Platelet-cytokine Complex Suppresses Tumour Growth by Exploiting Intratumoural Thrombin-dependent Platelet Aggregation

Abstract

Tumours constitute unique microenvironments where various blood cells and factors are exposed as a result of leaky vasculature. In the present study, we report that thrombin enrichment in B16F10 melanoma led to platelet aggregation, and this property was exploited to administer an anticancer cytokine, interferon-gamma induced protein 10 (IP10), through the formation of a platelet-IP10 complex. When intravenously infused, the complex reached platelet microaggregates in the tumour. The responses induced by the complex were solely immune-mediated, and tumour cytotoxicity was not observed. The complex suppressed the growth of mouse melanoma in vivo, while both platelets and the complex suppressed the accumulation of FoxP3(+) regulatory T cells in the tumour. These results demonstrated that thrombin-dependent platelet aggregation in B16F10 tumours defines platelets as a vector to deliver anticancer cytokines and provide specific treatment benefits.

Figure 1. Active thrombin is elevated in B16F10 melanoma and platelet aggregates were detected in the vasculature proximity.

(a) Plasma, freshly homogenized tissues of B16F10 and skin were measured for thrombin activity using SensoLyte assay kits. n = 3 mice. Error bars, SD. ****p < 0.001. One-way ANOVA followed by Dunnett’s test. (b) Sections from whole B16F10 tumours (upper panel) or kidney samples (lower panel) were stained for cleaved thrombin (red) and CD31 (green) and examined using a confocal microscope. Scale bars = 100 μm. (c) Thrombin activities of tissue sections from tumours and various normal organs were determined and normalized to the tissue volume. n = 4 mice. ***p < 0.005, ****p < 0.001. One-way ANOVA followed by Dunnett’s test. (d) Thrombin distribution in the surrounding tissues of B16F10 (upper panel) and unaffected skin (lower panel). The white dotted line indicates the tumour border. Scale bars = 100 μm. (e) Relative localization between vasculature (CD31, red) and platelet aggregates (CD41, green) in the B16F10 tumour. The arrowhead indicates a vessel-dissociated aggregate, and the arrows indicate vessel-associated aggregates. Scale bar = 100 μm. (f) In total, 174 clusters of intratumoural platelet aggregates from 7 tumours were examined for spatial relationship with the vasculature. Error bars, SEM.

Figure 2. Incorporation and release of IP10 in platelets.

(a) Fresh platelets were incubated with rm-IP10 in mTyrode’s buffer at 37 °C. The platelets were isolated at time points and lysates were measured for IP10 by ELISA. n = 3. NT, no treatment; NT-2 hr, NT incubated at 37 °C for 2 hr. (b) Platelet-IP10 complexes were stained for IP10 (red) with or without permeabilization and counterstained using phalloidin (green). Scale bars = 20 μm. 100× objectives. (c) More than 100 complexes were examined, and the IP10 signals were enumerated. ****p < 0.0001. Student’s t-test. (d,e) Complexes were treated with buffer, anti-CD41, or anti-CD41 plus thrombin. The released contents were assayed for PDGF (d) and IP10 (e) using ELISA. n = 3. (f) Complexes were stimulated with an increasing concentration of thrombin, and assayed for IP10 release using ELISA. n = 3. Error bars, SEM. nd, not detected, *p < 0.05, ***p < 0.005. One-way ANOVA followed by Dunnett’s test (for (a,d,e)) or Tukey’s test (f) was performed.

Figure 3. Tumour targeting of the platelet-IP10 complex.

(a) Platelet or platelet-IP10 complex or PBS was intravenously injected into B16F10-bearing mice through the tail vein for three consecutive days. The tumour sections were stained for platelet aggregates (CD41, green) and IP10 (red). Arrowheads show examples of aggregation of the platelet-IP10 complex (yellow). Scale bars, 100 μm. (b,c) Tumour-bearing mice were peritoneally injected with argatroban or PBS. PFA perfused tumour tissues were stained for CD31 (red) and CD41 (green). 10× objective. n = 4 mice. Scale bar = 100 μm. *p < 0.05. Student’s t-test. (d) PKH67-labelled platelets were injected with argatroban (arga) or PBS to pre-conditioned tumours. The sizes of the platelet microaggregates in PBS (n = 237) and argatroban (n = 113)-treated tumours were measured. (e) The platelet-biotinylated IP10 complex was constructed, at 2 hours after treatment with a single intravenous dose, the tumours were harvested and measured for intratumoural biotinylated IP10 using ELISA. n = 4–9 tumours per group. nd, not detected, *p < 0.05, ****p < 0.001. Mann-Whitney U-test for (d,e). Error bars, SEM.

Figure 4. Antitumour effects of platelet-IP10 complex in vivo.

(a) Schematic representation of the treatment schedules (Sch). Each block represents one daily unit, and the arrows represent ongoing monitoring. (b) The volume of the B16F10 tumour at 14 and 20 days post-tumour inoculation. n = 5 mice per group. (c) Tumour growth time course study on mice bearing 7-day old subcutaneous B16F10 tumour received indicated intravenous treatments. n = 10–12 mice per group. Error bars, SEM. *p < 0.05. One-way ANOVA was followed by Dunnett’s test in (b,c). (d) The tumour volumes in the platelet-IP10 treatment group were compared with those in the platelet control group at 22 days post-tumour inoculation. The data are pooled from (c) and two other independent experiments. n = 21–22 mice per group. **p < 0.01. Mann-Whitney U-test.

Figure 5. Cellular immune responses induced by the platelet-IP10 complex in vivo.

The mice were inoculated with B16F10 cells and treated following Schedule A, as described in Fig. 4a. The excised tumours were digested with 0.5% collagenase containing 2% FBS until dissociation. Intracellular immune cells were stained for CD4, CD8 and CD49b (a) and FoxP3 (b) gated at CD45⁺ and analysed using flow cytometry. In (c), the ratio of FoxP3⁺ to FoxP3⁻ cells is shown. n = 4–5 tumours. ns, not significant, ***p < 0.005. One-way ANOVA followed by Dunnett’s test was used throughout this figure.

Figure 6. Platelet modulates FoxP3 expression.

(a) Splenocytes were co-cultured with buffer or platelet for 24 (n = 7–10) or 48 hours (n = 3–7) in the presence or absence of thrombin, and foxp3 expression was analysed using qRT-PCR. (b) Splenocytes were treated as in (a) for 48 hours in the presence of thrombin, FoxP3 expression was analysed using flow cytometry. n = 9–10. Error bars, SEM. ****p < 0.0005. Student’s t-test.