An immunotherapy approach targets nanoparticles to dendritic cells of the immune system, leading to an antitumour immune response with antiviral-like features. Initial clinical tests of this approach show promise. See Letter p.396
Preventive vaccines are perhaps the most effective form of immunotherapy. But on page 396, Kranz et al.1 describe a vaccination strategy against cancer that targets existing tumours by recruiting immune mechanisms normally used against viral infection. The authors used nanoparticles carrying tumour RNA to simulate the intrusion of a viral pathogen into the bloodstream. When the nanoparticles reach lymphoid tissues, including the spleen and lymph nodes, they activate antiviral defence mechanisms in immune cells such as dendritic cells. The dendritic cells translate RNA obtained from the nanoparticles to express and present tumour antigens (molecules used by the immune response as attack targets) to the T cells of the immune system, priming these cells to launch an antitumour immune response.
Why is it so difficult to effectively vaccinate against cancer? One reason is that cancer cells are similar in many ways to normal cells and the immune system avoids attacking the self. Only relatively modest immune responses occur with vaccines containing antigens that are also expressed on healthy tissue. Strong immune responses can be expected only when cancer cells express antigens that are not usually expressed in normal adult cells.
Another reason is that the growth of a cancer is not accompanied by strong inflammatory signals such as those that occur during microbial infection and which initiate a strong immune response. This leads to tumour microenvironments in which immune cells tolerate, or even promote, cancer growth2. Antitumour vaccines must therefore work when the disease has already taken hold, and often when it has spread throughout the body. Last, and in a key contrast to preventive vaccinations against viruses, most cancers coexist and coevolve with our immune systems over years, resulting in an immunosuppressive tumour microenvironment that adds an extra obstacle for immunotherapy.
In vaccine approaches for a range of diseases, specialized antigen-presenting cells have a pivotal role. Dendritic cells in particular are extremely well suited to handling and presenting antigens to activate T cells. Cultured dendritic cells that have been loaded with antigens in vitro can boost immunity when given to patients with cancer, but up to now the clinical efficacy of this strategy has been limited3. Most of these vaccines use dendritic cells that have been derived in vitro from white blood cells called monocytes. Ex vivo activation of different dendritic-cell subsets that naturally circulate in the blood has also been investigated, using several types of dendritic cell including plasmacytoid dendritic cells, which produce high levels of the immune-response protein interferon-α (IFNα) upon viral infection4.
Immunologists have also explored vaccines aimed at directly activating the patient's own dendritic cells in vivo, which avoids laborious and expensive in vitro culture5. Such a vaccine requires at least three components: an 'address label' (a dendritic-cell-specific antibody or ligand molecule such as a carbohydrate)6,7,8 that targets the dendritic cell; a tumour antigen; and a compound that readies the dendritic cells to fully activate T cells (usually a ligand for a Toll-like receptor (TLR)). Nanoparticles containing antigen and TLR ligands, along with targeting antibodies or other ligands, have proved effective in animal models9, and initial clinical trials using conjugates of dendritic-cell-targeting antibodies bound to a tumour antigen are under way (see ref. 10 for examples).
Kranz et al. have developed a different type of nanoparticle vaccine that does not require antibodies or ligands to target the dendritic cells. Instead, they made nanoparticles consisting of RNA–lipid complexes11. They first demonstrated that, by making the nanoparticles slightly negatively charged by manipulating the RNA-to-lipid ratio, the particles can be directed to dendritic-cell-containing compartments in the spleen and other lymphoid tissues when intravenously injected into mice. By using nanoparticles that carried RNA encoding a fluorescent protein, the authors observed that the distribution within the body was more dependent on the overall charge of the nanoparticle than on the type of lipid used. Fluorescence was observed in antigen-presenting dendritic cells and in macrophages, another type of antigen-presenting cell (both of which express the molecular marker CD11c) in the marginal zone of the spleen and in other lymphoid organs. Fluorescence was not observed in mice depleted of CD11c-expressing cells. Plasmacytoid dendritic cells did not fluoresce but showed other signalling responses that suggest that they have taken up nanoparticles.
The researchers found that uptake of the nanoparticle RNA occurred by a cell-membrane-based process called micropinocytosis. Uptake was highest in macrophages. However, the highest expression of RNA-encoded fluorescent marker was observed in dendritic cells, indicating that they are more effective than macrophages at enabling the ingested RNA to reach the cytoplasm and be translated into protein.
Intriguingly, the authors observed two transient waves of IFNα after nanoparticle injection (Fig. 1): the first was produced by plasmacytoid dendritic cells and peaked at 2–3 hours after injection; it was followed by a macrophage-produced wave around 6–8 hours later. By testing an array of genetically modified mice, the authors show that IFNα secretion is mediated by the receptor TLR and that the first wave is necessary for precursor dendritic cells to mature and migrate to encounter T cells in the spleen and lymph nodes. This leads to a full-blown T-cell response (helped by the second wave of IFNα secretion) against a range of antigens in tumour models in mice, generating robust and long-lasting antitumour responses.
Kranz et al. extended their research to an initial clinical study in patients with melanoma, using nanoparticles carrying RNA encoding tumour antigens, and present results from the first three patients treated. Impressively, immune responses were observed — although it is still early days, and a larger, randomized trial will be needed to validate these findings. All three patients produced IFNα and developed strong T-cell responses against the immunizing antigens, even though a smaller dose of nanoparticles was used than in the mouse studies. The T-cell responses involved both CD4-type and CD8-type T cells; the activation of cytotoxic CD8 T cells is typical of an antiviral response, and having both types of T-cell response usually improves anticancer action.
The authors used intravenous injection to deliver the nanoparticles, but it would be interesting to explore other administration routes, which might alter their distribution. It would be worth examining the tissue distribution of radiolabelled nanoparticles in humans, as in the mouse experiments, to see if they also mainly target CD11c-expressing cells. Other immune-system cells that are marked by CD11c, such as neutrophils and monocytes, also have a high phagocytic capacity, and might therefore be able to take up nanoparticle materials and become activated. If so, it is not clear what contribution these other cell types might make to producing immune-system signals such as cytokines.
Kranz and colleagues' study highlights the role of IFNα in obtaining robust T-cell responses against tumours. Notably, CD8 (as well as CD4) T-cell responses were observed in both the mouse and human studies. Although CD8 T cells have long been known as the major class of immune cells acting in tumour eradication, different subtypes of dendritic cells stimulate different types of CD8 T cells3, and the contribution of CD4 T cells may have been underestimated. The responses in the three cancer patients are interesting given the different types of tumour antigen that were explored (including antigens that are not usually expressed in adult tissue and new antigens that arose owing to mutation within the tumour cells). This nanomedicine platform may give a strong boost to the vaccine field, and the results of forthcoming clinical studies will be of great interest. Footnote 1
Kranz, L. M. et al. Nature 534, 396–401 (2016).
Joyce, J. A. & Fearon, D. T. Science 348, 74–80 (2015).
Palucka, K. & Banchereau, J. Nature Rev. Cancer 12, 265–277 (2012).
Tel, J. et al. Cancer Res. 73, 1063–1075 (2013).
Tacken, P. J., de Vries, I. J., Torensma, R. & Figdor, C. G. Nature Rev. Immunol. 7, 790–802 (2007).
Bonifaz, L. C. et al. J. Exp. Med. 196, 1627–1638 (2002).
Bonifaz, L. C. et al. J. Exp. Med. 199, 815–824 (2004).
Tacken, P. J. et al. Blood 106, 1278–1285 (2005).
Cruz, L. J. et al. J. Control. Release 144, 118–126 (2010).
Kuhn, A. N. et al. Gene Ther. 17, 961–971 (2010).
Related links in Nature Research
About this article
Cite this article
De Vries, J., Figdor, C. Cancer vaccine triggers antiviral-type defences. Nature 534, 329–331 (2016). https://doi.org/10.1038/nature18443
Frontiers in Immunology (2020)
Journal of Veterinary Science (2019)
International Journal of Molecular Sciences (2018)
Co-delivery of tumor antigen and dual toll-like receptor ligands into dendritic cell by silicon microparticle enables efficient immunotherapy against melanoma
Journal of Controlled Release (2018)