Comparison of a tissue transglutaminase ELISA with the endomysium antibody test in the diagnosis of gluten-sensitive enteropathy.

BACKGROUND: Tissue transglutaminase (tTG) has recently been found to be the major if not the only autoantigen of gluten-sensitive enteropathy (GSE). OBJECTIVES: To further determine the significance of this finding for diagnostic (screening) and follow-up purposes, we performed tTG-based ELISAs, and compared the results to the endomysium antibody test (EMA). PATIENTS: We examined 120 serum samples from patients with celiac disease (CD) including 72 on a gluten-free diet (GFD) and eleven on a gluten challenge, 47 with dermatitis herpetiformis (DH) including 16 on a GFD and one on a gluten challenge, 96 with non-CD gastrointestinal diseases, and 117 with others; i.e. 380 serum samples altogether. Follow-up sera from 13 patients were included. METHODS: Results of an ELISA with guinea pig liver tTG were compared with the EMA test using monkey esophagus. Inhibition of endomysial staining was performed with sera positive on the EMA test but negative with the guinea pig tTG ELISA. RESULTS: The specificity and sensitivity of the tTG ELISA are high (98.6% and 92.5%). The serum IgA antibody titers against tTG decrease after introduction of a GFD. In one case, endomysial staining could not be inhibited. CONCLUSIONS: Our results show that the guinea pig tTG ELISA is suitable for use as a simple diagnostic screening and follow-up method for GSE. Further studies are necessary to identify possible additional minor antigens in GSE.

Effect of interferon-alpha (IFN alpha) on various human tumor xenografts.

The growth of some human tumor xenografts (3 out of 8, melanoma, non-Hodgkin's lymphoma, squamous cell carcinoma) was successfully--but moderately and temporarily--inhibited, when interferon-alpha (EGIS, Hungary) was given for 10 days. The route of administration (intratumoral or intraperitoneal) was usually not a decisive factor. An attempt to potentiate IFNa action with Zymozan or Cyclophosphamide did not succeed.

Immunogenic properties of alpha (1----6) dextran, its protein conjugates, and conjugates of its breakdown products in mice.

Mice were immunized with alpha (1-6) dextran, either as such or coupled to protein carriers, and their anti-dextran response was measured by a solid-phase radioimmunoassay and the Farr assay. Like earlier investigators we found that protein-conjugated dextran was more antigenic than plain dextran. Our novel findings were that (1) a standard dose (30 micrograms of dextran per injection) coupled to strongly antigenic protein (chicken serum albumin (CSA) was three times more antigenic than dextran coupled to weakly antigenic bovine serum albumin (BSA); (2) dextrans of low molecular weight (1000-10,000 daltons) coupled to CSA induced at least ten times stronger secondary responses than did a similarly coupled macromolecular dextran (5-40 million daltons); (3) variation of the CHO/protein ratio from 0.3 to 1 had little effect on the antigenicity of the dextran. Increase of the ratio from one appeared to decrease immunogenicity when BSA was the carrier but not when CSA was the carrier.

Standard sera in solid-phase immunoassays.

Solid-phase immunoassay-derived antibody titers are often converted to weight unit concentrations with the aid of standard sera containing known antibody concentrations. Systematic studies justifying this procedure have not yet been published. We therefore investigated the magnitude of errors associated with this conversion. Antibody concentrations of thirteen sera or ascites fluids were determined by quantitative precipitation or equilibrium dialysis, and one was then used as a "standard antibody" for the others in solid-phase radioimmunoassay (SP-RIA) or enzyme-linked immunosorbent (ELISA) assays. Antibody concentrations determined by the conventional solid-phase assay (the "standard serum" has the same specificity as the "sample") had up to fourfold errors. These errors could be reduced by basing the conversion on the combination of two standard sera instead of one. The possibility was studied of whether the conversion to weight units could be done with the aid of a standard serum directed to a different antigen than the sample antibody. Errors associated with the use of such a heterologous standard were not significantly greater than those found using the conventional conversion. A combination of two reference sera again reduced the errors. The use of such heterologous standard(s), however, requires checking the binding capacity of the antigen coats.

Pokeweed mitogen, Bordetella pertussis and breast milk cell factor induce preferentially the synthesis of different immunoglobulin classes.

Human tonsillar lymphocytes were cultured in the presence of different activators and [14C]-isoleucine. De novo synthesized, [14C]-labeled immunoglobulin was determined after separation of the different classes by immunoadsorbants carrying class-specific anti-human IgA, IgG, IgM, IgD or IgE. Pokeweed mitogen and whole killed Bordetella pertussis enhanced the synthesis and secretion of IgA, IgG and IgM. Maximum stimulation was found with pokeweed mitogen in IgM secretion (up to 5-fold), while Bordetella pertussis had the largest impact on IgA and IgG (4-5 fold increase). The human milk cell factor (demonstrated by Pittard and Bill., Cell. Immunol. 1979, 42, 437.) in the supernatant of cultured milk cells stimulated selectively the synthesis of IgA (4-fold).

IgG subclass distributions in anti-hapten and anti-polysaccharide antibodies induced by haptenated polysaccharides.

Mice were immunized with hapten [NIP, (4-hydroxy-5-iodo-3-nitrophenyl)acetyl or TNP (2,4,6-trinitrophenyl)] conjugates of Ficoll or pneumococcal polysaccharide type 14 (S14), and they were bled on days 10 or 14. Anti-hapten and anti-polysaccharide antibodies were determined from the sera or from fractions (IgM + IgA). IgG1, IgG2a, IgG3 or IgG2b separated by a gradual acid elution from protein A. Approximately one-half of both anti-hapten and anti-polysaccharide antibodies was found in the IgM + IgA fraction. The subclass distribution of the IgG antibodies was dependent on the antigenic determinants. Polysaccharide antibodies were mostly in the IgG3 fraction (36-62%) and in the IgG1 fraction (18-36%). Hapten IgG antibodies were mostly in the IgG1 fraction (38-74%): each of the other three subclasses contributed the average of 13%. These results provide the first evidence that antibodies to different determinants of one antigen have grossly different isotype distributions.

Mouse IgG antibodies have subclass associated affinity differences.

Subclasses of IgG were separated from pools of mouse sera by letting immunoglobulins absorb on protein A-Sepharose and by eluting with buffers of decreasing pH. Most donor mice were immunized with a conjugate of a hapten (NIP) and chicken gamma globulin 20 days previously. The results indicate that concentrations of IgG varied from 5.1 to 8.6 mg/ml in the pools of immune sera and was 3.0 mg/ml in one normal serum tested. One half of this was IgG1, ca. 20% of IgG2a and IgG2b each, and 10% IgG3 in the pools of BALB/c sera. IgG2a and IgG3 could not be separated from C57BL sera (due to allotype b), but their combined share of IgG appears to be higher than in BALB/c. Immune sera contained 0.5-1.6 mg/ml of anti-NIP antibodies. Of this 90-98% was IgG1 and the remainder was split between the other subclasses. Up to one half of the protein in the IgG1 fraction was anti-NIP antibody. This surprising finding was confirmed by demonstrating that nearly 50% of the u.v.-light absorption was specifically removed by a NIP-immunosorbent. Subclass-associated affinity-differences were observed. IgG1 anti-NIP had a greater average affinity than IgG2a anti-NIP antibodies. The difference was ca. 1.5-fold when the equilibrium dialysis was focusing on the high-affinity bracket of the total population (concentration of free hapten 16-200 nM). At higher hapten concentrations the trend was the same but the data are fewer. Antibodies in subclasses IgG2b and IgG3 appear to share the lower affinity of IgG2a.

The four subclasses of IgG can be isolated from mouse serum by using Protein A-Sepharose.

We confirmed the findings of Ey and colleagues that mouse IgG is absorbed by protein A-Sepharose at pH 8.0. Also confirmed was their finding that IgG1 mainly elutes from such a column by means of a buffer with pH 6.0 and that the corresponding pH values for IgG2a and IgG2b were 4.5 and 3.5. We made the new finding that the bulk of IgG2a bearing allotypes a or j eluted already at pH 5, in contrast to IgG2a bearing allotype b. Another new finding was that IgG3 was mainly eluted at pH 4.5 regardless of the allotype. All four subclasses of IgG could thus be physically separated if the allotype was a or j (the only known exception is allotype b). Separation of IgG2a and IgG3 was achieved even when the allotype was b by using a pH gradient for elution. IgG2a came out at a slightly higher pH than IgG3. Mouse IgG antibodies against group A streptococcal polysaccharide belonged mostly to IgG3 and, to a lesser extent, to IgG2a and IgG2b.

Evidence for transformation of glucagon-like immunoreactivity of gut into pancreatic glucagon in vivo.

The effect of gut glucagon-like immunoreactivity (GLI) devoid of pancreatic glucagon was studied in piglets. All glucagon-like peptides with an accessible C-terminal were removed from the gut extract by specific antibodies reacting with the C-terminal of the glucagon molecule. Endogenous secretion of pancreatic and gut glucagon was blocked by somatostatin infusion, and then the purified gut glucagon preparation was infused. The latter prevented the hypoglycemia resulting from somatostatin infusion, and increased the glucagon level detectable by C-terminal specific antibodies in the blood of the animals. This rise was significant statistically from the 30th min of GLI administration (26.7 +/- 7.2 pg/ml versus 137.0 +/- 67.0 pg/ml; P less than 0.05) and increased until the end of the infusion (90th min, 218 +/- 60 pg/ml; P less than 0.005). It has been suggested that, owing to the in vivo enzymatic degradation of the infused gut glucagon, biologically active "pancreatic" glucagon fractions are formed extracellularly.

Studies of pre- and postimmunization antibodies of sheep.

Serum samples were taken serially from three nonimmunized sheep over a long period of time. Antibodies to human serum albumin (HSA), ovalbumin (OA) and FITC were separated from the samples. Than, two of the animals were injected with HSA+ complete Freund's adjuvant, the third with adjuvant without antigen. Serial postimmunization serum samples were subjected to the same procedures as the pre-immunization ones. The specific antibodies increased in concentration, and only the postimmunization antibody population was able to precipitate. In the presence of the antigen, the postimmunization antibodies bound to the Fc receptors of lymphocytes to an increased degree. There was no difference between pre- and postimmunizaton antibody populations either in complement-activating capacity or in the quantity of antigen necessary for reaching antigen-antibody equivalence. Isoelectro-focusing showed no new bands which would indicate antibodies different from the pre-existing ones. However, changes were observed in the relative participation of the antibodies forming different bands. No sharp limit was observed between pre- and postimmunization antibody populations. The Ig increment demonstrated after immunization was accompanied by a similar increment in the specific antibodies tested in the animal that had not given antigen, but not in the others. The authors attribute a role to humoral antibodies already in the earliest phase of immune response.

Application of pepsin-digested horse antibodies for quantitative immunoelectrophoresis.

The majority of the precipitating antibodies in hyperimmune horse serum belong to the beta globulins, as demonstrated by reversed immunoelectrophoresis. As these antibodies migrate in agarose gel during electrophoresis in conventional pH = 8.6-8.9 buffers, horse antiserum cannot be used satisfactorily for quantitative immunoelectrophoresis. On pepsin digestion of horse antiserum F(ab)' 2 fragments are generated which migrate with the gamma globulins. This cleaved product works well in quantitative immunoelectrophoretic techniques.

Effect of the immunizing dose of soluble and immobilized immunglobulins on the production of antibodies.

The role of the immunizing dose on antibody production was studied in rabbits. The humoral immune response to soluble and insoluble human IgG was evaluated by the measurement of the amount of specific antibodies. The immobilizations of antigen was made by antigen-antibody precipitation and attachment of the protein to polyacrylamide and agarose matrices. It was found, that the antibody production did not decrease when the antigen dose was reduced from 100 microng to 300 microng; application of the antigen in immobilized form did not enhance the antibody response.

Behaviour of polyclonal human IgE in the course of fractionation.

Normal human pooled plasma was fractionated by a variety of methods. The IgE concentration of the different fractions was determined by a solid-phase radioimmunoassay. The results of these studies indicate, that polyclonal IgE behaves similarly to IgE of myeloma origin. A biospecific method was worked out to purify IgE from fraction III of the cold ethanol fractionation procedure.

Improved method for isolation of human IgM as a by-product.

The authors describe a simple method for isolating heterogeneous IgM from fraction I obtained as a by-product of the large scale fractionation of human plasma. The optimum circumstances of fractionation were determined and it was found that the majority of IgM is precipitated between 2 and 5% polyethylene glycol concentration and 1.1 and 1.8 M ammonium sulphate concentration. After a simple fractionation and gel filtration, a preparation of high purity can be obtained.

Comparison of antigen-type immunosorbents prepared by different ways.

7 types of IgG-containing immunosorbents were compared from different respects by using them for adsorption of antibodies. The human IgG was insolubilized by ethylchlorophormate and glutaraldehyde, thereafter it was bound to activated agarose and polyacrylamide gels and 3 new types of polyacrylamide immunosorbent were prepared by polymerizing the acrylamide in the protein solution and by coupling the protein to the matrix by different ways. Using batch technique, the ethylchlorophormate, the glutaraldehyde and some of the new polyacrylamide immunosorbents proved to be useful, but in almost every respect the immunosorbent prepared from Sepharose gel was the best.