Accumulation of pDC: roles in disease pathogenesis and targets for immunotherapy

Summary

Plasmacytoid dendritic cells (pDC) secrete type I interferons (IFN-I) in response to viruses and RNA/DNA/immunocomplexes. IFN-I confer resistance to viral infections and promote innate and adaptive immune responses. PDC also produce cytokines and chemokines that influence recruitment and function of T cells and differentiation of B cells. Thus, pDC have been implicated both in protective immune responses and induction of tolerance. In this article we discuss how the recruitment and accumulation of pDC may impact pathogenesis of several diseases and how pDC can be targeted for therapeutic interventions.

PDC are type I interferon-producing cells

Plasmacytoid dendritic cells (pDC) are bone marrow-derived leukocytes that secrete type I interferons (IFN-I) [1, 2]. PDC detect RNA and DNA from viruses and RNA/DNA/immunocomplexes through two endosomal sensors, Toll-like receptors (TLR) 7 and TLR9, respectively, which induce secretion of IFN-I through the MyD88-IRF7 signaling pathway [3-5]. PDC were first identified in humans as CD4⁺, CD68⁺ and IL-3R⁺ (CD123) plasma-like cells [6]. Initially, it was unclear what functions these cells perform in vivo, however, their prominent endoplasmic reticulum alluded to a role in cytokine secretion. Later, it was demonstrated that this unique subset could differentiate into antigen-presenting cells [7, 8] and specialize in the secretion of IFN-I, thus corresponding to the human natural IFN-producing cells (NIPC) [9, 10]. In 2001, cells that resembled human pDC were finally identified in the mouse [11].

PDC accumulation during infection and disease

PDC originate in the bone marrow from common lymphoid/myeloid progenitors and depend on fms-like kinase 3 ligand (Flt3L), STAT3 (signal transducer and activator of transcription 3) and the transcription factor E2-2 for development [12]. Unlike classical DC, which migrate to lymph nodes via the lymphatics, pDC enter the T cell areas of lymph nodes directly from the blood through high endothelial venules [13, 14]. In homeostatic conditions pDC also inhabit mucosal tissues and organs albeit at low numbers. PDC accumulation in lymphoid tissues, mucosa and organs occurs during several human pathologies, particularly in lymph nodes of patients affected by sarcoidosis, Mycobacterium tuberculosis infection [15], Kikuchi's disease [16] and in the skin of patients affected by psoriasis [17, 18], systemic lupus erythematosus (SLE) [19] and lichen planus [20, 21]. PDC accumulation has also been observed in brain lesions of patients with multiple sclerosis [22], in the salivary glands of patients with Sjogren's syndrome [23] and the synovia or inflamed muscle tissue/skin of people afflicted with rheumatoid arthritis [24, 25] or dermatomyositis [26, 27], respectively. PDC are overrepresented in the blood of patients with type I diabetes around the time of diagnosis [28]. PDC also infiltrate tumors [29-36] and are recruited to infected sites during viral infections such as Herpes Zoster Virus (HZV) [37], Hepatitis C virus (HCV) [38] and Herpes simplex virus [39].

The accumulation of pDC has also been observed in animal models of disease. During influenza [40-42] and respiratory syncytial virus (RSV) [43, 44] infections pDC are recruited to the lungs and draining lymph nodes of mice. PDC numbers increase in the pancreatic lymph nodes around the onset of diabetes in nonobese diabetic (NOD) mice [45] and in the pancreas during lymphocytic choriomeningitis virus (LCMV) infection [46]. In mouse models of HSV infection pDC accumulate in the lymph nodes following footpad infection with HSV-1 [47] and in the vaginal mucosa during HSV-2 infection [48]. PDC are also recruited to the vaginal mucosa of rhesus macaques intravaginally infected with simian immunodeficiency virus (SIV) [49]. Furthermore, it has been reported that pDC infiltrate lymph nodes during SIV infection [50, 51]. PDC infiltration of tumors [52-54], tumor-draining lymph nodes (TDLN) [55, 56] and the CNS during EAE [57] has also been documented. What is the impact of pDC accumulation in the pathogenesis and progression of diseases? As we explain below, pDC may have either negative or positive effects (Figure 1).

Figure 1. PDC accumulation and disease.

PDC impact both innate and adaptive responses. Top and bottom panels illustrate the negative and positive effects of the major pDC-induced immune functions in the pathogenesis of autoimmune diseases, viral infections and cancer.

Drawbacks to accumulation of activated pDC

The accumulation of pDC contributes to pathogenesis in several viral models and disease settings. PDC infiltration and excessive IFN-I production are hallmarks of psoriasis and SLE [2, 58-63]. During psoriasis, pDC accumulate in the skin and produce IFN-I in response to self-DNA complexed with the antimicrobial peptide LL-37 [64]. Blocking IFN-I strongly inhibits the T cell-dependent progression of psoriasis thus implicating pDC as critical mediators of disease [18]. Because peripheral blood mononuclear cells from SLE patients have an IFN-α/β signature in the transcriptome which correlates with disease severity [65-68] and pDC infiltrate the skin and secrete IFN-I in response to self-DNA/RNA/immunocomplexes, pDC are often deemed as the culprits in promoting SLE as well. Additionally, pDC-derived IFN-I has been implicated in the initiation of type I diabetes in NOD mice [45]. PDC accumulate in the pancreatic lymph nodes and produce IFN-I in response to apoptotic β-cell debris thereby activating DC and autoreactive T cells. Thus, it would appear that pDC, upon activation and IFN-I secretion, aggravate, even perhaps instigate the diseases mentioned above, although it is still unclear if pDC are really the perpetrators.

Prolonged pDC activation and secretion of IFN-I have been also associated with the progressive loss of CD4⁺ T cells and the chronic activation of CD8⁺ T cells in HIV infection [69, 70]. Additionally, pDC may participate in HIV pathogenesis by recruiting T cells to sites of HIV replication where they can become infected. PDC preferentially secrete the chemokines CXCL9 (MIG), CXCL10 (IP-10), CCL3 (MIP-1α), CCL4 (MIP-1β) and CCL5 (RANTES) [71], which can attract naïve and activated CD4⁺ and CD8⁺ T cells to sites of infection [72, 73]. It has been shown that pDC accumulate in the vagina of rhesus macaques that are intravaginally infected with simian immunodeficiency virus (SIV) [49]. This accumulation resulted in increased levels of MIP-1β, which attracted activated T cells that are susceptible to SIV infection, facilitating the generation of a local infection focus that can subsequently spread to the draining lymph nodes. PDC may also facilitate the recruitment of T cells to the liver during HCV infection. Liver biopsies from patients with HCV revealed infiltrates containing both pDC and T cells [38]. While cytotoxic T cells (CTL) are critical for eradicating many viral infections, in the case of hepatitis virus, robust CTL responses induce severe liver damage.

A pathogenic role for tolerogenic pDC

PDC have been shown to promote tolerance, particularly during cancer. While activated pDC appear to behave as immunogenic cells, unstimulated or alternatively stimulated pDC can alleviate protective immunogenic responses to tumor cells through the induction of T regulatory cells (Treg). PDC induce Treg by several mechanisms. First, they express the immunoregulatory enzyme indoleamine-2,3-dioxygenase (IDO) [74, 75] which promotes tryptophan catabolism, depleting the tryptophan pool that T cells need to generate effective responses. IDO-expressing cells in TDLN of breast cancer patients correlates with worse clinical outcome [55]. Likewise, studies performed in a mouse model of malignant melanoma have demonstrated that cells resembling pDC expressed IDO in TDLN [55] and activated Treg [56]. Second, activated human pDC express inducible costimulator ligand (ICOS-L) which promotes the generation of IL-10-producing Treg from naïve T cells [76]. In addition to infiltrating TDLN, pDC can be directly recruited to tumors by factors such as stromal derived factor-1 (SDF-1) [29, 34] and induce IL-10-producing Treg. Moreover, human pDC can directly suppress T cell responses through expression of granzyme B [77].

The ability of pDC to induce Treg can also impact responses to HIV infection. Human pDC exposed to HIV in vitro express IDO and promote the differentiation of naïve CD4⁺ T cells into Treg that suppress proliferation of effector T cells [78] and impair DC maturation [79]. This reason, in addition to those listed above, suggest that pDC accumulation during HIV infection may be detrimental.

Positive aspects of pDC accumulation

Although damaging in some cases, pDC-mediated recruitment of CTL might be essential in the control of several infections, such as murine hepatitis virus (MHV), RSV, and HSV-2, where pDC depletion dramatically impairs host anti-viral responses [43, 44, 48, 80, 81]. PDC induction of Treg is also beneficial in many situations. Despite inducing tolerance to tumor cells, pDC mediate tolerance to harmless antigens and alloantigens through the induction of Treg [82-84]. In homeostatic conditions, self-reactive T cells are kept in check by Treg. Genetic defect of the Treg-specific transcription factor Foxp3 results in Treg-deficiency and development of fatal autoimmune pathology [85]. PDC also reside in the thymus [86, 87] and may directly participate in the generation of Treg there [88, 89].

Despite the negative impact pDC may have during HIV infection, evidence suggests they may serve a protective role, at least early on. Initially, it was observed that pDC numbers were dramatically reduced in the blood of patients chronically infected with HIV. Loss of pDC correlates with high viral loads, decreased numbers of CD4⁺ T cells and the onset of opportunistic infections [90-98]. PDC stimulated in vitro with HIV secrete IFN-I and other immune mediators [99, 100] and can cross-present HIV-derived antigens to CD8⁺ T cells [101]. HIV-activated pDC may also contribute to host responses by inducing DC maturation through the secretion of IFN-I and TNF-α [99]. Furthermore, PDC-derived IFN-I induces an anti-viral state and limits replication of HIV in CD4⁺ T cells [102, 103]. Therefore, pDC may be capable of eliciting protective responses to HIV in vivo.

PDC can also facilitate protective anti-tumor responses. The critical decision for pDC to either promote tolerance or immunogenic responses to tumors ultimately depends on their activation or maturation state, much like classical DC [104]. While immature or alternatively activated pDC induce Treg, pDC activated with TLR ligands can initiate tumor regression in an NK cell-dependent manner [105]. The anatomical location of pDC also appears to be critical factor in whether pDC act as tolerogenic or immunogenic cells. This was clearly demonstrated in a model of oral tolerance [83]. PDC from mesenteric lymph node or liver but not spleen effectively mediated suppression of T cell responses to oral antigens. In contrast to spleen pDC, pDC isolated from Peyer's patches (PP) fail to produce IFN-I after TLR stimulation [106]. Treating spleen pDC with factors associated with mucosal tissues such as IL-10, TGF-β or prostaglandin E (PGE) prior to TLR stimulation recapitulated the phenotype of PP pDC. It should be noted that tumors produce several of these factors to evade detection by the immune system [107]. Therefore, pDC accumulation in tumor environments rich in these anti-inflammatory mediators may condition and render them ineffective at generating immunogenic responses.

PDC can also participate in the direct killing of tumor cells or virus-infected cells. CD2 is a cell adhesion molecule that distinguishes two human pDC subsets [108]. One of these subsets (CD2^hi) expresses lysozyme and displays cytolytic capacity against tumor cells while pDC kill virus-infected cells through FasL and TRAIL-dependent mechanisms [109-112]. While killing tumor cells and virus-infected cells are beneficial in most situations, it was recently shown that pDC mediate killing of CTL in the lymph nodes during lethal influenza infection [42].

Targeting pDC for therapeutic intervention

Although we have extensive knowledge of how pDC may influence immunity or tolerance, the present challenge in the field is to better understand what pDC actually do during immune responses in vivo and particularly, the selective pressures under which pDC have been maintained throughout evolution. PDC-less mice have been described [113, 114] and will be instrumental in addressing these issues. Another important challenge in the field is to target pDC for therapeutic purposes. Antibody-mediated depletion of tolerogenic and activated pDC may be advantageous in tumors and autoimmune diseases, respectively. Blood dendritic cell antigen-2 (BDCA-2) is a molecule expressed exclusively by human pDC [115, 116], which provides an attractive target for the development of human pDC-depleting antibodies. On the other hand, infusion of tolerogenic or activated pDC may be useful therapies for transplantation and cancer, respectively.

PDC express MHC class II and costimulatory molecules which enables them to present Ag to CD4⁺ T cells and cross-present Ag to CD8⁺ T cells [117]. Thus, the function of pDC as antigen-presenting cells could be exploited to induce immunity or tolerance. To achieve this, Ag must be conjugated with anti-pDC antibodies that selectively target them to pDC. This technology has been successfully used for classical DC, inducing effective immune responses [118-120]. Targeting Ag to human pDC has also been described to some extent using anti-BDCA-2 [116] and -Dendritic cell immunoreceptor (DCIR) antibodies [121]. In mice, Ag could be targeted to Siglec-H [122, 123] and PDC-TREM [124]. However, unlike Siglec-H, which is constitutively expressed by pDC, PDC-TREM is only expressed by TLR-activated pDC. Targeting Ag to pDC via these two molecules could provide valuable insight into the Ag-presenting capacity of unstimulated versus activated pDC and the tolerogenic or immunogenic responses that might ensue.