C-type lectins on dendritic cells: key modulators for the induction of immune responses

DCs and antigen-recognition receptors

Immature DCs are well positioned throughout the body, in skin and mucosal tissue, to sense and capture invading pathogens for efficient antigen presentation to naïve T-cells present in draining lymph nodes. DCs have the capacity to sample and migrate to lymph-node-residing T-cells to direct the differentiation of helper T-cells into distinct effector cell subpopulations. These include Th1, Th2 and Th17 effector T-cells against intracellular and extracellular pathogens, and Tregs (regulatory T-cells) which sustain tolerance to self-antigens and dampen excessive immune responses [7].

Immature DCs express an abundant variety of PRRs to interact with invading pathogens that recognize characteristic molecular patterns present within microbial carbohydrates, lipids and nucleic acids [8], including numerous TLRs [9,10] and CLRs [10]. TLRs relay information from the interacting pathogen to DCs through intracellular signalling cascades, thereby eliciting appropriate cellular processes, such as DC maturation and/or the induction of pro-inflammatory cytokines [IL (interleukin)-12 or IFN (interferon) γ] [8,11]. In contrast, CLRs recognize carbohydrate structures on pathogens and internalize pathogens for antigen processing and presentation, without induction of DC maturation [2,12]. So far more than 15 CLRs have been identified on DCs and macrophages [5]. Although TLRs and CLRs recognize different determinants and have distinct functions, various studies suggest that CLRs may also modulate immune reactions through cross-talk with other receptors and especially with TLRs [2–4,13,14]. This indicates that the outcome of an immune response is determined by the balance between triggering of the two receptor families.

We have studied the expression and function of the DC-specific C-type lectins, DC-SIGN (DC-specific intracellular adhesion molecule-3 grabbing non-integrin) and MGL (macrophage galactose/N-acetylgalactosamine-specific C-type lectin). DC-SIGN and MGL are expressed on different subsets of DC in vitro and in vivo. MGL is mainly expressed on tolerogenic DC, whereas DC-SIGN is highly expressed on immature monocyte-derived DCs. CLRs are highly expressed by immature DCs, whereas their expression drops as DCs mature.

Most CLRs function as antigen receptors that are involved in antigen capture and presentation [12]. Endocytosis by CLRs is guided by their intracellular internalization motifs, whereas some CLRs contain ITIM (immunoreceptor tyrosine-based inhibitory motif)- or ITAM (immunoreceptor tyrosine-based activation motif)-like motifs in their cytoplasmic domains, illustrating potential immuno-suppressive or -activation functions of these receptors [2]. CLRs possess different numbers of CRDs (carbohydrate-recognition domains) ranging from a single domain [e.g. DC-SIGN, Dectin-1, MGL or DCIR (DC immunoreceptor)] to eight or ten different CRDs [e.g. MR (mannose receptor) and DEC-205]. Some CLRs recognize N-linked glycans, whereas others interact specifically with O-linked glycans, but the exact specificity of the different CLRs for their glycan ligands remains to be determined. Although some CLRs recognize monosaccharides, such as mannose, fucose or galactose, others recognize more complex sugar moieties expressed on glycoproteins and glycolipids. The specificity of CLRs appears to be determined by multimerization of the receptor, the branching of the carbohydrate chains and the protein backbone of the glycoprotein that exposes the carbohydrate structure [1] (see http://www.functionalglycomics.org/static/index.shtml). O-linked structures are often exposed on collagens, mucins and some pathogens, whereas N-linked structures are present on the vast majority of glycoproteins in the body as well as on pathogens, such as retroviruses, that use the glycosylation machinery of the host for their survival and spread.

Several pathogens that target DC-SIGN, MR and Dectin-1 seem to subvert the function of these CLRs [4,15], either by inhibition of antigen presentation or by modification of T-cell responses. The physiological function of CLRs may be recognition of glycosylated self-antigens for homoeostatic control [16,17]. The in vivo localization of CLRs on immature DCs in peripheral tissues is consistent with an important function in the clearance of self-antigens and in tolerance induction. DC-SIGN and the MR are highly expressed by DCs in placenta at the maternal–fetal interface, a site where the maintenance of immune tolerance plays a central role [18]. Indeed, both in vitro and in vivo antigen targeting to CLRs on immature DCs leads to tolerance, suggesting that tolerance is maintained by immature DCs that have captured self-glycoproteins through CLRs, leading to the induction of Treg [19]. Thus the continuous interaction of carbohydrate determinants on self-glycoproteins with CLRs on resident APCs may be important for homoeostatic control.

CLRs and recognition of cancer

In particular, the activation and antigen presentation by DCs plays a key role in the initiation of anti-tumour responses. However, tumour cells use several strategies to dampen both antigen presentation and T-cell responses initiated by DCs. TAAs (tumour-associated antigens) interacting with DCs may modify DC function (such as production of immunosuppressive cytokines) or inhibit DC migration to tumour sites.

TAAs are often auto-antigens that become deregulated, and changes in glycosylation often accompany oncotransformation. Such TAAs include the CEA (carcinoma embryonic antigen) and MUC1 (mucin 1) [20–22]. Both antigens are expressed on cells of normal colonic mucosa and epithelial cells; however, during oncotransformation, aberrant glycosylation of CEA is a common phenomenon that accompanies colon carcinoma progression [23–25]. These changes include increased expression of the Lewis blood group family of antigens, particularly Le^x and Le^y, that are often associated with a poor prognosis. How these post-translational modifications contribute to tumour cell dissemination and disease severity is not fully understood. Previous findings have demonstrated that DCs recognize these modified glycosylations on CEA or MUC1 through CLRs such as DC-SIGN and MGL respectively, while these CLRs do not interact with ‘normal’ CEA or MUC1 from colon tissue [26,27]. Strikingly, tumour-associated CEA from five adenocarcinoma cell lines as well as from colorectal cancer patients contains Le^x and Le^y antigens, and our recent findings demonstrate that DC-SIGN, in particular, is involved in the recognition of tumour-associated CEA by DC [28]. Immature DC expressing DC-SIGN, but not mature DC, are localized within the tumour, indicating an immune-silenced environment [29]. The fact that TAAs, such as CEA, are secreted during metastasis indicates that secretion of modified glycosylated CEA that targets DC-SIGN can lead to systemic tolerance in colon cancer patients. Indeed, initial experiments demonstrate that CEA-containing serum from breast carcinoma patients reduces IL-10 and IL-12 production by DCs and DC-induced T-cell polarization.

Targeting antigens to DCs to improve anti-tumour responses

Several studies have demonstrated that targeting antigens to CLRs can result in strong anti-tumour responses. An elegant study demonstrated that in vivo targeting of CLRs in the mouse can induce antigen-specific auto-immunity when the antigen was coupled to an anti-DEC-205 antibody [30]. In contrast, when antigen coupled to DEC-205 was simultaneously combined with a strong DC activator such as a TLR stimulus, a strong antigen-specific immune activation was induced that resulted in an anti-tumour response and clearance of the tumour [31–34]. Furthermore, other CLRs have also been used for targeting purposes, resulting in strong induction of immunity. However, differences can be observed between CLRs that are targeted with antigens, as some are more prone to induce CD4 T-cell responses, whereas others process antigens in such a way that strong CD8⁺ T-cell responses are induced.

Clearly, this indicates that the interplay and balance between CLR and TLR signalling is crucial for the outcome of any immune response and may lead to the induction of either tolerance or immunity.

Concluding remarks

Although the function of many CLRs remains unknown, it is becoming clear that CLRs have a key role in the maintenance of homoeostasis. Identification of self-ligands and their specific carbohydrate structures will be essential to understand the cellular function of many CLRs. Lessons concerning the function of CLRs can be learned from studying pathogens or tumour antigens that specifically target CLRs to evade immune surveillance.

In clear contrast, several studies have demonstrated the potency of targeting antigens to CLR to improve antigen-specific immune responses. This can only be accomplished when adjuvant triggers DC maturation simultaneously. The fact that CLRs are highly specific internalization receptors that facilitate and enhance antigen loading on MHC class I and II has initiated new directions to optimize DC vaccination strategies to enhance anti-tumour CD4⁺ and CD8⁺ T-cell responses. Although this strategy shows its potency in mouse model systems, it will depend on future research in humans as to whether it will find its way to the clinic to treat cancer.