|Home | About | Journals | Submit | Contact Us | Français|
Candida albicans is the most common cause of invasive fungal infections in humans. It is unclear how C. albicans escapes from phagocytic attack and survives in the hostile blood environment during life-threatening systemic infections. Using a conditional overexpression or suppression genetic strategy, we discovered that HYR1 gene reduced phagocytic killing activity of C. albicans in vitro and increased tissue fungal burden in vivo. Concordant with its positive regulation by the transcription factor Bcr1p, autonomous expression of HYR1 complemented the hypersusceptibility to phagocyte-mediated killing of a bcr1 null mutant of C. albicans in vitro. As for C. albicans, heterologous expression of HYR1 in Candida glabrata rendered the organism more resistant to neutrophil killing activity. Vaccination with a recombinant Hyr1p significantly protected mice against hematogenously disseminated candidiasis (P = .001). Finally, anti-rHyr1p polyclonal antibodies enhanced mouse neutrophil killing activity by directly neutralizing rHyr1p effects in vitro. Thus, Hyr1 is an important virulence factor for C. albicans, mediating resistance to phagocyte killing. Hyr1p is a promising target for vaccine or other immunological or small molecule intervention to improve the outcomes of disseminated candidiasis.
Approximately 60,000 cases of disseminated candidiasis per year occur in the United States , resulting in billions of dollars of healthcare expenditures. Given the 40% mortality rate of such infections, there is a need to identify new prophylactic or therapeutic targets for intervention.
The primary host defense mechanism against disseminated candidiasis is phagocytic killing of the organism [2, 3]. Only phagocytic cells are capable of directly killing Candida in vitro . Additionally, within 30 min of intravenous inoculation of Candida in mice, rabbits, dogs, or humans, yeasts are retained within the reticuloendothelial system, especially in the liver [5–9]. The liver, rich in Kupffer macrophages, is capable of clearing 99.9% of yeast in the portal system during a single pass , underscoring the effectiveness of phagocytic defense mechanisms against the fungus. Hence, resistance of Candida albicans to phagocyte killing is an important virulence function of the organism.
Cell surface glycosyl phosphatidylinositol (GPI)-anchored -proteins are at the critical interface between pathogen and host, making these proteins likely participants in host-pathogen interactions . Furthermore, these proteins are the first targets encountered by host defense mechanisms, which make them attractive vaccine candidates. By screening a conditional overexpression or suppression system focusing on GPIanchored proteins in C. albicans, we identified HYR1 as a potential virulence factor. HYR1 is a hyphae coexpressed gene, the null mutant strain of which does not display any morphologic abnormality in vitro . Here we report that HYR1 mediates resistance to phagocytic killing in vitro, modulates tissue fungal burden in vivo, and is a potential vaccine target to ameliorate the severity of disseminated candidiasis.
To generate a conditional HYR1 expression strain, an HIS1-TR promoter cassette  was inserted in front of 1 allele of the HYR1 gene of strain THE4, which yielded strain CAAH. The URA3 at the HIS1 locus in strain CAAH was looped out, which generated CAAH-1. The second allele of HYR1 in CAAH-1 was disrupted by a recyclable URA3 cassette, which generated strain CAAH-2, followed by looping out of URA3, which yielded strain CAAH-3. A 3.9-kb Nhe I-Pst I fragment containing the URA3-IRO1 gene was inserted into its original locus on the CAAH-3 genome, which yielded CAAH-31. Primers used are listed in Table 2.
The semiquantitative reverse-transcription polymerase chain reaction (RT-PCR) used for detection of gene expression in vitro was described previously. Primers used to detect EFB1 expression were EFB1a and EFB1b; primers used to amplify HYR1 were HYR1 specific1 and HYR1 specific2 (Table 2). To study the impact of neutrophils on C. albicans HYR1 expression, 1 × 106overnight cells of SC5314 grown in yeast peptone dextrose (YPD) were either cocultured with 1 × 107HL-60 derived neutrophils or cultured alone in Roswell Park Memorial Institute (RPMI) 1640 medium, plus 10% pooled human serum. Samples were obtained at 30-min intervals for 3 h until RNA was extracted, and semiquantitative RT-PCR was performed.
Human or mouse neutrophils were isolated, HL-60 cells were differentiated into neutrophils or macrophages, and the phagocyte killing assay was performed as described elsewhere [12–14]. Briefly, phagocytes were incubated with fungi for 1 h and then sonicated and quantitatively cultured. Percentage of killing was calculated by dividing the number of fungal colonies after coincubation with phagocytes by the number of fungal colonies incubated with media without phagocytes. Human neutrophils and HL-60 derived neutrophils or macrophages were tested at a 2:1 and 20:1 phagocyte:fungus ratio, respectively. For bcr1 and related mutants, the blastospores were pregerminated for 40 min in RPMI 1640 medium, plus 10% fetal bovine serum (FBS), at 37°C before performing the assay.
Candida glabrata BG14 was transformed with either an HYR1 expression vector pGRB2.2-HYR1 or an empty control plasmid pGRB2.2 . The HYR1 coding sequence was amplified by CG-Hyr1-a and CG-Hyr1-b (Table 2) and was cloned into Xba I and Xho I sites of pGRB2.2 by using In-Fusion 2.0 Dry-Down PCR Cloning Kit, per the manufacturer’s instructions (Clontech Laboratories).
HYR1 expression by wild-type C. albicans SC5314 was evaluated during hematogenously disseminated candidiasis, as described. Brains, livers, lungs, kidneys, and spleens of BALB/C mice were collected 6 and 24 h after infection. Reverse transcription was performed with RETROscript (Ambion) using primers listed in Table 2. For amplification of the mouse G3PDH housekeeping gene, primers G3PDHF and G3PDHR were used. For detection of HYR1 and EFB1 of C. albicans,2 rounds of polymerase chain reaction (PCR) were performed. Round 1 used the outer primer set (EFB1F and EFB1R for EFB1, or P2 and P5 for HYR1); round 2 used an aliquot (1 mL) of the round 1 PCR product as a template. The inner primer sets were as follows: EFB1nF and EFB1nR (for EFB1), or P2 and P4 (for HYR1). All PCR conditions were as follows: denaturing at 95°C for 2 min and amplification of 35 cycles at 94°C for 30 s (denaturing), 55°C for 30 s (annealing), and 72°C for 90 s (extension). For quantitative RT-PCR, complementary DNA was prepared as described above. Optimization of amplification efficiency and real-time RT-PCR SYBR green assays were performed as described elsewhere . Constitutively expressed ACT1 was used as a control for all reactions. Calculations and statistical analyses were performed as described in ABI PRISM 7000 Sequence Detection System User Bulletin 2 (Applied Biosystems).
Mice were given water with or without doxycycline (2 mg/mL) dissolved in 5% sucrose solution throughout the period of the experiment starting from day -3 relative to infection  and were given food and water ad libitum. Tissue fungal burden was performed using a modification of our previously described method . Briefly, mice were infected through tail vein. Kidneys, brain, and liver were harvested 1 day after infection, homogenized, serially diluted in 0.85% saline, and quantitatively cultured on YPD that contained 50 μg/mL chloramphenicol. Values were expressed as log colony-forming units per gram of tissue. All procedures involving mice were approved by the institutional animal use and care committee, following National Institutes of Health guidelines.
rHyr1p-N (from amino acids 25–350 of Hyr1p) was produced in an Escherichia coli pQE-32 expression system (Qiagen), and the 6XHis tagged protein was purified as described elsewhere , with the exception of using a HisPur Cobalt resin (Thermo Scientific) affinity column. Endotoxin was removed from rHyr1p-N ny using Detoxi-Gel Endotoxin Removing Columns (Thermo Scientific), and the endotoxin level was determined with Limulus Amebocyte Lysate endochrome (Charles River) per manufacturer’s instruction. Using this procedure, endotoxin was reduced to < 0.28 EU per dose used for vaccination.
All vaccinations were subcutaneous, at the base of the neck. Eight juvenile (10–12-week) C57BL/6 mice were vaccinated with 20 μ g of affinity-purified rHyr1p-N in complete Freund’s adjuvant and boosted in incomplete Freund’s adjuvant (IFA) at 3 weeks. Eight additional juvenile mice received adjuvant alone mixed with the preparation produced from E. coli cells transformed with the empty plasmid. Fourteen days after the boost, mice were infected via the tail vein with 5 × 105cells of wild-type C. albicansSC5314 .
The efficacy of rHyr1p-N in protecting against hematogenously disseminated candidiasis was also evaluated using alum (2% Alhydrogel; Brenntag Biosector), an adjuvant approved by the Food and Drug Administration (FDA) for use in humans. Additionally, to determine that rHyr1p-N was protective against other strains of C. albicans, we used another clinical isolate, strain 15563. For these experiments, 33 μg of affinity-purified rHyr1p-N was mixed with 0.1% alhydrogel and administered to BALB/c mice as above on day 0, boosted on day 21, and then infected on day 35 with C. albicans through tail vein injection. For all vaccination experiments, survival of mice for 35 days after infection was used as an end point.
Indirect immunofluorescence was performed using polyclonal anti-Hyr1p antisera generated by immunization of mice with rHyr1p-N. An inoculum of 1 × 107blastospores of hyr1 null mutant were incubated in RPMI 1640 medium for 90 min at 37°C and pelleted twice to absorb the antiserum.
C. albicans blastospores (1 × 105) were pregerminated in RPMI 1640 medium for 90 min at 37°C and transferred into a 4-well chamber slide (Nalge Nunc International). After incubation at 4°C for 30 min, the cells were blocked with 300 μ L of 1.5% goat serum, stained with polyclonal antiserum at a 1:100 dilution of phosphate-buffered saline as a negative control, and then by fluorescein isothiocyanate-labeled goat anti-mouse immunoglobulin G (IgG) at 1:200. The cells were imaged with confocal scanning laser microscopy .
Pooled anti-Hyr1p or control serum was collected from 5 mice that were vaccinated either with rHyr1p-N or with the preparation produced from E. coli cells transformed with the empty plasmid. The total IgG from both sera was isolated using Nab Spin Kit (Thermo Scientific). The F(ab)12 fragments were purified with Pierce F(ab)12 Preparation Kit according to the manufacturer’s instruction.
Candida cells were opsonized on ice for 45 min with 5% normal mouse serum (Santa Cruz Biotechnology) or 5% normal mouse serum plus 5% F(ab)12 prepared from either rHyr1p-N vaccinated or control mice IgG before mixing with mouse neutrophils. Mouse neutrophil killing assay was described as above.
Phagocyte-mediated killing activity and tissue fungal burdens among different groups were compared by using the Mann-Whitney U test for unpaired comparisons, as appropriate. The nonparametric log-rank test was used to determine differences in survival times. P values of < .05 were considered to indicate significant differences.
To study the function of HYR1, we constructed a conditional overexpression or suppression strain of C. albicans, CAAH-31. In CAAH-31, 1 allele of HYR1 was controlled by the tetracycline-regulated-promoter and the other allele was disrupted. By semiquantitative RT-PCR, we confirmed that HYR1 was abundantly expressed when blastospores of CAAH31 were grown in media without doxycycline, and it was not detected in the presence of doxycycline (Figure 1A). As expected, HYR1 was not detected in wild-type (THE31) blastospores, because HYR1 is a hyphal coexpressed gene (Figure 1A).
The HYR1 conditional overexpression strain, CAAH-31, and THE31 wild-type control strain had identical growth rates, regardless of the presence or absence of doxycycline (doubling time for wild-type control strain without doxycycline, 1.51 ± 0.29 h, and with doxycycline, 1.51 ± 0.38 h; doubling time for CAAH-31 strain without doxycycline, 1.39 ± 0.30 h, and with doxycycline, 1.35 ± 0.19 h). We also evaluated the impact of HYR1 overexpression on the normal accumulation of other GPI-anchored proteins on the cell surface. By direct immunofluorescence, we confirmed that HYR1 overexpression had no effect on the accumulation of the GPI-anchored protein Als1p (data not shown) .
During routine screening for virulence-associated phenotypes, we determined the impact of conditional expression of HYR1 on candidal killing by human phagocytes. HYR1-expressing C. albicans (CAAH-31, without doxycycline) was significantly more resistant to human neutrophil-mediated killing than was wild-type C. albicans (which does not express HYR1 in the blastospore phase) (P < .001) and HYR1-suppressing C. albicans (CAAH-31, with doxycycline) (P < .001) (Figure 1B). This phenotype was not due to doxycycline, because killing was not significantly different between the wild-type control strain and HYR1-suppressing C. albicans (with doxycycline).
We also performed candidal killing assays using the HL-60 cell line, which can be differentiated into either neutrophil-like cells or macrophage-like cells [13, 14]. Similar to that of freshly harvested human neutrophils, conditional expression of HYR1 reduced killing of C. albicans blastospores by both HL-60-derived neutrophils (Figure 1C) and macrophages (Figure 1D) in vitro.
Because HYR1 is a downstream gene of the positive transcription regulator Bcr1p , we hypothesized that disruption of bcr1 would exacerbate susceptibility to neutrophil killing under conditions promoting wild-type C. albicans to express HYR1 (ie, during hyphal formation). We therefore induced C. albicans to form germ tubes by incubating the cells in RPMI 1640 medium, plus 10% fetal bovine serum, at 37°C for 40 min. This condition is known to induce expression of HYR1  and resulted in germ tubes short enough that extensive hyphae were not formed, thereby enabling quantification of colony-forming units in our kill assay. We compared neutrophil killing of the bcr1 null mutant (CJN702), a BCR1 complemented (CJN698), and a wild-type C. albicans strain (DAY185). The bcr1 null mutant was hypersusceptible to neutrophil-mediated killing compared with the BCR1-complemented and wild-type control strains (Figure 2A). Furthermore, the hypersusceptibility to killing of bcr1- deficient C. albicans was fully complemented by autonomous expression of HYR1 in the bcr1 mutant but not by other cell surface encoding genes regulated by Bcr1p  (Figure 2A and 2B).
To further define the virulence phenotype mediated by HYR1 and to reduce the background, we expressed the gene heterologously in C. glabrata BG14  using a plasmid pGRB2.2 carrying a constitutive PGK1 promoter . We also generated a C. glabrata BG14 transformed with the empty plasmid, as a negative control. Expression of HYR1 in C. glabrata resulted in a 75% reduction in killing by HL-60-derived neutrophils in vitro, compared with the C. glabrata transformed with an empty plasmid (Figure 2C).
Because conditional expression of HYR1 conferred Candida blastospores resistant to neutrophil-mediated killing, we used RT-PCR to study the expression of wild-type C. albicans HYR1 in response to HL-60-derived neutrophils in vitro. HYR1 was expressed as early as 30 min after exposure to medium containing serum and maintained high expression for 2.5 h during culture (Figure 3A). However, when C. albicans was exposed to HL-60-derived neutrophils in culture, even in the presence of serum, HYR1 expression was inhibited for up to2hin the coculture (Figure 3A).
To determine whether C. albicans expressed HYR1 during hematogenously disseminated candidiasis, 5 major organs- brain, liver, lung, spleen, and kidney-were harvested from mice infected with C. albicans wild-type strain 6 and 24 h after infection. An improved nested-RT-PCR assay  was used to assess HYR1 expression in vivo. HYR1 expression was detected in all 5 organs (Figure 3B).
Overexpression of HYR1 increased fungal burden in the liver and spleen and suppression of HYR1 decreased it (Figure 4A). Furthermore, we confirmed that the overexpressing strain had significantly higher levels of HYR1 than did the wild-type strain in livers; the suppressing strain demonstrated a trend toward reduced levels of HYR1 expression compared with the control strain (Figure 4B). In contrast, HYR1 expression did not significantly alter fungal burden in the kidney (data not shown), an organ lacking resident phagocytes .
On the basis of sequence analysis, Hyr1p is predicted to be a cell surface protein . To confirm this, we generated a recombinant N-terminus of Hyr1p in E. coli transformed with an expression clone that includes amino acids 25–350 of the coding sequence (rHyr1p-N). Serum from mice immunized with rHyr1p-N was preabsorbed against an hyr1 null mutant of C. albicans , followed by indirect immunostaining on wild-type hyphae. We found that the cell wall of wild-type C. albicans hyphae was heavily stained (Figure 5), confirming that it is cell-surface-expressed and hence exposed to the immune system.
Because Hyr1p is a cell surface protein, which confers resistance to candidal killing by phagocytes, we sought to determine its potential as a vaccine candidate. Mice were vaccinated with rHyr1p-N plus adjuvant or adjuvant alone. Vaccination with rHyr1p-N mixed with complete or incomplete Freund’s adjuvant or alum markedly improved survival of mice compared with those vaccinated with either adjuvant alone (Figure 6A and 6B).
The protective effect of the rHyr1p-N vaccine suggested that the anti-rHyr1p serum might be able to neutralize the protective function of Hyr1p in C. albicans. To determine whether antirHyr1p antibodies could directly inhibit Hyr1p function, we isolated and prepared F(ab)12 fragments from total IgG of mice immunized with either rHyr1p-N or the control (preparation produced from E. coli cells transformed with the empty plasmid). We found that F(ab)12 from immune but not control serum was able to restore neutrophil killing of the HYR1 conditional expressing strain to levels equivalent to that of the suppressing strain (Figure 6C).
In this study, we demonstrated that HYR1, a hyphal coexpressed gene [11, 25], encodes a candidal phagocyte resistance factor. Conditional expression of HYR1 in Candida blastospores caused the fungus to be more resistant to killing by phagocytes compared with wild-type C. albicans blastospores. Additionally, the function of HYR1 in C. albicans was recapitulated by heterologously expressing the gene in C. glabrata in vitro. We also found that if a strain was deficient in Bcr1p, a transcription factor that positively regulates HYR1 expression , it exhibited enhanced susceptibility to phagocyte-mediated killing. The hypersusceptibility to phagocyte killing of the bcr1 null mutant was fully complemented by autonomously expressed HYR1, but not other genes that encode GPI proteins positively regulated by Bcr1p. Hence, HYR1 is a downstream gene of BCR1 in a phagocyte killing resistance pathway.
It is interesting that HL-60-derived neutrophils were able to inhibit the expression of HYR1 in wild-type C. albicans during the initial contact between the phagocytes and Candida. Thus, a dynamic interaction occurred between host phagocytes and C. albicans, in which the phagocytes mediated a delay in the expression of a phagocyte-resistance gene in the wild-type fungus. This is consistent with an earlier finding that human neutrophils delayed the formation of C. albicans hyphae and the expression of hyphae coexpressed genes .
Resistance to phagocyte killing is a complex phenotype that is likely attributable to multiple factors. Phagocytes can kill Candida extracellularly or intracellularly. Recently, C. albicans cell surface superoxide dismutases were characterized as virulence factors that help the fungi escape from phagocyte killing by degrading host-derived, highly reactive oxygen species . Neutrophils typically attach to and spread over the surfaces of hyphal forms of fungi, because extended hyphae are too large for phagocytes to ingest completely. Because C. albicans expresses HYR1 in the hyphal form, it is possible that Hyr1p contributes to resistance to phagocyte killing by preventing surface contact to the phagocyte. Alternatively, Hyr1p might interfere with oxidative or nonoxidative killing mechanisms of phagocytes. These possibilities are currently the focus of ongoing research.
The in vitro phenotype of HYR1 overexpression was recapitulated in vivo. During murine-disseminated infection, overexpression of HYR1 led to a significant increase (P < .002), and suppression of HYR1 led to a significant decrease (P < .03), in tissue fungal burden compared with the wild-type strain in organs with resident phagocytes. The lack of phenotype in the kidney likely reflects the fact that kidneys do not have resident phagocytes, and that neutrophil influx into the kidneys during lethal disseminated candidiasis does not occur until 124 h of infection . Nevertheless, vaccination with rHyr1p-N resulted in considerable protection against hematogenously disseminated candidiasis. The efficacy seen for the rHyr1p-N vaccine appeared to be greater than that previously seen in mice vaccinated with the rAls3p-N, the latter of which is being advanced into clinical testing. Hence, rHyr1p-N is a promising vaccine candidate for disseminated candidiasis. Additional investigations into optimal immunization protocols are ongoing.
Our data also demonstrate that the mechanism of protection by rHyr1p-N vaccine was the direct inhibition of the Hyr1p neutrophil resistance function, rather than opsonization of the organism by anti rHyr1p-N antibodies. We used F(ab)12 fragments rather than whole antibodies to avoid the possibility of enhanced opsonophagocytosis by the immune serum, because F(ab)12 fragments’ lack of Fc receptor. Nevertheless, the rHyr1p-N induced immune fragments but not control F(ab)12 fragments restored neutrophil killing of the fungus, which indicated direct antibody-mediated neutralization of the virulence function of Hyr1p.
We also showed the advantage of using a conditional overexpression or suppression approach to explore potential virulence functions of a given gene. Large-scale forward genetic approaches to explore virulence genes in vitro require functional screening assays and are limited to screening for genes that are expressed while conducting the assay. When a gene is not expressed significantly in the wild-type strain under conditions used for the in vitro screening assay, the forced overexpression of the gene still allows the detection of a gain-offunction phenomenon in the assay, as in the case of HYR1 during Candida blastospore growth. When a gene is strongly expressed, conditional suppression of the gene yields a loss-offunction phenotype in the same assay. Furthermore, when overexpression and suppression are used simultaneously, the phenotype can be amplified, as in the case of the effect of HYR1 on liver and spleen fungal burdens in vivo. The ability to detect a phenotype by comparing gene overexpression and suppression enables conditional gene expression to overcome the limitations of functional redundancy, which particularly plague forward genetic approaches when members of a gene family are being studied.
In summary, we used a conditional gene expression approach to identify Hyr1p as a surface-expressed, virulence factor for C. albicans. HYR1 overexpression mediated the resistance to neutrophil killing in vitro and increased the tissue fungal burden in vivo. Finally, we demonstrated that HYR1 is a promising vaccine target that merits additional development as a prophylactic or therapeutic strategy for disseminated candidiasis.
We thank Dr Hironobu Nakayama for providing the TR-expression system, Dr Brendan Cormack for plasmid pGRB2.2, and Dr Al Brown for the hyr1 null mutant.
Financial support: Public Health Service (grants R21 AI066010 and R03 AI083251 to Y.F.; grant R01 AI067703 to A.P.M.).
Presented in part: 48th(Washington, DC, 25–28 October 2008; abstract M-1583) ticipants in host-pathogen interactions . Furtherand 49th(San Francisco, California, 12–15 September 2009; abstract G1–880) more, these proteins are the first targets encountered Interscience Conference on Antimicrobial Agents and Chemotherapy.