The phenomena of wound contraction and scar retrac-Ž w x. tion are known since old ages for review see 1. In the first part of our century, the work of Carrel and Lecomte du Nouy has contributed to the notion that the forces producing wound contraction are generated within the wx granulation tissue itself 2. These forces were generally considered to depend on extracellular matrix rearrangements; however, Abercrombie and co-workers reported, in the fifties, that fibroblasts exert tractional forces in vitro w x3. Similarly, Hoffmann-Beerling showed that addition of ATP to permeabilized fibroblasts in culture produces a wx contraction of their cytoplasm 4. In this context, and in the context of emerging work on cytoskeleton morphology wx and function 5, the ultrastructural observation made in our laboratory in 1971 that during granulation tissue evolu-Ž. tion fibroblasts acquire smooth muscle SM cell features, such as the presence of cytoplasmic microfilament bundles w x6, lead to the proposition that these cells are the source of the force producing wound contraction, and probably connective tissue retraction during fibrotic phenomena. Shortly thereafter, it was shown that strips of granulation tissue isolated and placed in a pharmacological bath would contract and relax under the influence of substances which are known to be capable of contracting and relaxing SM cells wx Ž 7, 8. It is noteworthy particularly because this observa-