Biotin carboxylases in mitochondria and the cytosol from skeletal and cardiac muscle as detected by avidin binding

Biotin carboxylases in mammalian cells are regulatory enzymes in lipogenesis and gluconeogenesis. In this study, endogenous biotin in skeletal and cardiac muscle was detected using avidin conjugated with alkaline phosphatase and applied in high concentrations to muscle sections. The avidin binding was subsequently visualized by histochemical demonstration of the alkaline phosphatase activity. All cardiac muscle cells showed high affinity for avidin with only the nuclei and the intercalated discs remaining unstained. In skeletal muscle a diffuse reaction could be detected in the sarcoplasm of the muscle fibres. A granular reaction was noted in the same fibres that showed activity for succinic dehydrogenase. The specificity of the coloured reaction product in the muscle sections was investigated and is suggested to be caused by avidin binding to biotin moieties in mitochondria and the cytosol. Mitochondrial and cytosolic preparations of skeletal muscle were electrophoresed in sodium dodecyl sulphate gels. After blotting and incubation with conjugated avidin, two bands with molecular weights of 75 kDa and 130 kDa respectively were evident in the mitochondrial preparation. It is suggested that the 75-kDa band represents comigration of the biotin-containing subunits of propionyl-CoA carboxylase and methylcrotonyl-CoA carboxylase. The 130-kDa band may represent the biotin-containing pyruvate carboxylase. In the cytosolic preparation a 270-kDa band was stained in blots that had been incubated with conjugated avidin; this band is suggested to represent acetyl-CoA carboxylase. A 190-kDa cytosolic band might be a cleavage product of acetyl-CoA carboxylase. We propose that using alkaline phosphatase-conjugated avidin it is possible to detect the mitochondrial and cytosolic biotin-dependent carboxylases in striated muscle.     

Multiple biotin-containing proteins in 3T3-L1 cells.

Extracts of 3T3-L1 cells prepared after labelling the monolayer cultures with [3H]biotin contained numerous protein bands that were detected by fluorography of dried SDS/polyacrylamide electrophoresis gels. All labelled proteins in the extracts could be removed by avidin affinity chromatography. The biotin-containing subunits of acetyl-CoA carboxylase, pyruvate carboxylase, methylcrotonyl-CoA carboxylase and propionyl-CoA carboxylase, with molecular masses of approx. 220, 120, 75 and 72 kDa respectively, were detected together with minor bands at 100, 85 and 37 kDa that did not appear to be partial degradation products. Additional labelled bands increased in amount during incubation of cell extracts or did not occur in extracts prepared with trichloroacetic acid, 9.5 M-urea or proteolytic inhibitors, and were tentatively classified as partial degradation products. The unknown bands were not removed by incubation of cell monolayers for 24 h, a treatment that gave degradation rate constants of 0.47 day-1 for acetyl-CoA carboxylase and 0.28 day-1 for pyruvate carboxylase. Upon two-dimensional electrophoresis, pyruvate carboxylase, methylcrotonyl-CoA carboxylase and propionyl-CoA carboxylase had isoelectric points of 6.4, 7.2 and 6.4 respectively. Several additional discrete spots with isoelectric points below 6.2 were also present. All the unknown biotin-containing proteins banded with intact mitochondria during density-gradient centrifugation. We conclude that several unknown biotin-containing proteins are present in the mitochondria of 3T3-L1 cells, whereas others are partial breakdown products of mitochondrial proteolysis.     

Localization of Endogenous Biotin-Containing Proteins in Mouse Bergmann Glial Cells

A peroxidase-conjugated avidin–biotin complex was used to detect endogenous biotin-containing proteins in mouse cerebellum. By this method, Bergmann glial cells were found to be strongly labelled in the adult mouse cerebellum. Developmentally, cells in the granular layer, probably astrocytes, appeared to be labelled around postnatal 10-day (P10). Their labelling decreased after P20, although the positive-labelling remained in the Bergmann glial cells up to the adult stage. The findings were confirmed by using a Alexa Fluor 488-conjugated streptavidin technique. The labelling was not affected by routine hydrogen peroxide treatment, but it was eliminated by avidin–biotin blocking. By another transblot method, the reactive proteins in the mouse cerebellum were found to be 120?kDa (the strongest one) and 75?kDa. For electron microscopy, a gold-conjugated anti-biotin antibody was immunoreacted to the mitochondria of Bergmann glial cells. These results suggest that endogenous biotin-containing proteins are abundant in the Bergmann glial cells. Therefore, the avidin–biotin complex method is useful for detecting Bergmann glial cells, probably because of the difference of biotin metabolism in the cerebellar glial cells.     

The Bacillus thuringiensis insecticidal toxin binds biotin-containing proteins.

Brush border membrane vesicles from larvae of the tobacco hornworm, Manduca sexta, contain protein bands of 85 and 120 kDa which react directly with streptavidin conjugated to alkaline phosphatase. The binding could be prevented either by including 10 microM biotin in the reaction mixture or by prior incubation of the brush border membrane vesicles with an activated 60- to 65-kDa toxin from Bacillus thuringiensis HD-73. The ability of B. thuringiensis toxins to recognize biotin-containing proteins was confirmed by their binding to pyruvate carboxylase, a biotin-containing enzyme, as well as to biotinylated ovalbumin and biotinylated bovine serum albumin but not to their nonbiotinylated counterparts. Activated HD-73 toxin also inhibited the enzymatic activity of pyruvate carboxylase. The biotin binding site is likely contained in domain III of the toxin. Two highly conserved regions within domain III are similar in sequence to the biotin binding sites of avidin, streptavidin, and a biotin-specific monoclonal antibody. In particular, block 4 of the B. thuringiensis toxin contains the YAS biotin-specific motif. On the basis of its N-terminal amino acid sequence, the 120-kDa biotin-containing protein is totally distinct from the 120-kDa aminopeptidase N reported to be a receptor for Cry1Ac toxin.     

Transblot and cytochemical identification of avidin-interacting proteins in mitochondria of cultured cells

Cell lysates prepared from 3T3-L1 cells were analyzed by western blotting using the avidin-biotin complex system and anti-Bax antibody. The antibody interacted with bands of proteins with estimated molecular weights of 120, 74, 72, and 25 kDa. However, only the 25-kDa band was detected with the anti-Bax antibody when the direct immunoblotting method was used. Peroxidase-conjugated avidin interacted with the 120-, 74-, and 72-kDa bands. This interaction was not limited to 3T3-L1 cells, because peroxidase-avidin also interacted with these three proteins in MC3T3-E1, YROS, Saos-2, MG63, SCCKN, and SCCTF cells although the staining intensity was different in each cell type. Avidin-peroxidase also interacted with these three proteins in the mitochondria-containing fractions prepared from 3T3-L1 cells. FITC-streptavidin was also localized in mitochondria in the cultured cells. The localization of avidin/streptavidin-interacting proteins in mitochondria was confirmed by using double staining with FITC-streptavidin and Mito-tracker.     

Endogenous biotinylated proteins in Dictyostelium discoideum.

Biotin-dependent enzymes are involved in carboxylation, decarboxylation and transcarboxylation reactions. Here, we have used sodium dodecyl sulfate polyacrylamide gel electrophoresis and electroblotting followed by probing with avidin to identify biotin-containing polypeptides in Dictyostelium discoideum. Twenty biotinyl polypeptides were visualized, with a 23 kDa protein appearing transiently. Based upon the molecular mobility of the biotinyl polypeptides, D. discoideum may contain the biotin-dependent enzymes acetyl CoA carboxylase, proprionyl CoA carboxylase, pyruvate carboxylase, and 3-methylcrotonyl CoA carboxylase.     

Author index for volume 202

A procedure is described for a simple two-step purification of human liver propionyl-CoA carboxylase. The method is based on acid and carbon tetrachloride extraction to remove other biotin carboxylases followed by an 800-fold purification through biotin-pretreated, monomeric avidin-Sepharose 4B-CL with elution of active enzyme using a biotin gradient. The enzyme had a sedimentation coefficient of 17.4 S and polyacrylamide gel electrophoresis after reduction and alkylation revealed two nonidentical polypeptide chains of 75,000 and 60,000 M_r. The heavier chain was identified as the biotin-containing subunit by electrophoresis after avidin binding.     

Identification of pyruvate carboxylase in 3T3-L1 cells by transblot and its correlation with differentiation into adipocytes.

The 3T3-L1 preadipocytes treated with insulin, dexamethasone and 3-methyl-1-isobutylxanthine (IBMX) two days before reaching monolayer undergo differentiation into adipocytes. Cell lysates were prepared from these cells under various conditions and analyzed by SDS-PAGE and transblot. Peroxidase-conjugated avidin used to detect endogenous proteins interacted strongly with a protein with an estimated molecular weight of 120 kDa, corresponding to pyruvate carboxylase, in the differentiated 3T3-L1 cells. On the other hand, this protein was not detected in undifferentiated 3T3-L1 cells.     

Formation of malonyl coenzyme A in rat heart. Identification and purification of an isozyme of A carboxylase from rat heart

Acetyl-CoA carboxylase is thought to be absent in the heart since the latter is highly catabolic and nonlipogenic. It has been suggested that the high level of malonyl-CoA that is found in the heart is derived from mitochondrial propionyl-CoA carboxylase, which also uses acetyl-CoA. In the present study, acetyl-CoA carboxylase was identified and purified from homogenates of rat heart. The isolated enzyme had little activity in the absence of citrate (specific activity, less than 0.1 units/mg); however, citrate stimulated its activity (specific activity, 1.8 units/mg in the presence of 10 mM citrate). Avidin inhibited greater than 95% of activity, and addition of biotin reversed this inhibition. Further, malonyl-CoA (1 mM) and palmitoyl-CoA (100 microM) inhibited greater than 90% of carboxylase activity. Similar to acetyl-CoA carboxylase of lipogenic tissues, the heart enzyme could be activated greater than 6-fold by preincubation with liver (acetyl-CoA carboxylase)-phosphatase 2. The activation was accompanied by a decrease in the K0.5 for citrate to 0.68 mM. These observations suggest that the activity in preparations from heart is due to authentic acetyl-CoA carboxylase. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the preparation from heart showed the presence of one major protein band (Mr 280,000) and a minor band (Mr 265,000) while that from liver gave a major protein band (Mr 265,000). A Western blot probed with avidin-peroxidase suggested that both the 280- and 265-kDa species contained biotin. Antibodies to liver acetyl-CoA carboxylase, which inhibited greater than 95% of liver carboxylase activity, inhibited only 35% of heart enzyme activity. In an immunoblot (using antibodies to liver enzyme) the 265-kDa species, and not the major 280-kDa species, in the heart preparation was specifically stained. These observations suggest the presence of two isoenzymes of acetyl-CoA carboxylase that are immunologically distinct, the 265-kDa species being predominant in the liver and the 280-kDa species being predominant in the heart.     

Immunoblotting of hydrophobic integral membrane proteins

For diagnosis and research purposes it is frequently desirable to measure by immunoblotting small amounts of proteins in complex mixtures such as tissue biopsy homogenates. Standard immunoblot procedures that give excellent results for soluble proteins unexpectedly gave low and irreproducible signals with some hydrophobic membrane proteins. We found that this was due to inefficient electrophoretic transfer to nitrocellulose, which could be corrected by modification of the transblot buffer. Hydrophobic integral membrane proteins of peroxisomes as well as other rat and human liver proteins were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose filters. The nitrocellulose-bound proteins were detected both by staining and by immunoblotting with an antiserum against the 22-kDa integral membrane protein of peroxisomes plus 125I-labeled protein A. A modified transblot buffer with 0.7 M glycine and 25 mM Tris (pH 7.7) but no methanol allowed use of a much shorter transfer time and strikingly improved the electrophoretic transfer of membrane proteins such that a peroxisomal integral membrane protein could be easily detected in human liver biopsy homogenates.     

3H]biotin-labeled proteins in cultured human skin fibroblasts from patients with pyruvate carboxylase deficiency

Biotin containing carboxylases in cultured human skin fibroblasts were radioactively labeled by addition of [8,9-3H]biotin to biotin-depleted cell cultures. Three major bands were visualized by fluorography after sodium dodecyl sulfate-polyacrylamide gel electrophoresis of the fibroblast proteins. These bands corresponded to pyruvate carboxylase (Mr = 125,000), the biotin-containing subunit of methyl crotonyl-CoA carboxylase (Mr = 75,000) and the biotin-containing subunit of propionyl-CoA carboxylase (Mr = 73,000) as judged by molecular weight markers, purified carboxylase protein standards, and interaction with monospecific antisera. Four out of 5 cell lines from patients with classical pyruvate carboxylase deficiency (less than 5% of normal activity) labeled with this technique displayed a normal band in the position of pyruvate carboxylase while one cell line showed complete absence of any labeled protein in this area. These results demonstrate heterogeneity in the etiology of pyruvate carboxylase deficiency.     

Characterization and purification of the hyaluronan-receptor on liver endothelial cells.

In order to characterize the proteins on liver endothelial cells that bind hyaluronan (HYA), liver endothelial cells were surface-iodinated with 125I, solubilized by Triton X-100 and passed through a column containing HYA coupled to agarose. The column was washed and eluted with HYA-oligosaccharides. Sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) of the eluted material, followed by autoradiography, showed a major band with a molecular mass of 100 kDa, that upon reduction gave major bands of 20 and 35 kDa, and minor doublet bands at 60 and 80 kDa. Two-dimensional electrophoresis of liver endothelial cell membrane proteins revealed that the 100 kDa protein has a pI of 6.6-6.8. The protein was purified by preparative SDS-PAGE of liver endothelial cell membrane proteins. The 100 kDa protein was excised from the gel and used for immunization of rabbits. Antiserum from immunized rabbits specifically recognized only the 100 kDa protein on immunoblots of liver endothelial cell membrane proteins separated by SDS-PAGE. The binding of 3H-HYA to liver endothelial cells and liver endothelial cell membranes could be specifically inhibited by Fab-fragments of the antibodies. When we tried to isolate the receptor in large scale by affinity chromatography of proteins from purified liver endothelial cell membranes, the 100 kDa protein could often not be detected on immunoblots or by silver staining following SDS-PAGE of the eluted material. Instead, proteins with molecular masses of 55 and 15 kDa were detected, but the antibodies reacted specifically with these proteins. Thus the 100 kDa protein is apparently susceptible to cleavage into distinct subcomponents.     

Localization of endogenous biotin-containing proteins in mouse Bergmann glial cells

A peroxidase-conjugated avidin-biotin complex was used to detect endogenous biotin-containing proteins in mouse cerebellum. By this method, Bergmann glial cells were found to be strongly labelled in the adult mouse cerebellum. Developmentally, cells in the granular layer, probably astrocytes, appeared to be labelled around postnatal 10-day (P10). Their labelling decreased after P20, although the positive-labelling remained in the Bergmann glial cells up to the adult stage. The findings were confirmed by using a Alexa Fluor 488-conjugated streptavidin technique. The labelling was not affected by routine hydrogen peroxide treatment, but it was eliminated by avidin-biotin blocking. By another transblot method, the reactive proteins in the mouse cerebellum were found to be 120 kDa (the strongest one) and 75 kDa. For electron microscopy, a gold-conjugated anti-biotin antibody was immunoreacted to the mitochondria of Bergmann glial cells. These results suggest that endogenous biotin-containing proteins are abundant in the Bergmann glial cells. Therefore, the avidin-biotin complex method is useful for detecting Bergmann glial cells, probably because of the difference of biotin metabolism in the cerebellar glial cells.     

Purification, characterization, and ontogeny of acetyl-CoA carboxylase isozyme of chick embryo brain.

Acetyl-CoA carboxylase catalyzes the first committed step in the synthesis of fatty acids. Because fatty acids are required during myelination in the developing brain, it was proposed that the level of acetyl-CoA carboxylase may be highest in embryonic brain. The presence of acetyl-CoA carboxylase activity was detected in chick embryo brain. Its activity varied with age, showing a peak in the 17-18-day-old embryo and decreasing thereafter. The enzyme, affinity-purified from 18-day-old chick embryo brain, appeared as a major protein band on polyacrylamide electrophoresis gels in the presence of sodium dodecyl sulfate (Mr 265,000), indistinguishable from the 265 kDa isozyme of liver acetyl-CoA carboxylase. It had significant activity (Sp act = 1.1 mumol/min per mg protein) in the absence of citrate. There was a maximum stimulation of only 25% in the presence of citrate. Dephosphorylation using [acetyl-CoA carboxylase] phosphatase 2 did not result in activation of the enzyme. Palmitoyl-CoA (0.1 mM) and malonyl-CoA (1 mM) inhibited the activity to 95% and 71%, respectively. Palmitoylcarnitine, however, did not show significant inhibition. The enzyme was inhibited (greater than 95%) by avidin; however, avidin did not show significant inhibition in the presence of excess biotin. The enzyme was also inhibited (greater than 90%) by antibodies against liver acetyl-CoA carboxylase. An immunoblot or avidin-blot detected only one protein band (Mr 265,000) in preparations from chick embryo brain or adult liver. These observations suggest that acetyl-CoA carboxylase is present in embryonic brain and that the enzyme appears to be similar to the 265 kDa isozyme of liver.     

Chloroplast envelope proteins are encoded by the chloroplast genome of Chlamydomonas reinhardtii.

To characterize envelope proteins encoded by the chloroplast genome, envelopes were isolated from Chlamydomonas reinhardtii cells labeled with [35S] sulfate while blocking synthesis by cytoplasmic ribosomes. One and two-dimensional gel electrophoresis of envelopes and fluorography revealed four highly labeled proteins. Two with masses of 29 and 30 kDa and pI 5.5 were absent from the stroma and thylakoid fractions, while the others at 54 kDa, pI 5.2 and 61 kDa, pI 5.4 were detected there in smaller amounts. The 29- and 30-kDa proteins were associated with outer envelope membranes separated from inner envelope membranes after chloroplast lysis in hypertonic solution. A 32-kDa protein not labeled by [35S]sulfate was found exclusively in the inner membrane fraction, suggesting the existence of a phosphate translocator in C. reinhardtii. To identify envelope proteins exposed on the chloroplast surface, isolated active chloroplasts were surface-labeled with 125I and lactoperoxidase. The 54-kDa, pI 5.2 protein as well as a protein corresponding to either of the 29- or 30-kDa proteins described above were among the labeled components. These results show that envelope proteins of C. reinhardtii are encoded by the chloroplast genome and two are located on the outer envelope membranes.     

Dietary dependent distribution of acetyl CoA carboxylase between cytoplasm and mitochondria of rat liver

Biotinyl proteins in cytoplasm and mitochondria of rat liver were examined by fluorography and the quantity of acetyl CoA carboxylase was determined after sodium dodecyl sulfate-denatured proteins were incubated with [14C] methyl avidin and separated by polyacrylamide gel electrophoresis. Results show that one-half of the total acetyl CoA carboxylase in liver of fed rats was associated with mitochondria in a relatively inactive form. Fasting shifted the distribution of the enzyme toward the mitochondrial fraction and refeeding previously fasted rats shifted the distribution towards cytoplasm. Thus, acetyl CoA carboxylase can be added to the list of ambiquitous enzymes whose subcellular distribution varies with physiological conditions.     

Superoxide dismutase, peroxidase, and germin-like protein activity in plasma membranes and apoplast of maize roots

Summary. The analysis of plasma membranes from maize roots by native gel electrophoresis revealed the existence of Mn-containing 120 kDa and CuZn-containing 70, 40, and 15 kDa superoxide dismutase (SOD) isoform activities. Isoelectric focusing of the plasma membranes differentiated anionic SOD isoforms with a pI of about 5 and cationic SOD isoforms at pI 8.6. Solubilization of the plasma membrane proteins further separated the cationic SOD into pI 8.6, 8.2, 8.4, and 7.2 isoforms. Double staining for both SOD and peroxidase activities showed an overlap of these activities only in the case of the high-molecular-mass (ca. 120 kDa) isoforms. High-temperature treatments demonstrated that the 120 kDa isoform was active even at 100 °C, indicating that it was a germin-like protein with superoxide-dismutating activity, different from the peroxidase with a similar molecular mass and the lower-molecular-mass CuZn-containing superoxide dismutases. These results are compared to those obtained from whole-tissue extract and apoplastic fluid. Correspondence and reprints: Maize Research Institute, POB 89-Zemun, 11081 Belgrade, Serbia and Montenegro.     

Capturing SDS-treated biotinylated protein and peptide by avidin functional affinity electrophoresis with or without SDS in the gel running buffer

Avidin functional affinity electrophoresis (AFAEP) is a new method of affinity electrophoresis. In this technique, bifunctional linker glutaraldehyde is added to the polyacrylamide gel solution to embed avidin within the gel matrix by interaction with the amino/amide groups. Samples are heated with triglycine sodium dodecyl sulfate (SDS) sample buffer to ensure that biotinylated proteins biotinylated peptides are negatively charged and migrate electrophoretically toward the cathode through the avidin zone regardless of their pI values. The AFAEP method allows the capture and concentration of biotinylated proteins or biotinylated peptides irrespective of the use of SDS in both the sample buffer and the gel running buffer.     

Determination of the quantity of acetyl CoA carboxylase by [14C]methyl avidin binding

Conditions are described under which monomeric [14C]methyl avidin binds to SDS-denatured biotin enzymes and remains bound through polyacrylamide gel electrophoresis. The location of radioactive proteins on the dried gel was determined by fluorography and their identity was established by subunit molecular weight. The relative quantity of bound radioactive avidin, stoichiometrically equivalent to the molar quantity of biotin protein, can be determined by scanning the fluorograph with a soft laser densitometer. To determine the absolute quantity of biotin protein, the radioactive areas of the dried gel were cut out, resolubilized, and assayed for radioactivity. Since the specific radioactivity of the [14C]methyl avidin was known, the quantity of avidin bound and therefore the quantity of biotin enzyme could be calculated. The method is illustrated by the analysis of purified acetyl CoA carboxylase and is applied to the analysis of biotin enzymes in isolated rat liver mitochondria.     

Heat-shock proteins and nucleolar hypertrophy in the liver of rat infused with methionine-free total parenteral nutrition

Infusing a methionine-free solution into rats for 7 days resulted in a marked enlargement of liver nucleoli. By the analysis of two-dimensional gel electrophoresis, a spot 'a' (76 kDa, pI 5.3) stained with Coomassie blue was observed to accumulate highly in liver cytosol from rat infused with methionine-free solution. Metabolically labeling experiments with [35S]methionine showed that 'a' was more heavily labeled in primary hepatocytes of rats infused with methionine-free solution than in those of control rat. To ascertain whether 'a' is one of stress proteins, primary hepatocyte cultures were incubated at 42 degrees C for 2 h. 'a' (76 kDa, pI 5.3) was slightly induced in control hepatocytes but not appreciably in hepatocytes from the treated rat. In contrast, two other spots 'b' (74 kDa, pI 5.6) and 'c' (74 kDa, pI 5.3) were highly induced at 42 degrees C in hepatocytes from control and treated rats. The antibody against the consensus sequence peptide of hsp70 family reacted with 'a' (76 kDa, pI 5.3) as well as 'b' and 'c'. Immunoblot analysis revealed that 'a' accumulates highly in hepatocytes of treated rats. These results indicate that infusion of methionine-free solution into rats induces one member of the hsp70 family in hepatocytes.