Methods for computer assisted analysis of lymphoid cell shape and motility, including Fourier analysis of cell outlines.
Locomotion of lymphocytes and other leukocytes is an essential feature of the immune system, and therefore the evaluation of the locomotor behaviour of a lymphocyte population is part of its functional analysis. Paradoxically, the locomotor status of leukocytes is usually assessed on the basis of static information, by counting the number of spherical versus non-spherical cells. In this paper we describe two methods for the measurement of shape changes in microscopic images of lymphoid cells. First we computed a simple shape change factor, coined incongruence factor, based on the degree of non-overlap of the contours of the cell at the beginning and at the end of a 1 min time interval. Second we have used Fourier analysis of the cell outline: a function describing the undulations of the cell outline is broken down into sinusoidal 'waves' of increasing frequency, each with its corresponding amplitude. The amplitude values for the first ten frequencies produced a satisfactory mathematical description of lymphoid cell shapes, and the change of these amplitudes over a 1 minute time interval produced a quantitative description of the shape alterations of the cells. We have used five approaches to evaluate the shape and shape changes in the following populations of mouse lymphoma cells: a constitutively low-motile T lymphoma cell line (BW5147), a high-motile hybridoma (BW-O-Li1) either on plastic or on a precultured fibroblast-like monolayer, BW-O-Li1 cells after penetration through the monolayer, and BW-O-Li1 cells after treatment with cytochalasin B. We compare the results from direct visual evaluation of cell shape, from computer assisted assessment of sphericity and from Fourier analysis of cell shape at one moment, with the two methods for quantitative shape change analysis. All approaches revealed a clear distinction between spherical low-motile populations, and non-spherical high-motile cells. Moreover, the incongruence factor proved to be a reliable single parameter of active cell deformation. In addition, the Fourier analysis of cell outlines produced useful measures of static shape and of dynamic shape change, at any user-defined level of accuracy.