| Abstract: | By use of digital autocorrelation and fast Fourier methods, dynamic light-scattering studies of in vitro reconstituted muscle F-actin were made over a wide range of concentrations, 0.01-2 mg/ml F-actin. Measurements of correlation function [g1(t)]2 showed that a transition from a dilute to a semidilute regime for the Brownian motions of filaments occurred at around 0.3 mg/ml F-actin. Beyond this concentration, profiles of successively measured [g1(t)]2 showed very poor reproducibility. This resulted from the existence of very slow components, which could not be measured with a high statistical accuracy even for a measuring time of 3600 s/run. On the other hand, subtraction of these components automatically by an electronic circuit, [g-1(t)]2, or by computer processing, [g1(t)]2, resulted in a fairly good reproducibility of the profiles. The decay characteristics of [g1(t)]2 (and [g-1(t)]2) were very similar to those of [g1(t)]2 for dilute solutions. A theoretical model will be discussed which could account for the above situation. The time sequence [n(t,T)] of photoelectron counts at a sampling time T of light scattered from semidilute solutions of F-actin was stored on magnetic tapes, and both power spectra S(f) and correlation functions [g-1(t)]2 were computed by taking the ensemble average over many short records with duration 1024T. Since both S(f) and [g-1(t)]2 lacked frequency components lower than 1/(2048T) Hz, their profiles were highly reproducible. An analysis of S(f) confirmed our earlier results which had shown an apparent contradiction to later results by a correlation method. A comparison of S(f) and [g-1(t)]2 based on the same [n(t,T)] clarified the reasons why the bandwidth gamma of S(f) largely differed from the bandwidth gamma of [g1(t)]2 and [g-1(t)]2. The temperature dependence of gamma suggested that F-actin would be flexible and that the flexibility parameter would change with temperature. |
