Cytotoxic Activity Of Bispecific Antibody-Redirected Human Regulatory T Cells: Fact Or Artifact

Regulatory T cells (Tregs) play an inevitable role in immune homeostasis by maintaining self-tolerance as well as regulating the magnitude of immune responses against foreign antigens. Over the last few years, the enormous potential of adoptive Treg transfer for treatment of auto- and alloimmunity including Graft-versus-Host disease (GvHD) has been validated in a vast number of in vitro and in vivo studies. For their clinical application, all modes of action should be well understood. Regarding their cytotoxic potential, only few and conflicting data exist. On the one hand, it is assumed that Tregs are capable of inducing apoptosis of effector T cells (Teff) utilizing granzyme/perforin or FasL expression. Others claim that Tregs are not capable of suppressing Teff via programmed cell death pathways but rather induce apoptosis by cytokine deprivation. However, it is of importance to clarify whether Tregs possess a cytotoxic potential particularly when activating the cells antigen-specifically using bispecific antibodies (bsAb).

In recent years, bsAb have emerged as promising tools for an antigen-specific immunotherapy of malignant diseases. Their tremendous potential for tumor therapy has been verified in a plethora of in vitro and in vivo studies as well as in first clinical trials. So far, our group was able to demonstrate that not only Teff but also Tregs can be redirected by CD3-engaging bsAb (Koristka et al., J Immunol. 2012; J Autoimmun. 2013). According to a recent presentation (Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research, 2012, abstract nr 4841), bsAb-redirected Tregs can act as killer cells and efficiently mediate cancer cell lysis. In order to shed light onto this controversial issue, we decided to analyze this question in more detail.

According to our investigations tumor cell elimination of bsAb-engaged Tregs is largely dependent on the purity of isolated Treg fractions. Tregs isolated on the basis of CD25 expression exhibited a remarkable killing capacity which is most probably due to contaminating CD25⁺FOXP3^- Teff, as highly pure (> 99 %), FACS-isolated CD4⁺CD25⁺CD127^low Tregs did not display any considerable cytotoxic effect upon cross-linkage to tumor cells via bsAb. The same applies for CD45RA-sorted, expanded Tregs. In comparison to autologous, expanded Teff, tumor cell lysis was negligible. Moreover, the lack of cytotoxicity was independent of the chosen target antigen, as redirecting Tregs with two different bsAb did not result in tumor cell eradication. Besides, upon polyclonal stimulation with conventional aCD3/CD28-coated beads Tregs were not capable of eliminating target cells. Furthermore, as opposed to autologous Teff, Tregs showed only a marginal upregulation of the degranulation marker CD107a when being activated either antigen-specifically via bsAb or polyclonally via beads.

Taken together, our findings clearly demonstrate that Tregs bear no considerable cytotoxic potential and hence do not contribute to cancer cell lysis, as recently claimed. On the other hand, the results show that Tregs can be activated by bsAb without the risk of cytotoxic effects against the recognized target cells. This provides the basis for the application of bsAb for a site-specific recruitment of Tregs aiming at attenuating Teff-mediated proinflammatory immune responses and tissue destruction in order to treat auto- and alloimmune diseases including GvHD.