Candida species are commensal microorganisms of vertebrate hosts that can cause infections ranging from non-life-threatening to invasive illnesses. Although candidaemia is the most common manifestation of invasive candidiasis, extensive visceral invasion with Candida can occur in all organs. The eyes, brain, liver, spleen, and kidneys are the most commonly affected [1]. Candidiasis is the fourth most common cause of nosocomial bloodstream infections in Brazil and the U.S.A., with a mortality rate of approximately 40% [1, 2]. A progressive increase in the number and severity of candidiasis over the past two decades has been observed worldwide, especially in immunocompromised patients and also in patients hospitalised with serious underlying diseases, during immunosuppressive therapy, or parenteral nutrition, as well as among patients exposed to invasive medical procedures. Yeasts of Candida albicans are the most frequently implicated in cases of invasive candidiasis infections. However, nowadays Candida non-albicans (CNA) species such as Candida glabrata, Candida krusei, and Candida parapsilosis have increased in importance and number among fungal infections [1].

Currently, the mainstay of chemotherapy employed for the treatment of fungal infections comprises drugs that affect the function or biosynthesis of membrane sterols [3]. The polyenes (such as amphotericin B) were the first antifungal class used to treat invasive fungal infections. The primary mechanism of amphotericin B is its binding to the signature 24-alkyl sterols present in fungal cell membranes, leading to a perturbation of the membrane selective permeability and, consequently, loss of the cellular content. Despite the specific fungicidal effect of polyenes, they display significant toxicity to mammalian cells [4]. Another important antifungal class comprises the azoles, such as ketoconazole, fluconazole (FLC), itraconazole (ITC), posaconazole, and voriconazole, which are the compounds most frequently used today, and whose specific target is the cytochrome P-450-dependent C14α-demethylase, a key enzyme of the ergosterol biosynthesis pathway [4]. Although azoles are one of the main classes of drugs used in the treatment of fungal infections, these drugs present several problems such as their fungistatic rather than fungicidal activity, variable drug bioavailability, lack of intravenous preparations, large number of drug-drug interactions, development of resistance, and potential cross-resistance between different azoles [5].

During the last two decades, some studies have described a new class of antifungals called azasterols, which are inhibitors of the Δ²⁴⁽²⁵⁾-sterol methyltransferase (24-SMT), another key enzyme of the ergosterol biosynthesis pathway, which is absent in the mammalian host cells [6–8]. This enzyme catalyses the S-adenosylmethionine-mediated incorporation of methyl groups at position 24 in sterols, which is an essential step for the biosynthesis of fungal sterols [6, 8]. 20-piperidin-2-yl-5α-pregnan-3β-20(R)-diol (AZA) and 24(R,S),25-epiminolanosterol (EIL) are steroid compounds with a nitrogen atom in the side chain (azasterols, Fig. 1), and are known inhibitors of 24-SMT in fungi [9], Trypanosoma cruzi [10], and Leishmania amazonensis [11, 12]. Antifungal activities of these inhibitors were also described against Pneumocytis carinii [13] and Paracoccidioides brasiliensis [14].

The purpose of the present study was to (i) examine the susceptibilities of a collection of 70 yeasts of the genus Candida to AZA and EIL; (ii) determine the fungicidal activities of these compounds; and (iii) detect the main morphology and ultrastructural alterations of the yeasts after drug treatment.

Although C. albicans is the predominant species in candidiasis, CNA species have increased in frequency in recent years. The reasons for the emergence of CNA species are not fully understood, but some medical conditions may frequently run the risk of developing candidaemia due to the CNA species: C. parapsilopsis has been associated with vascular catheters and parenteral nutrition; C. tropicalis with cancer and neutropenia; and C. krusei and C. glabrata with previous treatments with FLC and ITC [2].

Previous studies have described a high susceptibility of C. albicans isolates to azoles and AMB, whereas CNA isolates are usually less susceptible and may be intrinsically resistant to FLC and ITC [2, 15–17]. As reported by other investigators [2, 18, 19], none of our Candida isolates showed MIC ≥ 2 μg.ml^-1 for AMB. MIC values found for ITC and FLU were similar to those previously reported by different groups [2, 15–17]. However, in the present study, FLC-resistant Candida strains were only observed among CNA species (6.8% of the isolates). However, ITC-resistance was found in C. albicans (1.5%) and CNA isolates (7.7%).

The observation of a high positive correlation between MIC values of FLC and ITC (r = 0.79 for MIC₅₀ and r = 0.71 for MIC₉₀), in this study, suggests that cross-resistance may be occurring. However, no correlation was observed between MIC values of the azoles and 24-SMTI, indicating lack of possible cross-resistance.

The general finding for our Candida spp. isolates was that they were mostly susceptible to AZA and EIL, because the MIC₅₀s were lower than 2 μg.ml^-1 for 73% and 88% of the isolates after treatment with AZA and EIL, respectively. Interestingly, some FLC- and ITC-resistant strains were susceptible to 24-SMTI. However, residual growth of Candida after treatment with AZA was similar to that observed for FLC and ITC. No residual growth was observed after treatment with EIL. The fungicidal action of 24-SMTI was more prominent against CNA species than against C. albicans isolates. A concentration of 4.0 μg.ml^-1 of 24-SMTI was enough to kill 100% of C. lusitanae, C. zeylanoides, and C. rugosa, and 50% of C. glabrata. In contrast, this same concentration killed only 4.7% and 9.5% of C. albicans isolates, considering AZA and EIL respectively.

Previous studies have shown that azasterol derivatives have antifungal activity against a variety of species [7]. 15-azasterol, in concentrations ranging from 0.01 μg.ml^-1 to 4.08 μg.ml^-1, inhibits the growth of Saccharomyces cerevisae and C. albicans, with a concomitant accumulation of sterol intermediate molecules [20, 21]. The range of MIC and MFC values for 15-azasterol analogues against these fungal species varied from 0.8 to 3.1 μg.ml^-1 and 3.1 to 6.3 μg.ml^-1, respectively [7] and are similar to the values obtained in the present study.

Other azasterol derivatives have been shown to inhibit S. cerevisae 24-SMT, leading to the accumulation of zymosterol [22]. Recent work demonstrated that AZA displays antifungal activity against Paracoccidioides brasiliensis [14] and Pneumocystis carinii [13]. Concentrations of 5 μM (2.05 μg.ml^-1) inhibited 100% of the growth in P. brasiliensis, and the treatment of P. carinii with the IC₅₀ of 0.3 μM (0.12 μg.ml^-1) led to growth arrest and accumulation of 24-desakyl sterols, indicating an inhibition of 24-SMT [13]. In addition, previous studies have also shown an anti-protozoan activity of AZA and EIL on T. cruzi epimastigotes and intracellular amastigotes [10], L. amazonensis promastigotes and intracellular amastigotes [11, 12], Toxoplasma gondii [23], and Giardia lamblia [24], with MICs in the low μM to sub-μM range. For protozoans, EIL was reported to be more active than AZA. In contrast, we found in this study that AZA was more active than EIL against Candida spp. isolates.

Treatment of C. albicans yeasts with AZA and EIL caused dramatic changes in their cellular and sub-cellular structure. The main alterations included changes in the cell wall shape and thickness, a pronounced disconnection between the cell wall and cytoplasm, with the presence of an electron-lucent zone between them, cell collapse and release of cellular content, mitochondrial swelling, and abnormalities in the nuclear structure. These findings are similar to those previously reported after treatment of Candida spp. with different azoles [25–28]. Borges and co-workers [27] reported that exposure of Candida albicans to ITC leads to primary alterations at the cell periphery and the appearance of vacuoles in the cytoplasm, which may be lipid inclusions. These changes were usually accompanied by an increase in the cell volume and impaired cell division. In addition, studies by Hazen and co-workers [28] revealed that Candida treated with FLC shows a distinct retraction of the membrane from the cell wall. On the other hand, C. albicans treated with low concentrations of AMB shows chromatin condensation and margination, separation of the nuclear envelope, and nuclear fragmentation [29]. High concentrations of AMB induce cellular changes characteristic of necrosis, showing many large vacuoles [29]. Additionally, Bahmed and co-workers [30] demonstrated an increase in cell wall thickness of Candida yeasts, which may be related to alterations in the cell wall composition induced by the treatment with AMB.

In addition, similar to our findings, the appearance of multivesicular bodies and myelin-like structures were reported after treatment of Leishmania [11, 12] and T. cruzi [31] with AZA and EIL.

Staining with Nile Red revealed the presence of lipid accumulation in the cytoplasm after treatment with 24-SMTI, confirming that these compounds induce a perturbation in lipid biosynthesis. Similar observations have recently been made as the result of treatment of Leishmania amazonensis with 24-SMTI, which induced several abnormalities in the lipid content, with the accumulation of steroid intermediate molecules [12].

In addition, staining of DNA with DAPI indicates a profound alteration in the cell cycle after treatment with AZA and EIL. Candida yeasts produced unfertile buds that remained closely associated with the mother cell, and appeared with or without various nuclei. The nucleus may also have an altered shape and/or with abnormal chromatin condensation that might be associated with apoptosis cell death, as previously described after treatment of C. albicans with AMB [29]; and also after treatment of Tritrichomonas foetus with hydrogen peroxide [32]. The presence of many cells with more than one nucleus may also indicate that ergosterol biosynthesis inhibitors are interfering with cytokinesis. In fact, it was previously found that ergosterol levels modulate the activity of protein kinases such as pp60v-src and also the levels of cAMP, both of which are directly related to the control of the cell cycle [33, 34].

In addition, some studies have shown that drugs such as griseofulvin and nocodazole, which interfere with the assembly of cytoskeleton components, induce alterations in the cell cycle and apoptosis cell death [35–37]. Thus, there are two possible explanations for the alterations in the cell cycle: (a) the depletion of ergosterol and other lipids, which are essential for the maintenance of the cytoplasmatic membrane structure and are also important key regulators of the cell cycle; and/or (b) alterations in the cytoskeleton structure, which can affect the cytokinesis process. The results obtained here suggest that AZA and EIL are probably interfering with sterol biosynthesis in Candida spp., as previously described for C. albicans [20], P. carinii [13], T. cruzi [3], and L. amazonensis [12]. On the other hand, we cannot exclude the possibility that these compounds may be acting in other pathways, inducing some secondary effects that could be related to the accumulation of other lipids or, as demonstrated in Crithidia deanei, that AZA can interfere with phospholipid biosynthesis [38]. Further studies are necessary to characterise the correlation between the depletion of ergosterol and the cell cycle in C. albicans.

The results presented herein demonstrate the potential usefulness of the 24-SMT inhibitors AZA and EIL as antifungal agents, including azole-resistant Candida strains. The specific in vitro and in vivo antifungal and antiprotozoal activity of azasterols has been known for years, and in most cases has been linked to their specific inhibition of 24-SMT, an enzyme absent in mammals [10–14, 39]. However, other studies have found that these compounds are also active against parasitic protozoa that lack endogenous sterol biosynthesis, such as T. gondii [23, 40] and Trypanosoma brucei [41], indicating that they may have other biochemical targets. Taken together, these results indicate azasterols as useful leads for novel antifungal agents, but optimisation of their selectivity, ADME, PK, and toxicological properties is required for their further advancement as drug candidates.