A CD200R-CD28 fusion protein appropriates an inhibitory signal to enhance T-cell function and therapy of murine leukemia

Abstract

Acute myeloid leukemia (AML), the most common adult acute leukemia in the United States, has the poorest survival rate, with 26% of patients surviving 5 years. Adoptive immunotherapy with T cells genetically modified to recognize tumors is a promising and evolving treatment option. However, antitumor activity, particularly in the context of progressive leukemia, can be dampened both by limited costimulation and triggering of immunoregulatory checkpoints that attenuate T-cell responses. Expression of CD200 (OX2), a negative regulator of T-cell function that binds CD200 receptor (CD200R), is commonly increased in leukemia and other malignancies and is associated with poor prognosis in leukemia patients. To appropriate and redirect the inhibitory effects of CD200R signaling on transferred CD8+ T cells, we engineered CD200R immunomodulatory fusion proteins (IFPs) with the cytoplasmic tail replaced by the signaling domain of the costimulatory receptor, CD28. An analysis of a panel of CD200R-CD28 IFP constructs revealed that the most effective costimulation was achieved in IFPs containing a dimerizing motif and a predicted tumor-T-cell distance that facilitates localization to the immunological synapse. T cells transduced with the optimized CD200R-CD28 IFPs exhibited enhanced proliferation and effector function in response to CD200+ leukemic cells in vitro. In adoptive therapy of disseminated leukemia, CD200R-CD28-transduced leukemia-specific CD8 T cells eradicated otherwise lethal disease more efficiently than wild-type cells and bypassed the requirement for interleukin-2 administration to sustain in vivo activity. The transduction of human primary T cells with the equivalent human IFPs increased proliferation and cytokine production in response to CD200+ leukemia cells, supporting clinical translation. This trial was registered at www.clinicaltrials.gov as #NCT01640301.

Conflict of interest statement

Conflict-of-interest disclosure: P.D.G. has an equity interest in Juno Therapeutics, Inc., and receives a consulting payment from Juno Therapeutics, Inc. P.D.G, S.K.O., and T.M.S are named as inventors on 1 or more patent and/or patent application related to this work. The remaining authors declare no competing financial interests.

© 2017 by The American Society of Hematology.

Figures

Graphical abstract

Figure 1.

CD200R-CD28 constructs are expressed at high levels on primary murine CD8+T cells. (A) Schematic representation of CD200R-CD28 constructs. CD200Rtm-CD28 contains CD200R ec and tm domains and a CD28 ic signaling domain. CD200R-CD28tm contains the ec domain of CD200R and the tm and ic domains of CD28. The remaining 3 constructs also incorporate a portion of the ec domain of CD28 to the tm-proximal cysteine to promote multimerization and enhance CD28 signaling. To account for the extra 9 ec aas, CD200R-3aas-CD28cys has a truncated portion of CD200R that removes 3 membrane-proximal aas, but preserves an N-linked glycosylation site. CD200R-9aas-CD28cys has an extracellular portion of CD200R that is truncated 9 membrane-proximal aas. The first, second, and last constructs are predicted to approximate the spatial distance between the T cell and an APC, as indicated by the dashed line. (B) Transgenic expression of CD200R-CD28 constructs on TCRgag T cells as detected by anti-CD200R antibody. (C) Expression of CD200R in transduced TCRgag T cells as in panel B, expressed as mean fluorescence intensity (MFI). Cumulative results of 6 independent experiments (P = not significant).

Figure 2.

CD200R-CD28 constructs promote proliferation, accumulation, and effector function of transduced T cells stimulated by CD200+tumor target cells in vitro. Splenocytes from naive TCRgag mice were stimulated in vitro with anti-CD3 (1 μg/mL), anti-CD28 (1 μg/mL), and recombinant human IL-2 (rhIL-2, 100 U/mL) and transduced with retroviral supernatant for 2 days. Cells were restimulated every 7 days with irradiated FBL and splenocytes and cultured with rhIL-2 (50 U/mL) for ≤3 stimulations. T cells were used for assays 5 to 7 days after the last stimulation. (A) Proliferation of CD200R-CD28 (red lines) and GFP empty vector control (blue lines) TCRgag T cells as measured by CTV dilution after stimulation with CD200– FBL (upper panels) or CD200+ FBL (lower panels) at a low E:T ratio of 25:1 for 3 days. (B) Cumulative proliferation of cells depicted in panel A. Proliferation was normalized across experiments by assessing the proportion of divided CD200R-CD28+ T cells relative to empty vector–transduced T cells (% dividedCD200R+/% dividedCD200R– [n = 3]). (C) Enrichment of transduced TCRgag T cells in a mixed population including nontransduced TCRgag T cells after 3 weekly cycles of stimulation with irradiated CD200+ FBL and splenocytes (n = 4-5/group). **P < .01 (Student t test). (D) Carboxyfluorescein diacetate succinimidyl ester cytotoxicity assay. TCRgag T cells were transduced with CD200R-9aas-CD28cys (red bars) or mock-transduced cells (black bars). Effector TCRgag T cells were incubated at the indicated effector to target ratio with a 1:1 mix of CD200+ FBL and nonspecific EL4 control targets for 4 hours. The relative frequency of FBL vs control tumor cells was determined by flow cytometry, and the percentage of specific lysis was determined by the frequency of FBL cells after T-cell culture relative to FBL incubated without T cells; cumulative of 3 independent experiments. *P < .05 (Student t test). (E) Cytokine production of TCRgag T cells transduced with GFP control (black lines) or CD200R IFP (red lines) relative to unstimulated T cells (gray filled) after coculture at a 1:1 ratio with CD200– (upper histograms) or CD200+ (lower histograms) FBL for 4 hours in the presence of GolgiPlug (BD Biosciences). Cells were fixed, permeabilized, stained for ic cytokines, and assessed by flow cytometry. (F) Summary of panel E. Stacked bar charts of cytokine production in TCRgag T cells in response to CD200– (left) or CD200+ (right) FBL stimulation at a 1:1 ratio. Data are presented as no cytokine (white), 1 cytokine (light gray), or 2+ cytokine (dark gray) production; cumulative results of 3 independent experiments. (G) Visualization of CD200R localization within T cell-FBL conjugates. TCRgag in vitro expanded effector T cells transduced with CD200R-9aas-CD28cys (upper panels) or CD200R-CD28cys (lower panels) were cocultured with CD200+ FBL at an E:T ratio of 10:1 at 37°C for 20 minutes to allow conjugate formation. Conjugates were loaded on a μ-Slide VI 0.4 chamber (Ibidi) for an additional 15 minutes. Cells were fixed and stained for CD200R (first panels), CD200 (second panels), and lipid rafts by CTxB (third panels; overlay in fourth panels). Conjugates were imaged by microscopy on the DeltaVision Elite (60×) and analyzed using ImageJ software. (H) Quantification of conjugates that exhibit CD200R staining within the synapse, as in panel G, expressed as the percentage of total T cell-FBL conjugates. Data are represented as mean ± standard deviation from 2 separate experiments for a total of 50 conjugates assessed. (I) pLCK Y394 expression of TCRgag T cells. T cells were transduced with CD200R-9aas-CD28cys (red line), CD200R-CD28cys (blue line), or GFP control (black line) and were left unstimulated (upper left histogram and gray filled in other histograms) or stimulated for 10 minutes with PMA/ionomycin, CD200– FBL, or CD200+ FBL, as indicated. FBL stimulation was performed at an E:T ratio of 10:1; representative of 2 independent experiments.

Figure 3.

T cells transduced with CD200R-9aas-CD28cysenhance adoptive immunotherapy of disseminated leukemia. (A-B) CD200R-CD28 IFP-transduced TCRgag T cells were generated as described in Figure 2. B6 mice were injected with 4 × 106 CD200+ FBL cells. Five days later, CD200R-9aas-CD28cys or empty vector control TCRgag T cells were injected into Cy-treated FBL-bearing B6 mice. The expression of surface markers on splenic CD200R-9aas-CD28cys TCRgag T cells (blue lines), control TCRgag T cells (red lines), and endogenous T cells (shaded) was assessed by flow cytometry at days 8 (A) and 15 (B) post–T-cell transfer; representative of 2 independent experiments. (C-D) Survival of mice treated with T-cell immunotherapy in the presence (C) or absence (D) of IL-2 injections. B6 mice were injected with 4 × 106 CD200+ FBL cells. Five days later, CD200R-9aas-CD28cys or empty vector control TCRgag T cells were injected i.p. into Cy-treated FBL-bearing mice at 105 cells/mouse (indicated by arrow). IL-2 was administered every 2 days for a total of 10 days (2 × 104 U/dose) in a cohort of mice. (C) Data from 1 experiment are shown (n = 3-4 mice/group). (D) Pooled data from 3 independent experiments are shown (No therapy [tx], n = 6; Cytoxan [Cy] only, or Cy plus T-cell treated: n = 9-10 mice/group). Statistical analyses are shown: Cy + CD200R-9aas-CD28cys T cells (red) vs Cy + empty vector T cells (blue), *P < .05; Cy + CD200R-9aas-CD28cys T cells (red) vs Cy only (black dashed line), ***P < .001); Cy + empty vector T cells (blue) vs Cy only (blacked dashed line), *P < .05).; and Cy only (blacked dashed line) vs No treatment (tx; black), ****P < .001; log-rank Mantel-Cox test.

Figure 4.

Coexpression of CD200R-CD28 enhances function in WT1-specific TCR-transduced human primary T cells. (A) Expression of CD200 on CD45dimCD34+ cells (black lines) from a healthy donor leukapheresis sample (left panel) or leukemic blasts from 2 separate donors (center and right panels) in relation to matched FMO control (gray shaded). (B) Schematic representation of CD200R-CD28 constructs. CD200Rtm-CD28 contains CD200R ec and tm domains and a CD28 ic signaling domain. CD200R-CD28tm contains the ec domain of CD200R and the tm and ic domains of CD28. The remaining 3 constructs also incorporate 12 aas of the ec domain of CD28 to the tm-proximal cysteine to promote multimerization and enhance CD28 signaling. To account for the extra CD28 residues, CD200R-9aas-CD28cys has a truncated portion of CD200R that removes the 9-aa membrane-proximal stem region, and CD200R-12aas-CD28cys has a truncated portion of CD200R that removes 12 membrane-proximal residues that include additional amino acids beyond the stem region. The first, second, and fifth constructs are predicted to approximate the spatial distance between the T cell and an APC, as indicated by the dashed line. (C-H) CD8+ T cells were enriched by magnetic beads from PBMCs harvested from healthy HLA-A2+ donors. CD8+ T cells were stimulated with anti-CD3/CD28 Dynabeads and transduced with lentiviral supernatant for 2 days. Transduced T cells were isolated by fluorescence-activated cell sorting and restimulated by rapid expansion protocol every 10 to 14 days in the presence of IL-2. (C) Diagram of construct combining IFP, TCRα, and TCRβ chains. The IFP constructs were inserted into single lentiviral vectors with the β and α chains of the HLA-A2–restricted WT1126-specific TCRC4. The first P2A sequence was codon optimized (CO P2A) to prevent genetic recombination with the second P2A sequence (P2A). Flow cytometry plots show expression of TCRC4 only (left plot) or TCRC4 + CD200R-CD28 (right plot) in primary human T cells as detected by anti-CD200R antibody and WT1126 HLA-A2 tetramer binding. (D) Proliferation of T cells as detected by dilution of CTV. Primary human T cells transduced with TCRC4 or TCRC4 and CD200R IFP were stained with CTV and stimulated with WT1126-pulsed T2 cells at an E:T ratio of 25:1 in the absence of IL-2 for 6 days. The percentage of cells that diluted CTV (proliferated) was determined by FlowJo proliferation analysis. Cumulative of 3 independent T-cell donors (**P < .01). (E) Representative histogram of CTV dilution in unstimulated (gray filled), TCRC4-transduced (black line), or TCRC4- and CD200R-9aas-CD28cys-transduced (orange line) T cells. (F) Intracellular cytokine production of CD8+ T cells. Primary human T cells transduced with TCRC4 only (upper panels) or TCRC4 and CD200R-9aas-CD28cys-transduced (lower panels) were unstimulated or stimulated with WT1126-pulsed T2 cells for 6 hours in the presence of GolgiPlug (BD Biosciences). T cells were fixed, permeabilized, and stained for ic cytokines and assessed by flow cytometry. (G) Summary of cytokine production in panel F (E:T ratio, 1:1). Data are presented as no cytokine (white), 1 cytokine (light gray), or 2+ cytokine (dark gray) production; cumulative results of 3 independent T-cell donors. (H) Cytotoxicity assay. Primary AML blasts were cocultured with primary human T cells transduced with TCRC4 alone (black symbols) or TCRC4 and the CD200R-9aas-CD28cys IFP (orange symbols) for 24 hours. Remaining viable blasts were quantified by flow cytometry and the percentage of lysis was determined after normalization with the tumor-only control well (assayed in triplicate; *P < .05, **P < .01).