Abstract
Although Candida species are widespread commensals of the microflora of healthy individuals, they are also among the most important human fungal pathogens that under certain conditions can cause diseases (candidiases) of varying severity ranging from mild superficial infections of the mucous membranes to life-threatening systemic infections. So far, the vast majority of research aimed at understanding the molecular basis of pathogenesis has been focused on the most common species—Candida albicans. Meanwhile, other closely related species belonging to the CTG clade, namely, Candida tropicalis and Candida dubliniensis, are becoming more important in clinical practice, as well as a relatively newly identified species, Candida auris. Despite the close relationship of these microorganisms, it seems that in the course of evolution, they have developed distinct biochemical, metabolic, and physiological adaptations, which they use to fit to commensal niches and achieve full virulence. Therefore, in this review, we describe the current knowledge on C. tropicalis, C. dubliniensis, and C. auris virulence factors, the formation of a mixed species biofilm and mutual communication, the environmental stress response and related changes in fungal cell metabolism, and the effect of pathogens on host defense response and susceptibility to antifungal agents used, highlighting differences with respect to C. albicans. Special attention is paid to common diagnostic problems resulting from similarities between these species and the emergence of drug resistance mechanisms. Understanding the different strategies to achieve virulence, used by important opportunistic pathogens of the genus Candida, is essential for proper diagnosis and treatment.
1. Virulence of Candida Pathogenic Fungi
The fungal kingdom includes many different virulent species that are able to infect, colonize, and cause diseases in humans [1, 2]. Virulence is defined as the relative ability of a microorganism to cause damage to the host during colonization, whereas virulence factors are defined as microbial attributes that mediate the destruction capacity. However, the virulence determinants are not merely proprietary and inherent attributes of the pathogen, but must be carefully considered with regard to host-pathogen interactions and the state of the host’s immunity and strength of its response to the pathogen. Therefore, host damage can result from both the activity of the microorganism itself and the action of the immune system [3, 4]. If there is no harm to the host or is not clinically relevant, such microorganisms may be considered commensals that do not provoke an excessive immune response. However, under certain conditions, the balance can shift to pathogenicity with increasing damage and activation of the host’s defense. In infections caused by Candida fungi, such conditions include a disturbance of the beneficial microflora of the host or immune disorders [5]. In susceptible hosts, Candida fungi can cause different types of infections, manifested by superficial infection of the skin and mucous membranes or deep tissue infections and candidemia [6–9].
Of several hundred known species of the genus Candida, only some are indicated to be capable of causing human infection, and they even differ significantly in their virulence and adaptation strategies, thus locating at divergent points in a continuous spectrum of the types of interactions between the host and the microorganism. Recent considerations indicate that the evolution of fungal virulence could have taken many individual directions and depended on a variety of factors, and only some of them have been recognized so far [2, 10–13]. The ability to colonize the human host by some Candida species arose from using the available properties related to their cell biology and required to adapt to specific niche conditions. The transition between the commensal and pathogenic states, closely correlated with modulation of host immunity, is based on the differential adjustment of fungal virulence to ensure the survival and multiplication of fungi, in addition to avoiding microbe removal and causing excessive host damage [2, 14]. Throughout coexistence with the human host, particular Candida species have developed different mechanisms for adaptation to colonization. Observable manifestations of these adaptations related to the biological functions of the cell, i.e., the ability to grow at higher temperature, changes in cell morphology, and the range and level of production of manifold molecules, may be similar to some extent, or they may differ significantly for individual Candida species. This could not only result in the distinct virulence potential and adaptation ability of individual Candida species but can also be of great importance in practice, affecting the correct diagnosis and effective treatment of candidiases.
So far, the most widespread and best-characterized species of the Candida genus is C. albicans, representing the so-called CTG clade, in which the CTG codon is translated as serine instead of leucine [9, 15]. The frequency and type of infections caused by selected Candida species from the CTG clade in different groups of predisposed patients, their geographic distribution, and the severity of the disease course can vary, depending on the specific susceptibility of individuals, concomitant risk factors, and environmental determinants [16, 17]. Candidiases comprise fungal manifestations on the skin and mucous membranes, such as oral or vaginal thrush, which are relatively mild disorders, although extremely bothersome, as well as severe infections of internal organs, systems, and the entire organism with widespread inflammation and sepsis. The highest distribution is still observed for C. albicans, and among the other species discussed in this review, the average incidence was significantly higher for C. tropicalis than for C. dubliniensis. C. albicans and C. tropicalis belong to a group of five Candida species that account for more than 90% of all candidial infections, both superficial and invasive [18]. The number of confirmed identified infections with C. auris has increased markedly in the last decade, since its first identification in 2009, making this species an emerging pathogen [19]. Interestingly, among the discussed species, C. auris is not considered a component of the commensal intestinal microflora, but rather as environmental contamination, and it rather inhabits the skin surface, thus facilitating transmission between individuals and the development of infections after gaining access to the interior of the human organism [20].
Since changes in the epidemiology of fungal infections have been observed in recent decades, manifested by an increase in candidiases caused by species other than C. albicans, it seems that despite their close relationship (Figure 1), these species maintain different ways of survival and reproduction in the host [21–25]. Therefore, the main objective of this review was to compare current knowledge on observable differences and similarities in virulence factors, cointeractions, and susceptibility to antifungal drugs by Candida species most closely related to C. albicans, namely, C. dubliniensis and C. tropicalis, but also by a relatively newly identified species of global concern, C. auris [26–30], to indicate the impact of differences between these individual Candida species on practical issues related to candidial infections, especially problems in proper diagnosis and treatment of candidiasis.
1.1. Variety of Virulence Factors in the Fungal CTG Clade
Candida species belonging to the CTG clade have evolved multiple virulence factors that are used at different stages of infection in interactions with host proteins and cells and in the evasion of the immune system. Although the set of virulence factors differs within the CTG clade and their expression depends on the type of strain and the stage of infection [31], they involve primarily the production of different classes of proteins, including adhesins and a wide spectrum of hydrolytic enzymes (Figure 2) [32–35]. Comparative genomic analyses indicated that for C. albicans, a significant enrichment compared to nonpathogenic yeasts was found for genes encoding secreted proteases and lipases, acid sphingomyelinases, cytochromes P450, and various transporters. Also, noticeable differences in genes related to oxidative metabolism and environmental sensing and response were indicated in comparison with nonpathogens. Understanding the differences in the genomes of individual fungal species closely related to each other, but differing in virulence, may allow the determination of the genetic basis of adaptation of yeasts to pathogenicity [36, 37].
The first landmark study comparing the evolutionary pathways of the main fungal pathogens belonging to the Candida clade, including C. albicans and C. tropicalis, identified 64 gene families with positive selection, which through their participation in filamentous growth (e.g., SWI, PMT, and CPH/STE), biofilm formation (e.g., CPH/STE and PGA/YWP), and drug sensitivity (e.g., ERG) play an essential role in the pathogenesis of these species [10]. The genomic comparative analysis of C. albicans and C. dubliniensis indicated that since both species lost a common ancestor some 20 million years ago, the former species was enriched with the virulence-related gene families, while the latter consistently lost pathogenicity-related genes, including hyphal-associated HYR1, genes for particular secreted enzymes, or IFA gene family encoding transmembrane proteins, which compromised its virulence [38]. Moreover, the difference in the number of members of the telomere-associated (TLO) gene family, represented by 14 genes in C. albicans and only two in C. dubliniensis, may be significant in understanding discrepancy in the pathogenicity of these species, as they may possibly act as transcriptional regulators involved in different cellular processes [11, 38].
The key stage in the development of a fungal infection is the adhesion of the pathogen to a variety of biotic and abiotic surfaces [39]. The possibility of C. albicans binding to various ligands on the surface of host cells, as well as to artificial surfaces, is associated with exposition of glycosylphosphatidylinositol- (GPI-) anchored adhesins: agglutinin-like sequence protein family (Als1-7 and Als9), hyphal wall protein family (Hwp1), Eap1, Iff4, Ecm33, and also noncovalently cell wall-associated proteins, i.e., Mp65 and Phr1 [40, 41]. Phylogenomic studies have identified three main families of genes that encode cell wall components, including genes for Als adhesins, Hyr/Iff proteins, and Pga30-like proteins that are highly enriched in pathogenic species, including C. albicans and C. tropicalis [10]. Analysis of predicted cell wall proteins in C. auris did not reveal highly developed families, in contrast to the expansion of genes for transporters and lipases, which may also indicate different mechanisms of virulence in C. albicans and C. auris [42, 43].
In addition to classic adhesive proteins, this function is also performed by some cytosolic proteins, called moonlighting proteins, and located on the surface of yeast cells, where they perform completely different functions than at the primary location [44, 45]. For example, C. albicans cell wall-associated glyceraldehyde-3-phosphate dehydrogenase (Tdh3), an enzyme originally involved in the glycolysis pathway, has a high affinity for extracellular matrix (ECM) components such as fibronectin and laminin [46] or other glycolytic enzymes, phosphoglycerate mutase 1 (Gpm1) which allows adhesion to umbilical vein endothelial cells (HUVEC) and keratinocytes (HaCaT) [47] and to vitronectin and fibronectin [35] and triosephosphate isomerase (Tpi1), which interact with several ECM proteins [48]. C. albicans enolase (Eno1) was also indicated as an abundant moonlighting protein responsible for binding of human high-molecular-weight kininogen (HK), prekallikrein, coagulation factor XII, and plasminogen [34, 49, 50]. Furthermore, moonlighting proteins are also involved in binding to abiotic surfaces; for example, Eno1 plays a role in silicone and polyvinyl chloride adhesion and therefore is involved in biofilm formation on medical devices [51, 52].
Less is known about the profile of adhesins in C. tropicalis. Studies by Punithavathy and Menon evaluated the presence of ALS genes in the C. tropicalis genome in clinical isolates from HIV and non-HIV patients [53]. The results indicated that the genes ALS2 (50% of isolates) and ALS3 (48%) are more common than ALS1 (28%). Recently, the use of DNA sequencing technology combining short-read (Illumina MiSeq) and long-read (Oxford Nanopore MinION) datasets allowed the identification of 13 distinct genomic loci in C. tropicalis, with predicted encoded proteins demonstrating Als structural features such as the N-terminal-binding domain, a central domain of tandemly repeated sequences and the C-terminal domain rich in serine and threonine. Detailed comparisons indicated that the predicted percent sequence identity compared to C. albicans was highest (30-50%) in the N-terminal domain of the C. tropicalis protein. Moreover, the predicted ALS genes showed variability in relative expression depending on the culture conditions [54]. Furthermore, Himratul-Aznita et al. [55] showed that the HWP1 adhesin-encoding gene is also present in the C. tropicalis genome and has a sequence identical to that of C. albicans. Bioinformatic studies conducted by Willaert et al. demonstrated the presence of the Flo adhesin family (Flo11 type), previously discovered in brewer yeasts [56]. Some reports indicated that the levels of proteins present on the surface of C. tropicalis vary with the growing conditions. For example, in plasma-containing medium, three times higher levels of Rbt1, Als-like, and also, Tdh3 were indicated [57]. Furthermore, Hyr1 adhesin was shown to have a significant affinity for ECM proteins and HK [58, 59]. In addition, C. tropicalis moonlighting proteins such as malate synthase Mls1, fructose-1,6-bisphosphatase Fbp1, Eno1, and Gpm1 may play a role in adhesion to components of ECM or plasminogen [48, 50, 58].
Several studies suggested that C. dubliniensis exhibits lower virulence than C. albicans [23, 60]. As adhesion is one of the key virulence factors, it was assumed that there may be differences in the adhesin profile of these two species. It seems that C. dubliniensis, despite its similar evolutionary origin, might have gradually lost several genes that are still present in C. albicans [38]. The percentage of homology of amino acids compared to C. albicans is less than 50 [61–63]. To date, several genes encoding proteins homologous to C. albicans adhesins have been identified in the genome of C. dubliniensis, including ALS-like sequences and HWP1 [38]. However, a detailed phylogenetic analysis suggests that C. dubliniensis has evolved a new Als family. That is, ALS3 located on the R chromosome in C. albicans was shown to be completely absent at its corresponding position in C. dubliniensis, similar to ALS5 [38], while ALS1 and ALS2 do not show an ortholog sequence in relation to C. albicans, suggesting that these proteins are species-specific acquisitions [38]. Studies by Oh et al. demonstrated that the genes encoding Als proteins in C. dubliniensis strain CD36 differ in relative expression depending on culture conditions, with the greatest changes observed for ALS2 and ALS6 [54]. In the case of HWP1, there is a homologue in the C. dubliniensis genome, but this gene is very divergent due to the presence of major deletions [63].
Studies conducted in recent years show that C. auris is as virulent as C. albicans [64]; however, the range of virulence attributes used differs between species [65]. The genome of C. auris contains genes that encode several orthologs of adhesins found in C. albicans, which are well characterized as virulence factors [66]. Namely, C. auris biofilm-forming cells possess orthologs of the ALS protein, as well as proteins such as IFF4, CSA1, PGA26, PGA52, and HYR3, which are expressed during biofilm formation [67]. Genomic analyses showed that there are differences in the adhesin profile between C. auris strains. The less virulent Clade II strain of C. auris (East Asian type), responsible for ear infections, lacks significant fragments of subtelomeric regions encoding adhesins [43].
The virulence of the CTG clade is also associated with the production of a variety of hydrolytic enzymes that contribute to the invasion of host tissues by damaging cell surface structures and degrading a variety of host proteins. In the case of C. albicans, the following classes of enzymes can be distinguished: aspartic proteases (Saps), phospholipases (PL) and lipases (LIP), esterase, and hemolysin, where most are extracellularly secreted enzymes [68]. These enzymes are differentially regulated and produced during different stages of infection [68, 69]. In recent years, aspartyl proteases have been most extensively studied. These enzymes are involved in every stage of C. albicans infection, and their functions are not limited to substrate hydrolysis. Saps contribute to the adhesion and invasion of tissues and damage a variety of human proteins, including albumin, hemoglobin, keratin, collagen, laminin, fibronectin, mucin, and immunoglobulins [70]. The multigene Sap family in C. albicans consists of at least 10 members. Sap1-8 are secreted into the extracellular space, while Sap9 and Sap10 are GPI-anchored proteins. PLs are also a broad group of enzymes that catalyze the hydrolysis of ester bonds in glycerophospholipids. These enzymes were recognized as factors that allow penetration of host cells and interaction with host signaling pathways [68, 71]. PLs can be divided into several subclasses—PLA, PLB, PLC, and PLD—depending on the phospholipid substrate [68, 72]. PLs are particularly localized at the ends of penetrating hyphae and at initial budding sites. Lipases are a relatively poorly studied group of C. albicans enzymes, and their functions in infections are not precisely known. Genetic studies have shown the presence of ten (LIP1-10) genes in the genome of C. albicans [73].
Very high proteolytic activity was also found in C. tropicalis. A study by Zaugg et al. indicated the existence of the SAPT gene family (secreted aspartic proteinase tropicalis) in the genome of C. tropicalis (SAPT1-4) [74]. Although SAPT2-4 genes are upregulated during the yeast-hyphal transition and are involved in oral mucosal invasion and damage [75, 76], earlier studies have shown that the high invasiveness of C. tropicalis may not be related to the specific expression of SAPT1 [77, 78]. On the other hand, studies indicate a low activity of phospholipases in the majority of C. tropicalis isolates [79, 80]. However, 95.8% of the C. tropicalis strains isolated from the blood of patients admitted to the intensive care unit were characterized by high hemolysin and esterase activity [81]. Hemolysin secretion by C. tropicalis caused the release of hemoglobin from red blood cells for further use as an iron source; however, the production of hemolytic factor strictly depends on culture conditions [82].
C. dubliniensis is phylogenetically very closely related to C. albicans; however, it shows significantly reduced virulence in animal models of infection that may be related to the absence of certain proteolytic enzymes. The study by Loaiza-Loez et al. showed that C. dubliniensis expresses SAP orthologs (CdSAP1-4 and CdSAP7-9), but not SAP5 and SAP6 [83]. The absence of SAP5 may partially explain the lower ability to invade host tissues because this enzyme is involved in the degradation of E-cadherin [83]. In addition, most of C. dubliniensis isolates also lack phospholipase activity [84].
Substantial proteinase activity was detected in 96% of C. auris strains [85]; however, hydrolytic enzymes have not been well characterized to date. Of the Sap family, SAP5 has been shown to be present in the C. auris genome [67]. Research by Wang and coworkers indicated that C. auris exhibits a high Sap activity at 25°C, 37°C, 40°C, and even 42°C while C. albicans showed significantly reduced Sap activity above 37°C [86]. Studies on C. auris isolates from patients with invasive infections and colonization revealed also that more than half of the isolates had phospholipase activity (67.3%) [87]. Among this group of enzymes, phospholipase PLB3 was identified in the C. auris genome [67]. In addition, more than 60% of the clinical isolates were also indicated to possess hemolytic activity [87].
In the microbial world, another important group of virulence factors includes toxins. Studies on prokaryotic microorganisms indicate that toxins are molecules that target host cells and manipulate cell signaling or induce cell death [88]. The first toxin discovered in the fungal pathogen was candidalysin, isolated from C. albicans [89]. Candidalysin is an amphipathic, α-helical peptide, derived from a polypeptide (Ece1) encoded by the ECE1 gene [90]. Recent studies indicate that candidalysin is one of the key molecules involved in the destabilization of the epithelial plasma membrane and the induction of necrotic cell death [91, 92]. Orthologs of candidalysin from C. albicans have also been found in C. dubliniensis and C. tropicalis [92]. Interestingly, studies on oral epithelial cells and artificial lipid membranes presented by Richardson et al. [92] indicated that candidalysins produced by C. dubliniensis and C. tropicalis have stronger and faster cytolytic and immunostimulatory effects than candidalysin of C. albicans. Importantly, both species cause less damage to epithelial cells than C. albicans, which may be due to lower Ece1 production, processing, or secretion during contact with the host. Since candidalysin is a classic virulence factor that involves a strong host immune response, species that prefer a more commensal mode of existence are unlikely to exploit its potential to that extent [93]. In the case of C. auris, the ECE1 gene has not been detected in the genome and, according to current knowledge, this species does not produce candidalysin [92].
1.2. Other Mechanisms Related to Virulence
Morphological change to filamentous form is considered strongly associated with the significant virulence of C. albicans and C. tropicalis [94]. In the case of C. albicans and C. dubliniensis, the different ability to produce hyphae under specific niche conditions indicates different adaptations to exist as a commensal or pathogen, as the transition to the filamentous form for which C. albicans is more predisposed is considered as increasing virulence potential, while the reduced filamentation is believed to be an adaptation of C. dubliniensis to a less virulent lifestyle [2, 23]. However, C. auris exists predominantly as blastospores [86, 95, 96], which indicates that in addition to the morphological transformation into hyphae, which is considered considerably important in virulence, there are other relevant factors that determine a species becoming the pathogen [97].
Phenotypic switching is another feature that allows yeasts to adapt to different niches [98, 99]. C. albicans, C. dubliniensis, and C. tropicalis are diploid yeasts that can exist in two different cell phenotypes. The first, called white, includes smooth and shiny cells, while the second, called opaque, includes flat and rough cells with a curved and elongated shape [100], although opaque cells, unlike white cells, did not release a strong chemoattractant for human polymorphonuclear leukocytes (PMN), thus gaining invisibility for host immune cells and protection against phagocytosis [101, 102]. On the contrary, in the case of C. auris, which is the haploid, three phenotypes were observed: white, opaque, and sectored, with all cells having shiny and smooth surfaces and the same size [95]. Interestingly, in the case of white cells, a higher rate of self-propagation was observed in liquid medium, suggesting increased stability of this phenotype [95].
An important factor contributing to the evolution of pathogenicity is reproduction, which includes three main models—sexual, parasexual, and asexual reproduction. While most fungi reproduce sexually, C. albicans has been shown to have parasexual cycles in which, after the fusion of diploid cells, rather than meiosis, coordinated loss of chromosomes occurs, resulting in viable progeny; this pathway of reproduction increases genetic diversity and contributes to adaptation to stressful environments, but also may mediate drug resistance [103]. The evolutionary role of reproduction, especially parasexual, in other medically important species of the genus Candida has become the subject of recent studies that discuss in detail the regulation of its mechanisms and importance in the evolution of pathogenicity [104–106]. Although C. albicans, C. tropicalis, and C. dubliniensis, due to their close relationship, share common features of the parasexual cycle, such as mating between diploid cells, the presence of a stable tetraploid form, coordinated loss of chromosomes, and the ability of diploid progeny to reenter the parasexual process, they also developed their unique attributes of those processes [104, 106]. For example, in the case of C. tropicalis, the possibility of generating hexaploid progeny has been reported by mating homothallic and heterothallic tetraploid products with diploid cells and the possibility of pheromone-assisted a-a homothallic mating without the need to white-to-opaque switch [104, 106]. In turn, C. dubliniensis, compared to C. albicans, shows a much higher white-to-opaque frequency and does not show clumping dependent on mating, which results in less efficient intraspecific mating than interspecies mating [104]. To date, parasexual reproduction has not been reported for C. auris; however, the presence of a complete mating-type locus suggests that this species may also be capable of parasexuality [105].
To maintain complete virulence, Candida species also need calcineurin activity, which is calcium/calmodulin-dependent serine/threonine-specific protein phosphatase composed of catalytic subunit A (Cna1) and a regulatory subunit B (Cnb1) [107, 108]. The main target of calcineurin-dependent dephosphorylation in fungal cells is the Crz1 transcription factor Crz1; however, there may be other substrates available, and signaling triggered by calcineurin activity influences many different aspects related to fungal biology, as well as their attainment of virulence [109]. For C. tropicalis and C. dubliniensis, calcineurin was shown to be crucial for hyphal growth, while for C. albicans, the role of this phosphatase in filamentation is discussed [110–112]. Furthermore, in the case of C. tropicalis and C. dubliniensis, it was also demonstrated that calcineurin is important for cell wall integrity, resistance to echinocandins and azoles, and it is essential for virulence in a murine systemic infection model [110, 111]. When comparing the latter species with C. albicans, it is considered that calcineurin is more important in controlling acidic pH homeostasis, while for C. albicans, calcineurin is more essential in serum survival compared to C. dubliniensis [112]. Orthologous genes for calcineurin and Crz1 were indicated for C. auris; however, further detailed studies of their role are required for this species.
2. Formation of Mixed Species Biofilms within the CTG Clade: Competition or Assistance during Infection?
Although most available reports focus on studying biofilms produced by one Candida species, in vivo analysis shows that these are complex communities where multiple species can coexist in the same niches of the host organism [113–118]. The classical division of communication between them is divided into two paths. The first is synergism that promotes the formation of biofilms through coaggregation and jointly increases protection against the host’s immune system. The second is antagonism based on competition for nutritional resources and mutual inhibition of growth [119]. Therefore, it is necessary to identify interactions involving several different species, as this can be of great importance in combating infections associated with the formation of a mixed fungal biofilm.
The studies presented by Pathirana et al. [118] focused on the analysis of biofilms produced under induced flow conditions that mimic the host oral environment in which C. albicans interacted with C. dubliniensis or C. tropicalis shown that after 4 hours of development of a dual-species biofilm, the percentage of adherent C. dubliniensis and C. tropicalis cells was higher compared to that of a single-species biofilm. Moreover, these species were quantitatively dominant over C. albicans, covering a larger biofilm area. However, as the biofilm reached maturity, C. albicans acquired quantitative dominance over both species. Interestingly, the coexistence of C. albicans and C. dubliniensis increased the area covered by mature biofilms compared to monospecies biofilms. In contrast, the presence of C. albicans inhibited hyphae production by C. tropicalis, and a mixed biofilm was formed more slowly than monospecies biofilms. Furthermore, it was found that the C. tropicalis cell clusters formed in sites not occupied by C. albicans cells showed high dispersibility. The authors of the study postulate that the observed interactions between C. albicans and C. dubliniensis are synergistic mainly due to the rapid growth rate of a typical mature biofilm [118], which may contribute to protection against the host immune system or to increased resistance to the antifungal drugs [120, 121]. These conclusions are consistent with the results obtained by Kirkpatrick et al. [122], who showed that C. dubliniensis could withstand competitive pressure from C. albicans and cocreate a biofilm in the urethral catheter model. Interestingly, C. dubliniensis cells were also shown to not adhere to C. albicans blastospores [118]. On the contrary, under the conditions of joint planktonic growth, C. albicans showed a significant competitive advantage over C. dubliniensis [122]. In vivo studies based on an oral-gastric infection model of murine infants showed that despite similar colony forming units (CFU), values were obtained for both species after 2 days of coinfection; the number of C. dubliniensis yeast cells was undetectable after 6 days [60].
Interestingly, in the mixed biofilm of C. albicans and C. tropicalis formed under induced flow conditions, the mature dual-species biofilm covered a much larger surface area compared to that produced by C. tropicalis itself, despite the inhibition of the morphological change of C. tropicalis cells by C. albicans. Thus, it appears that the in vivo coexistence of the two species may be beneficial to C. tropicalis [118], by the increased protection against antifungal agents noted in comparison to monospecies biofilms [123]. On the contrary, the results presented by de Barros et al. [124] suggested an antagonistic relationship between C. tropicalis and C. albicans, which was manifested by a decrease in the number of C. albicans cells and a decrease in metabolic activity in mixed biofilms. Moreover, in the presence of C. tropicalis, a significant reduction in the expression of genes involved in the morphogenesis and biofilm formation of C. albicans such as ALS3, BCR1, CPH1, EFG1, HWP1, and UME6 was shown [124]. When comparing these reports, it seems that the experimental conditions are of critical importance in the analysis of interspecies interactions [118, 124]. C. albicans and C. tropicalis show apparent differences in the adhesion and growth of biofilms in different media, which can favor the growth of only one of them. Furthermore, the stationary growth of the biofilm differs compared to the biofilm formed in the induced flow [118, 124]. However, the effect of reduced pathogenicity of C. albicans in mixed biofilm formed with C. tropicalis was confirmed by analyses carried out on Galleria mellonella larvae, where contact with dual-species biofilm resulted in significantly higher larvae survival compared to the control group, in which infection was caused by C. albicans alone [124].
Although C. auris coexists in vitro with other species [125, 126], no studies on its interaction in mixed biofilms have been reported yet.
3. Communication of Candida Species with Host and Environment
To survive and reproduce, pathogenic yeasts have developed sophisticated mechanisms that allow them to adapt to changing niches in the host body where oxygen deficiency, nutrient deficiency, or oxidative stress can occur so as to pursue existence more as a commensal or pathogen. Despite the close relationship between Candida species, adaptation mechanisms, including intercellular communication through quorum sensing molecules, changes in fungal cell metabolism resulting from stress, and response to the emergence of host immune surveillance, show significant differences (Figure 3).
3.1. Quorum Sensing Molecules Produced by C. tropicalis, C. dubliniensis, and C. auris
Within the structures of biofilms, except for the physical cell-cell interactions, the key process in detecting and reacting to changing external conditions is performed by quorum-sensing molecules. By enhancing or inhibiting the activity of many genes, released signaling molecules regulate the change in morphological form, maintaining the structure of the cell wall or phagocytic and the response to heat shock [127, 128]. This strategy benefits biofilm communities by preventing overpopulation or competition for nutrients and contributing to the spread of pathogen cells to distal sites of infection [129–131]. In the case of C. albicans, the best-characterized quorum sensing molecule so far is the secondary product of sterol biosynthesis, farnesol, which, by inhibiting the hyphal formation process and the formation of biofilms, contributes to culture growth in the form of single cells [127, 132, 133]. Despite the growing importance of quorum sensing molecules in the regulation of communication within complex communities [128, 134, 135], limited information is available on other Candida species.
In the case of C. dubliniensis, the first studies showed that the use of spent medium inhibited morphological transformation to the hyphae, as did exogenous farnesol, suggesting that this species secreted farnesol or another compound with similar activity [136]. Similar results were obtained for the two reference strains of C. dubliniensis, which showed that the presence of exogenous farnesol did not affect growth and growth rate, but was effective in blocking both the morphological changes of yeast to pseudohyphae in RPMI 1640 medium and the formation of true hyphae in the presence of FBS, wherein the observed inhibition of the morphological transformation showed some dose dependency [137]. Interestingly, achieving 100% inhibition of C. dubliniensis morphological transformation required the use of doses of farnesol lower than was necessary for the reference strain of C. albicans [137]. Finally, the analysis of supernatants from both planktonic culture and C. dubliniensis biofilm, based on high performance liquid chromatography and gas chromatography coupled with mass spectrometry, showed that, in addition to farnesol, there are also four other compounds in the fungal external environment: isoamyl alcohol, 2-phenylethanol, 1-dodecanol, and E-nerolidol; however, these alcohols were secreted with different profiles depending on the stage of cell growth and the composition of the medium [138–141]. It was shown that synthetic alcohol mixtures, at physiological concentrations, produced a similar effect to the supernatants obtained from cultures of mature biofilms, resulting in inhibition of C. dubliniensis filamentation by approximately 50%. Interestingly, only two compounds—isoamyl alcohol and 1-dodecanol—showed an effect of inhibiting the growth of planktonic cells [138].
In the case of C. tropicalis, it was shown that this yeast, after exposure to exogenous farnesol, showed a significant reduction in the production of the biofilm on the polystyrene surface, and an additional enhancement of this effect was evident when cells were grown in nutrient-poor solid medium, and the temperature was reduced to 30°C [141]. The analysis revealed the presence of quorum sensing molecules such as farnesol, tyrosol, dodecanol, isoamyl alcohol, or 2-phenylethanol [140, 142]. Analogously to C. albicans, farnesol was shown to inhibit pseudohyphae formation, mainly in the initial stage of adhesion, and reduce the number of viable cells in the mature biofilm of C. tropicalis. The opposite effect was demonstrated for 2-phenylethanol, whose presence promoted the production of pseudohyphae and stimulated an increase in the number of cells and the density of the biofilm [142]. In turn, the addition of high concentrations of exogenous tyrosol significantly reduced biofilm formation and reduced mature biofilms of C. tropicalis, manifested in a reduction in its metabolic activity [143, 144]. Moreover, the early stage of biofilm formation by C. tropicalis cells was inhibited by nerolidol, which was identified by a significant decrease in biofilm biomass and mitochondrial activity [140].
Metabolic profiling showed that the main compounds secreted by C. auris are aromatic alcohols such as tyrosol, phenylethyl alcohol, benzyl alcohol, and isoamyl alcohol, with only two compounds—phenylethyl alcohol and tyrosol revealed a significant gain [145]. According to the authors of the study, the lack of detection of farnesol may not only suggest its lower production compared to C. albicans species but also indicate that the metabolites inhibiting the development of the filamentous form of C. auris differ from those described for C. albicans [145]. Studies by Srivastava and Ahmad [146] demonstrated that the presence of exogenous farnesol in the initial stages of biofilm development significantly inhibits cell adhesion. However, along with the extension of the cell culture time of C. auris without the presence of farnesol, the created biofilms became more resistant to the action of this molecule, and to inhibit the development of the mature biofilm, it was necessary to use a higher concentration of the metabolite. By downregulating biofilm-related genes, such as IIF4, PGA7, PGA26, PGA52, and HYR3, farnesol reduces the thickness of the biofilm and reduces the viability of C. auris cells. Furthermore, farnesol has been shown to block the effect of drug resistance-promoting efflux pumps [146, 147]. A detailed report based on the analysis of genome-wide gene transcription using transcriptome sequencing (RNA-Seq) showed that the presence of farnesol causes significant changes in the expression of more than 700 genes, manifested by increased superoxide dismutase production, inhibition of manganese, zinc and iron transport by a simultaneous increase in copper content in fungal cells, as well as modulation of metabolism towards β-oxidation [147].
3.2. Environmental Stress Response and Related Changes in Fungal Cell Metabolism
Adapting to the novel environment related to the initiation or further dissemination of infection requires Candida fungi to respond immediately and effectively to the stress associated with the colonization of a new infectious niche, consisting of fluctuations in the availability of nutrients, limits in access to essential microelements, physicochemical changes in the surroundings, the influence of other microorganisms, and activity of host defense cells [148, 149]. One of the most important factors that influences Candida cell metabolism during infection is the availability and type of carbon source, where glucose is preferred. However, some environments, that is, blood or vaginal secretions, have limited glucose concentrations, or alternative carbon sources are available, including lactate, acetate, ethanol, glycerol, fatty acids, amino acids, and N-acetylglucosamine [150, 151]. When settling in a new niche, fungal cells must change their metabolism to adapt to the conditions in the existing environment and to face a combination of the different stresses encountered, i.e., thermal, osmotic, oxidative, and nitrosative stress, destructive impact of different substances on the cell wall, phagocytes, and antifungal drugs action [149, 152, 153]. In studies by Heaney et al., it was shown that different Candida species, including multidrug-resistant C. auris, displayed rather similar sensitivity to combinatory stress, including high salt concentration, alkaline, and thermal stress; however, for individual stresses, there were some important species and strain-dependent differences [154]. In C. tropicalis, high resistance to osmotic stress could also be correlated with the activity of ion efflux pumps, which could also explain the inherent drug resistance of environmental strains, especially those that inhabit marine environments [155].
Carbon assimilation by Candida cells takes place via glycolysis, the glyoxylate cycle, and gluconeogenesis, while the latter two appear to play an important role during the initiation of infection and phagocytosis of fungal cells by host immune cells, while further development of infection and tissue colonization causes metabolism to go through the glycolysis pathway [156–159]. Lactate-grown C. albicans cells were more virulent and resistant to osmotic stress, cell wall disruptors such as Calcofluor White or Congo Red, and antifungals caspofungin, tunicamycin, and amphotericin B than glucose-grown cells, while they were more sensitive to miconazole [160]. Additionally, C. dubliniensis cells grown in the presence of lactate instead of glucose were more resistant to osmotic stress and amphotericin B and C. tropicalis to amphotericin B [160]. Omic analysis showed that in C. albicans, the glycolytic pathway was enriched in proteins with higher abundance, while in C. auris, proteins involved in the tricarboxylic acid cycle were more abundant in yeast grown in Sabouraud broth [161].
Genes encoding key enzymes of the glyoxylate cycle in C. albicans, namely, isocitrate lyase (ICL1) and malate synthase (MLS1), were upregulated during phagocytosis and identified as necessary for complete fungal virulence [156]. Furthermore, many genes involved in β-oxidation of fatty acids were upregulated during the phagocytosis of C. albicans, which may prove that the acetyl coenzyme A that drives the glyoxylate cycle is the product of this process [157]. Orthologs of the ICL1 and MLS1 genes are specified for C. dubliniensis (Cd36_04240 for ILC1 and Cd36_09130 for MLS1) and C. auris (B9J08_003374 for ILC1 and B9J08_002919 for MLS1), and for C. tropicalis, both genes for Icl1 and Mls1 were reported to be regulated by the carbon source, and their expression is repressed in the presence of glucose and increased when cells grown with acetate [162, 163]. For C. albicans, disruption of the glyoxylate cycle could lead to disturbed drug efflux pumps belonging to the ABC superfamily, disorder of plasma membrane homeostasis, and decrease in chitin biosynthesis, which can increase the susceptibility of fungi to azole drugs and echinocandins [164]. Contact with macrophages may represent a glucose-depleted location for C. tropicalis, and a similar mechanism of metabolic adaptation may be demonstrated as for C. albicans during phagocytosis [156, 165]. Using the complex model of ex vivo whole blood infection, the C. tropicalis genes ICL1 and MLS1 were shown to be strongly upregulated under these conditions [97]. In the case of C. dubliniensis cells phagocytized by macrophages, upregulation of genes encoding Icl1 and Mls1 was also demonstrated [166]. Interestingly, during such model infection caused by C. auris, unlike C. albicans, C. dubliniensis, and C. tropicalis, no upregulation of the ICL1 and MLS1 genes was observed, while downregulation of genes involved in β-oxidation including CRC1 encoding mitochondrial carnitine carrier protein, FOX2 for 3-hydroxyacyl-CoA epimerase, and the POX1-3 gene for predicted acyl-CoA oxidase was detected. For C. tropicalis, the latter two were upregulated, suggesting differences in the use of various carbon sources between these species. In response to reactive oxygen species during engulfment by host phagocytes C. albicans upregulated the superoxide dismutase genes, i.e., SOD5, while C. tropicalis gene AHP1 encoding alkyl hydroperoxide reductase and genes for putative after fungal infection reduces intestinal colonization by C. tropicalis [167].
For C. albicans virulence access to microelements during infection is crucial, as this species produces proteins responsible for binding zinc ions (Pra1p; pH-regulated antigen 1) and several proteins involved in iron acquisition, i.e., ferric reductase Frp1 [168], genes for these proteins were upregulated for both C. albicans and C. tropicalis in the whole blood infection model, but not for C. auris [97]. Although another protein gene was involved in iron acquisition, FRE3 encoding ferric reductase was strongly upregulated in C. auris, but not in C. tropicalis in this model [97, 169]. For C. auris, significant upregulation of genes encoding ferrichrome siderophore transporters (i.e., SIT1) was observed, similarly to C. albicans and contrary to C. tropicalis; additionally, the C. auris gene CSA1 encoding a protein belonging to the heme-binding protein family was also upregulated under these conditions [97, 170, 171]. As evidenced by the published data, there are also differences in this regard between the discussed species in obtaining micronutrients in various infection niches where access to them is often limited by the host.
3.3. The Effect of C. tropicalis, C. dubliniensis, and C. auris on the Host
The immune system response to fungal infections has been extensively studied for several years. The recognition of Candida spp. by host cells contributes to the development of a strong inflammatory state at the site of infection. The most important immune cells involved in the first line of the antifungal response are phagocytic cells such as neutrophils and macrophages. Due to differences in the level of virulence of individual species in the CTG clade, some variations in the host response are noticeable.
An important signaling pathway involved in the response to Candida spp. is Dectin-1/caspase-associated recruitment domain adapter 9 (CARD9)/IL-17 axis [172]. CARD9 is expressed in myeloid cells (neutrophils, macrophages, and dendritic cells), although the expression of CARD9 was observed in T and NK cells. Upon activation of the CARD9 receptor, it activates NF-kB and then induces the production of cytokines, including IL-6, IL-1β, IL-23, and TNF-α, which, in turn, are involved in inducing an IL-17/Th17 response. IL-17 is an important activator of the antifungal response, as it activates signaling through the IL-17 receptor that induces other proinflammatory cytokines, antimicrobial peptides, and neutrophil chemokines that are important for antifungal activity [173]. A study by Whibley et al. showed that C. tropicalis, unlike C. albicans, requires CARD9 and TNF-α, but not IL-17, signaling. CARD9-dependent TNF-α production plays an important role in the induction of the neutrophil response [174]. C. tropicalis induces robust formation of neutrophil extracellular traps (NET) even in the absence of filamentous structure [175]. Furthermore, C. tropicalis phagocytosis is more efficient compared to C. albicans, most likely due to differences in the structure of the cell wall [176].
Differential induction of the immune response was also observed for C. albicans and C. dubliniensis. C. dubliniensis induced increased neutrophil migration and phagocytosis compared to C. albicans, but in turn activated less NET release despite a high level of reactive oxygen species, myeloperoxidase, and lactoferrin excretion [177]. A possible explanation for this phenomenon is that C. dubliniensis is phagocytosed much more efficiently than C. albicans. Considering studies showing that C. albicans escape from the interior of neutrophils by changing the morphological form, capture and immobilization with NETs appear to be an essential neutrophil defence mechanism against this species, but not for C. dubliniensis [177]. Significant differences in response to these two related species determined by cytokine production were also shown. For example, C. dubliniensis increases the production of IL-8, a key cytokine responsible for enhancing neutrophil migration, but not IL-17A [177]. On the contrary, a comparison of macrophage responses to contact with C. dubliniensis and C. albicans showed reduced production of G-CSF, GM-CSF, IL-1α, IL-1β, IL-10, IL-16, serine protease inhibitor-1, and TNF-a, when exposed to C. dubliniensis. This research suggests that C. dubliniensis primarily activates the early immune response [177].
Research on C. auris began in 2009, and therefore, the immune system response to this species is poorly understood. One of the studies on the pathogenicity of C. auris indicates that this pathogen avoids the induction of antifungal neutrophil responses. The results obtained in the Zebrafish model indicated that C. albicans induces the recruitment of 50% more neutrophils than C. auris [178]. Furthermore, microscopic observations showed that only a small fraction of neutrophils (15-20%) attempted to phagocytose C. auris, while at least half of the neutrophils were able to neutralize C. albicans by engulfing [178, 179]. The results presented by Johnson et al. [178] indicated that neutrophils did not form NETs in contact with C. auris. It has been suggested that mannosylation pathways may be the reason for avoiding an immune response by C. auris. The C. auris cell wall efficiently masks strong immunogenic patterns such as β-glucan and chitin by forming a layer of mannoproteins. Mannosylation of C. auris may be crucial for protection against neutrophil phagocytosis and activation of other types of immune response. The mannosylation pattern in C. auris differs from that of C. albicans, which may be important in its ability to be recognized by receptors such as Dectin-1 and correlate with the variation in neutrophil response in these two species [179]. Slightly different results were obtained in the work of Bruno et al. [180] where it was shown that the production of one of the neutrophil response markers—myeloperoxidase in C. auris-infected mice was comparable to C. albicans. PBMCs incubated with clinical isolates (in particular clades I and IV) of C. auris produced significantly higher amounts of proinflammatory cytokines TNF-α, IL-6, and IL-1β compared to C. albicans [180]. In these studies, it has been shown that the early (4 hours) PBMC cytokines were mainly induced by β-glucans, and it was similar to the response induced by C. albicans. On the contrary, the late host response (24 hours) was mainly activated by C. auris mannoproteins characterized by a specific structure that includes a unique M-α-1-phosphate side chain [180]. The results obtained so far indicated that the immune response induced by C. auris is complicated, and further work is required.
As mentioned above, candidal coinfection has been reported to be a serious problem that increases the risk of severe COVID-19 disease and increases the number of deaths [181, 182]. The high susceptibility of COVID-19 patients to fungal infections is associated with changes in the host immune system caused by the presence of the virus. It was shown that these patients had a higher level of proinflammatory cytokines such as IL-1, IL-2, IL-6, and TNFα. They also experienced dysregulation of monocytes, which showed a phenotypic shift from CD16+ to CD14+. In addition, BALF macrophages and neutrophil counts were increased, and the type I interferon (IFN-I) response was reduced or delayed, making it difficult to remove the virus. From the point of view of fungal infections, lymphopenia is also important, i.e., a decrease in the absolute number of T lymphocytes, which in patients with severe COVID-19 was manifested by a significant decrease in the number of CD4+ T cells, CD8+ T cells, NK, and B cells [183, 184].
4. Methods of Species Identification: Problems of Diagnostics
Despite advances in yeast identification techniques, diagnosing fungal infections remains a challenge. The specific nature of the fungi and the fact that yeast infections occur less frequently than bacterial infections make it difficult to quickly identify yeast pathogens [185]. Another problem is the correct differentiation of individual Candida species due to the fact that they are quite closely related and show similarities with each other. The selection of the appropriate diagnostic method is crucial for the efficient application of antifungal treatment. Previously, C. dubliniensis has often been misidentified as C. albicans [21, 186], and the first infections caused by C. auris were incorrectly assigned to C. sake, C. haemulonii, S. cerevisiae, and Rhodotorula glutinis [187, 188].
Various types of diagnostic methods can be used at present, and all have some advantages and disadvantages (Table 1). Currently, tests based on the comparison of fungal phenotypes could only play a complementary role when more specific methods are available.
Candida yeasts may exist in the form of single oval blastospores, which was reproduced by budding, and their size ranges between 2 and 8 μm and depends on the species. C. tropicalis cells are the largest (4-8 μm), slightly smaller are C. dubliniensis (3-7 μm) and C. albicans cells (4–6 μm), and the smallest are C. auris cells (2-3 μm) [86, 95, 96, 186, 189]. Moreover, they show the ability to change their morphological form in response to different external conditions. The yeasts can convert to the form of hyphae, which are tightly connected and form shared walls, and pseudohyphae, which are formed by budding without separation from the mother cell [22, 32, 60, 94, 186, 189–192]. These forms are similarly developed by C. albicans and C. tropicalis [94]. Additionally, large, spherical, thick-walled cells—chlamydospores—are produced only by C. albicans and C. dubliniensis [193]. In the case of C. auris, most reports indicated that it occurs in the form of blastospores [86, 95, 96], although there were reports indicating that some clinical isolates can form filaments and large aggregates of pseudohypha-like cells to which mother and daughter cells are attached [27, 96, 194, 195].
The methods based on chromogenic media are easy to perform and useful in preliminary tests, but due to the possible misidentification of closely related species, they require confirmations by more advanced methods. Several different commercially available biochemical systems based on fungal carbohydrate assimilation profiles are also often used to distinguish Candida species in addition with methods based on spectroscopy, including mass spectrometry characterized by significant precision [196–199]. The application of other spectroscopic methods in yeast identification, including rapid evaporative ionization MS (REIMS), vibrational spectroscopy, or surface-enhanced resonance Raman spectroscopy (SERRS) was described in detail in the review of Arastehfar et al. [200]. In addition to physicochemical techniques, methods based on the molecular identification of targeted DNA regions are implemented to identify Candida species. They meet the highest accuracy expectations, even for rare and uncommon species. Unfortunately, their disadvantage is still high costs and the fact that they are not widely available. Comprehensive descriptions of the different molecular diagnostic methods used in the identification of Candida species are included in several reviews [200–203] and will not be discussed in detail here.
5. Treatment of Infections and Drug Resistance
5.1. Methods for Treating Infections Caused by C. tropicalis, C. dubliniensis, and C. auris
Currently, several main classes of antifungal drugs are used to treat Candida infections, including echinocandins, azoles, allylamines, polyene antibiotics, and nucleoside analogues (mainly 5-fluorocytosine). One of the first classes of drugs introduced into widespread use in the treatment of mycoses is polyenes, complex macrolides with amphipathic characteristics, derived from the fermentation products of Streptomyces bacteria [164, 233]. Polyenes such as nystatin, amphotericin B, candicidin, and natamycin work by disrupting the fungal plasma membrane by binding to ergosterol and forming canals responsible for increased membrane permeability, adsorption to the membrane surface, and the formation of “sterol sponges” and by induction of oxidative stress leading to cell damage [164, 234, 235]. Another class of antifungals was azoles which can be divided into imidazoles and triazoles, depending on whether they have two or three nitrogen atoms in a five-membered ring. These two groups include clinically important drugs such as clotrimazole, tioconazole, and ketoconazole of the first category and fluconazole, itraconazole, and terconazole of the second [236]. The mechanism of action of azoles is the disruption of the fungal plasma membrane by altering the conversion of lanosterol to ergosterol by inhibiting the cytochrome P-450-dependent enzyme lanosterol demethylase, a product of the ERG11 gene [237–239]. Also, allylamines, i.e., terbinafine, affect ergosterol synthesis by inhibiting squalene epoxidase, a product of the ERG1 gene, although they are mainly used for the treatment of superficial candidiasis [237, 240, 241]. One of the newest group of antifungals is echinocandins, including anidulafungin, caspofungin, and micafungin—lipopeptides being derivatives of the fungal fermentation products that affect candidial cell wall integrity acting as inhibitors for beta-1,3-glucan synthase [238, 239, 242]. Azoles, echinocandins, and polyenes alter Candida morphogenesis (yeast to pseudohyphae and hyphae transformation); however, the vast majority of information about this phenomenon is related to C. albicans [243]. Nucleoside analogs, such as 5-fluorocytosine, are absorbed into the cytoplasm by cytosine permease, where after conversion to 5-fluorouridine by cytosine deaminase, they inhibit DNA or RNA synthesis [244, 245]; however, at higher doses, nucleoside analogs are toxic to humans, so they are not the first-choice drugs [246, 247]. The most recent drug introduced into use is triterpenoid ibrexafungerp, which is a glucan synthase inhibitor that weakens the fungal cell wall. It is an orally administered preparation used to treat vulvovaginal candidiasis that demonstrates antifungal activity against C. albicans, C. glabrata, C. tropicalis, C. parapsilosis, C. dubliniensis, and C. auris [248, 249]. There are also other new drugs in clinical trials, including fosmanogepix, which is an inhibitor of the inositol acyltransferase Gwt1 involved in the early steps of GPI anchor biosynthesis, the novel triazole opelconazole and the novel echinocandin rezafungin, which are described in detail by Hoenigl et al. [250]. Both fosmanogepix and rezafungin have antifungal activity against C. albicans, C. auris, C. dubliniensis, and C. tropicalis, while for opelconazole, this activity has been demonstrated so far for the two first Candida species [250]. These are promising preparations that may significantly enrich the arsenal of currently available antifungal agents and overcome the resistance problem of some Candida strains to available antifungals.
In the therapies currently used against superficial and invasive candidiasis, different antifungals from the above-mentioned classes and their combinations are used; however, due to the emerging resistance among species other than C. albicans, problems with selecting the appropriate therapy often occur. As previously described, C. albicans and C. dubliniensis demonstrate common antifungal susceptibility patterns in vitro, and most of the C. dubliniensis isolates tested were susceptible to commonly used antifungal agents, although in several studies, some strains resistant to fluconazole were identified [251–255]. Interestingly, the in vitro test showed differences in the tolerance of fluconazole or echinocandins between C. albicans and C. dubliniensis. In the case of C. dubliniensis, incomplete growth inhibition in supraMIC, known as a trailing effect, was observed with echinocandins but not with fluconazole, while for C. albicans, the reverse has been observed [256]. C. tropicalis strains are largely susceptible to azole antifungals, polyenes, flucytosine, and echinocandins; however, worldwide recorded resistance to fluconazole or amphotericin B is increasing alarmingly [17, 189]. In 2015, Kawai et al. demonstrated that C. tropicalis biofilms are susceptible to liposomal amphotericin B in high concentrations (starting at 8 μg/ml). The same study with micafungin showed that this drug reduced biofilm formation when used at low concentrations (0.125-1 μg/ml), but surprisingly, biofilm metabolic activity increased in the presence of higher concentrations of micafungin (2-64 μg/ml) [257]. This could be the effect of slow resistance development against drugs present in the growth environment, possibly by reconstruction of the cell wall structure and modification of the expression of the FKS gene. However, Marcos-Zambrano et al. demonstrated that exposing mature C. tropicalis biofilm to micafungin resulted in changes in biofilm formation: the number of hyphae and pseudohyphae was significantly reduced, although the application of liposomal amphotericin B (16 μg/ml) caused only a few biofilm alterations [258].
In general, C. auris characterizes a significantly higher tolerance level for fluconazole than C. albicans, since approximately 90% of C. auris strains demonstrate intrinsic resistance to fluconazole; however, a considerable number of strains appear to be susceptible to echinocandins, some also to amphotericin B and voriconazole [259–261]. The use of echinocandin in the treatment of candidiases caused by C. auris is recommended, but there are also reports of strains resistant to these drugs [260–262]. Therefore, combinations of antifungal drugs can be used to treat infections with C. auris, some of which show a synergistic effect [263–265]. Antifungal susceptibility is higher for C. auris planktonic cells but extremely lower for biofilms. C. auris biofilm resistance may be related to high cell density and other resistance mechanisms, rather than to biofilm matrix formation and hyphal production as is the case of C. albicans [261, 266]. Interestingly, the use of farnesol to modulate C. auris biofilm formation and efflux pump activity, also in combination with classic antifungals such as echinocandins, might be a possible novel approach to combat biofilms [146, 267, 268]. Miltefosine, a medicament used in the therapy of leishmaniasis or infections caused by Cryptococcus spp. [269], has recently been revealed as a promising drug for the treatment of C. auris infections. Applied together with amphotericin B showed a synergistic effect for several isolates tested; however, more studies are required on its efficiency for C. auris candidiases [270–272].
5.2. Mechanisms of Drug Resistance
Resistance of Candida fungi to antifungal drugs is primarily related to quantitative or qualitative changes in target enzymes, reduced contact of the drug with its target, overexpression of genes encoding multidrug efflux pumps/transporters, and biofilm formation with a dense extracellular biofilm matrix [189, 273, 274].
In C. tropicalis, resistance to fluconazole was shown to be correlated with amino acid substitutions Y132F and S154F in lanosterol 14-alpha-demethylase (Erg11p) [275, 276]. Furthermore, the resistance of C. tropicalis to azoles could also depend on the expression level of ERG11 and UPC2, the second of which encodes the zinc transcription factor involved in the regulation of ergosterol biosynthesis genes [277, 278]. Overexpression of the ERG11 and UPC2 genes was observed in drug-resistant isolates of C. tropicalis in the studies by Wang et al. [278] and Choi et al. [279]; however, not by Jin et al. [276], demonstrating the complexity and variety of resistance phenomena in C. tropicalis. In the case of the C. tropicalis biofilm, overexpression of ERG11 and MDR1 was associated with resistance to fluconazole [280]. Additionally, the presence of large amounts of dense biofilm extracellular matrix material was indicated as responsible for the C. tropicalis biofilm resistance for amphotericin B and fluconazole through impaired drug penetration [123]. The deletion of 132 nucleotides in ERG11 and substitution S258F in ERG11 play a role in the cross resistance of C. tropicalis to azoles and polyenes [281]. The lower susceptibility to antifungals might be also caused by overexpression of MDR1 (multidrug resistance) and CDR1 (Candida drug resistance) genes encoding plasma membrane MDR/MFS multidrug efflux pump and multidrug transporter of ABC superfamily, respectively. The higher expression of MDR1 combined with mutations in ERG11 was noticed for C. tropicalis fluconazole-resistant strains [276], as well as the overexpression of CDR1 and MDR1 genes [279]. Analogously to C. albicans, resistance to caspofungin in C. tropicalis might be determined by specific mutations in FKS1 gene, encoding elements of beta-1,3-glucan synthase, i.e., change of amino acid L to W in position equivalent to position 644 in C. albicans Fks1p was reported for C. tropicalis drug-resistant strain [282]. Resistance to flucytosine could also be developed by C. tropicalis and may be determined by the mutation in the FCY2 gene that encodes purine-cytosine permease [283].
Similarly to C. albicans, C. dubliniensis resistance to fluconazole was related to different molecular mechanisms, including upregulation of the genes CdCDR1, CdMDR1, and CdERG11 and the presence of mutations in the CdErg11 protein [284]. In the case of itraconazole-resistant C. dubliniensis isolates, the expression level of the CdERG11 gene was at least four times higher than for susceptible isolates, and also, the increased expression of the CdCDR1 gene was shown, but not of the CdMDR1 gene, as itraconazole is not a substrate for this protein [285–287]. Furthermore, the resistance of C. dubliniensis to itraconazole was correlated with mutations in the ERG3 gene, which encodes sterol C5,6-desaturase [287]. The resistance of C. dubliniensis also to 5-flucytosine was characterized, and a single-point mutation (substitution of serine at the 29 position by leucine) in the CdFCA1 gene for cytosine deaminase has been identified as a potential factor in the appearance of flucytosine resistance [288, 289].
C. auris is a multidrug-resistant species, and the mechanisms of its drug insusceptibility are multifactorial and overlapped [290]. Some of them are similar to those previously indicated for C. albicans, while novel molecular mechanisms and their combinations are also identified, making the outbreak of multiresistance in C. auris an extremely important problem, which was described in great detail in several reviews previously; for a comprehensive overview, see [273, 274, 291, 292]. In the case of C. auris fluconazole-resistant strains, amino acid substitutions Y132F, K143R, or Y132F in ERG11 gene have been identified as responsible for resistance to antifungal [260, 293] and additionally nonsense mutations E429 in ERG11 and W182 in ERG3 genes [294], overexpression of ERG11 [260], mutation in gene encoding zinc-cluster transcription factor TAC1b resulting in CDR1 overexpression [295], and aneuploidy including ERG11 and TAC1b genes duplication and formation of an isochromosome [296, 297]. Additionally, the sequestration of fluconazole by biofilm extracellular matrix increasing fungal resistance was demonstrated for C. auris biofilms, similarly to C. albicans [298, 299].
The balance between expression ratios of FKS1/FKS2 genes seems to be one of the main elements of echinocandin-resistance progression in C. auris [300]. Two mutations in hot spot 1 of the product of the FKS1 gene were identified for C. auris in positions S639F and S639P and additionally deletion mutation ΔF635 [260, 293, 301, 302]. In the C. auris strain resistant to fluconazole and amphotericin B, two mutations were detected: Y132F in the ERG11 gene and D709E in the CDR1 gene in the studies carried out by Reslan et al. [303]. The resistance of C. auris to flucytosine is related to the mutation F211I in the FUR1 gene that encodes uracil phosphoribosyltransferase [304].
5.3. Alternative Methods for Treating C. tropicalis, C. dubliniensis, and C. auris Infections
The increased prevalence of multidrug resistance in yeasts implicates an urgent need to develop new therapeutic strategies to combat infections caused by species other than C. albicans. In addition to the widely described chemical modification of the surface of medical devices by coating them, for example, with a chitosan hydrogel [305], other methods are still being sought that would allow combating biofilms produced within various niches of the host organism.
A promising alternative treatment for Candida infections is the combination of quorum sensing molecules manipulating fungal morphology with traditional antifungal drugs, which are the subject of clinical trials in C. albicans [143, 144, 306, 307]. Combining farnesol with traditional antifungal drugs such as amphotericin B, fluconazole, itraconazole, caspofungin, and natamycin has been shown to reduce minimum inhibitory concentrations (MIC) against resistant C. tropicalis cells [142, 308, 309]. In the case of fluconazole, it has been shown that the synergistic effect of farnesol is further enhanced when the quorum sensing molecule is previously incorporated into liposomes [310]. Similar effects were observed for exogenous tyrosol, which, when combined with traditional antifungal drugs such as amphotericin B, fluconazole, and itraconazole, significantly reduces biofilm formation and its metabolic activity; however, in the case of mature biofilms, similar effects were observed only for amphotericin B [144]. In the case of C. dubliniensis, which has the ability to develop resistance to fluconazole [285, 311, 312], the presence of farnesol increases the permeability of the cell membrane to exogenous compounds, which in turn significantly increases the sensitivity of yeast cells to fluconazole [136]. In addition, in the treatment of infections caused by C. auris, farnesol is considered one of the supportive treatments. In vivo studies in a mouse immunocompromised candidiasis model showed that both preexposure to farnesol and daily farnesol treatment of infected mice significantly reduced the yeast load of the kidneys [268]. In addition, a significant synergy was observed between triazoles, echinocandins, and farnesol against C. auris biofilms [267, 268].
Another alternative method that has successfully passed preclinical and clinical trials in the case of C. albicans infection is the use of yeast-based probiotics [313–316]. It has been shown that Saccharomyces boulardi secreted into the medium active compounds, such as caproic acid, which reduce the virulence of C. albicans mainly by inhibiting the production of hyphae [314]. Also, in the case of other Candida species, it has been shown that two strains of probiotic yeast—Saccharomyces cerevisiae and Issatchenkia occidentalis—were shown to be able to inhibit adhesion to both abiotic surfaces and epithelial cells, inhibit morphological changes and biofilm formation, i.e., not only by C. tropicalis and C. auris species but also by mixed biofilms coproduced with C. albicans [315, 316]. Interestingly, the action of S. cerevisiae and I. occidentalis was more potent than that of S. boulardi, which best inhibited the biofilm of C. albicans [314–316]. Furthermore, in vivo studies using the Caenorhabditis elegans model showed that exposure of nematodes to probiotic yeast protected them from infection, while the use of probiotic yeast after fungal infection reduced intestinal colonization by C. tropicalis [316].
There are also reports in the literature showing that photodynamic therapy using a nontoxic dye photosensitizer, which when activated with visible light, produces reactive oxygen species having a cytotoxic effect on pathogen cells, is a promising method of treating superficial infections of the mucous membranes and skin [317–319]. In the case of Candida non-albicans species, it was demonstrated that the use of photodithazine formulated in hydrogel reduced the viability and biomass of biofilms produced by C. tropicalis [320], while curcumin-mediated photodynamic therapy in combination with LED light significantly reduced the metabolism of C. dubliniensis biofilms [321]. Furthermore, in the case of C. auris, blue and red lights in combination with photosensitizers were shown to inhibit biofilm formation and disrupt mature biofilms, with only blue light itself having antibiotic properties [322].
6. Conclusions
There are many different niches in the host organism that fungi can colonize, and various sophisticated and still insufficiently recognized factors may determine the outcome of the Candida–host relationship. Due to the wide variation within the genus Candida and the multitude of factors affecting virulence, there are still cases in clinical practice that, despite the available knowledge, show unpredictable disease evolution. Therefore, it is difficult to unequivocally answer the question of why some species may prefer a more commensal lifestyle, while others show increased pathogenicity [10, 323].
Currently, it is assumed that the four Candida species discussed, although so closely related, follow different pathways of adaptation during the colonization of the human host. Understanding the mechanisms leading to these divergences and based on the evolutionary history of individual species, resulting also in the differences in their genetic background, is undoubtedly very important to elucidate the emergence of pathogenicity among these fungi. However, the similarities observed, as well as differences in biology and virulence, between different species may have a direct impact on the practical aspect of our fight against fungal infections, which is a tough challenge during the diagnosis and treatment of candidiasis.
First, differences in virulence attributes and host invasion mechanisms may determine the design of distinct therapeutic strategies for infections caused by particular species. Moreover, as some niches may be shared by several Candida species that interact with each other, it is reasonable to consider the aspects of their virulence not only in isolation from each other but also in relation to other species to apply appropriately selected alternative therapy. Second, similarities in the biology of closely related species can make a correct diagnosis of candidiasis more difficult, especially when it is based on fast and inexpensive simple diagnostic tests. It seems, particularly when mixed infections are suspected, that it is necessary to use advanced diagnostic techniques based on physicochemical methods or molecular biology. Third, awareness of differences in antifungal drug resistance between species may outweigh the success of therapy and also lead to the search for new alternative treatments, especially in multiresistant species such as C. auris.
Therefore, in the case of such a multifactorial process as the development of infections caused by various pathogens of the genus Candida, the correct, multifaceted, and comprehensive approach to genetic and phenotypic differences is crucial not only to understand the entire infection process, but most of all to fight the invaders.
Conflicts of Interest
The authors declare that there is no conflict of interest regarding the publication of this paper.
Acknowledgments
This work was financially supported by the National Science Centre of Poland (grant no. 2019/33/B/NZ6/02284 awarded to MR-K).