Regulatory T cells suppress myeloma-specific immunity during autologous stem cell mobilization and transplantation

Abstract

Autologous stem cell transplantation (ASCT) is the standard of care consolidation therapy for eligible patients with myeloma but most patients eventually progress, an event associated with features of immune escape. Novel approaches to enhance antimyeloma immunity after ASCT represent a major unmet need. Here, we demonstrate that patient-mobilized stem cell grafts contain high numbers of effector CD8 T cells and immunosuppressive regulatory T cells (Tregs). We showed that bone marrow (BM)-residing T cells are efficiently mobilized during stem cell mobilization (SCM) and hypothesized that mobilized and highly suppressive BM-derived Tregs might limit antimyeloma immunity during SCM. Thus, we performed ASCT in a preclinical myeloma model with or without stringent Treg depletion during SCM. Treg depletion generated SCM grafts containing polyfunctional CD8 T effector memory cells, which dramatically enhanced myeloma control after ASCT. Thus, we explored clinically tractable translational approaches to mimic this scenario. Antibody-based approaches resulted in only partial Treg depletion and were inadequate to recapitulate this effect. In contrast, a synthetic interleukin-2 (IL-2)/IL-15 mimetic that stimulates the IL-2 receptor on CD8 T cells without binding to the high-affinity IL-2Ra used by Tregs efficiently expanded polyfunctional CD8 T cells in mobilized grafts and protected recipients from myeloma progression after ASCT. We confirmed that Treg depletion during stem cell mobilization can mitigate constraints on tumor immunity and result in profound myeloma control after ASCT. Direct and selective cytokine signaling of CD8 T cells can recapitulate this effect and represent a clinically testable strategy to improve responses after ASCT.

Graphical abstract

Figure 1.

Mobilized grafts from patients with myeloma contain high numbers of effector/effector memory CD8 T cells and Tregs. BM and PBSC were collected from the same patients with myeloma (n = 14-18). We performed FACS analysis to determine the phenotype of T cells in the BM and PBSC. Data from the BM and PBSC were combined and unbiased clustering was performed to visualize each population using FlowSOM. A heatmap was generated based on the expression levels of the flow cytometry markers. (A) Representative FACS plots and frequency of naïve and memory subsets in CD8 T cells. (B) Heatmap of marker expression and the frequency of each CD8 T-cell population. (C) Representative FACS plots and frequency of naïve and memory subset of conventional CD4 T cells. (D) Heatmap of marker expression and the frequency of each CD4 T-cell population. (E) Representative FACS plots of CD25⁺CD127⁻CD4 Tregs and the proportion of TIGIT⁺ Treg and TIGIT⁺CD39⁺ Treg. Each red dot and blue square represent a single BM and PBSC sample, respectively. Paired BM and PBSC are connected by solid line. (F) Frequency of TIGIT⁺ Treg and TIGIT⁺CD39⁺ Treg in PBSC from healthy donors (HD) and patients with myeloma (Pts). Wilcoxon matched pairs signed rank test for paired comparison and Welch test for 2 samples comparison (n = 18 per group). (G) Months to progressive myeloma after stem cell transplantation in cytogenetic high-risk patients with at least 12 months of follow-up relative to Treg within the stem cell grafts. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

Figure 2.

CD8 T cells from the BM egress into the PB during stem cell mobilization. (A) T cells were harvested from the BM and spleen of MM-bearing mice (CD45.1) and transferred to naïve B6 mice (CD45.2). Recipient mice were euthanized 1 week after adoptive transfer for analysis. The absolute count of total and CD44⁺ memory CD8 T cells (CD45.1) in the BM, blood, and spleen (n = 7 per group from 2 independent experiments). (B) BM T cells from MM-bearing mice (CD45.1) were harvested and transferred to naïve B6 mice (CD45.2). B6 mice were treated with G-CSF⁺AMD3100 or PBS 1 week after T-cell transfer and then euthanized to measure the absolute counts of transferred T cells (CD45.1) in the BM, blood, and spleen. Donor T cells are shown as fold change compared with the number in the nonmobilized group (n = 7 per group from 2 independent experiments). (C) Female MataHari HY-specific TCR transgenic CD8 T cells (CD45.1) were transferred into male (HY⁺) VkMYC (Vk28158)-bearing female RAG2/IL2rg KO mice, followed by G-CSF⁺AMD3100 or control vehicle injection. Total number of CD45.1⁺ MataHari CD8 T cells is shown (n = 4 per group from 2 independent experiments). Data represents mean ± SEM. Welch t test or Mann-Whitney test for 2 samples comparison. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001.

Figure 3.

Bone marrow Tregs with an immunosuppressive phenotype are mobilized into the blood during SCM. FoxP3-GFP-DTR mice were treated with G-CSF and AMD3100 or PBS for 5 days. BM and blood were collected from the treated mice after mobilization. The number and phenotype of Tregs in the BM, blood, and spleen were assessed by flow cytometry. (A) Total number of CD8 T cells, conventional CD4 T cell, and Tregs in the BM and blood. Tregs were identified using expression of green fluorescent protein (GFP). (B) Representative FACS plots of Tregs in the BM and blood from naïve mice. The proportion of TIGIT⁺Nrp-1⁺ and TIGIT⁺KLRG-1⁺ are shown. (C) Total number of TIGIT⁺Nrp-1⁺ and TIGIT⁺KLRG-1⁺ Tregs in the BM and blood with and without mobilization. (D) Heatmap depicts the expression levels of each marker on Tregs in mobilized and nonmobilized blood. The total number of each population is shown. (E) CD8/Treg and CD8/immunosuppressive Tregs (Pop3) with or without mobilization. Data are combined from 2 independent experiments. Data represents mean ± SEM. Mann-Whitney test for 2 sample comparison. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

Figure 4.

Treg depletion during SCM increases the total number of polyfunctional CD8 T cells in donor grafts. FoxP3-DTR mice were treated with diphtheria toxin (DT) or vehicle control every other day for 1 week. Mice in the mobilization group were treated with G-CSF and AMD3100. The mice were euthanized and T cells in the BM, blood, and spleen were harvested and analyzed by flow cytometry. (A) Fold changes in the number of CD8 and CD4 T cells. The fold change was calculated using the mean T-cell numbers from the non-Treg-depleted, nonmobilized group (grey triangles). (B) Representative FACS plots and frequency of memory CD8 T cells in the blood. (C) Percentage of activated DNAM-1⁺TIGIT⁺ and CD38⁺CD101⁻ CD8 T cells in the blood and spleen. (D) Frequency of IFNγ⁺TNF⁺ CD8 T cells in the blood and spleen (n = 3-8 per group from 2 independent experiments). Grey triangle: non-Treg depleted and nonmobilized, Green triangle: non-Treg depleted and mobilized, Blue square: Treg depleted and nonmobilized, Red circle: Treg depleted and mobilized. Data represents mean ± SEM. One-way ANOVA for multiple sample comparisons. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

Figure 5.

Myeloma-experienced CD8 T cells generated during SCM with Treg depletion mediate potent myeloma-specific immunity after ASCT. MM-bearing FoxP3-DTR donor mice were treated with DT or vehicle control during mobilization with G-CSF and AMD3100. T cells were harvested from spleens containing mobilized cells unless otherwise specified. (A-C) Phenotype and function of antigen-experienced memory (CD44⁺CD49d⁺) and virtual memory (T_VM, CD44⁺CD49d^–) CD8 T cells in spleens from mice treated with DT (Treg-depleted) or vehicle control (non-Treg-depleted) during mobilization (n = 5 per group from 2 independent experiments). (D-G) MM-bearing recipient mice were lethally irradiated and transplanted with mobilized splenocytes including 5 × 10⁶ T cells from Treg-depleted or non-Treg-depleted donors (D-E; n = 10 per group) or 10 × 10⁶ BM cells and 2 × 10⁶ CD4 T cells from nonmobilized B6 donors and 2 × 10⁶ CD8 T cells from mobilized Treg-depleted or non-Treg-depleted MM-bearing donors (F,G; n = 15 per group from 2 independent experiments). Recipient mice were monitored for survival and tumor burden using M-band (G/A ratio) level. Dotted line indicates a statistically determined threshold for myeloma relapse. (H) M-band and survival after tumor rechallenge in long-term survivors (>100 days after ASCT) who were transplanted with CD8 T cells from Treg-depleted donors. Naïve B6 mice were injected with the same tumor as a nonimmune control. (I) Mice from (H) were then rechallenged with a different clone of Vk∗MYC (Vk12598) and their survival was monitored. Data represents mean ± SEM. Mann-Whitney test for 2 sample comparison and log-rank test for survival data. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001.

Figure 6.

An IL-2/IL-15 mimetic generated protective myeloma-specific CD8 T cells during SCM. MM-bearing B6 or FoxP3-GFP-DTR mice were treated with DT, NL-201 or vehicle control during mobilization with G-CSF and AMD3100. CD8 T cells were harvested from mobilized spleen and transplanted with BM and CD4 T cells from naïve B6 into irradiated MM-bearing recipient mice. (A) Total memory CD8 and CD8 T-cell/Treg ratio in the spleen of mice treated with NL-201 and vehicle during mobilization (n = 6-9 per group from 3 independent experiments). (B-C) Phenotype and cytokine production of antigen-experienced memory (CD44⁺CD49d⁺) and virtual memory (T_VM, CD44⁺CD49d^–) CD8 T cells in mobilized grafts (n = 4 /group from 2 independent experiments). (D) M-band and survival after ASCT (n = 9-13 per group from 2 independent experiments). (E) Mice with long-term tumor control after ASCT were rechallenged with the same tumor clone. Naïve B6 mice were injected with tumor cells as controls. Mice were monitored for tumor burden using M-band and survival (n = 4-6 per group from 1 experiment). (F-G) MM-bearing HULK donor mice (IFNγ-YFP × IL-10-GFP × FoxP3-RFP reporter) were treated with NL-201 or the control vehicle during mobilization. Transplantation was performed as above. (F) Representative FACS plots and the frequency of IFNγ-YFP⁺ CD8 T cells in the mobilized graft (n = 3 per group from 2 independent experiments). (G) Representative FACS plots and the total number of each subset of CD8 T cells from the mobilized graft 2 weeks after transplantation (n = 9-10 per group from 2 independent experiments). Mann-Whitney test for 2 sample comparison, One-way ANOVA for multiple samples comparison, and log-rank test for survival data. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001.