Corneal Virulence of Candida albicans Strains Deficient in Tup1-Regulated Genes

Abstract

purpose. To evaluate the role of rbt genes downstream of Tup1p, a transcription factor regulating fungal filamentation, in experimental Candida albicans keratitis.

methods. Corneas of BALB/c mice were scarified and topically inoculated with 10⁵ or 10⁶ colony-forming units (CFU) of a wild-type human isolate of C. albicans (SC5314), a mutant strain with a transposon-induced homozygous disruption of the rbt1 gene (Tn7-rbt1), its control (DAY286), homozygous rbt knockout mutants deficient in rbt1 (BCa7-4) or rbt4 (BCa11-3), or their parental control (CAF2-1). Eyes were scored daily for clinical severity of fungal keratitis and were examined histopathologically.

results. With a 10⁵ CFU inoculum, the CAF2-1 control and its mutant derivatives (BCa7-4 and BCa11-3) produced significantly lower keratitis scores than did the moderately severe keratitis induced by control strains SC5314 and DAY286 and the Tn7-rbt1 mutant (P < 0.05). At a 10⁶ CFU inoculum, all strains induced severe disease except for the rbt4-deficient mutant. Fungal keratitis caused by Tn7-rbt1 was as severe as that of control strains (P > 0.2), and the BCa7-4 mutant initially caused severe disease that gradually waned (P < 0.02). However, the BCa11-3 mutant produced moderate disease that was significantly less severe than that induced by control strains (P < 0.04) and resolved within 1 week.

conclusions. The rbt4 gene of C. albicans is a potential virulence factor in posttraumatic corneal infection. Genetically regulated hyphal morphogenesis appears to be involved in the initial pathogenesis of experimental keratomycosis.

The yeast Candida albicans is a normal commensal and a pathogenic opportunist. Virulence factors of C. albicans infection include adhesion to host cells, ¹ secretion of enzymes, ² and phenotypic transformation from yeasts to hyphae. ^{3 4 5} A gene controlling the morphogenic switch to hyphal forms encodes a transcription factor involved in repression of filamentous growth, identified through sequence identity with the Saccharomyces cerevisiae tup1 gene. ⁶ Deletion of tup1 in C. albicans results in constitutive growth of pseudohyphae that do not injure or invade host cells. ^{6 7}  

Genes downstream of tup1 are designated rbt (repressed by Tup1p). ⁸ The rbt1 gene encodes a cell-wall protein involved in fungal adherence, and the rbt4 gene codes for a hyphally expressed protein. ^{9 10} Mutations in these rbt genes attenuate the severity of systemic candidiasis in experimental animals. ⁸ Braun et al. ⁸ also suggested a role for rbt genes in fungal keratitis using a rabbit model. We evaluated these genes as potential virulence factors in a murine model of posttraumatic keratomycosis. ¹¹  

We examined disease severity induced by various C. albicans yeast strains, including a wild-type strain (SC5314), a homozygous mutant in the rbt1 gene from a transposon-generated mutant library (Tn7-rbt1), ¹² its parental control (DAY286), a homozygous rbt1 knockout (BCa7-4), a homozygous rbt4 knockout (BCa11-3), and their parental control (CAF2-1). ⁸ Using an in vivo model that offers quantifiable assessment of fungal keratitis, our results indicate that the rbt4 gene contributes to the initiation of corneal infection by C. albicans. 

Materials and Methods

Fungi

The C. albicans yeast strains used in this study include SC5314, DAY286, Tn7-rbt1, CAF2-1, BCa7-4, and BCa11-3 (Table 1) . SC5314 is a wild-type, virulent C. albicans strain isolated from a human. ¹³ DAY286 is the parental control for the Tn7-rbt1 (rbt1 ^−/−) mutant. ¹² CAF2-1 is the parental control strain ¹⁴ used to generate the BCa7-4 (rbt1 ^−/−) and BCa11-3 (rbt4 ^−/−) knockout mutant strains. ⁸ C. albicans strains were routinely grown on YPD (1% yeast extract, 2% peptone-tryptone, and 2% dextrose) medium at 30°C. The yeasts were harvested for inocula in their exponential growth phase and diluted with sterile phosphate-buffered saline (PBS). 

In Vitro Growth Kinetics

Proliferation kinetics for each strain was measured in either YPD medium or SD medium (0.67% Difco yeast nitrogen base without amino acids and 2% dextrose; Difco, Detroit, MI) lacking uracil supplementation (SD-ura) at 33°C. Cultures were incubated at 33°C to mimic the temperature of the cornea. Each C. albicans strain was grown overnight in 5 mL of either YPD or SD-ura liquid medium. The concentration was estimated by using a spectrophotometer to measure the optical density (OD) at a wavelength of 600 nm and a conversion factor of one OD₆₀₀ unit equivalent to 3 × 10⁷ colony-forming units (CFU)/mL. ¹¹ Three cultures were analyzed for each yeast strain. For growth in YPD, 10³ CFU of yeast was added to 5 mL YPD liquid medium and incubated at 33°C with continuous shaking for 2 days. For growth in SD-ura, 10⁵ CFU was added to 5 mL SD-ura medium and incubated at 33°C with continuous shaking for 2 days. Concentrations were measured during the exponential phase of growth at 9, 24, 32, and 48 hours after inoculation (PI). Doubling times for each culture were calculated from the plotted data, and the mean for each strain was reported. Statistical analyses of the results are based on Student’s t-test. 

Animals

Immunocompetent adult female BALB/c mice (Harlan Sprague-Dawley, Houston, TX) 6 to 8 weeks old were anesthetized, and their corneas superficially scarified as described previously. ¹¹ The scarified cornea received a 5-μL inoculum of each C. albicans yeast strain that totaled either 10⁵ or 10⁶ CFU or 5 μL of sterile PBS for the negative control. Animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the research protocol was approved by the Baylor College of Medicine Institutional Animal Care and Use Committee. 

Clinical Scoring

Mice were monitored daily for up to 8 days PI and were scored for disease severity with the aid of a dissecting microscope. ¹¹ The same investigator estimated keratitis severity scores daily throughout 8 days of follow-up without consideration of experimental group, using written descriptions of severity categories and photographic examples of mild, moderate, and severe keratitis. Co-investigators masked to treatment group confirmed categorization by reviewing digital images of representative eyes infected with different fungal strains. A grade of 0 to 4 was assigned for three criteria (area of opacity, density of opacity, and surface regularity) to yield a possible total score of 0 to 12. The area of corneal opacification was graded 1 if the nontranslucent zone was ≤25% of the cornea, 2 for 26% to 50% involvement, 3 for 51% to 75%, and 4 for >75%. The density of a visible corneal opacity was graded 1 if a cloudy zone was barely detectable, 2 for an observable haze obscuring iris details, 3 for a marked opacity blurring the iris, and 4 for an opaque leukoma obstructing view of the iris or pupil. Corneal surface irregularity was scored 1 for a slightly distorted surface, 2 for mild changes in corneal thickness, 3 for moderate ulceration, and 4 for a descemetocele or perforation. A normal-appearing cornea received a score of 0 in each category. A total score of ≤5 was categorized as mild eye disease, 5 to 9 was considered moderate, and >9 was severe. Three to five mice were used for each strain and at each concentration. Score results at each day PI were compared by the Kruskal-Wallis one-way analysis of variance. 

Histology

Mice used for histologic examination were killed, and the globes were enucleated 1 day PI. The eyes from C. albicans–infected and mock-infected mice were formalin-fixed, paraffin-embedded, and sectioned at a thickness of 8 μm for histologic study. The sections were deparaffinized and stained with periodic acid-Schiff (PAS; Sigma-Aldrich, St. Louis, MO) or modified Grocott’s methenamine silver (GMS; Richard-Allan Scientific, Kalamazoo, MI). 

Results

The doubling time during the exponential growth phase of each strain was significantly faster (P < 0.001) in YPD than in SD-ura media (Table 1) . In YPD, the doubling times for control and mutant strains were similar to those of wild-type (P > 0.6). When strains were grown in SD-ura media, growth rates of both parental controls were similar to those of wild-type (P = 0.2), whereas both rbt1 ^−/− mutants grew more slowly compared with their respective controls, DAY286 (P = 0.05) and CAF2-1 (P = 0.003). BCa11-3, the rbt4 ^−/− mutant, grew more slowly than wild-type (P = 0.02) but similar to its parental control (P = 0.2). 

Keratitis severity scores induced by the three rbt mutants were compared with their respective parental controls at a 10⁵ CFU inoculum (Fig. 1) . Tn7-rbt1 and its parental control DAY286 had similar scores at each day (P > 0.1). Compared with CAF2-1, no significant difference was observed with BCa7-4 (P > 0.7) or BCa11-3 (P > 0.2). At a 10⁶ CFU inoculum (Fig. 2)keratitis severity was similar for DAY286 relative to SC5314 (P > 0.4). Tn7-rbt1 scores remained similar to DAY286 at an inoculum of 10⁶ CFU (P > 0.4). BCa7-4 slowly decreased in severity compared with CAF2-1. For the first 4 days after inoculation, a significant difference was not found between these two strains (P > 0.1), but keratitis scores of BCa7-4 were less than CAF2-1 (P = 0.05) and had a threefold reduction by the eighth day (P = 0.02). BCa11-3 also differed from its parental control at 10⁶ CFU on the first day after inoculation, and disease scores remained significantly lower throughout the next week (P < 0.04). At an inoculum dose of 10⁶ CFU, BCa7-4 and Tn7-rbt1 produced similar disease scores as their parental controls at each day (P > 0.1). 

The dose–response relationship between 10⁵ CFU and 10⁶ CFU inocula was also compared for each strain. Fungal keratitis was generally more severe at 10⁶ CFU compared with 10⁵ CFU, although the relative increase between the two concentrations varied among yeast strains. Keratitis scores for CAF2-1 were as much as eight times lower than SC5314 at 10⁵ CFU but were similar to those of SC5314 at 10⁶ CFU. BCa7-4 and BCa11-3 mutants had increased relative severity for the first 4 days between the two inocula, and this relative increase in scores was higher than that for wild-type. Beyond 5 days PI, the relative difference in scores from 10⁵ CFU to 10⁶ CFU for the knockout mutants decreased. 

Histologic examination of infected eyes harvested 1 day PI revealed budding yeasts. Hyphal forms were present in the cornea for strains SC5314 (Fig. 3A) , Tn7-rbt1, and BCa7-4. Both parental control strains (DAY286 and CAF2-1) produced similar fungal morphology and invasion as did the wild-type. Some pseudohyphae were seen in BCa11-3-infected animals but were limited to the superficial cornea (Fig. 3B) . Neutrophilic infiltration and stromal thickening were prominent in corneas infected with control and rbt1-deficient strains, whereas the rbt4-deficient strain induced minimal corneal inflammation. Mock-infected eyes did not show signs of infection or inflammation but did have microscopic wounding changes from the scarification procedure. 

Discussion

The emerging pathogenomics of C. albicans enables the identification of fungus-specific genes contributing to infection. ¹⁵ Virulence factors are becoming known for disseminated and mucocutaneous candidiasis, ¹⁶ but few studies have examined the molecular pathogenesis of fungal ocular infection. Recently, homozygous mutants for Tup1p-regulated genes rbt1 and rbt4 were shown to attenuate virulence in a mouse model of systemic candidiasis and in a rabbit model of C. albicans keratitis. ⁸ We examined these fungal rbt1 and rbt4 mutants in a mouse keratitis model and also studied a transposon-induced rbt1 homozygous mutant. 

C. albicans strains lacking rbt1 had disease severity similar to that of parental controls at a lower fungal inoculum, suggesting that this gene may not be an important determinant of pathogenicity in murine corneal infection. At a higher fungal inoculum we confirmed that a rbt1 knockout mutant had partial attenuation for producing corneal disease but found that a rbt1-deficient transposon mutant remained virulent. A fungal strain deficient in rbt4 had significantly less severe corneal infection at both inoculum concentrations and showed partial attenuation in virulence. Histopathologically, the rbt4 ^−/− strain also appeared relatively avirulent. 

These findings suggest that the rbt4 gene of C. albicans contributes to the pathogenesis of posttraumatic corneal infection. The rbt4 gene codes for a hyphally expressed, secreted protein that is hypothesized to damage host cells. ⁸ Rbt4 is induced in the presence of serum and by depletion of the tup1 gene product. ⁸ The link between rbt4 and fungal filamentation complements our previous findings that hyphal morphogenesis plays a key role in the early events of fungal keratitis after exposure to C. albicans. ¹⁷  

A secondary effect apparently modulated keratitis severity produced by the knockout and parental constructs used in this study, and differences in fungal genotypes could explain why rbt1-deficient strains varied in relative disease severity at two inoculum concentrations. One parental strain produced keratitis similar to wild-type at both inocula, whereas the other was similar to wild-type only at the higher inoculum. Both strains were derived from the wild-type isolate by using a technique that makes strains heterozygous for the ura3 gene. ¹⁴ The level of ura3 gene activity does not influence the process of germ tube formation but does lower the growth rate. ¹⁸ Furthermore, auxotrophic C. albicans strains have a lower level of pathogenicity in mice, depending on the location of the ura3 gene. ¹⁹ Our in vitro assay for proliferation kinetics confirmed slower growth in uracil-deficient medium. Decreased gene expression and reduced growth rate may lessen virulence in an in vivo model when relatively low concentrations of these strains are used. 

An additional finding of this study was the dose–response relationship of experimental C. albicans keratitis. At a lower inoculum, the CAF2-1 control did not produce keratitis as severe as the wild-type, SC5314. However, at a higher inoculum, CAF2-1 caused disease similar in severity to that caused by SC5314. Our results demonstrate the importance of the fungal inoculating dose when evaluating disease severity. 

Keratitis models offer a graded and nonlethal outcome for dissecting fungal virulence pathways. ^{20 21} The use of mutant microbial genotypes is a valuable tool for providing insight into the molecular pathogenesis of microbial infection of the cornea. Because virulence factors may differ by infection site, tissue-specific studies are needed to understand the mechanisms of ocular fungal infections. 

 Supported by Institutional Research Training Grant EY07001 and Core Grant EY02520 from the National Eye Institute, Research to Prevent Blindness, Inc., the Lions Eye Bank Foundation, and the Sid W. Richardson Foundation.

 Submitted for publication August 2, 2006; revised January 9 and February 21, 2007; accepted April 20, 2007.

 Disclosure: B.E. Jackson, None; B.M. Mitchell, None; K.R. Wilhelmus, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Genotype and Growth Rate of Candida albicans Strains

Strain Name	Strain Type	Genotype	Doubling Time ± SD			Reference
							YPD (h)	SD-ura (h)*
SC5314	ura3 +/+	Wild-type	1.60 ± 0.09	2.22 ± 0.08	Gillum et al. 13
DAY286	ura3 −/+	ura3::λimm434 his1::hisG pARG4::URA3::arg4::hisG,	1.61 ± 0.08	2.31 ± 0.06	Davis et al. 12
		ura3::λimm434 his1::hisG arg4::hisG
CAF2-1	ura3 −/+	ura3::λimm434	1.56 ± 0.06	2.34 ± 0.10	Fonzi and Irwin 14
		URA3
Tn7-rbt1	rbt1 −/−	ura3::λimm434 his1::hisG arg4::hisG rbt1::URA3	1.65 ± 0.11	2.51 ± 0.11, †	Davis et al. 12
	ura3 −/+	ura3::λimm434 his1::hisG arg4::hisG rbt1::ARG4
BCa7-4	rbt1 −/−	ura3::λimm434 rbt1::hisG-URA3-hisG	1.62 ± 0.06	2.72 ± 0.03, †	Braun et al. 8
	ura3 −/+	ura3::λimm434 rbt1::hisG
BCa11-3	rbt4 −/−	ura3::λimm434 rbt4::hisG-URA3-hisG	1.61 ± 0.11	2.43 ± 0.05	Braun et al. 8
	ura3 −/+	ura3::λimm434 rbt4::hisG

 SD, standard deviation; YPD, yeast peptone dextrose medium; SD-ura, synthetic dextrose without uracil medium.

*  For each fungal strain, in vitro growth in SD-ura medium was significantly slower (P < 0.001) than in YPD medium.

†  Strains that grew significantly more slowly (P ≤ 0.05) in SD-ura medium than the respective parental control.

Figure 1.

 

Mean mouse keratitis scores after inoculation with 10⁵ CFU C. albicans. Each yeast strain was inoculated onto scarified corneas of BALB/c mice at day 0 and monitored for 8 days. Each day, the mouse eyes were scored for the area, density, and depth of ulcerative keratitis. SC5314 is a wild-type yeast strain. DAY286 is the mutant control for the Tn7-rbt1 strain, and CAF2-1 is the parental control for BCa7-4 and BCa11-3. All mock-inoculated eyes had a zero score for each time point.

Figure 2.

 

Mean mouse keratitis scores after inoculation with 10⁶ CFU C. albicans. Corneas were scarified and inoculated on day 0. Eyes were scored daily for 8 days.

Figure 3.

 

Histopathology of murine keratomycosis. Keratitis at 1 day after infection with 10⁶ CFU of C. albicans, stained with Grocott’s methenamine silver. (A) In eyes infected with wild-type C. albicans, fungal hyphae (arrows) extended through the epithelium and into the anterior stroma, associated with intense leukocyte infiltration in the cornea and anterior chamber. (B) In eyes exposed to a rbt4-knockout strain, very little yeast was found within the epithelium, and minimal inflammation occurred. Original magnification, ×100.

The authors thank Aaron P. Mitchell, PhD, Columbia University, New York, and Michael C. Lorenz, PhD, University of Texas, Houston, for providing the DAY286 and Tn7-rbt1 strains and Alexander D. Johnson, PhD, University of California, San Francisco, for providing the CAF2-1, BCa7-4, and BCa11-3 strains.