A novel method of studying wound healing

In order to study wound healing, it is often necessary to administer various wound-active substances by the systemic route. It is unclear whether the observed effects are the result of local or systemic influence of the agent administered. Furthermore, high systemic doses are often required to achieve activity at the wound level. Direct intrawound administration of substances is traumatic and disruptive to the fragile wound environment and increases the risk of infection. We devised a system for continuous atraumatic delivery of substances directly to subcutaneously implanted polyvinyl alcohol sponges, an adaptation of a well-established model of wound healing. Sponge-catheter constructs were fashioned by feeding identical lengths of silicone catheters through two 40-mg sponge disks (on edge). The distal sponge was fixed 0.5 cm from the distal, ligated end of the catheter and centered over two l-mm holes in the catheter tubing. The proximal sponge was fixed over nonperforated catheter with its edge 2 cm proximal from the close edge of the distal sponge. Each construct was connected to a mini-osmotic pump (infusion rate 1 mul/h) loaded with an appropriate infusate and inserted subcutaneously on the dorsum of anesthetized male Sprague-Dawley rats. Hydroxyproline (OHP) content of sponges, a measure of collagen deposition, was determined at 7 days postwounding. Infusion of India ink confirmed selective delivery to the distal sponge. Saline infusion alone significantly elevated OHP content compared to noninfused sponges (450 +/- 43 vs 328 +/- 36 mug OHP/100 mg sponge, P < 0.05). Infusion of S-methylisothiourea (a selective iNOS inhibitor, 84 mug/sponge/24 h) successfully inhibited NO production (35.9 +/- 3.1 vs 49.6 +/- 3.6 muM, P < 0.05) and decreased sponge OHP content (385 +/- 60 vs 568 +/- 70 mug OHP/100 mg sponge, P < 0.05) without the toxic side effect (i.e., weight loss) seen with systemic administration, Infusion of an adenoviral solution containing mouse iNOS cDNA resulted in successful transduction of wound cells demonstrating the ability to deliver genes to a healing wound model. The data demonstrate that manipulation of wound physiology is possible by local delivery of low doses of wound-active compounds to the wound site. This promises 60 be a powerful tool for the study of both normal and impaired wound healing. (C) 2001 Academic Press.