Immune Checkpoint Blockade Enhances Mutated Calreticulin-Induced T Cell Immunity in Myeloproliferative Neoplasms

Somatic mutations in the calreticulin (CALR) gene are key drivers of cellular transformation in myeloproliferative neoplasms (MPN) and are the second most common driver mutations in essential thrombocythemia (ET) and myelofibrosis (MF) patients. Each CALR mutation leads to a shift in the reading frame resulting in the formation of an altered protein with an identical 36-amino acid sequence in the C-terminus. We hypothesized that the recurrence and uniformity of the novel sequence would provide the mutated CALR an attractive candidate as a MPN-specific tumor antigen that might elicit anti-tumor immune responses. Therefore, we investigated the immunogenicity of the altered CALR by employing in silico peptide binding prediction algorithms to interrogate the binding affinities of mutated CALR peptides to human leukocyte antigens (HLA). These analyses indicated that the mutated epitopes can bind to class I and II HLA alleles with high affinity supporting the hypothesis that mutant CALR could be immunogenic in vivo . We then tested the immunogenicity of the overlapping peptides spanning the mutated region in vitro by utilizing intracellular staining (ICS) and Enzyme-Linked ImmunoSPOT (ELISPOT) assays. Naïve T cells derived from the peripheral blood mononuclear cells (PBMNCs) isolated from healthy donors displayed effector functions after priming with the mutated peptides but not with the corresponding wild type peptides . Incubation withthe mutated CALR peptides induced T cells proliferation, upregulation of CD137 (4-1BB) and production of interferon (IFN)-γ and tumor necrosis factor (TNF)-α. Moreover, as suggested by in silico peptide-HLA binding predictions, both CD8⁺ and CD4⁺ T cell precursors displayed increased IFN-γ production in response to stimulation with mutated CALR peptides. There was an overall 4±3-fold increase in IFN-γ production by CD4⁺ T cells and a 2.5±0.5-fold increase by CD8⁺ T cells after stimulation with mutated peptides as compared to control self-antigens (n = 14 healthy donors out of which 11 showed CD4⁺ and 5 showed CD8⁺ T cell responses). These results demonstrate that mutated CALR elicits T cell responses in vitro supporting its potential immunogenic functions in vivo .

We next investigated the presence of mutant CALR-specific T cell immunity in MPN patients with CALR mutations. PBMNCs from a total of 18 CALR-mutated MPN patients with ET (7), MF (5) or MF arising from ET (6) were evaluated. MPN T cells were then stimulated with the pooled mutant CALR peptides and T cell responses were assessed by ELISPOT. Out of 18 patients, increased IFN-γ production upon stimulation was observed in only 2 patients (one with ET and one with ET-MF). The less than robust MPN T cell responses suggested that immune suppressive mechanisms might be operating. To this end, we evaluated the expression of cell surface inhibitory receptors, programmed cell death (PD)-1 and the cytotoxic T-lymphocyte-associated antigen (CTLA)-4, by T cells from MPN patients. Both CD8⁺ and CD4⁺ T cells from MPN patients exhibited increased levels of PD-1 and CTLA-4 when compared to the healthy controls (HC). For instance, 9.3±8.2% of MPN patients CD4⁺ T cells expressed PD-1 (n = 12) as compared to 2.5±1.7% of HC CD4⁺ T cells (n=10) (p value = 0.0482). Similarly, there were twice as many PD-1 expressing CD8⁺ T cells in MPN patients as compared to HC. We next assessed MPN T cell responses against mutated peptides in the presence of neutralizing PD-1 or CTLA-4 antibodies. PD-1 inhibition restored T cell immunity in cultures derived from 3 MPN patients and CTLA-4 inhibition restored response in one patient (n=16). Of note, following the PD-1 or CTLA-4 inhibition, the number of IFN-γ producing cells and the amount of IFN-γ produced on a per-cell-basis increased upon stimulation with mutant-CALR peptides as compared to control peptides. These results indicate that the limited MPN T cell responses to CALR mutated peptides were due to T cell exhaustion caused, at least in part, by activation of the PD-1/CTLA-4 pathway.

In summary, our studies establish mutated CALR as a MPN-specific tumor antigen thereby providing a rationale for the development of immunotherapies targeting mutated CALR in patients. Moreover, our data indicate that immune checkpoint signaling might be responsible for T cell exhaustion in some MPN patients thereby supporting the use of checkpoint blockade inhibitors for treatment of these patients.