Autoantibodies of sera from patients with primary biliary cirrhosis recognize the alpha subunit of the decarboxylase component of human branched-chain 2-oxo acid dehydrogenase complex.

BACKGROUND/AIMS: The major antigens for anti-mitochondrial autoantibodies in primary biliary cirrhosis (PBC) are the lipoyl-containing components of 2-oxo acid dehydrogenase complexes. Autoantibodies against the E1alpha subunit of the pyruvate dehydrogenase complex (PDH) also have been found, but those against the E1alpha subunit of branched-chain 2-oxo acid dehydrogenase complex (BCKADH) have not been detected. We investigated the occurrence of BCKADH-E1alpha-specific autoantibodies by employing the purified human antigen. METHODS: The reactivities of PBC sera against purified antigens were assessed by ELISA and by immunoblotting analysis. The specificity of immunoreactivity was confirmed by absorption tests and affinity-purified antibodies. RESULTS: Fourteen out of 27 PBC sera reacted with BCKADH-E1alpha, and these same sera also reacted with BCKADH-E2. No PBC sera reacted with BCKADH-E1beta. The reactivity of PBC sera with BCKADH-E1alpha was removed only when the sera were pre-absorbed with this antigen. However, reactivities to BCKADH-E2 and PDH-E1alpha were retained. Affinity-purified antibodies to BCKADH-E1alpha reacted with BCKADH-E1alpha, but not PDH-E1alpha. Thus, it was confirmed that anti-BCKADH-Elalpha did not cross-react with either BCKADH-E2 or PDH-E1alpha. CONCLUSIONS: BCKADH-E1alpha-specific autoantibodies were found in the sera of PBC patients. The antibodies seem to occur subsequent to the anti-BCKADH-E2 antibody production, supporting the concept of intermolecular determinant spreading.

cDNA cloning of the chicken branched-chain alpha-keto acid dehydrogenase complex. Chicken-specific residues of the acyltransferase affect the overall activity and the interaction with the dehydrogenase.

Branched-chain alpha-keto acid dehydrogenase complex is a macromolecule comprising three catalytic components: a dehydrogenase (E1) with alpha(2)beta(2) structure, an acyltransferase (E2) and a dihydrolipoamide dehydrogenase (E3). In the mammalian complex, the E2 component with 24 identical subunits forms a structural core, to which multiple copies of E1 and E3 bind noncovalently. We isolated cDNA clones encoding E1 alpha, E1 beta and E2 subunits from a chicken-liver cDNA library and performed nucleotide sequencing. Amino-acid sequences deduced from the nucleotide sequences revealed that chicken E1 alpha and E1 beta chains had substantially homologous sequences with the corresponding mammalian polypeptides, except for the N-terminus. Chicken E2 conserved three functional domains, a lipoyl-bearing domain, an E1/E3 binding domain and an inner-core domain, but contrasted strongly with mammalian E2 in respect of containing 11 additional residues in two interdomain linkers: nine sequential residues in one linker and two residues in the other. Replacement of many residues was also observed in the chicken linkers. When E2 activity for catalyzing the overall reaction was measured by activity reconstitution in combination with E1 and E3, chicken E2 was markedly less effective than mammalian E2. The capability of chicken E2 for binding E1 was also reduced when determined by the binding assay using sucrose density gradient centrifugation. Chicken E1 was functionally as well as structurally indistinguishable from mammalian E1. Thus the reduced catalytic activity of chicken E2 must arise from its reduced E1-binding capacity, which results from the characteristic structure of interdomain linkers in chicken E2.

CpG motif-containing DNA fragments from sera of patients with systemic lupus erythematosus proliferate mononuclear cells in vitro.

OBJECTIVE: To characterize DNA in sera of patients with systemic lupus erythematosus (SLE) in terms of size, guanine plus cytosine (G+C) content (by percentage), CpG dinucleotide (CpG) (percentage), and effects on mononuclear cells (MNC). METHODS: Nine DNA clones were sequenced. Oligodeoxynucleotides with the characteristic CpG motif (TTCGAA or PuPuCGPyPy) were examined for their proliferative effect on MNC by [3H]thymidine incorporation, expression of HLA-DR and intercellular adhesion molecule (ICAM)-1 on monocytes by flow cytometry, and mRNA levels encoding interleukin 12 (IL-12) and interferon-gamma (IFN-gamma) by semiquantitative reverse transcription polymerase chain reaction. RESULTS: The size of DNA clones ranged from 87 to 318 bp (mean +/- SD, 177+/-68) and enrichment in G+C and CpG ranged from 34.7 to 69.7% (48.1+/-10.7) and 0.63 to 12.8% (4.0+/-4.1), respectively. Three of 9 clones contained the characteristic CpG motif. Oligonucleotides proliferated MNC, and augmented HLA-DR and ICAM-1 expression in company with an increase of mRNA encoding IL-12 and IFN-gamma. CONCLUSION: Circulating CpG motif-containing DNA fragments in SLE increased mRNA encoding IL-12 and IFN-gamma, which in turn increased HLA-DR and ICAM-1 on monocytes, resulting in MNC proliferation. This mechanism could contribute to the pathogenesis of SLE.

VEGF-dependent signaling in retinal microvascular endothelial cells.

We examined the effect of vascular endothelial growth factor (VEGF) on intracellular signal transduction pathways using isolated bovine microvascular endothelial cells (BREC). When cell growth was determined by [3H]thymidine incorporation, it was significantly stimulated by VEGF stimulation. In situ hybridization results also demonstrated that c-fos expression was enhanced by the stimulation. Although BREC expressed Flt-1 and Flk-1 as VEGF receptors at similar levels, VEGF stimulation preferentially enhanced the activity of Flt-1 tyrosine kinase. This stimulation initiated an increase in the level of GTP-form Ras and the activation of mitogen activated protein kinase (MAPK). On the other hand, BREC expressed the Janus kinase (Jak) family members Jak1, Jak2, and Tyk2, and the signal transducers and activators of transcription (Stat) family members Stat1, Stat3, and Stat6. These molecules were tyrosine phosphorylated under culture conditions used, and the phosphorylation of Tyk2 and Stat6 was specifically enhanced by VEGF stimulation. These results demonstrate that, in addition to Ras/MAPK pathways, the Flt-1/Tyk2/Stat6 pathway is important in VEGF signaling in BREC. These signal transduction systems may regulate the growth of retinal endothelial cells.

Effect of growth factors on expression of integrin subtypes in microvascular endothelial cells isolated from bovine retinas.

We isolated microvascular endothelial cells from bovine retinas (BREC) by a colony isolation method, and investigated the effects of basic fibroblast growth factor (bFGF) and vascular endothelial growth factor (VEGF) on growth as well as integrin subtype expression of the cultured BREC. Either of the growth factors stimulated growth of the cultured BREC. In non-stimulated BREC, alpha v beta 3 integrin expression was predominant over the other integrin subtypes tested (alpha v beta 5, alpha 2 beta 1 and alpha 5 beta 1), but no enhancement of the alpha v beta 3 expression occurred by either growth factor stimulation. On the other hand, bFGF addition stimulated alpha v beta 5 and alpha 5 beta 1 expression while VEGF induced the alpha 2 beta 1 expression as well. These results suggest that external stimuli such as bFGF and VEGF alter integrin expression profiles of BREC and probably influence interactions of endothelial cells with extracellular matrix during angiogenesis which must be involved in pathogenesis of ischemic retinal disorders. In this respect, this BREC culture system could be useful to study the mechanism of neovascularization in these disorders.

Nucleotide sequence of the promoter region of chicken cytosolic phosphoenolpyruvate carboxykinase gene.

The nucleotide sequence of the promoter region and its flanking regions which span - 1855 to +2083 in the chicken cytosolic phosphoenolpyruvate carboxykinase gene was determined. The transcription initiation site was located at 119 nucleotides downstream of the previously reported chicken kidney transcription initiation site of this gene. The nucleotide sequences of exons 1, 2, and 3 were highly homologous to the corresponding exons of the rat gene. Homology of the sequence - 1 to - 500 to that of rat gene was 52% and most of the hormone-responsive sequences in rat gene, such as the glucocorticoid-responsive region, were not conserved in the chicken gene, in accord with the species-specific responsiveness to starvation. In contrast, in the region of - 1 to - 300, some sequence motifs conserved both in the chicken and rat genes were found at essentially the same positions in the promoters. Such sequence motifs included a cAMP-responsive element (CRE), a nuclear factor-1 (NF-1/CTF)-binding site, and a hepatocyte nuclear factor-1 (HNF-1)-binding site. Transient expression of the reporter luciferase gene ligated to the 3' end of this chicken sequence (-1855 to +7) was observed in a primary culture of chick hepatocytes when dibutyryl cyclic AMP was added to the culture medium.

Antibodies to E1 and E2/Protein X components of pyruvate dehydrogenase complex in sera of patients with primary biliary cirrhosis.

AIMS/METHODS: Using purified E1 component of pyruvate dehydrogenase complex (PDC) from bovine heart, we measured the levels of anti-E1 antibodies in PBC sera using ELISA and determined the degree of inhibition that these antibodies exerted on E1 enzyme activity. We also estimated levels of anti-E2/Protein X (Pro-X) antibodies in PBC sera using purified E2 and Pro-X of PDC which were copurified with E1. RESULTS/CONCLUSIONS: Anti-E1 antibodies were detected in 87.5% (35/40) of PBC sera. Some of these sera inhibited E1 enzyme activity but inhibition did not correlate with levels of anti-E1 antibodies. A high positive correlation (r = 0.918) was found between levels of anti-E1 and anti-E2/Pro-X antibodies, suggesting that anti-PDC antibody production was stimulated by PDC itself. Levels of IgG class anti-E2/Pro-X antibodies were significantly higher in sera of symptomatic PBC patients than in those of asymptomatic PBC patients. It was also found that patients who were positive for only IgM class anti-E2/Pro-X antibodies had early-stage PBC.

Comparison of kinetic property between human seminal and renal gamma-glutamyltransferase.

The initial-velocity kinetics, optimal pH, acceptor specificity and the influence of metal ions, EDTA and urea were studied on the human seminal gamma-glytamyltransferase (GGT) in comparison with the human renal GGT. The activity was measured with glycylglycine as an acceptor and with gamma-glutamyl-4-nitroanilide or gamma-glutamyl-3-carboxy-4-nitroanilide as a donor. Because the double-reciprocal plots showed paralled lines, the reaction of seminal GGT proceeds in nonsequence (Ping Pong Bi Bi) mechanism. The acceptor Michaelis constants for the seminal GGT were about 2 times higher than those for the renal enzyme with gamma-glutamyl-3-carboxy-4-nitroanilide as well as gamma-glutamyl-4-nitroanilide as donors, which the donor michaelis constants for seminal GGT were very similar to those for renal enzyme. The optimal pH and pK values were 8.2-8.6 and about 7.7, respectively. There was little difference in the specificity for various acceptors between the seminal and renal enzyme. Glycylglycylglycine was an effective acceptor other than glycylglycine, showing 80% of the activity with glycylgycine. Various substrates including metal ions tested were practically neither inhibitory nor stimulatory for seminal and renal GGTs.

Binding properties of the components of branched-chain 2-oxo acid dehydrogenase complex on ELISA.

Binding characteristics among three catalytic components of rat liver branched-chain 2-oxo acid dehydrogenase complex (BCKADH) were investigated by ELISA. Dihydrolipoamide dehydrogenase (E3) was bound to solid-phase dihydrolipoamide acyltransferase (E2). The binding curve was hyperbolic giving a calculated half-maximal binding concentration of 167 ng/ml for E3. Specificity of the binding of E3 to E2 was certified by a competition experiment measuring binding of biotin-labeled E3 in the presence of unlabeled E3. The decarboxylase component (E1), which is the other catalytic component of the complex, prevented the E3 binding to E2. The specific binding between E2 and E3 was verified in the opposite direction with immobilized E3. E2 also bound to solid-phase E1 in a specific manner, and addition of E3 prevented the E2 binding to E1. No binding between E1 and E3 was observed. Thus, E1 and E3 prevented each other's binding to E2, suggesting that E1 and E3 may recognize overlapped binding sites on the E2 polypeptide or that they may, at least in part, sterically interact with each other on their binding to E2. The reconstitution experiment of the complex also supported such a mutually exclusive binding of E1 and E3 to the E2 core.

Differential processing of human and rat E1 alpha precursors of the branched-chain alpha-keto acid dehydrogenase complex caused by an N-terminal proline in the rat sequence.

The N-terminal sequences of the E1 alpha, E1 beta and E2 subunits of the human branched-chain alpha-keto acid dehydrogenase complex have been determined by microsequencing. The N-terminal of human E1 beta and E2 subunits (Val and Gly, respectively) are identical to those of the corresponding rat and bovine subunits. However, the N-terminus of the human E1 alpha subunit (Ser) is identical to bovine, but differs from the rat E1 alpha (Phe) subunit. Comparison of the N-terminal sequences of human and rat E1 alpha subunits shows that the serine residue at the +1 position in the human sequence is replaced by a proline residue in the rat sequence. The presence of the proline residue apparently causes a 5'-shift by one residue in the cleavage site by the mitochondrial processing peptidase in the rat sequence, when compared to the human sequence. The results provide evidence that the mitochondrial processing peptidase cannot cleave an X-Pro bond, similar to trypsin, chymotrypsin and microsomal signal peptidases.

The source of the oxygen atom in the alpha-hydroxyglycine intermediate of the peptidylglycine alpha-amidating reaction.

Peptidylglycine alpha-amidating activity catalyses the oxidation of a C-terminally glycine-extended peptide to a desglycine alpha-amidated peptide at the expense of ascorbate and O2 in the presence of Cu2+. The reaction involves oxidative N-dealkylation within the terminal glycine residue, with retention of the glycine N atom and release of the remainder as glyoxylate. Recent studies by us and others have revealed that the reaction consists of two steps via a carbinolamide as an intermediate (peptidyl alpha-hydroxyglycine), and also that two separate enzymes derived from a common precursor protein catalyse these steps, formation of the carbinolamide and its conversion into alpha-amide and glyoxylate. As for the mechanism of carbinolamide formation, two distinct pathways can be considered: direct mono-oxygenation at the glycine alpha-C atom and dehydrogenation leading to an imine followed by hydration. To draw a distinction between them, we carried out the reaction with D-Tyr-Val-Gly as the substrate either in the H2(18)O-enriched medium or under an atmosphere of 18O2, and isolated the alpha-hydroxylglycine intermediate. The fast-atom-bombardment mass-spectral analysis demonstrated that the hydroxy O atom comes from O2, but not from H2O, indicating that the alpha-hydroxylation should be a monooxygenase reaction.

Both induction and activation of the branched-chain 2-oxo acid dehydrogenase complex in primary-cultured rat hepatocytes by clofibrate.

Clofibrate administration to rats caused both the activation and induction of the branched-chain 2-oxo acid dehydrogenase complex in the liver; the former phenomenon occurred within the first 6 h after clofibrate administration whereas the latter occurred after 12 h. Essentially the same results were obtained with primary cultures of rat hepatocytes in the presence of 0.5 mM clofibrate, though about three-fourths of the enzyme complex in control cells (without clofibrate addition) was inactivated during a culture for 44 h, with little reduction of the enzyme amount. This was also confirmed by immunotitration analysis with antibodies raised against the purified decarboxylase and transacylase components of the enzyme complex. On the other hand, the activity of dihydrolipoamide dehydrogenase (a constituent of the complex) was little affected by clofibrate administration. The half lives of the decarboxylase and transacylase components in the primary cultures were estimated to be in the range of 22-26 h, and were unchanged in the presence of clofibrate, when determined with the use of cycloheximide and by a pulse-chase experiment. On the contrary, the rates of synthesis of these two enzyme components had increased to about 1.9-fold after 32 h cultivation in the presence of clofibrate. Thus, the increase in the synthesis of both the components resulted in induction of the complex.

Purification and immunological characterization of a new form of gamma-glutamyltransferase of human semen.

A new form of gamma-glutamyltransferase was purified from human seminal plasma. The purified enzyme was composed of two non-identical subunits with apparent molecular masses of 150 and 95 kDa on polyacrylamide gel electrophoresis (PAGE) in the presence of sodium dodecyl sulfate (SDS), and showed a molecular mass of 500 and 250 kDa on gel filtration in the absence and presence of 1% Triton X-100, respectively. This enzyme was different from human renal gamma-glutamyltransferase not only in apparent molecular masses, but also in amino acid compositions of both the subunits to each other. Experiments with the antisera raised against the purified enzyme revealed that the enzyme was different from the renal, hepatic and testicular enzymes in reactivity to the antibody though partially related to those enzymes. Ouchterlony double diffusion analysis indicated that both human seminal plasma and prostatic extract contained two types of gamma-glutamyltransferase, one is that we purified and the other the renal type. Hence, it is most likely that gamma-glutamyltransferase accounting for most of the enzyme activity in semen results from prostata followed by secretion to seminal plasma.

Regulation of hormonal induction of chick liver cytosol-specific phosphoenolpyruvate carboxykinase.

Administration of glucagon, epinephrine, or dibutyryl cAMP to chicks induced cytosol-specific phosphoenolpyruvate carboxykinase in liver. In vitro translation assay with poly(A)+RNA indicated that this induction was due to the increase in phosphoenolpyruvate carboxykinase-coding mRNA synthesis which resulted from an increased level of hepatic cAMP. Either hydrocortisone or alpha-adrenergic agonist was ineffective for the induction by itself, but showed a significant effect when administered together with one of the inducing agents given above. In particular, hydrocortisone enhanced the synthesis of phosphoenolpyruvate carboxykinase-specific mRNA without changing the profile of the time courses of the induction and of hepatic cAMP level. Those observations suggest that the fundamental machinery required for induction of cytosol-specific phosphoenolpyruvate carboxykinase in liver is shared in common between rat and chick, and that the absence of appreciable induction of cytosol-specific hepatic phosphoenolpyruvate carboxykinase in starved chicks is due to neither lack nor impairment of the hormone-mediated induction mechanism, but is due to the difference in usage of the genetic information between the two animal species.

Regulation by induction of branched-chain 2-oxo acid dehydrogenase complex in clofibrate-fed rat liver.

The activity of branched-chain 2-oxo acid dehydrogenase complex increased 3.0-fold in liver of rats fed on 0.1%(w/w) clofibrate. Immunotitration experiments with antibodies against the constituent enzymes of the complex revealed that this increase resulted mainly from the increased amounts of only two(a decarboxylase and a lipoate acyltransferase) of three components of the complex and that the other component(dihydrolipoamide dehydrogenase) remained unchanged in its content, irrespective of clofibrate administration. The increases of both enzyme components were associated with increases in their mRNA levels which were estimated by in vitro translation with poly(A)+ RNA.

Control of the activity of branched-chain alpha-keto acid dehydrogenase complex in primary cultured rat hepatocytes.

When isolated rat hepatocytes were cultured in the medium containing 0.5 mM clofibrate (dissolved in ethanol) for 6 hours, the activity of branched-chain alpha-keto acid dehydrogenase (BCKADH) complex in the cells increased to more than two times that of control cells. Immunochemical determination with an anti-BCKADH IgG revealed that the activity increase occurred without increase in the enzyme amount. On the contrary, BCKADH activity in control cells (cultured in the presence 0.4% ethanol) decreased to two-third the initial activity at the end of 6-hour incubation whereas the activities of other enzymes tested were practically unaltered either in control or in clofibrate-treated cells. When the control cell extract was incubated with rat liver protein phosphatase, the BCKADH activity increased near to that of clofibrate-treated cells, but the same treatment of clofibrate-treated cell extract produced practically no increase in the activity. These observations indicate that clofibrate addition brought about the increase of BCKADH activity through activation-inactivation mechanism based on dephosphorylation-phosphorylation, and that this mechanism might function on liver BCKADH complex in vivo as a short term control of the activity in some conditions.

Cell surface heparan sulphate and adhesive property of sublines of rat ascites hepatoma AH7974.

Two variants (74AD and 74FL) established from rat ascites hepatoma AH7974 were examined for the production of glycosaminoglycans in culture. There was no difference between the adhesive (74AD) and the floating (74FL) variants in quantity of glycosaminoglycans produced by their cultivation in minimum essential medium supplemented with 10% foetal calf serum. However, they were distinctly different in the distribution patterns of heparan sulphate. In 74FL, about 70% of total heparan sulphate was found in the culture medium in soluble form, whereas in 74AD, only 7% was found in the medium and the rest was in the cell-substratum complex. In a serum-free medium, 74AD cells grew without adhering to the substratum. After cultivation, more than 90% of total heparan sulphate was found in the cell-associated fractions and the rest in the substratum fractions. No heparan sulphate was detected in the culture medium. On the other hand, 74FL cells released heparan sulphate to the serum-free medium as much as to the serum-containing medium. The increase in amount of heparan sulphate in the culture medium of 74FL cells was supposed to be caused by failure of the cells to deposit heparan sulphate at the cell surface and not caused by increased production. Cell-substratum adhesion mechanisms involving cell surface heparan sulphate (heparan sulphate proteoglycan) and some serum intermediate(s) are discussed for 74AD cells.

Purification and characterization of human liver branched-chain alpha-keto acid dehydrogenase complex.

Human liver BCKADH complex was purified. On SDS-polyacrylamide gel electrophoresis, the purified enzyme complex gave three major bands having molecular weights of 51,000, 46,000, and 36,000, and one minor band with a molecular weight of 55,000. The minor band corresponded in molecular weight to lipoamide oxidoreductase which was purified separately. The purified BCKADH represented only approximately 20% of the maximum activity when assayed without addition of exogenous lipoamide oxidoreductase, indicating that lipoamide oxidoreductase component was readily dissociable from the complex. The BCKADH effectively oxidized all of KIV, KIC, and KMV, yielding apparent Km values in the range of 14-17 microM for those alpha-keto acids. Vmax values obtained were 0.86, 0.61, and 0.51 mumole NADH produced/min/mg of protein for KIV, KIC, and KMV, respectively, in the presence of excess amount of lipoamide oxidoreductase. This ratio of Vmax values was practically identical to those of specific activities obtained with respective branched-chain alpha-keto acids at each purification step. The enzyme complex also oxidized pyruvate and alpha-ketoglutarate to a lesser extent. Kinetic experiments gave Km values of 0.98 and 2.9 mM for pyruvate and alpha-ketoglutarate, respectively, with Vmax of 0.43 and 0.08 mumole NADH produced/min/mg of protein. NAD and CoASH were absolutely required for the reaction. Km values for NAD and CoASH were estimated to be 47 and 25 microM, respectively.

Purification, resolution, and reconstitution of rat liver branched-chain alpha-keto acid dehydrogenase complex.

Branched-chain alpha-keto acid dehydrogenase (BCKADH) was solubilized as an enzyme complex from rat liver mitochondria by sonic treatment. Dehydrogenase (E1) and dihydrolipoyltransacylase (E2) components of the complex were purified in an associated form and resolved into individual components in the presence of 1 M NaCl, while lipoamide dehydrogenase (E3) component was dissociated from the complex during purification. Analysis by gel electrophoresis in dodecyl sulfate revealed the E1 comprised two different subunits with apparent molecular weights of 36,000 and 45,500, presumably in an equal molar ratio, while E2 consisted of a single subunit with an apparent molecular weight of 51,000. The BCKADH complex was reconstituted by combining E1, E2, and E3, and the formation of the complex was confirmed by analysis by sucrose density gradient centrifugation. The reconstituted enzyme complex oxidized not only alpha-ketoisovalerate (KIV), alpha-ketoisocaproate (KIC), and alpha-keto-beta-methylvalerate (KMV), but also pyruvate and alpha-ketoglutarate. Apparent Km values were 10-12 microM for the branched-chain alpha-keto acids, 2.2 mM for pyruvate, and 2.5 mM for alpha-ketoglutarate.