Boolean Immunotherapy: Reversal of Fortune

In this issue of Molecular Therapy, Leen et al. describe an approach to exploit the inhibitory tumor microenvironment to enhance antitumor immune responses.¹ Tumors produce a range of inhibitory cytokines and signals that reduce the ability of infiltrating T cells to recognize and destroy the tumor (reviewed in refs. 2,3,4). One such tumor-derived signal is the cytokine interleukin 4 (IL-4), and in this report Leen and colleagues express on T cells a chimeric cytokine receptor comprising a fusion between the extracellular binding domain of the IL-4 cytokine receptor and the intracellular signaling portion of the receptor for IL-7, a cytokine that potentiates the growth and activity of effector T cells. The approach has potential application to the field of adoptive immunotherapy using genetically engineered T cells, in that it presents a strategy to reverse the immunoinhibitory components of the tumor microenvironment and instead use them to activate T cells and augment cell-mediated antitumor efficacy.

The new study describes an approach to cancer therapy that involves the transfer of ex vivo–expanded T cells with antitumor activity into a tumor-bearing host in a process called adoptive cell therapy (ACT). ACT had been shown to be clinically effective in autologous lymphocyte infusions for viral-associated malignancies⁵ and in melanoma using autologous tumor-infiltrating lymphocytes,⁶ but did not generate significant interest until it was shown that genetic modification of normal T cells with antitumor antigen receptors is effective in a variety of tumor types. The accelerated pace of development of gene-modified T-cell ACT has been largely driven by excellent clinical responses, first reported by the National Cancer Institute⁷ and later confirmed by others,⁸ that chimeric antigen receptors (CARs) targeting the B-cell antigen CD19 could yield durable complete responses in a high proportion of patients. In general, these results have not been repeated in solid tumors, whose inhibitory microenvironment has been suggested as a major element in this lack of success. Beyond the inhibitory environment, the lack of truly tumor-specific antigens on solid tumors has led to safety issues with regard to “on-target antigen but off-target tumor” toxicities. The data presented in the report by Leen et al. begin to address both negative issues and turn them into positive attributes.

The investigators fused the extracellular binding domains of the IL-4 receptor α-chain (IL-4Rα) with the intracellular signaling domains of the IL-7 receptor α-chain (IL-7Rα) (Figure 1) to create a chimeric cytokine receptor.⁹ IL-4 is one of several tumor-associated cytokines that can have inhibitory effects on antigen-reactive T cells, for example, by initiating a gene expression cascade that can be associated with immune suppression. Thus, when IL-4 binds to its natural receptor it activates transcription factors, which in turn can lead to the suppression of the important effector cytokine interferon γ. Conversely, signaling through the normal IL-7 receptor transmits signals that promote T-cell growth and survival. When the investigators introduced the chimeric IL-4Rα/IL-7Rα gene into T cells and then exposed them to IL-4, microarray analysis demonstrated that the gene expression profile resembled that obtained from T cells that had been exposed to IL-7. Thus, a potentially negative cytokine signal was reversed and turned into a positive signal!

Figure 1.

Reversing tumor inhibitory signals. Diagram of the chimeric cytokine receptor approach to the reversal of tumor-derived inhibitory cytokines. Upon interleukin 4 (IL-4) cytokine binding to the chimeric receptor, a positive T helper type 1 cell (Th1)-like growth signal is transmitted to the antitumor antigen T cell via STAT5-mediated signaling events.

To test the activity of their approach in a tumor system, the authors introduced the IL-4Rα/IL-7Rα gene into naturally occurring T cells specific against Epstein-Barr virus (EBV). They then implanted EBV⁺ tumor cell lines secreting IL-4 into immunodeficient mice. The tumor-bearing animals received control EBV-specific T cells or the same EBV-specific T cells engineered with the IL-4Rα/IL-7Rα gene. Animals receiving the IL-4Rα/IL-7Rα-engineered cells had significantly greater reduction in tumor volume than controls, and a third of them survived long-term (as long as 100 days), whereas none of the control treatment group survived beyond day 55.

One obvious concern in using these types of hybrid receptors is that the growth of the cells could become autonomous. Fortunately, however, the proliferation and activation of the chimeric cytokine receptor–engineered T cells remained both antigen- and cytokine-dependent.

As described in the discussion of the report by Leen et al., such engineering approaches have the potential to introduce protein-based Boolean operators into cells that can sense the environment and make “AND,” “OR,” and “NOT” decisions. In this case, the T cell becomes optimally activated when it “sees” both its antigen via the anti-EBV T-cell receptor “AND” IL-4 produced by the tumor, which is recognized by the chimeric receptor. Absent one or the other signal, activation and expansion do not occur. Other examples of Boolean “AND” operators have been developed that included systems that require two tumor antigens for optimal T-cell activation,¹⁰ and negative/positive signals in T cells have also been reversed using PD-1/CD28 chimeric proteins.¹¹ Clearly this is just the start of the use of these types of approaches to optimize antitumor specificity and activity of ACT.

IL-4 is only one of a number of potentially immunosuppressive factors that an antitumor T cell is likely to encounter in the tumor microenvironment. It remains to be studied how the IL-4Rα/IL-Rα-engineered T cells would respond if they faced a combination of inhibitory factors, such as transforming growth factor β (TGF-β), and IL-10. TGF-β and IL-10 are more widely expressed in the tumor microenvironment than IL-4, and developing similar approaches targeting these cytokines might be a more relevant approach to increase the specificity and reactivity of ACT. Finally, the generality of this approach requires demonstration with other tumor antigen–specific T cells as well as T cells engineered with artificial antigen receptors (e.g., anti-CD19 CARs).

Cancer immunotherapy using ACT has been shown to be effective in a select subset of malignancies, but this does not include the most common cancers such as breast, colon, and lung cancer, where these approaches are largely ineffective and could be dangerous because they may have to target tumor antigens that are also expressed on normal tissues. The reverse-inhibitory Boolean operator genes such as those reported by Leen et al. are a first step in attempts to develop both potent and safe T cells for cancer ACT.

Conflict of Interest

The author is an employee of bluebird bio Inc. The Baylor College of Medicine, Celgene, and bluebird bio have entered into an agreement for the potential commercial development of anticancer T cells engineered with chimeric antigen receptors.