Current Aspects in the Biology, Pathogeny, and Treatment of Candida krusei, a Neglected Fungal Pathogen

Introduction

Candidiasis is the infection caused by members of the fungal genus Candida, which can be a superficial or a deep-seated disease. The latter is often associated with high morbidity and mortality rates, in particular in hospitalized or immunosuppressed patients. A rate of 3–28 patients out of 1000 intensive care unit admissions in European hospitals develop candidemia;¹ and in the United States of America the scenario is similar, as the Transplant-Associated Infection Surveillance Network reported that 3.8% of solid organ transplant recipients developed invasive candidiasis.² Studies that have enrolled cancer patients admitted into hospitals placed in Europe or the Middle East showed a 36–39% mortality rate after one month of hospitalization,^3,4 and these figures suffer minor modifications when the mortality associated to systemic candidiasis among intensive care unit patients is analyzed, which has been calculated in 48%.⁵ This rate though can scale to figures in the range of 63–75%, depending on the hospital and the patient’s staying ward.⁶ Thus, there is no doubt that candidemia represents a global healthcare problem and a significant burden on patients.

Candida albicans is the most frequent etiological agent of candidiasis, although other Candida species are also relevant in the clinical setting, causing about 35–65% of candidemia cases.^7,8 These include Candida tropicalis, Candida parapsilosis, Candida glabrata, Candida guilliermondii, Candida dubliniensis, Candida auris, and Candida krusei.^6,9 Collectively, these species are the causative agents of non-albicans candidiasis and infections by C. krusei are characterized because of their high mortality rate (40–58%) and poor response to standard antifungal therapies.^10,12

Due to the clinical relevance of this organism and the significant amount of information generated in recent years; here, we provide a literature revision on the C. krusei basic and clinical aspects.

Biological and Fungal Aspects

C. krusei produces cylindrical yeast cells that may have up 25 µm of length (Figure 1). They usually resemble long-grain rice, which contrasts with the spheric or ovoid shape of other Candida species.¹³ Like C. albicans, C. krusei shows thermodimorphism, producing hyphae when growing at 37°C and blastoconidia and pseudohyphae when incubated at lower temperatures (Figure 1).^14,15 The colony morphology is the typical one of other Candida species, with no obvious features that could provide a hint about the species: round, creamy, and smooth whitish colonies of about 5–8 mm diameter when grown at 25–28°C in rich culturing media, such as malt yeast extract glucose agar, yeast extract peptone glucose agar, or Sabouraud agar (Figure 1). Even though colony morphology switching has been reported,¹³ no systematic attempt to classify the morphological variations has been reported, like those in C. parapsilosis, C. tropicalis, and C. albicans.^16–18 One interesting contrast with other medically relevant Candida species is the presence of sexual reproduction in C. krusei, being Issatchenkia orientalis the teleomorph.¹³

Figure 1 Candida krusei cell and colony morphology. (A) Yeast cells were grown in YPD broth until reach the exponential phase and then stained with calcofluor white, to label chitin. Scale bar = 10 µm. The arrowheads indicate the mother cells. (B) Cell filamentation was stimulated in RPMI medium incubated at 37°C. Scale bar = 20 µm. (C) A C. krusei colony grown on a YPD plate. Scale bar = 5.0 mm. Images from panels A and B were taken with a Zeiss Axioscope-40 microscope and an Axiocam MRc camera.

Like other fungal cells, a cell wall, intracellular vesicles, endoplasmic reticulum, mitochondria, ribosomes, and intracellular glycogen-like granules have been described when cells are inspected under transmission electron microscopy.¹⁹ Importantly, the microscopical examination indicates that these are mononuclear cells.¹⁹ Thus far, the study of the C. krusei organelles has not been reported in detail, with the exception of the cell wall. This bias in the study of C. krusei components is likely to be related to the relevance of this structure during the interaction with the host and because it is a target of some antifungal drugs, as revised in the following sections. The early study of the C. krusei cell wall by transmission electron microscopy showed the presence of three major layers: the outermost is an electron-dense layer that includes flocculent material surrounding the cell, followed by an electron-transparent layer in the middle with the appearance to be composed of fluffy material and scatter granules, and an innermost electron-dense layer closer to the plasma membrane.¹⁹ Recently, our group characterized the basic components of the C. krusei cell wall and found that contains the same polysaccharides found in the C. albicans wall: chitin, β-glucan, and mannans.²⁰ Even though both species have similar levels of cell wall β-glucan, the chitin content is a 4.1-fold higher in C. krusei than in C. albicans, and mannan is 34% less abundant in C. krusei when compared to the C. albicans mannan content.²⁰ In agreement with this observation, the cell wall protein content, and O-linked and N-linked mannans decorating the C. krusei wall glycoproteins were lower than those found in C. albicans.²⁰ The structural study of the C. krusei N-linked mannans indicated that the outer chain is short and lightly branched with α-1,2-mannose units,²¹ which supports our observations and contrasts with the structure of the C. albicans N-linked mannans, where the outer chain is highly branched with α-1,2-mannose units and capped with either α-1,3-mannose or β-1,2-mannose residues.²² In regard to the C. krusei O-linked mannans, these are oligosaccharides composed of α-1,2-mannose units that can contain from two to four sugar residues,²³ which again contrasts with the O-linked mannans found on the C. albicans surface, that may contain up to seven α-1,2-mannose residues.²⁴ Like other Candida species, the C. krusei mannans are modified with mannose residues bound via phosphodiester links, named phosphomannan, although the content of this is about the half of the phosphomannan found in the C. albicans cell wall.²⁰

Like in other Candida species, the C. krusei structural polysaccharides chitin and β-1,3-glucan are localized underneath other cell wall components, and this impairs the proper sensing of these polysaccharides by the host immunity.²⁰

Thus far, only one report dealing with the C. krusei cell wall proteome has been reported, but this was performed with walls from cells growing in the presence of oxidative stressors.²⁵ Interestingly, only moonlighting proteins were identified, which could be a result of contaminants from intracellular compartments, since cells were disrupted with an ultrasonic homogenizer.²⁵ Nonetheless, the presence and abundance of canonical cell wall proteins found in other Candida species remain to be established.

The metabolism of this fungal species is another aspect poorly studied to date. This yeast cell is capable of using exclusively glucose as carbon source,¹³ which is a trait exploited in its identification by zymograms and chromogenic culturing media.²⁶ This has also been taken in advantage to produce and accumulate glycerol in fermentative processes with potential industrial applications,²⁷ and to prepare traditional meals and alcoholic beverages used by some African communities.^28,29 The fermentation process involving C. krusei is positively affected by the presence of lactic acid bacteria, which promote tolerance to short-term changes in the extracellular pH.³⁰ Interestingly, and contrary to this restricted carbohydrate assimilation, C. krusei has been isolated from the decaying wood from Ficus religiosa; suggesting this could be an environmental niche of this fungal species.³¹ Even though arabinitol is produced by many yeast-like cells and the presence of this metabolite has been reported in serum from patients with invasive candidiasis, C. krusei is incapable of producing this five-carbon polyol.³²

In regard to the C. krusei genome, it has been reported this species does not belong to the CUG clade of the Candida genus, it is a diploid and heterozygous organism, with the genetic information distributed in five chromosomes.^33–35 The C. krusei genome sequence of a clinical isolate (strain 81-B-5) showed that the nuclear genome sequence size is 10.9 Mbp, the mitochondrial genome contains 51.3 kbp, the single nucleotide polymorphism rate was calculated in 1/340 bases, being higher than that reported for C. albicans isolates.³⁶ Moreover, it was reported a GC content of 38.42%, a 2.15% of repeat content with no significant similarity to the repeat sequences found in C. albicans, and a total of 4949 protein-encoding genes.^33,35 The number of intron-containing genes in the coding regions was calculated in 205.³⁵ Different from other Candida species where high variability in the mating loci has been reported,³⁶ these subtelomeric loci are complete in C. krusei.³³ A recent study reported the sequencing of 32 strains of C. krusei, Pichia kudriavzevii, Issatchenkia orientalis, and Candida glycerinogenes and found that they are the same species with more of 99% identical DNA sequences.³⁵ Since the analysis of single nucleotide polymorphisms could not segregate between clinical and environmental strains, it was suggested that infections by C. krusei are acquired from the environment.³⁵ Finally, the analysis of the genome sequences supported the re-classification of these organisms in the Pichia genus, being a distant relative of the Candida species.³⁵

Recent Understanding of the Host-Pathogen Interaction

Virulence Factors

Since C. krusei and C. albicans belong to the same taxonomical genus, it has been assumed they share biological traits that help them to interact with the host, a wrong rationale that applies not only to C. krusei but other medically relevant non-albicans Candida species.^20,37-39 Therefore, the study of C. krusei virulence factors is a research area with limited information, if compared with the vast amount of reports dealing with C. albicans virulence. In this section, we will provide the most relevant information about C. krusei virulence factors and a genomic comparison to predict putative orthologs of well-known factors already described in C. albicans.

Virulence is classically determined by the ability to damage cells, tissues, organs, or a whole organism, and in mycology, both the in vitro and in vivo systems are thoroughly used. In the murine model of systemic candidiasis, C. krusei was incapable of killing both female or male mice, contrasting with the high mortality rate associated with C. albicans.^40,41 When the colony-forming units of these organisms were analyzed in infected spleen, liver, kidneys, and lungs, a gradual reduction in the C. krusei burden was observed during the observation period, with a fungal clearance at day 21 post-infection, contrasting again with the fungal loads in C. albicans-infected organs, which were constants or slightly reduced during the same observation period.⁴⁰ Therefore, C. krusei displays lower virulence than C. albicans in the murine model of systemic candidiasis. This in vivo system is regarded as the gold standard to assess the virulence of Candida species and isolates, but in recent years, logistical issues to include large numbers of animals per experimental condition and the increased ethical concerns about the use of these animals in basic research have limited their inclusion in the experimental design and stimulated the search for alternative models to study fungal virulence. Caenorhabditis elegans is an invertebrate model that has been used as an alternative for studying the Candida species virulence.⁴² Upon administration of fungal cells by feeding, C. krusei and C. albicans showed similar ability to kill C. elegans and were ranked as the most lethal species in this experimental setting.⁴² Like C. albicans, C. krusei was capable of producing aspartyl proteinases, phospholipase, hemolysins, and to develop biofilms, providing a possible explanation to the lethal behavior in the C. elegans system.⁴² Moreover, this invertebrate model has been useful in establishing the effect of the antibacterial drugs cefepime, imipenem, meropenem, and vancomycin on Candida spp. virulence. C. albicans, C. parapsilosis, C. krusei, and C. tropicalis incremented the proteolytic activity and killing of C. elegans upon incubation with these drugs, whereas amoxicillin potentiated the virulence of C. krusei and C. tropicalis.⁴³ The wax moth Galleria mellonella has been proved as a good model to study infections caused by C. krusei. Upon injection into the hemocele, fungal cells decreased hemocyte density, induced melanization and animal dead.⁴⁴ The virulence in this host is similar to that observed in the murine model, with C. krusei showing low to moderate ability to kill G. mellonella (median survival of larvae was 7 days), which contrast with the high mortality associated with C. albicans infection (median survival of larvae was 2 days)⁴⁵ In addition, these larvae have helped to propose that Lactobacillus paracasei, Lactobacillus fermentum, and Lactobacillus rhamnosus, acid bacteria used as probiotics, have a prophylactic effect on the larvae, increasing the animal survival upon administration of a C. krusei lethal dose.⁴⁶ Drosophila melanogaster is another invertebrate model that has been used to evaluate the C. krusei virulence. Adult flies with mutations in the toll signaling pathway were highly susceptible to infection with either C. albicans or C. krusei, demonstrating this model is useful for virulence assessment.⁴⁷ Moreover, these data strongly suggest that C. krusei lethality depends on the immunological status of the host, as in immunocompetent animals this Candida species was not capable of killing the host.^40,41

Cell and tissue adhesion are part of the early stage of the Candida-host interaction and will lead to the establishment of both commensalism and pathogenesis. Adhesion is paramount to establish colonization and tissue invasion in the oral epithelium, as this tissue is in constant contact with saliva, which cleans the epithelial surface. C. krusei binds to human buccal epithelial cells but not as efficient as C. albicans and C. tropicalis.⁴⁸ Accordingly, exposure of the epithelial cells to the minimum inhibitory concentration of nystatin affected the Candida species adhesive properties, with C. krusei showing a 64% reduction of adhesion to epithelial cells, a value higher to that found in C. albicans (54%).⁴⁸ Similar to epithelial cells, endothelial cells are also a surface where C. krusei can adhere, but not as efficiently as C. albicans.⁴⁹ In agreement with these observations, C. krusei showed 11-fold lower colonization potential of the rat oral surface than C. albicans.⁵⁰ However, C. krusei adheres in great numbers to acrylic surfaces.⁵⁰ Interestingly, when the C. krusei phenotypical switching was induced with phloxine B, a 30-fold increment in adhesion to saliva-coated glass surface was observed.⁵¹ The Eap1, Iff4, Mp65, Phr1, Int1, Ecm33, and ALS gene family members are the major C. albicans adhesins.^52–58 The putative functional orthologs of the genes encoding for Phr1 and Int1 were identified within the C. krusei genome but no those encoding for Eap1, Iff4, Ecm3 or any of the Als family members (Table 1). Interestingly, three putative orthologs of C. albicans Mp65 were identified (Table 1).

Table 1 Prediction of Some Virulence Factors in Candida krusei

The cell wall hydrophobicity is an important aspect of interaction with the host components. A study that involved 20 C. krusei isolates showed these had higher wall hydrophobicity than C. albicans cells, and there was a correlation between hydrophobicity and adhesion to HeLa cells, but not to acrylic surfaces.⁵⁹

Several studies have demonstrated that C. krusei, like other Candida species, is capable of secreting hydrolytic enzymes that could degrade host macromolecules, contributing to nutrient acquisition, to degrade immune effectors, or in the dissemination within the host tissues. In two independent studies, using clinical isolates, C. krusei did not show phospholipase activity, contrasting with C. albicans strains that had a strong presence of this enzyme activity.⁶⁰ However, other studies have shown that this enzyme activity is found in C. krusei.^61–63 Whether this discrepancy is due to different methodologies used to measure phospholipase activity or reflects the phenotypical plasticity of the species remains to be addressed. Nonetheless, the C. krusei genome contains two putative orthologs of the PLB gene family that encodes for the major C. albicans secreted phospholipase activity (Table 1).^64,65 Interestingly though, no putative orthologs of the genes encoding for secreted lipases belonging to the C. albicans LIP gene family⁶⁶ were found within the C. krusei genome (Table 1).

Other hydrolytic activities like proteinase, hemolytic factors, and DNase have been reported in C. krusei.^60,67,68 A study that included clinical isolates from Turkish patients diagnosed with candidiasis, found that about half of the C. krusei isolates formed biofilms, 22% showed coagulase activity and all the isolates were capable of hemolyzing red blood cells.⁶⁹ In C. albicans, most of the secreted proteolytic activity is associated with members of the SAP gene family.^70,71 A search for putative orthologs of members of this gene family in C. krusei identified SAP1-SAP5, but not SAP6-SAP10 (Table 1). These genes are likely to account for the secreted proteolytic activity reported in C. krusei clinical isolates.

C. krusei is capable of forming biofilms on polyethylene, polyvinylchloride, and glass.^67,72 These fungal biofilms are particularly sensitive to fluconazole when generated on polystyrene surfaces.⁷³ This represents a promising observation that remains to be confirmed in vivo. In agreement with the reported ability to generate biofilms, the C. krusei genome contains key genes involved in the establishment of these multicellular communities. The chaperone Hsp90, besides contributing to the establishment of apical growth during the C. albicans dimorphism, is required to neutralize the cellular stress generated during biofilm formation;⁷⁴ while Efg1, Brg1, Zap1, and Bcr1 are transcriptional factors required for biofilm formation in both in vitro and in vivo conditions.^75–77 Putative orthologs for these genes were found within the C. krusei genome (Table 1). However, for the case of Rob1, a transcriptional factor required for biofilm formation in C. albicans,⁷⁶ no putative ortholog was found within the C. krusei genome (Table 1), suggesting that the regulatory network of biofilm formation in both organisms could share master regulators but their contribution might be species-specific.

As mentioned, C. krusei belongs to the members of the genus capable of forming true hyphae. The C. albicans dimorphism has been associated with tissue invasion and the expression of several virulence factors that are morphology specific.⁷⁸ The Hgc1 is a hypha-specific G1 cyclin essential for the establishment of the apical growth and is negatively regulated by the transcriptional repressors Nrg1 and Tup1; while Cph1 and Efg1 are transcriptional factors required to sustain hyphal growth.^79–81 The C. krusei genome contains putative orthologs of these genes (Table 1), suggesting the central regulatory network that controls dimorphism is similar in both species.

In this regard, it is noteworthy to mention that C. krusei is capable of inhibiting the C. albicans filamentation and biofilm formation.⁸² Even though there is no formal explanation yet for these observations, the production of signaling molecules by C. krusei to avoid polarized growth in C. albicans, competition for nutrients, adherent surfaces, and space between the biofilms are the main hypotheses currently under study.⁸²

Gliotoxin is an immunosuppressive compound from the fungal metabolism and has been suggested as a fungal virulence factor, including in C. albicans.⁸³ However, a chemical analysis based on highly sensitive HPLC and tandem mass spectrometry of 100 clinical isolates of Candida spp, including C. albicans and C. krusei did not detect intracellular or extracellular gliotoxin production, suggesting this compound does not participate in the pathogenesis of Candida spp.⁸³ In the same line, the C. albicans cytolytic peptide toxin named candidalysin, which is essential for mucosal infection and encoded by ECE1,⁸⁴ has no putative ortholog within the C. krusei genome.

It is noteworthy to mention that C. krusei has been isolated from bat feces in a Brazilian urban region, and showed the ability to secrete proteases, to form biofilms, and kill laboratory animals, suggesting the animal depositions could be an environmental source for C. krusei infections.⁸⁵ Similarly, C. krusei has been isolated from droppings of healthy breeding rheas, chickens and hens,^86,87 and from the vestibule and vagina of healthy female horses.⁸⁸

The C. krusei-Immune System Interaction

Similar to other pathogens, the C. krusei recognition and interaction with components of the host immunity is required to establish a response that could protect against the infection. Both the innate and adaptive branches of immunity are essential to control fungal pathogens, including C. albicans and other causative agents of candidiasis.^89,90

As part of the humoral factors that belong to the innate immunity, some cells produce antimicrobial peptides that show antifungal properties. The human β-defensin 2 is produced by epithelial cells, while the human neutrophil peptides 1–3 are α-defensins synthesized by circulating white blood cells.⁹¹ Both kinds of antimicrobial peptides were stimulated by C. albicans and C. krusei, being the former a stronger inductor than C. krusei cells.⁹¹ With no doubt, this study showed the ability of C. krusei to stimulate both local and systemic responses against this pathogen. Similarly, it was reported that C. krusei is 1.4 times more sensitive to lactoferrin, a secreted antimicrobial protein, than C. albicans cells; and this difference has been suggested to be relevant to modulate the fungal oral carriage.⁹²

The peripheral blood mononuclear cells (PBMCs) are a heterogeneous group of immune cells with the ability to produce cytokines upon the interaction between pathogen-associated molecular patterns and their pattern recognition receptors.

Different to the human PBMCs-C. albicans interaction, where low levels of TNFα, IL-6, IL-1β, or IL-10 were stimulated, the immune cells interacting with C. krusei produced higher levels of these four cytokines.²⁰ Both C. albicans and C. krusei heat-killed cells, which expose inner wall components like β-1,3-glucan and chitin at the cell surface, stimulated higher levels of TNFα, IL-6, IL-1β, or IL-10, when compared to live cells.²⁰ Interestingly, a difference between C. krusei and C. albicans was observed when O-linked mannans were removed from the cell wall: C. albicans recognition by PBMCs was not affected, indicating this wall component is dispensable for cytokine stimulation; while in C. krusei, cells with no O-linked mannans on the surface were capable of stimulating higher cytokine levels, most likely because of the unmasking of β-1,3-glucan and increased recognition via dectin-1.²⁰ Like in C. albicans, this immune receptor is essential to control C. krusei infections, as dectin-1 knock out mice are more susceptible to C. krusei and showed poor ability to establish a protective anti-C. krusei immunity.⁹³

In contrast with C. albicans cells, C. krusei yeast cells induced lower levels of complement components C3 and factor B, and the granulocyte-macrophage colony-stimulating factor, but a significant amount of IL-12 (p70).^94,95 This differential ability to stimulate IL-12 (p70) could be part of the C. albicans strategies to avoid the establishment of an effective type I immune response against this pathogen, a situation likely to occur fo the case of C. krusei.⁹⁴

Contrary to this observation though, C. albicans is more readily phagocytosed by neutrophils than C. krusei cells (37% vs 9%),⁹⁶ and more susceptible to the neutrophil-expressed antimicrobial protein S100A12 than C. krusei,⁹⁷ underscorings that the differences in pathogenicity and control by the innate immune system are difficult to be reduced to the analysis of a handful of biological parameters.

Like other pathogens, C. krusei is capable of interacting with macrophages, but interestingly the outcome is variable. Rat alveolar macrophages phagocytosed C. glabrata and C. albicans in a similar rate, but this was significantly slower for the case of C. krusei, due to reduced attachment.⁹⁸ This differential recognition was abrogated though when the fungal uptake was performed with opsonized cells.⁹⁸ For the case of primary human PBMC-derived macrophages, the results are the opposite: C. krusei is more readily phagocytosed than C. albicans, C. auris, C. tropicalis, and C. guilliermondii.²⁰ In mice, both a macrophage-like cell line and primary macrophages are capable of uptaking C. krusei yeast cells, with around 10 to 20% of the immune cells ingesting yeasts after 2 h, without opsonization.⁹⁹ However, C. krusei was capable of surviving and undergoing filamentation inside the phagocytic cells, induced defects in the phagolysosome maturation, yeast transfer between infected macrophages, macrophage fusion, and death of the immune cells.⁹⁹ These data clearly show that the origin of the immune cells has to be taken in to account before drawing general conclusions.

Pyroptosis, an inflammasome-mediated macrophage death process, is activated upon interaction with C. albicans cells.¹⁰⁰ This caspase-1-, ASC-, and NLRP3-dependent pathway is triggered in lower levels by C. krusei cells and does not restrict the fungal replication.¹⁰⁰

The interaction between C. krusei with dendritic cells has particular outcomes too. The C. krusei mannan but not the cell wall component isolated from C. albicans, C. tropicalis or C. glabrata induced strong cytokine production by these immune cells and led to apoptosis.¹⁰¹ These effects on dendritic cells were mediated by TLR2 and activation of a MyD88-dependent pathway, which controlled the production of the polarizing cytokines IL-12 and IL-6, and thus the Th1/Th17 switching.¹⁰¹

Interestingly, human PBMCs tend to proliferate in the presence of either voriconazole or caspofungin and produce increased levels of IL-2, IFN-γ, and IL-6 when stimulated with either C. albicans or C. krusei, with no effect on the stimulation of TGF-β and IL-10.¹⁰² These data suggest that antifungal therapy has a positive immunomodulatory effect on human PBMCs, an observation that should be further explored and taken into account during the treatment of candidiasis and other fungal infections.

Another promising study on new immunomodulatory approaches for the treatment of candidiasis caused by C. krusei involves chromogranin A, a mammalian-expressed soluble protein of the adrenal medullary chromaffin granules and neurons. Chromogranin A N-46, a 46 amino acid portion of the chromogranin A N terminal has shown to have antifungal properties.¹⁰³ In line with this observation, treatment with this peptide (60 mg/kg/day) had positive effects on mice infected with C. krusei. Treated animals showed increments in the body weight and survival, along with higher counts of circulating monocytes, lymphocytes, and neutrophils.¹⁰³

Candidiasis Caused by C. krusei

The list of the etiological agents of candidiasis is vast and new species have been added in recent years; however, most of the cases are caused only by five species, named C. albicans, C. glabrata, C. tropicalis, C. parapsilosis, and C. krusei.¹⁰⁴ Even though C. albicans is the most common cause of candidemia worldwide, infections caused by C. krusei are an emergent public health threat. Data from the ARTEMIS DISK registry gathered from 1997 to 2007 indicated that the frequency of C. krusei-associated infections was stable, ranging from 1.7 to 3.2%.¹⁰⁴

A report from a tertiary care academic hospital in Montreal, which included 190 cases of candidemia, found that C. krusei was the causative agent in 7% of the patients, but this increased up to 13% in the period of 2003–2006. ¹⁰⁵ The most likely explanation for this observation in a specific period of time was the systematic prophylaxis with fluconazole in risk groups to develop candidemia.¹⁰⁵ Similarly, a Cancer Center of Texas, USA, and a tertiary health care center localized in Haryana, India reported that 8% and 9% of candidemia cases were associated with C. krusei, respectively, but in these cases, the studies were conducted in a pediatric population.^106,107 Another study from the USA, but in this case performed in Ihowa, reported that 3.4% of candidemia cases were due to infection with C. krusei, and these showed the lowest 90-day survival rate.¹⁰⁸ A study that enrolled 26 short-stay university hospitals in the Paris area reported that the candidemia incidence caused by C. krusei in patients not admitted to ICU was 5.2% in patients with hematological diseases, 3.7% in patients with an oncological condition, and 1.2% in patients with no malignancy diagnosed; while in patients in ICU, the figures were 5.1%, 4.4%, and 2.3% for patients with a hematological disease, oncological conditions, and no malignancy, respectively. ¹⁰⁹ Similarly, studies conducted in the Republic of Korea, Australia, Spain, the USA, India, Taiwan, Saudi Arabia, and Portugal reported that 2.6%, 4%, 6%, 5%. 3.3%, 4%, 6%, and 5% of candidemia cases were caused by C. krusei, respectively, with 25% of the Portuguese isolates resistant to posaconazole.^110–118 Contrary to these figures though, a study based on a tertiary care hospital in North China found that only 0.9% of the candidemia cases were caused by C. krusei.^110–118 Similar to this Chinese study, reports from two Greek, one Brazilian, one Swiss consortium of hospitals, and one Mexican tertiary hospital found that 1.8%, 0.2%, 1%, 2%, and 2.2% of candidemia cases were caused by C. krusei, respectively,^119–123 and a study conducted in neonates admitted to ICU, participating in the National Nosocomial Infection Surveillance system from 1995 to 2004 in the USA reported only 0.15% of candidemia cases associated to C. krusei.¹²⁴

As an outlier report, a study carried out in the University of Texas M. D. Anderson Cancer Center with clinical records of patients admitted from 1993 to 2003 found that C. krusei was the causative agent of 24% and 2% candidemia cases in patients with hematological malignancies and solid tumors, respectively.¹²⁵ The authors of this study proposed that this prevalence in patients with hematological conditions is due to the prevalent use of fluconazole as a prophylactic antifungal agent, especially in patients with hematological malignancies and recipients of bone marrow transplantation.¹²⁵ The figures above reported were similar in the period of 2001 to 2007 in the same cancer center, where 17% of candidemia cases were caused by C. krusei.¹²⁶

The risk factors for fungemia due to C. krusei include the recent surgery report (< 30 days), artificial implants, splenectomy, neutropenia, the presence of oncological conditions such as solid tumors, acute leukemia, or lymphoma as an underlying disease; stem cell transplantation, preexposure to fluconazole, echinocandins or antibacterial agents, specifically vancomycin or piperacillin-tazobactam.^{104,109,110,125,127,128} At first glance, it is difficult to relate the use of antibacterial agents with the risk to acquire an infection caused by C. krusei or other Candida species. It has been proposed that vancomycin can alter the ecology of the normal skin microbiota, promoting colonization by Candida species and thus increasing the potential to develop a systemic infection; while anti-anaerobic antibacterial agents such piperacillin-tazobactam, may promote overpopulation of yeast species and colonization of the gastrointestinal tract.¹²⁸ In neonatal patients, among the risk factors associated with C. krusei fungemia are parenteral nutrition, recent fluconazole exposure, use of broad-spectrum antimicrobials, and the presence of a percutaneous inserted central catheter.^129,130

Besides the systemic disease, C. krusei is also associated with superficial infections. This organism can cause bronchopneumonia and vulvovaginal candidiasis but is a rare etiological agent in the latter, being isolated only in 0.1% of cases and has a good response to nystatin.^131,132 C. krusei has been also found infecting the tonsils, where only surgical removal of the organ offered a permanent cure, causing septic arthritis, ulcers, urinary tract infections, and vasculitis^133–136

In veterinary, this organism can also cause infections and deteriorate the health conditions in animals. C. krusei was reported as the causative agent of bovine bronchopneumonia and mastitis in Japan, China, Turkey, Algeria, Canada, Polony, and the United Kingdon;^137–143 while in cats could be responsible for the failure of long-term gastrostomy tubes.¹⁴⁴ For the case of bovine mastitis, it has been suggested that wheat silage, rather than unappropriated milking is the source of the C. krusei cells affecting the udder.¹⁴⁵ Despite it is part of the normal microbiota of birds, C. krusei has been associated with gastrointestinal diseases in white-crowned parrots (Pionus senilis),¹⁴⁶ and acute necrotizing ventriculitis in Eclectus parrot (Eclectus roratus). ¹⁴⁷

C. krusei Identification

Since C. krusei belongs to a fungal genus that contributes with several species as etiological agents of candidiasis, the methods for identification of C. krusei have been originally developed to discriminate C. albicans from other species. Nonetheless, the following strategies have been refined to identify other Candida species, including C. krusei.

Biochemical reactions are the most common strategies for speciation of Candida isolates, and these have been taken into account to develop chromogenic media that easily differentiate Candida species, depending on the color and morphology of the colonies growing on the plates. C. krusei generates purple fuzzy, large rough colonies with flat pale edges when grown on HiCrome Candida; while the color change to pink fuzzy when cultured on CHROMagar Candida (CHROMagar) or CHROMagar-Pal’s plates.¹⁴⁸ In Brilliance™ Candida Agar (formerly Oxoid Chromogenic Candida Agar, OCCA) this species grows like dry, irregular pink-brown colonies.¹⁴⁹ Even though this colony color could be informative for species identification, it could be mistaken with the one generated by other species that develop pinkish colonies, such as C. parapsilosis, Candida kefyr, and Candida haemulonii.^148,149 Another medium for quick detection of Candida species is CandiSelect™ 4 (BioRad), where C. krusei generates large turquoise-blue colonies with a characteristically rough morphotype, a dry appearance, and an irregular outline. However, C. tropicalis and C. glabrata also grow like smooth, turquoise colonies, making the species identification troublesome.¹⁵⁰ Due to these limitations, these media are used for preliminary species identification, and additional phenotypic or genotypic tests should be included for the proper identification of C. krusei and other Candida species. The API ID32C method is currently the gold standard for phenotypic characterization of these organisms, but another alternative is Micronaut-Candida (Bornheim), a microplate-based system that contains 21 biochemical reactions, and 14 carbohydrate assimilation tests (melibiose, D-xylose, L-rhamnose, gentibiose, D-glucose, inositol, cellobiose, saccharose, trehalose, galactose, maltose, lactose, raffinose, and a control reaction), and urease test with its control. Results are generated in 24 h and interpretation assisted by the Micronaut software.¹⁵¹ This strategy proved to be as good as the API ID32C method for the C. krusei identification.¹⁵¹ The Vitek 2 system (bioMérieux) is an automated alternative for C. krusei identification based also on biochemical reactions and has 100% specificity to identify this species.¹⁵²

Another alternative for Candida identification based on phenotypic traits is the analysis of volatiles using offline gas chromatography and mass spectrometry. The p-xylene, 2-octanone, 2-heptanone, and n-butyl acetate are signature volatiles of the C. krusei presence in in vitro analyses.¹⁵³

Among the molecular strategies, PCR is one of the techniques thoroughly analyzed and applied for the identification of C. krusei and other Candida species. It was reported that a single primer pair aiming to amplify a fragment of L1A1 gene, which encodes for a cytochrome P-450 lanosterol-14α-demethylase, was capable of detecting fungal DNA in clinical specimens with a sensitivity of 200 fg of DNA but the amplicon size was not useful for Candida species discrimination, as this was in the range of 336 to 350 bp for all the analyzed species.¹⁵⁴ However, combined with restriction enzyme analysis using HincII, NsiI, and Sau3A it was generated a species-specific pattern of restriction fragments.¹⁵⁴ Similarly, the amplification of the gene coding for the small ribosomal subunit 18S-rRNA and restriction with AluI, BanI, BbsI, DraII, Eco147I, and NheI generated a species-specific profile able to discriminate C. krusei from other Candida species.¹⁵⁵ The PCR-RFLP method amplifying the ITS1-5.8S-ITS2 rDNA region followed by restriction with MspI was reported as another alternative to identify this organism.¹⁵⁶

Alternatively, a PCR method using a primer pair that amplifies the polymorphic species-specific repetitive sequence C. krusei repeated sequence 1 (CKRS-1) of the non-transcribed intergenic regions of rRNA genes showed 100% specificity and a sensitivity to detect 10 to 100 fg of purified DNA.¹⁵⁷ Another alternative for C. krusei identification by PCR is the amplification of part of the gene encoding for topoisomerase II. A nested PCR reaction using a combination of degenerate and specific primers was reported to amplify a characteristic 227 bp DNA fragment from the C. krusei genome, with 100% specificity and a sensitivity of 40 fg of genomic DNA.¹⁵⁸

The real-time PCR has been also adapted for C. krusei identification. The amplification of the 5.8 rRNA gene demonstrated that was possible to obtain positive reactions with a detection limit of 10 CFU/mL blood, with 100% specificity.¹⁵⁹

More recently, it has been developed the CanTub-simplex PCR based on the amplification of the gene encoding for β-tubulin in a real-time platform, where the amplicons melting temperature is species-specific.¹⁶⁰ Following the same rationale, amplification of the internally transcribed spacer region 2 and analysis of melting peaks and curves, allowed the discrimination of C. krusei from the other 15 Candida species included in the study.¹⁶¹

The current molecular alternatives developed for C. krusei identification also include microarrays. Based on the nucleotide sequences of the internal transcribed spacer regions (ITS1 and ITS2) of the rRNA gene a chip to identify 32 fungal pathogens was recently reported.¹⁶² Even though the results of the microarray and the automated system Vitek 2 were concordant in 96.7% of cases for all the pathogens tested, for the case of C. krusei the specificity was of 100%.¹⁶²

The multianalyte profiling system has also been adapted for the identification of Candida species. This consists of mixed polystyrene beads covalently linked to specific DNA probes for C. albicans, C. glabrata, C. tropicalis, C. parapsilosis, C. krusei, and C. dubliniensis, which are incubated with amplicons containing the ITS2 region of Candida species rRNA gene previously amplified with universal fungal primers. The beads conjugated with the corresponding amplicons are analyzed in the multianalyte profiling system flow cytometer that measures the fluorescence produced by the different pairs of amplicons and beads.¹⁶³ This approach was 100% specific and showed a sensitivity limit of 0.5 pg of DNA.¹⁶³

The strategies for C. krusei identification also include immunological tests, although these are not as specific as the molecular methods. The Krusei color test (Fumouze) is a latex beads agglutination assay performed with red particles coated with a monoclonal antibody that specifically reacts with a C. krusei antigen found on the cell surface. Although all the C. krusei strains used in the study agglutinated the latex beads, false-positive reactions were observed with C. famata, C. glabrata, C. guilliermondii, C. kefyr, C. parapsilosis, and C. tropicalis.¹⁶⁴

Therapy

The treatment of Candida infections includes the use of several kinds of family compounds, named polyenes, azoles, echinocandins, nucleoside analogs, and allylamines. Fluconazole is one of the most common antifungal drugs used for empirical treatment of candidiasis; however, C. krusei is a species intrinsically resistant to this drug, with more than 95% of clinical and veterinary isolates being fluconazole-resistant.^141,165 The mechanisms behind this observation are not fully understood yet, but the flux pump activity of the ATP-binding cassette transporter Abc1 and reduced fluconazole affinity to Erg11 have been associated with this phenotypic trait.^165,166 It has been also proposed that both proteins could be part of the resistance mechanisms observed in some itraconazole-resistant strains.¹⁶⁵ The in vitro acquisition of resistance to voriconazole has been reported after exposing daily C. krusei to 1 µg/mL of the drug. In these cells, drug resistance was associated with increased expression of the ABC1 gene and point mutations within ERG11.¹⁶⁷

Several studies conducted with clinical samples have shown that most of the C. krusei strains are susceptible to voriconazole, itraconazole, posaconazole, anidulafungin, micafungin, 5-flucytosine, and amphotericin B; but intermediate resistance to caspofungin has been reported in some isolates.^168–170 Although C. krusei is a rare etiological agent of vaginitis, the use of local clotrimazole, ciclopirox olamine, terconazole, and boric acid is recommended.^171,172

In veterinary infections though, isolates resistant to fluorocytosine, itraconazole, ketoconazole, and amphotericin B have been reported.¹⁴¹ The pharmacological alternatives to treat candidiasis caused by C. krusei in animals include sulphamethoxypyridazine in cases of bovine mastitis.¹⁷³

There is a vast amount of studies addressing the search and design of compounds with antifungal activity, as well as the use of herbal derivatives with anti-Candida activity, with the potential of being explored as new alternatives to control candidiasis. Among the most relevant new alternatives are VT-1161 and VT-1129, a new generation of CYP51 inhibitors, a lanosterol 14-α-demethylase that belongs to the cytochrome P450 family and has a role in ergosterol biosynthesis, which showed the inhibition of C. krusei growth at concentrations of ≤ 2 μg/mL after 24 h of incubation.¹⁷⁴ Another alternative that is currently under investigation is the use of nanoparticles to deliver antifungal drugs into the fungal cells. It has been recently demonstrated that lipid core nanocapsules containing fluconazole were capable of reducing the effective dose of this antifungal drug and reverted the resistance to fluconazole observed in several C. krusei strains.¹⁷⁵ The use of palmatine, in combination with either fluconazole or itraconazole, has shown antifungal synergism, in a mechanism that inhibits the efflux pumps, with the consequent increment in the intracellular azole content.¹⁷⁶

A trypsin inhibitor from Tecoma stans (yellow elder) leaves has been isolated, characterized, and shown to have anti-Candida activity.¹⁷⁷ The minimal inhibitory concentration for this compound against C. krusei cells was 100 µg/mL, whereas the minimal fungicidal concentration was 200 µg/mL. ATP depletion and lipid peroxidation are thought to be the mechanisms behind its antifungal activity. In addition, it showed no cytotoxicity against human PBMCs,¹⁷⁷ being a promising candidate to move forward in the search for new antifungal compounds to treat candidiasis. Flavonoid and tannic fractions from Psidium guajava L. contain high levels of phenolic compounds and showed anti-C. krusei activity that synergizes with fluconazole and affects the morphological transition.¹⁷⁸ Similarly, flavonoids from Plinia cauliflora leaves, which mainly contain glycosylated quercetin and myricetin showed inhibitory activity against C. krusei (minimal inhibitory concentration of 19 µg/mL) and low cytotoxicity effect on human cells.¹⁷⁹ The water-insoluble fraction from Uncaria tomentosa (cat’s claw) showed a synergistic effect with either terbinafine or fluconazole in a mechanism that involves the action of the plant proanthocyanidins on the fungal cell wall.¹⁸⁰

These studies show that traditional medicine, mainly based on herbology, could be a source of a new generation of antifungal drugs.

Animals are also a source of molecules with antifungal activity. The 2-lysophosphatidylcholines isolated from deer antler extracts showed fungistatic activity, suppressing the morphological transition in C. albicans, C. krusei, C. guilliermondii, and C. parapsilosis, in a mechanism mediated through the mitogen-activated protein kinase pathway.¹⁸¹

Concluding Remarks

In recent years, there is a significant amount of information gathered about C. krusei biological and clinical aspects, underscoring the relevance of this organism as an emergent species, most likely because of its intrinsic resistance to fluconazole. The C. krusei genomic sequencing has opened new doors for basic research in this organism that could be translated into clinical applications. The genes prediction, along with their organization within the genome, the proteomic, transcriptomic, and metabolomic analyses could unveil species-specific genes related to virulence or drug resistance, information that could be later exploited in the diagnosis or treatment of the infection. The isolation of this organism from vegetal material and animal dropping points out to the environment as the source of candidiasis caused by C. krusei and provides the background to expand this kind of analysis to get a closer panorama of the C. krusei ecological distribution. The further expansion of our current knowledge on the C. krusei-host interaction would discover singularities in this species, which might be exploited for the design of alternative strategies to control the disease caused by this and other Candida species. This is of particular interest because thus far no vaccine is currently available to prevent candidiasis.¹⁸² However, new immunotherapeutic approaches, and the ultimate development of a vaccine against Candida species will rely on the deep knowledge of the immunity against these organisms.

Even though there are phenotypical and molecular strategies to identify C. krusei available in the clinical setting, faster, cheaper and more accurate alternatives are desirable for the early diagnosis of C. krusei and other Candida species. We provided some examples of natural products that could have antifungal activity, and these efforts should be replicated and look into the mechanisms behind the antifungal effect, as these compounds could be part of a new generation of drugs to treat candidiasis and other fungal infections. Finally, the search for the mechanisms behind the resistance to fluconazole in this species would provide useful information for the design of new treatment alternatives.