Histone deacetylase inhibitors may reduce pathogenicity and virulence in Candida albicans


  • Editor: Richard Calderone

Correspondence: Claudio Passariello, Dipartimento di Scienze di Sanità Pubblica, Sezione di Microbiologia Farmaceutica, Università‘La Sapienza’, P.le Aldo Moro 5, 00185, Rome, Italy. Tel.: +39 6 49914885; fax: +39 6 4451324; e-mail: claudio.passariello@uniroma1.it


Candida albicans is able to establish mucosal and invasive diseases by means of different virulence factors that are frequently regulated by epigenetic mechanisms, including the acetylation-deacetylation of histones and of other regulatory proteins. The focus of our work was on understanding the possible effects of several histone deacetylase inhibitors (HDACi) on the expression of phenotypes that are associated with virulence and pathogenicity in C. albicans, such as adhesion to epithelial cells and the yeast to hypha transition. Some of the HDACi used for experiments caused a 90% reduction in the adherence of C. albicans to human cultured pneumocytes and significantly inhibited serum-induced germination. Inhibition of germination was correlated with a significant reduction in transcription of EFG1. Inhibition appeared less evident when an HDA1-deficient strain was tested. These results suggest that selective and specific HDACi could prove to be a valid approach for selected at-risk patients in the combined treatment of infections caused by C. albicans.


Candida albicans is normally isolated from the human mucosae of the oral cavity, gut, and vagina. At these sites it behaves as an opportunistic pathogen, and causes disease usually following environmental changes in the host. In immunocompromised, transplant-receiving and cancer patients, C. albicans is frequently the fatal agent of systemic, invasive infections (Shepherd et al., 1985; Dupont, 1995; Fridkin & Jarvis, 1996). Although an opportunist, C. albicans is armed with a number of virulence factors to establish disease (Calderone & Fonzi, 2001). These factors include proteins devoted to recognition of the host (Sundstrom, 2002), enzymes enabling the invasion of the host (Shepherd et al., 1985; Cutler, 1991; Dupont, 1995), complex adaptive systems that allow response to environmental changes (Calderone & Fonzi, 2001), and proteins that help it evade host immune defense via molecular mimicry.

Most of these factors, including adhesiveness and dimorphism, have been extensively studied both in vitro and in animal models. Adhesins are believed to play a pivotal role in the pathogenesis of candidiasis by enabling C. albicans to colonize human mucosae. Such a role is confirmed by observations on the reduced pathogenic potential of strains with mutated adhesin genes or with mutations in genes that tightly regulate strategic adhesins (Buurman et al., 1998; Fu et al., 2002).

The ability of C. albicans to rapidly change morphology and cell wall structure and composition in response to environmental chemical and physical stimuli (De Bernardis et al., 1998; Munro et al., 1998; de Groot et al., 2004; Klengel et al., 2005) significantly influences its adhesiveness (Calderone et al., 2000) and its ability to cause systemic infections (Saville et al., 2006).

Dimorphism also plays a relevant role in the virulence process of C. albicans. In fact, non-filamentous mutants of C. albicans are avirulent in mouse models (Lo et al., 1997; Kelly et al., 2004) and when filamentation is inhibited, the ability of C. albicans strains to cause systemic infections is strongly reduced (Saville et al., 2006).

Morphogenetic interconversions in fungi are dependent on signal transduction pathways, including the cAMP-dependent protein kinase A (PKA) pathway (Borges-Walmsley & Walmsley, 2000). The tight regulation of cAMP levels has been shown to be essential for the normal growth and virulence of C. albicans (Bahn et al., 2003). The transcriptional regulator Efg1p plays a key role within this pathway (Stoldt et al., 1997; Bahn & Sundstrom, 2001). Efg1p in fact is a potent morphogenic regulator whose presence in C. albicans is required to allow induction of hyphae in the presence of serum (Lo et al., 1997; Stoldt et al., 1997). This regulatory protein is a downstream target of the cAMP-PKA pathway (Bockmühl & Ernst, 2001). Mutants lacking EFG1 grow as a yeast, and serum induces only short pseudohyphae rather than true septated hyphae as in wild-type strains.

In eukaryotes, developmentally regulated genes are frequently controlled at the epigenetic level by several classes of proteins including the histone deacetylases (HDACs). HDACs selectively regulate chromatin structure, thus affecting its folding patterns and interactions with DNA-binding proteins (Grozinger & Schreiber, 2002). A wide array of information has been made available on these enzymes in both superior eukaryotes and fungi (Grunstein, 1997; Bernstein et al., 2000; Sterner & Berger, 2000), and this knowledge has led to the synthesis of HDAC inhibitors (HDACi) (Marks et al., 2001; Grozinger & Schreiber, 2002; Mai et al., 2005a, b).

In fungi, important cellular functions were shown to be controlled by different HDACs (Perez-Martin et al., 1999; Srikantha et al., 2001; Kurdistani & Grunstein, 2003) and HDACi were shown to affect a number of cellular functions in these organisms. Trichostatin A (TSA), a well characterized HDACi, was shown to significantly affect the frequency and balance of white-opaque switching in C. albicans cells (Klar et al., 2001) and to enhance the sensitivity of C. albicans to azoles (Smith & Edlind, 2002). Recently, we have reported that TSA, suberoylanilide hydroxamic acid (SAHA), and two newly synthesized HDACi can also significantly inhibit the acquisition of resistance and trailing growth in C. albicans (Mai et al., 2007).

Targeted deletion of HDA1, encoding a deacetylase sensitive to TSA, was shown to have the same selective effect as TSA, supporting the hypothesis that acetylation of histones plays a selective role in regulating the white-opaque switching process (Klar et al., 2001).

In recent years, more data have been collected that demonstrate a role for some HDACs in regulating the functions of different intracellular proteins (Grunstein, 1997; Grozinger & Schreiber, 2002; Quivy & Van Lint, 2004).

In the light of these considerations, our work was aimed at demonstrating that HDACi actually inhibit HDACs of C.albicans in vivo, inducing relevant and interesting biological effects. Consequently, experiments were designed to elucidate the possible effects of different new HDACi on two important phenotypes that are associated with virulence and pathogenicity in C. albicans, adhesion to epithelial cells, and yeast to hypha transition.

Materials and methods

Strains and media

Four C. albicans strains were used in this study: one reference strain (American type culture collection (ATCC) 10231) isolated originally from a bronchial infection, two isolates from symptomatic HIV-positive patients (AIDS 32, a bronchial isolate, and AIDS 68, a vaginal isolate), and one vaginal isolate from a healthy woman (Cav 47). For experiments, strains were grown on Sabouraud agar plates (Difco Laboratories, USA) for 48 h at 25°C. Following growth, cells were suspended in liquid media for the different experiments, as described below in detail.

Inhibitors and inhibition activity testing

Four newly synthesized HDACi were tested during this study (1a, N-hydroxy-6-(6-benzyl-3,4-dihydro-4-oxopyrimidin-2-ylthio)hexanamide; 1b, N-hydroxy-5-(6-phenyl-3,4-dihydro-4-oxopyrimidin-2-ylthio)pentanamide; 1c, N-hydroxy-5-(6-(3-chlorophenyl)-3,4-dihydro-4-oxopyrimidin-2-ylthio)pentanamide; and 1d, N-hydroxy-5-(6-(4-chlorophenyl)-3,4-dihydro-4-oxopyrimidin-2-ylthio) pentanamide) (Fig. 1). They belong to the uracil-based hydroxamate (UBHA) series previously reported by us (Mai et al., 2005a, b, 2006). The synthesis of compounds 1a and 1b (Mai et al., 2005a, b, 2006) and of compounds 1c and 1d (Mai et al., 2007) has been already reported. The new HDACi were tested together with two well characterized HDACi, TSA and SAHA (Mai et al., 2001).

Figure 1.

 Chemical structures of the four newly synthesized HDACi.

When these UBHA compounds were tested for their ability to inhibit three maize deacetylases – HD2, HD1-B (class I), and HD1-A (class II) – they showed comparable mean 50% inhibitory concentration (IC50) values for the three enzymes (Mai et al., 2007). These values were higher than those obtained with TSA, but were significantly lower than those obtained with SAHA (Mai et al., 2007).

Minimal inhibitory concentrations (MIC) of UBHA compounds were determined for all tested strains according to a standard broth microdilution method (NCCLS, 2002). The four UBHA compounds showed MIC values ≥64 μg mL−1. Growth curves for all tested strains were constructed in YP medium, containing 1% (w/v) Yeast extract (Difco) and 2% (w/v) Peptone (Difco), alone or with the addition of each UBHA compound at the concentration used for all subsequent experiments (i.e. 4 μg mL−1). Analysis of growth curves showed that at these concentrations the UBHA compounds did not significantly affect the growth rate of the four C. albicans strains (e.g. 1d in Fig. 2).

Figure 2.

 Growth curves obtained by growing the four Candida albicans strains in parallel in YP medium alone (black diamond, continuous line) or in the presence of 1d (4 μg mL−1) (empty diamond, dotted line) at 37°C up to ≤48 h. Data are means of three independent experiments and are reported as log of the CFU mL−1 values obtained for each culture at the different time intervals. Comparable curves were obtained for all tested strains with the remaining HDACi.

Adhesion assays

The influence of HDACi on the efficiency of adhesion of C. albicans ATCC 10231 to epithelial cells was evaluated using monolayers of cultured A549 type II alveolar pneumocytes derived from a pulmonary adenocarcinoma (ATCC CCL-185). A549 cells were cultured in RPMI1640 medium (EuroClone, Italy), containing 1% glutamine, and 10% fetal calf serum (FCS) (HyClone, USA) (if not otherwise specified) and incubated at 37°C in 5% CO2. For quantitative determinations, 7.0 × 104 cells were seeded in 24-well plates and incubated overnight at 37°C. For adhesion assays, C. albicans cells grown on Sabouraud agar plates as described above (with or without each HADCi at 4 μg mL−1) were suspended in sterile saline at a density of 7.0 × 104 cells mL−1. Epithelial cells were washed twice with serum-free RPMI1640 medium and fed with the same medium, prior to addition of the C. albicans cells (100 CFU cell−1). After incubation for 1 h at 37°C in 5% CO2, the medium was removed, and the cells were rinsed five times and processed for enumeration of adherent Candida cells.

For enumeration of adherent Candida cells, epithelial cells were lysed in 1 mL of 0.025% Triton-X100 for 10 min. Samples were then centrifuged at 10 000g for 5 min at 4°C, suspended in phosphate-buffered saline (PBS), appropriately diluted, and plated for counts on Sabouraud agar plates. Mean values of five replicate counts for each sample were used for subsequent calculations.

Microscopic evaluation of adhesion

6.0 × 105 A549 cells were seeded in 35-mm tissue culture plates and incubated overnight at 37°C. Candida albicans cells from the suspensions in sterile saline described above were added to epithelial cells in medium without FCS (100 CFU cell−1), and incubated 1 h at 37°C. The medium was then removed and the cells were rinsed five times, fixed for 20 min in ice-cold methanol (50% in PBS) and finally stained with carbolfuchsin. Samples were then extensively washed, air-dried, and observed at × 400 magnification.

Germ tube formation assays

The influence of HDACi on germ tube formation was assayed in the presence of FCS as follows. Candida albicans strains, grown on Sabouraud agar plates as described above (with or without each HADCi at 4 μg mL−1), were suspended in YP medium at an OD540 nm=0.3. FCS (20% v/v final concentration) was then added to the suspensions, which were incubated for 2 h at 37°C. Germ tubes and total cells were then counted microscopically, and the results were expressed as percent germ tubes.

Semiquantitative reverse transcriptase-PCR (RT-PCR)

To evaluate the effect of treatment with HDACi on the levels of transcription of the EFG1 regulator we employed a semiquantitative RT-PCR method (Singh et al., 2005). Cells of C. albicans ATCC 10231 were obtained from germ tube formation assays, 2 h after challenging them with 20% FCS. Total RNA was isolated, using the sodium dodecyl sulphate (SDS)-hot phenol procedure described by Schmitt et al. (1990) and quantified with the Qubit Quantitation System (Invitrogen). First-strand cDNAs were generated from 2 μg of total RNA in a 30-μL reaction using Sensiscript Reverse Transcriptase (Qiagen) with oligo(dT) primers. For PCR amplification, 1 μL of cDNA was used as template. Gene-specific primers (Table 1) were added to the reaction for PCR amplification (15 cycles) with recombinant Taq polymerase (Invitrogen).

Table 1.   Primers used in the RT-PCR amplification of C. albicans specific gene transcripts
Primer nameSequence

The same RT-PCR method was used to assess the ability of all tested strains to express the two main HDAC genes, HDA1 and RPD3, using a specific set of primers for each gene (Table 1). Cells were obtained from fresh cultures grown both in liquid and on solid media and processed using RT-PCR as described above.

Equal volumes of each amplification reaction were analyzed by gel electrophoresis (0.7% agarose). The bands were visualized by staining with ethidium bromide and the intensities of bands were analyzed densitometrically using the Kodak Digital Science model 120 electrophoresis documentation and analysis system.

Western blot analysis

To evaluate the effect of treatment with HDACi on the acetylation of Histone H3, C. albicans ATCC 10231 was grown at 30°C in YP medium containing glucose 2% (w/v) (with or without each HDACi at 4 μg mL−1) to an OD600 nm of 0.8. A volume equivalent to 12 OD units of each culture was subjected to extraction of total proteins by the TCA lysate preparation method (Kao & Osley, 2003). SDS-polyacrylamide gel electrophoresis (PAGE) and Western blots were performed using standard protocols. A volume equivalent to 0.8 OD units of each sample was subjected to SDS-PAGE using 15% mini-gels. Duplicate gels were run in parallel, one gel being used for Western blots and the other being Coomassie stained. Separated proteins were transferred to nitrocellulose membranes that were processed to detect acetylated Histone H3 using Rabbit polyclonal anti Histone H3 (acetyl K18) (1/1000)(Abcam ab1191) primary antibody, and peroxidase-conjugated anti-rabbit IgG (Abcam) secondary antibody. Peroxidase activity on membranes was detected by the Sigma Fast-DAB system (Sigma Chemical Company). Digital images of membranes were obtained using the GS-800 calibrated densitometer (Bio-Rad Laboratories).

Statistical analysis

All the experiments were performed at least three times in triplicate. The data are reported as means ± SD. Two-tailed t-test statistics were applied to the data at significance levels of 0.05 and 0.01 respectively.


Adhesion assays

An in vitro model of pulmonary infection was used to evaluate the effect of different HDACi on the adherence of C. albicans to epithelial cells. Epithelial cells were grown to about 90% confluence and excess C. albicans cells, in medium without FCS, were added to evaluate the influence of pretreatment with HDACi on the adhesiveness of C. albicans. When C. albicans ATCC 10231 cells were pre-incubated in the presence of SAHA, TSA, 1c, or 1d, they showed a significant reduction (about 90%) in efficiency of adhesion to cultured pneumocytes (Fig. 3). Pre-incubation with 1a and 1b induced only a slight reduction in the efficiency of adhesion, and only with 1b was the reduction statistically significant (Fig. 3). Comparable adhesion assays were also analyzed after fixation and staining to evaluate the morphology of C. albicans cells under experimental conditions and to the exclusion of C. albicans cells adherent to the free plastic surfaces. Microscopic observations performed on A549 monolayers challenged with C. albicans ATCC 10231 confirmed that treatment with SAHA, TSA, 1c, or 1d, but not with 1a or 1b, significantly reduced the amount of C. albicans cells adherent to epithelial cells (Fig. 4). Moreover, it appeared that the percentage of germ tubes observed was significantly higher in positive controls (adhesion assays with no HDACi pretreatment) and in samples treated with 1a or 1b than in samples treated with SAHA, TSA, 1c, or 1d (Fig. 4). In all samples, the number of C. albicans cells adherent to the plastic surface was not significant.

Figure 3.

 Inhibition of adhesion of Candida albicans ATCC 10231 to A549, human cultured pneumocytes, following treatment with different HDACi. Results are reported as CFU well−1 and are means of three separate experiments performed in triplicate (individual SDs are also reported). *0.01<P<0.05, **P<0.01.

Figure 4.

 Inhibition of adhesion to human cultured pneumocytes by different HDACi. Adhesion assays were performed with Candida albicans ATCC 10231 cells cultured in different conditions and 6.0 × 105 A549 cells seeded in 35-mm tissue culture plates (a) Candida albicans cultured in medium without additions; (b) C. albicans cultured in medium containing SAHA; (c) Candida albicans cultured in medium containing TSA; (d) Candida albicans cultured in medium containing 1a; (e) Candida albicans cultured in medium containing 1d; (f) noninfected A549 cell. Original magnification × 400.

Germ tube formation assays

These previous observations prompted us to evaluate the effect of different HDACi on serum-induced germination. A standard method was used to evaluate the effects of different HDACi on the percentage of germ tubes formed by different C. albicans strains, following induction with 20% FCS at 37°C for 2 h. Under these experimental conditions, all strains were efficiently induced to germinate, the % germ tubes formed, ranging from 43% to 72% (Fig. 5). Pre-incubation with four of the six tested HDACi (i.e., SAHA, TSA, 1c, or 1d) caused a significant reduction in the percentage of germ tubes formed after 2 h (Fig. 5). Pre-incubation with 1a or 1b was significantly less effective on all of the C. albicans strains (Fig. 5). SAHA always affected germ tube formation less than TSA, 1c, or 1d (Fig. 5). Strain Cav 47, although it showed a comparable percentage of germ tubes formed when challenged with FCS, was evidently less affected by HDACi than the remaining strains (Fig. 5).

Figure 5.

 Inhibition of germ tube formation by six different HDACi on four different Candida albicans strains. Germ tube formation assays were performed on C. albicans cells pretreated with each HDACi for 48 h at 25°C and then suspended in YP medium at an OD540 nm=0.3, in the presence of 20% FCS. Results are reported as percent of germ tubes formed at 2 h after induction with FCS. Results are means of triplicate experiments (individual SDs are also reported). *0.01<P<0.05, **P<0.01.

Semiquantitative RT-PCR assays

As both adhesion to epithelial cells and germination in C. albicans are regulated by the cAMP-dependent pathway, we decided to evaluate whether treatment with HDACi could affect the transcription of a gene coding for a key regulatory protein involved in this pathway, such as Efg1p. Transcription was investigated by a semiquantitative RT-PCR method using C. albicans ATCC 10231 cells from germ tube formation assays. As expected, C. albicans ATCC 10231 cells, when induced to germinate by treatment with 20% FCS, showed significant amounts of intracellular EFG1-specific transcripts. Pre-treatment with TSA, 1c, or 1d caused a significant reduction in the abundance of intracellular EFG1-specific mRNAs (Table 2), whereas pretreatment with SAHA caused only a minor reduction in the abundance of EFG1 mRNAs (Table 2). 1a and 1b did not significantly affect the expression of these genes (Table 2).

Table 2.   Results of the densitometric analysis of semiquantitative RT-PCR assays performed on samples of C. albicans ATCC 10231 induced by FCS
GeneSample/control signal ratio (± SD) of specific gene transcripts
  • SAHA, TSA, 1a, 1b, 1c, and 1d: Candida cells pretreated for 48 h at 25°C with the correspondent HDACi and induced to germinate for 2 h at 37°C.

  • Results are reported as mean relative signal ratios (± SD), assuming values obtained with controls as the reference and normalizing values for EFG1 using ratios obtained for TEF1 with that sample.

  • *


  • **


EFG10.63 (± 0.11)**0.3 (± 0.08)1.27 (± 0.03)1.2 (± 0.05)0.21 (± 0.06)**0.25 (± 0.05)**
TEF11.24 (± 0.11)1.11 (± 0.1)0.82 (± 0.09)0.79 (± 0.07)1.38 (± 0.12)*1.02 (± 0.04)

The four C. albicans strains were also analyzed by RT-PCR for their ability to express two primary HDAC genes, HDA1 and RPD3, under different culture conditions. RPD3-specific amplification products could be obtained for all tested strains under all culture conditions. Densitometric analysis of digital images of agarose gels indicated that for strain AIDS 68 RPD3 transcripts were at least two times more abundant than for the other strains. Strains AIDS 32 and AIDS 68 showed significantly higher levels of expression of HDA1 (3.9 ± 0.09 and 3.7 ± 0.12 times) when compared to strain ATCC 10231. All attempts to amplify HDA1-specific sequences from strain Cav 47 failed, although amplification of the same sequence from a genomic preparation of the same strain yielded a specific amplification product. These data suggest that Cav 47 does not express this gene to a degree comparable to that of the other strains under our experimental conditions. No specific investigations were performed at the genomic level to explain any transcriptional differences.

Western blot analysis

To determine whether biological effects of HDACi were actually dependent on the inhibition of HDACs in vivo, the hyperacetylation of histone H3 was evaluated in C. albicans ATCC 10231 following treatment with the different HDACi. Results of this experiment confirmed that the HDACi utilized in these experiments actually inhibited HDACs in C. albicans in vivo. In fact, in samples obtained from C. albicans cells treated with HDACi the amounts of acetylated histone H3 were increased as compared with those of the non-treated control (Fig. 6). Under these experimental conditions, the amount of acetylated histone H3 in cells treated with SAHA showed only minor differences as compared to non-treated controls (Fig. 6).

Figure 6.

 Induction of hyperacetylated histone H3 following treatment with different HDACi. The in vitro effects of the different tested HDACi were investigated in Candida albicans ATCC 10231 grown in liquid medium for about 6 h to a final OD600 nm of c. 0.8, in the presence of each HDACi. Volumes corresponding to 0.8 OD of the clarified total cell lysates were separated by SDS-PAGE in 15% gels and were analyzed with Western hybridization (a) using an anti-acetylated histone H3 antibody. Comparable Coomassie-stained polyacrylamide gels were used as controls for protein load (b).


HDACi are members of an interesting family of compounds that is actively studied for various therapeutic applications in man (Marks et al., 2001; Grozinger & Schreiber, 2002; Mai et al., 2005a, b). At present, due to their relevant toxic effects, they have deserved attention principally as anti-cancer agents. Recently, however, new HDACi have been synthesized that are characterized by lower toxic effects as a consequence of higher selectivity and specificity for their targets.

In this context, we decided to investigate the possibility of developing inhibitors able to interfere with the virulence of pathogenic fungi. Previous studies have evaluated the effect of HDACi on the expression of selected phenotypes in C. albicans (Klar et al., 2001; Srikantha et al., 2001; Smith & Edlind, 2002). However, these studies dealt mainly with the physiological role of HDACs, or they used HDACi characterized by extremely high toxicity and very low specificity or selectivity. In contrast, the UBHA compounds used in this study showed no significant effect on either cell cycle or apoptosis induction in the human leukaemia U937 cell line (Mai et al., 2006, and unpublished results). Under the same experimental conditions, SAHA (used as reference drug), showed a S/G2/M block of the cell cycle and massive apoptosis (98%) (Mai et al., 2006).

Also, the selectivity profile of the UBHA compounds is quite different from that of SAHA. In fact, SAHA inhibited human HDAC1 (class I HDAC) and HDAC4 (class IIa HDAC) to the same extent, and it increased the acetylation levels of both histone H3 and μ-tubulin (Mai et al., 2006) in the same human cell line. In contrast, 1b inhibited human HDAC1 (50% inhibition at 5 μM), but not HDAC4 (0% inhibition at 5 μM), and it highly increased the acetyl-tubulin level in U937 cells (Mai et al., 2006). In the same assays, 1d was inactive against HDAC1 (0% inhibition at 5 μM), strongly inhibited HDAC4 (83% inhibition at 5 μM), and again induced α-tubulin acetylation (unpublished results).

Four different strains of C. albicans were included in this study as representatives of different colonization and pathogenic potentials. All these strains except Cav 47 (a strain defective for the expression of HDA1) were shown to express the two main HDAC genes, HDA1 and RPD3. Although different levels of expression of both genes were observed for the four strains, no correlation was apparent between gene expression and the origin of strains, or with the percent of germ tubes formed when challenged for 2 h with FCS.

Adhesion to epithelial cells and the transition from yeast to hypha were selected as models to study the influence of HDACi on the adaptative potential of C. albicans to environmental changes during colonization of mucosal membranes and dissemination via the bloodstream. Adhesion to epithelial cells influences the initiation of infection at the level of mucosae (Buurman et al., 1998), and adhesiveness is significantly enhanced by the potential to germinate (Kimura & Pearsall, 1980). The experiments described in this paper demonstrate that different HDACi, including leading compounds, such as TSA and SAHA, and de novo synthesized UBHA compounds, show significant activity in reducing adhesion to cultured pneumocytes and also FCS-induced germination in C. albicans. The observed effects could not be ascribed to direct toxicity of HDACi as demonstrated by calculation of MIC values and construction of growth curves for each strain in the presence of the different HDACi.

In C. albicans the expression of several adhesins (Fu et al., 2002; Li & Palecek, 2003) and HDACi-sensitive white-opaque switching (Klar et al., 2001) are regulated by the pathway comprising the Efg1p transcriptional regulator. Enhancement of intracellular concentrations of the EFG1 transcript was demonstrated to be essential for initiation of the morphological transition program, although later on during hyphal induction, levels of the EFG1 transcript decline dramatically. Moreover, mutants deficient in hyphal formation conferred by inactivation of both EFG1 alleles, were shown to be avirulent (Lo et al., 1997). Previous reports demonstrate that an increase in the intracellular concentration of cAMP plays a central role in promoting the yeast to hypha transition (Niimi et al., 1980; Chattaway et al., 1981; Zelada et al., 1996). These intracellular biochemical events were shown to be controlled by the Efg1p transcriptional regulator, and to influence significantly both the morphology and the virulence of C. albicans (Bahn & Sundstrom, 2001).

Data on the biological effects of the studied UBHA compounds prompted us to analyze the effects of HDACi at the transcriptional level on both these genes. In consideration of the elevated number of samples to be analyzed and of the preliminary nature of these observations, we decided to perform transcriptional studies by a semiquantitative approach. This approach theoretically enabled us to disclose only major differences in gene expression levels. In fact, the possibility to evaluate amplification reactions only out of their linear phase causes differences to be underestimated. Further analyses using quantitative methods on a larger number of genes will consequently be necessary to assess the exact nature of transcriptional alterations induced by selected inhibitors.

Consistent with our hypothesis, treatment with the different HDACi resulted in transcriptional down regulation of EFG1, proportionally with the ability to inhibit germ tube formation. These HDACi were consequently able to affect a step that is considered crucial in giving C. albicans its potential to cause systemic infections in vivo (Stoldt et al., 1997; Saville et al., 2006).

In contrast to our data, Smith & Edlind (2002) previously reported that TSA had no effect on germination in C. albicans. However their experiments were performed by adding TSA at the time of induction, while we decided to pretreat C. albicans strains with each HDACi, instead of adding HDCAi directly to the FCS-containing medium. We chose to pretreat with HDACi because serum induction is extremely rapid, and the time required for HDACi to be active is not known. In these experiments we made no attempt to determine the minimum pretreatment time required to obtain an inhibition of germination, although preliminary observations indicated that only a short pretreatment would be required.

The four active inhibitors in this study showed comparable effects on all four C. albicans strains with the exception of strain Cav 47, in effecting different levels of expression of the two principal HDAC genes. Strain Cav 47, although induced to germinate by FCS, was significantly less sensitive to HDACi than the other strains. These data suggest that Hda1p is actively involved in the cellular events that are affected by HDACi and that lead to an inhibition of germ tube formation. Further studies possibly performed on HDA1 deleted strains will be necessary to elucidate these aspects.

It is surprising that the four UBHA compounds and SAHA, which are all structurally related, showed comparable activities regarding human and maize HDAC enzymes yet yielded such different effects on both germination and adhesion in Candida. The reasons for better activity being shown by the two chlorine-containing UBHA compounds (1c and 1d) are not known, although they are possibly multifactorial. A different susceptibility of the Candida HDACs to the inhibitors is suggested by evaluation of H3 acetylation induced by the different inhibitors. In fact, the above discussed selectivity profile, as determined on human HDACs allowed to expect SAHA, 1a, and 1b to show comparable activities, and 1c, and 1d to be less active. A variable penetration ability of the different HDACi through the fungal membrane could also be suspected to be conferred by different lipophilicities (Clog P of 1a and 1b: 0.753; Clog P of 1c and 1d: 1.481), although the ability to induce histone H3 hyperacetylation shows that they are all able to penetrate the fungal cell wall and membrane. These preliminary observations suggest that it may be necessary to develop methods that make it possible to determine the selectivity profile of HDACi directly on the different purified HDACs of C. albicans.

Our experimental data support the hypothesis that observed effects of HDACi here described are a consequence of the direct inhibition of HDACs enzymes of C. albicans, although they do not elucidate which HDACs are the main targets of these compounds in vivo. These experimental data suggest that HDACi selective for specific Candida sp. enzymes can be designed, and this approach could prove to be instrumental in developing new therapies for mucosal and invasive fungal infections, at least for selected at risk groups. Overall, these data show that the chemical manipulation of HDACi structures can lead to the identification of new compounds with fungal cell selectivity that could be suitable for therapeutic application in humans. Moreover, these compounds could be developed to enhance specificity for fungal HDACs enzymes. Structural and biochemical studies performed on the purified or recombinant enzymes may lead to the development of new inhibitors with favourable pharmacologic and kinetic characteristics, making them convenient for in vivo application in humans. The main targets of these drugs should be the combined treatment of mucosal and systemic infections in patients within well defined risk categories. It must be noted that data reported in this paper suggest that the administration of HDACi as anti-cancer agents could be beneficial also in reducing the incidence of fatal disseminated candidiasis in these subjects.

Much work is still needed to make HDACi sufficiently selective and to minimize any toxic effects in humans, but preliminary data on several new HDACi, including the UBHA compounds that were studied here, make it conceivable that through the design and application of specific screening protocols, HDACi could become a valuable addition to the arsenal of antifungal drugs.


This work was supported by MIUR Project PRIN 2005 ‘protocol 2005061134_004’ granted to CP, PRIN 2006 ‘protocol 2006038137_0056’ and AIRC 2006 granted to AM. The authors are grateful to Dr Rick Veeh for his patient and competent critical revision of the manuscript, and to Alessandra Virga for skilfull assistance in performing transcriptional studies.