Rapid Screening Method for Compounds That Affect the Growth and Germination of Candida albicans, Using a Real-Time PCR Thermocycler

Abstract

We propose a screening method for compounds affecting growth and germination in Candida albicans using a real-time PCR thermocycler to quantify green fluorescent protein (GFP) fluorescence. Using P_ACT1-GFP and P_HWP1-GFP reporter strains, the effects of a wide range of compounds on growth and hyphal formation were quantitatively assessed within 3 h after inoculation.

TEXT

Candida albicans is a polymorphic, opportunistic human pathogen that can cause infections in immunocompromised individuals. The yeast-to-hypha switch (germination) is important in the virulence process (21) and biofilm formation (15, 18); therefore, compounds that inhibit germination are the subject of many studies. Germination can be quantified using standard microscopy-based approaches, but these methods tend to have observer bias. Alternatively, observer-independent approaches such as gene expression studies have been used. Ever since codon-optimized green fluorescent protein (GFP) became available for C. albicans (7), GFP has been used as a reporter for real-time gene expression within individual cells both for in vitro and in vivo models (2, 3, 6, 8, 19). The aim of the present study was to determine expression of the GFP gene in a real-time PCR thermocycler and to use this as a sensitive, rapid, high-throughput, and observer-independent approach to analyze compounds that affect growth and germination of C. albicans. To accomplish this, we used a set of reporter strains containing fusions of GFP with the promoters of the ACT1 (2) and HWP1 (20) genes.

C. albicans strains were grown on yeast nitrogen base (YNB) agar (pH 7.0) supplemented with 0.5% glucose at 30°C for 48 h and incubated overnight in 5-ml cultures at 30°C with shaking (150 rpm). C. albicans SC5314 (11) and CAI4-P ura3/URA3 (gift from B. Distel) were used as nonfluorescent controls. HB12 (provided by P. Sundstrom) is a CAI4-derived strain expressing GFP under the control of the HWP1 promoter (20) and is referred to as the P_HWP1-GFP strain for clarity. P_ACT1-GFP (2) was introduced into strain CAI4 (ura3/ura3) using electroporation (14). GFP expression in P_ACT1-GFP and P_HWP1-GFP strains was verified by fluorescence microscopy (Leica DM4000B) with a 40× objective using GFP filters. The P_HWP1-GFP strain showed GFP expression exclusively in hyphae (Fig. 1A and B), whereas the P_ACT1-GFP strain showed a similar level of fluorescence irrespective of growth morphology (Fig. 1C and D). The germination efficiency of both GFP reporter strains, determined as described before (12), was comparable to that of the SC5314 and CAI4-P strains (data not shown).

Fig. 1.

Light and fluorescence microscopic images of C. albicans grown under hypha-inducing conditions. (A to D) C. albicans P_HWP1-GFP strain by phase-contrast microscopy with hyphae, pseudohyphae, and yeast cells evident (A), whereas with fluorescence microscopy (B), only true hyphae are visualized. In contrast, the C. albicans P_ACT1-GFP strain visualized with both phase-contrast microscopy (C) and fluorescence microscopy (D) show all cell morphologies. All images are shown at the same magnification. Bar, 40 μm.

Cultured cells were washed once with phosphate-buffered saline (10 mM potassium phosphate, 150 mM NaCl [pH 7]) and resuspended in YNB to an optical density at 600 nm (OD₆₀₀) of 1. Serial dilutions in YNB were made to obtain cell suspensions of the final OD₆₀₀ used in experiments. Duplicate cell suspensions were placed in PCR tubes to a final volume of 100 μl. A MyiQ2 two-color real-time PCR detection system, iCycler thermal platform (Bio-Rad), was programmed to maintain a constant temperature at 37°C and measure the fluorescence intensity at time zero and every 15 min for a total of 4.5 to 6 h. The iQ5 optical system software (version 2.1) used to control the MyiQ2 system was set to analysis mode with a subtracted background. Raw data were copied into an Excel file, and the following calculations were made: baseline values at time zero were subtracted from the values at all subsequent time points, and the averages and standard deviations were calculated for duplicate samples. Additionally, a line chart or line graph was created for fluorescence intensity (in arbitrary units [AU]) versus time (in hours), and depending on the obtained pattern of fluorescence increase for a specific period of time, that time period was isolated for further analysis; this involved fitting a linear trend line and calculating the slope. Where applicable, autofluorescence of test compounds was corrected for by subtracting the fluorescence intensity without cells (autofluorescence) for each individual concentration.

The C. albicans P_ACT1-GFP strain showed immediate increasing fluorescence intensity in time for OD₆₀₀s equal to 1, 0.5, and 0.25 (Fig. 2A). The strongest increase of the fluorescence signal over time was detected at the highest inoculation density (Fig. 2B). Further analysis of induction and inhibition of growth for this strain was performed using an OD₆₀₀ of 0.5. For the C. albicans P_HWP1-GFP strain, a lag phase of approximately 1 h was observed before there was a detectable increase in fluorescence over time (Fig. 2C). This reflected the induction of gene expression and the time required for GFP to fold into its active conformation; after this point the graphs increased linearly over almost the complete assay period. The strongest increase of fluorescence signal over time was detected at an OD₆₀₀ of 0.06 (Fig. 2D), and in contrast to P_ACT1-GFP, this did not occur at the highest inoculation density. This reflects the inoculum size effect described previously (17), where inhibition of germination occurs when the inoculation density was above 0.5 × 10⁶ cells/ml. To allow detection of decreasing and increasing rates of fluorescence over time, an inoculation density OD₆₀₀ of 0.03 was selected. P_HWP1-GFP expression in the thermocycler correlated well with the germination efficiency determined by microscopy (data not shown).

Fig. 2.

Expression of GFP in time is dependent on inoculation density. (A to D) The C. albicans P_ACT1-GFP strain (A and B) and P_HWP1-GFP strain (C and D) were inoculated to different densities (optical densities at 600 nm). The line charts or line graphs in panels A and C represent fluorescence intensity (in arbitrary units [AU]) over time as measured using an iCycler. Subsequent analysis of the maximum rate of fluorescence intensity increase (dF/dt) at different inoculation densities illustrates a different behavior for the C. albicans P_ACT1-GFP strain (B) than for the C. albicans P_HWP1-GFP strain (D).

The C. albicans P_ACT1-GFP strain was grown for 5 h in the iCycler with a range of concentrations of fluconazole (Sigma-Aldrich; from 40 mM in dimethyl sulfoxide [DMSO]) as a growth inhibitor. Fluconazole concentrations up to 0.4 mM presented no difference in the rate of fluorescence intensity, but from 0.8 to 3.3 mM, there was a striking acceleration of the rate of fluorescence (Fig. 3A, white bars). This was also observed for the P_HWP1-GFP strain but at lower fluconazole concentrations (Fig. 3A, gray bars). This increased ACT1 expression at sub-MIC concentrations could be explained by either increased protein expression related to the general stress response, a phenomenon often observed in biofilm-related drug susceptibility (16), or alternatively, the fluorescence intensity of GFP could increase due to the general stress response. At 1.6 mM fluconazole concentrations and higher, there was no further increase in fluorescence intensity observed. These data indicate that germination was inhibited by fluconazole at lower concentrations than growth as reported previously (22), which could reflect the suboptimal aeration conditions in the assay.

Fig. 3.

(A) The maximum rate of fluorescence intensity increase (dF/dt) in a C. albicans P_ACT1-GFP strain (white bars) and P_HWP1-GFP strain (gray bars) in the presence of increasing concentrations of fluconazole. (B and C) Effects of inhibitors of germination on the maximum rate of fluorescence increase (dF/dt) measured using the iCycler. The C. albicans P_HWP1-GFP strain was grown with serial dilutions of farnesol (B) and 3OC₁₂HSL (C). (D to F) Effects of inducers of germination on the maximum rate of fluorescence increase (dF/dt). The C. albicans P_HWP1-GFP strain was grown with serial dilutions of fetal bovine serum (D) and tyrosol (E and F), and the maximum rate of fluorescence increase was calculated. (E) Increase of GFP-derived fluorescence for the first 3 h, illustrating the very rapid and time-dependent nature of the observed effect.

The C. albicans P_HWP1-GFP strain grown with farnesol (from 100 mM stock in methanol; Sigma-Aldrich) showed a concentration-dependent decrease of GFP expression starting at 25 μM with no GFP expression being detected at 200 μM (Fig. 3B). In contrast, the C. albicans P_ACT1-GFP strain grown in the same conditions did not show inhibition of GFP expression compared with the control (data not shown). This is consistent with previous observations that farnesol inhibits germination without impacting growth (13). The C. albicans P_HWP1-GFP strain grown with N-(3-oxododecanoyl) homoserine lactone (3OC₁₂HSL; 25 mM stock in DMSO; a gift from M. M. Meijler) showed a decrease in GFP fluorescence starting at 12.5 μM (Fig. 3C), in contrast to the nearly 10-fold-higher concentrations that have been previously reported (9, 12).

Fetal bovine serum (Sigma-Aldrich) was added to the C. albicans P_HWP1-GFP strain up to 2%; larger amounts could not be used due to increased background fluorescence (10). An increase in fluorescence over time was observed, especially for 1 and 2% fetal bovine serum (Fig. 3D). Similarly, for the amino sugar N-acetyl-d-glucosamine (0.4 M stock in demineralized water; Sigma-Aldrich) (4), a 2-fold induction of GFP expression was observed at a 4 mM concentration (data not shown). The C. albicans P_HWP1-GFP strain grown with tyrosol (100 mM stock in demineralized water; Fluka) (5) revealed an increase of fluorescence signal in the very early stages of growth (between 0.25 and 0.5 h) for two concentrations, 0.06 and 0.125 mM (Fig. 3E and F). This effect disappeared within 1 h after inoculation. Our assay also allowed for analysis of compounds that accelerate hyphal formation in an elusive manner, like tyrosol (1, 5).

In conclusion, this study shows that it is feasible to quantify a GFP-derived fluorescence signal within living cells using a universally available device such as an thermocycler with a fluorescence detector. Microscopic analysis of germination is dependent on the observer, and it is not easy to distinguish hyphae from pseudohyphae. Alternatively, quantification of mRNA levels of germination-induced genes by Northern blotting or quantitative PCR (qPCR) analysis is elaborate and expensive and requires relatively large numbers of cells and test compounds. The advantages of our method are that it is fast, observer independent, and inexpensive and it can be performed in 96- or 384-well plates to minimize compound usage.