FGL2-targeting T cells exhibit antitumor effects on glioblastoma and recruit tumor-specific brain-resident memory T cells

Abstract

Although tissue-resident memory T (T_RM) cells specific for previously encountered pathogens have been characterized, the induction and recruitment of brain T_RM cells following immune therapy has not been observed in the context of glioblastoma. Here, we show that T cells expressing fibrinogen-like 2 (FGL2)-specific single-chain variable fragments (T-αFGL2) can induce tumor-specific CD8⁺ T_RM cells that prevent glioblastoma recurrence. These CD8⁺ T_RM cells display a highly expanded T cell receptor repertoire distinct from that found in peripheral tissue. When adoptively transferred to the brains of either immunocompetent or T cell-deficient naïve mice, these CD8⁺ T_RM cells reject glioma cells. Mechanistically, T-αFGL2 cell treatment increased the number of CD69⁺CD8⁺ brain-resident memory T cells in tumor-bearing mice via a CXCL9/10 and CXCR3 chemokine axis. These findings suggest that tumor-specific brain-resident CD8⁺ T_RM cells may have promising implications for the prevention of brain tumor recurrence.

Fig. 1. Restricted functional antitumor activity of T-αFGL2 cells in vitro.

a Schematic of the vector encoding the FGL2-blocking scFv. b Schematic of a T cell transfected with the FGL2-blocking scFv vector (T-αFGL2). c Representative flow cytometry histograms demonstrating expression of the FGL2-blocking scFv on mouse T cells following transduction. d Flow cytometry plots depicting no difference in proportion of tumor cells (DBT-GFP+) cocultured with T-Ctr or T-αFGL2 cells at E:T ratio of 4:1 for 72 h (top panel); flow cytometry plots depicting higher granzyme B expression and no difference in TNFα or IFNγ expression in T-αFGL2 cells (compared with T-Ctr) cocultured with DBT tumor cells at E:T ratio of 1:1 for 24 h. Data are mean ± SD from three independent experiments. NS not significant, two-tailed t-test. e Representative micrographs of FGL2 expression in GBM and the indicated normal human tissue samples. The images of normal human tissue shown are representative results from two samples. Micrographs are representative of two sections per tissue sample. Scale bars, 100 μm.

Fig. 2. Antitumor activity of T-αFGL2 in vivo.

a Representative micrographs of FGL2 expression in the brains of mice with glioma. Slides stained with mouse IgG were used as negative controls. Images are representative of results from three samples with multiple fields of view for each sample. b Schematic of the orthotopic DBT glioma model. On days 3–5 after DBT tumor cell inoculation, mice were treated with temozolomide (TMZ), followed by infusion of T-Ctr or T-αFGL2 cells on days 6 and 13. c Representative bioluminescence images of DBT-luc tumor growth in the orthotopic glioma model shown in (b). d Flux vs. time [p/s] data (mean ± SEM) for the orthotopic glioma model shown in (b) (n = 5 mice). e Kaplan–Meier survival curves of mice shown in (b) (n = 10 mice for NT group, n = 13 for T-Ctr group, n = 20 for T-αFGL2 group), log-rank test. The experiments were repeated three times with similar results. f Representative H&E staining of brains in (b) collected on day 14 after tumor cell inoculation. Images are representative of results from three samples with multiple fields of view for each sample. Data are representative of three mice. g Schematic of the orthotopic GL261 glioma model. h Kaplan–Meier survival curves of mice in (g) (n = 6 for T-Ctr group, n = 7 for T-αFGL2 group), log-rank test. The experiments were repeated twice with similar results.

Fig. 3. T-αFGL2 treatment induced brain-resident tumor-specific memory T cells.

a Schematic of rechallenge with tumor cells. On day 70 after first tumor cell inoculation, T-αFGL2–treated survivors bearing orthotopic DBT gliomas were rechallenged with DBT cells injected either subcutaneously (s.c.) or intracranially (i.c.). b Kaplan–Meier survival curves (left) and representative bioluminescence images (right) of mice in (a) on day 0 and day 7 after second tumor cell inoculation (i.c.) (n = 6 mice). ***P = 0.0005, log-rank test. c Tumor volume (left) and representative bioluminescence images (middle) of mice on day 0 and day 7 after second tumor cell inoculation (s.c.), and representative tumors (right) collected on day 11 after second tumor cell inoculation (s.c.) from the flanks of Balb/c mice (n = 3 mice/group; data are mean ± SD). d Representative flow cytometry plots depicting increase of CD8⁺ T cells in the brains (BIL) of long-term survivors treated with T-αFGL2 (T-αFGL2 survivor). LN lymph node. e Ratio of CD8⁺ T cells to CD4⁺ T cells in the brains of naïve mice and T-αFGL2 survivors and the LNs of T-αFGL2 survivors. Data are mean ± SD, one-way ANOVA with Tukey’s test for comparing multiple treatments. f CD8⁺ T cell numbers in the brains of naïve mice and T-αFGL2 survivors (n = 3 naïve mice, n = 5 T-αFGL2 survivors; data are mean ± SD), two-tailed t-test. The experiments were repeated three times with similar results.

Fig. 4. T_RM-like cells can be adoptively transferred.

a Schematic of experimental design. On days 70 and 100 after the first tumor cell inoculation, T-αFGL2-treated survivors were rechallenged with DBT tumor cells intracranially (i.c.). On day 7 after the third challenge (day 100) with tumor cells, the mice were euthanized, and their brains, draining lymph nodes (dLN), and peripheral blood (PB) were collected. b Representative bioluminescence images of naïve Balb/c mice coinoculated i.c. with 3 × 10³ DBT glioma cells and 3 × 10³ T cells. Images show gliomas in mice coinoculated with CD8⁺ T cells in the brain (BIL-CD8⁺T), CD4⁺ T cells in the brain (BIL-CD4⁺T), CD8⁺ T cells in peripheral blood (PB-CD8⁺T), or CD8⁺ T cells in draining lymph nodes (dLN-CD8⁺T). CD8⁺ T cells and CD4⁺ T cells were sorted by flow cytometry on day 7 after the third challenge in T-αFGL2–treated survivors. c Kaplan–Meier survival curves for mice in (b) (n = 9 in BIL-CD8⁺T group, n = 6 in BIL-CD4⁺T and dLN-CD8⁺T groups, n = 8 in PB-CD8⁺T group), log-rank test. The experiments were repeated 3 times with similar results. d, Representative bioluminescence images of naïve Balb/c mice and mice bearing transplanted BIL-CD8⁺T cells on days 0 and 4 after i.c. rechallenge with DBT cells on day 30 after BIL-CD8⁺T cell transplantation. e Kaplan–Meier survival curves of mice in (d) (n = 9 mice/group), log-rank test. Data shown are representative of three independent experiments.

Fig. 5. T_RM-like cells were CD8⁺T cells and exhibited T_RM phenotypes.

a Schematic of experimental design. On day 7 after the third tumor cell rechallenge, 3 × 10⁴ T_RM-containing brain-infiltrating lymphocytes (T_RM-BIL) from T-αFGL2–treated survivors were sorted by flow cytometry and co-inoculated i.c. along with 3 × 10³ DBT cells into naïve SCID mice; 35 days after transplantation, the SCID mice bearing transplanted T_RM-BILs were rechallenged with 3 × 10³ DBT cells i.c., combined with antibodies blocking CD8, CD4, or asGM1 i.p. b Kaplan-Meier survival curves of mice in (a) (n = 6 mice/group), log-rank test. c Kaplan–Meier survival curves of mice treated with anti-CD8, anti-CD4, or anti-asGM1 antibodies in (a) (n = 3 mice/group), log-rank test. The experiments were repeated twice with similar results. d Representative H&E staining of brains from c collected on day 14 after tumor cell rechallenge. Similar observations were made in three mice per group and representative images are shown. e Representative flow cytometry plots and graph showing ratio of CD69⁺CD103⁺ T cells, CD69⁺CD103⁻ T cells, and CD69⁺CD62L⁻ T cells in brain and PB of T-αFgl2-treated survivors (n = 4 for detection of CD69⁺CD103⁺ T cells and CD69⁺CD103⁻ T cells, n = 3 for detection of CD69⁺CD62L⁻ T cells; data represent mean ± SD), two-tailed t-test. The experiments were repeated twice with similar results.

Fig. 6. T_RM cells showed the presence and expansion of unique T cell clones.

a Schematic of TCRα/β deep sequencing of CD8⁺ T cells from the brains and draining lymph nodes (dLNs) of T-αFGL2-treated survivors. Cells were sorted via flow cytometry on day 20 after the third challenge with intracranially (i.c.) injected DBT tumor cells. b Representative tree maps (top row) of TCRα-T_RM -CD8⁺T, TCRβ -T_RM -CD8⁺T, and TCRβ-dLNs-CD8⁺T cell clones. Each spot represents a unique entry: V-J-CDR3, and the size of a spot denotes its relative frequency; 2D map of V and J usage of TCRα-T_RM -CD8⁺T, TCRβ-T_RM -CD8⁺T, and TCRβ-dLNs-CD8⁺T cell clones (bottom row). The experiments were repeated twice with similar results. c, Schematic of experimental design. On day 1, 3 × 10⁴ CD8⁺T_RM cells and 3 × 10³ DBT cells were coinoculated i.c. into naïve Balb/c mice; on days 0, 5, 10, 15, 20, and 25, the mice were treated with IgG or MHC-I blocking antibodies (100 µg/mouse, i.p.). d Representative bioluminescence images of Balb/c mice on days 0, 14, and 28 after i.c. coinoculation with CD8⁺T_RM and DBT cells. e Kaplan–Meier survival curves of mice in (f) (n = 4 mice/group), log-rank test. The experiments were carried out twice with similar results. f Representative H&E staining of brains from (d) collected on day 40 after transplantation. Similar observations were made in three mice per group and representative images are shown.

Fig. 7. T-αFGL2 treatment increased the CD69⁺CD8⁺T_M cell subset.

a Schematic of the experimental design. Four days after the second infusion of T-Ctr or T-αFGL2 cells, brains were collected to isolate brain-infiltrating lymphocytes (BILs), which were then stained with antibodies conjugated to metal isotopes. Single-cell mass cytometry (CyTOF) data were clustered to identify common populations across the treatment groups (n = 4 mice per group). The experiment was carried out once. b Analysis of CD45⁺ cells from the brain, colored by relative expression of CyTOF markers. Cell populations are indicated on the right. c Frequencies of total CD8⁺ T cell population and subsets of CD8⁺ T cells and CD4⁺ T cells (n = 4 mice per group; data represent mean ± SD), two-tailed t-test. d Composition of the CD8⁺ T cell compartment in T-Ctr and T-αFGL2-treated DBT-bearing mice showing increased frequency of CD69⁺CD8⁺ T_M cells in the T-αFGL2 group. e, Fold expression of Ki67, CD69, CD223, and CD279 on the CD69⁺CD8⁺T_M subset and the CD69⁻CD8⁺T_EM subset. f Schematic of experimental design. On day 1, 3 × 10⁴ CD8⁺ T_RM cells and 3 × 10³ DBT cells were coinoculated i.c. into naïve Balb/c mice; on days 0, 5, 10, 15, and 20, the mice were treated with either IgG or CD69 blocking antibodies (150 μg/mouse, i.p.); on day 60, Balb/c mice bearing transplanted CD8⁺T_RM were rechallenged with 1 × 10⁴ DBT cells (i.c.). g Representative bioluminescence images of Balb/c mice on days 0 and 7 after i.c. rechallenge with DBT cells in (f). Data are representative of two independent experiments. h Kaplan–Meier survival curves of mice in (f) (n = 3 mice/group), log-rank test. The experiments were carried out twice with similar results.

Fig. 8. T-αFGL2-induced CD69⁺CD8⁺T_M cells were associated with CXCL9/10-CXCR3 axis.

a Expression levels of CCR2, CSF1R, CXCR2, CXCR3, and CX3CR1 on CD69⁺CD8⁺T_M cell populations; CyTOF analysis was conducted on day 2–4 after the second T cell treatment. (n = 5 mice per group; data represent mean ± SD), p value from two-tailed t-test, FDR false discovery rate. b Representative flow cytometry plots and graphs showing that T-αFGL2 treatment increased the proportion of CXCR3⁺CD69⁺CD8⁺T cells among total CD8⁺T cells in glioma-bearing brains (n = 5 mice/group; data represent mean ± SD), two-tailed t-test. c Kaplan–Meier survival curves of GL261-bearing WT mice and CXCR3^−/− mice treated with T-Ctr or T-αFGL2 (n = 5 mice/group), log-rank test. d Number of CD69⁺CD8⁺T_M cells per GL261-bearing brain on day 5–7 after the second T cell infusion (n = 5 per group; data represent mean ± SEM), one-way ANOVA with Dunnett’s test for comparing multiple treatments. e, f Quantitative analysis of CXCL9 and CXCL10 protein levels in DBT tumors from mice 4–6 days after the second infusion of T-Ctr or T-αFGL2 cells (n = 5 per group; data represent mean ± SEM), two-way t-test. g Percentages of CD69⁺ cells among CD44⁺CD8⁺T cells (n = 5 per group; data represent mean ± SEM), one-way ANOVA with Dunnett’s test for comparing multiple treatments. NS not significant. h Schematic illustration of cellular and molecular events underlying T-αFGL2–induced tumor-specific brain-resident CD8⁺T_RM cells. T-αFGL2 cells block FGL2 in the tumor microenvironment, resulting in CXCL9/10 induction. The CXCL9/10-CXCR3 engagement boosts recruitment of CXCR3⁺CD69⁺CD8⁺T cells, which are candidate tumor-specific brain-resident CD8⁺T_RM cells.