The search for a fundamental mechano-chemical process that results in net cell motion has led investigators to fit neutrophil tracking data to well described physical models in hopes of understanding the functional form of the driving force. The Ornstein-Uhlenbeck (OU) equation for mean square displacement describes a locally persistent and globally random process and is often used as a starting point for analysis of neutrophil displacements. Based upon the apparently close fit of neutrophil tracking data to this equation and the nature of its derivation, biologists have inferred that the motor of the neutrophil is best represented as a random process. However, 24 of 37 neutrophil paths that we investigated preferentially display programmatic rather than Markov short term correlations between displacements or turn angles. These correlations reflect a bimodal rather than a uniform distribution of subpath correlations in the two variables, and are strongly sampling rate-dependent. Significant periodic components of neutrophil shape change are also detected at the same time scale using either Fourier or elliptical Fourier transform-based descriptors of the neutrophil perimeter. Oscillations in neutrophil velocity have the same period. Taken together, these data suggest a nonstochastic, and perhaps periodic, component to the process driving neutrophil movement.
展开▼
