Cytoplasmic protein in hepatocytes is sequestered and degraded by two general classes of lysosomes, overt autophagic vacuoles (macroautophagy) and dense bodies (microautophagy). Volumes of the apparent space in each class that contain the internalized protein, together with estimates of cytoplasmic protein concentration, were used as a basis for predicting rates of protein degradation by the lysosomal system in livers of fed, 48-hr starved, and starved-refed mice. Assuming that the turnover of all sequestered protein is equal to that previously determined in overt autophagic vacuoles (0.087 min'), we obtained close agreement between predicted and observed rates in the three conditions studied. The two autophagic components, though, exhibited different patterns of regulation. Microautophagy followed a downward course through starvation and into refeeding, a trend that explained fully the fall in absolute rates of protein degradation during starvation. By contrast, macroautophagy remained constant throughout starvation but was virtually abolished with refeeding. Whereas regulation of the latter can be explained largely by immediate responses to the supply of amino acids, present evidence together with results of others indicate that microsequestration could be linked to functional and quantitative alterations in the smooth endoplasmic reticulum. Both types of regulation contributed equally to the marked suppression of proteolysis during cytoplasmic regrowth.Protein and other cytoplasmic constituents in hepatocytes are internalized within the lysosomal system in two general ways: (i) by the formation of overt autophagic vacuoles, a highly visible mechanism involving the sequestration of organelles within smooth-surfaced membranes (1)(2)(3)(4)(5), and (ii) by the engulfment of small bits of cytoplasm that subsequently appear in dense bodies (1, 3-5), a microautophagic process that we believe accounts for basal protein turnover (5). Although both are ongoing functions, we have shown in perfused livers of fed animals that the addition or deletion of amino acids can regulate overt autophagy rapidly over a wide range without appreciably affecting microautophagy (3, 5). On the other hand, the induction of cytoplasmic growth, such as that produced by starvation and refeeding, is associated with decreases in rates of protein breakdown (6, 7) and sequestered protein (7) to levels that can be explained only by the suppression of both autophagic functions. Nothing is presently known, however, of the way microinternalization is regulated.The present investigation is an attempt to quantify by stereologic methods the contributions of each process to protein breakdown in livers of fed, starved, and starved-refed mice and to compare degradation rates predicted by this approach with experimentally determined rates. Close agreement was found under all conditions, although activities of the individual autophagic components were affected in distinctly independent ways. Rates of overt autophagy were interpretable from the gene...

Previous studies have demonstrated that autophagic vacuoles induced in hepatocytes of the perfused rat liver can be effectively separated from secondary lysosomes in self-generating gradients of colloidal silica/polyvinylpyrrolidone. Here we show that, as autophagy is increased by graded amino acid deprivation, the lysosomal marker N-acetyl-beta-D-glucosaminidase shifts progressively from dense to buoyant gradient fractions, an effect attributable to enhanced lysosome-autophagosome fusion and the formation of additional autolysosomes. Both the amount of the shift and the aggregate volume of autolysosomes were directly proportional to the increase in protein degradation above basal. Because the total quantity of protein sequestered in the vacuoles is also a direct function of deprivation-induced proteolysis, the relative specific activity of the marker enzyme in autolysosomes (enrichment) would be expected to remain constant over the full range of deprivation; numerical estimates of enrichment, in fact, did not differ from the mean (43.9) by more than 5%. The close correspondence between cytoplasmic sequestration and fusion was maintained, despite a greater than 3-fold increase in the average volume of individual vacuoles. This suggests the existence of an intrinsic regulatory mechanism which limits the number of fusions per vacuole, on the basis of information derived from vacuolar volume.

No abstract

Lysosome-enriched samples from normal fed rat livers have been shown to separate into dense and buoyant bands when layered on self-generating gradients of colloidal silica. The hypothesis that the buoyant peak is comprised largely of autophagic vacuoles that are selectively restrained by sedimenting mitochondria was tested in livers subjected to conditions known to accelerate autophagy or suppress it. In all cases, a sharp peak of beta-N-acetyl-D-glucosaminidase activity was evident when autophagy was present but was undetectable when it was eliminated or virtually eliminated. Moreover, the sensitivity to osmotic shock of buoyant peak particles, but not those of the dense band, was increased after autophagic stimulation. The total amount of enzyme that moved to buoyant fractions with the induction of autophagy generally correlated with the degree of autophagic acceleration. However, with maximal autophagy, the buoyant peak was broad, and only 70% was capable of being dislodged and returned to the dense peak by dispersing samples in the gradient medium and/or reducing the sample size. On the other hand, at roughly half-maximal stimulation and lower, the buoyant distribution was sharply defined and completely reversible. These findings support the above hypothesis and suggest that the buoyant peak could be used as a measure of autophagy in the physiological range of regulation.