Anti-cytoskeletal autoantibodies: diagnostic significance for liver diseases, infections and systemic autoimmune diseases.
Explore the source record for details and available documents.
Biomedical subjects
Publications and source records attributed to B H Toh.
Explore the source record for details and available documents.
Aconitase has been purified from membranes prepared from both the human gastric carcinoma cell line Okajima and from porcine gastric mucosa by chromatography on concanavalin A-Sepharose and carboxymethyl-Sepharose, and preparative polyacrylamide gel electrophoresis. Automated Edman degradation of the intact proteins yielded no N-terminal amino acid sequence due, presumably, to N-terminal blockage. Sequence analysis of tryptic peptides derived from S-carboxymethyl porcine and human aconitases established the positions of 95 and 64 amino acid residues, respectively. The amino acid sequence data for porcine aconitase was in perfect agreement with the previously reported cDNA-deduced amino acid sequence [Zheng et al. (1990) J Biol Chem 265:2814-2821]. Comparison of the human amino acid sequence data with the cDNA-deduced amino acid sequence of porcine aconitase indicated that these two proteins have 95% amino acid sequence identity within the sequenced region.
1. An iron-binding glycoprotein has been purified to homogeneity from porcine gastric mucosa. 2. The molecular weight (80,000), amino acid composition, carbohydrate content, N-terminal amino acid sequence, tryptic map, stoichiometry of iron binding (2 mol/mol), visible absorption spectrum of the ferric complex and chromatographic behaviour of the gastric protein are all strikingly similar to the corresponding properties of porcine serum transferrin. 3. The quantity of the gastric protein (1.3 mg/g wet weight) present in the gastric mucosa suggests that it is not serum transferrin (plasma concentration 1.8 mg/ml) contaminating the tissue. 4. A role for transferrin in the uptake of dietary iron by the gastrointestinal tract is proposed.
Serum levels of alpha-lactalbumin were assayed using a monoclonal antibody specific for this breast specific molecule. Elevated levels were found in 87% (48/55) of sera from women in the third trimester of pregnancy (29.1 +/- 7.4 ng ml-1), from 64% (62/97) of patients with breast cancer (23.4 +/- 5.6 ng ml-1), and from 70% (56/80) of patients with gynaecological cancers (19.4 +/- 6.7 ng ml-1). These alpha-lactalbumin levels were significantly higher (P less than 0.001) than those for men and non-pregnant women (11.0 +/- 2.3 ng ml-1) and for patients with other, non-gynaecological cancers (13.4 +/- 3.6 ng ml-1). The alpha-lactalbumin levels were higher in patients with stage IV breast cancer than those with stage I-III breast cancer. The overall sensitivity and specificity of the radioimmunoassay were 72% and 75% respectively. These findings suggest that the assessment of serum levels of alpha-lactalbumin may be useful as a marker for monitoring breast cancer.
Autoantibodies in the sera of patients with pernicious anemia recognize, in addition to the alpha subunit of the gastric H+/(+)-ATPase, an abundant gastric microsomal glycoprotein of apparent Mr 60,000-90,000. Herein we have colocalized the glycoprotein and the alpha subunit of the gastric H+/K(+)-ATPase to the tubulovesicular membranes of the parietal cell by immunogold electron microscopy. Moreover, the glycoprotein and the alpha subunit were coimmunoprecipitated, and copurified by immunoaffinity chromatography, with an anti-glycoprotein monoclonal antibody. The pig glycoprotein was purified by chromatography on tomato lectin-Sepharose, and five tryptic peptides from the purified glycoprotein were partially sequenced. The complete amino acid sequence, deduced from the nucleotide sequence of overlapping cDNA clones, showed 33% similarity to the sequence of the beta subunit of the pig kidney Na+/K(+)-ATPase. We therefore propose that the 60- to 90-kDa glycoprotein autoantigen is the beta subunit of the gastric H+/K(+)-ATPase and that the alpha and beta subunits of the proton pump are major targets for autoimmunization in autoimmune gastritis.
The cytoplasmic tubulovesicular and canalicular membranes of gastric parietal cells are intimately involved in hydrochloric acid secretion. To characterise the glycoproteins of these membranes, we examined a panel of lectins for reactivity with parietal cells in paraffin sections of rat, dog and pig stomach. The poly-N-acetyllactosamine-specific lectin from Lycopersicon esculentum (tomato) and from Solanum tuberosum (potato), and the galactose-specific lectin Ricinus communis agglutinin (RCA120), showed strong cytoplasmic binding of parietal cells of all three species, with a pattern indicative of an intracellular membrane network. Binding to parietal cells was confirmed by double-labelling studies with parietal cell auto-antibodies from patients with autoimmune gastritis. Mucous cells and mucin also bound these lectins strongly. Other gastric cell types did not stain with either tomato or potato lectin, but stained weakly with RCA120. Electron-microscopic examination of lectin binding sites using biotinylated tomato lectin or RCA120 and streptavidin-gold, revealed specific binding to the luminal face of parietal cell tubulovesicular and canalicular membranes as well as the contents of mucous cell secretory granules. Tomato lectin and RCA120 reacted by lectin blotting with a major species of apparent molecular weight 60-90 X 10(3) Mr from rat, dog and pig gastric membranes. A tubulovesicular membrane fraction, enriched 10-fold for K(+)-dependent phosphatase activity, was also enriched three-fold for tomato lectin binding as assessed by a solid-phase lectin assay. The 60-90K (K = 10(3) Mr) component, in 125I-labelled detergent extracts of dog tubulovesicular membranes, bound to an affinity support of tomato lectin-Sepharose and was specifically eluted with N,N',N'-triacetylchitotriose. Digestion with N-glycanase collapsed the 60-90K component into a sharp 35K band. We conclude that: (1) a 60-90K membrane glycoprotein localised on the luminal face of tubulovesicles and canaliculi of parietal cells interacts strongly with tomato lectin and RCA120; and (2) the glycoprotein is composed of a 35K core protein glycosylated with N-glycans probably containing poly-N-acetyllactosamine sequences with terminal galactosyl residues. The properties of this 60-90K glycoprotein are identical to a major parietal cell autoantigen recognised by sera of patients with autoimmune gastritis.(ABSTRACT TRUNCATED AT 400 WORDS)
Thirty-four human sera containing parietal cell autoantibodies (PCA) specifically immunoprecipitated two antigens, with apparent molecular masses of 60-90 kDa and 100-120 kDa under nonreducing conditions and 60-90 kDa and 120-150 kDa under reducing conditions, from porcine gastric membrane extracts. A third antigen of 92 kDa was only observed in immunoprecipitates analyzed under reducing conditions. By immunoblotting, 24 of the 34 PCA-positive sera reacted with only the 60-90-kDa antigen, three reacted with a broad 60-120-kDa smear, one reacted only with a 92-kDa antigen and six did not react. Reactivity with the 60-90-kDa antigen was observed with gastric membranes from dog, pig, rat, and rabbit. Twenty PCA-negative sera did not react with these components by immunoprecipitation or immunoblotting. PCA reactivity with the 60-90-kDa antigen was abolished when the gastric membranes were (a) digested with Pronase, (b) reduced with 100 mM dithiothreitol, (c) treated with sodium periodate, or (d) digested with N-glycanase. The 60-90-kDa and 100-120-kDa components were insensitive to neuraminidase treatment. N-glycanase digestion of 125I-labeled antigens purified by immunoprecipitation and preparative sodium dodecyl sulfate-polyacrylamide gel electrophoresis collapsed the 60-90-kDa antigen to a sharp 34-kDa band; the 100-120-kDa component was unaffected. These observations suggest that (i) parietal cell antigens comprise three components of 60-90, 92, and 100-120 kDa; (ii) the epitopes differ in conformational sensitivity; (iii) the 60-90-kDa antigen is a conserved molecule comprising a 34-kDa core protein extensively glycosylated with N-linked oligosaccharides; (iv) sialic acid residues are not present in the 60-90- and 100-120-kDa molecules, and (v) the carbohydrate and protein moieties of the 60-90-kDa molecule are required for antibody binding.
The effect of incubation temperature on the background staining of Western blots with monoclonal antibodies to a human milk protein, alpha-lactalbumin (Mr 14,500), is presented. Human milk proteins were electrophoretically separated and transferred to nitrocellulose membranes which were then blocked with bovine serum albumin, "BLOTTO", casein, or Tween 20. They were subsequently incubated with mouse monoclonal antibody to human alpha-lactalbumin, biotinylated anti-mouse antibody, strepavidin-biotinylated horseradish peroxidase complexes and a substrate containing diaminobenzidine and nickel chloride. Reduction of incubation temperature from 37 degrees C to 22 degrees C and 4 degrees C was found to decrease the extent of non-specific background staining independent of the type of blocking reagent used. Good specific staining with minimal background was found using 0.1% Tween 20 in phosphate-buffered saline, pH 7.2, as blocking agent and incubation temperatures of 4 degrees C.
Human alpha-lactalbumin (alpha-LA), a milk protein normally produced by, and restricted to, functionally differentiated breast epithelial cells has been shown immunohistochemically to be a good marker for breast cancer. Antibodies with specificity only for human alpha-LA have not previously been reported. The present study documents the production and characterization of an IgG1 murine monoclonal antibody with specificity restricted only to human alpha-LA. This monoclonal antibody, designated ET-1, was purified by affinity chromatography using human alpha-LA. Specificity of ET-1 for human alpha-LA was established by enzyme linked immunosorbent assay (ELISA), absorption studies, immunoprecipitation and by immunoblotting. ET-1 binds to human alpha-LA with a KD of 7.41 x 10(-8) mol/l and does not cross-react with bovine alpha-LA. This unique reactivity of ET-1, which is not inherently shared by polyclonal antisera, should enable the antibody to be used for the development of a sensitive immunoassay for circulating human alpha-LA.
Autoantibodies to neurons and to the cytoskeleton were demonstrated in the serum of a patient with small cell carcinoma of the lung (SCCL) and paraneoplastic sensory neuropathy. The serum reacted by immunofluorescence with the nuclei of neurons, and with cytochalasin B-sensitive "stress fibres" of cultured cells. The serum also reacted by immunofluorescence with the nuclei of some cultured cell lines. Immunoblotting experiments with brain tissue, with SCCL, HeLa and other cultured cells showed that the serum reacted with 1-4 bands of 35-39 kDa apparent m.w. Two dimensional immunoblotting showed that these molecules had a neutral pI. Antibodies, affinity-purified by elution from the 35-39 kDa bands, gave staining of the nuclei of neurons and of cultured cells and immunoblotted the same 35-39 kDa antigens. These observations show that the anti-neuronal autoantibody reacts not only with neurons and with SCCL but also with some cultured cell lines. Molecular mimicry has been invoked as the basis for the reactivity of this autoantibody with exposed epitopes on SCCL and on neurons.
Studies of T lymphocyte subsets were carried out in a group of 50 scleroderma patients, of whom 46 were also HLA phenotyped. The total lymphocyte count was depressed in 22 patients, and CD4 (helper cells) numbers were normal. CD8 (suppressor-cytotoxic) cells were reduced in 27 patients, and the CD4/CD8 number ratio increased above normal in three additional patients, resulting in 30 patients being classified as CD8-deficient. In the 46 patients HLA phenotyped, DRw8 was significantly increased in the entire patient group, but when the patients were subdivided into CD8-deficient (n = 29) and CD8-normal (n = 17), the increase in DRw8 was confined to the CD8-deficient patients. B18 was also increased in patients with limited sclerosis, while DR4 and DRw53 were significantly decreased and DR5 significantly increased in patients with more extensive skin sclerosis. These findings suggest that scleroderma is a heterogeneous condition and that this heterogeneity is reflected in different HLA profiles in patients subtyped according to their clinical profile and subpopulations of T cells.
Five out of 51 sera (10%) from patients with progressive systemic sclerosis reacted by immunoblotting with tissue concanavalin A binding glycoproteins of 50 kD and 45 kD Mr, whereas only one out of 133 control sera gave the same reaction (P less than 0.001). The antigens were localized to the microsomal fractions of tissues and were eluted from concanavalin A affinity columns by the competing sugar alpha-D-methymannoside but not by lactose, fucose or N-acetylglucosamine. Serum reactivity with these antigens was seen with a variety of human and porcine tissues and with human, porcine, bovine and canine spleens. Immunoblotting with cultured human fibroblasts and with human lymphocytes showed reactivity with the 50-kD component (gp50) only. No reactivity was seen with bovine or human endothelial cells or with HeLa, Hep-2 or mouse 3T3 cells. The gp50 antigen in human fibroblasts and lymphocytes and in human spleen had identical isoelectric points by two-dimensional immunoblotting, suggesting that they are the same molecule. These observations suggest that circulating autoantibodies to a 50-kD glycoprotein of fibroblasts and lymphocytes are present in some patients with progressive systemic sclerosis. The microsomal localization of the molecule suggests that it may have a role in the pathogenesis of progressive systemic sclerosis.
The origins of autoantibodies remain largely unknown. Burnet suggested that self-reactive clones of B lymphocytes are deleted during ontogeny and that 'forbidden' clones of autoantibody-producing B lymphocytes arise as a consequence of somatic mutation of the variable regions of immunoglobulin genes. The alternative view, which is gaining increasing support, is that autoreactive B cells are present in all normal individuals but are held in check by control mechanisms, e.g. by the idiotypic network of Jerne. Autoantibodies segregate with clusters of autoimmune diseases and are useful diagnostic markers of these diseases. The autoantigens targeted by these autoantibodies are largely unknown. The precise molecular characteristics of these autoantigens are currently under active investigation by immunochemical and recombinant DNA methods. Precise definition of these autoepitopes should lead to sensitive immunoassays and studies directed at determining the pathogenicity of the autoantigens. The new knowledge gained in this way may lead to successful restoration of tolerance to self antigens.
BALB/c mice were immunized with canine gastric mucosal cells enriched to 70% for parietal cells, to produce monoclonal antibodies (MoAb). Three MoAb, FMM-4C5, FMM-4C9 and FMM-2B2, were obtained which reacted by indirect immunofluorescence with gastric parietal cells and kidney tubules, predominantly distal kidney tubules, with a pattern similar to that of the M2 autoantibodies of primary biliary cirrhosis (PBC). The antibodies also reacted with tissues from rabbit, rat, pig, human and with rod-shaped structures in acetone-fixed monolayer cultures of human fibroblasts and HEp 2 cells. FMM-4C9 and FMM-2B2 reacted with tissues from BALB/c mice but FMM-4C5 did not. Immunoblots of FMM-4C5 with mitochondrial fractions showed that the antibody recognized a 63 kD antigen from dog stomach, rat kidney and rat liver, and a 72 kD antigen from human placenta; mouse preparations were not reactive. The antigen co-migrated with that recognized by serum from cases of PBC and some cases of progressive systemic sclerosis. Absorption of the mitochondrial fraction with PBC sera removed reactivity by immunoblotting with the murine autoantibody and vice versa. Two dimension immunoblots showed that the murine and human antibodies recognized an identical series of paired 'spots'. FMM-4C5 also reacted by immunoblotting with a rat recombinant mitochondrial polypeptide which has disease-specific reactivity with PBC sera. Absorption with recombinant polypeptide removed anti-mitochondrial activity by immunoblotting and immunofluorescence. These observations suggest that the MoAb FMM-4C5 recognizes part of the same 72 kD molecule recognized by human PBC sera. The murine monoclonal antibodies should be useful probes for further studies of the structure, function and possible pathogenicity of the 72 kD autoantigen.
Rubella virus (RV)-host cell interactions were examined by indirect immunofluorescence staining using antibodies to viral products and cytoskeletal components as probes. The patterns of immunofluorescence observed with human convalescent sera indicated that in infected Vero cells RV-specified proteins were distributed throughout the rough endoplasmic reticulum with some possible accumulation in the region of the Golgi complex. Viral RNA synthesis, detected with anti-double stranded RNA, appeared to be confined to small, intensely stained foci irregularly distributed in the cytoplasm. When cells were infected at a higher multiplicity, these foci appeared to aggregate into linear arrays. Infection with RV had a profound effect on the organization of actin in both Vero and BHK 21 cells, as shown by anti-actin antibodies. Actin microfilaments were observed to disintegrate progressively into amorphous aggregates of apparently monomeric actin as the infection proceeded. Because of the role actin microfilaments may play in cell mitosis it is postulated that this effect may be related to the inhibition of cell division reported to be associated with the congenital rubella syndrome.
Four monoclonal antibodies reactive by immunofluorescence and by flow microfluorimetry with canine and porcine gastric parietal cell membranes were produced by fusion of mouse NS-1 myeloma cells with splenocytes from mice immunized with a population of canine gastric mucosal cells containing 60-70% parietal cells. One of these, an IgM antibody designated 2C1, reacted with the surface membranes of parietal cells by immunofluorescence, flow microfluorimetry, and immunogold electron microscopy; competed with 125I-labeled gastrin for binding to gastric cells; and inhibited by 56% maximal gastrin stimulation of [14C]aminopyrine uptake in parietal cells. The antibody immunoprecipitated 125I-labeled samples of a 78-kDa gastrin-binding protein purified from membrane extracts of porcine gastrin mucosa but did not recognize the same protein labeled covalently with 125I-labeled gastrin-(2-17)-hexadecapeptide. These observations suggest that the previously identified 78-kDa gastrin-binding protein is the gastrin receptor and that the antibody 2C1 is directed against the gastrin binding site of the gastrin receptor.
Sera from 82 patients with acute or chronic hepatitis and 40 chronic carriers of hepatitis B were examined by ELISA and immunoblotting for reactivity with the glycolytic enzyme aldolase. The results of the ELISA tests, expressed as a percentage of a positive control, were compared to those obtained with sera from 39 patients with rubella, 11 with cytomegalovirus infection and 74 healthy subjects. The ELISA reaction with sera, expressed as mean +/- standard deviation was, for 15 patients with hepatitis A, 58.3 +/- 20.5%; 15 with hepatitis B, 59.5 +/- 42.18; 23 with hepatitis non-A, non-B 51.1 +/- 34.4%; 11 with HBsAg positive chronic active hepatitis, 70.1 +/- 31.5%; and 17 with autoimmune chronic active hepatitis, 66.8 +/- 21.4%. All values were significantly (p less than 0.05-p - less than 0.001) higher than those obtained with sera from carriers of hepatitis B surface antigen, 25.6 +/- 27.2%; rubella, 21.1 +/- 20.0%; cytomegalovirus infection, 19.2 +/- 27.8%; or healthy subjects, 20.9 +/- 16.2%. In two randomly selected sera, reactivity with aldolase by ELISA was neutralized by absorption with the enzyme. Selected sera showing reactivity by ELISA reacted by immunoblotting with aldolase. The findings suggest that acute or chronic liver damage may provoke the production of autoantibodies to aldolase.
A murine monoclonal antibody HuPIA3, produced by immunization with human platelet membranes, reacted by radioimmunoassay with platelets, and inhibited ristocetin- and collagen-induced platelet aggregation and release of 14C-serotonin. The antibody also inhibited ristocetin-induced aggregation of washed, formaldehyde-fixed platelets by von Willebrand factor. On cultures of human and rodent fibroblasts, and on frozen sections of rabbit liver and rat kidney, the antibody gave a diffuse, homogenous immunofluorescence staining of cell nuclei which could be abolished by treatment with 0.1 M HC1 or 2 M NaCl and restored by reconstitution with histones, suggesting a reaction with nuclear histones. Absorption of the antibody with histones abolished nuclear staining and abrogated the inhibitory effect of the antibody on ristocetin- and collagen-induced platelet aggregation and 14C-serotonin release. Conversely, absorption with platelets removed antibody reactivity for platelets and for cell nuclei. In addition, the antibody reacted with H1 histones by radioimmunoassay, and immunoblotting studies showed that the antibody reacted with a protein of 199,000 daltons on platelets and with H1 histones (31,000 dalton and 32,000 dalton). These observations suggest that the antibody recognizes epitopes found on surface molecules of platelets as well as on H1 histones of cell nuclei.