Chapter 21: Dendritic Cells and Adaptive Immunity

Host defense is mediated by innate and adaptive immune responses, and dendritic cells (DCs) play essential roles in linking together innate and adaptive immunity.^1,2,3 The innate immune response provides rapid resistance to pathogens, but the potency of innate responses does not increase following initial exposure. Adaptive responses, mediated by B and T lymphocytes, generate immune memory, resulting in more rapid and more potent responses following antigen reexposure (Chaps. 75 and 76).

DENDRITIC CELLS AND INNATE IMMUNITY

DCs provide innate immune resistance through production of cytokines, including interleukin (IL)-12 and type I interferons, and by activating other innate lymphocytes such as natural killer (NK) cells, NKT cells, and γδ T cells (Chap. 75). Innate responses are most often initiated by “pattern recognition receptors” (Chap. 20), which respond to evolutionarily conserved molecules found in microbes, parasites and viruses.^4,5 Pattern recognition receptors include toll-like receptors (TLRs), nucleotide-binding oligomerization domain-like receptors, retinoic acid-inducible gene 1-like receptors, and numerous C-type lectins. Pattern recognition receptors recognize a wide array of ligands, such as single- or double-stranded RNA, lipopolysaccharides, and other microbial constituents. DCs express pattern recognition receptors and respond to pattern recognition receptor agonists by becoming potent immunostimulatory cells and by presenting captured antigens to T cells in the context of major histocompatibility antigens. Pattern recognition receptors on DCs can also be activated via noninfectious stimuli induced by tissue damage and malignant cells, including uric acid crystals and heat shock and chromatin proteins. Such pathways are likely important for activating DCs toward tumor-associated antigens following transplantation or in disease states such as cancer or allergy.

Activated or matured DCs can prime resting NK cells, which then reciprocally act back on DCs, enhancing DC maturation and initiating adaptive immune responses.⁶ NK cells can also negatively regulate DC function by killing immature DCs, a feature that is important in NK mediated graft-versus-leukemia (GVL), without graft-versus-host disease (GVHD), following hematopoietic stem cell transplantation.⁷ Such immune crosstalk between NK cells and DCs, is emblematic of the exquisite interdependence and complexity of the immune response.^8,9,10

DENDRITIC CELLS AND ADAPTIVE IMMUNITY

Adaptive immunity imparts immune memory to the host, which allows a more rapid and more effective immune response to rechallenge with the same antigen. Adaptive immunity is mediated by B and T lymphocytes, but its induction is largely regulated by DCs. Immune memory increases the frequency and function of antigen-specific lymphocytes, leading to enhanced levels of protective antibodies, cytokines, and killer molecules. Through an elaborate series of gene rearrangements, hypermutation, and selection, antibody on B cells and T-cell receptors on T cells provide remarkable diversity and specificity. This array might be thought of as the largest combinatorial library of specificities in the world!

DCs function as sentinels of the immune system (Table 21–1)^1,3,11 and provide a critical bridge between innate and adaptive immunity. Indeed, the presence of antigen and lymphocyte is rarely sufficient to induce adaptive immune responses. Rather, a third party, the DC system of antigen-presenting cells provides the pivotal immune initiation signal required to induce an adaptive immune response. DCs sense a wide range of environmental stimuli, producing cytokines, such as IL-12 and type I interferons that help stimulate immune responses. DCs express most types of TLRs, with specific subsets of DCs differentially expressing distinct TLRs; for example, plasmacytoid DCs express significant levels of TLR-7 and TLR-9.¹² DCs can also respond to endogenous stimuli, ranging from inflammatory cytokines, including tumor necrosis factor (TNF)-α, IL-1, or interferons, to byproducts of cell death or tissue damage. DCs capture microbes and tumor cells, processing their component antigens for presentation to the adaptive immune system. DC activation, via innate immune receptors, leads to DC maturation and initiation of adaptive immunity. In addition to antigen processing and presentation, sentinel DCs produce chemokines and cytokines. They migrate to lymphoid tissues, where they recruit naïve antigen-specific lymphocytes and instruct their subsequent development.

A variety of DC subsets exist and the biology of DC activation can vary significantly according to the exposure. During immunization, DCs initiate clonal expansion of T cells and can directly and indirectly influence the growth of B cells. In addition, DCs can modulate differentiation of lymphocytes, such that the properties of the lymphocytes are appropriate to the invading pathogen.¹³ For example, under the influence of DCs, T cells can preferentially produce interferon-γ (T-helper [Th] type 1 cells) to activate macrophages to resist infection by intracellular microbes; or IL-4, IL-5, and IL-13 (Th2 cells) to mobilize white cells to resist helminths; or IL-17 (Th17 cells) to mobilize phagocytes at body surfaces to resist extracellular bacteria.

An essential counterpart to adaptive immunity is adaptive tolerance, which is the silencing of cells with receptors reactive to self or harmless environmental antigens. T cells develop tolerance centrally in the thymus and peripherally in lymphoid organs.^14,15 DCs play a role in inducing tolerance, especially with regard to T cells. DCs can silence self-reactive T cells, through either deletion or functional inactivation. DCs can also drive T cells to suppress immunity by expressing IL-10 (T regulatory [Tr]1 cells) or FOXP3.¹⁶

DCs are “antigen-presenting cells.” An antigen-presenting cell is any cell that uses its major histocompatibility complex (MHC) products (or other antigen-presenting molecules, such as the CD1 molecules that present glycolipids and lipoglycans) to bind and display (i.e., “present”) fragments of antigen to lymphocytes. DCs are more specialized or professional than other antigen-presenting cells. This is because DCs have efficient and regulated pathways for antigen uptake and processing, and DCs possess dozens of features that allow them to initiate and control immunity (Table 21–2). For example, when DCs mature in response to infection, hundreds, even thousands, of gene transcripts can be upregulated or downregulated.^17,18

ANTIGEN UPTAKE AND PROCESSING

DCs express a wide array of endocytic receptors, which enhance the efficiency of antigen capture, processing, and presentation. Many endocytic receptors are predicted to be C-type lectins, and in some cases their natural ligands have not been identified. DCs also express Fcγ and Fcε receptors, which recognize immune complexes, as well as scavenger receptors. Recognition of pathogens by DC receptors can have two outcomes. One outcome is immune activation followed by antigen presentation and development of a productive immune responses. Alternatively, pathogens may use DC receptors to evade the host immune response. For example, DC-SIGN (CD209) a lectin expressed on DCs, is used by HIV-1 and cytomegalovirus (CMV) to reach T cells and endothelial cells, respectively^19,20; by Dengue virus to replicate within DCs²¹; and by Mycobacterium tuberculosis to trigger production of the suppressive cytokine IL-10.²²

Following uptake, efficient processing of antigen yields peptides that bind to MHC class II and class I products. “Exogenous” antigens refer to molecules processed directly following uptake, whereas “endogenous” antigens are processed following biosynthesis in the antigen-presenting cell. Classic pathways of antigen presentation emphasize processing of “exogenous” antigens for presentation on MHC II–peptide complexes to CD4+ T lymphocytes, whereas “endogenous” antigens were targeted for presentation on MHC class I–peptide complexes to CD8+ T cells. However, it is now clear that there is significant overlap in these pathways such that exogenous antigens may also be presented on MHC I products to CD8+ T cells. This pathway is termed cross-presentation and is important for initiation of antitumor immune responses. Cross-presentation is well developed in DCs, especially those found in lymphoid tissues, and leads to either tolerance or activation of CD8+ T lymphocytes, depending upon the DC maturation stimulus.²³ For instance, products of dying cells, from transplants, tumors, foci of infection, and self-tissues are endocytosed by DCs then presented on MHC Class I to CD8+ cells.^24,25 Importantly, cross-presentation of antigens onto MHC class I can involve the proteasome and transporters for antigenic peptides, which are used in the presentation of endogenous antigens.

DCs are a major cell type involved in cross-presentation of proteins,^26,27 and probably of lipids.^28,29 Cross-presentation has been documented involving nonreplicating microbes, dying cells, ligands for the DEC205 receptor, and immune complexes, including antibody-coated tumor cells. This pathway allows DCs to induce tolerance or immunity to antigens not synthesized de novo in these cells. Fcγ receptors, in addition to mediating presentation, can influence DC maturation, either enhancing maturation through activating forms of the receptor or preventing maturation through inhibitory forms.³⁰ Such consequences of antibody binding to DC Fc receptors, with regard to DC maturation and cross-presentation, are important features for consideration when trying to understand the use of antibodies as therapeutic agents.

Some DCs subsets also express the CD1 family of antigen-presenting molecules. For example, CD1a typically is found on epidermal Langerhans cells in skin, whereas CD1b and CD1c are expressed on dermal DCs. CD1 molecules present glycolipids, whereas microbial glycolipids are the best studied to date with regard to CD1a, CD1b, and CD1c. CD1d molecules on DCs also efficiently present the synthetic glycolipid α-galactosylceramide.³¹ This process leads to activation of distinct lymphocytes with a restricted T-cell repertoire, the NKT cells.³² NKT cells have significant potential as effector cells because they can produce large amounts of interferon-γ and lyse tumor targets.

A newer “nonclassical” pathway for antigen presentation involves presentation of “endogenous” proteins on MHC class II.³³ This pathway involves autophagy and is also well developed in DCs.³⁴ It allows nuclear, mitochondrial and cytoplasmic proteins to be presented from digestive compartments, including as a first example, the Epstein-Barr nuclear antigen 1.³⁵

MATURATION OF DENDRITIC CELLS

Immature DCs efficiently take up antigen but do not induce immunity, defined as the production of immune effectors and the establishment of memory. For immune induction to occur, DCs require additional stimuli that lead to an intricate differentiation process called “maturation.” Maturation involves changes in endocytic and antigen processing machineries, production of chemokines and cytokines, and expression of many cell-surface molecules, including those of the B7, TNF, and Notch ligand families. DCs have an endocytic system that is tightly regulated and devoted to presentation of captured antigens, rather than clearance and scavenging.

In the case of DCs derived from marrow and monocyte precursors, DC maturation is accompanied by exquisite changes in the endocytic system with attendant consequences for antigen processing and presentation. During maturation, antigen uptake is dampened as a result of inactivation of a rho-guanosine triphosphatase termed cdc42.³⁶ At the same time, machinery associated with antigen processing is augmented. Lysosomal processing is activated by assembly of an active proton pump, which acidifies the lysosome so that processing of antigens and the MHC class II associated invariant chain can proceed. MHC–peptide complexes form within the endocytic system of the maturing DCs,^37,38 then traffic in distinct nonlysosomal compartments to the cell surface. Internalization and degradation of MHC II also occurs via ubiquitination is also inhibited in mature DCs.³⁹ DC maturation also increases presentation on MHC I via, in part, the formation of an “immunoproteasome,” a combinatorial form of proteasome that increases the spectrum of peptides destined to be presented on MHC I.⁴⁰

A hallmark of DC maturation in response to several stimuli is upregulation of costimulatory molecules such as CD80 and CD86, resulting at least in part from production of inflammatory cytokines, particularly TNF-α.³² Importantly however, CD86 upregulation should not be equated directly with immune activation, which requires other DC functions, such as those triggered by CD40 ligation, including production of cytokines such as IL-12 or type I interferons, and/or engagement of other receptors such as CD70.³²

DCs enhance antibody formation by several mechanisms. The classical pathway involves induction of antigen-specific CD4+ helper T cells, which then help B-cell growth and antibody secretion. However, DCs can also directly affect B cells to enhance immunoglobulin (Ig) secretion and isotype switching, including production of the IgA class of antibodies, which contribute to mucosal immunity.^41,42 DCs can induce a B-cell class switch in a CD40-independent manner, through production of ligands such as B-lymphocyte stimulator (B-cell activating factor belonging to the TNF family [BAFF]) and a proliferation-induced ligand (APRIL), including T-cell–independent induction of IgA antibodies to commensal organisms.⁴³ Plasmacytoid DCs stimulate antibody responses to influenza virus in culture.⁴⁴ Production of antibodies by any of these mechanisms may lead to interaction with DC FcγR and thereby an adaptive response by T cells.

The primary function of DCs is to survey host tissues and initiate responses upon encountering danger signals. To optimally perform these functions, DCs are present in every tissue of the body, and are enriched in lymphoid tissues. Immature DCs are strategically positioned along body surfaces (skin, airway, gut) and in the interstitial spaces of many organs, such as the heart and kidneys. DCs are able to extend their processes through the tight junctions in epithelia, without altering the epithelial barrier, allowing them to sample antigens from harmless environmental antigens and commensal microorganisms. In the steady state, DCs migrate continuously from tissues into afferent lymphatics and probably blood. By electron microscopy, DCs in lymphoid tissues often are termed interdigitating cells, appearing as large stellate cells with a lucent “empty”-appearing cytoplasm.

One method of categorizing DCs is to describe them based upon the tissue in which they reside. Lymphoid tissue–resident DCs have undergone intensive study, especially in murine models. The scarcity of DCs in nonlymphoid tissues, on the other hand, limited understanding of their importance for many years and continues to render study of these cells challenging. Because circulating DCs are the most accessible subset in humans, they have received intense study; however, blood DCs may not reflect accurately the biology of tissue resident cells. Briefly, all DCs subsets are CD45+CD11c+class II+ cells that lack markers associated with T-cell, B-cell, erythroid, granulocytic, and NK lineages. However, such a definition is not complete as some macrophages express this phenotype yet are distinct from the DC lineage. Additional confidence in a DC lineage is provided by expression of Flt-3 (FMS-like tyrosine kinase 3, CD135), which mediates signals important in the differentiation of this lineage, c-kit and/or CCR7.

An alternative method for classifying DC subsets separates the major lineages as classical dendritic cells (cDCs) versus plasmacytoid dendritic cells (pDCs). cDCs are the most plentiful and incorporate lymphoid tissue–resident and nonlymphoid tissue–resident DCs. Although both pDCs and cDCs derive from a common progenitor, pDCs are distinct from cDCs in appearance (they resemble plasma cells) and they reside primarily in blood and lymphoid tissue, but are not found in nonlymphoid tissues. pDCs are important mediators of innate immunity, because of their capacity to rapidly produce high levels of interferon (IFN)−α upon encounter with pathogens that engage the TLR-7 and TLR-9 receptors, which are plentiful in this subset. cDCs are frequently further subdivided into lymphoid-resident cDCs versus nonlymphoid-resident cDCs. In mice, most lymphoid-resident cDCs are CD8α+CD11b− whereas most nonlymphoid-resident cDCs are CD103+CD11b−. CD8α+CD11b− lymphoid tissue–resident cDC have a similar origin, phenotype, and transcriptional profile as CD103+CD11b− nonlymphoid-resident cDC. In contrast, CD11b+ cDCs can be found in both lymphoid and nonlymphoid tissue, and this subset is notable for its ability to derive from monocytes in response to granulocyte-monocyte colony-stimulating factor (GM-CSF), monocyte colony-stimulating factor (M-CSF), and other inflammatory mediators. CD8α+ and CD8α− cDCs show a remarkable division of labor in terms of the nature of induced host response. While CD8α+ DCs efficiently cross-prime CD8+ T-cell immunity through MHC class I antigen presentation,⁴⁵ CD8α− DCs stimulate predominantly CD4+ T-cell response through MHC class II presentation.⁴⁶ These differences may be explained in part by the fact that CD8α+ DCs have high endosomal pH, low antigen degradation, high antigen export to cytosol, and more presynthesized stores of MHC class I molecules.^47,48

In humans, essentially all DCs lack lineage markers for T cells, B cells, NK cells, and erythroid and granulocytic lineages, and express CD45 and MHC class II. Human pDCs are described as Lin−classII+CD303(BDCA2)+CD304(BDCA4)+, whereas human cDC blood subsets generally correlate with lymphoid tissue cDC subsets. Human cDCs are either CD1c(BDCA1)+ or CD141(BDCA3)+, with the CD1c+ subset being significantly more plentiful than the CD141+ subset. In humans, BDCA3+ DCs have been proposed as human equivalents of murine CD8α+ DCs, and may have superior cross-presentation capability compared to other human DC subsets.⁴⁹ However, the capacity for cross-presentation is not restricted to this subset of human DCs. The DEC205/CD205 lectin receptor is also a useful marker for DCs in human lymph nodes.⁵⁰ Nonlymphoid tissue–resident DCs in humans remain incompletely characterized, with the exception of Langerhans cells. First described more than 100 years ago, Langerhans cells are cDCs found in the epidermis, characterized by a “tennis racket” appearance by light microscopy as a result of internalized Langerin, which localizes in “Birbeck granules.”⁵¹ Langerhans cells are characterized by expression of CD1a as well as EpCAM and CD207(Langerin). Uncontrolled proliferation of Langerhans cells is responsible for the clinical disorder known as Langerhans cell histiocytosis (LCH).

Further insights into the functional diversity of DC subsets has come from discovery of distinct transcriptional factors that underlie their development and functional properties.⁵² For example, z-DC, was identified as a transcriptional factor regulating the development of classical DCs.^53,54 Functional specialization of DC subsets is likely to impact the next generation of vaccines, as discussed in “Dendritic Cells in Immunotherapy” below.

One of the most important discoveries in DC biology has been identification of a clonogenic DC progenitor within the marrow that is committed to the DC lineage and unable to give rise to other cell types. Such progenitors, described as pro-DCs can generate plasmacytoid and CD8α+ and CD8α− classical DCs, but no other cell subset.^55,56 Subsequent work confirmed that the committed DC progenitor derives from a more primitive cell that is capable of giving rise to monocytes and macrophages.⁵⁷ Progenitors can also be identified that specifically gave rise to cDCs but not pDCs.⁵⁶ In mice, administration of Flt-3 ligand expands progeny of DC-committed progenitors. This series of discoveries dispelled the notion that DCs in vivo were largely derived from inflammation driven monocyte differentiation and firmly established DCs as a distinct hematopoietic lineage. Still, in states of inflammation, DCs can be induced from mature monocytes in the presence of inflammatory mediators such as GM-CSF and M-CSF.

In view of their central role as antigen-presenting cells, DCs have been targeted to both boost T-cell immunity in the setting of resistance against pathogens/tumors or suppress T cells in the setting of autoimmune disease.⁵⁸ Although both are relevant to hematologic diseases, much of the current effort has been in the context of vaccination to boost T-cell immunity against tumors. Immunity to standard subcutaneously injected protein vaccines is also likely impacted by the biology of lymph node–resident and migratory DC subsets.⁵⁹ Two broad strategies have been attempted to harness the ability of DCs to boost T-cell immunity. One approach involves adoptive transfer of antigen-loaded DCs.⁶⁰ In most of these studies, DCs were generated ex vivo from precursors such as blood monocytes, although some of these studies also utilized more primitive progenitors, such as CD34+ hematopoietic stem cells or circulating DCs.⁶¹ One of the first studies involved the injection of idiotype-pulsed DCs against lymphoma.⁶² Most of these studies were small, and while DCs were well tolerated, they led to only modest clinical effects in terms of objective tumor regression. Although these studies provided important fundamental insights into DC biology, such as the link between DC maturation and their immunogenicity,^63,64 they did not exploit the biology of naturally occurring DCs and their subsets. These studies also emphasized the need to combine vaccine-based approaches with strategies to overcome suppressive elements in the tumor bed, including inhibitory immune checkpoints, and address the vaccine-induced induction of regulatory T cells.⁶⁵ Recent promising clinical results with T-cell immune checkpoint blockade such as with antibodies against PD1/PDL1 in human cancer is setting the stage for the next generation of combination therapies with vaccines.^66,67 In addition to T cells, DCs are also being explored to activate other immune cells such as innate NKT cells. Combination of NKT-cell–targeted vaccine with low-dose lenalidomide led to synergistic immune activation and tumor regressions in early myeloma.⁶⁸

Another strategy involves targeting antigens directly to DCs in situ. Coupling antigens to antibodies against DEC205, an antigen-uptake receptor on DCs, leads to enhanced activation of T-cell immunity in several models.⁵⁸ Even in this setting, DC activation is essential to elicit immunity and targeting antigens to DEC205 in steady state results in the induction of tolerance.⁶⁹ Early clinical studies with targeting antigens to DEC205 combined with approaches to activate DCs demonstrate clear induction of humoral and cellular immune responses.⁷⁰ Promising but preliminary results in patients who receive T-cell checkpoint blockade following the vaccine supports the investigation of combination approaches.⁷⁰ Other emerging approaches to targeting DCs in situ include nanoparticles. These technologies hold promise for flexibility and personalization of the vaccine, but the nature of optimal DC subset or adjuvant/DC maturation stimulus remains to be clarified.⁷¹ It is notable that potent human vaccines, such as the yellow fever vaccine, simultaneously engage multiple DC subsets,⁷² raising the prospect that to optimize the generation of T-cell immunity a combination of DC subsets will need to targeted in vivo.⁷³

A setting where the biology of DCs plays a central role in hematology is allogeneic stem cell transplantation. The immunologic activity of donor T cells in allogeneic stem cell transplantation is a critical factor for eradicating residual malignancy, a process termed GVL, but can also lead to detrimental GVHD. There is considerable evidence that the induction of GVHD is dependent on the remaining host antigen-presenting cells of which DCs are the most potent.^74,75 The role of DCs in mediating the GVL effect is less-well understood, although a role for DCs has been implicated.⁷⁶ DC subsets, particularly pDCs, have also emerged as critical regulators of autoimmune hematologic disorders such as immune thrombocytopenia in children.⁷⁷

Consequently, targeting DCs may be of benefit in the setting of autoimmunity as well as GVHD.