the hybridization fragments are shown at right. (C) Nou selection of caSAT1-containing (Nou^R) transformants (left) and screening for Nou^S colonies after FLP activation (right). Arrows indicate small Nou^S colonies.

Figure 2. Lipase activity of C. parapsilosis WT and lipase mutants.

(A) Effect of C. parapsilosis WT and lipase mutants on rhodamine B–FCS agar plate. Lipolytic activities are demonstrated by the red-stained regions around the colonies. 1, WT strain; 2, homozygous mutant; 3, heterozygous mutant; and 4, reconstituted mutant. (B) Rhodamine B–FCS agar plate irradiated with UV light. Fluorescent halos indicate lipolytic activity. The experiment was performed 3 times with similar results. (C) Relative extracellular lipolytic activity of C. parapsilosis WT, heterozygous CpLIP1-2/Δcplip1-2 (HE), homozygous Δcplip1-2/ Δcplip1-2 (KO), and reconstituted Δcplip1-2/CpLIP2 (RE) mutants. Lipolytic activities of supernatants of overnight cultures in YNB–olive oil medium were measured using paranitrophenyl palmitate. Error bars indicate SD. *P < 0.0001 (ANOVA).

Figure 3. Growth curves of C. parapsilosis WT and lipase mutants in YPD, YNB, YNB–olive oil, and in YNB-intralipid media.

Growth curves of the WT strain and heterozygous CpLIP1-2/Δcplip1-2, homozygous Δcplip1-2/Δcplip1-2, and reconstituted Δcplip1-2/CpLIP2 mutants in YPD (A), YNB (B), and YNB–olive oil (C) media. (D) CFU in YNB-intralipid solution. Error bars indicate SD. The experiment was repeated, and similar results were found.

Figure 4. Comparison of biofilm formed by C. parapsilosis WT and homozygous Δcplip1-2/Δcplip1-2, heterozygous CpLIP1-2/Δcplip1-2, and reconstituted Δcplip1-2/CpLIP2 mutants.

(A) Light microscopy, (B) fluorescent microscopy, and (C) confocal microscopy images of WT biofilm on polyethylene surface after incubation for 48 hours at 37°C in YNB medium. The biofilm formed by the homozygous Δcplip1-2/Δcplip1-2 mutant is shown by (D) light microscopy, (E) fluorescent microscopy, and (F) confocal microscopy. (G–I) Comparison of biofilm formed by C. parapsilosis WT and lipase mutants on different surfaces. Relative XTT activity of biofilm formed on (G) polyethylene, (H) silicone, and (I) polystyrene surfaces. XTT activity was measured at 492 nm. *P < 0.02 (ANOVA). Error bars indicate SD. The experiment was repeated 3 times, and equivalent results were obtained. Scale bars: 50 μm (A, B, D, and E); 20 μm (C and F).

Figure 5. Reconstituted human oral epithelium at 48 hours infected with C. parapsilosis WT.

(A and B) Uninfected control tissues, (C and D) WT, (E and F) heterozygous mutant CpLIP1-2/Δcplip1-2, (G and H) homozygous mutant Δcplip1-2/ Δcplip1-2, and (I and J) reconstituted Δcplip1-2/CpLIP2. Insets show histopathological alterations. Top insets in C and E show apoptotic cells; middle insets in C and E show intercellular edema; bottom inset in C shows cleft formation and tissue separation; and bottom inset in E shows vacuolization. Left panels demonstrate hematoxylin and eosin staining, and right panels periodic acid Schiff staining. Arrows indicate yeast cells, and arrowheads indicate pseudohyphae. Scale bars: 10 μm (A–J); 5 μm (all insets). (K) Relative LDH measured in the tissue culture supernatant by reconstituted human oral epithelium after 48 hours of coculture with WT, heterozygous mutant CpLIP1-2/Δcplip1-2, homozygous mutant Δcplip1-2/ Δcplip1-2, and reconstituted mutant Δcplip1-2/CpLIP2. Error bars indicate SD. The assay was repeated 3 times. *P < 0.02 (ANOVA).

Figure 6. Phagocytosis and killing of C. parapsilosis cells by macrophage-like cell lines J774.16 and RAW 264.7.

Fluorescent microscopic image of phagocytosis of C. parapsilosis (A) WT cells and (B) Δcplip1-2/ Δcplip1-2 lipase-negative cells by J774.16 macro­phages. Dead C. parapsilosis (arrowheads) show bright red and yellow fluorescence. Living cells show green fluorescence (arrows). Scale bars: 10 μm. (C and D) Phagocytosis by (C) J774.16 and (D) RAW 264.7 macrophage-like cells. Bars represent the mean measurements from 3 wells (at least 5 fields each). *P < 0.02. (E and F) Intracellular killing of C. parapsilosis WT and mutants by macrophage-like cells. Killing by (E) J774.16 and (F) RAW 264.7 macrophage-like cells. Bars represent mean percentage of killing determined by CFU counts. Error bars indicate SD. **P < 0.01 (ANOVA).

Figure 7. Endothelial damage assay of C. parapsilosis WT, heterozygous mutant CpLIP1-2/Δcplip1-2, homozygous mutant Δcplip1-2/Δcplip1-2, and reconstituted Δcplip1-2/CpLIP2 mutant.

C. albicans SC5314 cells were used as a positive control. HUVEC lysis was measured after 8 hours by LDH assay. Error bars indicate SD. The assay was repeated 3 times.

Figure 8. Intraperitoneal and intravenous infection of mice with WT and homozygous Δcplip1-2/Δcplip1-2, heterozygous CpLIP1-2/Δcplip1-2, and reconstituted Δcplip1-2/CpLIP2 mutants.

(A) CFUs in the kidney, spleen, and liver 2 days after intraperitoneal infection and (B) 4 days after infection. (C) CFU counts determined from mice intravenously infected in the kidney, spleen, and liver 2 days after infection. Each box represents 1 mouse. *P ≤ 0.0049 (Kruskal-Wallis); ^#, no detectable CFU. Data are representative of 2 independent experiments.