Large dense core vesicles (LDV) were purified from bovine splenic nerve homogenates by the sucrose-D2O density gradient method. Vesicles were subjected to a 50% increase and decrease in osmolarity from the control 330 mosmol 1(-1) by adjusting sucrose or potassium phosphate buffer during pre- and/or postfixation. Control vesicles with a mean diameter of 717 A readily swelled to approximately 1050 A and shrunk to approximately 600 A in the hypotonic and hypertonic media, respectively, with either sucrose or phosphate buffer. The dense core responded similarly but to a lesser degree. Prefixation in glutaraldehyde had little effect on vesicle sensitivity to subsequent tonicity change, not did the fixative per se exert an obvious osmotic effect. Thus, final vesicle size was largely determined by the OsO4 postfixation medium and principally by the vehicle rather than the fixative. In controls there was a mixture of spherical to oblate vesicles mostly filled with an electron-dense matrix. Upon swelling, more vesicles became spherical and nearly all had a prominent translucent halo between core and membrane. Upon shrinking, more vesicles became oblate, the halo was obliterated and the electron-density of the matrix increased. Frequency distributions of vesicle diameters at different tonicity clearly indicated that the diameter of LDV could overlap the 400-500 A range characteristic of small dense core vesicles; however, there was no suggestion of a population of the latter in the purified LDV fraction. Implications are discussed concerning the biochemical and morphological identification of 'light' and 'heavy' density peaks of noradrenaline and dopamine beta-hydroxylase from mixed vesicle populations and the possible relevance of changes in vesicle shape to a functional state in situ.

Knowledge of the vesicular origin of circulating dopamine beta-hydroxylase (DbetaH) is indispensable for any attempts to explain the parallelism or lack of it between circulating enzyme and catecholamines as they may relate to physiological stress, forms of hypertension, neurological disorders, and the response to pharmacological agents. The present study represents an effort to evaluate and to place in proper perspective data based on the DbetaH activity found in the region of the light vesicle peak of noradrenaline (NA), which is used as a quantitative measure of a population of small terminal vesicles. Distributions of vesicles and subvesicular components are compared with DbetaH and NA in sucrose-D2O density gradients used to prepare relatively pure fractions of large dense cored vesicles (LDV) from bovine splenic nerve. Although NA in sedimentable particles of the light vesicle peak is likely to be a valid measure of a small vesicle population, the following is demonstrated: (1) A substantial fraction (25%-37%) of the total sedimentable DbetaH activity can be proven to distribute in the region of the light vesicle peak from a tissue with an insignificant small vesicle population. Based on studies of vesicles from sequential nerve segments, this enzyme activity probably corresponds to a population of "immature" LDV which are undergoing axoplasmic transport and have not synthesized their full complement of transmitter. (2) Physical lysis which depletes the matrix of LDV causes redistribution of DbetaH activity from the heavy vesicle peak into the region of the light vesicle peak. Analogously, DbetaH associated with exocytosed LDV and retrograde transport particles is also likely to contaminate the region of the light vesicle peak. (3) Based on available data, it can be calculated that each small dense cored vesicle could contain only 0.1-0.5 molecules of DbetaH and that a contamination of only 0.016% LDV can account for all of the DbetaH reported to occur in the light vesicle peak of normal rat vas deferens preparations.