Fc‐conjugated C‐type lectin receptors: Tools for understanding host–pathogen interactions

The use of soluble fusion proteins of pattern recognition receptors (PRRs) used in the detection of exogenous and endogenous ligands has helped resolve the roles of PRRs in the innate immune response to pathogens, how they shape the adaptive immune response, and function in maintaining homeostasis. Using the immunoglobulin (Ig) crystallizable fragment (Fc) domain as a fusion partner, the PRR fusion proteins are soluble, stable, easily purified, have increased affinity due to the Fc homodimerization properties, and consequently have been used in a wide range of applications such as flow cytometry, screening of protein and glycan arrays, and immunofluorescent microscopy. This review will predominantly focus on the recognition of pathogens by the cell membrane‐expressed glycan‐binding proteins of the C‐type lectin receptor (CLR) subgroup of PRRs. PRRs bind to conserved pathogen‐associated molecular patterns (PAMPs), such as glycans, usually located within or on the outer surface of the pathogen. Significantly, many glycans structures are identical on both host and pathogen (e.g. the Lewis (Le) X glycan), allowing the use of Fc CLR fusion proteins with known endogenous and/or exogenous ligands as tools to identify pathogen structures that are able to interact with the immune system. Screens of highly purified pathogen‐derived cell wall components have enabled identification of many unique PAMP structures recognized by CLRs. This review highlights studies using Fc CLR fusion proteins, with emphasis on the PAMPs found in fungi, bacteria, viruses, and parasites. The structure and unique features of the different CLR families is presented using examples from a broad range of microbes whenever possible.

| 633 WILLMENT Fc receptors, or altering the in vivo half-life by changing the pHdependent affinity of binding to the neonatal receptor FcRn (Boesch et al., 2014;Strohl, 2009).
Producing Fc fusion proteins requires the transfection of the purified plasmid containing the gene of interest into a recipient cell line (See Mayer et al., 2018 for an overview). Numerous mammalian cells lines such as human embryonic kidney (HEK) 239T and Chinese hamster ovary (CHO) cells are routinely used to generate stable or transiently expressed fusion proteins due to their ability to glycosylate, both the Fc portion and the fused ligand-binding domain, although with slight differences between these cell lines which have ramifications for purification and function (Alain et al., 2008;Sheeley et al., 1997). The levels of glycosylation, especially with sialic acid residues not generally common on the Fc portion, can affect the structure and the pI of the fusion protein, furthermore it may impact on its use as a therapeutic agent and on its ability to interact with ligands (e.g. Dendritic cell [DC] immunoreceptor [DCIR] (Bloem et al., 2014) and as reviewed by Alain et al., 2008 andStrohl, 2009) oligosaccharides, and are often non-native in structure impacting on F I G U R E 1 (a) Cartoon representation of types of Fc CLRs: (a) IgG Fc fusion protein with a N-terminal type II Group V C-type lectin receptor for example Dectin-1 (yellow CRD) with (blue spiral) stalk and two predicted glycosylation sites (green circles), the hIgG Fc CH2 domain (light purple), and CH3 domain (dark purple) showing the glycosylated region (position N297 in unmodified IgG) (green circle), linker hinge region (black lines), and the disufide bridge (double line). (b) The Mannose receptor Fc fusion protein contains multiple CRD domains (CRD4-7) cloned in tandem. (c) Heterodimeric Fc fusion proteins for example a MCL (green CRD) and Mincle (grey CRD) containing protein could be used to examine function of proteins normally associated in complexes on the cell surface. (d) Pre-complex Fc fusion proteins using a fluorescently tagged (red star) anti-Fc antibody. (e) A modified IgG Fc fusion protein containing the IgM tail piece, microtp (µtp) (red lines), and modified CH2-CH3 domains (yellow ovals) generates multivalent Fc protein complexes (modified from Czajkowsky et al., 2012) ligand-binding efficiency and precision (Strand et al., 2013), for example Fc DCIR expressed in CHO Lec8 cells, has reduced sialylation and glycosylation levels, resulting in altered ligand-binding properties (Bloem et al., 2013).
The CLR fusion protein is generally expressed with an in frame Ig signal peptide sequence for secretion through the cotranslational translocation pathway where the nacent peptide is cleaved as it crosses the ER membrane and the protein is then appropriately folded in the ER lumen. The efficiency of the signal sequence for both high titres and cleavage specificity can vary depending on the combination of cell type expressing the fusion protein and the precise amino acid sequence of the signal (Haryadi et al., 2015). Numerous non-adherent cells expression systems based on HEK293T cells have been optimized for enhanced protein yields and expression in media without serum proteins, which allows for higher purity preparations free from contaminating bovine IgG, and consequently a more accurate assessment of concentration. Purification of the Fc fusion protein requires centrifugation of the culture supernatant containing the secreted soluble protein to remove cellular components, filtration to remove debris, enrichment over a Protein A resin affinity column, a low pH elution step, and re equilibration in an appropriate buffer (Flanagan et al., 2007;Mayer et al., 2018). A secondary purification step post protein A purification, such as those used to purify alternatively tagged (e.g. HIS or Flag) versions of fusion protein, further improves purity (Rodrigues et al., 2020).
The necessity for multimeric structures for full function of CLRs must be considered when designing screening strategies using Fc fusion proteins. If the interaction between the PRR and the glycan is of low affinity, a property shared by many of the CLRs, Fc fusion structures with multiple CRDs will have a higher binding affinity (Mnich et al., 2020)  increased affinity can be prepared in numerous ways; coated on protein-A/G microbeads, engineered as hexameric Fc platforms based on the IgM C-terminal microtp (µtp) sequence, or the use of secondary antibodies recognizing the Fc portion to pre-complex the proteins (Cao et al., 2009;Czajkowsky et al., 2012). Multivalent complexes would more closely mimic the cellular context of the CLR where often multiple CRDs from the same receptor are simultaneously present, such as the tetrameric DC-specific ICAM-3 grabbing non-integrin (DC-SIGN) receptor complexes located in lipid raft clusters, simultaneously binds multiple sites on ligands, thereby increasing affinity. Screening pathogenic ligands displayed at a high glycan density can improve Fc CLR affinity and binding kinetics, either due to the increased glycan length allowing some ligand slippage but maintaining overall binding, or if multiple CRDs are present then switching of the ligand between the various CRDs may occur maintaining the overall association (Mnich et al., 2020).
The structure of the CLR ligand-binding domain(s) and how they interact with their ligands (viral, bacterial, fungal and parasite) will be discussed more fully below.

| The CLR-binding domain
A diverse range of ligand-binding domains have been utilized as Fc fusion proteins to probe pathogens and detect endogenous ligands.
This review will discuss one of the major types of glycan-binding domains, the C-type lectins, grouped according to their distinct carbohydrate recognition domain (CRD), which generally binds in a Ca 2+ -dependent manner. The other three major glycan-binding domain structures, the Siglecs with an immunoglobulin fold, galectins with a β sandwich fold, and those with a Ricin (R)-type-binding domain, will not be addressed (see Taylor & Drickamer, 2019 for a review of mammalian glycan receptors). There are comprehensive online resources which describe and annotate lectins from a variety of kingdoms: plant, viral, fungal, algal, amphibian, reptile, bacterial, and mammalian (https://www.unile ctin.eu/unile ctin3 D/search) (Bonnardel et al., 2018).

| C-type lectin receptors
The CLRs are a large superfamily of proteins that have one or more carbohydrate recognition domain (CRD) binding in a Ca 2+ -dependent manner, the C-type lectin domain. This superfamily of receptors has 17 groups (Group I-XVII) based on their domain organization and phylogeny (see Brown et al., 2018;Taylor & Drickamer, 2019 for recent reviews and https://www.imper ial.ac.uk/resea rch/anima llect ins/ctld/mamma ls/human vmous edata.html for a database of human and mouse proteins.) The CLRs' roles in modulation of the immune response, shaped by their expression, cellular context, and signaling capacity, are covered by numerous reviews (Brown et al., 2018;Brubaker et al., 2015;Cao, 2016;Iborra & Sancho, 2015).
The secondary structure of the CLR CRD is maintained through six conserved cysteine residues forming disufide bonds, creating a hydrophobic core. Conserved amino acid motifs such as EPX (Glu-Pro-X = Asn, Ser or Asp) found in DC-SIGN, Langerin (all Group II), and the MR (group VI) preferentially allow binding to glycans with equatorial 3-and 4-hydroxyl groups such as mannose, fucose, glucose, and N-acetyl-D-glucosamine (GlcNAc) residues. CRDs with a QPD (Gln-Pro-Asp) motif, such as human mannose galactose lectin (MGL) (Group II), preferentially bind glycans with axial 4-hydroxyl groups such as galactose and N-acetyl-galactosamine (GalNAc) residues (See Foster et al., 2016;Mnich et al., 2020 and for reviews).
High Ca 2+ concentrations are important for glycan binding by receptors with multiple Ca 2+ -binding sites, such as SIGNR1 (Group II), while receptors containing only one Ca 2+ -binding site (e.g. DCIR-2) have a basic amino acid present in the secondary site . Not all CRDs bind their ligands in a Ca 2+ -dependent manner and are termed the C-type Lectin-like domains (CTLD) an example of which is Dectin-1 (Group V) (See Table 1). Some CLRs have additional cysteine residues in the stalk region, located between the transmembrane (TM) region and the CRD, which are necessary for homo-dimerization (e.g. Lox-1), or a positively charged residue, e.g. Arg in Dectin-2, within the TM facilitating hetero-dimerization β 1,3 glucans β 1,3 and 1,6 linked-glucans Brown and Gordon (2001); Viriyakosol et al. (2005); Graham et al. (2006); Sato et al. (2006)

TA B L E 1 (Continued)
with the signaling molecule FcRgamma (Ohki et al., 2005;Sato et al., 2006). The presence of these residues and any glycosylation sites, within the extracellular regions, can impact on the expression and function of the glycan-binding Fc fusion proteins (see below).
A brief overview of selected Group II, V, and VI receptors is given highlighting their use as Fc CLR proteins.

The Group V CLRs
The Group V CLRs are type II proteins, with a single extracellular  Table 1). The majority of these CLRs are cloned such that their C-terminal end is fused to the Fc region (e.g. the Fc Dectin-1 fusion protein includes the stalk region (Graham et al., 2006)) (See Figure 1). This allows two CRDs to be simultaneously presented as a dimeric complex for ligand binding, which is not always sufficient to detect low-affinity binding without further pre complexing, for example Fc Lox-1 (Cao et al., 2009).
Dectin-1. The β-glucan-binding receptor Dectin-1, with high specificity for β 1,3-linked glucans in a Ca 2+ independent manner, was first identified in a screen for receptors binding to the Saccharomyces cerevisiae-derived particle zymosan. The receptor has been used as an Fc fusion screening tool in numerous studies involving fungi, bacteria, parasites, and endogenous ligands (See Table 1).
The receptor is widely expressed as a monomer on a range of cell inhibitable manner, yet neither pathogen contains β-1,3-glucans (Cheng et al., 2017;Heyl et al., 2014;Rothfuchs et al., 2007). The endogenous ligands such as the proteins, vimentin, and galectin-9 are recognized by Dectin-1 in a β-glucan inhibitable manner (Daley et al., 2017;Thiagarajan et al., 2013). A second ligand-binding site on Dectin-1 exists and is able to interact with the endogenous ligands IgG and Annexins in a β-glucan independent manner. Dectin-1 binds IgG at a core fucose linked to an N-terminal asparagine with an adjacent aromatic amino acid, extending the range of known ligand structures (Bode et al., 2019;Manabe et al., 2019).
Fungi are a rich source of β-glucan (see below and Erwig & Gow, 2016 for a review) consequently Fc Dectin-1, due to its high affinity for β 1,3 glucan, has been used to probe the fungal cell wall (e.g. Candida albicans [See Figure 2]) for alterations in exposure due to growth conditions, by flow cytometry, EM, and immunofluorescent microscopy (Lenardon et al., 2020;Pradhan et al., 2019;Vendele et al., 2020).
Comparisons of Fc Dectin-1 binding to a variety of Candida sp. show the following trends in β-glucan exposure with binding to C. glabrata < C.
tropicalis equivalent to C. parapsilosis < C. albicans, when all cultured in a similar manner (Thompson, Griffiths, et al., 2019). Detection of ex-  Table 1). The fungal cell wall has a multi-layered structure with chitin and β1,6-glucan in the inner layer above the cell membrane (CM), glycosylated integral and GPI anchored proteins (blue ovals), and the outer cell wall composition varying depending on the fungal strain and/or morphology. CRDs separated by a short linker (Cao et al., 2009). There are limited data available for Fc Lox-1 fusion proteins as tools for binding microbial pathogens, potentially due to the need for pre-complexing; however, in vitro cell-expressed receptors have shown interactions with S. aureus and E. coli (Shimaoka et al., 2001). Significantly, using a double CRD Lox-1 fc fusion protein, anionic oligoglycerol dendronized sulfates targeting the dimeric basic spine region were detected, and these compounds inhibit binding to E.coli, an endogenous ligand on red blood cells, and to oxLDL (Kumari et al., 2019).
independent binding homodimeric C-type lectin receptor. Its Fc CLEC-2 protein would need to be pre-complexed or expressed as a heterodimeric receptor to fully explore its PRR ligand-binding functions (see Figure 1). Nevertheless, a homodimeric Fc CLEC-2 used in two glycan array screens bound a limited range of glycans some of which are present on both host and pathogen surfaces (see Table 1). Binding of Fc CLEC2 to agarose-immobilized T. gondii oocysts was detected by immunofluorescence microscopy and although the Fc fusion binds C. jejuni in an ELISA-based prescreen this was not repeatable by flow cytometry, thus highlighting the need for alternative confirmatory assays when assigning function (Fabian et al., 2021;Mayer et al., 2018).
The CLEC-2 CRD-binding sites for podoplanin and rhodocytin are distinct from the canonical sugar-binding site of C-type lectins, the β3 and β4 strands, and while both ligands share the identical binding site, they each interact with a further unique site. A number of critical arginine residues in these sites were identified using CLEC12AandCLEC12B. These two receptors are understudied in the context of glycan binding, but have been used as Fc-fusion proteins in the screening of microbial ligands often used as baseline binding control proteins due to their limited interactions with pathogens detected to date (Mayer et al., 2018;Stappers et al., 2018). CLEC12A (myeloid inhibitory C-type lectin MICL, DCAL-2, KLRL-1, and CLL) and CLEC12B (Mah) are both inhibitory receptors, due to the presence of immunoreceptor tyrosine inhibitory motifs within their cytoplasmic signaling tail. CLEC12A plays a role in homeostasis, as it binds dead cells and uric acid crystals, and limits inflammation in rheumatoid arthritis (Neumann et al., 2014;Redelinghuys et al., 2016). In microarray screens mild-to-weak binding for both Fc CLEC12A and Fc CLEC12B was observed to a limit range of glycans (see Table 1 for a list of these glycan structures). More recently CLEC12A has been found to bind hemozoin, an insoluble crystal derived from Plasmodium (Raulf et al., 2019), again indicating that crystal structures could be recognized by CLEC12A. Interestingly, Fc CLEC12A has unidentified ligand(s) on both T. gondii oocysts and C.

CLEC1 (MelLec) and CLEC9A (DNGR). MelLec (CLEC1a) has
been recently identified as the first CLR able to bind melanin, recognizing the naphthalene-diol unit of 1,8-dihydroxynaphthalene conidial spores (Stappers et al., 2018). No other MelLec microbial ligands have been identified to date, but further structures may still be identified using complexed Fc fusion proteins to enhance binding affinity. Significantly, a human MelLec fusion protein bound with high affinity to a mammalian histidine-rich glycoprotein (HRG), but the nature of the interaction is unresolved (Gao et al., 2020).  (Table 1).
As a cell membrane expressed receptor CLEC5A's binding to L.
monocytogenes is enhanced when the receptor is complexed with TLR2 . CLEC5A is further an example of a CLR's full binding function dependent on the presence other receptors.

The Group II CLRs
The Group II family of C-type lectin receptors are similar in structure to the Group V cluster, they are type II proteins with a TM region, stalk, and a single C-type lectin-like domain, and have a short  Table 1 Table 1 (Table 1). Langerin is a trimeric CLR, due to the coiled neck region, with a distinct glycan-binding pattern mostly resolved using bacterial expressed tagged proteins  residues binding in an adjacent wide shallow groove, necessary for higher affinity binding (Jégouzo et al., 2015). They have distinct functions as DC-SIGN is an endocytic receptor which releases its ligand at low pH and DC-SIGNR is an adhesion receptor (Guo et al., 2004). Both receptors have an extended neck region, which contains seven repeats of highly conserved stretches of amino acids, facilitating homo-oligomerization necessary for high avidity binding to the high mannosylated HIV gp120. In contrast, binding to their endogenous ligand ICAM-3 is of low affinity and independent of oligomerization state (Snyder et al., 2005). The DC-SIGN tandem-neck repeat region is constant in size, whereas the numerous isoforms of DC-SIGNR vary in neck length with consequences for function (Khoo et al., 2008). The full length DC-SIGNR's neck regions allow the CRDs to be more closely positioned than the DC-SIGN CRDs, when these receptors are membrane expressed. This impacts on the surface area available to interact with ligand baring molecules and pathogens (Feinberg et al., 2021).

MCL (Dectin
The complexed structure of these receptors, in particular DC-SIGN, impacts on their function as soluble glycan-binding molecules. When an interacting ligand contains appropriately distanced glycans, the HMW complexes bind with higher affinity than monomeric structures (Mitchell et al., 2001). In contrast, however, the significantly enhanced binding of a Man9GlcNAc2 oligosaccharide over mannose by both receptors is not facilitated by oligomeric CRD structures implying that multiple binding sites on each individual CRD are needed to interact with the oligosaccharide. Many of these binding sites would not be exposed when the CRDs are clustered into HMW complexes. In general, binding profiles of either DC-SIGN or DC-SIGNR when used as pre-complexed Fc proteins, dimeric Fc protein or bacterial expressed proteins are comparable, with some differences in binding to low-affinity ligands observed Geissner et al., 2019;Tateno et al., 2008).  (Table 1). Bacterial expressed and Fc fusion proteins of these two receptors were used to demonstrate that they bind preferentially to mannose over glucose and GlcNAc, with DC-SIGNR showing greater affinity for mannose than fucose, and it has a higher affinity than DC-SIGN for GlcNAc. The higher affinity for L-fucose than mannose is due to the presence of a Val351 in the primary-binding site of DC-SIGN. This residue interacts with fucose necessary for binding to the Le X and LDNF glycans on the soluble egg antigen (SEA) from S. mansoni (Figure 2), but not for binding of HIVgp120 (Guo et al., 2004;van Die et al., 2003).
There are eight mouse homologs of human DC-SIGN, of which SIGNR1 and SIGNR3 are the most extensively studied as glycanbinding Fc CLRs, with subtle differences in ligand-binding profiles detected (Galustian et al., 2004;Lee et al., 2011) (See Table 1).
Membrane-expressed SIGNR1 binds the capsular polysaccharide The binding of Fc SIGNR3, but not Fc SIGNR1, is inhibited by the presence of a core xylose on biantennary N-glycan, indicating that only SIGNR1 might be able to detect and respond to non-self glycan motifs often found on parasites. The absence of response to L. infantum in SIGNR3 KO mice confirms the lack of recognition observed in vitro (Brzezicka et al., 2016;Lefèvre et al., 2013). In contrast, Fc SIGNR3 detects fungi in microbiota samples and the KO mice display increased inflammation in a colitis model (Eriksson et al., 2013).

MGL (CLEC10A), ASGPR (CLEC4H). These closely related receptors
are predominantly known as self-glycan-binding receptors with roles in homeostasis, dependent upon their cellular context (Brown et al., 2018). Both receptors have the galactose, GalNAc and glucose-binding QPD motif and although they share homology their carbohydrate-binding specificities differ considerably (See Table 1).  (Zhang, Tian, et al., 2016) and assays using cell membrane-expressed receptors have demonstrated binding to Hepatitis A, B, and C viruses.
Human MGL is a trimeric receptor, which exclusively recognizes terminal GalNAc residues (Tn and sialyated Tn antigens (Figure 2) found on tumor cells, mucins, and CD45. The murine homolog MGL2 (CD301b) is similar in function and binds Tn antigens and galactose (Singh et al., 2009 Table 1 and

| DE TEC TI ON OF PATHOG EN S
The immune system has to recognize and differentiate between PAMPs, DAMPs, and self-associated molecular patterns (SAMPs) to maintain the balance between immunity and homeostasis; however, dysregulation of these processes often results in autoimmunity, with many of the glycan-binding receptors highlighted in this review im-

| Recognition of viral PAMPs
Viruses  Figure 2) or incompletely processed N-linked glycans with exposed GlcNAcβ1-2Man making them detectable by the immune system (Taylor & Drickamer, 2019 One of the extensively studied viral families is the Retroviridae, of which the enveloped virus HIV-1 is a member. The HIV viral glycoprotein (gp120) (Figure 2) that mediates viral entry is highly glycosylated containing approximately 30%-40% high mannose-type glycans (Balzarini, 2007). Deglycosylation with PNGasF, but not denaturation, of gp120 destroys binding to Fc DC-SIGNR. Furthermore, this binding is pH dependent with loss of function occurring for both recombinant DC-SIGN and Fc DC-SIGNR at low pH, implicating these receptors in viral uptake and processing in endosomal compartments (Snyder et al., 2005). assays. These assays were verified using soluble Envelope (sE) protein DV antigen, produced both in mosquito and human cells with binding Ca 2+ , mannose, fucose, and galactose dependent. Transfer blots probed with the Fc protein (far eastern blots) confirmed the specificity for binding to the sE glycans (Miller et al., 2008). Fc proteins of CLEC5A, DC-SIGN, and DC-SIGNR bind to DV (serotype 2) (Chen et al., 2008).

| Recognition of fungal PAMPs
Fungi have long been understudied as pathogens considering their considerable burden on human health (Erwig & Gow, 2016). Fungi cause significant disease with more than a million deaths and significantly more life-threatening infections per annum with most fungi-opportunistic pathogens causing disease in patients with compromised immune systems (Bongomin et al., 2017). The most well studied yeasts are S. cerevisiae, C. neoformans, Candida spp., and Malassesia spp. and of the molds Penicillium, Mucor, and Aspergillus.
(see Drummond et al., 2014;Garcia-Rubio et al., 2020 for reviews).  Table 1). The outer walls of fungi are usually highly glycosylated, may be covered in a hydrophobic layer (e.g. A. fumigatus conidia), or surrounded by a capsule (e.g. C. neoformans).
The distribution of these components within the cell wall can vary between strains, for example Dectin-1 binding to H. capsulatum strain G186A is prevented by a layer of α-glucans, whereas strain G217B or a glucan synthase mutant Δags1 is readily bound by cellexpressed Dectin-1 (Wang et al., 2014). The amount of α-1,3-glucan content can differ for Histoplasma, Blastomyces, and Paracoccidiodes during laboratory passage and will consequently impact on the Fc PRR-binding levels. Variations in β-glucan exposure can be influenced by many factors for example the Histoplasma Eng1 glucanase, which is usually highly expressed during the pathogenic phase cleaves exposed β-glucan only allowing detection by Dectin-1 of βglucan in the septum of budding cells. In contrast, the Eng1 mutant yeast is completely bound by Fc Dectin-1 as observed by immunofluorescent microscopy (Rappleye et al., 2007).  (Rapaka et al., 2007). Binding of Fc MCL to Cn serotypes (B, and AD) by flow cytometry and by ELISA to extracted GXM from these same serotypes, as well as evidence from infected KO mice being more susceptible, provides evidence for a ligand for MCL (Huang et al., 2018). These highlighted experiments, using Fc CLRs and relevant controls to probe the glycan ligands on Cryptococcus, demonstrate the importance of testing a range of serotypes and that fungal mutants can be utilized to fully probe the fungal cell wall structure.
C. albicans, with a large number of well characterized mutants, is a model organism for studying fungal interactions with the immune system and consequently the cell wall structure has been extensively studied. The Candida cell wall (Figure 2) has an inner layer composed of β-glucan interspersed with chitin and outer layer containing fibrils of N-mannans linked to cell wall proteins, O-linked mannans, and phosphomannans (Erwig & Gow, 2016;Garcia-Rubio et al., 2020;Lenardon et al., 2020).
The four Candida species that cause the most hospital acquired infections are C. albicans, C. glabrata, C. tropicalis, and C. parapsilosis with C. auris being a more recent growing concern (Bongomin et al., 2017;Chowdhary et al., 2017). These species have been assessed for binding to Fc Dectin-1, detecting the abundant β-1,3-linked glucans (approx. 50%-60% of the dry weight), showing that C.a had the least β-glucan exposure (Thompson, Griffiths, et al., 2019). There are 17 closely related C.a strain groups (clades), based on multi locus sequence typing, and although they have varying amounts of exposed β-glucan these differences only give a discernible phenotype during in vivo Dectin-1 KO infection models and not in vitro-based assays. Fc Dectin-1 bound more exposed β-glucan on ATCC 18804 (clade 5) than SC5314 (clade 1) hyphae, and in this study both strains increased their surface exposed β-glucan when cultured under hypoxic conditions (Marakalala et al., 2013). In contrast, the amount of exposed β-glucan, as detected by FcDectin-1, of hypoxic SC5314 or iron-depleted yeast cells is reduced, and exposure is increased under low pH, a range of antifungal drug treatments, and during depletion of manganese and zinc micronutrients (Pradhan et al., 2019).
Interestingly SC5314 yeast with higher chitin levels and more exposed β-glucan were less virulent in Dectin-1 KO mice (Marakalala et al., 2013). Although the plant lectin wheat germ agglutinin (WGA) is routinely used to probe chitin content, as it binds β-1,4-GlcNAclinked residues, it has been utilized as a an antifungal therapeutic Fc fusion protein (Liedke et al., 2017). No Fc PRRs currently recognize chitin, although a few PRRs (e.g. Mannose receptor, NOD2, and TLR-9) have been implicated in shaping the immune response to C.
The use of Fc CLRs to probe the cell wall content and structure of aspergillus currently lags behind that of Candida, despite it being implicated in severe asthma, causes allergic bronchial pulmonary aspergillosis, chronic and invasive aspergillosis. The aspergillus outer cell wall varies considerably between the conidium and hyphae Latgé & Chamilos, 2019 for reviews).
Aspergillus exposure of β-glucans bound by a soluble un mutated murine Fc fused with Dectin-1 shows low-binding levels on resting conidia and hyphae, with more uniform binding on swollen conidia and early germlings (Steele et al., 2005). The importance of the exposed β-glucan during antifungal immunity is highlighted from two different assays using Fc Dectin-1. Firstly, intranasally introducing a mutated human Fc region fused to Dectin-1 at the same time as A.f conidia into mice significantly decreases the immune response (CXCL1/KC, TNF-alpha, CCL3/MIP-1, IL-6, and GM-CSF) and results in much higher fungal burdens (Steele et al., 2005). Secondly, the exposed β-glucan bound by a mIgG Fc portion fused to Dectin-1 targets the microbe for enhanced clearance via Fc binding receptors, this decreased fungal burden and enhanced survival of the infected mice (Mattila et al., 2008;Rodriguez-de la Noval et al., 2020). The aspergillus GAG layer masks the β-glucan layer as can be observed with the Δuge3 mutant, lacking GalNAc, has increased binding to FcDectin-1 on both swollen conidia and hyphae, and the mutant is less virulent due to the increased detection of the exposed β-glucan by the immune system (Gravelat et al., 2013). There is some incon- Melanin, located below the rodlet layer in A. fumigatus conidia, is bound by FcMelLec in a punctate manner and removal of the RodA protein then shows a uniform melanin layer over the surface of the conidia, which disappears as the conidia swells and germinates (Stappers et al., 2018). This study found melanin on F. pedrosoi and NaOH-treated C. cladosporioides.

| Recognition of bacterial PAMPs
The global burden of bacterial infections is significant, with 1.4 million people dying from TB infections alone in 2019, furthermore there was a 10% rise in antimicrobial resistance, underscoring the importance of combating these pathogens (WHO, 2020). Understanding how bacteria interact with the immune system requires analysis of the first point of contact, the outer cell wall of the bacteria, which can be probed with Fc CLRs to define the exposed glycans (Mayer et al., 2018). A range of bacterial cell wall structures will be highlighted with emphasis on those, such as Mtb which has a complex outer cell wall structure (Figure 2), with known CLR ligands (Table 1).
An inner peptidoglycan ( Streptococcus, Enterococcus, and Lactococcus contain large amounts of L-rhamnose, a sugar not found in humans and therefore could act as a PAMP, as part of their anchored CWPs; however, there are currently no known mammalian PRRs recognizing the rhamnan backbone, which is often decorated by sugars like GlcNAc, GalNAc and glucose to avoid immune recognition (Mistou et al., 2016).
CLEC5A, for example, binds to L. monocytogenes independently of the terminal sugars (rhamnose, galactose, glucose) to TA, with inhibition studies suggesting binding occurs to the GlcNAc-MurNAc disaccharide backbone .
Glycosylation of outer exposed bacterial proteins is a further source of immune recognition and occurs in numerous bacterial species (e.g. mycobacteria, L. monocytogenes, B. anthracis) and although many PRRs recognize these glycoproteins, their full role in detection of these sugars is not resolved and is a neglected area of research (See Mehaffy et al., 2019;Weidenmaier & Peschel, 2008; Taylor & Drickamer for reviews). Bacterial biofilms are another layer of protection against immune recognition; however, they are also a rich source of carbohydrates such as the mannose-rich P. aeruginosa PSI polymer, recognized by both Fc DC-SIGN and Fc MMR CRD4-7.
The PSI polymer interferes with endocytic function of these receptors when cell membrane expressed (Singh et al., 2020).
The unique outer membrane of mycobacteria contains unique molecules (see Figure 2 and Dulberger et al. (2020) for a review on the mycobacterial cell envelope). Some glycans are linked to the mycolic acids such as trehalose monomycolate (TMM) and trehalose dimycolate (TDM) also known as cord factor, the latter recognized by Mincle and MCL (Furukawa et al., 2013). Lipids such as the phosphatidylinositol mannosides (PIMs), which depending on their structure and acylation, are recognized by a range of CLR receptors (e.g. DCAR specifically recognizes the tri and tetra-acylated dimannosyl PIM2 (Toyonaga et al., 2016)). CLRs can recognize multiple mycobacterial-derived ligands, for example the outer membrane lipomannans (LMs) and inner membrane-associated lipoarabinomannans (LAMs) are both recognized by Dectin-2 (Yonekawa et al., 2014) and the mannose-capped lipoarabinomannans (ManLAM) and the larger structure PIMs associate with the MMR and DC-SIGN, with the PIM binding dependent on the terminal carbohydrates and the degrees of acylation (Driessen et al., 2009;Torrelles et al., 2006).  Although mechanisms for eukaryotic protein glycosylation are highly conserved, and many are identical to those of the hosts' glycans (see Figure 2), parasites display some unusual monosaccharides residues and end caps, uncommon oligosaccharide linkage configurations, and protozoa have glycosylphosphatidylinositol (GPI)anchored oligosaccharides often differing in composition to the hosts (Veríssimo et al., 2019). An example is T. brucei, which displays high-mannose structures, often decorated with β 1-2-linked xylose or α-1,3-fucose modifications, readily recognized by the host immune system as foreign. Arrays of chemically synthesized core-xylosylated glycans, synthetic N-glycan, and fragments of glycolipids and Oglycans have been probed with Fc CLRs (Brzezicka et al., 2015).  properties of some CLR receptors, pre-complexing of the Fc fusion either through secondary antibody (Figure 1) or bead coupling is necessary to achieve higher sensitivity required to detect lowaffinity interactions by ELISA (Lee et al., 2011;Mayer et al., 2018;Rodrigues et al., 2020). The spatial location of PAMPs can be detected in pathogens by electron microscopy, for example detection of Fc Dectin-2 binding to the inner cell wall of C. albicans (Vendele et al., 2020) and Fc mDectin-1 detection of β-glucan contributed to resolve the nanomolecular structure of the C. albicans cell wall (Lenardon et al., 2020). Direct stochastic optical reconstruction microscopy (dSTORM) imaging using Fc hDectin-1 detected nanoscale exposure of β-glucan in C.

| Fc CLR : A SSAYS , D IAG NOS TI C S AND THER APEUTI C S
albicans and C. glabrata and was able to detect increased size and density in C.a clinical isolates (Graus et al., 2018). These studies have visually demonstrated the spatial and temporal exposure of β-glucan available to interact with immune cell membrane expressed Dectin-1 and this has aided our understanding how pathogens through modification of their outer glycans modulate the downstream immune response (Pradhan et al., 2019).
Fc CLRs, apart from their function in detecting glycan content in microbes, have therapeutic and diagnostic uses (Czajkowsky et al., 2012).

| LIMITATIONS AND CONS IDER ATIONS US ING Fc CLR s
As evidenced by the CLRs discussed in this review, the binding properties of each CLR must be fully understood before undertaking assays screening pathogens for glycan binding. A number of factors need to be considered; the basic biochemical properties such as Ca 2+ dependence for binding, requirement for receptor multimerization to ensure sufficient avidity of binding, species differences in ligand recognition (e.g. CD23), and analysis of the glycosylation state of the CLR and/or fusion protein and how this differs depending on the expression protocol (Taylor & Drickamer, 2019). Glycan-binding results should always be validated by more than one method, ideally including the use of soluble inhibitors to verify specificity of the CLR for example MGL, and not the Fc portion of the fusion, specifically recognizes C. jejuni LOS as verified by inhibition of binding with an excess of GalNAc, blocking with the anti-MGL antibody, the necessity for divalent cations for binding and the loss of reactivity for bacterial glycosylation mutants (Van Sorge et al., 2009). The mutation of essential amino acids within the glycan-binding site (e.g. the Mincle QPD motif (Furukawa et al., 2013)) is a desirable control for both CLR-binding specificity and the higher background-binding levels inherent in Fc fusion protein assays.
Many bacteria, such as S. aureus and S. pyogenes, express Fcbinding proteins making the true detection of expressed ligands using Fc fusion proteins unreliable. An alternative approach is to cleave the Fc portion post-purification, for example using a thrombin cleavage or tobacco etch virus cleavage sequence cloned between the receptor and the Fc portion (Guan et al., 2010;Rodrigues et al., 2020). CLRs such as DC-SIGN, Dectin-2, and the MMR which bind oligomannosylated IgG (Boesch et al., 2014), may bind glycosylated antibodies present on the surface of serum blocked pathogens, they could interact with the secondary antibodies used for pre-complexing or interact with the Fc region itself.
Further limitations in using Fc proteins for pathogen ligand screens are not as obvious: for example protein aggregations may give false positive results (Mayer et al., 2018). The proportion of HMW complexes in Fc fusion preparations should be assessed and size-exclusion chromatography used to standardize the Fc CLR production and usage for consistent and comparative binding assays.
Differences were observed in glycan array-binding assays using dimeric or multivalent complexes for example Fc DC-SIGN binds glycans from both C. albicans and S. cerevisiae, whereas tagged probes only recognize those from C. albicans. The tagged probes allowed subtle difference in binding between DC-SIGN and mSIGNR1 to be detected, with only the latter receptor able to bind the β-mannosecapped α-mannose side chains found in C. albicans (Takahara et al., 2012). In contrast to these data, DC-SIGN binding, tested both as an Fc fusion protein and as a tetrameric soluble protein produced in bacteria, to an array of over 140 glycans-containing bacterial, parasite, mammalian, and plant polysaccharides showed very similar patterns of glycan recognition (Geissner et al., 2019).
Detection of pathogens usually requires two or more different cell surface expressed CLRs to elicit an appropriate immune response, which is shaped by the cell type and its location, and may and Cryptococcus (Walsh et al., 2017;Wang et al., 2014). In contrast, SIGNR1 binds Dextran when cell membrane expressed but not as an Fc fusion protein due to the polysaccharide being a very low-affinity ligand (Galustian et al., 2004).
Fc CLRs might not detectably bind to ligand-baring pathogens due to a range of factors. Live pathogens may mask or alter their PAMPs in response to cellular growth conditions, or cell wall ligand density is too low by comparison to the high concentrations of purified compounds used in assays (Inoue & Shinohara, 2014;Mnich et al., 2020;Pradhan et al., 2019). Ligand binding may be pH sensitive, with some receptors able to bind and function at a wide range including low pH for example Fc Dectin-1 is able to bind from pH4 to pH8, in contrast glycan binding by Fc CD23 protein is pH sensitive, with optimal binding at pH6.5 (Faro-Trindade et al., 2012;Jégouzo et al., 2019).
A further limitation of Fc CLRs is the inability to simultaneously detect multiple Fc fusion proteins probing unique ligands on the same pathogen. Directly conjugating fluorophores is often problematic as conjugation, using amine labeling chemistry, can render the ligandbinding domain of CLRs non-functional (unpublished data), but has been successfully achieved with DC-SIGN (Zheng et al., 2017).
Alternative approaches such as the use of the fusion proteins precomplexed to fluorescent beads or antibodies, or harnessing systems using variations on fluorescently labeled streptavidin-binding biotin tagging have been successfully used ) (See Figure 1). These examples highlight some of the challenges using Fc CLRs, instead of cell membrane expressed receptors, as tools for probing ligand content on pathogens.

| CON CLUS IONS
The Fc fusion proteins of the CLR family, with their diverse ligandbinding repertoire, are extremely valuable tools for the detection and analysis of the content and structure of pathogen cell walls.
Their use reveals the PAMPs exposed for recognition by the host's immune system. The temporal detection of exposed PAMPs has been extrapolated into understanding host-pathogen interactions and how pathogens both evade and engage with host cells.
In the last few years significant progress has been achieved understanding the role of CLRs in immunity and homeostasis; however, there are some outstanding areas of research that still need to be considered. While substantial advances have been made for fungi and viruses, the role of CLRs in the detection and immune response to bacterial, protozoa, and helminth glycans is relatively not very well characterized (Brown et al., 2018;Taylor & Drickamer, 2019;Veríssimo et al., 2019). For example, there are currently unknown receptors for rhamnose, arabinofuranose, and xylofuranose all expressed by various microorganisms (Zheng et al., 2017). Fc fusion proteins of numerous PPRs have yet to be made (e.g. the CL-P1 collectins) and many of the CLRs reviewed here have not been extensively examined or were screened on the first generation arrays with limited ranges of glycans. Furthermore the ligand binding profiles of many of the Fc CLRs, especially those with low-affinity binding properties, which are usually cell membrane expressed as multimeric structures, should be reassessed using complexed Fc probes in screens with high-density purified glycans (Mnich et al., 2020;Zheng et al., 2017). Recent advances in chemical synthesis and enhanced purification methods of specific glycans should resolve false positive CLR ligand interactions detected due to the impurity of many glycan preparations, especially those from bacteria and viruses (Geissner et al., 2019;Zamze et al., 2002). The use of Fc CLRs as well as verification by cellular binding assays, controlled with relevant CLR CRD mutations, glycan inhibition, and/or antibody blocking, should further resolve discrepancies regarding recognition of specific pathogens by CLRs.
Due to our increased understanding of the role that various cell wall structures play in pathogenicity, Fc CLRs have been used as basic diagnostic and therapeutic tools and have the potential to be engineered, modifying sensitivity and specificity, enhancing these functions. Exciting future work using artificial design or strategically modifying binding motifs in CLRs may give rise to probes with unique binding properties able to screen distinct glycans and give rise to a new generation of research, diagnostic, and therapeutic tools.

ACK N OWLED G M ENTS
Funding was provided by the MRC Centre for Medical Mycology (MR/N006364/2) at the University of Exeter, UK.

CO N FLI C T O F I NTE R E S T
The author has no conflict of interest to declare.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data sharing is not applicable to this article as no new data were created or analysed in this study.