Neutrophil chemotaxis. Analyses of neutrophil chemotaxis i a new fluid gradient chamber.

Abstract

The leukocytes have, as part of their function in immune system, the ability to leave the bloodstream, penetrate the vascular wall and migrate towards extravascular targets. This behaviour of the leukocytes has been extensively studied both in vivo and in vitro since the 19th century but the mechanisms underlying leukocyte migration are not yet fully understood. The Fluid Gradient Chamber introduces new techniques for the creation and characterization of controlled spatio-temporal chemotactic gradients. A theoretical model based on normal mixture distribution was developed and used for data analysis. The new methodology has been used in detailed population studies of migrating leukocytes which has enabled an identification of discrete subpopulations of human granulocytes, classified on the basis of their migratory behaviour under chemokinetic and chemotactic stimulation with f-MLP. Studies of the time history of the chemotactic response showed that random migration occurs immediately following the exposure to a chemotactic gradient whereas true chemotaxis develops 20 - 30 minutes later. The initial slow response to the chemotactic gradient represents an adaptation of the cells that later respond promptly to changes in gradient direction. Studies of moving gradient have established that a temporal gradient (concentration change over time) of chemotaxin is necessary for maintaining chemotactic movement. It is suggested that concentration change over time may act as a positive feedback signal for cell migration. Comparison of our theoretical model for data analysis, based on normal mixture distribution, with the well established model of Keller & Segel based on partial differential equation showed good agreement between the models. On the contrary, simulation of temporal gradient (moving spatial gradients), using the Keller &Segel model did not reflect the experimental effects of temporal gradient, earlier demonstrated with the normal distribution model.