Response to Harper et al.

Abstract

We thank Harper et al. for their comments on our publication. The authors of this correspondence point to some differences between our publication (Schumacher et al., 2013xSchumacher, D., Strilic, B., Sivaraj, K.K., Wettschureck, N., and Offermanns, S. Cancer Cell. 2013; 24: 130–137Abstract | Full Text | Full Text PDF | PubMed | Scopus (77)See all ReferencesSchumacher et al., 2013) and others with regard to the in vitro phenotype of Munc13-4-deficient platelets. Based on our experimental data, we think that Munc13-4 deficiency primarily blocks the release from dense granules, but not from α-granules of platelets. It is, however, well known that platelet activation involves multiple positive feedback loops and that the release of ATP/ADP from dense granules can promote platelet aggregation and α-granule release via the autocrine and paracrine activation of P2Y1 and P2Y12 receptors (Gachet, 2006xGachet, C. Annu. Rev. Pharmacol. Toxicol. 2006; 46: 277–300Crossref | PubMed | Scopus (195)See all References, Kahner et al., 2006xKahner, B.N., Shankar, H., Murugappan, S., Prasad, G.L., and Kunapuli, S.P. J. Thromb. Haemost. 2006; 4: 2317–2326Crossref | PubMed | Scopus (126)See all References). Based on this concept, it is indeed expected that loss of ATP/ADP release from platelet dense granules also affects, under certain experimental conditions, the release of α-granules from activated platelets, as well as the aggregation of platelets. Whether these secondary effects of a blocked dense-granule release are seen in vitro depends on the experimental conditions used to study platelet function, such as platelet concentration or the concentration and type of applied stimuli. At intermediate thrombin concentrations, we actually saw a tendency of reduced PF4 release and integrin αIIbβ3 activation which, however, was not significant (Figures 2A and S2C in Schumacher et al., 2013xSchumacher, D., Strilic, B., Sivaraj, K.K., Wettschureck, N., and Offermanns, S. Cancer Cell. 2013; 24: 130–137Abstract | Full Text | Full Text PDF | PubMed | Scopus (77)See all ReferencesSchumacher et al., 2013). In the experiments mentioned by Harper et al., we used relatively low platelet concentrations, reducing the likelihood that sufficient amounts of ATP/ADP are released to induce secondary effects. Finally, at least in the case of tumor-cell-induced platelet activation, it is also possible that the small amount of ATP released from tumor cells (see Figure 1D in Schumacher et al., 2013xSchumacher, D., Strilic, B., Sivaraj, K.K., Wettschureck, N., and Offermanns, S. Cancer Cell. 2013; 24: 130–137Abstract | Full Text | Full Text PDF | PubMed | Scopus (77)See all ReferencesSchumacher et al., 2013) obscured a somewhat reduced α-granule release.Whether the secondary effects of dense-granule release are of relevance for platelet-dependent tumor cell metastasis is not known. Nevertheless, we agree that solely on the basis of our experiments performed with Munc13-4-deficient platelets and animals, we would not be able to exclude that defects other than reduced dense granule secretion also contribute to the observed defects in transendothelial tumor cell migration and in vivo metastasis. We therefore mentioned in the discussion of this publication the possibility that “in addition to ATP, other platelet-derived factors, such as transforming growth factor β, promote, in particular, later stages of tumor cell extravasation (Labelle et al., 2011xLabelle, M., Begum, S., and Hynes, R.O. Cancer Cell. 2011; 20: 576–590Abstract | Full Text | Full Text PDF | PubMed | Scopus (299)See all ReferencesLabelle et al., 2011).”However, our study goes beyond the description of the role of dense-granule secretion in platelet-dependent tumor cell transendothelial migration and metastasis. We also identified a downstream mechanism by demonstrating that the endothelial ATP receptor P2Y2 mediates platelet-dependent tumor cell transmigration and that loss of P2Y2, which is not present on platelets, strongly reduces tumor cell metastasis from primary tumors, as well as after intravenous injection of tumor cells. If the above described secondary effects played a major role in platelet dense-granule secretion-dependent tumor cell transendothelial migration and metastasis, we would expect the phenotype of Munc13-4 deficiency to be more severe than the phenotype of P2Y2 deficiency. However, both platelet Munc13-4 deficiency and endothelial knockdown of P2Y2 abolished platelet-dependent transendothelial tumor cell migration. More importantly, tumor cell metastasis in vivo, when studied in bone-marrow chimeras, was reduced both in P2Y2-deficient mice transplanted with wild-type bone-marrow and in P2Y2-deficient mice transplanted with Munc13-4-deficient bone marrow to the same degree, and no additive effect could be seen.Thus, while we acknowledge that loss of dense-granule secretion in Munc13-4-deficient platelets can lead under certain experimental conditions to a secondary impairment of α-granule release, we think that our data show that dense-granule-derived ATP acting on P2Y2 receptors is the primary mechanism through which blockade of ATP release from Munc13-4-deficient platelets affects tumor cell transendothelial migration and metastasis.