Saccharomyces cerevisiae mannoproteins form an external cell wall layer that determines wall porosity

A beta-glucanase (Z-glucanase) from Zymolyase was freed from a protease (Z-protease) by affinity chromatography on alpha 2-macroglobulin-Sepharose columns and used to solubilize proteins from isolated cell walls of Saccharomyces cerevisiae. The cell wall proteins were labeled with 125I and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and autoradiography. The bulk of the labeled material had very low mobility. Its mannoprotein nature was demonstrated by precipitation with specific antibodies and by conversion to a band with an average molecular weight of 94,000 after incubation with endo-beta-N-acetylglucosaminidase. The intact mannoproteins were hydrolyzed by Z-protease, but were resistant to the enzyme when the carbohydrate was first removed by endo-beta-N-acetylglucosaminidase. In intact cells, lysis of cell walls by Z-glucanase required a previous incubation with z-protease, which led to solubilization of most of the 125I-labeled proteins. Other proteases that did not attack the cell wall mannoproteins were unable to substitute for Z-protease. The specific effect of Z-protease is consistent with the notion that mannoproteins form a surface layer of the cell wall that penetrates the wall to some depth and shields glucans from attack by Z-glucanase. Mannoproteins, however, do not appear to cover the inner face of the cell wall, because isolated cell walls, in contrast to intact cells, were completely solubilized by Z-glucanase in the absence of protease. The function of mannoproteins in determining cell wall porosity was highlighted by the finding that horseradish peroxidase (Mr, 40,000) causes lysis of cells that had been treated with Z-protease. Depletion of mannoproteins by Z-protease also resulted in the disappearance of a darkly stained surface layer of the cell wall, as observed by electron microscopy. Other agents that facilitate cell lysis by Z-glucanase, such as 2-mercaptoethanol, digitonin, and high concentrations of salts, caused little or no solubilization of mannoprotein. We assume that they perturb and loosen the structure of the mannoprotein network, thereby increasing its porosity. The implications of our results for the construction of the yeast cell wall and the anchoring of mannoprotein to the cell are discussed.

Isolation of cDNA and genomic DNA clones encoding type II collagen

A cDNA library constructed from total chick embryo RNA was screened with an enriched fraction of type II collagen mRNA. Two overlapping cDNA clones were characterized and shown to encode the COOH propeptide of type II collagen. In addition, a type II collagen clone was isolated from a Charon 4A library of chick genomic fragments. Definitive identification of the clones was based on DNA sequence analysis. The 3' end of the type II collagen gene appears to be similar to that of other interstitial collagen genes. Northern hybridization data indicates that there is a marked decrease in type II collagen mRNA levels in chondrocytes treated with the dedifferentiating agent 5-bromodeoxyuridine. The major type II collagen mRNA species is 5300 bases long, similar to that of other interstitial collagen RNAs.

Activation of chitin synthetase in permeabilized cells of a Saccharomyces cerevisiae mutant lacking proteinase B

Digitonin treatment at 30 degrees C of a Saccharomyces cerevisiae mutant lacking proteinase B permeabilized the cells and caused rapid and extensive activation of chitin synthetase in situ. The same result was obtained with a mutant generally defective in vacuolar proteases. By lowering the temperature and using different permeabilization procedures, we showed that increases in permeability and activation are distinct processes. Activation was inhibited by the protease inhibitors antipain and leupeptin, but by pepstatin or chymostatin. Metal chelators were also inhibitory, and their effect was reversed by the addition of Ca2+ but not by Mg2+. Antipain added together with Ca2+ after incubation of the cells in the presence of a chelating agent prevented reversal of inhibition, a result that was interpreted as indicating that antipain acts either on the same step affected by Ca2+ or on a subsequent step. Efforts to obtain activation in cell-free extracts were unsuccessful, but it was possible to extract the synthetase, once activated, by breaking permeabilized cells with glass beads. Treatment of the cell-free extracts with trypsin led not only to increased activity of chitin synthetase, but also to a change in the pH-activity curve and a diminished requirement by the enzyme for free N-acetylglucosamine. These observations suggest that the modification undergone by the synthetase during endogenous activation is different from that brought about by trypsin treatment.

Serum copper concentration and hepatic enzyme induction during long-term therapy with anticonvulsants

We evaluated hepatic enzyme induction by measuring urinary D-glucaric acid and serum gamma-glutamyltransferase in a group of 40 adult epileptics of both sexes who were receiving long-term treatment with phenobarbital and (or) phenytoin. Total concentrations of copper and ceruloplasmin in their serum and the oxidase activity of ceruloplasmin were significantly greater than in the control group. However, non-ceruloplasmin copper and specific oxidase activity of the ceruloplasmin (activity per gram) were unchanged. A highly significant relationship was found between gamma-glutamyltransferase and (a) copper (r = 0.682, p less than 0.001), (b) ceruloplasmin (r = 0.523, p congruent to 0.001), and (c) the oxidase activity of the ceruloplasmin (r = 0.598, p less than 0.001). There is also a significant correlation of hemopexin with ceruloplasmin (r = 0.531, p congruent to 0.001) and the oxidase activity of the ceruloplasmin (r = 0.598, p less than 0.001). These results suggest that hypercupremia in patients undergoing long-term anticonvulsant therapy is a direct result of hepatic enzyme induction caused by the drugs that induce synthesis of ceruloplasmin.

Serum copper concentration and hepatic enzyme induction during long-term therapy with anticonvulsants

We evaluated hepatic enzyme induction by measuring urinary D-glucaric acid and serum gamma-glutamyltransferase in a group of 40 adult epileptics of both sexes who were receiving long-term treatment with phenobarbital and (or) phenytoin. Total concentrations of copper and ceruloplasmin in their serum and the oxidase activity of ceruloplasmin were significantly greater than in the control group. However, non-ceruloplasmin copper and specific oxidase activity of the ceruloplasmin (activity per gram) were unchanged. A highly significant relationship was found between gamma-glutamyltransferase and (a) copper (r = 0.682, p less than 0.001), (b) ceruloplasmin (r = 0.523, p congruent to 0.001), and (c) the oxidase activity of the ceruloplasmin (r = 0.598, p less than 0.001). There is also a significant correlation of hemopexin with ceruloplasmin (r = 0.531, p congruent to 0.001) and the oxidase activity of the ceruloplasmin (r = 0.598, p less than 0.001). These results suggest that hypercupremia in patients undergoing long-term anticonvulsant therapy is a direct result of hepatic enzyme induction caused by the drugs that induce synthesis of ceruloplasmin.