Abstract
As a cultured fibroblast moves, particles or pieces of debris attached to its surface move backwards with respect to both the cell and the substrate1,2, as also do concanavalin A (Con A) receptors in the membrane of the leading lamella3. One suggested explanation is that membrane components from the cytoplasm are assembled and introduced into the surface membrane of the moving cell close to its leading edge and flow backwards to a 'sink'; there the membrane is disassembled, and its components enter the cytoplasm and flow forward within the cell, to be re-introduced later into the surface membrane1,4. However, because cell-surface receptors can be redistributed as a result of cross-linking by external ligands5,6 it has been proposed that the backward flow of Con A receptors and particles may result from such a cross-linking rather than from a flow of intact cell membrane7,8. To investigate these alternatives, I have studied moving fibroblasts by means of a cell membrane label that does not induce the redistribution of its receptors. My results do not seem compatible with the membrane flow model.