Mannan detecting C-type lectin receptor probes recognise immune epitopes with diverse chemical, spatial and phylogenetic heterogeneity in fungal cell walls

During the course of fungal infection, pathogen recognition by the innate immune system is critical to initiate efficient protective immune responses. The primary event that triggers immune responses is the binding of Pattern Recognition Receptors (PRRs), which are expressed at the surface of host immune cells, to Pathogen-Associated Molecular Patterns (PAMPs) located predominantly in the fungal cell wall. Most fungi have mannosylated PAMPs in their cell walls and these are recognized by a range of C-type lectin receptors (CTLs). However, the precise spatial distribution of the ligands that induce immune responses within the cell walls of fungi are not well defined. We used recombinant IgG Fc-CTLs fusions of three murine mannan detecting CTLs, including dectin-2, the mannose receptor (MR) carbohydrate recognition domains (CRDs) 4–7 (CRD4-7), and human DC-SIGN (hDC-SIGN) and of the β-1,3 glucan-binding lectin dectin-1 to map PRR ligands in the fungal cell wall of fungi grown in vitro in rich and minimal media. We show that epitopes of mannan-specific CTL receptors can be clustered or diffuse, superficial or buried in the inner cell wall. We demonstrate that PRR ligands do not correlate well with phylogenetic relationships between fungi, and that Fc-lectin binding discriminated between mannosides expressed on different cell morphologies of the same fungus. We also demonstrate CTL epitope differentiation during different phases of the growth cycle of Candida albicans and that MR and DC-SIGN labelled outer chain N-mannans whilst dectin-2 labelled core N-mannans displayed deeper in the cell wall. These immune receptor maps of fungal walls of in vitro grown cells therefore reveal remarkable spatial, temporal and chemical diversity, indicating that the triggering of immune recognition events originates from multiple physical origins at the fungal cell surface.

remarkable spatial, temporal and chemical diversity, indicating that the triggering of immune 48 recognition events originates from multiple physical origins at the fungal cell surface. 49

Author Summary 50
Invasive fungal infections remain an important health problem in immunocompromised 51 patients. Immune recognition of fungal pathogens involves binding of specific cell wall 52 components by pathogen recognition receptors (PRRs) and subsequent activation of immune 53 defences. Some cell wall components are conserved among fungal species while other 54 components are species-specific and phenotypically diverse. The fungal cell wall is dynamic 55

Introduction 69
Fungi are associated with a wide spectrum of diseases ranging from superficial skin and 70 mucosal surface infections in immunocompetent people, to life-threatening systemic infections 71 in immunocompromised patients [1,2]. The global burden of fungal infections has increased 72 due to infection related or medically imposed immunosuppression, the use of broad-spectrum 73 antibiotics that suppress bacterial competitors, and the use of prosthetic devices and 74 intravenous catheters in medical treatments [3,4]. Patients that are pre-disposed to fungal 75 diseases include those with neutropenia, those undergoing stem cell or organ transplant 76 surgery or recovering from surgical trauma as well as HIV infected individuals and those with 77 certain rare predisposing mutations in immune recognition pathways [3][4][5][6]. 78 Innate immunity is the primary defence mechanism against fungal infections and involves host 79 Pattern Recognition Receptors (PRRs) that recognise specific Pathogen-Associated 80 Molecular Patterns (PAMPs), which are mostly located within the cell wall [7][8][9]. These 81 profiles to different Candida species and C. albicans strains with no clear association between 161 Fc-lectin binding, phylogenetic relatedness and relative virulence. 162

Differential expression of ligands for C-type lectin receptors during growth and 163 morphogenesis 164
We next tested the stability of the epitopes recognised by CTL-Fc-lectins during the growth 165 and morphogenesis of C. albicans. Samples in the lag phase, early, mid and late exponential 166 as well as stationary phases were sampled during batch growth (Fig. 3 A). During the period 167 of exponential growth of yeast cells, dectin-2-Fc ligand exposure was somewhat reduced (Fig.  168 3 B) whilst β-glucan for dectin-1-Fc was increased (Fig. 3 B). In contrast, epitopes for mannose 169 receptor assessed by CRD4-7-Fc binding appeared to be exposed throughout all growth 170 phases of batch growth (Fig. 3 B). Therefore PAMP binding was affected by growth phase and 171 potentially growth rate of the target pathogen. 172

C. albicans filamentation was induced and hyphal cells were fixed at different time points to 173
test Fc-lectin binding (Fig. 4). Mannan-recognising dectin-2-Fc and CRD4-7-Fc demonstrated 174 strong binding to early germ tubes grown in serum-containing medium ( Fig. 4 A, B). However, 175 binding of both mannan-detecting lectin probes gradually decreased over prolonged periods 176 of hyphal growth ( Fig. 4 A, B). In particular, decreasing binding of CRD4-7-Fc to the mother 177 yeast cell was observed and was virtually absent on germ tubes that were older than 2 h ( Fig.  178 4 B). In contrast, although germ tubes lacked bud scars, which have exposed β-1,3 glucan, 179 dectin-1-Fc demonstrated the opposite pattern with binding gradually increasing to the lateral 180 cell walls of maturing filamentous cells (Fig. 4 C). These results reinforce previous 181 observations that nascent mannan epitopes are gradually modified as the yeast cells and 182 hyphae progress through different growth stages [61]. 183 The binding of mannan-detecting C-type lectins to yeast, pseudohypha, hypha and the 184 recently described goliath cells [62] of C. albicans was examined (Fig. 5). Dectin-2-Fc 185 demonstrated low binding affinity to C. albicans fixed yeast cells which could be detected by 186 flow cytometry ( Fig. 1-3) but not by microscopy ( Fig. 5 A). Fixation was used to capture and 187 immobilise cells at specific morphogenetic stages, but control experiments showed that 188 paraformaldehyde fixation did not influence Fc-lectin binding patterns (data not shown). 189 Dectin-2-Fc exhibited punctate binding pattern on hyphae with strong staining observed at the 190 hyphal tip (Fig. 5 A). In contrast, CRD4-7-Fc demonstrated high intensity punctate binding to 191 both yeast cells and hyphae (Fig. 5 B). As predicted, dectin-1-Fc recognised yeast cells mainly 192 at the bud scars while some punctate binding was also detected along hyphae (Fig. 5 C) (Fig. 5). As before, dectin-1-Fc bound mainly to the bud scars of goliath cells (Fig. 5). 198 Negative control protein CR-Fc did not show any binding to any of the C. albicans 199 morphologies ( Fig. 5 D), and binding to yeast cells was not detected by flow-cytometric 200 analyses. These data demonstrated differences in the specificities of dectin-2-Fc and CRD4-201 7-Fc towards the fungal cell surface components, and were in accord with knowledge of the 202 glycan-binding specificities of dectin-1 [64] and of . 203

Spatial distribution of mannan epitopes in the inner cell wall 204
To elucidate precise localisation of ligands for mannan-recognising Fc-lectins within the cell 205 wall, immunogold labelling of dectin-2-Fc, CRD4-7-Fc and CR-Fc-stained embedded sections 206 of cells was analysed by TEM (Fig. 6). We observed clustered dectin-2-Fc binding to both 207 yeast and hyphae inner cell walls of C. albicans with little labelling of the outer mannoprotein-208 rich fibrils (Fig. 6 A). CRD4-7-Fc recognised ligands within the plasma membrane as well as 209 outer glycoprotein fibrils (Fig. 6 B). CR-Fc gave no staining (Fig. 6 C). The differential 210 specificities of dectin-2-Fc, CRD4-7-Fc and CR-Fc recognition were again demonstrated. 211 Flow cytometry and microscopy were used to compare binding strengths of Fc-lectins to 212 C. albicans yeast cell wall after mild heat treatment (at 65°C) which mechanically perturbs the 213 C. albicans (Fig. 8 D) that was similar to that observed for CRD4-7-Fc labelling. The binding 241 to hDC-SIGN-Fc epitopes progressively decreased in maturing, more elongated C. albicans 242 hyphae (Fig. 8 E). Reminiscent of CRD4-7-Fc binding, hDC-SIGN-Fc demonstrated high 243 intensity binding to C. albicans cells in different morphologies with binding sites distributed 244 along the plane of the plasma membrane and in the outer cell wall glycosylated fibril layer. 245 Chemical nature of C-type lectin receptor targets in fungal cell walls 246 To assess the nature of the targets of the CTL receptors, the binding of the Fc-lectin probes 247 was examined using a collection of cell wall mutants including isogenic nulls with truncations 248 compared to wild type ( Fig. 9 A). This is compatible with previous observations that dectin-2-264 Fc recognised inner cell wall mannans whilst CRD4-7-Fc labelled the outer mannoprotein 265 fibrillar layer (Fig. 6 A, B). Dectin-1-Fc was used as a control and demonstrated increased 266 binding in all backgrounds deficient in N-and O-mannan attributable to the increased exposure 267 of inner cell wall component β-glucan (Fig. 9 A). Therefore, the cell wall glycosylation status 268 had a major impact on the ability of CTL probes to bind C. albicans. 269 To gain further insight into the carbohydrate recognition by the CTL receptors, we analysed 270 these initially using a microarray comprised mostly of fungal-type saccharides and compared 271 their binding profiles (Fig. 10, S Table 2). Both CRD4-7-Fc and hDC-SIGN-Fc showed strong 272 binding to the C. albicans N-mannoprotein that is characterised by an α-1,6-mannose 273 backbone with oligomeric α-1,2-, α-1,3-, and β-1,2-manno-oligosaccharide branches ( Fig. 10  274 A, B, position 13, S Table 2). The two proteins bound also to other mannan-related 275 saccharides from S. cerevisiae and M. tuberculosis, that share an α1,6-mannose backbone 276 with α-1,2-manno-oligosaccharide branches ( Fig. 10 A, B, S Table 2). In contrast, no binding 277 was detected with dectin-2-Fc to any of Manα-1,2-Man-containing polysaccharides in the 278 conditions of the microarray analysis, which suggests it may have less capacity to bind -279 mannans of this type compared to CRD4-7-Fc and hDC-SIGN-Fc (data not shown). Dectin-1-280 Fc showed, as predicted, strong and highly specific binding to glucans with a β-1,3-glucosyl 281 backbone ( Fig. 10 C, S Table 2). 282 Glycan microarrays of 474 sequence-defined oligosaccharide probes (S Table 3 B) were also 283 applied to compare the binding specificities of dectin-2-Fc, CRD4-7-Fc and hDC-SIGN-Fc. 284 The signal intensities observed with dectin-2-Fc were the lowest overall among the three Fc 285 probes. Dectin-2-Fc binding was detected to Man 9 GN 2 derived probes that resemble the core 286 N-mannan structures within the inner C. albicans cell wall ( Fig. 11 A, S Table 3 A); this 287 relatively weak and restricted binding is in agreement with previous glycan array studies [37, 288 69]. No binding of dectin-2-Fc was detected to oligo-mannose sequences smaller than 289 Man 7 GN 2 . Additionally, binding was detected of dectin-2-Fc to 3'sialyl LNFPIII and a number 290 sulphated glycans as with hDC-SIGN-Fc. 291 CRD4-7-Fc showed mannose-related binding of high intensity to oligo/high-mannose N-glycan 292 sequences, fucosylated probes including Fuc-GlcNAc and Man 3 FGN 2 , as well as β-1,4-293 oligomannoses ( Fig. 11 B, S Table 3 A). hDC-SIGN-Fc gave strong binding signals with N-acetylglucosamine containing sequences including chitin-derived glycans ( Fig. 11 C, S Table  295 3 A) and also to glucan oligosaccharide sequences with differing glucosyl linkages as was 296 also observed in recent studies [70,71] (Fig. 11 C, S Table 3 A)]. hDC-SIGN-Fc also gave 297 binding to a broad range of N-glycans including oligo/high-mannose sequences having α-1,2-298 , α-1,6-and α-1,3/1,6-linked mannose, high-mannose sequences capped by Glc residues and 299 a number of N-acetylglucosamine-terminating N-glycans; binding was also detected to β4-300 linked mannose oligosaccharides as with CRD4-7-Fc. Collectively, glycan array binding 301 results are consistent with the dectin-2-Fc ligand being based on Man 9 GN 2 found in the core 302 N-mannan triantennary structure within the C. albicans inner cell wall, whereas CRD4-7-Fc 303 and hDC-SIGN-Fc have a broader binding profile compared to dectin-2-Fc and can recognise 304 oligo-mannose structures terminating in α-1,3 and α-1,6-mannose resembling to some extent 305 those found in the outer-chain mannan structures. This is compatible with the binding patterns 306 observed using colloidal-gold TEM. 307 308 Discussion 309 CTL receptors provide a first line defence against fungal pathogens and orchestrate both 310 innate and adaptive immunity through the recognition of fungal PAMPs. A large number of 311 CTL receptors have been proposed to bind fungal cell wall epitopes such as mannans, β-1,3-312 glucan and chitin [34,37,47,50,52]. Previous studies suggested dectin-2 to recognise high 313 mannose residues that are present on fungal cell surfaces while MR was proposed to bind 314 branched N-mannan structures, with fucose, N-acetylglucosamine sugar residues and 315 mannose-capped lipoarabinomannan (ManLAM) on the mycobacterial cell wall [34,37,72]. 316 hDC-SIGN has been demonstrated to recognise galactomannan, mannose-and fucose-317 containing glycoconjugates and N-linked mannose rich components in C. albicans cell wall 318 [38,73]. There is however limited knowledge of the structural arrangement, spatial distribution 319 and variation in mannoside architecture of the cell wall of fungal pathogens, which this study 320 addresses. We deployed recombinant CTL-Fc proteins as probes to explore the distribution, 321 regulation and chemical structure of fungal ligands for these receptors within the cell wall, in 322 particular for dectin-2, MR and hDC-SIGN. We demonstrated that there is considerable intra-323 and inter-species variability in the expression of key mannan epitopes. We mapped the 324 patterns of binding to these ligands during growth and cellular morphogenesis and observed 325 marked spatial segregation and an unexpected clustering of certain mannan epitopes. 326

327
The analysis of C. albicans cell wall mutants corroborated the occurrence of inner cell wall 328 epitopes for dectin-2-Fc and superficial mannan ligands for MR and hDC-SIGN-Fc. An och1Δ 329 mutant, which has a defect in its ability to synthesise outer chain N-glycans, had been shown 330 previously to have exposed inner cell wall components [67]. This mutant also exhibited 331 increased binding by dectin-2-Fc and has been shown to have an elaborated α-1,2-and α-332 1,3-mannan side chains to the Man 8 GlcNAc 2 core triantennary complex which is a ligand that 333 promotes CRD4-7-Fc and hDC-SIGN-Fc binding. Other mutants with decreased N-mannan 334 outer chains, including mnn2-26Δ, pmr1Δ [66,68] bound less CRD4-7-Fc and hDC-SIGN-Fc 335 and more dectin-2-Fc commensurate with an increased exposure of the inner wall layers. 336 Moreover, the mnn4Δ mutant, which lacks cell wall phosphomannan that confers a negative 337 charge on the wall [65] had reduced binding by CRD4-7-Fc and hDC-SIGN-Fc but increased 338 dectin-2-Fc affinity. These observations complement previous studies suggesting that cell 339 surface glycosylation profoundly influences the efficiency of pathogen recognition by immune 340 receptors [9,11,13,23,74]. Glycan microarray data with the Fc-lectins and the fungal-type 341 saccharides are consistent with CRD4-7-Fc and hDC-SIGN-Fc but not dectin-2-Fc recognizing 342 α-1,6-Man backbone with oligomeric α-1,2-, α-1,3-, and β-1,2-Man branches which comprise 343 C. albicans outer wall N-mannan branches. This contrasts with the weak binding detected of 344 dectin-2-Fc to the high-mannose Man 9 GN 2 probe with terminal Manα-1,2-Man sequences, 345 which resembles C. albicans core N-mannan within the inner cell wall. This is in agreement 346 with published data that showed the Manα-1,2-Man sequence on the Man 9 GN 2 to be a primary 347 target for dectin-2 binding [37,69], and that this receptor could adopt a geometry of the binding 348 and Manα-1,2-Manα-1,3-Man-(D2 branch) trisaccharide sequences [37,69]. This prompts a 350 hypothesis that dectin-2 binds to internal sequences in fungal mannan polysaccharides, 351 expressing higher density of the ligands and achieving multivalent binding. binding to hyphae was found to vary during hypha elongation. It is possible that the efficiency 369 by which invading hyphae induce or escape immune surveillance ultimately influences 370

virulence. 371
We also demonstrated that C. albicans growth phase influenced expression and exposure of 372 cell wall components. Batch growth of C. albicans yeast culture revealed increased dectin-1-373 Fc binding during the exponential growth phase. This is likely to be due to increased β-glucan 374 exposure and number of bud scars on actively dividing cells. As batch culture of cells 375 transitioned into the stationary phase, a reduction in dectin-1-Fc and increased dectin-2-Fc 376 binding was observed that was likely to be related both to the degree of mannan shielding of 377 β-glucan as the cell wall was reorganised. CRD4-7-Fc demonstrated similar binding patterns 378 during all growth phases of the yeast batch culture, highlighting the diversity in availability of 379 the cell wall manno-oligosaccharide sequences and suggesting that the CRD4-7-Fc (mannose 380 receptor) ligand is superficial in the cell wall. This suggestion was supported by colloidal gold-381 TEM images showing CTL binding patterns at the ultrastructural level, and is in accord with a 382 previous study which demonstrated that MR was not required for host defence in a systemic 383 candidiasis mouse model [82]. 384 The hyphal cell wall is also modified during the process of hyphal extension [17,61]. In this 385 study we mapped the distribution of CTL epitopes on hyphal cell walls over a period of We observe that dectin-2-Fc weakly recognises C. albicans yeast cells but has increased 416 binding on goliath cells (Fig. 4 A), potentially due to cell expansion and concomitant exposure 417 of inner cell wall epitopes or a novel arrangement of mannoproteins on the surfaces of goliath 418 cells that may expose more dectin-2-recognising mannan epitopes. Dectin-2-Fc also bound 419 well to hyphae of C. albicans, in particular at the hyphal tip (Fig. 4 A). The cell wall of the 420 hyphal apex is thinner and the polysaccharides are less crystalline and less cross-linked. This 421 may facilitate access of high molecular weight PRRs to inner cell wall ligands. Both mannose 422 receptor (CRD4-7-Fc) and hDC-SIGN-Fc probes demonstrated similar binding to different 423 morphologies and their binding was not greatly affected by mild heat treatments suggesting 424 their epitopes are superficial. As predicted, dectin-1-Fc binding was concentrated at the bud 425 scars of yeast and goliath cells, but binding was punctate on hyphal cells that lack bud scars. 426 Punctate binding patterns have been reported elsewhere, in particular for dectin-1 binding 427 revealed by super-resolution microscopy in which binding became increasingly associated 428 with highly granular multi-glucan surface exposures in response to caspofungin treatments 429 [87]. 430 Using multiple approaches, we provide evidence that individual mannan-recognising CTLs 431 recognise different mannan structures that are displayed in distinct binding patterns. CRD4-7-432 Fc (MR) and hDC-SIGN-Fc recognised α-1,6-mannose backbone with oligomeric α-1,2, α-1,3, 433 and β-1,2-mannan branches while dectin-2-Fc bound Man 9 GN 2 , which has close similarities 434 to the core N-mannan structure in C. albicans inner cell wall. Heat-killing and 435 immunofluorescent staining of C. albicans cell wall mutants as well as immunogold-labelling 436 and TEM microscopy supported the conclusion that MR and DC-SIGN recognise outer chain Membranes of eukaryotic cells are known to organise some proteins into specialised 447 microdomains which compartmentalise cellular processes and serve as organising centres 448 which assemble signalling molecules, facilitate protein and receptor trafficking, vesicular 449 transport and signalling events [88,89]. Such microdomains have also been described in 450 bacterial membranes and a recently published study suggested that similar microdomains 451 might also exist in fungal cell walls [90]. Therefore, clustered dectin-2-Fc binding could be 452 related to the accumulation of specialised glycosylated proteins at certain parts of the inner 453 cell wall. Nevertheless, these attractive hypothesis should be addressed by future studies. 454 In conclusion, we have demonstrated that mannan epitopes are differentially distributed in the 455 inner and outer layers of fungal cell wall in a clustered or diffuse manner. Immune reactivity of 456 fungal cell surfaces was not correlated with relatedness of the fungal species, and mannan-457 detecting receptor-probes discriminated between cell surface components generated by the 458 same fungus growing under different conditions. Moreover, individual mannan-recognising 459 CTL probes conferred specificity for distinct mannan epitopes. These findings indicate that the 460 fungal cell wall structures are highly structured but dynamic, and that immune recognition is Tween 20 (0.05 %) and 100 µ of TMB (Thermofisher) were added to develop. Reaction was 480 stopped with 50 µl 2N H 2 SO 4 and plates were assayed on a spectrophotometer at 450 nm with 481 the necessary λ correction at 570 nm . Fc chimeric proteins were purified via affinity based Fast 482 Protein Liquid chromatography using Prosep® Ultra resin (Millipore). Fc conjugated proteins 483 were eluted with 0.1 M glycine (pH 2.5) before neutralisation with 1 M Tris buffer (pH 8) and 484 then dialysed in 1 X PBS overnight. Fc conjugated protein concentration was quantified using 485 NanoVue Spectrophotometer (GE Healthcare). 486

QC of purified Fc-lectins 487
Purified proteins were checked via SDS-PAGE gel analysis using 4-12 % Bis-Tris SDS-PAGE 488 gels under reducing conditions (S Fig. 1). ELISA was carried out for confirmation of binding to 489 original target using ELISA protocol described above (S Fig. 1). ELISA plates were coated 490 with live C. albicans yeast cells, 25 µg/ml S. cerevisiae mannan (SIGMA), 100 µg/ml 491 C. albicans yeast β-glucan or PBS overnight (S Fig. 1). Fc chimeric proteins were added at 5 492 µg/ml and serial doubling dilutions were performed to confirm concentration-dependent 493 binding. 494

Comparison of fungal strains under different parameters 495
For comparison of fixed and heat-killed C. albicans (CAI4-CIp10) cells, a single colony was 496 inoculated into 10 ml YPD (1% yeast extract, 2% glucose, 2% peptone) and incubated 497 overnight at 30°C, 200 rpm. Overnight culture was washed in 1 X PBS and 2.5 x 10 6 cells 498 were either fixed with 4 % paraformaldehyde or kept at 65°C in a heat block for 2 h prior to 499 staining. For comparison of C. albicans (CAI4-CIp10) culture overtime, OD 600 of overnight 500 culture was measured and culture was diluted to OD 600 of 0.1 in 50 ml YPD in 250 ml flasks. 501 Cells were collected at OD 600 0.2, 0.4, 0.6, 1 and 18, washed in 1 X PBS, fixed with 4 % 502 paraformaldehyde at RT for 45 min, washed and then stained. For comparison of different 503 C. albicans isolates and cell wall mutants, cells were fixed at OD 600 ~ 0.5 prior to staining (S 504 Table 1). For comparison of different Candida species and S. cerevisiae, cells were fixed in 505 stationary phase, OD 600 ~ 18 (S Table 1). Samples were stained as described below and 506 analysed on BD Fortessa flow cytometer or in 3D on an UltraVIEW® VoX spinning disk 507 confocal microscope. Three independent biological replicates were performed per sample. 508

Conditions for generating different morphologies of C. albicans 509
Single colonies of C. albicans were inoculated into 10 ml YPD and incubated overnight at 510 30°C, 200 rpm. To induce hypha formation, cultures were diluted 1:1333 in milliQ water and 511 then adhered on a poly-L-lysine coated glass slide (Thermo Scientific, Menzel-Gläser) for 30 512 min prior to incubation in pre-warmed RPMI + 10 % FCS at 37°C for 45 min-3 h 15 min 513 depending on the tested parameter. Slides were than washed in DPBS and fixed with 4 % 514 paraformaldehyde. C. albicans pseudohyphae were produced using published conditions with 515 modifications [91]. Overnight culture was collected by centrifugation, washed twice with 0.15 516 M NaCl, resuspended in 0.15 M NaCl and incubated at room temperature for 24 h to induce 517 starvation. After 24 h, cells were transferred into RPMI 1640 at a concentration of 1x10 6 and 518 incubated at 30°C 200rpm for 6 h prior to fixation with 4 % paraformaldehyde. Fixed cells were 519 stained and imaged as described below. To induce goliath cell formation, a single colony of 520 C. albicans was inoculated in 4 ml SD media (2% glucose, 6.7g/L yeast nitrogen base without 521 amino acids) and incubated for 24 h at 30°C 200 rpm [62]. Following incubation, 600 µl of 522 culture were washed in three times milliQ water and resuspended in 600 µl milliQ water prior 523 to OD 600 measurement. To elicit zinc starvation, and hence goliath cell formation, washed cells 524 were inoculated into 4 ml of Limited Zinc Medium (LZM) at OD 600 0.2. LZM culture was 525 incubated for 3 days prior to fixation with 4% paraformaldehyde and staining [62]. 526

Immunofluorescent staining of Fc-lectins binding to fungal cells 527
Yeast cells were counted using an Improved Neubauer haemocytometer and 2.5 x 10 6 cells 528 were transferred into V-bottomed 96-well tissue culture plates. Plates were centrifuged at 4000 529 rpm 5 min and supernatants were removed. Samples of 1 µg/ml dectin-1-Fc in PBS, 1 % (v/v) 530 For staining filamentous cells, Fc protein in PBS + 1% FCS or BB buffer at the same 538 concentration as above was added on top of poly-L-lysine slides. Samples were analysed 539 using a BD Fortessa flow cytometer where 10000 events were recorded for each sample from 540 three independent experiments. Median fluorescence intensity for asymmetric peaks and 541 mean fluorescence intensity for symmetric peaks was calculated for each sample using 542 FlowJo v.10 software. Alternatively, 5 µl of yeast cells were added on a poly-L-lysine coated 543 glass slides (Thermo Scientific, Menzel-Gläser) prior to imaging in 3D on an UltraVIEW® VoX 544 spinning disk confocal microscope (Nikon, Surrey, UK). 545

Fc-lectins for Transmission Electron Microscopy (TEM) 547
Yeast and hyphal C. albicans samples were prepared using high-pressure freezing by 548 EMPACT2 high-pressure freezer and rapid transport system (Leica Microsystems Ltd., Milton 549 Keynes, United Kingdom). Cells were freeze-substituted in 1% acetone (w/v) OsO 4 by using 550 a Leica EMAFS2 prior to embedding in Spurr's resin and polymerizing at 60°C for 48 h. 551 Ultrathin sections were cut using a Diatome diamond knife on a Leica UC6 ultramicrotome 552 and sections were mounted onto formvar coated copper grids. Subsequently, sections on 553 formvar coated copper grids were blocked in blocking buffer (PBS + 1% (w/v) BSA and 0.5% 554 (v/v) Tween20) for 20 min prior to incubation in three washes in binding buffer (150 mM NaCl, 555 10 mM Tris pH 7.4, 10 mM CaCl 2 in sterile water, 1% FCS) for 5 min. Sections were then 556 incubated with Fc chimeric proteins (5 µg/ml for yeast and 10 µg/ml for hyphae) for 90 min 557 before six washes in binding buffer for 5 min. Fc protein binding was detected by incubation 558 with Protein A conjugated to 10 nm gold (Aurion) (diluted 1:40 in PBS + 0.1% (w/v) BSA) for 559 60 min prior to six 5 min washes in PBS + 0.1% (w/v) BSA followed by three, 5 min washes in 560 PBS, and three, 5 min washes in water. Sections were stained with uranyl acetate for 1 min 561 prior to three 2 min washes in water and were left to dry. TEM images were taken using a 562 JEM-1400 Plus using an AMT UltraVUE camera. 563

Glycan microarray analyses of Fc-lectins 564
The binding specificities of the Fc-lectin receptors were analysed using two types of 565 carbohydrate microarrays: (1) a microarray designated 'Fungal and Bacterial Polysaccharide 566 Array' featuring 19 saccharides (polysaccharides or glycoproteins) and one lipid-linked 567 neoglycolipid (NGL) derived from the chitin hexasaccharide (S Table 2); and (2) a screening 568 microarray of 474 sequence-defined lipid-linked glycan probes, of mammalian and non-569 mammalian type (S Table 3 B) essentially as previously described [92]; these probes are a 570 subset of a recently generated large screening microarray containing around 900 glycan 571 probes (in-house designation "Array Sets 42-56", which will be published in detail elsewhere). 572 The Fc-lectin binding was performed in both types of arrays essentially as described [28]. In 573 brief, after blocking the slides with 0.02% v/v Casein (Pierce) and 1% BSA (Sigma) diluted in 574 HBS (10 mM HEPES-buffered saline, pH 7.4, 150 mM NaCl) with 10 mM CaCl 2 , the 575 microarrays were overlaid with the Fc-lectins precomplexed with the biotinylated goat anti-576 human IgG (Vector) for 2 hours. The Fc-lectin-antibody complexes were prepared by 577 preincubating the -Fc-lectin with the antibody at equimolar ratios for 1 hour, followed by 578 dilution in the blocking solution to give the final Fc-lectin concentration: dectin-2-Fc 10 µg/ml, 579 CRD4-7-Fc 20 µg/ml and hDC-SIGN-Fc 2 µg/ml. Dectin-1-Fc used as control for the Fungal 580 and Bacterial Array was analysed non-precomplexed at 20 µg/ml in the blocking solution 0.5% 581 v/v casein (Pierce) in HBS. The binding was detected with Alexa Fluor-647-labeled 582 streptavidin (Molecular Probes, 1 μg/ml). All steps were carried out at ambient temperature. 583 Details of the glycan library including the sources of saccharides, the generation of the 584 microarrays, imaging, and data analysis are in the Supplementary glycan microarray 585 document (S Table 4) in accordance with the Minimum Information Required for A Glycomics 586 Experiment (MIRAGE) guidelines for reporting glycan microarray-based data [93].