Glucose acquisition and carbon metabolism are intrinsically related to capsule polysaccharide synthesis. Increased concentration of glucose in growth media results in increased capsule production and secretion, as does the substitution of glucose in media with mannitol, highlighting the dependence of capsule production on carbon acquisition and metabolism (Cleare and Casadevall, 1999; Guimaraes et al., 2010). In addition to the enzymes involved in synthesis and modification of these sugar donors (Griffith et al., 2004; Moyrand and Janbon, 2004), the transferases that incorporate these sugars into the polysaccharide chains of capsule are required for appropriate capsule structure and subsequent biological functions (Wills et al., 2001; Cottrell et al., 2007).

Cryptococcal capsule is composed of polysaccharide chains organized into fibers which are attached to and extending from the fungal cell wall (Cleare and Casadevall, 1999; Maxson et al., 2007b; Frases et al., 2009). Polysaccharide is secreted from the cell and is incorporated at the capsule edge (Zaragoza et al., 2006). These polysaccharide chains decrease in concentration moving outward from the cell wall, creating a somewhat permeable outer layer and a dense inner layer that prevents cell wall recognition by antibodies and complement (Gates et al., 2004; Bryan et al., 2005; Charlier et al., 2005; Maxson et al., 2007a; Maxson et al., 2007b; Frases et al., 2009; Pontes and Frases, 2015). This architectural organization confers a hydrophilic quality to the capsule which provides the ability to form a protective hydrogel or biofilm. It also contributes to the overall negative charge of the cell, which is believed to suppress phagocytosis (Nosanchuk and Casadevall, 1997; Martinez and Casadevall, 2015; Aslanyan et al., 2017; Vij et al., 2020).

The two primary polysaccharides making up this barrier are glucuronoxylomannan (GXM) and glucuronoxylomannogalactan (GXMGal), with GXM being the predominant component accounting for ~90% of polysaccharide composition (Cherniak and Sundstrom, 1994; Cherniak et al., 1998b; Mcfadden et al., 2006; Zaragoza et al., 2009). Synthesis of capsule polysaccharide involves the polymerization of simple sugars into an elongated carbohydrate chain within the cell prior to transport across the cell wall. One of the essential components of GXM is UDP-glucuronic acid, converted from UDP-glucose by UDP-glucose dehydrogenase. Mutants of UDG1 (UDP-glucose dehydrogenase) lack any detectable capsule, as well as the ability to grow at 37°C, the normal temperature at which C. neoformans thrives (Griffith et al., 2004; Moyrand and Janbon, 2004). GXM is composed of an α(1,3) mannose backbone, while GXMGal is composed of an α(1,6) galactan backbone – both modified through the addition of side groups including xylose, mannose and glucuronic acid residues that form six different conformations designated as motifs 1-6 (M1-M6) (Cherniak et al., 1998b; Kozel et al., 2003; Heiss et al., 2009). These modifications alter the secondary structure of the polysaccharides, reduce molecular flexibility and create heterogeneity among different strains as well as within a population of a single strain (Todaro-Luck et al., 1989; Belay et al., 1997; Mcfadden et al., 2007).

Both types of host antigen presenting cells (e.g. macrophages and dendritic cells) express pattern-recognition receptors which interact with pathogen-associated molecular-patterns (PAMPs) allowing for the generation of an immune response (Takeuchi and Akira, 2010). The toll-like receptor (TLR) family of PAMPs are membrane-bound and their binding triggers pro-inflammatory responses from within the cell including the expression of TNF-α and IFN-γ (O’mahony et al., 2008; Campuzano and Wormley, 2018). Cell-surface TLR2 and TLR4 recognize capsular polysaccharides GXM and GXMGal, while endosomal TLR9 recognizes fungal genomic DNA, specifically unmethylated CpG motifs (Nakamura et al., 2008). Activation of TLR9 results in recruitment of phagosomes and is required for elimination of the fungal pathogen by the adaptive immune system, however TLR2 and TLR4 are not involved in host defense (Nakamura et al., 2008; Ramirez-Ortiz et al., 2008; Miyazato et al., 2009; Zhang et al., 2010; Biondo et al., 2011; Qiu et al., 2012). C-type lectin receptors (CLRs) are also important for fungal recognition. These receptors also bind polysaccharides, but more specifically they can bind α-mannans (Dectin-2), mannose (CD-206/mannose receptor), mannosylated mannoproteins (DC-SIGN), and chitin (CD-206/MR) (Campuzano and Wormley, 2018). Interestingly, CD206 is a pattern recognition receptor that recognizes terminal mannose residues such as those on C. neoformans, and its expression is also upregulated in alternatively activated (M2) macrophages. Binding of C. neoformans to CD206 by these cells results in phagocytosis and subsequent intracellular growth of the yeast. The critical adaptor molecule CARD-9 responds to CLR activation and is required for M1 activation of macrophages (Campuzano et al., 2020). The resulting signaling cascade leads to M1 activation, dendritic cell maturation, and increased ROS and cytokine production ( Figure 1 ).

The addition of xylose side groups among the remaining mannose residues varies between serotypes (Cherniak et al., 1980; Merrifield and Stephen, 1980). Side chains of glucuronic acid, along with the xylose residues, create two hydrophilic fringes that hide the mannan backbone from opsonization. Antigenic differences due to these substitutions and their locations have allowed for the categorization of Cryptococcus variations by serotype: A, B, C, D, and hybrids AD and BD (Cherniak et al., 1980; Ikeda et al., 1985), however molecular techniques have determined that serotyping alone is insufficient for determining cryptococcal species (Hagen et al., 2015). For example, the structure of C. neoformans serotype D has gaps between the additions of sugar groups that allow for exposure of mannan for immune recognition. In contrast, serotype A contains sugar groups arranged in a continuous band, creating an even more occluded surface against antibody binding (Kuttel et al., 2020). Urai and colleagues showed that infection with cryptococcal strains containing additional xylose, such as C. gattii (e.g. strain JP02) result in significantly lower immune recognition and stimulation compared to those with fewer xylose additions such as C. neoformans (e.g. strain H99) (Urai et al., 2015). The arrangement of acetyl groups along the mannan backbone further reduces polysaccharide flexibility and provides a structural conformation that is preferentially bound by anti-capsular antibodies (Kuttel et al., 2020). An immunostimulatory comparison of capsular polysaccharides showed O-acetylation of GXM is associated with increased recognition by antigen presenting cells as well as increased secretion of pro-inflammatory cytokines (Urai et al., 2015). Due to the lack of recognition, deacetylated mutants are more capable of evading phagocytosis resulting in hypervirulence compared to wild-type (Janbon et al., 2001). O-acetylation of capsule polysaccharides also plays an important role in virulence as it impairs chemotactic recruitment of neutrophils and their endothelial adhesion (Ellerbroek et al., 2004).

The exact mechanism of capsular growth has not been elucidated, but clearly occurs as a result of specific environmental stimuli (Zaragoza and Casadevall, 2004). Many of the capsule-inducing stressors involve nutrient deprivation, which is interesting considering the large metabolic requirements of polysaccharide accumulation (Trevijano-Contador et al., 2017). Contact with phospholipids, such as those found on macrophages or amoebae, is a trigger for capsular growth, along with other environmental stimuli that signal a threat to Cryptococcus, including elevated pH, CO₂, and serum (Granger et al., 1985; Chrisman et al., 2011). Inositol catabolism has recently been shown to be essential to capsule growth and structure. Readily available inositol as a carbon source increases the size of the capsule, leading to greater virulence. Of clinical significance, inositol is abundant in the brain, allowing the yeasts to colonize this host organ (Wang et al., 2021) ( Figure 2 ).

GXM has the ability to self-aggregate through divalent metal cations (such as calcium) which leads to the enlargement of capsule in a concentration dependent manner (Mcfadden et al., 2006; Nimrichter et al., 2007; Frases et al., 2008). The exopolysaccharides added during growth are different in size and mass compared to those in the steady-state capsule (Yoneda and Doering, 2008; Frases et al., 2009). In situations where capsule size must be reduced in adaptation to environmental changes, capsule does not appear to degrade, and no capsule degrading enzymes have been identified in C. neoformans. Instead, cells invest energy in producing buds with smaller capsules by modifying cell wall attachments of capsule during bud formation rather than degrade capsular polysaccharides (Zaragoza et al., 2009).

As the first structure to interact with the immune system, the capsule must not only be able to mitigate the immune response through structural changes, but also create a protective barrier that allows for growth and dissemination. The capsule is able to dramatically increase its size quickly after introduction of C. neoformans into the lungs (Feldmesser et al., 2001). Murine capsule studies have shown both size and compositional differences in capsules produced in different organs that can be attributed to the specific environments encountered in different tissues (Casadevall et al., 1992; Rivera et al., 1998; Garcia-Hermoso et al., 2004; Charlier et al., 2005). Capsular heterogeneity, even within a single population, can result in changes in immunoreactivity to specific monoclonal antibodies (Kozel et al., 2003; Garcia-Hermoso et al., 2004; Mcfadden et al., 2007). Capsule inhibits phagocytosis by monocytes in the absence of opsonization, preventing internalization of the yeast necessary for antigen presentation and subsequent T-cell activation (Vecchiarelli et al., 1994a; Vecchiarelli et al., 1994b; Syme et al., 1999; Vecchiarelli et al., 2003). Phagocytic inhibition may be due to the glucuronic acid residues in capsule imparting a negative charge across the cell surface, potentially causing an electrostatic repulsion from the host immune cells (Nosanchuk and Casadevall, 1997). Capsule structure may alter antibody binding in terms of affinity and specificity, as well as localizing antibodies deep within the capsule, reducing opsonization effectiveness. Additionally, C. neoformans also has been shown to metabolize opsonins through degradation by secreted proteases (Chen et al., 1996). Once opsonized and phagocytosed, capsule protects the yeast from oxidative damage (Zaragoza et al., 2008) while also interfering with the maintenance of pH within the phagolysosome (De Leon-Rodriguez et al., 2018). The phagolysosomal membrane may be maintained, releasing C. neoformans via non-lytic phagosomal extrusion, or it may rupture, releasing yeasts into the cytosol where they can replicate undetected (Alvarez and Casadevall, 2006; Ma et al., 2006; Voelz and May, 2010).

Unlike macrophages, dendritic cells do not require activation for C. neoformans killing and can quickly kill yeasts within the phagolysosome through oxidative and non-oxidative mechanisms (Wozniak et al., 2006). Additionally, they are significantly more efficient in antigen presentation to T-cells (Syme et al., 2002). Cryptococcal antigens (e.g. mannoproteins) induce IL-12 and TNF-α secretion in dendritic cells, promoting a protective T_H1 response (Pietrella et al., 2005). However, the extracellular environment during cryptococcal infection naturally induces a non-protective T_H2 response, including increasing the accumulation of immature dendritic cells in proximal lymph nodes (Osterholzer et al., 2009).

Encapsulated yeasts as well as purified capsular polysaccharides have been shown to inhibit the expression of host cytokines and chemokines (Lupo et al., 2008). Extracellular C. neoformans alters macrophage NF-κB protein synthesis resulting in repression of macrophage cell proliferation and pathogen-induced apoptosis (Ben-Abdallah et al., 2012). Purified capsule polysaccharides can trigger apoptosis of both macrophages and T-cells, which reduces the secretion of pro-inflammatory mediators from these cells (Pericolini et al., 2006; Chiapello et al., 2008; Monari et al., 2008; Villena et al., 2008). Furthermore, GXM has the ability to directly inhibit T-cell proliferation and responses (Yauch et al., 2006). Intracellular C. neoformans is able to modify the polarization of macrophages by altering their cytokine profiles. As the yeasts reside in the macrophages, they work to maintain a naïve-type state of the macrophages. iNOS expression is also upregulated by the intracellular yeasts (Subramani et al., 2020). Therefore, Cryptococcus has the insidious ability to hide within the host by manipulating the cytokine environment of the host.

Host surfactant proteins A and D (SP-A, SP-D) are an important part of the innate immune response; they bind to microbial surfaces and modulate leukocyte functions (Crouch and Wright, 2001). SP-D can bind purified GXM with high affinity but has a lower affinity for whole cryptococcal capsule, which is interesting considering GXM is the primary component of capsule (Van De Wetering et al., 2004). In cryptococcal infection, SP-A does not play an important role in clearance of fungal infection, however SP-D facilitates aggregation of microorganisms, increasing mucociliary clearance as well as promoting phagocytosis and killing in phagocytic cells. However, binding of C. neoformans by SP-D can also inhibit this process, increasing fungal burden and promoting systemic dissemination (Van De Wetering et al., 2004). Interestingly, SP-D is found at higher levels in the lungs of patients with AIDS, possibly contributing to the progression of disease amongst this group (Jambo et al., 2007). GXM prevents SP-D mediated aggregation which would typically facilitate pathogen clearance. SP-D may also selectively bind secreted capsule over attached capsule thereby reducing the potential for opsonization of attached capsule and subsequent immune interaction with yeast cells (Van De Wetering et al., 2004). Paradoxically, fungi bound with SP-D are protected from macrophage destruction and have increased survival (Geunes-Boyer et al., 2009; Geunes-Boyer et al., 2012). In these instances, Cryptococcus provides an excellent example of a pathogen using host defenses to its advantage.

Changes in capsular structure that occur during infection are necessary for transmigration into the CNS and are visible via immunofluorescence as soon as six hours post-intravenous inoculation (Charlier et al., 2005). Counterintuitively, an intriguing study of capsular mutants showed that a C. neoformans hypercapsular mutant was less virulent and showed lower levels of brain dissemination than a hypocapsular mutant (Pool et al., 2013). It has been hypothesized that capsule size is indicative of starvation stress in Cryptococcus (Casadevall et al., 2019), therefore neurotropism may result from the nutritionally richer environment of the deep tissues as compared to the lungs. In contrast to this hypocapsular mutant data, acapsular strains are easily phagocytosed and destroyed by immature dendritic cells and result in the upregulation of antigen-presenting molecules, showing the necessity of capsule in virulence and survival of C. neoformans (Vecchiarelli et al., 2003).

Capsule attachment has been recently shown to be regulated by Rim101, a transcription factor primarily involved in pH response. GXM is able to be secreted in a rim101Δ mutant, but the capsule is much thinner, producing a hypocapsular strain. In a mouse inhalation model, the rim101Δ mutant was surprisingly hypervirulent, resulting in decreased mouse survival compared to wild-type and rim101Δ + RIM101 strains (O’meara et al., 2010). The rim101Δ strain was shown to produce a very different immune response than the wild-type strain. There was a much greater influx of inflammatory infiltrate in the lungs of the rim101Δ mutant, and additionally the cells composing the infiltrate differed from wild-type. The infiltrate consisted more heavily of neutrophils and eosinophils in the rim101Δ mutant-infected lungs, whereas in wild-type infection, the inflammatory cells mainly consist of lymphocytes and monocytes, with few neutrophils and eosinophils (O’meara et al., 2013). Cytokine levels and profiles differed as well: rim101Δ had higher cytokine levels than wild-type, with IL-12, IFN-γ-inducible protein 10 (IP-10), VEGF, and TNF-α all significantly increased with rim101Δ infection (O’meara et al., 2013).

Expression of mannoproteins MP98 and MP88 was also increased in the rim101Δ strain, suggesting that the altered capsule regulation correspondingly impacts the antigen profile of C. neoformans. Mannoproteins are structural components in the inner cell wall of C. neoformans and also play an antigenic role (Vartivarian et al., 1989). They are known to be the predominant antigens involved in activating cell-mediated immunity in cryptococcosis (Murphy, 1988). MP98 and MP88 are known to stimulate T-cell responses in the host reaction to Cryptococcus (Levitz et al., 2001; Huang et al., 2002), which may at least partially explain the hypervirulence phenotype in this mutant.

The cell wall of Cryptococcus is a two-layered structure that sits above the plasma membrane and serves to control cellular permeability in addition to mitigating mechanical and osmotic stresses (Baker et al., 2007; Ponton, 2008; Doering, 2009; O’meara and Alspaugh, 2012; Gow et al., 2017; Agustinho et al., 2018; Wang et al., 2018). In fungi, the cell wall is composed of glucans, chitin, chitosan, glycosylated proteins, and melanin (Casadevall and Perfect, 1998). Glucans are the most important structural polysaccharides of the cell wall and are synthesized in the plasma membrane by a series of enzymes called glucan synthases. The inner layer is mostly fibers of β-glucan and chitin arranged parallel to the plasma membrane, while the outer layer contains α-glucan and β-glucan (Sakaguchi et al., 1993; Doering, 2009; O’meara and Alspaugh, 2012). Alpha-1,3-glucan is required to anchor the capsule to the cell wall, and is required for virulence (Reese and Doering, 2003; Reese et al., 2007).

Chitin is synthesized by an enzyme called chitin synthase from N-acetylglucosamine (GlcNAc) and is deposited into the extracellular space next to the plasma membrane. In Cryptococcus species, chitin is a minor component of the cell wall but it does contribute strength and plays a crucial role in capsular architecture (Doering, 2009; Zaragoza et al., 2010). Chitin also has the ability to modulate host immune responses (Wiesner et al., 2015). The deacetylated form of chitin, chitosan, is more soluble and flexible than chitin, making up 3-5 times more of the cell wall (by dry weight) than chitin and is necessary for cryptococcal virulence (Upadhya et al., 2016). As the density of the cell wall changes, so does the ratio of chitosan to chitin (Banks et al., 2005). Chitosan is important for maintaining the integrity of the cell wall and normal capsule width; mutants deficient in chitosan display a “leaky” phenotype and are unable to retain melanin within the cell wall (Baker et al., 2007).

Changes to antigenic structures due to alterations in carbon metabolism have been elucidated in other pathogenic fungi. In Candida albicans, glucose is preferred carbon source, but alternative non-fermentable carbon sources such as lactate, acetate and citrate can be assimilated through gluconeogenesis, glyoxylate shunt, and β-oxidation. Interestingly, growth of Candida on alternative carbon sources also modifies the composition and architecture of the cell wall, resulting in altered host recognition and immune cell interactions (Ene et al., 2012; Ene et al., 2013). When grown in host-derived blood or serum, Candida exhibited a dramatic reduction in the length of mannan chains, as well as a reduction in cell wall structural complexity (Lowman et al., 2011). With lactate as a sole carbon source, biomass was significantly reduced and cell walls were thinner, with the β-glucan and chitin layer dramatically reduced. Although these components maintained the same proportions, these cells displayed alterations to cell wall porosity and hydrophobicity (Ene et al., 2012), both of which could potentially affect pathogenicity. The observed changes result in decreased host recognition and phagocytosis, while those yeasts that are engulfed show increased macrophage killing and escape (Brown et al., 2014). This phenotype results in increased virulence and fungal burden and the promotion of a non-protective M2 phenotype as indicated by the increased secretion of IL-10 (Ene et al., 2012; Ene et al., 2013) ( Figure 1 ). Since many of the cell wall components are recognized by pattern-recognition receptors on antigen presenting cells, alterations in the cell wall structure may result in the inability of PAMPs to bind their antigenic targets and generate a protective immune response (Lewis et al., 2012).