T-cell derived extracellular vesicles prime macrophages for improved STING based cancer immunotherapy

Abstract

A key phenomenon in cancer is the establishment of a highly immunosuppressive tumour microenvironment (TME). Despite advances in immunotherapy, where the purpose is to induce tumour recognition and hence hereof tumour eradication, the majority of patients applicable for such treatment still fail to respond. It has been suggested that high immunological activity in the tumour is essential for achieving effective response to immunotherapy, which therefore have led to exploration of strategies that triggers inflammatory pathways. Here activation of the stimulator of interferon genes (STING) signalling pathway has been considered an attractive target, as it is a potent trigger of pro-inflammatory cytokines and types I and III interferons. However, immunotherapy combined with targeted STING agonists has not yielded sustained clinical remission in humans. This suggests a need for exploring novel adjuvants to improve the innate immunological efficacy. Here, we demonstrate that extracellular vesicles (EVs), derived from activated CD4⁺ T cells (T-EVs), sensitizes macrophages to elevate STING activation, mediated by IFNγ carried on the T-EVs. Our work support that T-EVs can disrupt the immune suppressive environment in the tumour by reprogramming macrophages to a pro-inflammatory phenotype, and priming them for a robust immune response towards STING activation.

Keywords: CDN therapy; IFNγ; STING; T cells; cancer immunology; extracellular vesicles; macrophages.

FIGURE 1

Activated CD4⁺ T‐cell derived EVs sensitises macrophages to STING activation. (a and b) THP‐1 cells were treated with T‐EVs (1.5–2 × 10⁹) from activated CD4⁺ T cells or EV‐free media for 1 h prior to stimulation with either cGAMP (0.5 μg), Poly(I:C) (0.1 μg) or LPS (0.5 μg/mL). After 20 h, the supernatant was analysed for secretion of (a) functional type I IFN and (b) IL‐6. Data show mean + SEM and individual replica of T‐EVs from two distinct T cell donors, each in duplicates. Unpaired t‐test (two‐tailed). (c and d) THP‐1 cells were treated with T‐EVs (3 × 10⁹) from activated CD4⁺ T cells or EV‐free media for 1 h prior to stimulation with increasing amounts of cGAMP as indicated. After 20 h, the supernatant was analysed for secretion of (c) functional type I IFN and (d) CXCL‐10. Data show mean + SEM and individual replica, and are representative of two independent experiments. (e) THP‐1 cells and (f) human monocyte derived macrophages (MDMs) were treated with T‐EVs (3 × 10⁹) from activated CD4⁺ T cells or EV‐free media for 1 h prior to stimulation with cGAMP (0.5 μg). The production of type I IFN was determined after 20 h of stimulation. Data in (e) shows mean +SD of six independent experiments and data in (f) shows mean +SD of six different experiments all with T‐EVs from distinct T cell donors. Each datapoint indicate mean value of duplicates. Wilcoxon test (two‐tailed). (g) THP‐1 cells were treated with increasing amounts of T‐EVs or EV‐free media for 1 h prior to stimulation with cGAMP (0.5 μg) as indicated. The production of type I IFN was determined after 20 h. Data show mean +SEM and individual replica and are representative of two distinct T‐EV donors. UT, untreated, and received only EV‐free media and lipofectamine. T‐EV, EVs derived from activated CD4⁺ T cells.

FIGURE 2

T‐EVs modulate STING signalling independent of cGAS and intravesicular cGAMP. (a) Schematics of the cGAS‐STING pathway. (b) Western blot analysis of cell lysates from WT, cGAS^−/−, and STING^−/−, THP‐1 cells. Vinculin (VCL) was used as a loading control. Data represents 1 experiment. (c) WT, cGAS^−/−, and STING^−/− THP‐1 cells were stimulated with Poly(I:C) (0.1 μg). The production of type I IFN response was determined after 20 h of stimulation. Data show mean + SEM and individual replica of two independent experiments in duplicates or triplicates. ns, non‐significant; unpaired t‐test (two‐tailed). (d) WT, cGAS^−/−, and STING^−/− THP‐1 cells were stimulated with T‐EVs (3 × 10⁹) or EV‐free media for 1 h prior to stimulation with cGAMP (0.5 μg). The type I IFN production was determined after 20 h of stimulation. Data show mean + SEM and individual replica of two independent experiments in duplicates or triplicates. Unpaired t‐test (two‐tailed). (e) T‐EVs were analysed by mass spectrometry for presence of cGAMP. Data show cGAMP concentration in each of three distinct T‐EV samples as well as in samples spiked‐in with indicated amount of cGAMP. WT, Wild type. UT samples in (c) and (d) are identical.

FIGURE 3

T‐EVs prime THP‐1 cells for enhanced STING activation. (a) THP‐1 cells were stimulated with T‐EVs (3 × 10⁹) or EV‐free media. After 6 h of stimulation, the mRNA expression of STING was determined. Data shows mean + SEM and individual duplicates and is representative of two independent experiments. (b) THP‐1 cells were stimulated with T‐EVs (3 × 10⁹) for 1 h prior to stimulation with cGAMP (0.5 μg). At indicated time‐points after cGAMP stimulation the supernatant was harvested and the production of type I IFN response was determined. Data shows mean +SEM and individual replica of three independent experiments each in duplicates. Unpaired t‐test (two‐tailed). (c) THP‐1 cells were stimulated with T‐EVs (1 × 10⁹) or EV‐free media at the indicated time‐points prior to cGAMP stimulation (0.5 μg). Right before cGAMP stimulation, the T‐EVs were washed out. After 20 h of cGAMP stimulation, the supernatant was collected and the production of type I IFN was determined. Data show mean +SEM and individual replica of two independent experiments. Unpaired t‐test (two‐tailed). (d and e) THP‐1 cells were stimulated with T‐EVs (3 × 10⁹) or EV‐free media for 1 h prior to stimulation with cGAMP (0.5 μg). At indicated time‐points after cGAMP stimulation, the cells were harvested. Expression and phosphorylation of indicated proteins were analysed by Western blotting. Data are representative of three independent experiments. VCL was used as a loading control. VCL in (d) and (e) are identical.

FIGURE 4

T‐EVs from primarily activated CD4+ T cells enhances STING signalling. T‐EVs were isolated from CD4⁺ T cells that were either activated (a) with anti‐CD3 (1 μg/mL) and anti‐CD28 (1 μg/mL), or left non‐activated (NA) for 48 h in presence of only IL‐2 (10 ng/mL). The T‐EVs were analysed by tunable resistive pulse sensing using a qNano for (a) concentration and (b and c) size distribution. (a) data shows T‐EVs from each of four distinct donors. Paired t‐test (two‐tailed). Data in (b) show mean size of T‐EVs from each of the four distinct donors. ns, not significant, paired t‐test (two tailed). Data in (c) show mean +SD of T‐EVs from each of the 4 distinct donors. (d) Western blot of paired cell lysate and T‐EVs from two individual donors. E) THP‐1 cells were treated with T‐EVs (3 × 10⁹) from either A or NA CD4⁺ T cells for 1 h prior to stimulation with cGAMP (0.5 μg). The production of type I IFN was determined upon 20 h of stimulation. Data shows mean +SD and individual mean values of duplicates, from six different experiments. Paired t‐test (two‐tailed). (f) THP‐1 cells were stimulated with T‐EVs (3 × 10⁹) from either A or NA CD4⁺ T cells for 1 h prior to stimulation with cGAMP (0.5 μg). After 1 h of stimulation, the cells were harvested for Western blot analysis. Data are from one experiment using T‐EVs from two distinct donors. Vinculin (VCL) was used as a loading control. NA, non‐activated. A, activated. ns, not significant.

FIGURE 5

T‐EVs transfer proinflammatory cytokines to enhance macrophage function. (a) EVs from either activated (A) or non‐activated (NA) CD4⁺ T cells were lysed in 2.5% Triton‐X‐100 and the presence of cytokines was measured using mesoscale or ELISA for IFNγ. Data show mean +SD and individual concentrations of n = 5 (A) and n = 2 (NA). (b) THP‐1 cells were stimulated with either T‐EVs (1 × 10⁹), recombinant TNFα (10 ng/mL) or recombinant IFNγ (10 ng/mL) for 1 h prior to stimulation with cGAMP (0.5 μg). The production of type I IFN was determined after 6 h of stimulation. Data show mean +SEM and individual replica from two independent experiments. (c) THP‐1 cells were treated with anti‐TNFR1 (10 μg/mL) or IgG1 (10 μg/mL) for 30 min at 37°C. The cells were stimulated with T‐EVs (1 × 10⁹) for 1 h followed by stimulation with cGAMP (0.5 μg). The production of type I IFN was determined after 6 h of stimulation. Data show mean +SEM and individual replica of three independent experiments. (d) THP‐1 cells were treated with anti‐IFNGR1 (20 μg/mL) or IgG1 (20 μg/mL) for 30 min at 37°C. T‐EVs were treated with anti‐IFNγ (10 μg/mL) or IgG1 (10 μg/mL) and incubated for 15 min prior to use for stimulations. The cells were stimulated with the T‐EVs (1 × 10⁹) for 1 h followed by stimulation with cGAMP (0.5 μg). The production of type I IFN was determined after 6 h of stimulation. Data show mean +SEM and individual replica of two independent experiments with T‐EVs from three distinct donors in total. (e) THP‐1 cells were treated with a combination of anti‐IFNGR1 and anti‐TNFR1 as described for (c) and (d). Data show mean +SEM and individual replica of four independent experiments. Unpaired t‐test (two‐tailed). Due to experimental setup, some of the datapoints for panel b) cGAMP and T‐EV+cGAMP are identical to datapoints in panel (d) and (e) ‘No block’.

FIGURE 6

T‐EVs enhance antitumour efficacy of cGAMP. (a) Murine bone marrow derived macrophages (BMMs) were stimulated with murine T‐EVs (1 × 10⁹) or EV‐free media 1 h prior to suboptimal cGAMP stimulation (0.05 μg). The secretion of IFN‐beta was determined after 20 h of stimulation. Data show mean +SEM and individual replica of two independent experiments each in duplicates. (b) Murine BMMs were stimulated with murine T‐EVs (1 × 10⁹) or EV‐free media. At indicated time‐points after T‐EV stimulation, the cells were harvested and analysed by western blotting for phosphorylation of P65. Data is from one experiment. VCL was used as a loading control. (c) MC38 tumour bearing mice were treated with different amounts of murine T‐EVs, administered intratumourally (IT) two times with 3 days interval as indicated with black arrows on the figurers. Treatment started on day 9 after tumour cell inoculation. Data show mean ± SEM of tumour volume up to day 19 after tumour cell inoculation. n = 4 in all groups. (d) MC38 tumour bearing mice were treated with different amounts of cGAMP as indicated, administered IT three times with 3 days interval, as indicated with black arrows on the figurers. Treatment started on day 9 after tumour cell inoculation. Data show mean ± SEM of tumour volume up to day 15 after tumour cell inoculation. n = 6 in all groups. (e) and (f) MC38 tumour bearing mice were treated with either T‐EVs alone (1.5 × 10⁸), cGAMP alone (1 μg) or a combination. Mice were treated IT three times with 3 days interval as indicated with black arrows on the figurers, starting on day 9 after tumour cell inoculation. Data in (e) show mean ± SEM of tumour volume in mice treated with either Vehicle (n = 9), T‐EVs (n = 9), cGAMP (n = 9), or a combination of T‐EVs and cGAMP (n = 8) up to day 15 after tumour cell inoculation. Data in (f) show difference in tumour growth from initiation of treatment on day 9 and until day 15, calculated as ∆TG. Bars indicate mean ± SEM and individual value of ∆TG. TG, tumour growth. (g and h) MC38 tumour bearing mice were treated with either T‐EVs alone (1.5 × 10⁸), cGAMP alone (10 μg) or a combination. Mice were treated IT two times with 2 days interval as indicated with black arrows on the figurers, starting on day 10 after tumour cell inoculation. Data in (g) show mean ± SEM of tumour volume in each group up to day 19 after tumour cell inoculation. Data in (h) shows probability of survival up to day 64 after tumour cell inoculation. n = 9 in each group. (i) Mice that obtained complete tumour regression in (g) was re‐challenged with inoculation of MC38 cells subcutaneously in left flank. As controls was used C57B/6J mice (n = 4). Data show mean ± SEM of tumour volume in each group of either complete responders or control mice.