An Essential Role for Programmed Death-1 in the Control of Autoimmunity: Implications for the Future of Hematopoietic Stem Cell Transplantation

Keywords:

• autoimmunity
• hematopoietic stem cell
• immune reconstitution
• tolerance

Co-inhibitory receptors are expressed by lymphocytes and other immune system cells. Their central function is to control the activation of the lymphocyte responses by providing negative signals in conjunction with signals from lymphocyte antigen receptors (‘cosignaling‘). Recently, co-inhibitory molecules have been extensively studied due to the broad spectrum of their potential application in various novel immunotherapeutic approaches. The cosignaling aspect of these receptors is particularly appealing for clinical development, as nondepleting antibodies targeting these co-inhibitory receptors will only affect lymphocytes that are encountering (or have recently encountered) their cognate antigen; ligation of co-inhibitors alone, without antigen receptor signals, has no effect. Even though it is now 40 years since the initial concept of co-inhibitory signaling was pioneered [1] (and reviewed in [2]), only recently have these signaling pathways been specifically targeted for therapy. Agonistic monoclonal antibodies to co-inhibitors are a promising new approach in the prevention of transplant rejection and graft-versus-host disease, and may also find application in treating autoimmune disease. Monoclonals that instead block these co-inhibitors are currently undergoing clinical trials in cancer treatment, and one such co-inhibitory blocker, specific to cytotoxic T-lymphocyte antigen 4, has recently achieved US FDA approval. Programmed death-1 (PD-1; CD279), is another important co-inhibitor. It is involved in the regulation of immune responses and self-tolerance [3]. PD-1 has two known ligands: PD-L1 (B7H1; CD274) and PD-L2 (B7DC; CD273). A recent study suggested that PD-1 may have a specific role in controlling T cells during lymphopenia, although a requirement for PD-1 in the control of lymphopenia-driven autoimmunity was not examined [4]. Autoimmunity can only rarely be attributed to variants of a single gene alone. Instead, discovery of the causes of most autoimmunity will be found in synergisms between alleles of many different genes and their interactions with the environment.

Lymphopenia-induced homeostatic proliferation (LIP) of T cells has frequently been associated with autoimmunity. Although recipient conditioning followed by syngeneic hematopoietic stem cell (HSC) transplantation does lead to LIP of T cells, it does not generally lead to autoimmunity, suggesting that control mechanisms such as co-inhibition may be at play. PD-1 has previously been shown to have a relatively limited role in self-tolerance. Loss of PD-1 function led to autoimmunity directed at a single organ or a late-life lupus/arthritis in only a fraction of mice. This raised the question of factors that might synergize with a deficiency in PD-1 function to lead to a more severe autoimmunity with higher incidence. Surprisingly, transfer of established peripheral T cells lacking PD-1 into a lymphopenic setting did not cause severe autoimmunity [5]. However, by using HSC transplantation in a murine model, we recently demonstrated that PD-1 function is essential in newly generated T cells to prevent a rapidly lethal lymphopenia-induced multiorgan autoimmunity [5]. The severity and reproducibility (100% of mice) of the phenotype suggested that the most critical function of PD-1 resides in controlling the first waves of T cells newly exported from the thymus, as they respond to as-yet unidentified lymphopenia-potentiated signals. Similar to lymphopenia, PD-1 deficiency by itself does not usually lead to such a severe multiorgan autoimmunity. Our data suggested that this may be due to the fact that the only stage where newly generated T cells naturally predominate is during the fetal/neonatal period, a time when lymphoid tissue is not yet fully developed and therefore has a reduced ability to support LIP of PD-1-deficient T cells. Supporting this view, lymphocyte-deficient adults, but not neonatal recipients of PD-1-deficient HSCs, succumbed to autoimmunity; adult recipients could also be protected from disease if they lacked lymph nodes. In this article, we will discuss the implications of these findings in designing future cancer immunotherapies.

Co-inhibitory pathways have been exploited by tumors as an immune evasion mechanism [6]. There is a negative correlation between the expression of co-inhibitory molecules such as PD-L1 and prognosis in the cancer patients [7,8]. Strategies have been designed to overcome this hurdle by blocking co-inhibitory pathways and are considered to have high therapeutic potential. For example, blockade of PD-1 inhibited the dissemination of CT26 colon carcinomas and B16 melanomas by increasing the priming and homing of effector T cells [9]. Furthermore, adoptive transfer of PD-1-deficient CD8 T-cell receptor transgenic T cells caused the rejection of tumor cells due to their increased cytolytic activity and cytokine responses [10].

Dendritic cell vaccines have been used in animal models [11,12], as well as in humans, to induce antitumor immunity [13]. Although vaccines against cancer can induce a high number of antitumor T cells, it does not always lead to a protective immune response, and therefore clinical efficacy is poor. Accumulating evidence suggests that the dysfunction of these tumor-infiltrating lymphocytes is due to the upregulation of PD-1 expression [14,15]. Studies from melanoma patients revealed that PD-1 is highly expressed by endogenous antimelanoma-specific T cells, melanoma vaccine-induced specific T cells and tumor-infiltrating lymphocytes [16–18]. In the case of lymphoma, an increase in the absolute number of tumor cells was associated with increased PD-1 expression of CD4 T cells [19]. Interestingly, combining a tumor vaccine with anti-PD-1 augmented the antitumor responses and also vaccine efficacy [15,20].

Similar to animal studies [15,20], an ex vivo preclinical study reported that the combination of anti-PD-1 with dendritic cell–myeloma fusions enhanced antitumor responses [21]. Treatment of cancer patients with anti-PD-1 is an exciting new approach and the recent clinical trials have reported the beneficial effects of the antibody, with the potential for fewer autoimmune side effects than anti-cytotoxic T-lymphocyte antigen 4 [22,23]. Encouraged by these results, a clinical trial has already been started for multiple myeloma patients, which involves high-dose chemotherapy and autologous HSC transplantation, followed by dendritic cell–myeloma fusions with anti-PD-1 [101].

High-dose chemotherapy is used to treat advanced cases of cancers to eradicate cancer cells more effectively. However, the downside of the high-dose regimens is the killing of healthy cells, in particular hematopoietic cells, as this makes it necessary to give a HSC transplant to these patients. High-dose chemotherapy followed by HSC transplantation has become a standard regimen to treat cancers such as multiple myeloma and lymphoma [24–26]. Autologous stem cell transplantation has an advantage over allogeneic stem cell transplantation due to a lack of graft-versus-host disease [27], although the graft-versus-leukemia effect is lost with this approach.

A study by our group suggests the potential for the generation of significant autoimmune disease in patients where the immune system is ‘restarted’, such as patients undergoing therapies that are lymphoablative followed by autologous HSC transplantation with anti-PD-1 treatment [5]. Even without anti-PD-1 treatment, patients with alleles of PD-1 that demonstrate reduced function would be anticipated to be at higher risk of developing autoimmunity due to the newly generated T cells post-HSC transplantation. In these patients, lymphopenia due to ablation will allow the newly generated T cells to expand and, together with the reduced functioning of PD-1, these newly generated T cells can acquire the capacity to induce autoimmunity. The development of agonistic antibodies to PD-1 or other co-inhibitors might be helpful to mitigate the autoimmunity in the aforementioned settings.

In the case of clinical trials with anti-PD-1 treatment, even the patients with fully functional alleles of PD-1 would be at risk of developing autoimmunity, due to the disruption of PD-1 function. However, there are a few caveats to be considered in interpreting the murine data in relation to the approaches undergoing clinical trials. Firstly, the lymphopenia in most of the murine studies was absolute, employing recipient mice that completely lacked both T and B cells due to the absence of the Rag gene. In the subset of experiments where recipients were instead lymphoreplete (immunocompetent), conditioning with a supralethal dose of irradiation was required to observe the severe disease. The presence of a ‘competitor’ population of T cells substantially blocked disease, perhaps explaining why recipients with lower doses of irradiation did not succumb to severe disease [Thangavelu G, Anderson CC, Unpublished Data]. Lymphopenia induced in the clinical setting is not absolute, and the remaining T cells resistant to chemotherapy or other depletional regimens may inhibit the autoimmunity caused by newly generated T cells. From this viewpoint, any efforts to develop protocols that more fully ablate the patients’ hematopoietic system would be ill advised if combined with dysfunction of co-inhibitors (due to patient-specific alleles or blocking-antibody treatment). A second caveat is that the pattern of immune reconstitution differs substantially between mouse models and the human setting. Immune reconstitution post-HSC transplantation in humans is notoriously slow and incomplete. Any efforts to augment immune reconstitution, such as through the development of methods that increase the patient’s thymic function, would increase the output of newly generated T cells and the potential for autoimmunity. Despite these caveats, future clinical trials involving high-dose chemotherapy/lymphoablation with stem cell transplantation may need to consider polymorphisms regulating the PD-1 pathway in protocol development. Given the central role of these co-inhibitory signals upon immune system start-up, any attempts to ‘perfect’ ablation and immune reconstitution protocols will need to be carefully planned if severe autoimmunity is to be avoided.

Financial & competing interests disclosure

The authors are supported by studentships from the Muttart Diabetes Research and Training Centre and Alberta Diabetes Institute to G Thangavelu and a senior scholar award from the Alberta Heritage Foundation for Medical Research to CC Anderson. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.

Additional information

Funding

The authors are supported by studentships from the Muttart Diabetes Research and Training Centre and Alberta Diabetes Institute to G Thangavelu and a senior scholar award from the Alberta Heritage Foundation for Medical Research to CC Anderson. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.