The surface carbohydrates of the Echinococcus granulosus larva interact selectively with the rodent Kupffer cell receptor

Molecular & Biochemical Parasitology 192 (2013) 55–59

Contents lists available at ScienceDirect

Molecular & Biochemical Parasitology

Short communication

The surface carbohydrates of the Echinococcus granulosus larva interact selectively with the rodent Kupffer cell receptor Tsui-Ling Hsu a , Gerardo Lin b , Akihiko Koizumi c , Klaus Brehm d , Noriyasu Hada c , Po-Kai Chuang a , Chi-Huey Wong a , Shie-Liang Hsieh a,e , Alvaro Díaz b,∗ a

Genomics Research Center, Academia Sinica, 128 Academia Road, Section 2, Nankang, Taipei 115, Taiwan Cátedra de Inmunología, Departamento de Biociencias, Facultad de Química, e Instituto de Química Biológica, Facultad de Ciencias, Universidad de la República, Instituto de Higiene, Av. A. Navarro 3051, Montevideo CP 11600, Uruguay c Faculty of Pharmacy, Keio University, 1-5-30 Shibakoen, Minato-ku, Tokyo 105-8512, Japan d University of Würzburg, Institute of Hygiene and Microbiology, Josef-Schneider-Straße 2/E1, 97080 Würzburg, Germany e Institute of Microbiology & Immunology, Institute of Clinical Medicine & Infection, and Immunity Center, National Yang-Ming University, No. 155, Sec. 2, Linong Street, Taipei 112, Taiwan b

a r t i c l e

i n f o

Article history: Received 23 October 2013 Received in revised form 6 December 2013 Accepted 10 December 2013 Available online 17 December 2013 Keywords: Echinococcus Laminated layer Liver Kupffer cells Kupffer cell receptor Carbohydrate

a b s t r a c t The larvae of the cestodes belonging to the genus Echinococcus dwell primarily in mammalian liver. They are protected by the laminated layer (LL), an acellular mucin-based structure. The glycans decorating these mucins constitute the overwhelming majority of molecules exposed by these larvae to their hosts. However, their decoding by host innate immunity has not been studied. Out of 36 mammalian innate receptors with carbohydrate-binding domains, expressed as Fc fusions, only the mouse Kupffer cell receptor (KCR; CLEC4F) bound significantly to the Echinococcus granulosus LL mucins. The receptor also bound the Echinococcus multilocularis LL. Out of several synthetic glycans representing Echinococcus LL structures, the KCR bound strongly in particular to those ending in Gal␣1-4Gal␤1-3 or Gal␣1-4Gal␤1-4GlcNAc, both characteristic LL carbohydrate motifs. LL carbohydrates may be optimized to interact with the KCR, expressed only in liver macrophages, cells known to contribute to the tolerogenic antigen presentation that is characteristic of this organ. © 2013 Elsevier B.V. All rights reserved.

The larvae of the cestodes belonging to the genus Echinococcus colonize mammalian internal organs, primarily the liver. Larval Echinococcus granulosus, the causative agent of cystic echinococcosis (hydatid disease), which is the main focus of this article, can develop in other organs in addition to liver. It infects many ungulate species (including in particular livestock animals) and accidentally humans. This larva (hydatid) is a fluid-filled, bladder like, unilocular structure. Larval Echinococcus multilocularis, the causative agent of alveolar echinococcosis, infects rodents (and accidentally, humans) and develops almost exclusively in the liver as a labyrinth of interconnected vesicles. In all cases, Echinococcus larvae are bounded by a larval wall comprising a thin inner layer of cells (germinal layer) and an outer acellular, carbohydrate-rich coat called the laminated layer (LL) (reviewed

Abbreviations: ASGR, asialoglycoprotein receptor; Gal, galactose; GalNAc, N-acetylgalactosamine; GlcNAc, N-acetylglucosamine; IIRs, innate immune receptors; KCR, Kupffer cell receptor; LL, laminated layer; MGL, macrophage galactose-specific lectin. ∗ Corresponding author. Tel.: +598 24874320. E-mail addresses: [email protected], [email protected] (A. Díaz). 0166-6851/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.molbiopara.2013.12.001

in [1,2]). The LL, which is specific to the genus Echinococcus, is essentially a meshwork of highly glycosylated mucins [3]. For E. granulosus, the glycans decorating the LL mucins have been studied [4,5]. They are mucin-type O-glycans composed only of galactose (Gal), N-acetylgalactosamine (GalNAc) and N-acetylglucosamine (GlcNAc), and based on either core 1 (Gal␤1-3GalNAc1␣-Ser/Thr) or core 2 (Gal␤1-3(GlcNAc␤1-6)GalNAc1␣-Ser/Thr). The core Gal residue can be substituted with a variable number of Gal␤1-3 residues, thus giving rise to a (Gal␤1-3)n “main chain” (defined as comprising the mentioned core residue). This chain can be capped with a single Gal␣1-4 residue. In addition, the core 2 GlcNAc residue can be decorated with the Gal␣1-4Gal␤1-4 disaccharide, thus giving rise to the P1 blood antigen motif (Gal␣1-4Gal␤1-4GlcNAc). Also, the main chain can be ramified with GlcNAc␤1-6 residues, which in turn can be decorated with Gal␣1-4Gal␤1-4, thus again giving rise to P1 motifs. Glycans related to the above are found in a LL fraction from E. multilocularis [2,6]. Comparison of the two sets of glycans suggests an apparent species difference, as capping with Gal␣1-4 appears to take place in E. multilocularis directly on the core 1 Gal residue, while in E. granulosus it happens only on chains comprising at least two Gal␤1-3 residues. Thus the LL glycans in E. granulosus are longer than those in the E. multilocularis fraction [2,4–6].

56

T.-L. Hsu et al. / Molecular & Biochemical Parasitology 192 (2013) 55–59

Under normal conditions in E. granulosus infections (i.e. excluding the cases of hydatid rupture), the parasite components in contact with the host immune system are only the soluble secreted molecules (so far largely uncharacterized) and the LL. The LL mucins have only very short non-glycosylated domains [7], and the LL is apparently devoid of non-mucin proteins [3]. In consequence, the relatively simple glycome of the LL constitutes the major part of the “identity card” that this parasite presents to the host immune system. Decoding of pathogen- and, more generally, microorganism-derived molecules by the innate immune system shapes the type of response mounted by the adaptive immune system in each case. The interaction of microorganism-derived carbohydrates with host lectin receptors is an important part of this decoding process. This applies to platyhelminth pathogens such as the well-studied schistosomes [8]. Further, in the genus Taenia of larval taeniids, carbohydrates are known to deliver signals to innate immunity [9]. Hence, the identity of the host innate lectin receptors that bind the LL is a central question in the immunology of larval Echinococcus infections. Importantly, these infections are characterized by the induction of tolerogenic circuits in the hosts’ immune systems [1,10–12], through mechanisms that largely remain unexplored at the molecular level. We have commenced to address the decoding of the LL carbohydrates by host innate lectins, with the help of a substantial panel of recombinant innate immune receptors (IIRs), expressed as human IgG1 Fc fusions, and therefore in dimeric presentation [13]. A soluble preparation containing the whole of the E. granulosus LL mucins was coated onto ELISA plates and probed with the recombinant IIRs. A total of 40 IIRs were assayed, of which 36 contain a known carbohydrate recognition domain, and 31 belong to the C-type lectin family (Fig. 1). All the receptors are of human origin, except for one murine receptor (Supplementary Table S1). The monoclonal antibody E492, which recognizes the P1 motif mentioned above [5], was used to confirm adsorption of the LL mucins to the plates, while human IgG1 was used to estimate the level of non-specific binding. Strikingly, among all the IIRs tested, only the mouse Kupffer cell receptor (mKCR; systematic name CLEC4F) gave a signal that was clearly in excess of that of IgG1 (OD450 ∼1.1 vs ∼0.3) (Fig. 1). Another five lectins, namely asialoglycoprotein receptor isoform 1 (ASGR-1; CLEC4H1), macrophage galactose-specific lectin (MGL; CLEC10A), FCN2 (L-Ficolin), NCR2 (NKp44) and NCR3 (NKp30), gave borderline signals (OD450 ∼ 0.25–0.3). The possible biological significance of these interactions should be further analyzed using the native receptors and/or cells expressing them, as our recombinant system cannot be expected to imitate avidity effects, potentially very important in recognition of the LL mucins. In contrast to the results with the LL mucins, 9 out of 17 IIRs show clear binding to the polysaccharides extracted from Reishi, a fungus often used as dietary supplements, in the first application of the same profiling system [13]. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.molbiopara.2013. 12.001. As expected, KCR binding to the LL mucins was calciumdependent (Fig. 2a). The binding was specific, as it could be outcompeted by excess solubilized LL mucins. LL material subjected to strong non-specific proteolysis, which leaves only the highly glycosylated mucin domains [3], was as effective as a competitor as untreated material (Fig. 2b), confirming that the KCR binds the mucin glycans. The KCR bound to the LL mucins of E. multilocularis at least as strongly as to those of E. granulosus (Fig. 2c). In fact, the estimated apparent affinity was higher for E. multilocularis LL than for E. granulosus LL (Kd ∼4 vs Kd ∼13, respectively). Again, previous proteolysis caused very little change in KCR binding, for both species. Given the well-known affinity of the KCR for sugars terminated in

Gal (and GalNAc) [14,15], it is not surprising that the receptor bound to the Gal-rich Echinococcus LL glycans. Further, sugars presented as O-glycopeptides are better KCR targets than the same sugars carried on artificial spacers, suggesting that mucin O-glycans (endogenous or not) are likely KCR counterparts in vivo [14]. In addition, as glycans on the LL mucins are probably presented in a very dense array [2,7], bivalent interactions with the chimeric KCR in this study probably contribute to the overall avidities observed. Multimeric and very high avidity interactions are likely with the native KCR, which is trimeric [16], and further presented in multiple copies on the cell surface. The O-glycans present in the E. multilocularis LL mucin fraction mentioned above have been synthesized (with either a 2-(trimethylsilyl)ethyl group or a spacer and biotin tag derivatizing the reducing end anomeric hydroxyl) [17,18]. E. multilocularis LL O-glycans share most non-reducing terminal structures with the E. granulosus ones, or in cases are identical to them [2]. We used this tool to study the recognition of the specific carbohydrate motifs present in the LL by the KCR. In a direct binding format using the biotin-tagged glycans, strong signals were obtained for glycans “H” and “I” (Fig. 2d), which share the sequence Gal␣1-4Gal␤13GalNAc. A third glycan displaying similar binding strength was glycan “J”, which lacks the Gal␣1-4 “cap” on the main chain but instead carries a Gal␤1-4 residue on the core 2 GlcNAc residue, so that a terminal N-acetyl-lactosamine (LacNAc) motif results. Glycan “K”, which carries the P1 motif, showed slightly weaker but still strong binding. Substantially weaker binding was observed for glycan “G”, corresponding to non-decorated core 2. A slightly different panel of (nonbiotinylated) glycans was tested in a competition format using plates coated with solubilized E. granulosus LL (Fig. 2e). The strongest competitor was glycan “D”, corresponding to the same carbohydrate structure as biotinylated glycan “J” already mentioned as a strong binder. The glycans “A” and “F”, corresponding to biotinylated glycans “G” and “K”, respectively, were weaker competitors (the order of binding strength being inverted with respect to the previous assay format). Finally, only very weak competition was displayed by glycan “L” (Gal␤1-3Gal␤1-3GalNAc). Although the two assay formats evidently did not measure exactly the same parameters (bivalent interactions can be expected to contribute to the overall signal in the direct binding format only), and not all glycan structures were represented in both (biotin-tagged and non-tagged) panels, an overall picture results as follows. Out of the Echinococcus LL glycan motifs assayed, the KCR binds most strongly to terminal LacNAc and to the Gal␣1-4Gal␤1-3GalNAc sequence. The Gal␣1-4Gal␤1-4GlcNAc trisaccharide (P1 motif) is a slightly weaker target, while terminal Gal␤1-3Gal␤ is a very weak target. This scenario is in broad agreement with published data obtained by glycan array [14,15] showing that (i) both O-glycans and N-glycans carrying single or repeated terminal LacNAc motifs are among the best KCR targets identified, (ii) the disaccharide Gal␤1-3Gal␤ is not a KCR target. However, the binding to the glycan carrying the P1 trisaccharide observed by us contrasts with the lack of binding observed previously [14] to the same trisaccharide present in a different context (i.e. directly bound to a spacer, as opposed to being part of a core 2-based O-glycan). The previous glycan array data additionally indicate that KCR binds to glycopeptides carrying the non-decorated core 1, which is an abundant Echinococcus LL glycan [2,4,6], but is not represented in our panel. Conversely, the glycan arrays against which KCR was previously tested do not include structures with the Gal␣1-4Gal␤1-3GalNAc sequence, identified as a good KCR ligand by our data. In sum, the binding of the KCR to the native Echinococcus LL mucins must be based on the recognition of: (i) non-decorated core 1 (as suggested from the mentioned glycan array data), (ii) the P1 motif, and (iii) capped main chain structures, terminated in Galp␣1-4Gal␤1-3GalNAc/Gal. With respect to this third contribution, it is worth noting that our

T.-L. Hsu et al. / Molecular & Biochemical Parasitology 192 (2013) 55–59

57

Fig. 1. Innate immune receptor profiling of the E. granulosus LL. Hydatid walls (which are quantitatively essentially only LL) from natural E. granulosus pig infections were processed as described [3]. Briefly, the material was dehydrated, pulverized, re-hydrated into buffer containing EDTA, subjected to reduction and alkylation of disulphide bonds, extensively washed with water and dissolved by sonication. In addition, 2 M NaCl was included in the EDTA step to remove adsorbed host proteins. Nunc MaxisorpTM plates were coated with the resulting soluble mucin preparation (1 ␮g total dry mass/0.1 ml per well in a 96-well plate format) and blocked with Tris-buffered saline containing 0.05% (w/v) Tween 20, 1% (w/v) bovine serum albumin and 0.5× CarbofreeTM (Vector Labs, Burlingame, CA). Plates were probed with IIRs expressed as Fc fusion chimeras [13], dissolved at 0.5 ␮g/ml in Tris-buffered saline containing 2 mM MgCl2 , 2 mM CaCl2 , 0.05% (w/v) Tween 20 and 1× CarbofreeTM , followed by peroxidase-conjugated anti-human IgG. The E492 monoclonal antibody, which binds to the blood P1 antigen motif carried on the LL O-glycans [5], was included as a positive control (and in those wells an anti-mouse IgG conjugate used). Human IgG1 was included as a negative control. Bars represent SD of duplicates.

panel comprises glycans terminated in Gal␣1-4Gal␤1-3GalNAc, as present in E. multilocularis [6], but not glycans terminated in Gal␣1-4Gal␤1-3Gal, corresponding to the longer variants present in E. granulosus, as mentioned above. Thus the possibility that KCR binds to E. granulosus glycans carrying the Gal␣1-4Gal␤13Gal sequence, although likely, is a hypothesis. With respect to O-glycans bearing terminal LacNAc, although they are very good KCR ligands, their contribution to overall LL recognition may be negligible. Such glycans are conspicuously absent from the E. granulosus LL [4,5], and although present in E. multilocularis [6], their representation in the crude LL of this species is quantitatively very small (Romina Rovetta and Díaz, unpublished results). As we have suggested previously [4,5], there must be a strong selective force (unrelated to the KCR) impeding the exposure of LacNAc termini in the LL. The KCR is a type 2 C-type lectin featuring a long “neck” region mediating receptor trimerization, which is predicted to cause the lectin domains to project some 50 nm away from the cell surface [16]. It is expressed only in Kupffer cells, the resident liver macrophages, and in monocytes infiltrating the liver in inflammatory contexts [15,19]. Kupffer cells are in contact with the blood that circulates through the liver sinusoids and have a role in clearing particles, including microorganisms, in particular those carried from the intestine by portal circulation [20]. The KCR is present in rodents but not in humans [16]. As mentioned, E. multilocularis but not E. granulosus uses rodents as intermediate hosts. However, rodents are the probable ancestral intermediate hosts for the whole of the taeniidae including the genus Echinococcus [21]. Further, their ancestral target organ is liver. Hence, we propose that an interaction with the rodent KCR could be an ancestral, evolutionarily selected feature of the LL. Since the LL is not present in the early establishing parasite, the LL–KCR interaction cannot play a role in an initial homing to the liver. Macromolecular aggregates and/or particles from the LL are shed during larval growth [22]. The results of our present study, together with the known capacity of Kupffer cells to endocytose large macromolecules and particles

via the KCR, strongly suggests that components shed from the LL in Echinococcus-infected rodents must be taken up selectively by Kupffer cells. As the LL has a great capacity to adsorb conventional proteins, the KCR–LL interaction may help target protein antigens leaked from Echinococcus larvae to Kupffer cells. Non-conventional (i.e. extra-lymphoid tissue) antigen presentation in the liver sinusoids by Kupffer cells and other resident liver cells is tolerogenic [20]. Alternatively or complementarily, engagement of the KCR could deliver signals to Kupffer cells to potentiate a tolerogenic phenotype, although there is as yet no experimental support for this possibility. We had previously put forward the liver-targeting/conditioning hypothesis outlined above in a slightly broader formulation [1] that incorporated the ASGR in addition to the KCR. Indeed, the two receptors bind to largely overlapping sets of carbohydrates in glycan array experiments [14]. As the ASGR is expressed in hepatocytes of all mammalian species studied, the broader formulation of our hypothesis would be applicable to E. granulosus infections in its whole range of intermediate host species. In our experiments, isoform 1 of the ASGR gave borderline binding to the LL mucins, and isoform 2 did not bind significantly (Fig. 1). In spite of the similar specificities displayed by ASGR and KCR, there are also known differences that could cause them to interact to very different extents with the LL mucins. The ASGR is much more selective than the KCR for sugars terminated in GalNAc (absent from the E. granulosus LL) over those ending in Gal (abundant in the LL) [16,23]. In addition, the ASGR has a strong preference for glycans carrying three and four terminal Gal/GalNAc residues [24], very scarce in the E. granulosus LL [4,5], while the KCR may instead bind preferentially to multiple glycans carried on the same glycoprotein molecule [16], as it is the case in the LL mucins. Further, we observed that ASGR bound much more poorly to the P1 motif than KCR: the binding signal ratio between glycan “K” (P1 motif) and glycan “J” (a known ligand for both KCR and ASGR [14]) was 0.3 for both ASGR-1 and ASGR-2 vs 0.8 for KCR (data not shown). On the other hand, as our ASGR constructs gave rather weak binding even towards a known

58

T.-L. Hsu et al. / Molecular & Biochemical Parasitology 192 (2013) 55–59

Fig. 2. Interaction of KCR with the Echinococcus LL mucin O-glycans. Soluble E. granulosus and E. multilocularis LL mucins were prepared in parallel, by the same procedure as in Fig. 1. The starting material for E. multilocularis consisted of larvae produced by in vitro culture that were subjected to trypsin-EDTA digestion and sieved in order to separate away parasite cells, thus leaving behind the LL. Portions from each material were subjected to proteolysis with pronase (as in Ref. [3]) for 24 h, before the disulfide reduction/alkylation step. (a) Binding of KCR to E. granulosus LL mucins in the presence or absence of Ca2+ . (b) Competition of the binding of KCR to E. granulosus LL mucins by the same solubilized material, either untreated or treated with pronase. (c) Binding of KCR to E. granulosus or E. multilocularis LL mucins (indicated as “Eg” and “Em”, respectively), either untreated or treated with pronase. (d) Binding of KCR to synthetic biotin-tagged glycans [18] immobilized onto the ELISA plate via streptavidin. (e) Competition of the binding of KCR to E. granulosus LL mucins by soluble synthetic glycans [17]. In parts (d) and (e), Kd values and their 95% confidence intervals (between parentheses), estimated by non-linear fit assuming one-site specific binding with Hill slope using the GraphPad Prism package, are shown; Kd values in parts (d) and (e) have different units (ng/ml and ␮M, respectively) and are further not comparable given the different experimental formats. Glycans have been represented following the Consortium for Functional Glycomics nomenclature (http://www.functionalglycomics.org/static/consortium/Nomenclature.shtml), in which clear circles, clear squares and dark squares are Gal, GalNAc and GlcNAc residues respectively. Bars correspond to SD of duplicates.

ligand (glycan “J”), experiments with cells expressing the native ASGR are needed to resolve this issue. Even if the ASGR does not bind the LL, it is plausible that a liver-targeting mechanism may exist in non-rodent intermediate hosts: the existence of paralogs that may fulfill the role of the KCR, to our knowledge, has neither been ruled out in humans [25] nor studied in ungulates, the main E. granulosus intermediate hosts.

The human macrophage galactose-type lectin (MGL, CLEC10A), related to the KCR and ASGR, displayed borderline binding to the LL mucins. This situation is similar to that discussed above for ASGR-1, including a previously reported preference for GalNAc over Gal [26], and a lack of binding signal when confronted with the P1 motif in our experiments. Two MGL paralogs exist in mouse, not included in our IIR panel. Mouse MGL2 is specific for GalNAc [27], and therefore

T.-L. Hsu et al. / Molecular & Biochemical Parasitology 192 (2013) 55–59

its capacity to bind the LL carbohydrates can be expected to be similar to that of human MGL. Mouse MGL1, which is specific for the Lewisx motif, should not be expected to bind the LL [27]. The most striking finding of our study is the lack of interaction of the LL glycans with over 25 host innate immune lectin receptors. Although our analysis excludes several lectins including the galectins, which bind glycans terminated in ␤-Gal and can participate in pathogen recognition [28], it is possible that the LL glycans may not be recognised by systemic innate immune lectins. This scenario would be in agreement with our observation that the phenotypic changes elicited by particles from the E. granulosus LL in mouse bone marrow-derived macrophages and dendritic cells are independent of the LL carbohydrates [29]. Taenia larval surface carbohydrates are active with respect to systemic innate immunity [9]. As these are predominantly N-glycans, it is conceivable that the N-glycans present in Echinococcus structures/stages other than the LL (structurally related to their Taenia counterparts) may be more “signal-rich” for systemic innate immunity than the LL mucin O-glycans [1,9,30].

[10]

[11]

[12]

[13]

[14]

[15]

[16] [17]

[18]

Acknowledgments The authors are grateful to Sylvia Dematteis and Gustavo Mourglia-Ettlin for kindly providing E492 monoclonal antibody. This work was supported by Academia Sinica, Taiwan, and Universidad de la República, Uruguay [CSIC I + D grant 2008 no 404 to A.D.]. References [1] Díaz A, Casaravilla C, Allen JE, Sim RB, Ferreira AM. Understanding the laminated layer of larval Echinococcus II: immunology. Trends Parasitol 2011;27: 264–73. [2] Díaz A, Casaravilla C, Irigoín F, Lin G, Previato JO, Ferreira F. Understanding the laminated layer of larval Echinococcus I: structure. Trends Parasitol 2011;27:204–13. [3] Casaravilla C, Díaz A. Studies on the structural mucins of the Echinococcus granulosus laminated layer. Mol Biochem Parasitol 2010;174:132–6. [4] Díaz A, Fontana EC, Todeschini AR, Soulé S, Gonzalez H, Casaravilla C, et al. The major surface carbohydrates of the Echinococcus granulosus cyst: mucintype O-glycans decorated by novel galactose-based structures. Biochemistry 2009;48:11678–91. [5] Lin G, Todeschini AR, Koizumi A, Neves JL, Gonzalez H, Dematteis S, et al. Further structural characterization of the Echinococcus granulosus laminated layer carbohydrates: the blood antigen P1 -motif gives rise to branches at different points of the O-glycan chains. Glycobiology 2013;23:438–52. [6] Hulsmeier AJ, Gehrig PM, Geyer R, Sack R, Gottstein BPD, et al. A major Echinococcus multilocularis antigen is a mucin-type glycoprotein. J Biol Chem 2002;277:5742–8. [7] Parkinson J, Wasmuth J, Salinas G, Bizarro CV, Sanford C, Berriman M, et al. A transcriptomic analysis of Echinococcus granulosus larval stages: implications for parasite biology and host adaptation. PLoS Negl Trop Dis 2012;6: e1897. [8] Meevissen MH, Yazdanbakhsh M, Hokke CH. Schistosoma mansoni egg glycoproteins and C-type lectins of host immune cells: molecular partners that shape immune responses. Exp Parasitol 2012;132:14–21. [9] Terrazas CA, Gomez-García L, Terrazas LI. Impaired pro-inflammatory cytokine production and increased Th2-biasing ability of dendritic cells exposed to

[19]

[20] [21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

59

Taenia excreted/secreted antigens: a critical role for carbohydrates but not for STAT6 signaling. Int J Parasitol 2010;40:1051–62. Mejri N, Muller N, Hemphill A, Gottstein B. Intraperitoneal Echinococcus multilocularis infection in mice modulates peritoneal CD4+ and CD8+ regulatory T cell development. Parasitol Int 2011;60:45–53. Nono JK, Pletinckx K, Lutz MB, Brehm K. Excretory/secretory-products of Echinococcus multilocularis larvae induce apoptosis and tolerogenic properties in dendritic cells in vitro. PLoS Negl Trop Dis 2012;6:e1516. Pan W, Zhou HJ, Shen YJ, Wang Y, Xu YX, Hu Y, et al. Surveillance on the status of immune cells after Echinnococcus granulosus protoscoleces infection in Balb/c mice. PLoS ONE 2013;8:e59746. Hsu TL, Cheng SC, Yang WB, Chin SW, Chen BH, Huang MT, et al. Profiling carbohydrate–receptor interaction with recombinant innate immunity receptor-Fc fusion proteins. J Biol Chem 2009;284:34479–89. Coombs PJ, Taylor ME, Drickamer K. Two categories of mammalian galactosebinding receptors distinguished by glycan array profiling. Glycobiology 2006;16:1C–7C. Yang CY, Chen JB, Tsai TF, Tsai YC, Tsai CY, Liang PH, et al. CLEC4F is an inducible C-type lectin in F4/80-positive cells and is involved in alphagalactosylceramide presentation in liver. PLoS ONE 2013;8:e65070. Fadden AJ, Holt OJ, Drickamer K. Molecular characterization of the rat Kupffer cell glycoprotein receptor. Glycobiology 2003;13:529–37. Koizumi A, Hada N, Kaburaki A, Yamano K, Schweizer F, Takeda T. Synthetic studies on the carbohydrate moiety of the antigen from the parasite Echinococcus multilocularis. Carbohydr Res 2009;344:856–68. Koizumi A, Yamano K, Schweizer F, Takeda T, Kiuchi F, Hada N. Synthesis of the carbohydrate moiety from the parasite Echinococcus multilocularis and their antigenicity against human sera. Eur J Med Chem 2011;46:1768–78. Haltiwanger RS, Lehrman MA, Eckhardt AE, Hill RL. The distribution and localization of the fucose-binding lectin in rat tissues and the identification of a high affinity form of the mannose/N-acetylglucosamine-binding lectin in rat liver. J Biol Chem 1986;261:7433–9. Thomson AW, Knolle PA. Antigen-presenting cell function in the tolerogenic liver environment. Nat Rev Immunol 2010;10:753–66. Hoberg EP, Jones A, Rausch RL, Eom KS, Gardner SL. A phylogenetic hypothesis for species of the genus Taenia (Eucestoda: Taeniidae). J Parasitol 2000;86:89–98. Barth TF, Herrmann TS, Tappe D, Stark L, Gruner B, Buttenschoen K, et al. Sensitive and specific immunohistochemical diagnosis of human alveolar echinococcosis with the monoclonal antibody Em2G11. PLoS Negl Trop Dis 2012;6:e1877. Iobst ST, Drickamer K. Selective sugar binding to the carbohydrate recognition domains of the rat hepatic and macrophage asialoglycoprotein receptors. J Biol Chem 1996;271:6686–93. Lee YC, Townsend RR, Hardy MR, Lonngren J, Arnarp J, Haraldsson M, et al. Binding of synthetic oligosaccharides to the hepatic Gal/GalNAc lectin. Dependence on fine structural features. J Biol Chem 1983;258:199–202. Falasca L, Bergamini A, Serafino A, Balabaud C, Dini L. Human Kupffer cell recognition and phagocytosis of apoptotic peripheral blood lymphocytes. Exp Cell Res 1996;224:152–62. van Vliet SJ, van Liempt E, Saeland E, Aarnoudse CA, Appelmelk B, Irimura T, et al. Carbohydrate profiling reveals a distinctive role for the C-type lectin MGL in the recognition of helminth parasites and tumor antigens by dendritic cells. Int Immunol 2005;17:661–9. Tsuiji M, Fujimori M, Ohashi Y, Higashi N, Onami TM, Hedrick SM, et al. Molecular cloning and characterization of a novel mouse macrophage C-type lectin, mMGL2, which has a distinct carbohydrate specificity from mMGL1. J Biol Chem 2002;277:28892–901. van den Berg TK, Honing H, Franke N, van Remoortere A, Schiphorst WE, Liu FT, et al. LacdiNAc-glycans constitute a parasite pattern for galectin-3-mediated immune recognition. J Immunol 2004;173:1902–7. Casaravilla C., Pittini A., Rückerl D., Seoane P.I., Jenkins S.J., MacDonald A.S., et al., Unconventional maturation of dendritic cells induced by particles from the laminated layer of larval Echinococcus granulosus, submitted for publication. Khoo KH, Nieto A, Morris HR, Dell A. Structural characterization of the N-glycans from Echinococcus granulosus hydatid cyst membrane and protoscoleces. Mol Biochem Parasitol 1997;86:237–48.