Monoclonal antibody blockade of IL-2 receptor α during lymphopenia selectively depletes regulatory T cells in mice and humans

Abstract

Lymphodepletion augments adoptive cell transfer during antitumor immunotherapy, producing dramatic clinical responses in patients with malignant melanoma. We report that the lymphopenia induced by the chemotherapeutic agent temozolomide (TMZ) enhances vaccine-driven immune responses and significantly reduces malignant growth in an established model of murine tumorigenesis. Unexpectedly, despite the improved antitumor efficacy engendered by TMZ-induced lymphopenia, there was a treatment related increase in the frequency of immunosuppressive regulatory T cells (T(Regs); P = .0006). Monoclonal antibody (mAb)-mediated inhibition of the high-affinity IL-2 receptor α (IL-2Rα/CD25) during immunotherapy in normal mice depleted T(Regs) (73% reduction; P = .0154) but also abolished vaccine-induced immune responses. However, during lymphodepletion, IL-2Rα blockade decreased T(Regs) (93% reduction; P = .0001) without impairing effector T-cell responses, to augment therapeutic antitumor efficacy (66% reduction in tumor growth; P = .0024). Of clinical relevance, we also demonstrate that anti-IL-2Rα mAb administration during recovery from lymphodepletive TMZ in patients with glioblastoma reduced T(Reg) frequency (48% reduction; P = .0061) while permitting vaccine-stimulated antitumor effector cell expansion. To our knowledge, this is the first report of systemic antibody-mediated T(Reg) depletion during lymphopenia and the consequent synergistic enhancement of vaccine-driven cellular responses, as well as the first demonstration that anti-IL-2Rα mAbs function differentially in nonlymphopenic versus lymphopenic contexts.

Figure 1

Frequency of functional T_Regs increases after lymphodepletive TMZ. (A) C57BL/6 mice were treated with 5 Gy of TBI (n = 3), single-dose TMZ (200 mg/kg; n = 5), or multiday TMZ (60 mg/kg/5 day; n = 5), and then CBCs were compared with untreated controls (n = 5). Total lymphocyte counts in peripheral blood were monitored over the course of 28 days; a representative experiment shown (n = 3). Postlymphodepletion values were evaluated for statistical significance (day 2): untreated vs 200 mg/kg TMZ, 60 mg/kg/5 day TMZ, and TBI, P < .0001 for all comparisons; and TBI vs 200 mg/kg TMZ and 60 mg/kg/5 day TMZ, P = .2042 and P = .0168, respectively. (B-C) The frequency and absolute number of CD4⁺ and CD8⁺ T cells and CD4⁺CD25⁺Foxp3⁺ T_Regs in the peripheral blood of untreated (n = 5), lymphodepletive TMZ-treated (n = 5), and TMZ-lymphodepleted mice who also received OT-I transfer (n = 5) were monitored by FACS analysis 1 week after TMZ administration (frequency CD4⁺, *P = .0396 and **P = .0216; frequency T_Regs, *P = .0006 and **P = .0019). For FACS analysis of all murine T-cell populations (supplemental Figure 11, available on the Blood Web site; see the Supplemental Materials link at the top of the online article), total lymphocytes were first gated by side scatter and forward scatter. For CD4⁺ and CD8⁺ T-cell selection, total CD3⁺ T cells were selected from the lymphocyte gate by displaying side scatter by CD3⁺. CD4⁺ or CD8⁺ T cells were selected from the CD3⁺ T-cell population by displaying CD4⁺ by CD8⁺ and gating the desired population. For T_Regs, lymphocytes were first gated by forward and side scatter, and from this gate CD4⁺ T cells were selected from side scatter. CD4⁺ T cells were displayed by CD25 versus Foxp3. For CD4⁺CD25⁺Foxp3⁺ T_Regs, double-positive cells were selected and for CD4⁺Foxp3⁺ T_Regs all Foxp3⁺ T cells were selected. Representative experiments are shown; all experiments performed in at least triplicate. (D) C57BL/6 mice were treated with 200 mg/kg single dose TMZ or 60 mg/kg/5day. Three days after the completion of TMZ treatment, CD4⁺CD25⁻ responder T cells and CD4⁺CD25⁺ T_Regs were isolated and cultured for 3 days in the presence of α-CD3e beads (all wells in triplicate). Proliferation was assessed after a 16-hour incubation with [³H]thymidine. (E) To determine the ratio of effector cells to T_Regs, the absolute number of peripheral blood CD8⁺ cells was divided by the absolute number of peripheral blood T_Regs (n = 5/cohort), ratios from representative experiments (B-C). *P = .0354.

Figure 2

TMZ-mediated lymphodepletion induces homeostatic proliferation and enhances vaccine-induced immune responses. (A-B) Mice were given increasing doses of TMZ before vaccination as described in “Preclinical temozolomide treatment, OT-I transfer, vaccination and PC61 administration.” The frequency and absolute number of OVA-specific T cells were monitored in peripheral blood using an OVA-specific tetramer and anti-CD8 mAb. For all murine tetramer FACS analyses, cells were gated with the following strategy: total lymphocytes were gated from forward and side scatter, CD8⁺ T cells were selected from the lymphocyte gate, and the population of OVA⁺ T cells was selected from total CD8⁺ lymphocytes (supplemental Figure 11). No vaccine, n = 3; untreated, n = 5; 50 mg/kg TMZ, n = 5; 75 mg/kg TMZ, n = 5; and 200 mg/kg TMZ, n = 5. Increasing TMZ dose was associated with both increasing frequencies of OVA-specific T cells (Pearson correlation = 0.69; P = .0008) and increasing absolute numbers of OVA-specific T cells (Pearson correlation = 0.54; P = .014). This experiment was repeated twice with similar results. (C) Untreated (n = 3), TMZ-lymphodepleted (n = 3), or 5-Gy TBI-treated (n = 3) C57BL/6 mice received CFSE-labeled OT-I T cells and OVA vaccination. Proliferation of CD4⁺, CD8⁺, and CD8⁺OVA⁺ T cells was monitored by CFSE dilution. Representative FACS plots are shown.

Figure 3

Inhibition of high-affinity IL-2Rα during lymphopenia depletes T_Regs. (A-C) Untreated or TMZ-lymphodepleted C57BL/6 mice received OVA vaccination with or without concomitant αIL-2Rα mAb treatment (n = 5/group). One week after mAb administration and vaccination, mice were bled and CD4⁺CD25⁺Foxp3⁺ T_Reg levels in the peripheral blood were assessed by FACS analysis. Experiments performed in at least triplicate with similar results. (A-B) For frequency T_Regs: *P = .0154, **P = .0027, and ***P = .0001. By 2-way ANOVA with interaction, the magnitude of effect of anti–IL-2Rα mAb was greater among TMZ-treated mice as opposed to mice not treated with TMZ (P = .0024). For absolute number of T_Regs, *P = .0025. (C) In the untreated and TMZ-treated cohorts, the percentage depletion of CD4⁺CD25⁺Foxp3⁺ T_Regs after αIL-2Rα mAb administration was determined from the absolute number of T_Regs after αIL-2Rα mAb treatment in comparison with the absolute number of T_Regs in the cohort that did not receive IL-2Rα blockade.

Figure 4

Anti–IL-2Rα mAb blockade synergizes with TMZ-induced lymphodepletion to enhance antigen-specific immunity. (A) Forty-eight hours after TMZ treatment, lymphodepleted or untreated C57BL/6 mice received OVA vaccination with or without simultaneous anti–IL-2Rα mAb treatment (n = 5/group). Immune responses were monitored for 1 month by FACS analysis of peripheral blood OT-I T-cell levels. This experiment was repeated twice with similar results, and unpaired t tests were generated against the untreated + vaccine cohort; TMZ + vaccine and TMZ + vaccine + αIL-2Rα mAb cohorts were significantly different from untreated + vaccine for weeks 1 to 4. Baseline values were determined in untreated and TMZ-treated mice that did not receive vaccination. At week 4, the percentage of CD8⁺ T cells observed with TMZ significantly increased with the addition of anti–IL-2-Rα mAb (P = .036). (B) Representative FACS plots shown at 1 and 4 weeks after vaccination. (C) To determine the ratio of effector cells to T_Regs, the absolute number of CD8⁺OVA⁺ effectors in peripheral blood 1 week after vaccination was assessed by FACS analysis with OVA⁺ tetramers and divided by the absolute number of CD4⁺CD25⁺Foxp3⁺ T_Regs. (D) Peripheral blood was collected from untreated and TMZ-lymphodepleted (60 mg/kg/5 days) C57BL/6 mice treated with or without αIL-2Rα mAb 72 hours after the termination of TMZ administration. Cytokine levels in plasma were measured by Luminex according to manufacturer's instructions. For cytokine measurements, analysis of variance with interaction was conducted with TMZ and αIL-2Rα mAb as the main effects and their statistical interaction. TMZ + αIL-2Rα had a significant effect on IL-2 production (P = .0005), and TMZ alone had a significant effect on IL-7 production (P = .0001). Asterisks denote significance (P < .05) versus untreated. (E) Memory recall responses were evaluated 5 weeks after vaccination in treated mice by measurement of IFNγ secretion in peripheral blood lymphocytes using CBA. *P = .0082 and **P = .0198 by unpaired t test.

Figure 5

Administration of anti–IL-2Rα mAb during lymphopenia augments vaccine-induced antitumor efficacy. B16/F10.9-OVA tumors were implanted subcutaneously in C57BL/6 mice, and TMZ was administered 3 days later. One week after implantation, cohorts of mice received vaccination with or without αIL-2Rα mAb administration. Mice were additionally vaccinated on days 12 and 17 after implantation (n = 8 per group). Statistical significance by unpaired t test on day 22 (maximal tumor burden in untreated mice): untreated + vaccine versus TMZ + vaccine, P = .0729; untreated + vaccine + αIL-2Rα mAb versus TMZ + vaccine + αIL-2Rα mAb, P = .0024; and TMZ + vaccine versus TMZ + vaccine + αIL-2Rα mAb, P = .0268).

Figure 6

Patients with GBM possess an increased frequency of T_Regs after TMZ treatment that can be reduced by a single administration of an anti–IL-2Rα mAb. (A-B) Percentages and absolute numbers of T_Regs (CD4⁺CD25⁺Foxp3⁺) from leukapheresis (pre-TMZ and post-TMZ and daclizumab) and peripheral blood (post-TMZ) samples were determined by CBC counts and FACS analysis. T_Reg levels from preoperative samples (pre-TMZ, ∼ day − 100), before initial vaccination and daclizumab (post-TMZ, day 0) and ∼ 7 weeks after vaccination and daclizumab (post-TMZ and daclizumab, day ∼ 50) were assessed. (A) Frequency of T_Regs by paired t test, *P = .0236 and **P = .0061. (B) Total CD4⁺, CD4⁺Foxp3⁺, and CD4⁺CD25⁺Foxp3⁺ T-cell frequencies from leukapheresis and peripheral blood samples from a representative patient were determined before and after daclizumab administration (day 0). Gating strategies for flow cytometric analyses were as follows and as shown in supplemental Figure 12: (1) For CD4⁺ T cells, all cells were displayed using forward and side scatter, and the lymphocyte population was selected. Lymphocytes were then displayed by side scatter and CD4 on a dot plot. The CD4⁺ population was then selected out of this lymphocyte population. (2) For Foxp3⁺ of CD4⁺, an identical gating strategy as described in 1 was used to select CD4⁺ T cells. The CD4⁺ T cells were then displayed by dot plot against Foxp3, and all Foxp3⁺ cells were selected. (3) For CD25⁺Foxp3⁺ of CD4⁺, an identical gating strategy as described in (1) was used to select the CD4⁺ population. The selected CD4⁺ population was then displayed by dot plot as CD25 versus Foxp3. CD25 and Foxp3 double-positive cells were then selected out of the CD4⁺ population.

Figure 7

Anti–IL-2Rα mAb blockade does not prevent the vaccine-induced expansion of antigen-specific CD8⁺ T cells. CMV-pp65–specific CD8⁺ T cells were determined by flow cytometric analysis of PBMCs both before (pre) and 43 to 55 days after (post) administration of CMV-pp65 DC vaccine and daclizumab. (A) Data are shown as the average fold increase (post/pre) in the frequency of pp65⁺CD8⁺ T cells from 4 REGULATE patients (individual patient average of tetramer⁺CD8⁺ T cells across tested HLA alleles). (B) Representative FACS plot of pp65-tetramer⁺CD8⁺ T cells. Gating strategy: total lymphocytes selected from forward scatter and side scatter display, CD3⁺CD8⁺ T cells selected from total lymphocytes, and pp65⁺ T cells selected from total CD3⁺CD8⁺ population (supplemental Figure 12).