In vivo evidence for platelet-induced physiological angiogenesis by a COX driven mechanism

Abstract

We sought to determine a role for platelets in in vivo angiogenesis, quantified by changes in the capillary to fibre ratio (C:F) of mouse skeletal muscle, utilising two distinct forms of capillary growth to identify differential effects. Capillary sprouting was induced by muscle overload, and longitudinal splitting by chronic hyperaemia. Platelet depletion was achieved by anti-GPIbα antibody treatment. Sprouting induced a significant increase in C:F (1.42±0.02 vs. contralateral 1.29±0.02, P<0.001) that was abolished by platelet depletion, while the significant C:F increase caused by splitting (1.40±0.03 vs. control 1.28±0.03, P<0.01) was unaffected. Granulocyte/monocyte depletion showed this response was not immune-regulated. VEGF overexpression failed to rescue angiogenesis following platelet depletion, suggesting the mechanism is not simply reliant on growth factor release. Sprouting occurred normally following antibody-induced GPVI shedding, suggesting platelet activation via collagen is not involved. BrdU pulse-labelling showed no change in the proliferative potential of cells associated with capillaries after platelet depletion. Inhibition of platelet activation by acetylsalicylic acid abolished sprouting, but not splitting angiogenesis, paralleling the response to platelet depletion. We conclude that platelets differentially regulate mechanisms of angiogenesis in vivo, likely via COX signalling. Since endothelial proliferation is not impaired, we propose a link between COX1 and induction of endothelial migration.

Figure 1. Different forms of angiogenesis and experimental protocols.

A) Mechanisms of capillary sprouting and longitudinal splitting. B) Extirpation induced endothelial sprouting, with anti-GPIbα, anti-Ly6G/6C or anti-GPVI administration 24 h after surgery. C) Prazosin induced longitudinal splitting with anti-GPIbα. Sampling always occurred on the eighth day.

Figure 2. Platelet depletion differentially affects skeletal muscle angiogenesis.

A) Capillary sprouting produced a significant capillary to fibre ratio (C∶F) increase, abolished by platelet depletion following i.p. injection of anti-GPIbα(denoted by ‘-’). Depletion did not affect C∶F of untreated or contralateral limbs. B) Longitudinal splitting caused a significant increase in C∶F which was unaffected by platelet depletion. *P<0.05, **P<0.01, ***P<0.001 vs. untreated. C) Detail of representative images of lectin-stained mouse muscle cross-section showing fibres (asterisks) and capillaries (arrows) for extirpation with (i) or without (ii) platelets, and prazosin with (iii) or without (iv) platelets.

Figure 3. Granulocytes/monocytes do not mediate endothelial sprouting.

A) Representative whole blood smears showing lymphocytes (black arrows) and granulocytes/monocytes (white arrow) were visualised in control animals. B) Following i.p. administration of anti-Ly6G/6C antibody only lymphocytes were readily visualised at 4, 24, and 48 h. C, Granulocyte/monocyte depletion did not alter the endothelial sprouting response normally observed (P<0.01).

Figure 4. Platelets differentially affect angiogenesis via COX signalling.

A) Dual clopidogrel/ASA regimen inhibited sprouting angiogenesis. Single regimens identified ASA as the active agent, with clopidogrel unable to alter the angiogenic response. B) Longitudinal splitting was unaffected by ASA or lower dose (LD)-ASA. C) PGF1α dropped significantly with overload + LD-ASA. D) Induction of longitudinal splitting did not alter PGF1α levels. *P<0.05, **P<0.01 between columns (A), or vs. untreated controls (B).

Figure 5. VEGF overexpression cannot rescue angiogenic deficit.

A) Extensor digitorum longus muscle VEGF expression in wildtype and MUC1-VEGF mice. MUC1-VEGF human VEGF (•) plus mouse VEGF (○) resulted in a 45% increase in total content (▪) vs. wildtype. B) MUC1-VEGF mice had higher initial C∶F than wildtype. Capillary sprouting with or without platelets was no different from wildtype. *P<0.05 vs. C57 BL/6 controls, +P<0.05 vs. contralateral.

Figure 6. Platelets are observable in skeletal muscle vasculature.

Platelets were labelled in vivo prior to muscle sampling. A) Imaging of muscle sections for the presence of platelets occurred under fluorescent microscopy in conjunction with B) rhodamine Griffonia simplicifolia lectin-1 staining for visualisation of vasculature, and C) DAPI for localisation of nuclei. D) A merged image of the three other panels depicts platelets within a venule, a potential locus for angiogenesis.

Figure 7. Angiogenesis is independent of collagen-induced platelet activation.

A) Anti-GPVI resulted in ∼70% GPVI shedding. Shaded grey peak, GPVI shed; clear black peak, control; clear grey peak, GPVI shed IgG expression. B) GPVI shedding did not alter capillary sprouting, with significantly increased C∶F. *P<0.05 vs. contralateral.

Figure 8. Platelet depletion did not affect proliferation of capillary-associated cells or myocytes.

A) The total number of proliferating cells increased slightly in tissue after induction of capillary sprouting, but was not affected by platelet depletion. B) Capillary-associated cell proliferation was unaffected by platelet depletion. C) The increase in proliferating cell number with sprouting results from increased interstitial cell proliferation. D) The absence of platelets did not impair the proliferation of myocytes following muscle overload. E) BrdU labelling was performed in conjunction with rhodamine Griffonia simplicifolia lectin-1 staining (bottom left panel) for visualisation of vasculature, and DAPI (top right panel) to ensure BrdU staining observed was localised to nuclei. A merge (bottom right panel) of the three other panels with three BrdU-labelled interstitial cells (white arrows) is shown. Imaging was performed with fluorescent light microscopy as described in the methods at ×40 magnification. *P<0.05 vs. contralateral.