Viewpoints on DC
Some active areas of DC research and their medical potential
This Viewpoint series provides authoritative and detailed outlines of exciting areas of DC research. Some of the subjects that frequently come up include development of DC; distribution of DC in lymphoid and non-lymphoid tissues such as skin, intestine and lung; different forms or subsets of DC; and the role of DC in initiating tolerance and immunity. In this Preface, I will introduce the Viewpoints and consider some future challenges as well as the medical relevance of DC research.
The development of DC, at least in mice, can be described with increasing precision because of discoveries summarized in the Viewpoint by Liu and Nussenzweig 1: (i) in the steady state, DC arise from a bone marrow progenitor that is shared with monocytes and macrophages 2; (ii) this progenitor gives rise to two cell types in the steady-state bone marrow: monocytes and a common DC progenitor 3–5; (iii) the latter gives rise to committed preDC that express some MHC II and CD11c, leave the marrow and circulate briefly in the blood before populating lymphoid and non-lymphoid organs 6, 7; (iv) Flt3 ligand (Flt3L) drives DC development 8, so that Flt3 knockout mice have a DC deficit 9, while administration of Flt3L expands DC numbers at least ten-fold in mice 10 and in humans 11.
The discovery of distinct steps in DC development should make it possible to identify the relevant transcription factors and, in turn, new markers to improve the definition and understanding of the DC lineage. Already there is progress, e.g. plasmacytoid DC (pDC) peculiarly require the E2-2 transcription factor for their development 12, 13.
A major gap in this aspect of DC science relates to Flt3-independent development from monocytes. Monocytes are bipotential. They can differentiate into macrophages with numerous scavenging and effector capacities. Alternatively, monocytes can develop into poorly phagocytic but highly immunostimulatory DC. This differentiation of monocytes to DC has been studied mainly in vitro for years, using monocytes from human blood 14, 15. What about in vivo? During inflammation in mice, several recent reports describe how monocytes acquire some properties of DC, i.e. expression of MHC II and CD11c 16–19. Now it is important to determine whether monocytes fully differentiate into authentic DC in vivo. By authentic, I mean the monocytes must acquire such DC properties as distinctive motility, localization to T-cell areas, loss of responsiveness to M-CSF, and efficient capture and presentation of antigens for display on both MHC I and II in vivo.
Most research on DC development involve mice; the study of DC in the human system is needed. The expansion of DC numbers with Flt3L could have medical benefit. For example, Flt3L administration suppresses autoimmune diabetes in NOD mice 20, probably by expanding both DC and Treg as part of a homeostatic circuit 21.
Different types of DC in the steady state, prior to the introduction of an infection or other stimulus, are called “subsets”. This field was initiated with mouse spleen 22, 23 and human blood 24, but now other organs are increasingly being scrutinized. Guilliams et al. 25 summarize studies in the skin that likely extend to other tissues. They provide a useful proposal in which there are at least five types of DC in the steady state: two types of classical DC, pDC, Langerhans cells, and monocyte-derived DC. Five subsets are in fact less complex than some previous descriptions.
Pabst and Bernhardt 26 discuss myeloid cells in the intestinal lamina propria. Pabst and Bernhardt concentrate on recent studies in which they examined for the first time some fundamental properties of CX3CR1high and CX3CR1low populations 27. CX3CR1high cells, or at least a sizeable fraction of them, derive from blood monocytes 28, 29 and are in a state where they do not present antigens effectively or migrate to the T-cell areas of mesenteric lymph node. In contrast, CX3CR1low/neg cells, which can express CD103, behave like bona fide DC, are able to present antigens effectively and also migrate to the T-cell areas.
Swiecki and Colonna 30 focus on pDC and consider the increasing examples in which pDC are involved in immunosuppression and tolerance. Swiecki and Colonna 30 also provide a valuable outline of the consequences of high type I interferon production upon nucleic acid signaling, a hallmark of these DC; these include resistance to viral infection and development of autoinflammatory diseases 31–33. Swiecki and Colonna's Viewpoint emphasizes a gap: the potential value of learning to target antigens selectively to receptors on pDC so that their functions in the intact animal can be directly analyzed.
An obstacle is the current emphasis on the integrins CD11c and CD11b to identify DC subsets. These integrins remain very helpful but are insufficiently cell specific in some circumstances. In searching for cell surface markers, I suspect that it will be valuable to stress groups of innate receptors, especially lectins that bind microbes, and transcriptional controls on subset development and function. To illustrate, the Langerin lectin 34 and the E2-2 transcription factor 12 are more incisive markers (than CD11b and CD11c) for subsets of CX3CR1low DC and pDC, respectively.
Several Viewpoints address the comparison of DC subsets in different species, including humans. There is an odd situation in which different markers are being used to identify functionally similar subsets in mice and humans. BDCA2 identifies human pDC but BDCA2 is not present in mice, in which Siglec-H is used instead. Yet mouse and human pDC are functionally similar and both can make high levels of type I interferons upon challenge with nucleic acids. Similarly, CD8α and BDCA3 are different molecules that identify mouse spleen and human blood DC subsets, respectively. Yet these subsets likely function in a similar manner in both species, including efficient cross presentation of antigens 35, 36 and unique expression of the long-sought lymphotactin or XCR1 receptor 37, 38. A more precise definition of DC subsets will also emerge from systems analyses of DC transcriptional programs, which also indicates that there are corresponding subsets of DC in several species, particularly mice and humans 39.
What do all these subsets signify? My view stems from the fact that many current markers for DC subsets are molecules involved in innate immunity such as antigen-uptake receptors (DEC-205, Langerin, DCIR2 on classical DC), pattern recognition receptors (TLR7 and TLR9 in pDC) and control mechanisms for innate immunity (BDCA2 or CD302 in pDC). This suggests that each DC subset is specialized to respond to distinct microbial and other challenges. In a related vein, the targeting of antigens to distinct lectins on DC subsets in vivo is providing a new way to interrogate receptor and DC function in animals, and is setting the stage for targeted delivery of antigens to improve vaccine efficacy in the clinic 40, 41.
DC location in vivo
The examination of living tissues by two photon microscopy has been essential in DC science, as in other areas of immunology. For example, this approach has established the unique probing morphology of DC in living lymph nodes 42, and the early steps in clonal selection in the T-cell areas 43–48.
Kastenmüller et al. 49 discuss several current challenges where vital techniques will be helpful. One is to understand the distinct location of DC subsets in vivo that can take place within lymph nodes and spleen 50, 51. Another is to determine what DC learn from their close interaction with the so-called fibroblastic reticular cells in the stroma of lymphoid tissues. Stromal cells are likely to be distinct in different regions of the lymph node where B cells, T cells and macrophages are enriched. A third challenge, also emphasized in Germain's laboratory, is how DC orchestrate the interaction of two rare cells, the antigen-specific helper CD4+ T cells and killer CD8+ T cells.
The medical impact of the last mentioned interaction of antigen-specific CD4+ and CD8+ T cells is notable. “Helped” CD8+ T cells mobilize better to infection challenge sites 52, and are a goal for more effective T-cell-based vaccines in the future 53.
An obstacle in vivo is to be able to do more imaging of DC in large animals and humans, e.g. appropriately labeled, DC-targeting antibodies might be visualized by positron emission tomography (PET scanning).
DC and tolerance
The tolerance field has been energized by exceptional progress with Foxp3+Treg as suppressors of immune responses. Rescigno's Viewpoint54 addresses the valuable DC part of the equation. DC exert significant controls on Treg and, reciprocally, will likely be necessary in understanding how Treg work. During homeostasis, DC regulate the numbers of Treg 21, and when DC present specific antigens, they can expand antigen-specific Treg 55–58. Control of Treg seems to be carried out best by particular DC subsets such as the CD103+ DC (also marked by DEC-205/CD205, Langerin/CD207, occasionally CD8) 59–61.
A challenge in going forward will be to learn how to control Treg in an antigen-specific manner. Until now, most research on Treg has involved approaches to totally remove them and then observe the rapid development of various forms of autoimmunity and chronic inflammatory bowel disease 62. These valuable approaches document the essential role of Treg in suppressing autoinflammatory diseases and have revealed critical mechanisms. A major gap remains: to determine whether one can expand antigen-specific Treg and selectively suppress unwanted immune responses.
Early papers on antigen-specific Treg have involved TCR transgenic T cells. DC either expand transgenic natural Treg in the presence of IL-2 or induce adaptive Treg along with TGF-β 63–65. When DC generate natural and induced Treg specific for a single pancreatic islet autoantigen, the Treg suppress autoimmune diabetes, which involves multiple autoantigens 63–65. A clinically relevant goal now is to find out whether antigen-capturing DC expand specific Treg from the polyclonal repertoire.
If we could learn to expand antigen-specific Treg, as Rescigno 54 emphasizes in her Viewpoint, we could achieve an entirely new approach to suppress allergy, autoimmunity and transplant rejection. Clearly, medicine already has benefited enormously from suppressive drugs such as Immuran, steroids and Cyclosporin, but these traditional drugs do not suppress the specific antigens driving disease. DC allow for the unique antigen-specific features of the immune system to be exploited, with the aim to provide more durable therapies with less side effects.
DC and the induction of Th2 responses to allergens and parasites
Plantinga, Hammad and Lambrecht 67 delve deeply into pulmonary DC to study DC biology at a pivotal mucosal surface. They emphasize that different DC subsets exert different functions, from the induction of Treg specific for environmental antigens to the formation of both protective IgA and allergenic IgE responses.
Previous studies in the lung concluded that DC tolerize the immune repertoire to harmless environmental antigens in the steady state and as a result, the DC do not induce unwanted immunity when they present both environmental and pathogenic antigens during infection 66. As Plantinga et al. 67 summarize, pDC, and not just classical DC, contribute to this vital tolerizing function.
Plantinga et al. 67 further describe how the lung is a key organ to approach the function of DC in Th2-driven allergy, both at the induction and effector phases. One shortcoming in the field is that the majority of experiments still rely on OVA as antigen. In contrast to OVA, authentic allergens can directly influence DC function 68, 69. Beyond the lung, antigens from helminths also alter DC to induce Th2 immunity 70. If these advances in DC science were extended to a vaccine perspective, e.g. to induce allergen-specific suppressive Treg or helminth-specific protective Th2 cells, the medical impact would be considerable.
DC and immune therapy
Schuler in his Viewpoint71 rightly draws attention to the new evidence that vaccination, as well as direct T-cell intervention with anti-CTLA-4 blockade, have real clinical benefit in phase III studies of patients with cancer. This gives a substantial impetus to research on DC-based immune therapy. I would like to comment on two points.
One relates to the choice of antigens for immune therapy, from the many that are being considered 72. The goal is to identify protective or regression-inducing antigens. But this in turn means that we need to learn how to use any given antigen in a way that leads to strong antigen-specific helper and cytotoxic T cells. Without research in this area in patients, i.e. improving immunogenicity, we are compromised in our capacity to compare antigens for their capacity to contain metastases, regress lesions and improve survival. Importantly, DC charged ex vivo with antigen should allow for effective antigen processing across a spectrum of MHC haplotypes 73, thereby facilitating an immunogenicity emphasis to cancer research. Improved vaccine immunity would also complement other strategies, e.g. in addressing immune checkpoints such as CTLA-4 and PD1, and to interfere with immune evasion mechanisms such as Treg and myeloid-derived suppressor cells in tumors.
A second point is that the induction of cancer immunity via DC is currently weak relative to what many suspect will be needed for cancer resistance. In thinking about this, we need to acknowledge the current weak immunity elicited by AIDS vaccines in healthy volunteers, in spite of a considerable investment to date 74. In other words, eliciting T-cell immunity in humans is far from straightforward. Yet the underdeveloped and undersupported field of DC therapy already has allowed for the induction of some immunity despite the fact that the research has been in patients who are sick and with scientific obstacles in place, such as the limited migration of therapeutic DC to lymphoid tissues 75. I urge that immunology be given the opportunity to play a much larger role to help reduce cancer morbidity and mortality. Scientists with talent in DC and other areas of immunology are ready to collaborate and provide a needed immune arm to cancer treatment. The cancer field should not be overlooking the unique mechanisms that the immune system can bring to the treatment of cancer.
Thanks to the authors and to Judy Peng and Reinhold Förster for putting together this series of Viewpoints on active areas of DC biology. In spite of the diversity of subjects covered here, many key areas (and laboratories) could not be represented, such as antigen processing and presentation, and the function of DC in relevant organs such as the brain, aorta, kidney and genital tract. Nevertheless, progress of the kind illustrated in these Viewpoints will continue to illuminate DC as an integrated system for immune control. DC provide a framework to alleviate disease in unique immunological ways, particularly the specific vaccines and therapies that have begun to emerge.
The author receives funding support from NIAID and the Bill and Melinda Gates Foundation.
Conflict of interest: The author is a paid scientific consultant to Celldex Therapeutics, which is developing DC-targeted vaccines.