Abstract
Schafer et al. (Nature 352:444-448 (1991)) devised the tethered particle motion (TPM) method to detect directly the movement of single, isolated molecules of a processive nucleic acid polymerase along a template DNA molecule. In TPM studies, the polymerase molecule is immobilized on a glass surface, and a particle (e.g., a 0.23 microns diameter polystyrene bead) is attached to one end of the enzyme-bound DNA molecule. Time-resolved measurements of the DNA contour length between the particle and the immobilized enzyme (the "tether length") are made by determining the magnitude of the Brownian motion of the DNA-tethered particle using light microscopy and digital image processing. We report here improved sample preparation methods that permit TPM data collection on transcript elongation by the Escherichia coli RNA polymerase at rates (approximately 10(2)-fold higher than those previously obtained) sufficient for practical use of microscopic kinetics techniques to analyze polymerase reaction mechanisms. In earlier TPM experiments, calculation of tether length from the observed Brownian motion was based on an untested numerical simulation of tethered bead Brownian motion. Using the improved methods, we have now empirically validated the TPM technique for tether lengths of 308-1915 base pairs (bp) using calibration specimens containing particles tethered by individual DNA molecules of known lengths. TPM analysis of such specimens yielded a linear calibration curve relating observed Brownian motion to tether length and allowed determination of the accuracy of the technique and measurement of how temporal bandwidth, tether length, and other experimental variables affect measurement precision. Under a standard set of experimental conditions (0.23 microns diameter bead, 0.23 Hz bandwidth, 23 degrees), accuracy is 108 and 258 bp r.m.s. at tether lengths of 308 and 1915 bp, respectively. Precision improves linearly with decreasing tether length to an extrapolated instrumentation limit of 10 bp r.m.s. and improves proportionally to the inverse square root of measurement bandwidth (1.9 x 10(2) bp Hz-1/2 for 1090-bp tethers). Measurements on large numbers of individual polymerase molecules reveal that time-averaged single-molecule elongation rates are more variable than is predicted from the random error in TPM measurements, demonstrating that the surface-immobilized RNA polymerase molecules are kinetically heterogeneous.