To the Editor,

Multiple myeloma is an aggressive, malignant, and incurable disease characterized by neoplastic plasma cell clone proliferation [1]. Poor prognoses of MM patients may be related to T cell immunodeficiency [2]. Recent findings have indicated that aberrant expression of immune checkpoint (IC) proteins such as programmed cell death receptor-1 (PD-1) and T cell immunoglobulin mucin-domain-containing-3 (Tim-3) is a key reason for T cell immune suppression though the promotion of T cell exhaustion [2, 3]. Up-regulation of PD-1 and other IC proteins, such as Tim-3, on CD4 + and CD8 + T cells has been detected in PB from patients with MM [2, 4]. Immunotherapy based on targeting ICs, such as PD-1 blockade, improves the clinical outcome of solid tumors and lymphoma in clinical trials, and the underlying mechanism is thought to reverse the immunosuppressive status of T cells and restore their anti-tumor ability in patients [5]. However, even with PD-1 over-expression on exhausted T cells, the effects of PD-1 blockade appear to be limited and heterogeneous for MM patients [6, 7]. These observations may be related to different immunosuppressive microenvironments and the expression pattern of ICs between solid tumors and MM [2]. Recently, it has been reported that over-expression of TOX (thymocyte selection-associated HMG BOX), a crucial transcription factor involved in T cell exhaustion, is detected in CD8 + tumor-infiltrating lymphocytes (TILs) in bladder cancer, and this is related to PD-1 expression on T cells [8, 9]. To further characterize the alternative expression profile of IC proteins and co-expression with their regulatory factors, we analyzed the expression of TOX and TOX co-expression with PD-1, Tim-3, and CD244 in T cells by multi-color fluorescent flow cytometry in peripheral blood (PB) and bone marrow (BM) samples from 16 patients with MM (Additional file 1: Supplementary Methods and Additional file 3: Table S1). Significantly, the percentages of TOX + CD3 +/CD4 +/CD8 + T cell subsets were all increased, and higher numbers of TOX co-expressed with PD-1, Tim-3, or CD244 in CD3 +/CD4 +/CD8 + T cells were found in both PB and BM from patients with MM in comparison with healthy controls (Fig. 1A, B). This result is consistent with the finding of up-regulation of TOX in solid tumors and lymphomas [10]. However, as the heatmap shows in Fig. 1C, the frequency of TOX and co-expression with PD-1, Tim-3, and CD244 in CD3 +, CD4 +, and CD8 + T cells relatively varied between different MM patients, and did not appear to be associated with different stages of MM. Interestingly, a higher frequency of the TOX + T cell subset can be also found in stage I MM (Fig. 1C). The global distribution and frequency of different phenotypes of T cells in the BM and PB of patients with MM and HI can be represented by tSNE clusters (Fig. 1D). Our previous study demonstrated that the level of PD-1 + Tim-3 + CD3 +/CD4 +/CD8 + T cells was high in the BM when compared with PB [2]. In this study, we also compared the percentage of the TOX + T cell subsets in 16 pairs of PB and BM samples from MM patients. Overall, a high percentage of TOX + T cell subsets could be found in the BM in comparison with that in PB in most cases; however, these were not statistically significant (Fig. 1B, Additional file 2: Figure S1) except for TOX + Tim-3 + regulatory T (Treg) cells, which were significantly higher in BM than in PB (Fig. 2C). Interestingly, the numbers of TOX + Tregs and TOX + PD-1 +/Tim-3 + Tregs significantly increased in the PB and BM (Fig. 2A, B). Our previous study also revealed an increase in TOX + Treg cells in patients with lymphoma [11]. However, the role of Treg cells with higher TOX and PD-1 or Tim-3 in MM pathogenesis, prognosis, and treatment remains unclear. Unlike high expression of PD-1 on CD8 + T cells induces exhaustion, PD-1 expression on Tregs negatively impacts immunosuppressive functions [11], moreover, TOX promotes the exhaustion of antitumor CD8 + T cells by preventing PD-1 degradation due to the binding of TOX to PD-1 in the cytoplasm and maintaining abundant PD-1 expression at the T cell surface [12]. Suggesting that the role of TOX may contribute to maintain PD-1 expression on Treg cells which may enhance negative immune regulatory in MM. In this case, whether targeting TOX has a dual inhibitory function remains an open question.

Fig. 1
figure 1

Distribution and frequency of TOX expression and co-expression with PD-1, Tim-3, and CD244 in T cell subsets in PB and BM from patients with MM. A The analytic logic of flow cytometry detection of TOX expression and co-expression with PD-1, Tim-3, and CD244 in CD3 +, CD4 +, and CD8 + T cell subsets in PB from a patient with MM and a healthy individual (HI). B Comparison of the percentage of TOX + CD3 + T cells (median: BM: 13.2 vs 2.06, P = 0.008, PB: 15.9 vs 6.145, P < 0.001), TOX + CD4 + T cells(median: BM: 25.2 vs 4.57, P = 0.004, PB: 21.85 vs 12.95, P = 0.019), TOX + CD8 + T cells (median: BM: 16.2 vs 2.23, P = 0.002, PB: 21.6 vs 7.51, P < 0.001), TOX + PD-1 + CD3 + /CD4 + /CD8 + T cells (median: BM: 5.575/9.19/5.6 vs 0.51/1.02/0.6, P = 0.008, P = 0.002, P = 0.004, respectively, PB: 5.04/7/4.985/ vs 1.495/3.76/1.555, P < 0.001, P < 0.001, P < 0.001, respectively), TOX + Tim-3 + CD3 + /CD4 + /CD8 + T cells (median: BM: 0.425/0.51/0.58 vs 0.00706/0.03/0.00747, P = 0.002, P = 0.002, P = 0.002, respectively, PB: 0.295/0.34/0.455 vs 0.063/0.0865/0.068, P < 0.001, P = 0.004, P < 0.001, respectively), and TOX + CD244 + CD3 + /CD4 + /CD8 + T cells (median: BM: 11.35/6.535/14.8 vs 1.71/0.81/1.85, P = 0.004, P = 0.004, P = 0.002, respectively, PB: 11.2/3.645/20.25 vs 4.36/2.435/6.755, P < 0.001, P = 0.171, P < 0.001, respectively) in BM and PB from patients with MM and HIs. C Heatmap representing the frequency of TOX +, TOX + PD-1 +, TOX + Tim-3 +, and TOX + CD244 + cells in T cell subsets in PB from 16 patients (stage I (2 cases), stage II (7 cases) and stage III (7 cases) with MM compared with HIs. D tSNE clusters of the global distribution and frequency of different phenotypes of T cells in the BM and PB of patients with MM and HIs. Note: P1–P16: MM patients who are numbered according to collection time

Full size image
Fig. 2
figure 2

Distribution and frequency of TOX expression and co-expression with PD-1, Tim-3, and CD244 on Treg cells in BM and PB from MM patients. A The analytic logic of flow cytometry detection of TOX expression and co-expression with PD-1, Tim-3, and CD244 in Treg cells in PB from a patient with MM and a HI. B Comparison of the percentage of TOX + Treg cells (median: BM: 47.35 vs 10, P = 0.008, PB: 48.55 vs 28.45, P = 0.010), TOX + PD-1 + Treg cells (median: BM: 11.75 vs 0.75, P = 0.002, PB: 9.3 vs 4.21, P = 0.002), TOX + Tim-3 + Treg cells (median: BM: 2.045 vs 0, P = 0.002, PB: 1.04 vs 0.22, P = 0.004), and TOX + CD244 + Treg cells (median: BM: 0.48 vs 0, P = 0.303, PB: 0.37 vs 0.325, P = 0.838) in BM and PB from patients with MM and HIs. C Comparison of the percentage of TOX +, TOX + PD-1 +, TOX + Tim-3 +, and TOX + CD244 + Treg cells between PB and BM from 16 patients (P1–P16) with MM

Full size image

Taken together, first, our findings indicate increased TOX expression in T cells in MM patients. Second, TOX co-expression with PD-1, Tim-3, and CD244 in T cells may be involved in promoting T cell exhaustion and impairing their function in MM. Third, higher TOX + Treg subsets in the BM, may contribute to mediating the BM immunosuppressive microenvironment, which may also be a reason why the effects of PD-1 blockade are relatively different in different MM patients. Understanding the exhausted phenotype pattern of T cells in different MM patients may help guide precision immunotherapy for MM patients.

In summary, we characterized the distribution of TOX expression in T cell subsets in MM patients. Increased TOX concurrent with PD-1, Tim-3, and CD244 in T cells may be considered a potential target for reversing T cell exhaustion and improving T cell function in MM.