Proteolysis of N-ethylmaleimide-modified aldolase loaded into erythrocyte ghosts: prevention by inhibitors of calpain.

1. When rabbit muscle aldolase labelled with tritium and inactivated by N-ethylmaleimide (NEM) was loaded into erythrocyte ghosts, significant proteolysis of the loaded protein occurred. The major product of this proteolysis, separated by electrophoresis under dissociating conditions, was found to be approx. 2 kDa smaller than the parent protein. 2. Proteolysis was detectable during erythrocyte ghost loading at 0 degrees C, reaching a plateau after approx. 12 min. Subsequent incubation at 37 degrees C to allow resealing of the ghosts resulted in additional proteolysis, and up to 20% of the loaded protein was converted to the smaller 38 kDa derivative. 3. EDTA, EGTA, leupeptin and chymostatin, each inhibitors of calcium-activated neutral proteinases (calpains), were the most effective inhibitors of the proteolysis of NEM-inactivated aldolase in ghosts. Other proteinase inhibitors were ineffective, while phenylmethanesulphonyl fluoride was only partially effective. 4. Inhibition of the proteolysis by EGTA was prevented by CaCl2, supporting the involvement of erythrocyte calpain. 5. Pretreatment of ghosts with EGTA prior to loading of NEM-modified aldolase followed by microinjection of the protein into HeLa cells did not result in a different rate of its overall breakdown to acid-soluble products. EGTA is suggested as a useful agent for the erythrocyte ghost-mediated microinjection of calpain-sensitive proteins.     

Expression of thermotolerance following microinjection of glutathione disulfide

Asynchronous Chinese hamster ovary cells were microinjected with glutathione disulfide (GSSG). Successfully injected cells were scored by coinjecting FITC-dextran with GSSG, followed by fluorescent microscopy. After microinjection, cells were incubated for 2.5 h at 37 degrees C to permit thermotolerance development and then heated at 45 degrees C for 40 min. Cellular heat sensitivity was quantitated by counting the number of grains per cell after labeling heated cells with tritiated amino acids and processing for autoradiography. The data show that microinjection of GSSG induced thermotolerance which increased the number of grains per cell up to 500% of controls. Cells that were exposed to similar concentrations of GSSG in culture medium without microinjection or microinjected without GSSG did not develop thermotolerance.     

Mechanisms of intracellular protein catabolism. Intracellular fate of microinjected polypeptides translated in vitro.

Erythrocyte-mediated microinjection was used to introduce [35S]polypeptides translated in vitro into 3T3-L1 cells. Such [35S]polypeptides are not degraded after loading into erythrocytes and are stable for the first 2 h after microinjection into growing 3T3-L1 cells. Similarly, little or no degradation of microinjected [35S]polypeptides is observed in either growing or confluent 3T3-L1 cells over a 70 h period. Microinjection of reticulocyte lysate alone does not affect the rate of degradation of long-lived endogenous protein. Reductively [3H]methylated lysate haemoglobin is degraded after microinjection by a cytosolic mechanism. Microinjected 125I-labelled bovine serum albumin is rapidly degraded by a cytosolic mechanism at the same rate in the absence or presence of reticulocyte lysate. The data do not support the notion that the observed lack of degradation of microinjected [35S]polypeptides translated in vitro is due to the presence of proteolytic inhibitors in reticulocyte lysates which can inhibit the degradation of microinjected or cellular proteins.     

Microinjection of endogenous and exogenous proteins into primary cultures of rat hepatocytes and the degradation of the injected proteins

1. A method to microinject proteins into cells through packaging proteins to erythrocyte ghosts (erythrocyte-mediated microinjection) was modified partially in order to apply the method to primary cultures of rat hepatocytes. 2. Degradation of the microinjected proteins was examined employing the improved method. The mean half-life of the injected endogenous liver protein was 20 hr. The data suggested that the injected proteins are degraded through both lysosomal and non-lysosomal proteolytic pathways probably depending on their structure. 3. The present method to microinject exogenous proteins into primary cultures of rat hepatocytes can be employed usefully for the investigations of protein metabolism in liver.     

Degradation of proteins microinjected into IMR-90 human diploid fibroblasts

Erythrocyte ghosts loaded with 125I-labeled proteins were fused with confluent monolayers of IMR-90 fibroblasts using polyethylene glycol. Erythrocyte-mediated microinjection of 125I-proteins did not seriously perturb the metabolism of the recipient fibroblasts as assessed by measurements of rates of protein synthesis, rates of protein degradation, or rates of cellular growth after addition of fresh serum. A mixture of cytosolic proteins was degraded after microinjection according to expected characteristics established for catabolism of endogenous cytosolic proteins. Furthermore, withdrawal of serum, insulin, fibroblast growth factor, and dexamethasone from the culture medium increased the degradative rates of microinjected cytosolic proteins, and catabolism of long-lived proteins was preferentially enhanced with little or no effect on degradation of short-lived proteins. Six specific polypeptides were degraded after microinjection with markedly different half-lives ranging from 20 to 320 h. Degradative rates of certain purified proteins (but not others) were also increased in the absence of serum, insulin, fibroblast growth factor, and dexamethasone. The results suggest that erythrocyte- mediated microinjection is a valid approach for analysis of intracellular protein degradation. However, one potential limitation is that some microinjected proteins are structurally altered by the procedures required for labeling proteins to high specific radioactivities. Of the four purified proteins examined in this regard, only ribonuclease A consistently showed unaltered enzymatic activity and unaltered susceptibility to proteolytic attack in vitro after iodination.     

Analysis of BHK cell growth kinetics after microinjection of catalytic subunit of cyclic AMP-dependent protein kinase.

The effect of catalytic subunit (C) of cyclic AMP-dependent protein kinase on cell growth kinetics of BHK cells was assessed by microinjection with chicken erythrocyte ghosts as vehicles for introduction of the protein into the cytosol of large populations of cells. The advantage in using chicken erythrocytes for microinjection is that the inactive erythrocyte nuclei serve as a probe for identifying and analyzing microinjection events. By utilizing this procedure, BHK cells were microinjected with an amount of C that was 5- to 10-fold greater than their endogenous levels. Growth kinetics were analyzed by [3H]thymidine incorporation and autoradiography. Cells were stained after autoradiography to more clearly reveal the chicken nuclei, and at each time point, cells were categorized into four groups: (i) not microinjected, not in S phase, (ii) not microinjected, in S phase, (iii) microinjected, not in S phase, (iv) microinjected, in S phase. Those cells not microinjected served as internal controls. Two experimental protocols were used to test the notion that C is involved in blocking cell progression through G1 phase of the cell cycle. First, cells were arrested in G0 phase by serum deprivation, microinjected with C or control proteins, and stimulated to proceed to S phase by the addition of serum or purified growth factors. Second, cells were collected in mitosis, microinjected with C or control proteins, and stimulated to proceed to S phase by the addition of serum. The results of these studies indicate that a 5- to 10-fold increase in the intracellular concentration of C is not a sufficient signal to arrest cell growth in G1 phase. Thus, growth-inhibitory effects of cyclic AMP on BHK cells are unlikely to be the result of activation of cyclic AMP-dependent protein kinase.     

The synthesis and intracellular localization of adenovirus hexon protein studied by microinjection of mRNA into human cells

Polyadenylated RNA isolated 18 h after infection of HeLa cells with adenovirus type 2 was both translated in vitro and microinjected into the cytoplasm of human transformed amnion cells. The hexon polypeptide could be specifically immunoprecipitated from the products of cell-free translation with a rabbit-anti-hexon serum. The same serum when used in immunofluorescence assays of microinjected cells revealed hexon protein synthesized 6 h after microinjection. The intensity of the staining persisted up to 16 h after injection of messenger RNA (mRNA). Newly-synthesized hexon protein was characteristically located mainly in the nucleus. Essentially the same results were obtained when normal amnion cells were microinjected.     

Organization of enzymes of glycolysis and of glutathione metabolism in human red cell membranes.

1) The activities of 16 enzymes of glycolysis and of glutathione metabolism were determined in intact human red cell membranes (ghosts) which were prepared by hypotonic hemolysis. 2) Enzymes and hemoglobin of the ghosts were resolved by two toluene extractions. Only the four enzymes hexokinase, fructose-bisphosphate aldolase, glyceraldehyde-phosphate dehydrogenase and pyruvate kinase could not be released completely from the ghosts. 3) The residual membrane fraction, which was obtained after the toluene extraction of ghosts prepared at 30 imOsM, contained 0.02% of the original hemoglobin content of the red cell. Between 6.5 and 23% of the hemolysate activities of glyceraldehyde-phosphate dehydrogenase, phosphoglycerate kinase, pyruvate kinase and fructose-bisphosphate aldolase were detected in this fraction after mechanical disruption. 4) Sonication of intact ghosts increased the activities of fructose-bisphosphate aldolase, pyruvate kinase and phosphoglycerate kinase. 5) In "white" ghosts prepared at 5 imOsM phosphate buffer which contained 0.5% of the original hemoglobin the activities of fructose-bisphosphate aldolase and glyceraldehyde-phosphate dehydrogenase were detected at high levels. The activities of pyruvate kinase and phosphoglycerate kinase were low in these preparations. 6) The results indicate that one part of all enzymes is loosely attached to the inner surface of the membrane as is hemoglobin. A second part, the "cryptic enzyme activity", is available after resolving by toluene. A residual part of four enzymes is firmly bound to the membrane. Two of them (fructose-bisphosphate aldolase and glyceraldehyde-phosphate dehydrogenase) are oriented toward the inner surface of the membrane, whereas pyruvate kinase and phosphoglycerate kinase are hidden in the lipid core of the membrane.     

Red blood cell-mediated microinjection: methodological considerations

Red blood cell-mediated microinjection is a powerful approach to introducing proteins into the cytoplasm of cultured cells. In the course of our microinjection studies of intracellular protein degradation, we have encountered several potential problems with certain proteins. The microinjection procedure may be accompanied by denaturation of protein by radiolabeling procedures, binding of protein to red cell ghosts during loading, degradation of protein by the red cell ghost prior to microinjection, and adsorption of protein that leaks from red cell ghosts in the presence of fusogen to the fibroblast monolayer. We conclude with a list of points that must be considered prior to use of red cell-mediated microinjection to study a particular protein.     

Endocytosis and degradation of native, cathepsin D-degraded, and glutathione-inactivated aldolase by perfused rat liver

The uptake and degradation of 125I-labeled (a) native aldolase, (b) cathepsin D-inactivated aldolase, and (c) aldolase inactivated by oxidized glutathione were studied in perfused rat liver. All three forms of aldolase were removed from the perfusion medium and degraded by the liver, but the uptake of the glutathione-inactivated enzyme (half-life in perfusate = 10 min) was much faster than that of the native enzyme (half-life = 30 min) or the cathepsin-inactivated enzyme (half-life = 42 min). The degradation of the enzyme was almost totally inhibited by leupeptin, indicating that thiol proteinases in lysosomes play an important role in the digestion process. Degradation of native and cathepsin D-inactivated aldolase appeared to be slower than that of the glutathione-inactivated enzyme but studies in which liver was preloaded with aldolase by perfusion at 19 degrees C and then warming to 37 degrees C indicated that the rate of degradation of all three forms was similar. It is concluded that the liver is capable of distinguishing between the glutathione-altered aldolase and native or partially degraded aldolase with regard to endocytosis, but that all three forms are degraded at similar rates once within lysosomes.     

Degradation of erythrocyte-microinjected and scrape-loaded homologous cytosolic proteins by 3T3-L1 cells.

Homologous cytosol was introduced into 3T3-L1 cells by two different methods. Erythrocytes loaded with radiolabelled cytosolic proteins extracted from 3T3-L1 cells were fused with the aid of Sendai virus to 3T3-L1 cells, which were then seeded to confluent and non-confluent cultures. Cytosolic proteins were also introduced into cells by the technique of scrape-loading. In confluent cells, injected cytosolic proteins were recovered largely (54-93%) in a sedimentable (6 X 10(6) gav.-min) fraction from recipient cells irrespective of the method of introduction or of radiolabelling of the injected proteins [( 125I]iodination, reductive methylation with NaB3H4 and backbone labelling with L-[4,5-3H]leucine). The degradation of microinjected cytosolic proteins was in all cases inhibited by the lysosomotropic agent NH4Cl to a greater extent (32-75%) than that observed for endogenous cytosolic (less than or equal to 19%) proteins (labelled with L-[4,5-3H]leucine). In growing cells both endogenous total cell proteins and microinjected proteins were degraded at a slower rate than in confluent cell monolayers. The inhibition by NH4Cl of the degradation of both the endogenous and microinjected proteins is decreased compared with the inhibition observed in confluent monolayers. The results are discussed in terms of the cytoplasmic capacity to segregate microinjected homologous proteins before protein degradation can take place.     

Microinjection studies of protein transit across the nuclear envelope of human cells

Proteins of various molecular weights were conjugated with rhodamine and microinjected into the cytoplasm or nucleus of HeLa cells. The injected proteins were then localized within the cells at various times thereafter with fluorescence microscopy. Proteins below approx. 60 kD rapidly crossed the HeLa nuclear envelope. Some larger proteins also were able to pass into or out of the nucleus, while others were unable to do so, indicating the selective permeability of the HeLa nuclear envelope to large proteins. The nuclear protein HMG17 accumulated within the nucleus shortly after cytoplasmic microinjection.     

Modulation of the chymotrypsin-like activity of the 20S proteasome by intracellular redox status: effects of glutathione peroxidase-1 overexpression and antioxidant drugs

ATP- and ubiquitin-independent proteolysis by the 20S proteasome is responsible for the selective degradation of oxidized proteins. In vitro, the 20S proteasome shows an increased proteolytic activity toward oxidized polypeptides and the suc-LLVY-MCA peptide specific for its chymotrypsin-like activity. We have analyzed the effect of the intracellular redox status on the chymotrypsin-like activity of the 20S proteasome in human T47D cells overexpressing the detoxifiant enzyme seleno-glutathione peroxidase-1 (GPx-1). We report a 30% decreased activity of the chymotrypsin-like activity in cells overexpressing GPx-1. This phenomenon correlated with a 2-fold increase in IkappaB alpha half-life, a protein whose basal turnover is 20S proteasome-dependent. Following exposure to H2O2, these cells showed a seleno-dependently decreased accumulation of intracellular reactive oxygen species and 20S proteasome chymotrypsin-like activity. Similar results were obtained in HeLa cells transiently overexpressing human GPx-1. Moreover, exposure of HeLa cells to antioxidant compounds reduced the proteasome 20S chymotrypsin-like activity. In contrast, no effects were observed when HeLa cell extracts used to determine proteasome activity were incubated with either reduced or oxidized glutathione. These results suggest that GPx-1 activity or pro-reducing conditions can downregulate basal 20S proteasome activity. Hence, the intracellular redox status, probably through the level of oxidized proteins, is an important element that can either activate or down-regulate the 20S proteasome chymotrypsin-like activity in living cells.     

Specific inactivation of glutathione S-transferases in class Pi by SH-modifiers

Treatment of Class Pi glutathione S-transferases (GST) such as rat GST P (7-7), human GST pi and mouse GST MII with 0.05-0.1 mM N-ethylmaleimide (NEM) in 0.1 M Tris-HCl (pH 7.8) resulted in almost complete inactivation of these forms, whereas no or less inactivation occurred for GSTs in Class Alpha and Mu under the same conditions. Inactivated GST P lost its S-hexyl-GSH-Sepharose column affinity. About 0.8 mol of [14C]NEM was found to be covalently bound to 1 mol of GST P subunit when 80% of the activity was lost. Similar treatment with N-dimethyl-amino-3,5-dinitrophenyl maleimide, a colored analogue of NEM, followed by trypsin digestion, HPLC and amino acid sequence analysis revealed that one cysteine residue at the 47th position from the N-terminal of the GST P subunit was preferentially modified. Subunits of GST P and GST pi are known to have 4 cysteine residues at the same corresponding positions. The present results suggest that the 47th cysteine residue may be located in the vicinity of the active site of Class Pi GSTs.     

A putative protein-sequestration site involving intermediate filaments for protein degradation by autophagy. Studies with microinjected purified glycolytic enzymes in 3T3-L1 cells.

Several glycolytic enzymes (lactate dehydrogenase, pyruvate kinase, glyceraldehyde-3-phosphate dehydrogenase) were radiolabelled by [125I]iodination, conjugation with 125I-labelled Bolton & Hunter reagent and reductive [3H]methylation, and their degradative rates after microinjection into 3T3-L1 cells compared with that of the extracellular protein bovine serum albumin. Although the albumin remains largely cytosolic in recipient cells, the glycolytic enzymes rapidly (less than 30 min) become insoluble, as measured by detergent and salt extractions. The microinjected glycolytic enzymes appear to form disulphide-linked aggregates, are found in a cell fraction rich in vimentin-containing intermediate filaments and histones (nuclear-intermediate-filament fraction), and are degraded slowly by a lysosomal mechanism, as judged by the effects of inhibitors (NH4Cl, leupeptin, 3-methyladenine). 125I-labelled bovine serum albumin appears to be degraded rapidly and non-lysosomally. Prolonged treatment (96 h) of cultured cells with leupeptin results in the accumulation of pulse-labelled ([35S]methionine for 24 h) endogenous cell proteins in the detergent-and salt-non-extractable residue, but NH4Cl and 3-methyladenine do not have this effect. The findings are in terms of the interpretation of experiments involving microinjection of proteins to study intracellular protein protein degradation by autophagy.     

Ubiquitin metabolism in HeLa cells starved of amino acids.

Radio-iodinated ubiquitin (Ub) was introduced into HeLa cells by red blood cell-mediated microinjection. The half-life and solubility of Ub, as well as the molecular weight distributions of Ub conjugates, were then measured in HeLa cells grown in complete medium or in medium lacking amino acids and fetal calf serum. Ub metabolism was similar in the two sets of cells. Thus, the dramatic changes in Ub metabolism induced by thermal stress are not observed upon amino acid deprivation.     

The effect on butyryl-CoA dehydrogenase of reagents specific for nucleophilic sulphur.

The technique of erythrocyte-mediated microinjection has been successfully adapted for use with cultured muscle cells. Erythrocytes were fused with primary chick myotube cultures with poly(ethylene glycol), and fluorescent antibodies to haemoglobin demonstrated that this protein was injected into the sarcoplasm of myotubes. The microinjection treatment did not significantly alter protein metabolism in the muscle cells as monitored by rates of synthesis and degradation of muscle proteins. 125I-labelled ribonuclease A and bovine serum albumin were degraded with the expected exponential decay kinetics after microinjection into muscle cells, and the half-life of ribonuclease A (40 h) was approximately twice that of bovine serum albumin (17 h). The degradation of ribonuclease A in the muscle cells was enhanced 1.6-fold in the absence of horse serum and chick-embryo extract, whereas the degradation of bovine serum albumin was not altered during deprivation. These results are characteristic of the breakdown of microinjected ribonuclease A and bovine serum albumin in other cell types. Therefore, our experiments indicate the erythrocyte-mediated microinjection is a valid technique to study protein degradation in primary chick muscle cultures.     

Degradation of horseradish peroxidase after microinjection into mammalian cells

Horseradish peroxidase (HRP) has been microinjected into mammalian cells in tissue culture by the erythrocyte ghost-mediated technique. This protein was selected because it can be localized and quantified after injection by cytochemical and spectrophotometric methods. HRP labeled by reductive methylation retained full catalytic activity, was efficiently loaded into erythrocyte ghosts, and did not associate to a significant degree with ghost membranes. A combination of cytochemical staining and autoradiography established that HRP injected into rat L6 myoblasts, HE(39)L human diploid fibroblasts, or HeLa cells was intracellular and uniformly distributed throughout the cell, while cell lysis techniques showed that the catalytically active HRP was not membrane bound. Inactivation of labeled HRP after injection paralleled the disappearance of the 40-kDa polypeptide, and was always more rapid than its overall degradation. This difference was associated with a pool of water-insoluble radioactivity in the injected cells. This material was of smaller molecular size than the native protein: many labeled peptides were detected in the range of 10 to 38 kDa. By the use of inhibitors of autophagic proteolysis or lysosomal function it was established that HRP degradation was not subjected quantitatively to the same regulatory processes as the average endogenous protein labeled in the same cultures.     

Regulation of horse-liver glutathione reductase

• 1.1. The enzyme was rapidly inactivated by NAD(P)H, GSH, dithionite or borohydride, while activity increased in the presence of NAD(P)+ or GSSG. NADH was more efficient for inactivation than NADPH. Redox inactivation required neutral or alkaline pH, was maximal at pH 8.5, and depended on the presence of metal cations.
• 2.2. GSSG and dithiothreitol fully protected the enzyme from inactivation at concentrations stoichiometric with NAD(P)H. Ten-fold higher ferricyanide or GSH concentrations were required to obtain partial protection. NAD+ or NADP+ were quite ineffective.
• 3.3. GSSG fully reactivated the inactive enzyme at 38°C and neutral to acidic pH (5.5–7.5). Reactivation by dithiothreitol was accomplished in short periods of time at pH 8.5 although the activity was progressively lost afterwards. Ferricyanide and GSH also reactivated the enzyme to different extents.

     

Degradation of proteins microinjected into HeLa cells. The role of substrate flexibility

Increasing the flexibility of a protein enhances its susceptibility to defined proteases in vitro. To ascertain whether flexibility also affects protein stability in vivo, radioiodinated proteins with similar structures, but dissimilar flexibilities, were introduced into HeLa cells using red cell-mediated microinjection. Intracellular proteolysis was then measured as the rate of release of 125I-tyrosine into the medium. Ribonuclease A was considerably more resistant to degradation by purified proteases or in reticulocyte lysate than its flexible derivatives ribonuclease S and S-protein. In contrast, all three proteins were equally stable within HeLa cells. Like the results obtained for RNases, the rates of degradation of trypsin inhibitors, trypsin analogs, and their complexes correlated with flexibility in reticulocyte lysate. However, the intracellular half-lives of anhydrotrypsin and various proteinaceous trypsin inhibitors were not affected upon formation of enzyme-inhibitor complexes. Furthermore, trypsinogen was degraded more slowly than the structurally similar anhydrotrypsin in HeLa cells, although trypsinogen has additional segmental flexibility in its activation domain. Electrophoretic analyses revealed that trypsin-inhibitor complexes remained intact following injection into HeLa cells, and that neither free inhibitors nor anhydrotrypsin formed Triton-stable complexes with soluble cytoplasmic proteins. The observation that the components of the trypsin-inhibitor complexes were degraded simultaneously indicates that neither constituent unfolded prior to the onset of proteolysis. These studies provide evidence that RNases, trypsin, and trypsin inhibitors are degraded by an intracellular proteolytic pathway(s) which recognizes surface features of the folded proteins.