Abstract
Aspergillus fumigatus is a ubiquitous saprophytic fungus responsible for organic material decomposition, and plays an important role in recycling environmental carbon and nitrogen. Besides its important role in the environment, this fungus has been reported as one of the most important fungal pathogens in immunocompromised patients. Due to changes in CO2 concentration that some pathogens face during the infection process, studies have been undertaken to understand the pathogenic roles of carbonic anhydrases (CAs), well-known CO2 hydration catalytic enzymes. As a basis for a discussion of the possible roles of CAs in A. fumigatus pathogenicity, this review describes the main characteristics of the A. fumigatus infection and the challenges for its treatment. In addition, it gathers findings from studies with CA inhibitor drugs as anti-infective agents in different pathogens.
Introduction
Aspergillus fumigatus
It is a given that the fungal community contributes to ecosystem dynamics. Aspergillus fumigatus, for instance, is a widespread fungus that plays its role as a saprophyte very well. Its natural ecological niche is the soil wherein it survives and grows on organic debris, decomposing the organic matter and returning important inorganic nutrients such as carbon and nitrogen to the environment (Latgé, 1999; Mullins et al., 1976).
A. fumigatus sporulates abundantly, releasing thousands of small airborne conidia (2–3 µm in size) that can survive a wide range of environmental conditions, eventually reaching the human bronchioles or alveoli. In immunocompetent individuals, inhalation of A. fumigatus conidia will rarely cause severe consequences for their health, since the innate immune system generally eliminates the conidia efficiently. On the other hand, when inhaled by immunocompromised patients, A. fumigatus becomes an opportunistic pathogen and a serious public health problem. With A. fumigatus, the virulence is multifactorial, resulting from the combination of the biological characteristics of the fungus and the immune status of the patient, wherein the latter seems to be more important (Abad et al., 2010; Balloy & Chignard, 2009; Tekaia & Latgé, 2005).
Aspergillosis is the general term that describes a wide spectrum of diseases caused by Aspergillus spp. Such diseases manifest themselves as non-invasive saprophytic or angioinvasive diseases (Hogan et al., 1996; Hope et al., 2005). Angioinvasive diseases occur after conidia inhalation and alveolar infection. Once hyphae are formed, invasion of the pulmonary vascular endothelial cells develops (a process called angioinvasion), and an extensive dissemination to other organs through the bloodstream may be established (Filler & Sheppard, 2006; Hope et al., 2005; Pfaller & Diekema, 2010).
Several species of Aspergillus such as A. flavus, A. niger, A. terreus and A. versicolor can cause aspergillosis, but A. fumigatus reigns as the main causative agent. It was responsible for 90 % of the invasive aspergillosis (IA) cases in the 1980s, for 50–60 % between the 1990s and 2000s, and for 72.6 % between 2004 and 2008 (Pfaller & Diekema, 2010; Steinbach et al., 2012).
A. fumigatus infection can cause allergic bronchopulmonary aspergillosis in immunologically hypersensitive individuals, chronic pulmonary aspergillosis in patients with pulmonary disorders and IA in immunocompromised patients, which represents the most severe and life-threatening disease form, with a 40–90 % mortality rate (Fortún et al., 2012; Lin et al., 2001). The most frequent site of IA infection is the lung (Steinbach et al., 2012), but IA also has clinical forms of extrapulmonary dissemination manifested throughout the host bones (Winterstein et al., 2010), joints (Yu et al., 2010), heart (Kalokhe et al., 2010), brain (Thakar et al., 2012), oesophagus (Akyol Erikci et al., 2009), stomach, liver (Scott et al., 2007), intestine (Mohite et al., 2007), eyes (Pushker et al., 2011), kidneys (Jung et al., 2007) and skin (cutaneous aspergillosis) (Tahir et al., 2011) (Fig. 1).
Clinical forms of aspergillosis. Parallel between primary and secondary cutaneous aspergillosis, bronchopulmonary aspergillosis and the process of angioinvasion leading to extrapulmonary dissemination.
It is important to mention that invasive cutaneous aspergillosis is characteristic of primary and secondary infections. In primary cutaneous aspergillosis, the infection starts through contact of contaminated material in traumatized skin (i.e. intravenous and central access, occlusive dressings, burns and surgical wounds). In a neutropenic host, angioinvasion can occur after the epidermis is disrupted, which can lead to vessel thrombosis, tissue necrosis and dissemination (Walsh, 1998). In secondary cutaneous aspergillosis, the infection occurs through haematogenous dissemination from another contaminated tissue source (Fig. 1). Although considered rare, some recent outbreaks of cutaneous aspergillosis have been reported (Kim et al., 2010a; Tahir et al., 2011; van Burik et al., 1998).
Immunocompromised populations that are more susceptible to IA include patients with haematological malignancy, solid organ and haematopoietic stem cells, transplant recipients, patients with a solid tumour, human immunodeficiency virus/AIDS patients, patients with inherited immunodeficiency disorders and others (Steinbach et al., 2012). Among these, special attention is given to transplant recipients since they are the ones at highest risk of fungal infections (Neofytos et al., 2009, 2010; Steinbach et al., 2012). In addition, cytomegalovirus disease, graft-versus-host disease, advanced age, poor transplant function, use of corticosteroids, episodes of protein malnutrition, uraemia, hyperglycaemia, leukopenia, use of tubes or catheters, and multiple or acute rejection episodes are the factors that can increase the risk of fungal infection in these patients (Asano-Mori, 2010; Patel & Paya, 1997).
Other than the patient condition, another important risk factor is the presence of Aspergillus spp. in the hospital environment. Nosocomial aspergillosis can occur as a result of increasing airborne conidia and water contamination, especially during hospital renovation and construction periods (Haiduven, 2009; Vonberg & Gastmeier, 2006). Although prophylactic efforts such as the use of high-efficiency particulate air filtration have shown encouraging results in decreasing nosocomial aspergillosis, infection in the hospital setting still remains a risk for transplant patients and further measures need to be taken (Eckmanns et al., 2006; Garnaud et al., 2012).
IA treatment
Fungi have emerged as a major cause of human disease, especially among patients who are immunocompromised or hospitalized with serious underlying diseases. Although the number of invasive fungal infections has increased substantially recently, the number of antifungal drugs with novel targets, developed over the past few decades, has been restricted to the echinocandins class (Pfaller & Diekema, 2010; Shapiro et al., 2011).
Since A. fumigatus is the main pathogen involved with IA, we will provide a brief discussion about the therapy for this disease and its drug resistance data.
The sites of action of the available antifungal drugs include cell-wall synthesis, membrane function, ergosterol synthesis, nuclear division and nucleic acid synthesis (Shapiro et al., 2011). Among these, the drugs approved by the US Food and Drug Administration for treatment of IA are amphotericin-deoxychlate, amphotericin lipid formulations [amphotericin B lipid complex (ABLC), liposomal amphotericin B (L-AMB) and amphotericin B colloidal dispersion (ABCD)], itraconazole, voriconazole, posaconazole and the most recent, the echinocardin caspofungin (VandenBussche & Van Loo, 2010; Walsh et al., 2008).
Conventional polyenes (amphipathic drugs that bind strongly to ergosterol, disrupting the fungal cell membrane, e.g. amphotericin-deoxycholate) and the first generation of triazoles (antifungals that inhibit the synthesis of ergosterol by inhibiting the enzyme lanosterol 14 α-demethylase, e.g. itraconazole) are effective against several fungal pathogens, and these drugs were widely used some years ago. However, the toxicity of polyenes and the emergence of resistance among the triazoles have been limiting the use of these drugs (Bowyer et al., 2011; da Silva Ferreira et al., 2004; Denning, 1998; Jeurissen et al., 2012; Klepser, 2011).
Recent IA therapeutic studies have shown good results with the use of the second-generation triazoles voriconazole and posaconazole. Voriconazole has an excellent activity against Aspergillus spp. and is approved for IA first-line treatment, while posaconazole is approved for prophylaxis of invasive fungal infections and second-line treatment of IA (Karthaus, 2011; Messer et al., 2006; Walsh et al., 2008). Unfortunately, the appearance of A. fumigatus isolates that are less susceptible to several azoles, including the high-spectrum examples such as voriconazole and posaconazole, is increasing (Arendrup et al., 2008; Howard et al., 2006; Shapiro et al., 2011). This can be explained by the fact that they are the only agents available in oral form, being frequently used in prophylaxis and chronic infection (Howard & Arendrup, 2011; Manavathu et al., 2001). Also, azole exposure in the environment through azole fungicides, for example, can contribute to multi-azole resistance (Snelders et al., 2012; Verweij et al., 2009).
Use of lipid formulations of amphotericin B (ABLC and L-AMB) are also considered as an alternative primary therapy for some patients (Cornely et al., 2007; Herbrecht et al., 2002; Patterson et al., 2005; Walsh et al., 2008). These drugs have lower nephrotoxicity and infusion-related reactions compared to conventional amphotericin B (amphotericin-deoxychlate), which was the considered ‘standard therapy’ for IA (Barcia, 1998; Deray, 2002; Ostrosky-Zeichner et al., 2003; White et al., 1998).
Besides the triazoles and polyenes, a third treatment option for IA is the use of echinocandins (e.g. caspofungin). This class of drugs has its target as fungal cell-wall synthesis by inhibiting β-(1,3)-d-glucan synthase, and it seems to be a well-tolerated drug with rare resistant Aspergillus specimens (Kartsonis et al., 2003). However, it is not considered for monotherapy since it only has fungistatic activity against Aspergillus spp. Instead, it is considered an option when patients present with voriconazole-refractory aspergillosis or if they are intolerant to the other therapies (Denning et al., 2006; Maertens et al., 2004).
As opposed to azoles, little is known about the resistance of echinocandins. The MIC test is not considered an appropriate sensible test to evaluate the susceptibility of Aspergillus spp. to echinocandins. For susceptibility evaluation, the minimum effective concentration test seems to show better results for Aspergillus spp. However, it is still unclear how these results can be correlated with the treatment success of echinocandins, and resistance may be underdiagnosed due to technical limitations (Arendrup et al., 2008; Mayr et al., 2012). In addition, unusual IA outbreaks in patients receiving caspofungin therapy and the selection pressure caused by its use, presume possible future echinocandin resistance (Madureira et al., 2007; Mayr & Lass-Flörl, 2011; Phai Pang et al., 2012).
Carbonic anhydrases (CAs)
CAs (EC 4.2.1.1) were first studied in erythrocytes in 1933 (Meldrum & Roughton, 1933; Stadie & O’Brien, 1933), being the first enzymes known to contain zinc in their active site. These enzymes are widely found in all life kingdoms (Eukarya, Bacteria and Archaea) playing important roles in the global carbon cycle (Ferry, 2013). CAs catalyse a simple but essential physiological reaction: the hydration of CO2 to bicarbonate (HCO3−) and the corresponding dehydration of HCO3− in acidic medium with regeneration of CO2. In high concentrations of CO2, this reaction can occur spontaneously; however, thanks to these enzymes, this reaction is much facilitated (Supuran et al., 2003; Supuran, 2008a, 2010).
The rate of enhancement is about 10 000 times the natural rate of CO2 hydration, which is very important because of the physiological roles of CO2 and HCO3− (Lindskog, 1997). CO2 and HCO3− are substrates and products in several reactions, participating in processes that must be rapid, such as transport and metabolic processes (Berg et al., 2002; Imtaiyaz Hassan et al., 2013; Supuran et al., 2003).
CO2 is used by autotrophic organisms for energy production, whereas it is a product of aerobic metabolism in heterotrophic organisms. In heterotrophic organisms, CO2 is inhaled, released into the blood and transported to the lungs for exhalation. In the blood, CO2 reacts with water forming carbonic acid, which further becomes a bicarbonate ion and a proton. In this manner, the CAs are involved in several physiological and pathological processes throughout the different kingdoms of life including CO2 and pH homeostasis, CO2/HCO3− transport in various metabolizing tissues, respiration, bone calcification, retina and nervous system electric activity, CO2 fixation, biosynthetic reactions (gluconeogenesis, lipogenesis and ureagenesis), electrolyte secretion and cancer development (Bahn & Mühlschlegel, 2006; Henry, 1996; Imtaiyaz Hassan et al., 2013; Potter & Harris, 2003; Schlicker et al., 2009; Supuran & Scozzafava, 2007; Supuran, 2008b; Tripp et al., 2001).
The CAs evolved independently in five genetically distinct enzyme classes. All classes are metalloenzymes that use metal ions at the active site (Supuran et al., 2003). The α-CAs, which use Zn2+ ions at the active site, are normally monomers and rarely dimers. They are encountered in vertebrates, the cytoplasm of green plants, protozoa, algae, fungi and in some bacteria; there are 16 isoforms of α-CAs already found in mammals (except primates that possess only 15 CAs) (Cuesta-Seijo et al., 2011; Elleuche & Pöggeler, 2009a; Supuran, 2008a, 2010). The β-CAs, which use Zn2+ ions at the active site, are dimers, tetramers or octamers, and they are present in bacteria, fungi, algae, plants and some Archaea. The γ-CAs probably use Fe2+ ions at the active site, but they are also active with bound Zn2+ or Co2+ ions. The γ-CAs are trimers and are found in Archaea and bacteria; it is postulated that the γ-CA class evolved as an ancient enzyme playing an important role in the metabolism of early life (Rowlett, 2010; Supuran, 2008a, 2010). The δ-CAs, which use Zn2+ ions at the active site, are probably monomers and are present in marine diatoms. Similarly, the ζ-CAs, which use Cd2+ or Zn2+ ions at the active site, are probably monomers, and they are found in marine diatoms and cyanobacteria (Bahn & Mühlschlegel, 2006; Cox et al., 2000; Ferry, 2013; Ferry & House, 2006; Krungkrai & Supuran, 2008; Lane & Morel, 2000; Liljas & Laurberg, 2000; So et al., 2004; Tripp et al., 2001; Xu et al., 2008).
As discussed above, fungal CAs belong mostly to the β-class, but some α-CAs have been found in these micro-organisms as well. A genomic search study identified α-CAs in filamentous ascomycetes, with the exception of hemi-ascomycetous yeasts and basidiomycetes. Surprisingly, this search was able to find α-CA genes in species of Aspergillus such as A. terreus, A. oryzae, A. flavus and A. niger. However, the same was not observed for A. fumigatus, A. nidulans and A. clavatus (Elleuche & Pöggeler, 2009a) (Table 1, Fig. 2).
Phylogenetic tree of the zinc-coordinating region from fungal β-CAs. This tree was constructed by the neighbour-joining method. The topology was also evaluated by bootstrap analysis (mega5.2; 1000 repeats). The numerical values in the trees represent bootstrap results. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method; the bar indicates the number of amino acid substitutions per site. All positions containing alignment gaps and missing data were eliminated only in pairwise sequence comparisons (pairwise deletion option). Fungal protein sequences used for this figure are available at the National Center for Biotechnology Information (NCBI: ). All downloads were performed before 25 September 2013.
In fungi, the CAs seem to be located mainly in the cytoplasm and mitochondria. In silico analysis predicts the localization of two CAs in the mitochondria and another two in the cytoplasm of A. fumigatus. In A. nidulans, the single pair of CAs is predicted to be located only in the cytoplasm (Han et al., 2010). The existence of CAs with or without an N-terminal mitochondrial target is related to the duplication of genes during the evolution of filamentous ascomycetes fungi. It seems that the mitochondrial localization occurred after duplication of the genes and the appearance of isoforms, suggesting also that the translocation of the isoforms into the mitochondria may denote the gain of novel mitochondrial-specific functions (Elleuche & Pöggeler, 2009a). The mitochondrial location of fungal β-CAs was shown for the first time in the ascomycetous fungus, Sordaria macrospora. Three β-CA isoforms were found in this fungus; the cytoplasmic isoforms CAS1 and CAS3, and the mitochondrial CAS2, which was proven to be required for ascospore germination and vegetative growth (Elleuche & Pöggeler, 2009b).
CO2 sensing and metabolism via CAs play important roles in the proliferation, survival and differentiation of diverse pathogenic fungi infecting human hosts. The reason for this is the drastic difference in CO2 levels in an infected host (5 % CO2) and in the natural habitat (0.033–0.038 % CO2) (Bahn et al., 2005; Bahn & Mühlschlegel, 2006; Innocenti et al., 2008, 2010; Isik et al., 2008; Klengel et al., 2005; Krungkrai & Supuran, 2008; Mogensen et al., 2006; Rowlett, 2010; Supuran & Scozzafava, 2007).
The NCE103 gene encodes a plant-like β-CA in Saccharomyces cerevisiae, and it was shown to be present in several fungal species, including the pathogenic Candida albicans, Candida glabrata and Cryptococcus neoformans (Bahn et al., 2005; Elleuche & Pöggeler, 2010; Innocenti et al., 2009a, b; Klengel et al., 2005). Its inactivation in Saccharomyces cerevisiae (deletion mutant Δnce103) results in a high CO2-requiring (HCR) mutant, suggesting that a functional CA is an important prerequisite for Saccharomyces cerevisiae to grow under low CO2 concentrations. Furthermore, in lower concentrations of CO2, there is an increase in the CA expression in Saccharomyces cerevisiae (Amoroso et al., 2005; Götz et al., 1999).
Likewise in Saccharomyces cerevisae, β-CA deletion mutants of Candida albicans and Cryptococcus neoformans have the HCR phenotype, not being able to grow in the atmospheric concentration of CO2. Both Candida albicans and Cryptococcus neoformans seem to be similar in that their adenylyl cyclase is sensitive to physiological concentrations of CO2/HCO3−, suggesting that the link between cAMP signalling and CO2/HCO3− sensing is a conserved mechanism between the fungal species (Klengel et al., 2005; Mogensen et al., 2006).
Cryptococcus neoformans, a fungal pathogen of humans that causes fatal meningitis in immunosuppressed patients, has two potential CA homologues (Can1 and Can2), although only Can2 plays essential roles during both cellular growth in environmental ambient conditions and sexual differentiation of the pathogen (Bahn et al., 2005). High CO2 concentrations (5 %), encountered in infected hosts, induce capsule biosynthesis, an important virulence factor for this pathogen, through activation of adenylyl cyclase by bicarbonate (Mogensen et al., 2006). Furthermore, transcriptional analysis of Cryptococcus neoformans has demonstrated that CA-dependent genes are involved in fatty acid biosynthesis, organization of the polysaccharide capsule, sexual differentiation, the environmental stress response and the oxidative stress response (Kim et al., 2010b).
Physiological levels of CO2 induce filamentous growth, an important virulence factor, and promote white to opaque switching in Candida albicans, thus facilitating mating by activation of the transcription factor Flo8 (Du et al., 2012). The activation of transcription factor Flo8 can occur by direct stimulation of adenylyl cyclase activity or by another pathway, independent of adenylyl cyclase. In addition, CA is essential for the pathogenesis of Candida albicans in niches where the available CO2 is limited or not supplied by the host. The skin is an example of a biologically limited CO2 niche. On the skin, the concentration of CO2 is lower than inside the host because of the equilibrium with the atmospheric CO2 concentration (Allen & King, 1978). During Candida albicans skin infection, CA seems to play a role in the epithelial invasion by this fungus (Klengel et al., 2005). The variation of CO2 concentration in niches where Candida albicans is able to grow and infect reveals the importance of the transcription factor Rca1p, the first direct CO2 regulator of CA in yeast. The transcription factor Rca1p activates CA at lower concentrations of CO2 in Candida albicans, independent of the adenylyl cyclase and also seems to repress virulence-related genes, confirming the existence of a cAMP-independent CO2 signalling pathway. Thus, the CO2 sensing in fungi appears to be accomplished by combined pathway routes, and the relationship between them remains to be unveiled (Cottier et al., 2012).
The level of Rca1p transcripts in hypercapnia is twice as high as the level at atmospheric CO2 concentrations, while Rca1p orthologues in Saccharomyces cerevisiae and Candida glabrata, Cst6p and CgRca 1p, respectively, present no variation in expression between the different levels of CO2 concentration (Cottier et al., 2012, 2013). This expected evolutionary pliancy is also observed in other regulatory complexes between species (Lavoie et al., 2010). In all species, however, inactivation of the transcription factor results in CA induction loss in ambient CO2 levels.
A. fumigatus is also a pathogen that faces dramatic changes in CO2 concentration during the infection process. Four isoforms of β-CAs have been identified in A. fumigatus, namely cafA–D. Both cafA and cafB are constitutively and strongly expressed, while cafC and cafD are weakly expressed and are CO2 inducible. Similar to other pathogens discussed above, the double mutant ΔcafAΔcafB is not able to grow in ambient CO2 conditions, presenting an HCR phenotype. This phenotype is reversed when in an environment of 5 % CO2 (Han et al., 2010). Furthermore, defects of CAs affect A. fumigatus conidiation, which is in accord with results shown in Saccharomyces cerevisiae, where the CA seems to be involved in spore formation (Jungbluth et al., 2012). When single and double mutants of CAs were tested in a low-dose murine infection, the virulence was not affected, which was also observed with other pathogens. However, a possible role of CAs on A. fumigatus virulence has not been discarded since triple and quadruple deletion mutants of the CAs were not constructed (Han et al., 2010).
CA inhibitors (CAIs)
Sulfonamides were the first CAIs described. Since then, other chemical substances with similar activity have also started to be evaluated for prevention or therapeutic use as anti-convulsants, anti-glaucoma agents, vasodilators and diuretics (Friedberg et al., 1953; Gloster & Perkins, 1955; Gray et al., 1957; Millichap et al., 1955). Current studies suggest the potential use of CAIs as anti-cancer (Husain & Madhesia, 2012; Winum et al., 2012), anti-obesity and anti-pain agents (Supuran, 2010, 2011).
The cloning of the genome of many pathogens allows the search for anti-infective agents with a novel mechanism of action (diverse mechanism of action compared with clinically used drugs for which drug resistance was reported) through exploration of alternative routes for inhibiting proteins essential for the life cycle of the pathogens or for virulence (Del Prete et al., 2012; Supuran, 2011).
The identification of CAs in known micro-organisms such as Bacillus anthracis (Wilson et al., 2008), Helicobacter pylori (Morishita et al., 2008), Plasmodium falciparum (Krungkrai & Supuran, 2008), Mycobacterium tuberculosis (Nishimori et al., 2009), Haemophilus influenzae (Cronk et al., 2006), Vibrio cholerae (Del Prete et al., 2012), Salmonella enterica (Nishimori et al., 2011), Brucella suis (Joseph et al., 2010, 2011), Streptococcus pneumoniae (Burghout et al., 2010), Candida albicans (Innocenti et al., 2008; Klengel et al., 2005), Candida glabrata (Innocenti et al., 2010), Cryptococcus neoformans (Bahn et al., 2005; Innocenti et al., 2008; Klengel et al., 2005; Mogensen et al., 2006), Saccharomyces cerevisiae (Isik et al., 2008), Sordaria macrospora (Elleuche & Pöggeler, 2009a), A. fumigatus and A. nidulans (Han et al., 2010) has been driving studies of effectiveness and performance of CAIs as anti-infective agents.
P. falciparum presents decreased growth in the presence of CAIs. This important parasite, which is one of the responsible agents of malaria, has a purine and pyrimidine requirement during its intra-erythrocytic development. The pathway for pyrimidine synthesis in the host differs from the pathway in the parasite, being a desirable target for drug development. It is the CAs that catalyse the formation of bicarbonate, which is a substrate for the parasitic pyrimidine pathway. A ureido-sulfonamide derivative is an example of a CAI that was shown to be effective against in vitro pathogen growth (Krungkrai et al., 2008; Krungkrai & Supuran, 2008; Supuran, 2010).
In the case of H. pylori, a pathogen associated with gastroduodenal diseases, there are two classes of CA characterized: hpαCA (α-CA) and hpβCA (β-CA). CAs are necessary for H. pylori survival in the stomach, being associated with acid acclimation, urea and bicarbonate metabolism of this pathogen (Marcus et al., 2005; Stähler et al., 2005). Supiride, a benzamide derivative, presents in vitro and in vivo inhibitory activity against H. pylori α- and β-CAs and was effective in killing first- and second-line therapy-resistant strains (Morishita et al., 2008). Sulfonamides/sulfamates are also able to strongly inhibit the β-CA of H. pylori, and a group of anions and molecules that interact with zinc proteins have also shown inhibitory activity against hpαCA and hpβCA, indicating that CAIs can be a possible alternative for gastric disease therapy (Maresca et al., 2013; Nishimori et al., 2007).
The CAs of M. tuberculosis, a pathogen with a high number of resistant strains, have also been shown to be effectively inhibited by sulfonamides/sulfamates (Nishimori et al., 2009, 2010), and a new C-cinnamoyl glycoside containing the phenol moiety was the first CAI with anti-tubercular activity (Buchieri et al., 2013).
In fungal organisms, different chemotypes of CAIs have already been tested against β-CAs of Saccharomyces cerevisiae, Cryptococcus neoformans, Candida albicans, Candida glabrata and Malassezia globosa. Some of the compounds also show inhibitory activity against α-CAs, indicating the need for more selective β-CA inhibitors (Hewitson et al., 2012; Innocenti et al., 2009a, 2009b; Monti et al., 2012). In conformity with CA characterization studies, the sulfonamide inhibitor, ethoxzolamide, significantly reduces the growth of Cryptococcus neoformans and Candida albicans at ambient levels of CO2 (Klengel et al., 2005; Mogensen et al., 2006). In addition, M. globosa and other dermatophytic fungi have demonstrated fragmented hyphae under sulfonamide treatment. Results are comparable to treatment with ketoconazole, a clinically used antifungal agent (Hewitson et al., 2012).
To our knowledge, there is no CAI study for A. fumigatus. Libraries of compounds for inhibition screening and high-resolution X-ray crystal structures of possible related β-CAs are available. These methods help to identify potential specific inhibitors for drug development based on the interaction and structural analysis at a molecular level. Having said that, homology modelling studies could be done for A. fumigatus in order to bring a better understanding of its CAI profiles, like that performed for Nce103 (Candida albicans) based on the crystal structure of Can 2 (Cryptococcus neoformans) (Innocenti et al., 2009a; Schlicker et al., 2009).
In Fig. 3, we have shown the alignment of Cryptococcus neoformans and Candida albicans CA amino acids, reported in the literature, with the four A. fumigatus CAs, cafA–D. The zinc ligand Cys-His-Cys is conserved in all of them, showing their relatedness to the β-class. In α- and γ-CAs, zinc is bound to three conserved His residues. Asp/Arg pairs are also present in most of the β-CAs analysed. It is known that the fourth zinc coordination site allots its catalytic cycle in two different ways. In some β-CAs, either a water molecule or acetate is found at the fourth zinc coordination site; other β-CAs possess a conserved Asp in that region (Tripp et al., 2001).
Protein alignment for β-CAs from Candida albicans (NCBI Protein nos EEQ44200.1 and EEQ43380.1), Cryptococcus neoformans Can1 (AFR93887.2) and Can2 (AFR94409.1), Aspergillus fumigatus cafA (XP_751704.1), cafB (XP_001481412.1), cafC (XP_751882.1) and cafD (XP_001481413.1). Sequences were aligned using ClustalO () and the Pearson/fasta output was analysed using Protein Boxshade () to show identities (>50 %, black background) and similarities (>50 %, grey shading). Zn+2-binding amino acids are shown in red, Asp–Arg conserved pairs are shown in blue and plant-type amino acids are shown in yellow. Thirty amino acids were added at the N terminus of cafD, according to cDNA analysis by Han et al. (2010).
β-CAs can also be classified into two subclasses: ‘plant-type’ or ‘cab-type’ (Table 1). It seems that cafA (XP_751704.1) and caf B (XP_001481412.1) are plant-like β-CAs (clade A), cafC is a cab-like β-CA (clade D) and the subclass of caf D is unknown (Elleuche & Pöggeler, 2009a). Design of inhibitors has to consider the differences between the catalytic sites of β-CAs, since the affinity for the inhibitors seems to vary (Tripp et al., 2001).
In A. fumigatus, the lack of virulence changes in murine models by the infection of each of the four deleted β-CA mutants, and the appearance of the same HCR phenotype as Saccharomyces cerevisiae, Candida albicans and Cryptococcus neoformans in the double mutant ΔcafAΔcafB, suggests the existence of conserved pathways of CO2 sensing on fungal pathogens that face changes in CO2 concentration during infection. The CA enzymes of these pathogens seem to play a role when CO2 is not supplied at satisfactory levels by the host. It is important to highlight again that the role of CAs in the pathogenicity of A. fumigatus is not yet conclusive, since triple and quadruple deleted mutants were not constructed by more recent available studies (Han et al., 2010; Klengel et al., 2005; Mogensen et al., 2006).
As previously mentioned, A. fumigatus has four CAs belonging to two different subclasses, which could make the search for therapeutic CAIs to this organism a very challenging task. Yet, the previous steps still need to be taken to understand more about the pathways that might relate to CA/CO2 for this fungus, leading to the answers of the following questions: Is there a conserved pathway between the pathogens that face changes of CO2 during the infection? Do A. fumigatus CAs also play a role when CO2 is not supplied in satisfactory levels by the host during the infection process? If yes, would CAs be an important factor during primary cutaneous aspergillosis infection as observed in skin infections caused by Candida albicans? How do the different CAs interact with other metabolic pathways during the physiopathology of diseases caused by A. fumigatus? Would CAIs be effective as a synergistic and/or prophylactic drug in cases of aspergillosis in tissues where the CO2 concentration is limited, such as the eyes and the skin?
Despite these currently unanswered questions, the fact that several pathogenic micro-organisms such as A. fumigatus possess β-CA enzymes, which are not present in humans, increases the expectation of future studies with CAIs aiming to develop novel anti-infective agents.
Conclusion
The high number of immunocompromised patients and the use of new therapeutic strategies including broad-spectrum antibiotics, chemotherapies and transplants, have been awakening the scientific community over the last decades about the importance of a better biological comprehension of pathogenic micro-organisms in searching for new pharmacological targets against infections. A. fumigatus is one of the pathogens that still represents an important public health concern. The role of CAs in the virulence of pathogenic fungi such as A. fumigatus is not yet completely understood. Given the existing antifungal resistance data and the search for new drug targets, whether the CAIs are drugs that could be considered as new anti-infective drugs remains to be evaluated. By the same token, based on the studies presented in this review, we also expect that CAs will still deserve to be suitable subject matter in further studies involving human pathogens.
Acknowledgements
We would like to thank the Foundation for Research Support of the State of São Paulo and the National Council for Scientific and Technological Development, both agencies from Brazil, for supporting our research. The authors report no conflicts of interest.