Dendritic Cells in Myelodysplastic Syndromes

From Pathogenesis to Immunotherapy

Nathalie Kerkhoff; Hetty J Bontkes; Theresia M Westers; Tanja D de Gruijl; Shahram Kordasti; Arjan A van de Loosdrecht


Immunotherapy. 2013;5(6):621-637. 

In This Article

Normal DC Biology

DCs originate from the bone marrow and are known to act as regulators of immunity against pathogens and tolerance to self-antigens. They mainly function as professional APCs.[19,20] In an immature state, DCs sample their environment for antigens. Upon 'danger' signals or exposure to so-called pathogen-associated molecular patterns (PAMPs), DCs will capture and process antigens and, subsequently, their maturation is initiated. This complex process includes changes in morphology to allow migration to lymphoid organs, expression of essential receptors and costimulatory molecules on the cell surface, and a shift from a capturing state to a cytokine-secreting and antigen-presenting profile. DCs that have migrated to draining lymph nodes will present the processed antigen on MHC and MHC-like molecules to naive T cells, leading to the formation of effector cells and, hence, the required immune response (Figure 1).

Figure 1.

Dendritic cell biology. In an immature state, DCs sample their environment for foreign antigens. After capturing antigens they process them and undergo a process called maturation. Environmental cues influence DC priming and thereby the eventual outcome of effector T-cell type. During the maturation process, they migrate to lymph nodes where they present the antigens to naive CD4+ T cells or in the context of cross-presentation to CD8+ T cells. To instruct naive T cells, three signals are required. The first signal is provided by the interaction of the MHC molecule on the DC with the TCR on the T cell. Secondly, costimulation is essential to induce effector T cells. Without costimulation, T cells become anergic. The last polarizing signal is mediated by factors secreted by DCs and environmental cells. Different effector outcomes are observed in various MDS risk groups.
CLR: C-type lectin receptor; CTL: Cytotoxic T lymphocyte; DC: Dendritic cell; IFN: Interferon; MDS: Myelodysplastic syndrome;PAMP: Pathogen-associated molecular pattern; TCR: T-cell receptor; TLR: Toll-like receptor.

DC Subsets

Different human DC subsets residing in various tissue types are identified. DCs and their precursors are present in the blood, bone marrow, peripheral tissues and lymphoid tissue. The main focus of this review is on circulating DCs. In peripheral blood, four DC subsets have been described: plasmacytoid DCs (pDCs), two types of myeloid DCs (CD1c+ and CD141+), also called conventional DCs, and 'slanDCs' (Table 2). To phenotypically characterize the different DC subsets, a range of markers can be used. In general, all DCs express HLA-DR and lack other markers of leukocyte lineages (i.e., CD3, CD14 and CD19).[21–23]

To recognize PAMPs, DCs are provided with a broad repertoire of pattern-recognition receptors, such as Toll-like receptors (TLRs) and C-type lectin receptors. The expression pattern of these receptors varies between DC subsets, which implies functional differences. Selective expression of TLR7 and 9 distinguishes pDCs from other DC subsets.[24–26] These two receptors expressed on pDCs are important in the recognition of viruses. Engagement by ribonucleotide homologs, such as loxoribine and synthetic ssRNA sequences, and unmethylated CpG-rich DNA motifs, respectively, upregulates the expression of type I interferon (IFN) genes. In line with this, pDCs are the main producers of type I IFN-α/β;[27,28] pDCs are further characterized by the combined expression of the IL-3 receptor α-chain (CD123) and the type II C-type lectin receptor BDCA-2 (CD303).[29,30]

The myeloid DCs comprise a group of two different subsets. Both subsets express high levels of the integrin α-X chain of complement receptor 4 (CD11c). This marker is not restricted to DCs, but is also expressed by monocytes and granulocytes. Therefore, other markers are required to identify and distinguish the myeloid DC subsets. Myeloid DCs are generally divided in a BDCA-1 or CD1c+ subset and a very minor BDCA-3 or CD141+ subset.[22,29] Regarding TLR-expressing profiles, the CD141+ subset expresses very low levels of TLR2 and 7 and does not display TLR4 and 5. Furthermore, TLR3 and 10 are more highly expressed on CD141+ DCs than on the CD1c+ subset.[24,31] The two myeloid DC subsets have different functional capacities. After stimulation with TLR3-ligand (polyI:C), CD141+ DCs show increased upregulation of costimulatory molecules (CD80 and CD86) compared with CD1c+ DCs. Moreover, their cytokine and chemokine production varies. The CD141+ subset mainly secretes proinflammatory cytokines and chemokines (i.e., CCL3, CCL4, CCL5, IL-6, CXCL8, TNF-α and CXCL10) upon TLR3 triggering. However, the quantity of cytokines secreted is low compared with the production by CD1c+ DCs.[31] TLR3 triggering also induces the production of IFN-β, which is associated with an enhanced Th1 response and cross-priming.[32,33] CD141+ DCs secrete significantly higher amounts of IFN-β than CD1c+ DCs.[31,34] Additionally, myeloid DCs are well known for their production of high levels of IL-12p70, thereby creating a highly inflammatory environment, skewing towards Th1 responses. In vitro studies showed that CD141+ DCs are superior in IL-12p70 production than CD1c+ DCs. Moreover, they induce enhanced Th1 responses as assessed by IFN-γ production in a mixed leukocyte reaction (MLR).[31,34] Despite this proinflammatory profile, CD141+ DCs are also associated with tolerance induction. In their immature state, they express higher levels of the immune-regulatory enzyme indoleamine 2,3-dioxygenase, which inhibits T-cell proliferation.[35] In AML, high indoleamine 2,3-dioxygenase expression is correlated with enhanced Treg conversion and reduced overall and relapse-free survival.[36,37]

The other human DC subset in peripheral blood is known as slanDCs.[38] This CD16+ and M-DC8+ (6-sulfo LacNAc or slan) subset can produce high levels of IL-12 after maturation, and, in response to lipopolysaccharide, they secrete TNF-α and induce primary T-cell responses.[39,40] Furthermore, slanDCs induce proliferation and IFN-γ production of NK cells and also promote NK cell-mediated tumor-specific cytotoxicity.[41] This immunogenic profile assigns this subset as a potential target for immunotherapeutic strategies in cancer treatment as demonstrated by Bippes et al..[42]

Importantly, different DC subsets exhibit phenotypical and functional differences (Table 2). This indicates distinct properties and roles within the immune system and could lead to a variety of T-cell responses. In the perspective of MDS, this could also be relevant. In low-risk MDS, an anti-inflammatory response is required, whereas in high-risk MDS, an inflammatory response is desired in order to overcome the tolerogenic environment. DC subsets that have an outstanding ability to induce inflammation are, for example, more attractive for targeted therapies in high-risk MDS than DC subsets with less inflammatory features, which are preferred in low-risk disease.

DCs in the Induction of T-cell Polarization

Different stimuli induce naive T-cell polarization. At least four types of CD4+ T cells are known: Th1, Th2, Th17 and Tregs.[43] DC–T cell interactions are described as a three-signal process (Figure 1). As stated above, captured antigens will be processed by DCs. The first signal in T-cell polarization is provided by the ligation of the T-cell receptor by the peptide presented in the MHC class II molecule on the surface of DCs. Second, the polarization process is mediated by the costimulatory molecules, CD80 and CD86, and their interaction with CD28 on the T cell. Without costimulation, T cells become anergic or will be skewed to a tolerogenic state. Signal three is the main polarizing factor and is mediated by soluble and membrane-bound factors, such as different complement factors,[44] interleukins and chemokines. These factors are produced by DCs or by environmental cells and highly depend on the setting and type of PAMP that initially triggered DC priming.[45] Key players are the Th1-cell factors (i.e., IL-12, IL-23 and IL-27, type 1 IFNs and ICAM1),[46–49] Th2-cell factors (i.e., OX40L and IL-4),[50–52] Th17-cell molecules (TGF-β, IL-6, IL-21 and IL-23)[53–55] and the factors that induce Tregs (i.e., IL-10 and TGF-β).[56,57] Different transcription factors signaling downstream from T-helper-skewing cytokines mediate the differentiation into the various T-helper subsets. Upon IFN-γ binding, T-helper cells upregulate the transcription factor T-bet via STAT1 signaling, which is crucial for Th1 differentiation. Together with IL-12, this response is amplified in a STAT4-dependent manner.[58] For polarization into Th2 effector cells, STAT6 has to be activated via IL-4.[59] This activation results in the induction of the Th2 master regulator gene, GATA-3.[60,61] Further differentiation also needs the activation of STAT5 by IL-2.[62,63] The induction of a Th17 response is mainly mediated by TGF-β in the presence of IL-6, which leads to the expression of RORγt and IL-17.[64] The production of IL-17 relies mainly on activation of the signal transduction pathway of STAT3. Removal of this transducer results in complete loss of IL-17 cells.[65] Furthermore, IL-21 is responsible for the amplification of Th17 differentiation, whereas IL-23 serves as a stabilizer for the development of Th17 cells. Treg development requires the presence of TGF-β for the induction of the transcription factor Foxp3.[66]

Notably, recent understandings indicate that polarization is more plastic than the abovementioned signaling processes imply.[67] Transcription factors can be suppressed and re-expressed upon environmental changes.

In conclusion, the fate of naive T cells and the outcome of immune responses are determined by factors responsible for the initial priming of DCs and subsequent transcription pathways that are activated by either proinflammatory or anti-inflammatory cytokines. In addition, in MDS, naive T-cell polarization by different DC subsets may play an important role. It can be hypothesized that DCs instruct T cells to a Th1 response in low-risk disease, creating an inflammatory environment and, to a tolerogenic direction, in high-risk MDS. It could also be assumed that in low-risk disease the high inflammatory environment created by the tumor cells generates immunogenic DCs and, in high-risk disease, the immunosuppressive setting induces DCs, which secrete anti-inflammatory cytokines, and thereby causes the conversion of naive T cells into Tregs. As mentioned earlier, CD141+ DCs, for example, are capable of creating a tolerogenic milieu by skewing naive T cells into Tregs. However, the exact mechanisms of T-cell polarization by DCs in MDS and the role in pathogenesis are still unknown.

DCs in Cross-presentation

Efficient removal of intracellular pathogens and the immune defense against tumor cells are strongly reliant on the activity of CD8+ cytotoxic T lymphocytes (CTLs). Activation of naive CD8+ T cells necessitates stimulation signals from APCs, such as DCs. This particular subset of T cells recognizes peptides presented on MHC class I molecules. Peptides presented by MHC class I molecules are derived from endogenous proteins. By contrast, MHC class II molecules present antigens derived from exogenous proteins. However, as most pathogens and tumor-derived proteins are not endogenous to DCs, the endogenous pathway of antigen presentation needs to be circumvented to present exogenous antigens in MHC class I molecules, a phenomenon known as cross-presentation.[68,69]

The cross-presentation capacities of DCs vary between the different subsets. In mice, the CD8α+ DC subset displays the ability to cross-present exogenous antigens, whereas CD8α DCs lack this capacity because of the absence of the specialized cross-presenting machinery.[70] Genome-wide expression profiling demonstrated that the human BDCA3+ or CD141+ myeloid DC subset possesses a number of phenotypic and functional similarities with the murine CD8α+ DC subset.[71,72] Several human studies showed that the BDCA3+ myeloid DC subset excels in cross-presentation when compared with BDCA1+ or CD1c+ myeloid DCs and BDCA2+ pDCs.[34,73] This subset also uniquely expresses XCR1.[74] Its ligand, XCL1, is produced by activated NK cells and CTLs. XCR1+ DCs are attracted to the site of inflammation and thereby optimizes the interaction and cross-presenting activities between CTLs and DCs.[75] Exclusively, the immature BDCA3+ DCs express the C-type lectin receptor CLEC9A (DNGR-1), which is involved in the recognition and uptake of antigens from necrotic cells through binding to exposed actin.[76–78] Antigens delivered by CLEC9A are presented on both MHC class I and II molecules. Therefore, this receptor is a relevant factor in cross-presentation and may serve as a target for antigen delivery in the context of cancer vaccination therapies.[79,80] Despite the superior cross-presenting capacity of BDCA3+ DCs, several studies have demonstrated that BDCA1+ myeloid DCs[34,81,82] and BDCA2+ pDCs[83] are also able to cross-present antigens. However, the exact role in tumor defense mechanisms and the efficacy as a target for immunotherapy still need to be clarified.

In MDS, cross-presentation can have an important function in immunity against dysplastic cells. Tumor-associated antigens (TAAs) can be presented by DCs to CD8+ T cells, thereby inducing tumor-specific immune reactions. The capacity of the BDCA3+ myeloid DC subset, which expresses CLEC9A, to cross-present necrotic material, might be interesting, for example, in the perspective of low-risk disease where a high apoptotic rate is found. Normally, CLEC9A does not detect early stages of apoptosis. However, excessive apoptosis or impaired clearance of apoptotic cells enables this receptor to become capable of taking up apoptotic cell fragments. In theory, this mechanism may fail in low-risk disease due to impaired DC function and, hence, allow the accumulation of apoptotic cells, which in turn can maintain the high inflammatory immune response. Interventions that increase uptake of apoptotic cells and stimulate cross-presentation by BDCA3+ DCs might then be beneficial.