3-hydroxy-3-methylglutaryl coenzyme A reductase is sterol-dependently cleaved by cathepsin L-type cysteine protease in the isolated endoplasmic reticulum.

We have recently shown that 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase, an endoplasmic reticulum (ER) membrane protein, is degraded in ER membranes prepared from sterol pretreated cells and that such degradation is catalyzed by a cysteine protease within the reductase membrane domain. The use of various protease inhibitors suggested that degradation of HMG-CoA reductase in vitro is catalyzed by a cathepsin L-type cysteine protease. Purified ER contains E-64-sensitive cathepsin L activity whose inhibitor sensitivity was well matched to that of HMG-CoA reductase degradation in vitro. CLIK-148 (cathepsin L inhibitor) inhibited degradation of HMG-CoA reductase in vitro. Purified cathepsin L also efficiently cleaved HMG-CoA reductase in isolated ER preparations. To determine whether a cathepsin L-type cysteine protease is involved in sterol-regulated degradation of HMG-CoA reductase in vivo, we examined the effect of E-64d, a membrane-permeable cysteine protease inhibitor, in living cells. While lactacystin, a proteasome-specific inhibitor, inhibited sterol-dependent degradation of HMG-CoA reductase, E-64d failed to do so. In contrast, degradation of HMG-CoA reductase in sonicated cells was inhibited by E-64d, CLIK-148, and leupeptin but not by lactacystin. Our results indicate that HMG-CoA reductase is degraded by the proteasome under normal conditions in living cells and that it is cleaved by cathepsin L leaked from lysosomes during preparation of the ER, thus clarifying the apparently paradoxical in vivo and in vitro results. Cathepsin L-dependent proteolysis was observed to occur preferentially in sterol-pretreated cells, suggesting that sterol treatment results in conformational changes in HMG-CoA reductase that make it more susceptible to such cleavage.

Nondestructive quantification of neutral lipids by thin-layer chromatography and laser-fluorescent scanning: suitable methods for "lipidome" analysis.

A simple method for separation and quantification of neutral lipids was developed using thin-layer chromatography (TLC) and high-performance fluorescent scanning. Neutral lipid classes were separated using the double-developing TLC method and detected by rhodamine 6G and a laser-excited fluorescent scanner. The amount of lipids applied correlated with scanned intensity volume in a dose-dependent manner. The mass of each neutral lipid band was determined by comparing band intensities of unknown samples to dilution curves of authentic standards. After scanning the dye-sprayed TLC, acyl chain species of triglyceride (TG) extracted from TLC could be determined by gas chromatography. Using this method, we quantified the amounts of TG in mouse liver and found that the measured total mass of TG correlated with that obtained by enzymatic methods. Our method should provide the basic technique for "lipidome" analysis, designed to determine and compare total lipid classes and mass present in biological samples.

Accumulation and degradation in the endoplasmic reticulum of a truncated ER-60 devoid of C-terminal amino acid residues.

The accumulation and degradation in the endoplasmic reticulum (ER) of a truncated ER-60 protease, from which the C-terminal 89 amino acid residues have been deleted (K 417 ochre), was examined. K 417 ochre overexpressed in COS-1 cells is not secreted into the medium, but accumulates as insoluble aggregates in non-ionic detergent without degradation in unusual clump membrane structures. K 417 ochre, stably expressed, forms soluble aggregates in non-ionic detergent and is distributed in the reticular structures of ER. Under these conditions, K 417 ochre is not secreted into the medium but is degraded with a half-life time of more than 8 h. Since K 417 ochre/C all S, in which all the Cys residues of K 417 ochre are replaced by Ser, also forms aggregates, an inter-disulfide bond appears unnecessary for aggregation. In both types of aggregates, Ig heavy chain binding protein, calnexin, glucose regulated protein 94, calreticulin, ERp72, and protein disulfide isomerase are scarcely found. Since degradation of the stably expressed K 417 ochre was not inhibited by lactacystin, leupeptin, NH(4)Cl, or cytocharasin B, but was inhibited by N-acetyl-leucyl-leucyl-norleucinal, the self-aggregated abnormal protein in the lumen of ER is assumed to be degraded by an unknown protease system other than proteasome, lysosome or autophagy.

Catalytic cysteine residues of ER-60 protease.

ER-60 protease contains two CGHC motifs that appear to include an active site cysteine residue(s). Its proteolytic activity was lost with a double mutation of the C-terminal cysteines of the two motifs to alanine, but not with a single mutation of the C-terminal cysteine of either of the motifs to alanine. This suggests that these C-terminal cysteines independently constitute the catalytic active site. A mutation of both histidine residues in the two CGHC motifs to serine did not abolish the proteolytic activity, suggesting these histidine residues in the CGHC motifs do not constitute the catalytic dyad of ER-60 protease.

Purification and characterization of diacylglycerol lipase from human platelets.

Diacylglycerol lipase (DGL) was solubilized from human platelet microsomes with heptyl-beta-D-thioglucoside, and purified to homogeneity on SDS-PAGE using a combination of chromatographic and electrophoretic methods. The molecular mass of the purified DGL was estimated to be 33 kDa. Its apparent pI was pH 6.0, as determined by Immobiline isoelectro-focusing. The enzymatic activity of the partially purified DGL was investigated in the presence of a variety of inhibitors and reagents, as well as its pH and calcium dependence. Thiol reagents such as p-chloromercurubenzoic acid (pCMB), N-ethylmaleimide (NEM), and HgCl2 inhibited the activity, while dithiothreitol (DTT) and reduced glutathione (GSH) enhanced it. In addition, the enzymatic activity was inhibited by two serine blockers, phenylmethylsulfonyl fluoride (PMSF) and diisopropyl fluorophosphate (DFP), and by a histidine modifying reagent, p-bromophenacyl bromide (pBPB). These results suggest that cysteine, serine and histidine residues are required for the enzymatic activity of DGL. DGL was optimally active in the pH range of 7-8 and its activity did not change significantly in the presence of various calcium concentrations, even in the presence of 2 mM EGTA. This indicates that DGL can hydrolyze substrates with a basal cytosolic free Ca2+ level in the physiological pH range. A DGL inhibitor, RHC-80267, inhibited DGL activity in a dose-dependent manner with an IC50 (the concentration required for 50% inhibition) of about 5 microM. Unexpectedly, several phospholipase A2 (PLA2) inhibitors were potent inhibitors of DGL activity (IC50<5 microM), suggesting that the catalytic mechanisms of DGL and PLA2 may be similar. Finally, we show that DGL activity was inhibited by 2-monoacylglycerols (2-MGs), the reaction products of this enzyme. Among the three 2-MGs tested (2-arachidonoyl glycerol, 2-stearoyl glycerol, and 2-oleoyl glycerol), 2-arachidonoyl glycerol was the most potent inhibitor.

Autodegradation of protein disulfide isomerase.

Protein disulfide isomerase (PDI) and its degradation products were found in HepG2, COS-1, and CHO-K1 cells. Whether or not the products were formed through autodegradation of PDI was examined, since PDI contains the CGHC motif, which is the active center of proteolytic activity in ER-60 protease. Commercial bovine PDI was autodegraded to produce a trimmed PDI. In addition, human recombinant PDI also had autodegradation activity. Mutant recombinant PDIs with CGHC motifs of which cysteine residues were replaced with serine or alanine residues were prepared. However, they were not autodegraded, suggesting the cysteine residues of motifs are necessary for autodegradation.

Functions of characteristic Cys-Gly-His-Cys (CGHC) and Gln-Glu-Asp-Leu (QEDL) motifs of microsomal ER-60 protease.

The human ER-60 protease cDNA was expressed in Escherichia coli BL21 (DE3) cells using the pET-20b(+) T7 promoter. The recombinant ER-60 protease was obtained in a water-soluble form and purified through four sequential chromatographies. The ER-60 protease contains two CGHC motifs. When an alanine residue was substituted for the N-terminal cysteine residue in both motifs, the protease activity was not lost. However, when the C-terminal cysteine residue in both motifs was replaced by a serine residue, the cysteine protease activity, which was inhibited by p-chloromercuribenzoic acid (pCMB) but not by diisopropyl fluorophosphate (DFP), changed to serine protease activity, which was inhibited by DFP but not by pCMB. These results indicate that the C-terminal cysteine residue(s) of the CGHC motifs may constitute the active site(s) of ER-60 protease. The ER-60 protease has a C-terminal QEDL sequence, which was proved to serve as an ER-retention signal by deletion of the QEDL sequence. However, because QEDL could not serve as the ER-retention signal for protein disulfide isomerase or ERp72, it is suggested that amino acid residue(s) of ER-60 protease, other than the QEDL sequence itself, is complimentarily responsible for the ER retention of this protein.

Apolipoprotein B is intracellularly associated with an ER-60 protease homologue in HepG2 cells.

Two ALLN (N-acetyl-leucyl-leucyl-norleucinal)-sensitive endoplasmic reticulum (ER)-localized proteases (ER-60 and ER-72) were recently purified from rat liver. We used an antibody to rat ER-60 to investigate the possible role of this protease in apolipoprotein B (apoB) degradation. First, immunoprecipitation and immunoblotting experiments with the anti-rat ER-60 antibody suggested that HepG2 cells contain a homologue of ER-60 with an approximate molecular mass of 58-60 kDa. The ER-60 homologue was mostly associated with the luminal contents of HepG2 microsomes. Evidence from co-immunoprecipitation and cross-linking experiments appear to suggest that the ER-60 homologue in HepG2 cells is associated with apoB intracellularly. A small pool of apoB was recovered when HepG2 lysates were subjected to immunoprecipitation with anti-rat ER-60 antibody followed by a second immunoprecipitation with anti-apoB antibody. Furthermore, cross-linking of permeabilized cells with dithiobis(succinimidylpropionate) further demonstrated association of apoB with the ER-60 homologue in HepG2 cells. Three polypeptides with molecular masses of 78, 66, and 50 kDa were consistently found to be associated with apoB as well as the 58-kDa ER-60 homologue. The 78-kDa protein associated with both apoB and ER-60 appeared to represent immunoglobulin heavy chain-binding protein (BiP) based on immunoprecipitation with a monoclonal antibody. Cross-linking and immunoblotting experiments suggested the association of the 78-kDa BiP with both the 58-kDa ER-60 homologue as well as the 550-kDa apoB. In summary, the data suggests that HepG2 cells contain a 58-kDa protein which is homologous to the rat liver ER-60 in size, antigenecity, and intracellular localization. The ER-60 homologue in HepG2 cells appears to be closely associated with apoB, as well as other proteins possibly representing ER chaperones such as BiP. We hypothesize that the ER-60 homologue may be involved in the degradation of apoB in the ER lumen of HepG2 cells.

Selective incorporation of polyunsaturated fatty acids into organelle phospholipids of animal cells.

The selective incorporation of linoleic (18:2(n-6)) and docosahexaenoic (22:6(n-3)) acids into phospholipids of mitochondria, endoplasmic reticulum, and plasma membrane was investigated by changing the ratio of 22:6(n-3) against 18:2(n-6) in a medium, in which Chinese hamster V79-R cells were grown. In those organelles, 18:2(n-6) and its elongation product (eicosadienoic acid) (20:2 (n-6)) were predominantly incorporated into phosphatidylcholine. However, 22:6(n-3) was incorporated more selectively into phosphatidylethanolamine than 18:2(n-6) and 20:2(n-6).

A possible role of ER-60 protease in the degradation of misfolded proteins in the endoplasmic reticulum.

Wild-type human lysozyme (hLZM) is secreted when expressed in mouse L cells, whereas misfolded mutant hLZMs are retained and eventually degraded in a pre-Golgi compartment (Omura, F., Otsu, M., Yoshimori, T., Tashiro, Y., and Kikuchi, M. (1992) Eur. J. Biochem. 210, 591-599). These misfolded mutant hLZMs are associated with protein disulfide isomerase (Otsu, M., Omura, F., Yoshimori, T., and Kikuchi, M. (1994) J. Biol. Chem. 269, 6874-6877). From the observation that this degradation is sensitive to cysteine protease inhibitors, such as N-acetyl-leucyl-leucyl-norleucinal and N-acetyl-leucyl-leucyl-methioninal, but not to the serine protease inhibitors, 1-chloro-3-tosylamido-7-amino-2-heptanone and (p-amidinophenyl)methanesulfonyl fluoride, it was suggested that some cysteine proteases are likely responsible for the degradation of abnormal proteins in the endoplasmic reticulum (ER). ER-60 protease (ER-60), an ER resident protein with cysteine protease activity (Urade, R., Nasu, M., Moriyama, T., Wada, K., and Kito, M. (1992) J. Biol. Chem. 267, 15152-15159), was found to associate with misfolded hLZMs, but not with the wild-type protein, in mouse L cells. Furthermore, denatured hLZM is degraded by ER-60 in vitro, whereas native hLZM is not. These results suggest that ER-60 could be a component of the proteolytic machinery for the degradation of misfolded mutant hLZMs in the ER.

Selective incorporation of n-3 and n-6 unsaturated fatty acids into animal cell phospholipids.

The selective incorporation of n-3 and n-6 unsaturated fatty acids into phospholipids was investigated by changing the ratio of n-3 unsaturated fatty acids against n-6 unsaturated fatty acids in the medium where Chinese hamster V79-R cells were grown. Unsaturated fatty acids with lower degrees of unsaturation were abundantly incorporated into phosphatidylcholine independently of n-3 and n-6. Unsaturated fatty acids with higher degrees of unsaturation were more predominantly incorporated into phosphatidylethanolamine. When the difference in the numbers of double bond between unsaturated fatty acids was more than two, the unsaturated fatty acids with a higher degree of unsaturation was more selectively incorporated into phosphatidylethanolamine than the unsaturated fatty acids with a lower degree of unsaturation. By the analysis of molecular species, n-3 and n-6 unsaturated fatty acids with higher degrees of unsaturation were incorporated competitively into phosphatidylethanolamine. Finally, docosahexaenoic acid was incorporated into diacyl phosphatidylethanolamine more selectively than the other unsaturated fatty acids.

The 1,10-phenanthroline micelles-copper(I) complex catalyzes protein degradation.

The 1,10-phenanthroline micelles-copper(I) coordinated complex, not the mono-dispersed bis 1,10-phenanthroline-copper(I) one, rapidly degraded proteins in the presence of beta-mercaptoethanol in the neutral-to-mild acidic region under aerobic conditions. The degradation products derived from alpha-casein were peptides and amino acids.

Protein degradation by ERp72 from rat and mouse liver endoplasmic reticulum.

The endoplasmic reticulum (ER) resident protein, ER60, is a member of the protein disulfide-isomerase family and contains two copies of the internal thioredoxin motif, CGHC. Previously, ER60 was identified as a cysteine protease and named ER-60 protease (Urade, R., Nasu, M., Moriyama, T., Wada, K., and Kito, M. (1992) J. Biol. Chem. 267, 15152-15159; Urade, R., and Kito, M. (1992) FEBS Lett. 312, 83-86). Here, ERp72, the other member of the protein disulfide-isomerase family containing three CGHC motifs, was isolated from ER of rat and mouse livers through four sequential chromatographies on DEAE-Toyopearl 650, AF-heparin Toyopearl 650M, and TSK gel G3000SW twice. The purified rat protein was found to be homogeneous on sodium dodecyl sulfate-polyacrylamide gel electrophoresis, not being contaminated by ER-60 protease, as judged on immunoblot analysis using an anti-ER-60 protease antibody. The partial amino acid sequence of rat ERp72 was 93% homologous to that of mouse ERp72. The purified rat ERp72 degraded other ER resident proteins such as protein disulfide-isomerase and calreticulin. The purified mouse ERp72 also degraded those proteins. Though rat ERp72 did not basically require Ca2+ for the reaction, the degradation of protein disulfide-isomerase was enhanced, but the degradation of calreticulin was inhibited in the presence of Ca2+. The proteolytic activity of rat ERp72 was inhibited by cysteine protease inhibitors. Its sensitivity to protease inhibitors was the same as that of ER-60 protease. In addition, the proteolytic activity of rat ERp72 was inhibited by acidic phospholipids, also similar to ER-60 protease. Therefore, we propose that ERp72 be named ER-72 protease.

Inhibition by acidic phospholipids of protein degradation by ER-60 protease, a novel cysteine protease, of endoplasmic reticulum.

A protein (ER60) with sequence similarity to phosphoinositide-specific phospholipase C-alpha purified from rat liver endoplasmic reticulum (ER) degraded ER resident proteins and is really a protease [(1992) J. Biol. Chem. 265, 15152-15159]. Therefore, ER60 is called ER-60 protease. We now show that negatively charged phospholipids, phosphatidylinositol, phosphatidylinositol 4,5-bisphosphate and phosphatidylserine inhibit ER protein degradation by ER-60 protease. Phosphatidylcholine and phosphatidylethanolamine show no effect on the activity of ER-60 protease. With the use of protease inhibitors, ER-60 protease is shown to be a novel cysteine protease distinct from those of the cytosol and lysosomes.

Protein degradation by the phosphoinositide-specific phospholipase C-alpha family from rat liver endoplasmic reticulum.

A 60-kDa protein homologous to phosphoinositide-specific phospholipase C-alpha was purified to apparent homogeneity on sodium dodecyl sulfate-polyacrylamide gel electrophoresis from the rough endoplasmic reticulum of rat liver through three sequential chromatographies on DEAE Toyopearl 650, AF-heparin Toyopearl 650M, and TSK gel G3000SW. The purified protein was monomeric, with an M(r) of 60,000. Eight types of protein were further separated from the 60-kDa protein and named ER60A-ER60H according to the order of their elution from a TSK gel DEAE-5PW column. They were essentially identical in terms of immunochemical properties and the NH2-terminal amino acid sequence. The partial amino acid sequence of ER60F showed homology to that of phosphoinositide-specific phospholipase C-alpha. ER60A-ER60H showed no phosphoinositide-specific phospholipase C activity. However, ER60A-ER60H catalyzed cleavage of themselves and the endoplasmic reticulum proteins protein disulfide-isomerase and calreticulin. Proteolytic degradation was inhibited by p-chloromercuribenzoate. These results indicate that ER60A-ER60H comprise a group of endoplasmic reticulum resident proteins and show thiol group-related proteolytic activity.

Cardiolipins from rats fed different dietary lipids affect bovine heart cytochrome c oxidase activity.

Different molecular species of cardiolipin were prepared from tissue of rats fed diets containing different oils. The effects of these different species of cardiolipins on bovine heart cytochrome c oxidase activity were examined using reconstituted vesicles. Cytochrome c oxidase was stimulated effectively by cardiolipin molecular species rich in linoleoyl/linoleoyl species.

Cardiolipin molecular species in rat heart mitochondria are sensitive to essential fatty acid-deficient dietary lipids.

Feeding various dietary lipids did not alter the mass of phospholipids in rat heart mitochondria, but the phospholipids' fatty acid compositions changed. Although the compositional changes in mitochondrial phosphatidylcholine and phosphatidylethanolamine were negligible, linoleic acid [cis 18:2(n-6)] of cardiolipin was replaced by other fatty acids from dietary lipids. An analysis of the molecular species showed an even greater extent of change than was observed by fatty acid composition analysis. The quantity of 18:2(n-6)/18:2(n-6) molecular species [18:2(n-6) in C-1 position/18:2(n-6) in C-2 position] in cardiolipin decreased when rats were fed lipids that were depleted of an essential fatty acid but not deficient in essential fatty aids (eicosatrienoic acid did not appear). The decrease in 18:2(n-6)/18:2(n-6) species in cardiolipin was most pronounced in rats fed sardine oil, in which the ratio of (n-3) to (n-6) polyunsaturated fatty acids was 0.2. The O2 consumption rate of rat heart mitochondria decreased as the quantity of 18:2 (n-6)/18:2(n-6) cardiolipin species decreased.