Capillary growth is a hallmark of the adaptation of skeletal muscle to endurance exercise training, i.e. repetitive bouts of low-intensity muscular activity, and represents physiological angiogenesis within mature differentiated tissue. Vascular endothelial growth factor (VEGF) has been implicated as an endothelial mitogen in exercising muscle on the basis that it could be upregulated by local tissue hypoxia in active muscle fibres. VEGF mRNA is indeed increased 4-fold after a single treadmill run in rat muscle (Breen et al. 1996), with similar findings in human muscle biopsies taken after knee extensor exercise (Richardson et al. 1999). However, capillary growth during training normally takes several weeks to appear and translation of this immediate message into an angiogenically effective signal for capillary growth in the longer term is less well defined. Likewise, the relationship of muscle hypoxia and VEGF to capillary growth during training is not fully determined.

Muscle contractions evoked by indirect low-frequency electrical stimulation can mimic exercise and be used to induce much more rapid capillary growth, within a week in rats and rabbits. In this model, we investigated the temporal and spatial aspects of VEGF protein expression by immunohistochemistry in relation to angiogenesis in ankle flexor muscles, extensor digitorum longus (EDL) and tibialis anterior (TA). From in situ on frozen rat muscle sections, VEGF mRNA was evident within muscle fibres at sub-sarcolemmal and peri-nuclear sites, and also at capillary locations. Constitutive VEGF protein expression was confirmed in muscle fibres and also in vascular smooth muscle (VSM) of larger vessels, in interstitial fibroblasts and other cell types. Around 15 % of capillaries (identified by lectin or alkaline phosphatase staining) co-localised to VEGF immunostaining. Repetitive bouts of contractions, using implanted remote-operated battery stimulators at 10 Hz for 8 h per day followed by a 16 h rest period, increased capillary per fibre ratio (C:F) in rat EDL by day 7, but proliferation, detected by proliferating cell nuclear antigen (PCNA), was increased > 20-fold at capillary sites by day 2. VEGF immunostaining increased at all locations only after 3-4 days of stimulation, with 40 % of capillary sites positive for VEGF and remaining so while C:F continued to increase. Although capillary proliferation by day 2 occurred preferentially around glycolytic fibres, only a fraction of PCNA-positive capillaries also expressed VEGF and of these, their distribution was not related to fibre type. VEGF protein is therefore not always expressed in conjunction with angiogenic loci, and even when it is, it is not necessarily linked with muscle fibres that would be metabolically stressed. Hypoxia may not be the sole regulator of VEGF expression in activity-induced angiogenesis, supported by direct measurements of stimulated muscle P_O2 showing that any decrements are small and transient. An alternate stimulus to VEGF upregulation could be stretch of fibres during contraction, which has been shown to enhance VEGF in cultured cardiac myocytes and microvascular endothelial cells (Zheng et al. 2001).

The increased presence of VEGF immunostaining at capillary locations in stimulated muscles would indicate autocrine regulation since endothelial cells are both source and target for VEGF; its function here may be mitogenic and/or for stabilisation of vessels. However, after 2 days stimulation, capillary endothelial cells and adjacent fibroblasts both show proliferation (Egginton et al. 2001), and the latter are clearly also a source of VEGF at capillary sites and in the interstitium. The resident population of ED2-positive macrophages is increased in stimulated muscles, and the proportion of these that express VEGF is increased 3-fold after 2 days, providing another paracrine source. The role of VEGF in VSM of arterial vessels in activity-induced angiogenesis can only be speculated upon at present but upregulation in this location, possibly via adenosine (Gu et al. 1999) could contribute to the production of matrix metalloproteinases from VSM that either help to mobilise VEGF, or encourage migration of mesenchymal cells participating in the ‘arteriolarisation’ of capillaries. Capillary growth in stimulated muscles is indeed linked with elevated activity of endothelial cell-stimulating angiogenic factor, ESAF, which activates the pro-forms of metallo-proteinases gelatinase A, collagenase and stromelysin, and it is accompanied by arteriolar growth (Hansen-Smith et al. 1998).

VEGF does not function alone in activity-induced angiogenesis in skeletal muscle. We have shown that capillary growth in stimulated muscles can be suppressed by concurrent treatment with inhibitors of either prostaglandin production or nitric oxide synthase, and there is confirmatory evidence that the VEGF mRNA response to acute exercise can be attenuated by inhibition of the latter. Basic fibroblast growth factor does not appear to be involved in the angiogenesis, whereas exercise-induced increases in TGF-β1 mRNA accompany those of VEGF. There is, as yet, only preliminary evidence of exercise effects on VEGF receptor expression in skeletal muscle and this is clearly a key area for future investigation, along with the role of the soluble VEGF receptor as a regulator of VEGF availability.This work was funded by The Wellcome Trust and British Heart Foundation, UK, and the late Dr J. Barclay.

Breen, E.C., Johnson, E.C., Wagner, H., Tseng, H.-M., Sung, L.A. & Wagner, P.D. (1996). J. Appl. Physiol. 81, 355-361.

Egginton, S., Zhou, A.-L., Brown, M.D. & Hudlicka, O. (2001). Cardiovasc. Res. 49, 634-646.

Gu, J.-W., Brady, A.L., Anand, V., Moore, M.C., Kelly, W.C. & Adair, T.H. (1999). Am. J. Physiol. 277, H595-602.

Hansen-Smith, F., Egginton, S. & Hudlicka, O. (1998). Microcirculation 5, 49-59.

Richardson, R.S., Wagner, H., Mudaliar, S.R., Henry, R., Noyszewski, E.A. & Wagner, P.D. (1999). Am. J. Physiol. 277, H2247-2252.

Zheng, W., Seftor, E.A., Meininger, C.J., Hendrix, M.J.C. & Tomanek, R.J. (2001). Am. J. Physiol. 280, H909-917.