Platelet P2Y12 is involved in murine pulmonary metastasis

Abstract

The involvement of platelets in tumor progression is well recognized. The depletion of circulating platelets or pharmacologic inhibitors of platelet activation decreases the metastatic potential of circulating tumor cells in metastasis mouse models. The platelet ADP receptor P2Y12 amplifies the initial hemostatic responses activated by a variety of platelet agonists and stabilizes platelet aggregation, playing a crucial role in granule secretion, integrin activation and thrombus formation. However, the relationship between P2Y12 and tumor progression is not clear. In our study, the Lewis Lung Carcinoma (LLC) spontaneous metastatic mouse model was used to evaluate the role of P2Y12 in metastasis. The results demonstrated that P2Y12 deficiency significantly reduced pulmonary metastasis. Further studies indicated that P2Y12 deficiency diminished the ability of LLC cells to induce platelet shape change and release of active TGFβ1 by a non-contact dependent mechanism resulting in a diminished, platelet-induced EMT-like transformation of the LLC cells, and that transformation probably is a prerequisite of LLC cell metastasis. Immunohistochemical analyses indicated an obvious P2Y12 deficiency related attenuation of recruitment of VEGFR1+ bone marrow derived cell clusters, and extracellular matrix fibronectin deposition in lungs, which presumably are required for pre-metastatic niche formation. In contrast to the LLC cells, non-epithelial melanoma B16 cells induced platelet aggregation in a cell number and P2Y12-dependent manner. Also, a platelet induced EMT-like transformation of B16 cells is dependent on P2Y12. In agreement with the LLC cell model, platelet P2Y12 deficiency also results in significantly less lung metastasis in the B16 melanoma experimental metastasis model. These results demonstrate that P2Y12 is a safe drug target for anti-thrombotic therapy, and that P2Y12 may serve as a new target for inhibition of tumor metastasis.

Figure 1. P2Y12 deficiency decreases lung metastasis but has no effect on primary tumor growth in the LLC (Lewis lung carcinoma) spontaneous pulmonary metastasis model.

Each group of 9 mice was injected intradermally with 2×10⁶ LLC cells. Subcutaneous primary tumors were removed from each mouse and weighed 14 days after implantation. One month after removal of the subcutaneous tumors, each mouse was dissected, and its lungs were removed, weighed, and photographed. (A) Representative images of gross metastatic lungs from each group of mice. (B) Representative images of tissue sections from metastatic lungs obtained from WT and P2Y12 deficient mice. The lungs were paraffin-embedded prior to sectioning. The sections were stained with hematoxylin and eosin. (C) The mean weights of the primary tumors from the WT and P2Y12^−/− groups. Each bar represents the mean ± SEM of the weights of all the subcutaneous tumors from each group of mice. There was no statistically significant difference between the mean tumor weights of the WT and P2Y12 deficient groups (n=9). (D) The mean weights of metastatic lungs from WT and P2Y12^−/− groups. Each bar represents the mean ± SEM. P2Y12 deficiency significantly inhibited pulmonary metastasis (*p<0.05, n=9).

Figure 2. LLC cell induced platelet shape change is P2Y12-dependent.

(A) In order to investigate the direct interaction between LLC tumor cells and platelets in vitro, 2×10⁵ /ml, 6×10⁵ /ml or 1×10⁶ /ml of Ds-Red labeled LLC cells were added to a suspension of 3×10⁸/ml calcein–labeled WT or P2Y12^−/− platelets in total volume of 300μl, and aggregation and shape change were monitored for 20 minutes with stirring at 1000rpm using a Chrono-Log aggregometeras described. In contrast to the WT platelets, P2Y12 deficient platelets did not undergo shape change or aggregate in response to stirring in the presence of LLC cancer cells (n=3). (B) After stirring, a sample of the mixture of LLC cells and platelets was smeared on the slides, and bright-field and fluorescence images of were taken in the 10x objective field. No obvious direct interaction between LLC cells and platelets was observed.

Figure 3. Platelets promote invasiveness of LLC cells by inducing a P2Y12 and TGF-β1 dependent epithelial-mesenchymal-like transition (EMLT).

(A) Platelet P2Y12 regulated LLC cell invasion in matrigel. As evident from the images and bar graphs, P2Y12 deficient platelets significantly inhibited LLC cell invasiveness. LLC cells were added at the top of transwells coated with matrigel and those cells were treated with buffer, WT platelets and P2Y12^−/− platelets, respectively. The effects of this treatment were evaluated by counting the crystal violet stained cells at the bottom of the wells after 40 hours incubation with buffer or platelets. Each bar represents the mean number of invasive cells ± SEM, statistical significance was determined using a one-way AVONA (10 random fields per sample, n=3 samples, ***p<0.001). (B) LLC cells induced the release of active TGF-β1 from platelets. The LLC cells attached to each well were incubated with buffer, WT or P2Y12^−/− platelets for 48 hours. Cancer cells and platelets were removed from the conditioned medium by centrifugation. The amounts of the active form of TGF-β1 in the supernatant fractions were measured using ELISAs. Each bar represents the mean ± SEM, and the statistical significance was calculated using a one-way AVONA (***p<0.001, n=3). (C) The levels of active TGF-β1 were measured in the supernatant fractions from 300μl volumes of washed platelets stimulated by 20μM ADP, 2μg/ml collagen or 0.1U/ml α-thrombin. EDTA (final concentration of 5mM) was added to each platelet suspension before the supernatant fractions were prepared by centrifugation. The level of TGF-β1 was measured in each supernatant fraction using an ELISA. Each bar represents the mean ± SEM, and the statistical significance was calculated using a one-way AVONA (***p<0.001, n=3). (D) 10× and 20× phase-contrast micrographs of LLC cells that had been incubated with buffer, WT platelets, P2Y12^−/− platelets or 20ng/ml recombinant active TGF-β1 (as a positive control). In contrast to the P2Y12 deficient platelets, incubation of the LLC cells with WT platelets or active TGF-β1 obviously induced the LLC cells to undergo an EMT-like transition.

Figure 4. P2Y12 regulates/affects the formation of the pre-metastatic microenvironment in the lungs.

(A) Representative images of fibronectin expression in lung sections. The levels of tissue fibronectin were measured in lung sections prepared at 3 and 14 days after intradermal implantation of the LLC cells. Immunostaining analysis was used to measure tissue fibronectin. The levels of lung tissue fibronectin were obviously enhanced at day 14 in the WT group, but not in the P2Y12 deficient group. In the latter group, fibronectin was expressed mainly in the regions of micro-vessels and terminal bronchi. Arrows indicate the fibronectin positive areas in the lungs of each group, and enlarged insets in 40× magnification are shown. (B) Statistical analyses of the percentages of fibronectin positive areas in sections of the total lungs for all the animals in each group. The lungs from both groups of animals were approximately the same size. P2Y12 deficiency did not support metastasis-driven enhancement of fibronectin expression (3 sections for each mouse; n=5 for each group, *p<0.05). (C) The numbers of VEGFR1 positive cells were measured in lung sections prepared on days 3 and 14 after intradermal injection of the LLC cells. By day 14, VEGFR1⁺ cell clusters were apparent surrounding the distal bronchials and microvascular regions of the WT mouse lungs. In contrast, the P2Y12 deficient mice had few VEGFR1+ cell clusters in their lungs. Arrows indicate the VEGFR1⁺ cell clusters in the lungs of each group, and enlarged 40× insets in are shown. (D) Statistical analyses of the mean number of VEGFR1⁺ cells per lung section P2Y12 deficiency did not support metastasis-driven recruitment of VEGFR1⁺ cells to the lungs. Each bar represents the mean ± SEM, and statistical significance was determined using a one-way AVONA (3 sections for each mouse; n=5 for each group, *p<0.05).

Figure 5. B16 cells induced platelet aggregation in a cell number and P2Y12-dependent manner.

(A) In order to examine the ability of B16 melanoma cells to directly cause platelet aggregation in vitro, 300μl suspensions containing 2×10⁵/ml, 6×10⁵/ml or 1×10⁶/ml of Ds-Red labeled B16 cells, respectively and 3×10⁸/ml calcein–labeled WT or P2Y12^−/− platelets were prepared, and aggregation was monitored using a Chrono-Log aggregometer. B16 melanoma cells were able to induce visible platelets-cancer cells aggregates. The aggregation was dependent on cancer cell number and WT platelets, and consequently was not supported by P2Y12 deficient platelets (n=3). (B & C) In order to detect the direct interaction between the B16 melanoma cells and the platelets in vitro, after the 20 minute aggregation trial, aliquots of each mixture of B16 cells and platelets were smeared on glass slides, and bright-field and fluorescent images were recorded at 10× and 20× magnifications. The images clearly revealed that some of the B16 cells directly interacted with WT platelets resulting in platelet accumulation on the surface of the B16 cells. In contrast, the P2Y12 deficient platelets did not support extensive accumulation of platelets on the surface of B16 cells (n=3).

Figure 6. P2Y12 facilitates a TGF-β1-independent EMT-like transition of B16 cells, and experimental pulmonary metastasis by B16 cells.

(A) Phase-contrast 10× and 20× micrographs of B16 tumor cells incubated with buffer, WT platelets, P2Y12^−/− platelets or 20ng/ml active TGF-β1, respectively for 48 hours. Incubation with WT platelets obviously induced B16 cells to undergo an EMT-like transition, P2Y12 deficient platelets did not induce an EMT-like transition of B16 cells. Interestingly, the recombinant active TGF-β1 induced the B16 cells to undergo a less extensive EMT-like morphologic change than was induced by the WT platelets. (B) The level of active TGF-β1 in the B16 conditioned medium was also measured by an ELISA. B16 cells induced significantly less release of active TGF-β1 from P2Y12 deficient platelets than from WT platelets. Each bar represents the mean ± SEM, and statistical significance was determined using a one-way AVONA (**p<0.01, n=3). (C) Each mouse from both groups (n=7 for each group) were injected with 2 × 10⁵ B16 cells via a tail vein. Twenty days after injection of the B16 cells, the, lungs were dissected from each mouse, and photographed. Images of visible metastatic foci are apparent in the photomicrographs. (D) Statistical analyses of the number of metastatic foci at the surface of the lung lobes. Each bar represents the mean ± SEM, and statistical significance was determined using the Student’s t test (**p<0.01, n=7 for each group). (E) Representative histochemical images (10×) of lung sections from wild-type and P2Y12 deficient mice.