Adapting to survive: How Candida overcomes host-imposed constraints during human colonization

Successful human colonizers such as Candida pathogens have evolved distinct strategies to survive and proliferate within the human host. These include sophisticated mechanisms to evade immune surveillance and adapt to constantly changing host microenvironments where nutrient limitation, pH fluctuations, oxygen deprivation, changes in temperature, or exposure to oxidative, nitrosative, and cationic stresses may occur. Here, we review the current knowledge and recent findings highlighting the remarkable ability of medically important Candida species to overcome a broad range of host-imposed constraints and how this directly affects their physiology and pathogenicity. We also consider the impact of these adaptation mechanisms on immune recognition, biofilm formation, and antifungal drug resistance, as these pathogens often exploit specific host constraints to establish a successful infection. Recent studies of adaptive responses to physiological niches have improved our understanding of the mechanisms established by fungal pathogens to evade the immune system and colonize the host, which may facilitate the design of innovative diagnostic tests and therapeutic approaches for Candida infections.


Introduction
The human body is home to a large number of microbes that play essential roles in maintaining human health. However, under particular host-compromising conditions, they can shift from harmless commensals to opportunistic pathogens to cause inflammation and disease. Fungal communities, which can include Candida species, constitute an integral part of the human microbiota that, under normal conditions, asymptomatically colonize several niches, including the skin, oral cavity, gastrointestinal, and urogenital tracts [1][2][3]. The remarkable ability to alternate between local current microenvironments within internal host niches such as blood or tissues is often linked with their pathogenic potential. Therefore, environmental changes promoted either by alterations in host microbiota or the host immune system may resistance. We propose that the genetic circuits governing Candida adaptation to human niches can be exploited in search of new antifungal targets and diagnosis improvement.

Candida within the human host
The human host contains a variety of environmental niches in which Candida species can thrive. Adaptation to these sites requires rapid and coordinated changes in Candida metabolism and physiology in order to avoid or escape immune surveillance and to counteract several host-imposed constraints (for example, nutrient limitation, oxygen deprivation, pH fluctuations, changes in temperature, or oxidative, nitrosative, and cationic stresses). Moreover, Candida species interact with other microbial residents, establishing either cooperative or antagonistic relationships, which may affect their growth and influence the outcome of an infection.
Depending on the local environmental cues, some Candida species may exhibit different cellular morphologies. These include budding forms, which have been associated with commensalism, and the filamentous forms hyphae and pseudohyphae, often related with invasive and disseminated disease [14,15]. However, these cell types were found in infected tissues, suggesting they all promote pathogenicity. C. albicans has also the ability to switch into more functionally and genotypically distinct cell types, which may present improved fitness in specific host niches [15]. In particular, "white" yeast cells can switch to mating specialized "opaque" cells, and a subset of these can also transit into a third, "gray" morphology [16]. An additional distinctive group of cells, known as GUT (gastrointestinally induced transition), seems to display enhanced fitness in the gastrointestinal tract when compared with other cell types [17]. The morphogenic transitions depend on a highly dynamic cell wall that acts as an environmental barrier, and it is essential for host-pathogen interactions. The core skeleton of the cell wall is composed of the polysaccharide β-1,3-glucan, covalently linked to β-1,6-glucan and chitin. The outer layer contains glycosylated mannoproteins cross-linked to β-1,6-glucans. The relative amount of each component fluctuates between morphologies and in response to external challenges, impacting immune responses [18,19].

Nutrient availability and Candida metabolic flexibility
Of the many challenges pathogens face in the human host, possibly none is more important than nutrient availability because cells must assimilate nutrients in order to thrive. These might include sugars, carboxylic acids, peptides, amino acids, lipids, or phospholipids. The assimilation of glucose, lactose, and galactose is mediated via hexose transporters (HGTs), providing major sources of energy and carbon (Fig 2a). The well-studied yeast model Saccharomyces cerevisiae, which is relatively closely related to some Candida species, uses glucose as a preferred carbon source and only switches to nonfermentable nutrients when glucose becomes depleted [20]. This hierarchical utilization requires highly evolved networks integrating several signaling pathways in order to repress the assimilation of alternative carbon sources [21][22][23][24]. This is partly achieved by the ubiquitination of key gluconeogenic and glyoxylate cycle enzymes following the exposure to glucose [25]. Notably, these enzymes appear to lack ubiquitination sites in C. albicans, C. glabrata, C. parapsilosis, and C. tropicalis, and consequently, they are not subjected to glucose-induced degradation [26,27]. The evolutionary rewiring of key metabolic ubiquitination targets has been suggested to increase the ability of C. albicans to colonize and cause infection in the mammalian host because, unlike S. cerevisiae, this yeast is able to assimilate sugars and alternative carbon sources simultaneously [26][27][28]. The availability of glucose is thought to enhance C. albicans virulence owing to the fact that this sugar has been reported to induce hyphal morphogenesis at low physiological concentrations [29][30][31] and promote antifungal resistance [32,33]. Moreover, rapid glucose metabolism by C. albicans seems to be important during infection because immune cells, specifically macrophages, rely on glucose for survival [34]. This limitation is exploited by C. albicans, which elicits rapid macrophage death by depleting the available glucose [34].
In glucose-limiting conditions, other alternative carbon sources, such as N-acetylglucosamine (GlcNAc) and carboxylic acids, are thought to play a critical role to sustain Candida growth. When infecting tissues and organs, Candida up-regulates several pathways involved in the utilization of alternative carbon sources, such as gluconeogenesis, the glyoxylate cycle, and fatty acid β-oxidation, suggesting that glucose levels may not be sufficient to satisfy the energetic requirements of the cells [28,35-37]. In C. albicans and C. tropicalis, GlcNAc, a monosaccharide produced mainly by bacteria in the gastrointestinal tract, enters the cell through the Ngt1 transporter, and is then sensed by the transcription factors, Ngs1 and Rep1, which control the expression of genes involved in the uptake and catabolism of GlcNAc [38-40] (Fig 2b). Depending on the metabolic state of the cells, GlcNAc can either be converted to uridine diphosphate-N-acetylglucosamine (UDP-GlcNAc) or to fructose-6-phosphate, which then enters the glycolytic pathway (Fig 2b). In C. albicans, GlcNAc can also be used as a signal to induce the expression of several virulence genes involved in white-opaque switching [41], hyphal morphogenesis [38][39][40]42], and cell death [43]. Additionally, GlcNAc metabolism seems to sustain Candida survival when growing inside phagocytic cells. The export of intracellular ammonia, derived from GlcNAc catabolism, has been reported to promote the alkalization of the phagosome, enabling cells to survive and escape from the acidic environment of the phagolysosome [44]. This mechanism is dependent on the transport of GlcNAc and subsequent catabolism through Hxk1, Nag1, and Dac1 enzymes [44]. Hence, mutants lacking the Ngt1 transporter or GlcNAc catabolic enzymes are defective in neutralizing the phagosome [44]. The ability to manipulate ambient pH is reported for all species of the CTG clade, a phylogenetic group that translates the CUG codon into serine instead of leucine [45]. This is in contrast to what is found for the distantly related C. glabrata, whose genome does not appear to encode homologs of GlcNAc transporters or catabolic enzymes [44].
C. albicans can also raise the extracellular pH by metabolizing carboxylic acids [46]. This phenomenon is physiologically and genetically distinct from the GlcNAc-driven mechanism, as the metabolism of carboxylic acids, when used as the sole carbon source, does not generate ammonia or promote hyphal morphogenesis [44,46] (Fig 2c). These two transporter families are strongly induced after phagocytosis [51,52], and they modulate biofilm formation and resistance to antifungal drugs in both C. albicans and C. glabrata [54][55][56]. In particular, exposure to lactate has been shown to trigger the masking of β-glucan, a major pathogen-associated molecular pattern (PAMP), in several Candida species [57]. This affects the visibility of these pathogens to host immune defenses, which correlates well with the observed decrease in C. albicans uptake by macrophages and reduced phagocytic recruitment [57,58]. The β-glucan masking phenotype has been proposed to be dependent on Gpr1 and the transcription factor Crz1 [57]. These proteins control the expression of genes associated with the organization of the cell wall, ultimately contributing to the masking effect [57,59]. Therefore, the concomitant exposure of Candida cells to different carboxylic acids potentiates immune evasion and consequently Candida persistence.
The uptake of nitrogen is also critical for Candida survival. Different in vivo studies have demonstrated that genes involved in amino acid uptake and catabolism are strongly up-regulated in C. albicans, especially when phagocytosed by neutrophils and macrophages [36,60-62]. Indeed, several C. albicans and C. glabrata amino acid auxotrophic strains retain full virulence in mice, suggesting that these nutrients are readily available during infection [63-65]. Proteolytic enzymes, namely secretory aspartyl proteinases (SAPs), are of particular importance because they allow Candida to efficiently degrade the complement proteins and host connective tissues [66]. Once available, extracellular amino acids are then sensed by the SPS complex (composed of Ssy1, Ptr3, and Ssy5), which in turn activates the transcription factors, Stp1 and Stp2 (Fig 2d). While Stp1 controls the expression of extracellular proteases and peptide transporters, Stp2 regulates amino acid permeases, Ato transporters, and catabolic enzymes [67,68] (Fig 2d). Along with GlcNAc and carboxylic acids, the catabolism of amino acids represents a third independent mechanism by which Candida rapidly neutralizes acidic microenvironments [52,69]. Previous studies reported that C. albicans mutants lacking STP2 or ATO genes release less ammonia than wild-type controls, failing to efficiently neutralize the acidic phagosome and undergo hyphal morphogenesis, which consequently affects their ability to escape phagocytic cells [52,70]. Recent data, however, suggest that the phagosomal membrane is highly permeable to ammonia, and the observed alkalization is rather a direct consequence of proton leakage induced by hyphal growth [71,72]. The transport of methionine via the highaffinity permease Mup1 and its subsequent metabolism have been also shown to induce morphogenesis in a process that is dependent on Gpr1 and the cAMP-PKA (cyclic Adenosine Monophosphate-Protein Kinase A) signaling cascade [73,74]. The methionine-induced morphogenesis pathway triggers the activation of adenylate cyclase by the production of increased levels of polyamines such as spermine and spermidine. These compounds are generated by the intracellular conversion of methionine into S-adenosylmethionine (SAM) and its decarboxylation by Spe2, which donates aminopropyl groups for polyamine synthesis [73] (Fig 2d).

Environmental pH fluctuations shape Candida physiology and pathogenicity
Changes in ambient pH represent an additional stress that Candida and other pathogens face in the human host. While the pH of human blood and tissues is slightly alkaline (pH 7.4), the pH of the oral cavity and the gastrointestinal and genitourinary tracts is acidic (2 < pH < 6). Adaptation to differing ambient pHs is critical for survival and growth in these niches. In fungi, including Candida species, pH signaling is mediated by the Rim pathway [75]. In C. albicans, the external pH is sensed by Rim21/Dfg16, Rim9, and an arrestin-like protein Rim8. Under alkaline pH, Rim8 is hyperphosphorylated, a signal that triggers the endocytosis of the plasma membrane complex and the recruitment of the signaling protease Rim13. This protease then cleaves the C-terminal inhibitory domain of Rim101, resulting in its activation. The activation of Rim101 promotes the expression of target genes involved in morphogenesis [ (Fig 3).
On the other hand, the adaptation of C. albicans to acidic environments drives cell-wall remodeling by enhancing the exposure of two key fungal PAMPs (chitin and β-glucan) at the cell surface [85]. While pH-dependent β-glucan exposure is regulated by a noncanonical signaling pathway, the remodeling of chitin is coordinated by several transcription factors, including Rim101, Bcr1, and Efg1 (Fig 3) [85,86]. The exposure of β-glucan at the cell surface hyperactivates the immune system largely through the recognition of the immunostimulatory β-glucan by Dectin-1, which enhances the recruitment of neutrophils and macrophages to the site of the infection [85]. This pH-dependent β-glucan exposure was also observed in C. dubliniensis and C. tropicalis, but not in C. auris or C. glabrata [85,86]. Surprisingly, adaptation to acidic environments induces β-glucan masking in C. krusei, suggesting that the outputs of pHdependent signal transduction differ between these Candida species [85]. Additionally, the pH-dependent reorganization of the cell wall fluctuates over time in C. albicans, with β-glucan and chitin being masked after an initial period of exposure [86]. While the subsequent β-glucan masking is mediated by farnesol, this quorum-sensing molecule does not trigger the chitin cloaking [86]. These temporal fluctuations suggest dynamic cell-wall responses to environmental pH. Moreover, the early PAMP exposure appears to govern the outcome of the infection because subsequent remasking on the cell wall does not compensate for the initial induction of strong proinflammatory responses [86].

Adaptation to oxygen-limiting niches is critical for Candida virulence
Oxygen levels inside the human host can vary greatly. While some niches are rich in oxygen, such as exposed skin or oral mucosa, others are anoxic or hypoxic, including the Besides affecting the cellular metabolism and energy homeostasis, adaptation to hypoxia induces hyphal growth in C. albicans [94] and promotes immune evasion by triggering β-glucan masking at the cell surface [97]. β-glucan masking leads to reduced phagocytosis and attenuates local immune responses [97]. In contrast to lactate-induced β-glucan masking, hypoxia-induced masking does not depend on Gpr1 and Crz1. Instead, hypoxia-induced masking is mediated by mitochondrial and cAMP-PKA signaling [57,97]. Hypoxia induces the generation of mitochondrial superoxide [98,99], which is rapidly converted into diffusible hydrogen peroxide by superoxide dismutase 1 (Fig 4). Hydrogen peroxide has been proposed to somehow activate the cAMP-PKA pathway, which, in turn, triggers cell-wall remodeling and β-glucan masking [97]. However, the mechanism by which β-glucan masking is achieved at the cell surface remains unclear.

PLOS PATHOGENS
Hypoxia-induced β-glucan masking has been observed for some other pathogenic Candida species, namely C. tropicalis and C. krusei, but not in C. glabrata, C. guilliermondi, or C. parapsilosis [97]. Therefore, during their evolution, hypoxic signaling has become integrated with PAMP masking only in some Candida pathogens. The adaptation to hypoxic environments enhances the ability of these Candida species to colonize the host. For example, it was shown that the recruitment of polymorphonuclear leukocytes (PMNs) to sites of C. albicans infection in mice was the main cause of hypoxia [88] (Fig 4). However, because of the hypoxia-induced β-glucan masking by C. albicans cells, these PMNs are not able to efficiently phagocytose the fungus, generate reactive oxygen species (ROS), or release extracellular DNA traps, allowing C. albicans to survive. Continued exposure to hypoxia leads to accumulation of lactate, prolonging the masking effect. Additionally, it was also observed that the antifungal activity of fluconazole is considerably reduced under hypoxic conditions. We speculate that the molecular mechanism behind this observation might include Upc2, considering its dual role in activating hypoxia-induced β-glucan masking [97] and conferring azole antifungal resistance [100]. In contrast to C. albicans, C. tropicalis is not able to induce β-glucan masking in response to hypoxia, and this species is more susceptible to PMN attack [88]. This is in agreement with the fact that C. tropicalis mainly infects neutropenic patients [101]. The molecular mechanisms allowing hypoxic adaptation are not completely defined. Nevertheless, it is clear that some Candida species take advantage of low-oxygen environments, either generated during infection or imposed by the specific host niche, to thrive by avoiding immune surveillance and escaping from antifungal therapy.

Candida adaptation to temperature shifts is essential for full virulence
The human body temperature is considered to be a potent nonspecific defense against fungal infection, especially in febrile patients, because high temperatures considerably restrict fungal growth [102,103]. The human host presents fever as one of the first responses against a Candida infection, thereby exposing the fungal cells to temperatures ranging from 37˚C to 42˚C. These temperature fluctuations profoundly influence many physiological aspects of C. albicans, including morphology, mating, phenotypic switching, and drug resistance [104].
Changes in ambient temperature are sensed by a broad diversity of mechanisms. One of the most studied pathways is the evolutionarily conserved heat shock response, which mediates thermal homeostasis by controlling the levels of heat shock proteins (HSPs) [105]. HSPs are molecular chaperones sequestered in response to heat shock, rescuing proteins from unfolding or targeting damaged proteins for degradation. In C. albicans, the expression of HSP genes is activated by the heat shock transcription factor 1 (Hsf1), which becomes phosphorylated in response to temperature elevations, including thermal transitions that mimic fever [106,107]. After adaptation to the exposed temperature, Hsf1 phosphorylation returns to basal levels and several lines of evidence have suggested the existence of a negative feedback loop, in which Hsp90 negatively regulates Hsf1 [107-109]. Besides Hsf1, Hsp90 also controls the activation of other regulators that mediate long-term thermal adaptation (Fig 5). These include several mitogen-activated protein kinase (MAPK) signaling pathways, particularly the Hog1, Mkc1, and Cek1 pathways, which are intimately associated with cell-wall remodeling [110,111].
Other small HSPs such as Hsp12 and Hsp21 have also been identified as crucial for C. albicans to resist thermal stress [112,113]. HSPs and their associated signaling pathways have been widely implicated in antifungal resistance, emerging as potential antifungal targets to treat Candida infections [114]. Moreover, the activation of the Hsf1 transcriptional program in C. albicans has been associated with increased host cell adhesion, damage, and virulence, reinforcing the importance of this regulon in thermal homeostasis [115,116].

Candida and host microbiota: Avoiding antagonistic interactions in health and disease
The structure of human microbiota is dynamic, often defined by host and environmental factors and also by physical and metabolic interactions between species. While some of these interactions are cooperative, others are antagonistic, and the latter may represent a major obstacle for Candida. This concept gained experimental support through studies involving the depletion of commensal microbiota by continued use of broad-spectrum antibiotics, which resulted in Candida overgrowth [117,118]. This suggests that some commensal microbial colonizers antagonize Candida spp. (and other exogenous pathogens) in order to maintain a homeostatic balance in the host. Some of these interactions are driven by metabolic competition, while others are mediated by quorum-sensing molecules that influence fungal cell behavior and regulate important virulence traits. Although quorum-sensing systems have been explored in great detail for pathogenic bacteria, they are relatively poorly understood in fungi [119]. The C. albicans molecule farnesol was the first quorum-sensing compound to be identified in an eukaryote [120] and has been the object of intense research. Yet, its precise mode of action remains unclear.
Lactobacillus species and C. albicans are a well-documented example of infectious antagonism [121][122][123]. Lactobacilli are a dominant species of the microbiota of the gastrointestinal and urogenital tracts, and they actively reduce the amount of fungal microbes by producing many fungicidal compounds [121][122][123]. Other commensal bacteria such as Bacteroides thetaiotamicron or Blautia producta can antagonize C. albicans by stimulating intestinal cells to produce antimicrobial peptides [124]. The pathogenic bacterium Acinetobacter baumanii has been also reported to interact antagonistically with C. albicans by binding to hyphae to promote apoptosis [125]. The elucidation of these types of interaction is of particular interest in the quest for novel targets for antifungal therapy, as the inhibitory secreted factors produced by these antagonists appear to have high fungicidal activity.
The disruption of commensal interactions through alterations in immune competence, by changes in environmental host conditions, or via antibiotic therapy may favor the outgrowth and overrepresentation of pathogenic microbes, with these growing at the expense of those organisms that fail to adapt. While antagonist interactions might lower the risk of infection, synergistic interactions during dysbiotic states are associated with increased pathogenesis because microbes can also interact to enhance colonization and persistence. An illustrative example is the infectious synergism established between several Candida species (including C. albicans, C. dubliniensis, C. tropicalis, and C. krusei) and the gram-positive bacterium Staphylococcus aureus [126,127]. Candida not only provides a substratum for the attachment and colonization of S. aureus but also facilitates its invasion across mucosal barriers, thereby promoting persistence and systemic infection [128].

Host immune defenses: How Candida species counteract the immune response
Microbial pathogens are constantly surveyed by the innate immune system. Phagocytic cells such as dendritic cells, macrophages, monocytes, and neutrophils play important roles in

PLOS PATHOGENS
clearing fungal pathogens from the bloodstream and tissues. Loss of innate immune cells or defects in their antifungal activities have major implications for the host. Candida cells are recognized through key PAMPs, some of which are located in the cell wall; for example, β-glucans, chitin, and mannans. These components are sensed by the multiple pattern-recognition receptors (PRRs) expressed by phagocytic cells or secreted (for example, complement components). PPRs mediate binding of the pathogen to the phagocyte, and the PAMP-PRR interactions trigger intracellular signaling pathways within the immune cells that can induce phagocytosis and the production of proinflammatory cytokines and chemokines. In order to attenuate recognition and escape phagocytosis, Candida cells are able to actively mask cell-wall PAMPs [129] and secrete specific proteases that target complement opsonization [130]. Alternatively, some Candida species can induce their phagocytic uptake into endothelial and epithelial cells and use these cells as "safe houses" by preventing maturation of the phagolysosome and subsequent killing [131]. If none of these strategies is employed, Candida cells are likely to be internalized and subjected to a combination of toxic oxidative and nonoxidative mechanisms that attempt to kill an intra-or extracellular yeast cell. These oxidative mechanisms include the production of reactive oxygen and nitrogen species (ROS and RNS, respectively), while nonoxidative killing mechanisms include the release of antimicrobial peptides and the induction of processes related to micronutrient restriction. Of note, while C. albicans is sensitive to the combinatorial stresses imposed by phagocytes [132], C. glabrata has adapted to survive within the inhospitable environment of the phagosome. This pathogen mounts robust stress responses against the ROS implemented by the phagocytic cell and neutralizes the phagocytic environment, thereby escaping phagocytosis [133].

Oxidative, nitrosative, and osmotic/cationic stresses
Phagocytic cells attempt to kill pathogens in part by employing toxic ROS and RNS, either intracellularly or extracellularly, as a major antimicrobial defense mechanism. ROS are produced by the NADPH oxidase complex, a process known as respiratory burst, and include chemicals such as the superoxide anion (O 2 • ), hydrogen peroxide (H 2 O 2 ), and the hydroxyl radicle ( • OH). Furthermore, ROS production in response to C. albicans infection has been shown to lead to the recruitment of additional phagocytes, creating a toxic oxidative environment for the fungus [134]. Inside phagocytes, ROS can interact with nitric oxide (NO), generating toxic products such as peroxynitrite [135]. These toxic chemicals cause irreversible damage to the pathogen by interacting with proteins, lipids, and nucleic acids.
Candida species attempt to counteract these stresses by activating cellular responses that include the activation of genes encoding proteins involved in stress detoxification and reparation. These include catalase, superoxide dismutases, glutathione peroxidases, and thioredoxins (Fig 6a) [136-138]. In C. albicans and C. glabrata, these stress pathways are regulated largely by the Hog1 stress-activated protein kinase [136,139], the transcription factor Cap1 [140][141][142], and the Rad53 DNA damage checkpoint kinase [143]. Together with the transcription factor Cta4, these signaling pathways play key roles in orchestrating the responses to osmotic, oxidative, and nitrosative stresses in these species [144]. In this way, these regulators promote the fitness of C. albicans during systemic infection. Indeed, mutants lacking these genes display attenuated virulence in mice, as well as impaired tolerance to these stresses in vitro and phagocytic survival [145,146]. Curiously, the oxidative stress response is delayed if the fungus is simultaneously exposed to cationic and oxidative stress [147]. This is thought to contribute to the ability of phagocytic cells to efficiently kill invading pathogens (Fig 6a) [132]. Given the importance of these stress response pathways for Candida survival, key molecular players involved may represent attractive targets for antifungal development. Host-enforced micronutrient restriction results in reduced iron, copper, and zinc availability, but C. albicans responds by upregulating efficient metal-scavenging strategies. Host phagocytes also exploit the toxicity of copper and zinc by pumping these metals in excess into phagosomes to intoxicate internalized pathogens. NOS, nitric oxide species; PM, plasma membrane; ROS, reactive oxygen species; Sod, superoxide dismutase; XS, oxidative stress. https://doi.org/10.1371/journal.ppat.1008478.g006

Host-enforced micronutrient restriction
The limitation of micronutrients such as iron, copper, zinc, or manganese is an effective way of controlling the outgrowth of invading microbes. These micronutrients are essential for the survival of both host and pathogen because they function as cofactors for enzymes, transcription factors, and other proteins that play crucial biochemical and cellular functions. However, our immune system attempts to restrict microbial access to these essential elements via a mechanism known as nutritional immunity [148].
Iron has well-studied implications for Candida pathogenesis, being a crucial micronutrient for Candida growth, survival, and virulence [149]. During systemic candidiasis, the host restricts this metal by increasing the levels of iron-binding proteins, such as ferritin and hemoglobin alpha, and accumulating heme oxygenase (Fig 6b) [150,151]. Both C. albicans and C. glabrata have developed efficient iron-scavenging strategies that can overcome these host mechanisms. This contributes to their ability to survive phagocytosis and replicate inside macrophages by using their intracellular storages of iron [152,153]. C. albicans and C. glabrata cells exploit sophisticated iron-uptake systems to acquire either free iron [154,155] or iron from host iron-binding proteins, including hemoglobin [156], ferritin [82], and transferrin (Fig 6b). Additionally, the utilization of siderophores promotes resistance to macrophage killing: in C. glabrata, the Sit1 siderophore-iron transporter mediates iron acquisition, being critical for the survival of the yeast inside macrophages [152].
Copper is also involved in Candida virulence, both positively and negatively. The fungal reductive iron-uptake pathway includes multicopper oxidases, and hence, iron acquisition and mobilization depends on copper availability [157]. Interestingly, the host also uses copper as a defense mechanism against Candida by pumping excess quantities of this metal into Candidacontaining phagosomes (Fig 6b) [158]. However, C. albicans adapts to this potential killing mechanism by differentially modulating the expression of copper-and manganese-dependent SODs (Sod1 and Sod3, respectively) [159]. Sod1 is expressed when copper is in excess, but when copper levels decline, Sod3 is then expressed (Fig 6b) [159]. Thus, during infection, C. albicans is able to adjust copper uptake and management by using it as an enzymatic cofactor for SOD enzymes [159].
Zinc is an abundant micronutrient that has crucial roles in cellular functionality for both host and pathogen. The host attempts to limit zinc availability for the fungus by depleting extracellular zinc levels, mainly via calprotectin, an antimicrobial peptide expressed by neutrophils that binds zinc and manganese with high affinity (Fig 6b) [160]. Calprotectin promotes the antimicrobial activity of neutrophil extracellular traps (NETs), which are released by neutrophils after sensing large microbes such as C. albicans hyphae [161][162][163]. Zinc depletion also occurs inside immune cells as an antifungal mechanism to kill intracellular pathogens such as C. albicans and C. glabrata [164]. During infection, macrophages deplete intracellular zinc by pumping it into the Golgi apparatus via specific ZnT-type zinc transporters (Fig 6b) and increasing the expression of zinc-binding metallothioneins [165]. Additionally, macrophages up-regulate the zinc importer ZIP2 to increase the intracellular levels of zinc (Fig 6b) [166]. This combination of strategies depletes zinc from the extracellular environment while dealing with the increased metabolic demands associated with microbial clearing [166]. To overcome zinc depletion, C. albicans overexpresses ZRT1 and ZRT2 genes, encoding zinc uptake transporter systems Zrt1 and Zrt2 (Fig 6b). Both transporters are regulated by the zinc finger transcription factor Zap1 (also known as Csr1) [167,168] and by pH [79]. Zinc transporters play important roles in Candida pathogenesis because overexpression of Zrt2 increases C. albicans virulence [169]. In addition to functioning as a zinc transporter, Zrt1 also serves as a receptor for the Pra1 zincophore [79,168], a secreted protein that binds and sequesters zinc from host cells during C. albicans invasion (Fig 6b) [170]. Similarly to copper, zinc has also been reported to be pumped in higher amounts into the phagosome to intoxicate internalized pathogens, constituting an important mechanism of killing (Fig 6b) [171].

Environment-triggered biofilm formation and antifungal resistance
So far, we have described major molecular circuits required by Candida species to counteract several constraints they face in the human host. The ability of Candida to adapt to these stresses imparts the flexibility to colonize diverse host niches. The physiological capacity to respond efficiently to stress and survive hostile environments also endows the fungal cells with the advantage of being better prepared for future insults [172,173]. The generation of biofilms might represent another strategy to resist harsh conditions and persist in the human host.
The Candida species most frequently associated with the formation of biofilms, either in host tissues or implanted medical devices, are C. albicans, C. glabrata, C. tropicalis, and C. parapsilosis [174]. Biofilms represent three-dimensional communities of adherent cells, with distinct biological properties, that are embedded in a self-synthesizing extracellular matrix (ECM) composed predominantly of proteins, glycoproteins, carbohydrates, lipids, and nucleic acids [175]. The ECM helps to maintain the overall structural integrity of the biofilm, and it also acts as a physical barrier to drug penetration. Consequently, biofilm cells can survive drug concentrations more than a thousand times higher than those defined for planktonic cells [176]. This phenotype is partly associated with the sequestration of drugs by the biofilm ECM and partly with the occurrence of a subpopulation of so-called "persister cells". Persister cells exhibit a dormant-like physiology that has been demonstrated to make them highly resistant to antifungals [177]. These features contribute to the intrinsic resistance of Candida biofilms to conventional antifungal treatments, the host immune system, and other environmental perturbations, making biofilm-based infections a clinical challenge.
Genome-wide transcriptional profiling and proteomic approaches have identified hundreds of genes that are differentially expressed between C. albicans biofilm and planktonic cells. The up-regulation of glycolytic and sulfur amino acid genes, similar to what is observed when cells grow under hypoxia, suggests that Candida biofilms constitute a heterogeneous environment with hypoxic niches [178]. Moreover, more than 50 transcriptional regulators and 101 other genes have functionally validated roles in the formation of Candida biofilms [179][180][181]. Some of these play important roles in hyphal formation, adhesion, drug resistance, and matrix production (all intrinsic characteristics of biofilms), as well as in stress adaptation. It is not surprising, then, that adaptation to specific environmental niches modulates the ability of cells to form biofilms and, consequently, to resist antifungals [54,55,58,59,[182][183][184].

Final remarks and future perspectives
Candida cells regulate specific sets of genes, including many involved in an array of stresses and metabolic pathways, in order to thrive and persist in the human host. In addition to conferring metabolic flexibility and stress resistance, the physiological reprogramming has been associated with enhanced virulence through impaired immune recognition, increased biofilm formation, and/or acquired antifungal tolerance and resistance. Although remarkable progress has been made in the last few decades in our understanding of the impact of host-derived stresses on Candida physiology and pathogenicity, many details remain unclear. During an infection, Candida cells are exposed to multiple environmental constraints, sometimes imposed consecutively, and at other times imposed simultaneously. Yet, in vitro experiments are predominantly designed to study individual environmental signals, often at single time points, rather than combinatorial stresses over time. Much progress has been achieved using the first approach. While this has given us valuable insights, it rather oversimplifies biological reality. The analysis of combinatorial stresses and of the dynamism of these inputs would mimic host conditions more closely and reveal more detailed views of which stress or stresses prevail and dictate the outcome of different types of infection. The same principle applies to infection and biofilm models, in which usually interactions between only a few different microbial populations have generally been examined. Most knowledge in the field comes from studies of either C. albicans or C. glabrata. Yet, the regulatory circuits required to effectively respond to each constraint, including antifungal treatments, differ considerably between the different Candida species, illustrating how heterogeneous these pathogens are. With the unprecedented emergence of multidrug resistant species such as C. auris, there is an urgent need to develop new effective antifungals. The integration of omics data with in vivo models, which mimic host conditions more closely, is now a powerful strategy to unravel molecular processes underlying adaptive phenotypes. These platforms have already produced novel lines of research and improved the identification of new potential therapeutic targets for vaccine and antifungal drug development, enhancing our ability to develop novel strategies to fight Candida infections.