Introduction

Pathogenic fungi are a significant burden to public health, recently estimated to cause 2.5 million annually attributable deaths worldwide¹. Among these, Candida species are responsible for more than 100 million mucosal infections and 1.5 million systemic infections, representing the fourth most frequent cause of bloodstream infection in hospitalized patients^1–3. Candida albicans is the most common opportunistic species, which in healthy individuals often colonizes mucosal surfaces (oral cavity, gastrointestinal tract, as well as genital tract)⁴, but leads to invasive infection, post trauma or when immune system of the host is compromised. C. albicans is classified in the critical priority group of the WHO list of fungal pathogens⁵.

Fungal cells are enveloped by a cell wall, which is essential for protection, integrity, and to maintain cell morphology. During commensalism with human hosts, the cell wall enables fungi to survive host defenses, whereas during pathogenic interactions, the cell wall also plays a central role in the polarized growth of invasive hyphae, thereby facilitating host tissue invasion^6,7. The cell wall architecture is dynamic and therefore critical in modulating the host immune responses^8,9, which are driven by host immune receptors (pattern recognition receptors [PRRs])^10–12. Recognition of fungal pathogen-associated molecular patterns (PAMPs) by host PRRs triggers an array of antifungal responses, including phagocytosis, the production of reactive oxygen/nitrogen species (ROS/RNS), neutrophil extracellular traps, and the release of cytokines and chemokines that recruit immune cells to the site of infection and activate adaptive immunity^12–15. Being extracellular, thereby first to interact with the host immune system, fungal cell wall components function as the major PAMPs.

The cell wall of C. albicans is a complex structure with a fibrillar outer layer of mannoproteins and an inner polysaccharide layer composed of chitin, β-1,3-glucan and β-1,6-glucan, which are not present in mammalian cells¹⁶. Dectin-1, a C-type lectin receptor that recognizes fungal cell wall β- 1,3-glucan, has been demonstrated to play a dominant pro-inflammatory role in antifungal immunity^17–21. Chitin particles derived from the C. albicans cell wall lead to the selective secretion of the anti-inflammatory cytokine IL-10²². Mannans from C. albicans are recognized by multiple receptors expressed on different immune cells (Mannose receptor, DC-SIGN, TLR4, Dectin-2, Dectin-3), which are involved in yeast phagocytosis and appear to play a critical role in the induction of Th17 responses^23,24. C. albicans has developed strategies to sense environmental signals encountered within the host and to regulate the exposure of its cell wall PAMPs. For example, low levels of β-1,3-glucan exposure by C. albicans correlate with enhanced colonization of the gastrointestinal tract^25,26 and β-1,3-glucan is mainly hidden from immune recognition by the outer layer of mannan fibrils^11,27,28. Although these C. albicans cell wall polysaccharide PAMPs have been investigated in the context of host-fungal interaction, the immunomodulatory potentials of β-1,6- glucan, another major constituent of the C. albicans cell wall, remains under-investigated.

The composition and organization of the fungal cell wall result from an equilibrium between syntheses and remodeling of the constituent polysaccharides^23,29. β-glucans and chitin, which account for about 60% and 2-10% (depending on the morphotype), respectively, are cross-linked to form the skeletal armor of the C. albicans cell wall^15,30. Mannoproteins, which are predominantly glycosylphosphatidylinositol (GPI)-modified and covalently linked to the skeletal polysaccharides via β-1,6-glucan bridges^30,31, account for 30%–40% of the cell wall. β-1,6-glucan is composed of a linear chain of β-1,6-linked glucose residues with few short side-chains of β-1,3-glucoside or laminaribioside residues^32,33. According to analytical methods, the β-1,6-glucan content is highly variable, from 15 to 50%^34–37. Although β-1,6-glucan plays a crucial structural role in the C. albicans cell wall, the dynamics of this cell wall polysaccharide as a function of growth or stress conditions remains underexplored.

Genes required for the full β-1,6-glucan production have been identified in model yeast, Saccharomyces cerevisiae³⁸. Among them, KRE5 codes a putative α-glucosyltransferase³⁹ (GT-24) localized in the endoplasmic reticulum (ER); its deletion results in the absence of β-1,6-glucan in S. cerevisiae and about 80% reduction of cell wall β-1,6-glucan content in C. albicans^35,40. Functions of Kre1, Kre6, Kre9 and the related homologs are unclear. Their localization in the ER, Golgi, plasma membrane or extracellular culture medium, suggests an intracellularly initiating biosynthetic pathway of β-1,6-glucan^41,42. KRE6 and its functional homolog SKN1 share 66% identity and both encode transmembrane proteins containing a glycosyl-hydrolase domain (GH-16). KRE6 deletion in S. cerevisiae results in a 70% decrease in the cell wall β-1,6-glucan content³³. In C. albicans the double deletion of KRE6 and SKN1 leads to a slow growth, cell wall abnormalities, cell separation failure, and reduced virulence^43,44. Two more KRE6 homologs (KRE62 and SKN2) have been identified in C. albicans⁴⁵ but their function has not yet been investigated, and the C. albicans cell wall β-1,6-glucan biosynthetic pathway, namely the β-1,6-glucosyltransferase involved in β-1,6- glucan polymerization and other enzymes involved in β-1,6-glucan remodeling (elongation, branching or cross-linking with β-1,3-glucan) remains to be identified.

We standardized biochemical approaches to quantify as well as to characterize β-1,6-glucan of C. albicans, which allowed us to perform comparative analysis and to study the dynamic of this cell wall polysaccharide in response to stress conditions and to antifungal treatments. β-1,6-glucan biosynthesis appears to be a compensatory reaction when mannan elongation is defective, whereas the absence of β-1,6-glucan results in a significantly sick growth phenotype and complete cell wall reorganization. Moreover, β-1,6-glucan showed a distinct immunomodulatory response compared to other cell wall polysaccharides. Together, our study establishes overarching and critical roles for β- 1,6-glucans in the organization, structure, biosynthesis and immunomodulatory properties of the cell wall.

Results

β-1,6-glucan - a critical polysaccharide of C. albicans cell wall

Despite numerous studies^34–37, the composition and organization of C. albicans cell wall β-1,6-glucan remain unclear. To better understand the composition and organization of the cell wall β-1,6-glucan, we have fractionated cell wall and applied new biochemical strategies to quantify and characterize them. The C. albicans cell wall was separated into three fractions according to polymer properties and reticulations, namely: sodium-dodecyl-sulphate-β-mercaptoethanol (SDS-β-ME) extract, alkali- insoluble (AI) and alkali-soluble (AS) fractions (see methods). In Synthetic Dextrose (SD) medium at 37°C, the SDS-β-ME fraction was the smallest (18% of total cell wall) and mainly composed of mannoproteins (Fig. 1a). The AI fraction was the major one (46%) and was exclusively composed of glucose and glucosamine, the monosaccharide building blocks of β-glucans and chitin, respectively. The AS fraction (36% of total cell wall) was characterized by a high amount of mannose residues (54% of the fraction) resulting from the release of covalently cell wall-anchored mannoproteins (Fig. 1a), and this fraction also contained glucose residues originating from β-glucans. Overall, we found that the C. albicans cell wall is composed of 3% of chitin, 43% of proteins/mannoproteins (27% of mannans) and 54% of β-glucans, which was in agreement with earlier studies and validated our analytic approach (Fig.1a, 1b)^46–48.

Figure 1:

To differentiate β-1,3-glucans and β-1,6-glucans, we developed two methods, one for the AI fraction and other for the AS fraction (see methods, and Fig. S8) as β-glucan was found in both fractions. β- 1,6-glucan corresponded to 23% of the total cell wall, was mainly found in the AI fraction (Fig. 1a, 1b). β-1,6-glucan in the AI fraction could be solubilised by digesting β-1,3-glucan using an endo-β- 1,3-glucanase, confirming that it is covalently linked to the cell wall β-1,3-glucan as previously shown in S. cerevisiae^33,49. Further, gel filtration chromatography showed that the solubilised β-1,6-glucan was eluted in a single symmetric peak with Gaussian shape showing a homogeneous polymer with an apparent molecular weight range of 18-112 kDa (Fig. 1d), and the average molecular weight was estimated to be 58 kDa. The structure of purified β-1,6-glucan was analysed by ¹H and ¹³C-NMR (Table S2) and by HPAEC after endo-β-1,6-glucanase digestion (Fig. 1e). The NMR chemical shifts and coupling constants as well as the HPAEC profile of the digested products were highly comparable to those we described for S. cerevisiae³³, showing a similar structure: a main chain of β-1,6-glucosyl residues with side chains of β-1,3-glucosyl or β-1,3-laminaribiosyl residues. The relative intensity of the anomeric proton in NMR and HPAEC analysis after endo-β-1,6-glucanase digestion, indicated that an average of 6.4% of glucose units of the main chain is substituted and that 88-90% of the side chains is composed of a single glucose residue (Fig. 1c, 1e, Table S2). Altogether, our data showed that β-1,6-glucan is a major cell wall polymer in C. albicans, representing 40% of the cell wall β-glucans.

Environmental regulation of β-1,6-glucan synthesis

The composition and surface exposure of fungal cell wall polymers is dependent on environmental growth conditions^26,29. We applied chemical assays and our biochemical approaches to enable comparative analyses of cell wall robustness in relation to β-1,6-glucan dynamics in C. albicans.

First, we analysed the cell wall composition in relation to cell morphotype and growth phase (exponential vs. stationary phase, yeast vs. hyphae, planktonic vs. biofilm). No significant difference was observed between cell wall β-1,6-glucan and other polymers produced during the exponential (control) and stationary phases in yeast cells (Fig. 2). In biofilms, only a slight increase of mannan content (26% vs. 20%), but no change in relative β-glucan content was observed. In contrast, the cell wall isolated from C. albicans hyphae was characterized by an increase in chitin content (10% vs. 3%) as previously shown^29,48,50, as well as a reduction of β-1,6-glucan (18% vs. 23%) and mannan contents (13% vs. 20%) compared to the yeast form. The decrease in β-1,6-glucan content was reflected in a reduction of the β-1,6-glucan size (43 vs. 58 kDa), demonstrating that both content and structure of β-1,6-glucan are regulated during growth and cellular morphogenesis. In addition, a higher branching rate of β-1,3-glucan of the hyphal cell wall was also observed in the AI fraction (7.3% vs. 5.3%) (Fig. S3).

Figure 2:

We further investigated the effect on cell wall composition of environmental growth conditions, such as temperature, pH, carbon source and hypoxia, on cell wall composition (Fig. 2). Hypoxic condition (1% O2, 5% CO2) resulted in a slow growth but did not lead to a significant alteration of the cell wall composition. In contrast, all other tested conditions led to a modification of cell wall composition including the proportion of β-1,6-glucan. For example, growth at 30°C or in 1M NaCl led to a decrease in the cell wall β-1,6-glucan content and a decrease of β-1,6-glucan chain length. However, these decreases at low temperature were compensated by a higher content of less-branched β-1,3- glucan, whereas high salt led to an increased branching rate of both β-1,6-glucan and β-1,3-glucan (Fig. 2, Fig.S3). Buffered culture medium at pH 4 or 7.5 led to a slight decrease in the β-1,6-glucan molecular weight and an increase in its branching frequency in comparison to buffered medium at pH 5.4. However, the most pronounced alteration of the cell wall was observed in lactate containing medium, with a steep reduction of polysaccharides from the structural core, chitin (−43%), β-1,3- glucan (−48%) and mainly β-1,6-glucan (−72%), in agreement with the decrease in thickness of the inner cell wall layer previously described⁵¹. Lactate medium also led to a strong reduction in the average β-1,6-glucan molecular weight (19 vs. 58 kDa), a decrease of β-1,6-glucan branching (4.9 vs. 6.4%) and an increase of β-1,3-glucan branching (7.4 vs. 5.3%, Fig. S3).

The presence in the medium of short chain fatty acids (acetate, propionate, or butyrate) at physiological gut concentrations resulted in a growth defect but did not alter the global cell wall composition. The main difference was an increase of β-1,6-glucan molecular weight (64-68 kDa vs. 58 kDa) (Fig. 2b). Oxidative stress and drugs targeting the cell wall did not induce any major alterations (Fig. 2). Cell wall chitin content was reduced in the presence of nikkomycin, a chitin synthesis inhibitor, and increased in the presence of the cell wall damaging agents Congo Red or caspofungin, as previously observed^52–54, but with no impact on β-1,3-glucan and β-1,6-glucan contents. The major stress effect we observed was in presence of caspofungin, which caused an increase of the average molecular weight of β-1,6-glucan (70 vs. 58 kDa).

These data demonstrate that the content, size, and branching of β-1,6-glucan in the C. albicans cell wall is regulated by numerous environmental growth and stress conditions.

β-1,6-glucan dynamics responds to alterations in inner cell wall synthesis

We investigated the dynamics of β-1,6-glucan in C. albicans mutants in which the genes involved in the biosynthesis or remodelling of chitin and β-1,3-glucan were deleted. Four chitin synthases, located at the plasma membrane, have been described in C. albicans and we have analysed three deletion mutants. Chs1 is involved in primary septum formation and is essential for vegetative growth. We therefore analyzed the conditional PMRP1-CHS1/chs1Δ mutant under repressive condition. Chs2 is a non-essential class II enzyme and represents the major chitin synthase activity measured in cell membrane preparations⁵⁵. Chitin synthase 3 (Chs3) is responsible for most of the cell wall chitin production and chitin ring formation at sites of cell division⁵⁶ but it is not essential for growth and viability. Our comparative analyses did not show any alteration in the chitin content in the PMRP1-CHS1/chs1Δ and chs2Δ/Δ mutants, but a strong reduction was observed in the chs3Δ/Δ mutant (−92.3%) (Fig. S2) in accordance with previous data^57,58. In the PMRP1-CHS1/chs1Δ mutant, the cell wall had a 45.7% decrease in mannan content in the AS fraction and a 50% increase in β- 1,6-glucan in the AI fraction (Fig. S2, Fig. 3a). In the chs3Δ/Δ mutant, the low proportion of chitin led to an increase of both β-1,3-glucan (x1.6) and β-1,6-glucan (x1.6) in the AS fraction (Fig. 3a, Fig. S2), suggesting a defect in β-glucans cross-linking in the cell wall polysaccharide core. No significant change in β-1,6-glucan structure was noticed for these two mutants (Fig. 3b, 3c). Strikingly, no significant effect of the CHS2 deletion was observed on cell wall chitin content or overall wall composition.

Figure 3:

FKS1, FKS2 and FKS3 encode β-1,3-glucan synthases in C. albicans^48,59. As FKS1 is essential for vegetative growth⁵⁹, we used the heterozygous fks1Δ mutant, which displays a haplo-insufficiency phenotype⁶⁰. Only a decrease (−58.7%) in β-1,3-glucan content in the AS fraction (Fig. S2) was observed, without any other significant changes. Moreover, essentiality of FKS1 prevented analysis of β-1,6-glucan dynamics in a β-1,3-glucan-deficient mutant. PHR1 and PHR2 encode β-1,3- glucanosyltransferases of the GH-72 family involved in β-1,3-glucan elongation and branching at the extracellular cell wall level^61,62. In C. albicans their expression is strongly pH dependent: in vitro, PHR1 is expressed above pH 5.5, while PHR2 is expressed below pH 5.5⁶³. The phr1Δ/Δ mutant was thus grown at pH 7.5 and phr2Δ/Δ at pH 5.5. As expected for both mutants, defects in GH-72 β-1,3-glucan remodelase led to a decrease in β-1,3-glucan branching (−67.3% for phr1Δ/Δ, −31.5% for phr2Δ/Δ) (Fig. S4) and a compensatory increase in the cell wall chitin content (1.5x for phr1Δ/Δ, and 3.7x for phr2Δ/Δ) (Fig. S2). No change in the cell wall β-1,3-glucan content was observed in the phr1Δ/Δ mutant (Fig. S2). In contrast, deletion of PHR2 led to an increase of β-1,3-glucan content and a cross-linking defect (−31.5% estimated on the ratio β-1,3-glucan content in AS and AI fractions). Concomitantly, deletion of PHR2 led to a reduction of cell wall β-1,6-glucan content in both AI (−58.7%) and AS (−51.3%) fractions and to a reduction of molecular weight and branching of the β-1,6-glucan chains (Fig. 3a). In the phr1Δ/Δ mutant, only a significant increase of the molecular weight of β-1,6-glucan chain was observed (60 vs. 50 kDa) (Fig. 3).

Taken together, these data showed that mutational disruption of chitin-synthase or β-1,3-glucan- synthase genes did not significantly alter the cell wall β-1,6-glucan structure. However, defects in chitin content led to a decrease in the cross-linking of β-glucans in the inner wall (i.e. AI Fraction) that corresponds to the effect of nikkomycin treated C. albicans phenotype (Fig. S1); conversely, an increase in chitin content led to more cross-linking of β-glucans as observed in the FKS1 mutant or in the presence of caspofungin (Fig. S1 and S2). In contrast, disruption of cell wall β-1,3-glucan remodelling had a pH-dependent effect on β-1,6-glucan structure (Fig. 2 and 3).

A β-1,6-glucan mannan compensatory pathway links the regulation of the inner and outer cell wall layers

In C. albicans, mannosylation of cell wall proteins is required for the organization of the fibrillar outer layer of the cell wall that is composed of N-linked mannans^27,64. This fibrillar layer is anchored to the polysaccharide core composed of β-1,3-glucan and chitin, via β-1,6-glucan bridges³⁰. We analyzed mutants deficient in Golgi α-mannosyltransferases involved in O- and N-mannosylation. Mnt1 and Mnt2 are α-1,2-mannosyltransferases involved in catalyzing consecutive reactions in O-mannan elongation^23,65. As predicted, the cell wall of the mnt1/mnt2Δ/Δ mutant showed minor changes in overall mannan content (Fig. 3d, Fig. S2)⁶⁵. However, deletion of MNT1 and MNT2 genes led to a strong reduction of β-1,6-glucan molecular weight (18 vs. 43 kDa) and an increased branching of β- 1,6-glucan chains (8.6 vs. 5.3%) (Fig. 3e, 3f). A slight increase in the β-1,3-glucan branching was also observed in the mnt1/mnt2Δ/Δ double mutant (Fig. S4). Overall, these data suggested that β- 1,6-glucan and mannan syntheses may be coupled.

N-mannan elongation requires the activity of a set α-1,6-mannosyltransferases in the Mannan Polymerase I and II complexes for the formation of α-1,6-mannan backbone that is subsequently modified with α-1,2- and α-1,3-mannosyltransferases that assemble the N-mannan side chains²³. Och1 activity is responsible for the initial addition of α-1,6-mannose onto the relatively well conserved N-glycan triantennary complex of glycoproteins, and Mnn9 activity is essential for the polymerization of the α-1,6-mannose backbone. Mnn2, Mnn21, Mnn22, Mnn23, Mnn24 and Mnn26 activities are essential to the synthesis of side chains²³. Single och1Δ/Δ and mnn9Δ/Δ mutants and the sextuple mnn2/21/22/23/24/26Δ/Δ mutant resulted in a major decrease in mannan content in the AS fraction (−90.4% for och1Δ/Δ, −88.9% for mnn9Δ/Δ and −87.1% for mnn2/21/22/23/24/26Δ/Δ) and a compensatory increase in chitin content (x3.1 for och1Δ/Δ, x2.5 for mnn9Δ/Δ and x2.5 for mnn2/21/22/23/24/26Δ/Δ) in the AI fraction (Fig. S2), as previously described²³. A significant increase of cell wall β-1,3-glucan content was observed in the mnn9Δ/Δ and mnn2/21/22/23/24/26Δ/Δ mutants. There was also a significant increase in the β-1,6-glucan content (x3.2 for och1Δ/Δ, x2.4 for mnn9Δ/Δ and x2.4 for mnn2/21/22/23/24/26Δ/Δ respectively) concomitant with the elongation of β-1,6-glucan chains, and consequential increase in its molecular weight (Fig. 3d, 3e).

Overall, these data suggest that the β-1,6-glucan biosynthesis is stimulated via a compensatory pathway when there is a defect in O- and N-linked cell wall mannan biosynthesis. This novel compensatory response occurs by increasing the β-1,6-glucan content in N-mannan elongation- deficient mutants and increasing branching of shorter β-1,6-glucan chains in O-mannan-deficient mutants.

KRE6-family-members dependent β-1,6-glucan biosynthesis

Genes encoding proteins essential for β-1,6-glucan biosynthesis were first identified in S. cerevisiae as mutants resistant to Killer Toxin K1 (KRE)^38,66. In addition, a range of proteins in the endoplasmic reticulum and Golgi (Cwh41, Rot2, Kre5, Kre6, Skn1) are critical for the assembly of this polymer⁶⁶. We analyzed homologs of CWH41, ROT2, KRE5 in C. albicans. These three calnexin cycle proteins are involved in glucosylation/deglucosylation of N-glycan in the ER⁶⁷ had a reduced amount of cell wall mannan (Fig. S2). Strong cell wall phenotypes were observed for kre5Δ/Δ and rot2Δ/Δ mutants which exhibited significant increases in the highly branched β-1,3-glucan (x1.8 for rot2Δ/Δ, x2.2 for kre5Δ/Δ) (Fig. S4), chitin (x4.2 for rot2Δ/Δ, x5.2 for kre5Δ/Δ) and a strong decrease in β-1,6-glucan in the rot2Δ/Δ (−68.7%) and kre5Δ/Δ mutants (−81.5%) (Fig. 3g and Fig. S2). In addition, the β-1,6- glucan structure was also altered in both mutants, with a decrease in the average molecular weight (15-16 kDa instead of 43 kDa for the parental strain) and a reduction in the side chain branching (−54.2% for rot2Δ/Δ, and −76.5% for kre5Δ/Δ) (Fig. 3h, 3i). KRE1 encodes a plasma membrane protein whose function is not yet clear^68,69. The kre1Δ/Δ C. albicans mutant exhibited a reduced molecular weight of β-1,6-glucan chains (20 kDa instead of 43 kDa in the parental strain) (Fig. 3h) but was not affected in overall cell wall composition (Fig. S2).

In C. albicans, the KRE6 family is composed of four homologs: KRE6, SKN1, KRE62 and SKN2. These genes encode putative trans-β-glycosylases of the GH-16 family for which no biochemical activity has been described to date⁷⁰. To analyze the function of these four proteins in C. albicans, we used a transient CRISPR/Cas9 protocol to build the four single knockout mutants, the double kre6/skn1Δ/Δ and the quadruple kre6Δ/Δ/kre62/skn2/skn1Δ/Δ mutants. Among the single mutants, only kre6Δ/Δ exhibited a marked alteration in the cell wall composition (Fig. S2), with significantly lower β-1,6-glucan (−46.8%) (Fig. 3g) and mannan content (−42.8%) (Fig. S2), and higher β-1,3- glucan (x1.9) and chitin (x3.9) content. In addition, the β-1,6-glucan chain length was shorter (−50.5%) and more branched (x1.7) than in the parental strain (Fig. 3h, 3i) and β-1,3-glucan was also hyperbranched (x1.8) (Fig. S4). The kre6/skn1Δ/Δ mutant phenotype, β-1,6-glucan content and structure were similar to the single kre6Δ/Δ mutant (Fig. 3g, 3h, 3i). The quadruple kre6/kre62/skn2/skn1Δ/Δ mutant did not produce any measurable β-1,6-glucan (Fig. 3g, Fig. S2), which was confirmed by endo-β-1,6-glucanase digestion of the AI and AS fractions (Fig. S6). The absence of β-1,6-glucan was associated with an increased chitin content (9.5% vs. 2.5%), and highly branched β-1,3-glucan (48% vs. 20% in the parental strain) (Fig. S2, Fig. S4). Reintegration of KRE6 under the control of the ACT1 promoter restored the production of β-1,6-glucan with a structure similar to that of the parental strain (Fig. 3g, 3h, 3i).

Therefore, in C. albicans, deletion of all KRE6 homologous genes led to the absence of β-1,6-glucan production and structural alterations of the cell wall, showing that the β-1,6-glucan biosynthetic pathway is critical to the regulation of cell wall homeostasis.

KRE6 regulation of cell wall architecture

The single deletions of KRE62, SKN1 and SKN2 did not affect the growth of C. albicans in SD medium nor drug sensitivity (Fig. 4), whereas the deletion of KRE6 led to a significant increase in doubling time (x2.4, Fig. 4a, b). The double kre6/skn1Δ/Δ mutant had the same growth rate as kre6Δ/Δ (Fig. 4a, b). However, deletion of all genes of the KRE6 homologs family (quadruple kre6/kre62/skn2/skn1Δ/Δ mutant) resulted in a stronger growth phenotype with a longer lag phase and a higher doubling time (x3.5 vs. parental strain). Cells of the quadruple kre6/kre62/skn2/skn1Δ/Δ mutant were bigger than the parental strain (Fig. 4d). In addition, the single kre6Δ/Δ and all multiple mutants aggregated in SD liquid medium and had increased sensitivity to the cell wall perturbing agents Congo Red, Calcofluor White, as well as to the chitin synthase inhibitor nikkomycin Z and the N-glycosylation inhibitor tunicamycin. No growth of the quadruple mutant was observed in the presence of either of these drugs. Surprisingly, caspofungin enhanced the growth of mutants bearing KRE6 gene deletion (Fig. 4c). Among the single mutants, only skn1Δ/Δ was defective for filamentous growth; all multiple mutants also displayed defects in filamentous growth, and the severity of this phenotype was proportional to the number of KRE6 homologous genes deleted (Fig. 4d). The reintegration of KRE6 under PACT1 restored parental strain-like growth kinetics, filamentation and drug resistance (Fig. 4) and confirmed the above differences noted in the mutants of KRE6 and SKN1⁴⁴.

Figure 4:

Cell wall architecture, observed by transmission electron microscopy, showed that single skn1Δ/Δ, kre62Δ/Δ and skn2Δ/Δ mutants presented a bilayered structure with a classical fibrillar outer layer similar to the parental strain (Fig. 5). However, the outer layer of the skn1Δ/Δ and skn2Δ/Δ mutants was thicker than the WT (skn1Δ/Δ, 70 ± 12 nm; skn2Δ/Δ, 66 ± 12 nm vs. WT, 53 ± 10 nm). The cell wall of the kre6Δ/Δ and kre6/skn1Δ/Δ mutants had a heterogeneous and thicker inner layer (kre6Δ/Δ, 210 ± 89 nm; kre6/skn1Δ/Δ, 216 ± 68 nm vs. WT, 108 ± 23 nm) (Fig. 5). The external mannoprotein-rich fibrillar layer was absent in the quadruple kre6/kre62/skn2/skn1Δ/Δ mutant and the inner layer was thicker and of variable thickness (440 ± 158 nm vs. 108 ± 23 nm). A normal wall architecture was fully restored in the complemented strain.

Figure 5:

These results demonstrate that β-1,6-glucan has a pivotal role in determining the architectural arrangement of polysaccharides in the cell wall, which in turn influences growth, drug sensitivity, cell morphology and filamentation.

β-1,6-glucan stimulates human peripheral blood mononuclear cells in vitro

The major cell wall polysaccharides of C. albicans, β-1,3-glucan, mannan, and chitin function as PAMPs, and their recognition by host PRRs trigger immune responses. Although β-1,6-glucan is one of the major constituent polysaccharide in the cell wall of C. albicans, its immunomodulatory potential has been understudied. We extracted β-1,6-glucan from the cell wall of the parental strain and used this to stimulate peripheral blood mononuclear cells (PBMCs) and neutrophils isolated from the whole blood samples of healthy human donors. As controls, PBMCs were stimulated with the AI fraction (composed of β-1,3-glucan, β-1,6-glucan and chitin) as well as with the periodate-oxidized AI fraction (AI-OxP) devoid of β-1,6-glucan. A protein profiler was used to identify cytokines, chemokines and acute phase proteins differentially released by PBMCs and neutrophils stimulated with the three different cell wall fractions (AI, AI-OxP and β-1,6-glucan) (Fig. S12, S13). Then, differentially stimulated cytokines, chemokines and acute phase proteins were quantified by ELISA. Fractions containing β-1,3-glucan (AI and AI-OxP fractions) induced the secretion of pro- inflammatory cytokines including IL-6, IL-1β and TNF-α, chemokines such as IL-8, MCP-1, MIP-1β and RANTES, and the complement factor C5a (an acute phase protein), but anti-inflammatory cytokine IL-10 was released in a relatively small amount (Fig. 6a). However, removal of β-1,6-glucan by periodate oxidation (AI-OxP) led to a significant decrease in the IL-8, IL-6, IL-1β, TNF-α, C5a and IL-10 released, suggesting that their stimulation was in part β-1,6-glucan dependent. Although the response was much lower than that induced by the AI fraction and AI-OxP, β-1,6-glucan activated PBMCs and induced secretion of most cytokines and chemokines (Fig. 6a). Both AI and AI-OxP fractions stimulated neutrophils secreted IL-8 at a similar level (Fig. 6b) but purified β-1,6-glucan did not, suggesting that β-1,6-glucan and β-1,3-glucan stimulate innate immune cells in distinct ways. Because the structure of cell wall β-1,6-glucan is dependent on stress as well as environmental conditions of growth, we investigated the structure-function immune response of β-1,6-glucan from C. albicans. Shorter β-1,6-glucan chains (19 kDa) were purified from the cell wall of C. albicans cell wall grown with lactate as carbon source and larger β-1,6-glucan chains (70 kDa) were purified from cell wall produced in the presence of caspofungin (Fig. 2). Both fractions stimulated PBMCs and led to cytokine/chemokine secretion at a level similar to the control β-1,6-glucan, suggesting that β-1,6- glucan size was not a significant factor influencing the stimulation of PBMCs (Fig. S5).

Figure 6:

Discussion

Cell wall β-1,6-glucan is a conserved constituent in most of the major human fungal pathogens including Candida, Cryptococcus, Pneumocystis, Histoplasma and Malassezia species ^47,71,72. The cell wall of fungal pathogens is the target of several classes of antifungal drugs and is critical for immune recognition and activation, therefore has been the subject of numerous investigations^29,47. However, this polysaccharide has been neglected, perhaps because its analysis requires bespoke methodologies. Here, we established new biochemical tools for quantitative and comparative analysis of cell wall β-1,6-glucan dynamics and structure in C. albicans and used them to demonstrate that this polysaccharide is instrumental in determining the gross architecture of the fungal wall and its immune reactivity.

In SD medium at 37°C, β-1,6-glucan represents 23% of the total cell wall, with an average size of 58 kDa (Fig. 1). From our analytical data on C. albicans cell wall, we proposed an updated model underlining the true role of β-1,6-glucan as a major cell wall component (Fig. 7a). NMR has demonstrated that the structure of β-1,6-glucan in the yeast cell wall is similar between species^33,71, but there are variations in β-1,6-glucan content, chain length and branching. Malassezia restricta contains long β-1,6-glucan chains, averaging 80 kDa, whilst in S. cerevisiae and C. albicans they are moderate in size (30-60 kDa). We showed here that modulating environmental conditions induce significant changes in both cell wall structure and content. As described previously, lactate used as the sole carbon source led to a thinner inner cell wall layer and greater exposure of β-1,3-glucan^51,73. We observe here that lactate medium also led to a marked reduction in cell wall β-1,6-glucan content and chain length (Fig. 2, Fig. 7b). Our data showed that several factors related to host niches (morphology, temperature, salinity, pH and presence of organic acids) and the presence of antifungals had a significant impact on the structure, size and branching of β-1,6-glucan (Fig. 7b). This dynamic nature of the β-1,6-glucan component of the wall (Fig. 2, Fig. S1 and Fig. S2) is as important that the regulation and structural diversity of chitin, mannan and β-1,3-glucan⁶⁴.

Figure 7:

The dynamic regulation of mannan biosynthesis leading to structural changes in the outer mannoprotein layer of C. albicans is also under-investigated. C. albicans evolves in different environments as a commensal or pathogen⁴, mainly in the gut, oral cavity, genital tract and blood. How fungal cell walls adapt to changing environments was recently been listed as one of the top five unanswered question in fungal cell surface research⁷⁴. Our data demonstrate that in response to several environmental factors β-1,6-glucan and mannan productions are coupled. The in vivo importance of this novel dynamic process remains to be addressed but it could be critical in the articulation of host inflammatory responses^75,76.

Immunolabelling with an anti-β-1,6-glucan serum demonstrated that this polymer is exposed at the cell surface and is therefore accessible to receptors that stimulate human immune cells (Fig. S11 and Fig. 6). Similarly in Pneumocystis carinii, β-1,6-glucans are exposed to the cell surface and stimulate mouse macrophages leading to the release of TNF-α⁷². Our immunological data indicate that β-1,6-glucan is a bona fide fungal cell wall PAMP that activates the human immune system (Fig. 7d). Beads coated with pustulan (an insoluble linear form of β-1,6-glucan) activate neutrophils and induce massive ROS production and phagocytosis in a dectin-1-independent manner^77,78. We observed that the absence of human serum in the culture medium led to a strong reduction of β-1,6- glucan induced cytokine secretion from PBMCs and that β-1,6-glucan bind to C3b (Fig. S14). These data are in agreement with the binding of pustulan to C3b/C3d, as described previously^77,78. This allow us to assume that the opsonization by complement component C3 could partially mediated the recognition of β-1,6-glucan by immune cells, but needs to be validated experimentally. We observed that the size of soluble β-1,6-glucan did not alter immune stimulation in vitro (Fig. S5) whereas, despite its ability to bind soluble and particulate β-1,3-glucan polymers, Dectin-1 signaling is only activated by particulate β-1,3-glucans⁷⁹. The relevance of fungal β-1,6-glucan to the global host immune response and mechanism of recognition remain to be investigated.

In C. albicans, β-1,3-glucan exposure is regulated by external signals such as hypoxia, carbon source or iron deprivation via the up-regulation of β-1,3-glucanase activities that lead to β-1,3-glucan masking and immune evasion^75,80,81. Regulation of the mechanisms determining β-1,6-glucan exposure and the expression of secreted glucanases remains also unknown in C. albicans. The absence of β-1,6-glucan in the AI fraction led to a reduced inflammatory response (Fig. 6), suggesting that an endo-β-1,3-glucanase, such as Eng1, may enable the release of cross-linked β- 1,6-glucan from the cell surface, which could attenuate the host response by pruning β-1,6-glucans that are covalently bound to β-1,3-glucans from the surface ⁸¹.

Cell wall organization in C. albicans depends on the cross-linking of all polymers that function as a dynamic flexible network ^29,47. Their interconnectedness means that alteration in the synthesis of one polymer can lead to a compensatory change and cell wall remodelling. Our data provides one new example of this phenomenon – namely decrease in cell wall chitin content results in an increase of both β-1,3-glucans and β-1,6-glucans in the AS fraction of the wall. Chitin exists in the yeast cell wall in three forms: free (∼40%), and crosslinked to β-1,3-glucan (40-45%) or β-1,6-glucan (15-20%) and crosslinked β-glucan-chitin forms the core fibrillar cell wall structure resistant to alkali- solubilization^29,82–84. Therefore, a decrease in the chitin content might reduce the formation of fibrillar structure, thereby releasing the β-glucan contents into the AS-fraction. On other hand, in the FKS1 mutant or upon caspofungin treatment, cell wall chitin and β-glucan-chitin contents were increased, which could be attributed to an upregulated CRH1 expression due decreased β-1,3-glucan synthesis as in FKS1 mutant⁸⁵ or due to caspofungin treatment. CRH family members are involved in the β- glucan-chitin crosslinking⁸⁶. Interestingly, defects in the expression of CHS1 that encodes an essential chitin synthase required to construct the primary septum, was also compensated by an increased content of cell wall β-1,6-glucan but not of β-1,3-glucan (Fig. 3). This suggests that Chs1 has other functions beyond septum construction, that had previously been suggested by loss of cell integrity and swelling of cells upon CHS1 repression⁸⁷. In S. cerevisiae, deletion of GAS1, the ortholog of PHR1 and PHR2, led to a strong decrease of cross-linked β-1,6-glucan to β-1,3-glucan- chitin cell wall core⁶². Deletions of PHR1 and PHR2, holding a β-1,3-glucan remodelling activity, led to an increased chitin content (Fig. S2) and to a reduction of β-1,3-glucan branching (Fig. S4). However, a decrease in β-1,6-glucan content and β-1,6-glucan size was only observed in the PHR2 mutant (Fig. 3). As PHR1 and PHR2 genes are strongly regulated by external pH, the compensatory differences we described may be explained by pH dependent regulation of β-1,6-glucan synthesis (Fig. 3).

In C. albicans, the mannan moieties of glycoproteins are covalently cross-linked to the β-glucan- chitin core. Mutants defective in N-mannan elongation led to a marked increase in β-1,6-glucan content and size (Fig. 3). In yeast, Och1 and Mnn9 are α-1,6-mannosyltransferases that are involved in initial steps of α-1,6-mannan polymerization in the Golgi. Members of the Mnn2 α-1,2 mannosyltransferases are required for the addition of the side chains of the α-1,6-mannan backbone²³. Mutants in these genes led to defects in the outer fibrillar layer^27,88 and increased β-1,3- glucan exposure^23,89. The quantification of β-1,6-glucan content in these N-mannan deficient mutants had not been described previously in C. albicans^23,88 and underline the new finding that the regulation of mannan, β-glucans and chitin synthesis is coordinated and coupled. Screening with K1 toxin killer revealed a hypersensitivity of the Scmnn9 mutant and an increase in β-1,6-glucan cell wall content⁹⁰ reinforcing the hypothesis that β-1,6-glucans play a compensatory role when N-mannosylation is compromised (Fig. 7c). Thus, a defect in the fibrillar outer mannan layer led to an increase in the inner cell wall chitin and β-1,6-glucan contents (Fig. S2), suggesting their synthesis was regulated by a rescue mechanism associated with the reprogramming of specific transcriptional responses via the cell wall integrity (CWI) pathway⁹¹. Defects in O-mannosylation also affected β-1,6-glucan synthesis leading to shortened highly branched β-1,6-glucan chains, without significant changes in overall cell wall composition, suggesting a specific inhibition of the elongation of β-1,6-glucan chain and showing that size and branching of β-1,6-glucan can be regulated independently (Fig. 7c).

Fine details of the biosynthetic pathway of yeast β-1,6-glucans are not clear. In S. cerevisiae, deletion of KRE5 and ROT2 led to a reduction in cell wall β-1,6-glucans⁶⁶. In C. albicans, this reduction is characterized by the presence of short seldom-branched β-1,6-glucan chains (Fig. 3). Although no difference was observed in the CWH41 mutant, our data imply that canonical N-glycan processing in the ER is required for complete β-1,6-glucan synthesis⁹⁰, and that Kre5p is not the β- 1,6-glucan polymerase³⁵. Furthermore, our results confirm that the β-1,6-glucan biosynthetic pathway is similar in these phylogenetically distant yeasts, and that the N-glycosylation pathway is critical for β-1,6-glucan synthesis. Defects in early N-glycan synthesis at the ER level led to concomitant defects in β-1,6-glucan synthesis while defects in N-mannan elongation led to increased β-1,6-glucan synthesis. However, as Kre5, Rot2 and Cwh41 are part of the calnexin cycle involved in the control of N-glycoprotein folding in the ER, early N-glycan synthesis may have an indirect effect in ensuring the functionality of the enzymes rather than the synthesis of N-glycan moiety required for β-1,6-glucan synthesis. Our hypothesis is that this biosynthesis begins intracellularly. According to in vitro systems in S. cerevisiae, the initial step is the polymerization of a linear chain by a β-glucosyltransferase using UDP-glucose as the sugar donor^33,92 and probably an as yet unknown acceptor (Fig. S15). The second step would be the addition of glucoside and laminaribioside as a side chain. Finally, after release into the cell wall, β-1,6-glucan will be cross- linked to the other cell wall polymers. Among the KRE genes, only KRE6 may be directly involved in β-1,6-glucan synthesis. Kre6 and its homologs contain a β-glycosyl-hydrolase domain of the GH-16 family, which includes a large number of fungal enzymes such as Crh1 homologs that are involved in the cross-linking of chitin and β-1,3-glucan⁸⁶. In C. albicans, the quadruple deletion of KRE6, SKN1, SKN2 and KRE62, led to the complete absence of β-1,6-glucan in the cell wall (Fig. 3, Fig. 7e) but also in culture supernatants and in intracellular extracts (not shown), demonstrating a stop of β-1,6-glucan synthesis rather than miss-localization. The absence of β-1,6-glucan in the quadruple mutant led to global cell wall remodeling and the absence of the outer fibrillar N-mannan layer (Fig. 5 and Fig. 7e). Kre6 had the dominant role and its absence led to a hypersensitivity to cell wall- targeting antifungal drugs^34,44. One exception was observed for caspofungin which enhanced the growth of all KRE6 deleted mutants (Fig. 4c). Growth above the MIC (paradoxical growth) has been described at high caspofungin concentrations and is associated with high chitin content^43,93 as observed in these mutants. Our data show that Kre6 is required for production of normal amounts of β-1,6-glucan and that the remaining β-1,6-glucans present in kre6Δ/Δ was composed of short and highly branched chains. This suggests that the Kre6 family is required for the polymerization and/or elongation of β-1,6-glucan, but not glucan branching.

In S. cerevisiae, only two KRE6 family members have been identified, Kre6 and Skn1, and their absence led to strong growth defects and hypersensitivity to cell wall perturbing agents –similar to what we observed here in C. albicans ⁹⁴. In C. neoformans, among the six KRE6 homologs, KRE6 and SKN1 are required for the production of cell wall β-1,6-glucan⁹⁵ and although the KRE6 family is essential for β-1,6-glucan synthesis, its biochemical function remains unknown. Kre6 homologs are type II membrane proteins with a GH-16 domain at the luminal C-terminus and an unfolded cytosolic N-terminal tail⁹⁶. Yeast Kre6 and Skn1 proteins have moderate similarity with diatom TGS (1,6-β-transglycosylase) proteins^97,98 and can synthesize glucan from UDP-Glc⁹⁹. Kre6 homologs share structural similarities with β-glucanases and β-transglycosylases, arguing for a possible cell wall remodeling function. In S. cerevisiae, Kre6p has been localized in the endoplasmic reticulum, the plasma membrane and secretory vesicle-like compartments^66,96,100, and the cytosol tail was shown to be essential for the correct cellular localization and β-1,6-glucan synthesis⁹⁶. Surprisingly, in C. albicans the cytosol tail is not essential⁴⁴, suggesting some functional differences of this protein in these yeasts. In C. albicans, the double KRE6/SKN1 mutant is avirulent in a mice model of candidiasis⁴⁴, showing that the β-1,6-glucan biosynthetic pathway is a potential virulence target. However, the drugs Jervine (steroidal alkaloid) and a pyridobenzimidazole derivate (D75-4590), that may target Kre6 and inhibit S. cerevisiae growth^101,102, had no effect on C. albicans^101,102, presumably representing species-specific differences in drug profiles. Kre6 family members are present in all fungal pathogens producing cell wall β-1,6-glucans and further investigations are required to understand the enzyme activities and the diversity of structures related to the potential drug-targeting of these proteins.

Overall, this work highlights the central role that β-1,6-glucan plays in orchestrating cell wall synthesis during growth and in response to stress and sets the focus for further work that can exploit this central role in mitigating fungal infections.

Methods

Ethics statement

Human blood samples were taken from healthy donors from Etablissement Français du Sang Trinité (Paris, France) with written and informed consent, as per the guidelines provided by the institutional ethics committee, Institut Pasteur (convention 12/EFS/023).

Strains and growth conditions

The Candida albicans strains used in this study are listed in Table S1. The reference strain SC5314, used for all stress conditions, was grown overnight in 25 mL flasks, at 37°C, 180 rpm, in SD medium (2% glucose, 0.67% yeast nitrogen base with amino acids (YNB) (BD), pH 5.4). Overnight cultures were diluted to an OD600= 0.001, in 50 mL fresh SD medium, incubated at 37°C, 180 rpm, and then harvested either when reaching mid-exponential phase (OD600 ∼ 3-5) or at stationary phase (OD600 ∼ 11-12 after 30h). The following growth conditions were selected: hypha induction (SD containing 2% GlcNAc instead of 2% glucose, pH 7.2, no yeast-form was observed in this medium), temperature (30°C), osmolarity (1M NaCl), carbon source (SD containing 2% lactate (ChemCruz) instead of 2% glucose), oxygen limitation (hypoxia: 1% O2, 5% CO2, anaerobic chamber (Baker Ruskinn, I-CONIC, Bugbox M)) and oxidative stress (H2O2 1.5 mM, Prolabo). For studies of the effect of medium pH, SD medium was buffered with 100 mM MES (Sigma) at pH 4 or 5.4 or with MOPS (Sigma) at pH 7.5. For experiments on the effects of short-chain fatty acids (SCFAs), SD medium was supplemented with 100 mM MES and 50 mM acetate (Sigma) or 20 mM propionate (Sigma) or 15 mM butyrate (Sigma) and buffered to pH 5.4. Drugs were tested at sub-lethal concentrations: Calcofluor White (CFW, 400 µg/mL, Sigma-Aldrich), Congo Red (CR, 50 µg/mL, Sigma), tunicamycin (0.125 µg/mL, Sigma), nikkomycin (16 µg/mL, Sigma) and caspofungin (0.015 µg/mL, Sigma). Biofilm production was induced in SD medium at 37°C for 48h in 6 well plates as previously described with some modifications¹⁰³. Overnight cultures were diluted at OD600=0.2 in 6 well polystyrene plates (TPP) in 2 mL of minimal medium (SD: 2% glucose, 0.67% yeast nitrogen base without amino acids (BD), pH 5.4) supplemented by arginine 0.1 g/L, histidine 0.1g/L, uridine 0.02 g/L and methionine 0.2 g/L. The plates were incubated at 37°C, 60 min at 110 rpm for initial cell adhesion. Then, the plates were washed with 2 mL of fresh medium, and 4 mL of fresh medium was added. Plates were then sealed with Breathseal sealing membranes (Greiner Bio-one) and incubated at 37°C, 110 rpm. After 48 h, supernatant culture was aspirated, the wells were gently washed with PBS and biofilm was collected in PBS.

Cell wall mutant strains were grown at 30°C, 180 rpm, in SD medium supplemented with arginine (100 µg/mL), uridine (100 µg/mL) and histidine (20 µg/mL). The phr1Δ/Δ mutant was grown in SD buffered medium at pH 7.5 (with 100 mM MOPS), and the phr2Δ/Δ mutant at pH 5.5 (with 100 mM MES). The conditional mutant PMRP1-CHS1/chs1Δ was first precultured in YNB with 2% maltose as a carbon source to allow the MRP1 promoter expression and then it was grown, as all other mutants, in SD (carbon source glucose), under which condition the promoter is repressed.

Cell wall fractionation

After growth culture, the fungal biomass was collected by centrifugation (5 min, 3300 g), washed and disrupted in 200 mM Tris-HCl pH 8 using a cell disruptor (FastPrep, MPbio, 6 cycles, 6 m/s, 60 sec) with 0.3 mm glass beads (Braun). The cell wall was collected by centrifugation (5 min, 3300 g), washed twice with 200 mM Tris-HCl, pH 8, then twice with milliQ water and freeze-dried. The wall polymers were then fractionated as previously described¹⁰⁴. Briefly, non-covalently bound proteins were extracted by the following treatment: dried cell wall was treated twice in 50 mM Tris-HCl, 50 mM EDTA, pH 7.5, 2% SDS, 40 mM β-mercaptoethanol in boiling water for 10 min (10-20 mg of cell wall/mL). After centrifugation (10 min, 3300 g), three volumes of ethanol were added to the supernatant and kept overnight at +4°C. The precipitate (SDS-β-ME fraction) was collected by centrifugation (10 min, 3300 g) and dialyzed against water before freeze-drying. Other cell wall pellet was washed three times with 150 mM NaCl and extracted twice with a 1 M NaOH containing 1 mg/mL BH4Na (Sigma), at 10-20 mg of cell wall/mL at 65°C for 1 h. After centrifugation (10 min, 3300 g), the alkali-soluble extract (AS fraction, supernatant) was neutralized with acetic acid, extensively dialyzed against water and freeze-dried. The alkali-insoluble fraction (AI fraction, pellet) was washed three times with distilled water and freeze-dried. Of note, the cell disruption was a key step to remove glycogen from cell wall extracts (Fig. S7).

Cell wall polymers quantification

Cell wall polymers of the three cell wall fractions (SDS-β-ME, AI and AS fractions) were quantified as previously described¹⁰⁴. Briefly, total neutral sugars were quantified by the phenol-sulfuric acid colorimetric assay using glucose as a standard¹⁰⁵. Proteins were quantified by colorimetry using the Pierce^TM BCA Protein Assay Kit (ThermoScientific) according to the manufacturer’s instructions. Monosaccharides, glucose and mannose, were identified and quantified by gas liquid chromatography (Perichrom) as their alditol acetates obtained after hydrolysis (4 N trifluoroacetic acid (TFA, Sigma), 100 °C, 4 h)¹⁰⁶. Glucosamine was quantified by high performance anion exchange chromatography (HPAEC, Dionex, ICS-3000) on a CarboPAC-PA1 column (ThermoScientific, 250 mm x 4.6 mm) after acid hydrolysis (6 M HCl, 6 h, 100 °C) using glucosamine as a standard. Samples were eluted at a flow rate of 1 mL/min with 18 mM NaOH for 15 min, then with a linear gradient to 300 mM sodium acetate (AcONa) in 100 mM NaOH for 20 min and finally under isocratic conditions with 300 mM AcONa in 100 mM NaOH for 5 min. The column was equilibrated for 20 min with the initial buffer before injection. Samples were detected on a pulsed electrochemical detector.

Quantification of β-1,6-glucan in alkali-insoluble (AI) and alkali-soluble (AS) fractions

To quantify β-1,6-glucan in the AI fraction, a colorimetric assay was used. Oxidation with periodate allows specific degradation of vicinal carbons bearing hydroxyl groups. In the AI fraction, only β-1,6- glucans were sensitive to this treatment, leading to the formation of aldehyde functions, which were quantified in the presence of 4-hydroxybenzhydrazide (PAHBAH) (Sigma) (Fig. S8). Oxidation was carried out at 50°C for 6 h on 100 µL of AI fraction (0.5 mg/mL) with 100 µL of 25 mM m-IO4Na (Sigma-Aldrich). Excess of reagent was quenched with 100 µL sodium sulfite 200 mM (Sigma). Then, 100 µL of reaction mixture were added to 900 µL of PAHBAH reagent (0.63 g sodium sulfite, 5 mL NaOH 5 M, 5 mL sodium citrate 0.5 M, 5 mL NaCl2 0.2 M qsp 100 mL milliQ water, 1 g PAHBAH) and placed in a boiling water bath for 10 min. The quantification of produced aldehyde was measured by the absorbance at 405 nm. Pustulan (linear β-1,6-glucan, Elicityl OligoTech) was used as a standard. The β-1,3-glucan content of the AI fraction was calculated as the total glucose content after subtraction of the amount of β-1,6-glucan.

The presence of mannan in the AS fraction prevents to use the colorimetric assay of β-1,6-glucan after periodate oxidation. β-1,3-glucan and β-1,6-glucan of the AS fraction were therefore quantified by enzyme digestion. The AS fraction was digested with the endo-β-1,3-glucanase, LamA (from Thermotoga neapolitana expressed in E. coli)¹⁰⁷ or with an endo-β-1,6-glucanase (from Schizosaccharomyces pombe expressed in Pichia pastoris)¹⁰⁸. Digestions were carried out at 37°C during 24 h by treating 0.2 mg of AS fraction with 20 µL of LamA (specific activity: 10 µmol eq/min/µL) or 60 µL of the endo-β-1,6-glucanase (specific activity: 3.03x10^-6 µmol eq/min/µL), in 50 mM NaOAc, pH 6, in a final volume of 200 µL. Both enzymatic treatments released more than 95% of glucan. Products were quantified by the amount of released reducing sugars using the PAHBAH reagent¹⁰⁹ and identified by HPAEC on a CarboPAC-PA1 column (ThermoScientific). Soluble and purified β- 1,6-glucans from C. albicans and a commercial source of laminarin (Sigma) were used as standards. Products from enzymatic digestion were eluted under the following gradient: isocratic step of 98% of eluent A (50 mM NaOH) and 2% of eluent B (500 mM AcONa in 50 mM NaOH) for 2 min, 2–15 min of a linear gradient of AcONa in 50 mM NaOH (2% B to 20% B), 15–20 min of a linear gradient (20% B to 43% B), 20–22 min of a linear gradient (43% B to 100% B), 22–25 min under isocratic conditions with 100% B. The column was equilibrated for 20 min with initial buffer before injection.

Structural characterization of β-1,6-glucan

β-1,6-glucan are branched by β-1,3 linked glucose or laminaribiose on the main chain^32,33. Branching was estimated by enzymatic digestion by an endo-β-1,6-glucanase as previously described³³. A 0.1 mg sample of the AI fraction was digested for 48 h at 37°C in a final volume of 100 µL with 30 µL of endo-β-1,6-glucanase in 50 mM sodium acetate, pH 6, 5 mM NaN3. Degradation products were analysed by HPAEC on a CarboPAC-PA1 column as described above. Branching was estimated by the ratio of branched oligosaccharides on total degradation products, to which a coefficient of correction was applied. This coefficient (43.1) was calculated from a branching rate of 6.9% of purified β-1,6-glucan established by NMR (see below and table S2). Branching of β-1,3-glucan was determined in the same way after digestion of the AI fraction with LamA.

The size of the β-1,6-glucan chain was estimated by gel filtration chromatography. First, β-1,6- glucans were released by incubation at 37°C for 48 h of the AI fraction (0.5 mg) with LamA (100 µL) in 70 mM sodium acetate, pH 6, 5 mM NaN3, in a final volume of 700 µL. After 3 min centrifugation at 12000 g, the supernatant was concentrated under vacuum (SpeedVac concentrator). Then, the sample was submitted to a gel filtration on a Superdex^TM 200 column (300 × 10 mm, Cytiva), eluted with ammonium acetate 150 mM, pH 4 at a flow rate of 0.2 mL/min. Samples were detected by a refractive index detector (Iota-2, Precision Instruments). The column was calibrated with dextran molecular weight standards (6, 10, 40, 70 and 500 kDa; GE Healthcare).

Nuclear magnetic resonance (NMR)

NMR experiments were recorded at 318.15 °K using a Bruker (Billerica, USA) 800 MHz Avance NEO spectrometer with an 18.8 Tesla magnetic field equipped with a cryogenically-cooled triple resonance (¹H,¹³C,¹⁵N) TCI probe. Spectra were recorded using TopSpin 4.2 (Bruker) and analyzed with CCPNMR Analysis 2.5.2¹¹⁰. ¹H and ¹³C chemical shifts were referenced to external DSS (2,2- dimethyl-2-silapentane-5-sulfonate, sodium salt). After three cycles of exchange (dissolution/lyophilization) against D2O, 5 mg of purified polysaccharide were dissolved in 550 µL of D2O (99.99% 2H, Eurisotop, Saclay, France) and placed in a 5 mm tube (Norell HT). Resonance assignment, glycosidic bonds identification and J coupling determination were achieved following standard procedures from homonuclear ¹H 1D, 2D COSY¹¹¹, NOESY¹¹² (150 ms mixing time) and natural abundance heteronuclear ¹H-¹³C experiments: edited HSQC^113,114, CLIP-HSQC¹¹⁵, H2BC¹¹⁶, HMBC¹¹⁷ and HSQC-TOCSY¹¹⁴ (100 ms mixing time). The anomeric configuration of monosaccharide residues was established from the corresponding chemical shifts and ¹JH1-C1 coupling constants obtained from CLIP-HSQC spectra. Glycosidic bonds were identified using HMBC experiments and confirmed with NOESY data. ³JH1-H2 coupling constants were measured from ¹H 1D spectra. The relative amount of monosaccharide residues was estimated from the integrals on ¹H 1D spectra obtained at 333.15 °K using a total recovery delay of 12 s.

Transmission electron microscopy

Cells were grown as described above and volumes of culture media equivalent to a quantity of cells equivalent to OD600nm=1 were collected and centrifuged 5 min at 3300 g. Cells were chemically fixed at room temperature in the dark for 1 h with 2.5% glutaraldehyde (Sigma) in fresh culture media (10 mL), washed in 1X PHEM buffer (60 mM PIPES, 25 mM HEPES, 10 mM EGTA, 2 mM MgCl2, pH 7.3) and centrifugated at low speed. The pellets were then resuspended in a droplet of 1X PHEM buffer and taken up into cellulose capillary tubes. For each strain, the capillaries were cut into 3 mm pieces, placed in 6 mm type A specimen carrier (side 0.2 mm) filled with hexadecen, covered by 6 mm type B carrier (side flat) and frozen with a high-pressure freezing machine (EM ICE, Leica microsystems) as previously described¹¹⁸. The samples were then freeze-substituted in a mix of 1% OsO4, 5% H2O in pure acetone for 24 h at −90°C, 12 h at 5°C/h at −30°C, 12 h at −30°C, 3 h at 10°C/h until 0°C and 1 h at 0°C. All samples were washed in pure acetone, and gradually infiltrated under agitation in low viscosity Spurr resin/acetone mix from 30% to 100%, as previously described¹¹⁹. Samples were then embedded in pure Spurr resin (EMS), followed by polymerization for 48 h at 60°C. Ultrathin sections (70 nm) were cut with a Leica Ultracut S microtome, and stained with 4% uranyl acetate and then 3% lead citrate. Transmission electron microscopy (TEM) images were captured with a Tecnai Spirit 120 kV TEM equipped with a bottom-mounted Eagle 4kx4k camera (FEI, USA). For statistical analysis, internal and external cell wall thickness have been measured (between 37-40 measurements, on 7-13 cells per condition) using TEM Imaging and Analysis software (TIA, FEI, USA).

Immunolabeling

Rabbit polyclonal antibodies were produced by ProteoGenix (Schiltigheim, France) using pustulan- conjugated BSA as the antigen. After 7-8 weeks of immunization, the resulting serum was used without antibodies purification. The specificity of serum was tested on pustulan, curdlan and cell wall AI fractions from Aspergillus fumigatus (which did not contain β-1,6-glucan) and C. albicans by ELISA (not shown). For immunolabelling of β-1,6-glucans and β-1,3-glucans in the cell wall of C. albicans, cells were grown as described above. Then, 1 mL of culture was fixed at ambient temperature in the dark (1-2 h) with paraformaldehyde 3.8% (PFA, Electron Microscopy Sciences). Cells were then centrifuged 3 min at 8000 g and washed 3 times with 500 µL PBS. In each step, cells were resuspended vigorously to separate the cells and disperse aggregates. Then, cells were resuspended vigorously in 500 µL of PBS and 5% goat serum (Sigma) and then 20 µL were added to each well of a slide (Diagnostic Microscope Slides, 8 well 6 mm, ThermoScientific) coated poly-L- lysine 0.1% and incubated 1 h at ambient temperature to permit cell adhesion. Liquid was removed and PBS 1X and 5% goat serum was added again and incubated during 1 h at ambient temperature. The liquid was aspirated and 20 µL of first antibody was added in PBS 1X and 5% goat serum overnight at 4°C (dilution 1/100 for polyclonal anti-β-1,6-glucan produced in rabbit, 1/250 for monoclonal anti-β-1,3-glucan (named 5H5) produced in mouse and kindly provided by N. Nifantiev¹²⁰). Wells were washed three times with PBS 1X and 5% goat serum. The second antibody was then added in PBS 1X and 5% goat serum 1h at ambient temperature (dilution 1/200, Alexa Fluor^TM 488 goat anti-mouse IgG (AF488, molecular probes) or fluorescein goat anti-rabbit IgG (FITC, Invitrogen)). Wells were washed 3 times with PBS and 5% (v/v) goat serum and once with PBS. Then, one drop of Fluoromount-G (Invitrogen) was added to each well, wells were covered with a coverslip N°1.5 (Menzel-Gläser), and polymerization was done during 48 h. Cells were observed on an EVOS FL microscope (Life Technologies), with a magnification x100 objective with oil immersion.

PBMC and neutrophil isolation, stimulation by parietal fractions and cytokine quantification

Human peripheral blood mononuclear cells (PBMCs) from heathy donors were isolated by a density- gradient separation of Ficoll 400 (Eurobio, France). Isolated PBMCs were re-suspended in RPMI 1640 + GlutaMAX^TM medium (Gibco), supplemented by 20% normal human sera (NHS, Normal Human Serum-Pooled, Zenbio) and seeded in each well (100 µL containing 4x10⁶ cells/mL) of 96- well microtiter plates (TPP). Cell wall fractions suspended in RPMI 1640 + GlutaMAX^TM medium were then added, at a final concentration of 25 µg/mL. LPS (Sigma, 0.1 µg/mL) was used as a positive control. After 24 h incubation at 37°C in a 5% CO2 chamber (Cytoperm 2, ThermoFisher Scientific), the culture supernatants were collected and stored at −20°C until further analysis.

Neutrophils from healthy donors were isolated with EasySep^TM Direct Human Neutrophil Isolation Kit (STEMCELL Technologies) according to manufacturer’s instructions. Isolated neutrophils were re- suspended in RPMI 1640 + GlutaMAX^TM medium (Gibco) and seeded in each well (100 µL containing 1x10⁶ cells/mL) of 24-well microtiter plates (TPP). 200 µL of NHS at 3% diluted in RPMI 1640 + GlutaMAX^TM medium were the added. Cell wall fractions suspended in RPMI 1640 + GlutaMAX^TM medium were then added, at a final concentration of 25 µg/mL. LPS (0.1 µg/mL) was used as a positive control. After 16 h incubation at 37°C in a 5% CO2 chamber (Cytoperm 2, ThermoFisher Scientific), the culture supernatants were collected and stored at −20°C until further analysis.

Global cytokines, chemokines, and acute phase proteins (total 36) present in the supernatants were detected with using Proteome Profiler Human Cytokines Array Kit (R&D Systems) according to manufacturer’s instructions. Following, the cytokines, chemokines and acute phase proteins of interest identified through protein profiler were quantified using DuoSet ELISA kits (R&D Systems) according to the manufacturer’s instructions.

Native AI and periodate oxidized AI (AI-OxP) fractions and purified β-1,6-glucans from C. albicans were used. AI-OxP fraction was generated by the treatment of 40 mg of AI fraction with 100 mM m- IO4Na (4 mL), 3 days at 4°C. The reaction was stopped by addition of 200 µL of glycerol, oxidized fraction was washed by centrifugation with water, reduced with BH4Na (10 mg/mL in 0.1 mM NaOH, overnight) then submitted to mild acid hydrolysis (AcOH 10% 100°C, 1 h) and finally washed with water and freeze-dried. Purified β-1,6-glucans were produced by digesting the AI fraction (40 mg) with LamA as described above. LamA was removed by passing the sample through a C-18 SPE cartridge (Sep-Pak classic, Waters) eluted with 5% acetonitrile and 0.1% TFA. Then, samples were purified by gel filtration chromatography (Superdex 200, 60 cm x 16 cm, Cytiva) in ammonium acetate 150 mM, pH 4 at a flow rate of 0.5 mL/min. Samples were detected by a refractive index detector (Iota2, Precision Instruments) and dialyzed against water and then freeze-dried. The fractions were quantified by the phenol-sulfuric acid colorimetric assay as previously described.

Complement activation

Complement activation was done as previously described¹²¹. Briefly, cell wall fraction from C. albicans: AI, AI-OxP and purified β-1,6-glucans were coated on 96-well microtiter plates (100 µL/well) at three different concentrations 50 µg, 25 µg and 12.5 µg per well in 50 mM bicarbonate buffer, pH 9.6, overnight at ambient temperature. Then, supernatants were discarded, and wells were blocked with PBS 1X/BSA 1% for 1h at ambient temperature. Then 8 µL of NHS and 92 µL of Gelatin veronal buffer (GVB, 5 mM barbital, 145 mM NaCl, 0.1% gelatin, pH 7.4) supplemented by MgCl2 (0.5 mM) and CaCl2 (0.15 mM) were added, incubated for 1 h, and then washed three times with PBS 1X/Tween-20 0.05%. Complement activation was measured by quantifying deposited C3b upon adding the wells with mouse monoclonal anti-human C3b antibody (MA1-82814, ThermoFisher Scientific, diluted 1:1,000) diluted in PBS 1X/BSA 1% and incubating 1 h at ambient temperature, washing the wells with PBS 1X/Tween 20 0.05%, adding the wells with peroxidase-conjugated secondary anti-mouse IgG antibodies (Sigma-Aldrich, diluted 1:1,000) and incubating for 1h at ambient temperature. After washing, 100 µL of substrate solution (TMB, BioFX; 100 µL/well) was added and the reaction was stopped with 50 µL of 4% H2SO4. The absorbance was read at 450 nm after subtraction of reference wavelength at 540 nm using microplate reader (Infinite m200 pro, TECAN).

Generation of mutants by transient CRISPR/Cas9

Homozygous null mutants (kre6Δ/Δ; skn1Δ/Δ; kre62Δ/Δ ; skn2Δ/Δ ; kre6/skn1Δ/Δ; kre6/skn1/kre62/skn2Δ/Δ) were generated in C. albicans SC5314 using established transient CRISPR-Cas9 methods^122,123. A sgRNA cassette was constructed by PCR, first by synthesising the SNR52 promoter (using primers SNR52/F and SNR52/R/GOI) and the sgRNA scaffold (with the primers sgRNA/F/GOI and sgRNA/R), using plasmid pV1093 as a template¹²². These two products were fused by PCR, followed by nested PCR with primers SNR52/N and sgRNA/N. The CaCAS9 cassette was amplified from the plasmid pV1093 with primers CaCas9/F and CaCas9/R. Repair templates (RT), which contained the SAT1-Flipper marker and harboured 80 bp homology to the 5’ and 3’ ends of the target gene, were amplified from pSFS2A¹²⁴. The RT containing the hygromycin resistance gene HygB was amplified from pCrispr-gRNA1/hygR, derived from plasmid pV1090¹²² as follows: a BspHI fragment carrying the HygB marker was cut from pYM70¹²⁵, the ends filled in with the Klenow enzyme and the fragment ligated into pV1090 digested with StuI and BglII and treated with the Klenow enzyme. PCR reactions to amplify sgRNA and CaCas9 were carried out using Phusion High-Fidelity DNA polymerase (NEB), and Q5 polymerase (NEB) was used for RT amplification, according to the manufacturer’s instructions. All PCR products were cleaned-up and concentrated with NucleoSpin Gel and PCR Clean-up (Macherey-Nagel). PCR products (3 µg of CaCAS9 cassette, 1 µg of sgRNA cassette and 3 µg of the relevant RT) were transformed into C. albicans SC5314 using the lithium acetate method¹²⁶. Transformants were selected on YPD containing 150 µg/mL nourseothricin (Jena Bioscience) or YPD with 600 µg/mL hygromycin (Sigma). Disruption of both alleles of the target locus and integration of the RT were confirmed by PCR after DNA extraction (MasterPure Yeast DNA Purification Kit, Lucigen). The SAT1 gene was removed with the flipper system. Briefly, the transformants were cultivated overnight in 3 mL YP with 2% maltose at 30°C. Then, the culture was diluted to 250 cells/mL in sterile water and 200 µL were plated on YPD with 25 µg/mL nourseothricin and incubated 1-2 days at 30°C. The smaller colonies were replicated on YPD vs. YPD + 150 µg/mL nourseothricin. Transformants that grew only YPD were checked by PCR using primers Flanking_F/GOI and Flanking_R/GOI. The primers used for the construction and PCR check of these mutants are described in Table S3. Deletions were validated by PCR (Fig. S9).

The quadruple mutant kre6/skn1/kre62/skn2Δ/Δ was complemented by integrating a StuI-linearized plasmid bearing the KRE6 gene under the control of the constitutive promoter ACT1 at the RPS1 locus. This plasmid was obtained using the Gateway^TM technology (Invitrogen^TM) as described previously¹²⁷. KRE6 was PCR amplified on SC5314 genomic DNA with primers KRE6-FWD and KRE6-REV; the fragment was cloned in pDONR207¹²⁸ with the Invitrogen^TM Gateway BP clonase, and then transferred to the destination vector CIp10-PACT1-SAT1¹²⁹ using the Invitrogen^TM Gateway LR clonase. Successful cloning into the destination vector was confirmed by sequencing. Transformants were selected on YPD+100 µg/mL nourseothricin. Plasmid integration at the RPS1 locus was checked by colony PCR using CIpUL and CIpSAT.

Phenotypic analyses

Kinetic curves were generated in 96 well plate (TPP) in SD medium, 30°C, at initial OD620 nm= 0.01 using a Tecan Sunrise absorbance microplate reader. Optical density was measured every 10 min during 80 h with Magellan^TM Software. Doubling time was calculated from growth curves with GraphPad Prism software by generating a nonlinear regression (logistic curve) to determine the k coefficient, and then calculated as doubling time = ln(2)/k.

Spot test drug sensitivity assays were performed on complete SD medium agar plates containing either Congo Red (50 µg/mL), Calcofluor White (50 µg/mL), tunicamycin (1 µg/mL), nikkomycin (32 µg/mL) or caspofungin (0.5 µg/mL). Cell were precultured for 24 h in SD medium liquid, 30°C, and then diluted to OD600 nm=1 in sterile water, and then 10-fold serially diluted. Samples of 5 µL of each concentration were spotted onto agar plates. Pictures of plates were taken after 24 h, 48 h, 72 h and 6 days with a Phenobooth+ Colony counter (Singer Instruments). To observe macroscopic filamentation, 5 µL (OD600 nm=1) were spotted on YNB 2%GlcNAc, buffered at pH 7.2 with 100 mM MOPS agar, at 37°C, for 6 days. Pictures were taken using an I-phone camera.

To observe hyphal growth, cells were grown overnight in SD liquid 30°C, then diluted to OD600 nm=0.2 in 1 mL of 24 well plate (TPP) and growth for 6 h either in SD, 30°C (control, no-filamentation) or in YNB 2%GlcNAc, buffed at pH 7.2 with 100 mM MOPS, at 37°C (inducing filamentation). Cells were then observed with an Olympus IX 83 microscope and images were captured with a Hamamatsu ORCA Flash cooled CCD camera, using the cellSens imaging software.

Statistical analyses

GraphPad Prism 10 software was used for statistical analyses. Data were generated from at least three independent biological replicates and then expressed as means ± standard deviation. Specific tests are indicated in the figure legends. Normally distributed datasets (QQ plot or Shapiro-Wilk test), were compared with ordinary one-way ANOVA (Dunnett’s or Tukey’s multiple comparisons test) or two-tailed unpaired t-test. For immunology data, measurements were paired/dependant as the same donors were exposed to different cell wall fractions, thus we applied the nonparametric test for non- normally distributed data: Friedman test (with Dunn’s multiple comparisons test). P-values (P) < 0.05 were considered as significant and represented on graphs as followed: *, P < 0,05; **, P < 0,01; ***, P < 0,001; ****, P < 0,0001; ns : non-significant.

Acknowledgements

C.B is the recipient of a Ph.D. fellowship from the Laboratoire d’Excellence Integrative Biology of Emerging Infectious Diseases (ANR-10-LABX-62-IBEID). We acknowledge support from the French Government’s Investissement d’Avenir program (Laboratoire d’Excellence Integrative Biology of Emerging Infectious Diseases [ANR-10-LABX-62-IBEID]) and Institut Pasteur from Paris. V.A. and T.F. are supported by the Agence Nationale de la Recherche ANR-21-CE17-0032-01, FUNPOLYVAC grant. We thank Romain Laurian and Natacha Sertour for their advice for molecular biology, Cécile Gautier for her advice on Prism software and spotting assays, Catherine Comte, Lucia Oreus and Jamal Boutchlhit for material availability and Matt Edmondson for providing C. albicans strains. The authors thanks Suzan Noble, Mathias Richard and Neil Gow for the access to their deletion mutant collections. NG acknowledges support of Wellcome Trust Investigator, Collaborative, Equipment, Strategic and Biomedical Resource awards (101873, 200208, 215599, 224323). NG also thanks the MRC (MR/M026663/2) and the MRC Centre for Medical Mycology (MR/N006364/2) for support. This study/research is funded by the National Institute for Health and Care Research (NIHR) Exeter Biomedical Research Centre (BRC). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. The 800-MHz NMR spectrometer of the Institut Pasteur was partially funded by the Région Ile de France (SESAME 2014 NMRCHR grant no 4014526). CS is grateful for support for equipment from the French Government Programme Investissements d’Avenir France BioImaging (FBI, N° ANR-10- INSB-04-01).

Additional information

Author contributions

C.B., V.A., S.B.B, N.G. and T.F. designed experiments and analyzed data. C.B. performed most of the experiments. I.V. and M.C. constructed strains. C.S. and T.M. designed and performed electron microscopy experiments. I.G. performed NMR experiments and analysis. NG provided essential strains and materials. C.B., V.A., C.dE., N.G. and T.F wrote the manuscript. All authors have read and agreed to the published version of the manuscript.

Competing interests

The authors declare no competing interests.

Supplementary figures

Figure S1:

Figure S2:

Figure S3:

Figure S4:

Figure S5:

Figure S6:

Figure S7:

Figure S8:

Figure S9:

Figure S10:

Figure S11:

Figure S12:

Figure S13:

Figure S14:

Figure S15:

Table S1:

Table S2:

Table S3: