AJP - Endocrinology and Metabolism, Vol 267, Issue 5 E759-E768, Copyright © 1994 by American Physiological Society
Convolution: a method for data analysis in perifusion systems
R. M. Brand, A. R. Midgley and W. J. Williams
Department of Electrical Engineering and Computer Sciences, University of Michigan, Ann Arbor 48109-0404.
Considerable interest has developed in defining how imposed stimuli effect
dynamic changes in cellular metabolism. We have developed a miniature
perifusion system that can reveal alterations in extracellular protons
within seconds after application of metabolic perturbants. This perifusion
system contains two pH sensors: one before the cells records changes
between medium and test solutions, and one, located just past the cells,
records these alterations plus cellular modifications. Because distortion
occurs as chemicals pass through perifusion systems, the shape of pH
changes induced by switching from medium to test solutions is different at
each electrode. This study describes and validates convolution to correct
this distortion. Data from HeLa cells exposed to the metabolic uncoupler of
oxidative phosphorylation carbonyl cyanide m-chlorophenyl hydrazone have
been analyzed with convolution. Cellular response to oxidative
phosphorylation removal is comprised of multiple components, is consistent
with a rapid uncoupling, and is followed by cellular adaptation. Therefore
convolutional analysis can provide an important adjunct to the analysis of
data acquired by perifusion and can provide new insights into cellular
responsiveness and metabolism.
