CO₂ Regulates White-Opaque Switching in Candida albicans

Summary

To mate, Candida albicans, must undergo homozygosis at the mating type like locus, MTL [1, 2], then switch from the white to opaque phenotype [3, 4]. Paradoxically, when opaque cells are transferred in vitro to 37 °C, the temperature of their animal host, they switch en masse to white [5–7], suggesting that their major niche may not be conducive to mating. It has been suggested that pheromones secreted by opaque cells of opposite mating type [8] or the hypoxic condition of host niches [9, 10] stabilize opaque cells. There is, however, an additional possibility, namely that CO₂, which achieves levels in the host 100 times higher than in air [11–14], stabilizes the opaque phenotype. CO₂ has been demonstrated to regulate the bud-hypha transition in C. albicans [15], expression of virulence genes in bacteria [16] and mating events in Cryptococcus neoformans [17]. Here we have tested the possibility that CO₂ stabilizes the opaque phenotype. It is demonstrated that physiological levels of CO₂ induce white-to-opaque switching and stabilize the opaque phenotype at 37 °C. It exerts this control equally under anaerobic and aerobic conditions. These results suggest that the high levels of CO₂ in the host induce and stabilize the opaque phenotype, thus facilitating mating.

CO₂ Stimulates White to Opaque Switching

The CO₂ content of air is 0.03%, but in the animal host it ranges between 4.5 and 30% [11–14]. When white cells of five natural strains homozygous at the MTL locus (Table S1) were plated and incubated in air containing 5% CO₂ at 25 °C, the frequency of switching, measured as the proportion of total colonies that are white with opaque regions or completely opaque, was 4 to 16 fold higher than in air (Figure 1A, B). When plated and incubated in air containing 20% CO₂, the frequency was 10- to 105-fold higher than in air. CO₂ also induced white-to-opaque switching at 37 °C (Figure S1). Variability existed among strains both in air and in the test concentrations of CO₂, but switching was induced by high concentrations of CO₂ in all strains (Figure 1B).

Figure 1.

High concentrations of CO₂ induce switching from the white (Wh) to opaque (Op) phenotype in C. albicans. White cells of either α/α or a/a strains were plated on agar and then incubated in air, which contained 0.03% CO₂, air (95%) containing 5% CO₂, or air (80%) containing 20% CO₂ (Experimental Procedures in Supplemental Data). The agar contained phloxine B which stained opaque colonies or opaque regions of white colonies red [28]

(A) Representative fields of colonies for one α/α (WO-1, FC4) and one a/a strain (GH1012).

(B) Quantitation of the frequency of switching measured as the proportion (%) of total colonies that were white with opaque regions or predominately opaque. The total number of colonies is the sum of three experiments. The “% Op colonies or Wh colonies with Op regions” represents the mean for the three experiments ± standard deviation. The fold difference with air is presented for the means at 5 and 20% CO₂. The p-values calculated by the student’s two tailed T-test for switching frequencies in 5 and 20% CO₂ compared to those in air were all <0.05.

CO₂ Blocks Opaque to White Switching

When opaque cells of test strains were plated and incubated at 25 °C in air, two (WO-1, GH1012) exhibited moderate frequencies of switching, and three (P37005, P97099, P78048) high frequencies of switching (Figure 2A). In air containing 5% CO₂, the frequency of switching by opaque cells of all five strains was reduced to low or negligible levels (Figure 2A). In addition, 5% CO₂ blocked temperature-induced mass conversion of opaque to white [5–7] after a temperature shift from 25 ° to 37 °C (Figure 2B). At 37 °C in air, both white and opaque cells multiplied, but while white cells maintained their phenotype, opaque cells converted en masse to white (Figure 2C). In 5% CO₂, however, both white and opaque cells multiplied and maintained their respective phenotypes (Figure 2C).

Figure 2.

CO₂ stabilizes the opaque phenotype at 25 ° and 37 °C, and does not interfere with cell multiplication.

(A) Opaque cells of a/a and α/α cells were plated on agar and monitored for switching according to the general procedure outlined in the legend to Figure 1 and Experimental Procedures in Supplemental Data. The p-values calculated by the student’s two tailed T-test of the switching frequencies of P37005, P97099 and P78048 in 5% CO₂ compared to those in air were <0.05.

(B) Opaque cells were plated at 37 °C and analyzed for switching to white, according to methods outlined in Experimental Procedures in Supplemental Data. The p-values calculated by the student’s two tailed T-test of WO-1 and GH1012 in 5% CO₂ compared to those in air were <0.05.

(C) White or opaque cells were grown in liquid cultures at 37 °C in air or air containing 5% CO₂, and monitored for cell number and cell phenotype (Experimental Procedures in Supplemental Data). Cell phenotype was monitored by plating aliquots at time intervals and counting the proportion of colonies with regions of alternative phenotype.

Phase-Specific Gene Expression

To test whether CO₂-induced white-to-opaque switching was accompanied by associated changes in gene expression, two white-specific genes, WH11 [18] and EFG1 [19, 20], and two opaque-specific genes, OP4 [21] and WOR1 (TOS9) [22–24], were analyzed by northern blot hybridization. In air, the transcript levels of the two white-specific genes remained high, and the transcript levels of the two opaque-specific genes remained negligible (Figure 3). In 5 or 20% CO₂, however, the transcript levels of the two white-specific genes decreased to negligible levels, while those of the two opaque-specific genes increased dramatically (Figure 3).

Figure 3.

CO₂-induced switching from white to opaque is accompanied by down-regulation of white-specific and up-regulation of opaque-specific genes

Northern blot hybridization was performed for white- and opaque-specific genes in white cell cultures of strain WO-1 for 0.5, 1, 2 and 3 days on agar in air, air containing 5% CO₂ or air containing 20% CO₂. The ethidium bromide stained 18S ribosomal RNA bands are shown to demonstrate equal loading of lanes.

CO₂ Facilitates Mating

Because CO₂ induced white to opaque switching and stabilized the opaque phenotype, we predicted it would facilitate mating in white populations. To test this, crosses were performed in air and 5% CO₂, by mixing white cells of strain WUM5A (α/α, ura3⁻) and MMY278 (a/Δα, ade2⁻), and white cells of strains CHY477 (Δa/α, ade2⁻) and 3UM5A (a/a, ura3⁻) (Table S1). 5% CO₂ increased the mating efficiency in the two crosses by 583 and 395 fold, respectively (Table S3).

Roles of Carbonic Anhydrase, Adenylate Cyclase and Ras1

At low atmospheric CO₂ levels, carbonic anhydrase (CA) catalyzes the interconversion of CO₂ to carbonic acid (HCO₃⁻), which is necessary for obtaining CO₂ effects, presumably through HCO₃⁻ signaling; at high atmospheric CO₂ levels, diffusion across the plasma membrane is high enough so that the uncatalyzed formation of HCO₃⁻ is sufficient [15]. In C. albicans, deletion of the gene for carbonic anhydrase, NCE103, blocks growth in air (0.03% CO₂), but not in air supplemented with 5% CO₂, and affects hypha induction at low but not high CO₂ concentrations [15]. We tested whether carbonic anhydrase played a similar role in the induction of switching by CO₂ by analyzing the null mutant nce103Δ (Table S1). Since nce103Δ cells did not grow in air [15], white cells of the mutant were first grown on agar in air containing 1% CO₂, then plated on fresh agar plates and incubated in 1 or 5% CO₂ to test for stimulation. While over 90% of the colonies formed in 1% CO₂ (low CO₂) by white cells of wild type and complemented strains contained opaque regions, only 22 and 19% of the colonies formed in 1% CO₂ by white cells of strain nce103Δ contained opaque regions (Figure 4B). At 5% CO₂ (high CO₂), however, over 95% of the colonies formed by white cells of both control and mutant strains contained opaque regions (Figure 4B). Therefore, carbonic anhydrase was necessary for wild type levels of switching at low (1%) CO₂, but it was not necessary for maximum levels at high (5%) CO₂, which is similar to the effects described for CO₂-induced filamentation [15].

Figure 4.

CO₂ induction of switching from white to opaque in null mutants of carbonic anhydrase (nce103Δ), adenylate cyclase (cdc35Δ), Ras1 (ras1Δ), the master switch locus WOR1 (wor1Δ) and the transcription regulator Czf1 (czf1Δ)

(A) Agar cultures demonstrating that the nce103Δ mutant does not grow in air, but does grow in 5% CO₂.

(B) Representative fields of colonies formed by white cells of parent strain GH1013 (Table S1) and nce103Δ in air containing 1 or 5% CO₂. Fields of control strains nce103Δ+pACT-NCE103 (the complemented nce103Δ mutant) and nce103Δ +vector (nce103Δ transformed only with the vector) are also presented.

(C) The switching frequencies of the tested mutants, wild type controls and an a/α control, monitored according to methods outlined in the legend to Figure 1 and Experimental Procedures in Supplemental Data.

Klengel et al. [15] demonstrated that CO₂ induced the bud-hypha transition by stimulating adenylyl cyclase. This cAMP-dependent pathway in turn required Ras1 [25]. We therefore tested the CO₂ effect on switching in null mutants of the adenylyl cyclase gene (cdc35Δ) and RAS1 (ras1Δ) (Table S1). In both mutants, the frequency of switching was lower than that of the control strain in air (Figure 4C). In 1% CO₂, switching increased in the two mutant strains, but to levels far below those of control cells (Figure 4C). At 20% CO₂, however, switching was maximal, as in control cells (Figure 4C). Therefore, both adenylyl cyclase and Ras1, like carbonic anhydrase, enhanced switching in air and enhanced induction by low but not in high concentrations of CO₂.

WOR1 (TOS9) and CZF1

Since the master switch locus WOR1 (TOS9) regulates the white-to-opaque transition [22–24], and overexpression of the transcription factor Czf1 promotes that switch [26], we analyzed the induction of switching by CO₂ in the mutants wor1Δ and czf1Δ (Table S1). As expected, switching did not occur in air or at 20% CO₂ in white cells of wor1Δ (Figure 4C). CO₂ did, however, induce switching in white cells of czf1Δ, but the induced frequencies in 1 and 20% CO₂ were lower than that of control cells, as was the basal frequency in air (Figure 4C). As expected, CO₂ did not induce switching in a/α cells (Figure 4C).

CO₂ and O₂

Dumitru et al. [9] found that under anaerobic conditions at 37 °C, opaque cells switched to white at a lower frequency than under aerobic conditions, suggesting that hypoxia stabilized the opaque phenotype. However, opaque cells also divided at a slower rate under anaerobic conditions [9], and switching requires cell division [7, 18]. Therefore the observed reduction in frequency could be due to the decrease in growth rate. Ramirez-Zavala et al. [10] demonstrated that anaerobic conditions induced white cells to switch to opaque, and that induction was regulated by Czf1. Their method to remove O₂, however, generated 18% CO₂ in the culture jars, a concentration we found induce a maximal rate of switching (Figure 1) and a maximum level of phenotypic stabilization (Figure 2A). To address whether hypoxia alone can stimulate switching, white cells were plated and incubated in either N₂ (99.97%) containing no O₂ and 0.03% CO₂ (the CO₂ concentration in air) or N₂ (99.47%) containing 0.5% O₂ (a 40 fold reduction from that in air) and 0.03% CO₂. Rather than stimulating switching, the frequency was reduced 8 and 12 fold, respectively, from that in air (Table S4). When white cells were plated and incubated in either N₂ (90%) containing no O₂ and 10% CO₂, or N₂ (89.75%) containing 0.5% O₂ and 9.75% CO₂, the frequencies of switching were 98.9 and 97.9%, respectively (Table S4). These results demonstrate that hypoxia does not induce white to opaque switching and that CO₂ induces switching equally in the absence and presence of O₂.

Conclusion

We have, therefore, presented evidence that exposure of white cells to the high levels of CO₂ found in host niches induced switching from white to opaque, then maintained the opaque phenotype. These results provide an explanation for observations made twenty years ago that physiological temperatures caused mass conversion of opaque cells to the white phenotype [5–7, 18]. CO₂ induced white to opaque switching equally well under anaerobic and aerobic conditions, the former also a characteristic of most host niches. Our results indicated that hypoxia was not an inducer of white-to-opaque switching. At low, but not high, concentrations of CO₂, induction was selectively enhanced by carbonic anhydrase, adenylate cyclase and Ras1. Similar dependencies had been observed for CO₂ induction of hypha formation [15], which may not be surprising given other similarities between the white-opaque transition and the bud-hypha transition [27]. At both low and high CO₂ concentrations, the transcription factor Czf1 promoted the induction of switching. Czf1 had previously been shown to promote white-to-opaque switching [10, 26]. The observation that the basal rates of switching in air as well as the induced rates in low CO₂ were depressed in the mutants nce103Δ, cdc35Δ and ras1Δ suggested that what has been referred to as “spontaneous” switching from white-to-opaque in air [5] may in fact represent induced switching by the low level of CO₂ (0.03%). We therefore propose that in the host, it is the high level of CO₂ that induces the white-to-opaque switch, then stabilizes the mating-competent opaque cell, thus facilitating mating.