Abstract. Traction forces produced by moving fibroblasts have been observed as distortions in flexible substrata including wrinkling of thin, silicone rubber films. Traction forces generated by fibroblast lamellae were thought to represent the forces required to move the cell forwards. However, traction forces could not be detected with faster moving cell types such as leukocytes and growth cones (Harris, A. K., D. Stopak, and P. Wild. 1981. Nature (Lond.). 290:249-251). We have developed a new assay in which traction forces produced by rapidly locomoting fish keratocytes can be detected by the two-dimensional displacements of small beads embedded in the plane of an elastic substratum. Traction forces were not detected at the rapidly extending front edge of the cell. Instead the largest traction forces were exerted perpendicular to the left and right cell margins. The maximum traction forces exerted by keratocytes were estimated to be ,x,2 x 10 -8 N. The pattern of traction forces can be related to the locomotion of a single keratocyte in terms of lamellar contractility and area of close cell-substratum contact.T o move cells must exert traction forces upon the substratum. This involves the temporal and spatial regulation of numerous force generating molecular motors. Yet an understanding of how and where moving cells generate traction forces, represents a major gap in our knowledge of cell locomotion. This paper presents the first measurements of the traction forces generated by rapidly moving cells.Traction forces produced by moving fibroblasts were first observed as distortions in flexible substrata that caused wrinkling of thin, silicone rubber films (Harris et al., 1980). These traction forces act inwards, relative to the extending lamella and retracting edge, leading to compression of the substratum such that wrinkles are formed perpendicular to the direction of lamellar extension. Wrinkles were thought to be formed by an actomyosin-based contraction of the cytoskeleton which is transmitted to the substratum via focal adhesions located just behind the extending edge and trailing cell edge. Traction forces generated by fibroblast lamellae were thought to represent the forces required to move the cell forwards along the substratum. It was therefore surprising to find that traction forces generated by faster moving cell types such as leukocytes and growth cones could not be detected (Harris, 1981), since it was assumed that larger traction forces would be required for faster locomotion. However, slow moving cells such as fibroblasts form strong focal adhesions to the substratum whereas faster moving cells tend to form weaker close contacts (Couchman and Reese, 1979).In addition large numbers of actin stress fibers are found in slower moving cells, implying greater cytoskeletal contractility. Therefore rapid cell locomotion appears to rely on both weaker cell substratum adhesions and cytoskeletal contractility.To learn more about the traction forces required for rapid locomotion, we have modified the traction ...

Stem cells are thought to balance self-renewal and differentiation through asymmetric and symmetric divisions, but whether such divisions occur during hematopoietic development remains unknown. Using a Notch reporter mouse, in which GFP acts as a sensor for differentiation, we image hematopoietic precursors and show that they undergo both symmetric and asymmetric divisions. In addition we show that the balance between these divisions is not hardwired but responsive to extrinsic and intrinsic cues. Precursors in a prodifferentiation environment preferentially divide asymmetrically, whereas those in a prorenewal environment primarily divide symmetrically. Oncoproteins can also influence division pattern: although BCR-ABL predominantly alters the rate of division and death, NUP98-HOXA9 promotes symmetric division, suggesting that distinct oncogenes subvert different aspects of cellular function. These studies establish a system for tracking division of hematopoietic precursors and show that the balance of symmetric and asymmetric division can be influenced by the microenvironment and subverted by oncogenes.

The cytoskeletal activity of motile or adherent cells is frequently seen to induce detectable displacements of sufficiently compliant substrata. The physics of this phenomenon is discussed in terms of the classical theory of small-strain, plane-stress elasticity. The main results of such analysis is a transform expressing the displacement field of the elastic substrate as an integral over the traction field. The existence of this transform is used to derive a Bayesian method for converting noisy measurements of substratum displacement into "images" of the actual traction forces exerted by adherent or locomoting cells. Finally, the Monte Carlo validation of the statistical method is discussed, some new rheological studies of films are presented, and a practical application is given.