Search

Using the micro-scale modification of a newly developed RIA kit for insulin, we established methods for the determination of free and total insulin in serum of insulin-treated diabetics. Precipitation with polyethylene glycol 6000 or acid alcohol extraction of sera was carried out to remove or to dissociate antibody-bound insulin. Both assays permit precise and accurate measurement of either serum insulin fraction. In 50 diabetic sera with tracer insulin binding of 0—97%, free (after equilibration of the sera at 37 °C) and total insulin levels as well as insulin antibody binding parameters were determined. There was a good correlation of free to total insulin levels with maximally 10-fold higher values of total insulin. Both free and total insulin were found to be correlated with the ability of the serum to bind insulin. In detail, binding affinities (i.e. the reciprocal of equilibrium dissociation constants) and binding site concentrations were evaluated which were shown to be positively correlated with free and total insulin levels as well. From these data we conclude that insulin antibodies in the serum may accumulate therapeutic insulin and function as a depot for delivering insulin in insulinopenic episodes (Keilacker et al., 1982 and 1986). Measurement of Insulin in Human Sera Using a New RIA Kit. 1. Insulin Determination in the Absence of Insulin Antibodies — Conventional Assay and Micro Modification

A sensitive and versatile radioimmunoassay (RIA) for insulin was established using human insulin standard, a specific guinea pig anti-insulin antiserum and rabbit anti-guinea pig serum. Radioiodination was performed according to a modified chloramine T method. Tracer preparations were used for as long as 6 weeks after iodination. The standard curve ranges from 0.044 to 1.2 nmol/l. The intra-assay coefficient of variation (CV) was 3-5% and the inter-assay CV was 6-9% in the optimal range between 0.4 and 0.9 nmol/l. The average recovery of human insulin added to plasma or serum samples was 100.2 +/- 2.0% (n = 38) and 100.1 +/- 1.9% (n = 42), respectively. In addition to human insulin, porcine, canine, rabbit and bovine insulin can also be determined but not rat or mouse insulin. The cross-reactivity of the antiserum with porcine proinsulin was found to be 40% on the molar basis. The range of mean fasting plasma insulin concentrations in healthy subjects and under various pathological conditions were estimated.

A radioimmunoassay for the detection of monoclonal islet cell antibodies was developed using rat insulinoma cells as antigen carriers and 125I-labeled affinity-chromatographically purified anti-mouse Ig antibodies for detecting cell-bound mouse Ig. Prior to the assay cells had been attached to glass tubes by poly-dimethyl-diallyl ammonium chloride thus allowing to perform the assay as easy as a solid-phase immunoassay. Incubation protocol and cell number were chosen to ensure a high sensitivity of the assay. Results compared well with immunofluorescence findings. Of seven monoclonal islet cell antibodies tested for crossreactivity only one was displaceable by islet cell surface antibodies from diabetic sera. This antibody was induced by immunization with human islets whereas all others were from mice which had been autoimmunized with streptozotocin and complete Freund's adjuvant.

Autoantibodies against insulin, C-peptide, and glucagon were determined by radio-binding assay in 63 new-onset Type 1 (insulin-dependent) diabetic patients as well as in 70 controls. Plasma peptide binding was determined by means of 125I-labeled peptides and charcoal-dextran separation technique. Binding values exceeding the mean plus three standard deviations of the controls were considered as antibody-positive. Sixteen patients (25%) were positive for IAA, as 6 (10%) were positive for CAA and 2 (3%) for GAA. Of all control subjects, none were positive for either IAA or CAA, whereas 2 (2%) had GAA. The mean 125I-glucagon binding in the patients' group was, however, slightly enhanced and could be suppressed to normal values by excess unlabeled glucagon. The presence of IAA and/or CAA was significantly associated with more severe symptoms at diabetes manifestation. These results indicate that in new-onset Type 1 diabetics autoimmunity arises against all the insular peptides tested but is predominantly directed against those antigens secreted from the beta cells. Nevertheless, extremely low-binding GAA seem to be common in these patients. The determination of IAA/CAA might be useful in detecting a possible heterogeneity of Type 1 diabetes with regard to its clinical mode of manifestation.

The GABA-producing enzyme glutamate decarboxylase (GAD) is a prominent autoantigen in insulin-dependent diabetes mellitus (IDDM). Autoantibodies against GAD were found with a high prevalence in IDDM patients and in animal models for IDDM. The aim of this study was to detect autoantibodies against both isoforms of GAD in diabetic and non-diabetic but diabetes-prone BB/OK rats by Western blotting and to test their specificity to GAD by an immuno-trapping enzyme activity assay. Eighteen diabetic and 18 non-diabetic BB/OK rats (age 121 +/- 20 days) were investigated. In 10/18 (56%) of the diabetic and 13/18 (72%) of the non-diabetic BB/OK rats autoantibodies against at least one GAD-isoform were detected by Western blotting. In the immunotrapping enzyme activity assay, the mean value of the diabetic (1151 +/- 552 cpm, n = 11) and nondiabetic BB/OK rats (1978 +/- 1213 cpm, n = 10) was significantly (p0.01) increased compared to the LEW. 1A control rats (581 +/- 274 cpm, n = 12). 7/10 (70%) individual sera of the non-diabetic and 5/11 (45%) of the diabetic BB/OK rats were positive in this test. In conclusion, the prevalence of GAD autoantibodies in BB/OK rat is connected with the genetic susceptibility to IDDM but is not a predictor for the onset of the disease in BB/OK rats.

Summary: A mathematical theory of competitive labelled-ligand assays was developed with the intention of theoretically re-evaluating the optimal assay conditions and precision data of assay Systems established by experiment. Our theory is based upon the assumptions of a simple bimolecular reaction mechanism, homogeneous reactants, as well as kinetically indistinguishable labelled and non-labelled ligands. The general case of two-step (non-equilibrium) assay was considered including the one-step (equilibrium) assay as a special case. The solution of the System of corresponding kinetic differential equations was used to mathematicall y construct Standard curves. Furthermore, intraassay precision profiles and indices as well as detection limits were calculated considering solely the pipetting error, , as a source of experimental error. A procedure was outlined to mathematicall y determine the optimal incubation conditions for any assay System targeted to a given analyte concentration, P, at which the Standard deviation of assay results is to be minimized. Estimates of both the content of binding sites and the equilibrium constant, K, of the specific binding agent are necessary, and these can be derived from Scatchard plots. For six RIA Systems, of which three were one-step and three were two-step assays, experimental assay conditions and precision data were compared with theoretical predictions. Experimentally determined antibody binding site concentrations agreed fairly well with those independently evaluated by mathematical optimization. Mean precision indices, defined as constituting an average over the complete precision profile, were fotind to be within the theoretically predicted range, i. e. two- to threefold the pipetting error. Detection limits (Standard deviation near concentration 0) differed from theoretical values at most by a factor of two in the case of two-step assays and were nearly identical with theoretical values for one-step assays. Generally, they were Of the order of , approaching a lower limit by the order of , when P falls to the order of K. Comparing the advantages of the one-step and two-step technique of competitive labelled ligand assays, the following results were obtained: The one-step method provides a more favourable pfecision profile, especially a better detection limit, and a higher specificity of analyte recogiution. The quantity of reagents needed (specific binding agent as well as labelled ligand) is three to four times lower than in the two^step method. On the other hand, the higher amount of reagents employed for the two-step technique results in a considefably higher measuring signal, which is important where activity of the labelled ligand is low. We conclude that mathematical modelling of labelled-ligand assays should be useful in re-evaluating assay conditions and precision data obtained by experiment. Furthermore, it permits some general assertions concenung the principal limits of assay precision as well as the advantages and disadvantages pf different assay protocols.

Following optimization of the reaction conditions, e.g. concentration of oxidizing agents, reaction time, volume of reaction mixture, and pH, chloramine T and the new iodination reagent, Iodogen, were compared for their effectiveness in radioiodination of insulin, glucagon, human growth hormone (hGH), and rabbit anti-mouse IgG. The radioactive peptide hormones prepared were analyzed for the presence of aggregate and breakdown products by polyacrylamide gel electrophoresis (PAGE) at pH 8.9, the rabbit anti-mouse IgG was tested for the presence of low molecular weight damage products by gel filtration on Sephadex G-50. The results demonstrate that with respect to iodine incorporation, specific activity, and immunological reactivity either method can be used to prepare under carefully controlled conditions a wide range of tracers with high specific activity at minimal oxidation damage. These tracers are shown to be highly suitable in radioimmunoassays after previous purification by PAGE and gel filtration, respectively.

Based on the production of a high specific antibody against ovine prolactin a heterologous prolactin RIA (self proposed method) is introduced. From this work both a homologous and a heterologous system have been developed and used for the PRL-RIA kit production. Substances necessary for setting up the kits, their investigations and applications are described. Statistical considerations about precision, sensitivity, specificity and accuracy show the validity of these heterologous PRL-systems.

Isolated rat pancreatic islets were incubated with either rat or human serum. Undiluted serum of both rat and man provided a strong insulin release, probably caused by a leakage of damaged B-cells. Heat inactivation of the serum only partially reversed this effect. Serum, which was diluted by 50% with buffer, had a much smaller cytotoxic activity. The effect on insulin leakage was further depressed when the complement system of the serum was inactivated by heat or zymosan treatment or storage of the serum before use. When rat islets were incubated in rat or human serum pretreated in this way, the glucose effect on insulin secretion could be determined.

A sensitive and versatile radioimmunoassay (RIA) for insulin was established using human insulin standard, a specific guinea pig anti-insulin antiserum and rabbit anti-guinea pig serum. Radioiodination was performed according to a modified chloramine T method. Tracer preparations were used for as long as 6 weeks after iodination. The standard curve ranges from 0.044 to 1.2 nmol/l. The intra-assay coefficient of variation (CV) was 3-5% and the inter-assay CV was 6-9% in the optimal range between 0.4 and 0.9 nmol/l. The average recovery of human insulin added to plasma or serum samples was 100.2 +/- 2.0% (n = 38) and 100.1 +/- 1.9% (n = 42), respectively. In addition to human insulin, porcine, canine, rabbit and bovine insulin can also be determined but not rat or mouse insulin. The cross-reactivity of the antiserum with porcine proinsulin was found to be 40% on the molar basis. The range of mean fasting plasma insulin concentrations in healthy subjects and under various pathological conditions were estimated.

An islet cell tumor, characterized by proinsulin level significantly elevated above normal human pancreas, has been found to contain insulin- and glucagon-degrading activity. Examination by chromatography on Sephadex G-75 of the degradation products formed from insulin showed A chain, and B chain rich-A chain aggregate as previously found with rat pancreatic islets. There was, however, little conversion of A chain to low molecular weight components indicating that insulinoma peptidase that has been found to degrade glucagon at about pH 6.8 degraded that A chain to a markedly lower rate. In contrast to the insulin-degrading activity, which was activated by glutathione in the presence of EDTA, the peptidase activity was not affected by the thiol compound. The activity of the peptidase was markedly inhibited by chelating agents, i.e., EDTA and o-phenanthroline, whereas chymotrypsin and trypsin inhibitors, i.e., TOS-PheCH2Cl, TOS-LysCH2Cl, soybean and pancreas trypsin inhibitor were found to have no effect.

We investigated equilibrium plasma binding patterns of insulin in 45 juvenile diabetics treated with conventional insulin preparations. Insulin binding parameters were evaluated by Scatchard analysis of the binding data. Stable diabetics had significantly lower equilibrium dissociation constants than labile, thus suggesting an enhanced insulin depot effect due to stronger insulin binding. Correlation of insulin binding data with a glycemic control index yielded a positive relationship between insulin antibody binding and the degree of glycemic control. Insulin neutralization as detected by a relationship between maximum binding capacity of high affinity antibodies and insulin requirement could only be found if patients with poor diabetes control were excluded. Similarly, the well-known promoting influence of residual beta-cell functional capacity (assessed by C-peptide levels) on diabetic stability was observed only after exclusion of patients with higher insulin antibody binding. These data suggest that insulin antibodies are influencing insulin treatment of diabetics in a dual way. They may neutralize therapeutic insulin but at the same time they exert an insulin-sparing action by improvement of diabetes control. Occasionally the latter effect may abolish the correlation between diabetes control and beta-cell functional capacity.

A radioimmunoassay for the detection of monoclonal islet cell antibodies was developed using rat insulinoma cells as antigen carriers and 125I-labeled affinity-chromatographically purified anti-mouse Ig antibodies for detecting cell-bound mouse Ig. Prior to the assay cells had been attached to glass tubes by poly-dimethyl-diallyl ammonium chloride thus allowing to perform the assay as easy as a solid-phase immunoassay. Incubation protocol and cell number were chosen to ensure a high sensitivity of the assay. Results compared well with immunofluorescence findings. Of seven monoclonal islet cell antibodies tested for crossreactivity only one was displaceable by islet cell surface antibodies from diabetic sera. This antibody was induced by immunization with human islets whereas all others were from mice which had been autoimmunized with streptozotocin and complete Freund's adjuvant.

Using the micro-scale modification of a newly developed RIA kit for insulin, we established methods for the determination of free and total insulin in serum of insulin-treated diabetics. Precipitation with polyethylene glycol 6000 or acid alcohol extraction of sera was carried out to remove or to dissociate antibody-bound insulin. Both assays permit precise and accurate measurement of either serum insulin fraction. In 50 diabetic sera with tracer insulin binding of 0-97%, free (after equilibration of the sera at 37 degrees C) and total insulin levels as well as insulin antibody binding parameters were determined. There was a good correlation of free to total insulin levels with maximally 10-fold higher values of total insulin. Both free and total insulin were found to be correlated with the ability of the serum to bind insulin. In detail, binding affinities (i.e. the reciprocal of equilibrium dissociation constants) and binding site concentrations were evaluated which were shown to be positively correlated with free and total insulin levels as well. From these data we conclude that insulin antibodies in the serum may accumulate therapeutic insulin and function as a depot for delivering insulin in insulinopenic episodes (Keilacker et al., 1982 and 1986).

Glucagon antibodies were produced in rabbits using three immunogenic glucagon conjugates. Glucagon was coupled to bovine serum albumin by difluorodinitrobenzene (DFDNB), carbodiimide (ECDI) or glutardialdehyde. Rabbits immunized with glucagon conjugated to albumin using DFDNB produced sensitive antisera for radioimmunoassay quite specific for pancreatic glucagon. The affinity constant of the best antiserum was approximately 10(11) l/mol. Antisera raised against the two other glucagon-conjugates were significantly less affine. All of these antisera showed inverse binding curves of 126J-glucagon caused by positive cooperatively in dependence upon the antigen/antibody ratio. The race of rabbits used for immunization was without influence on the immune response.

Radioiodinated polypeptide hormones, such as insulin, glucagon, human growth hormone, and human C-peptide are employed for radioimmunoassays for investigations of hormonal alterations in states of disturbed carbohydrate metabolism. Iodination was performed using chloramine T. Iodination products of these polypeptide hormones and, for preparation of standard material, native human C-peptide from cadaver pancreases were fractionated by polyacrylamide gel electrophoresis at pH 8.9. Disc electrophoresis in 24 cm-long gel rods resulted in stable tracers with high specific activity as well as homogeneous standard material being highly suitable for radioimmunoassays.

Three different methods of legal abortion were applied to 23 patients, who had asked for termination of pregnancy, between the eleventh and 14th weeks of pregnancy. In vitro studies into human foetal pancreas were taken into consideration, when the choice was made. Carbohydrate tolerance was normal in all cases (50 g oGTT). Timing of intra-uterine foetal death, in the context of the three above methods (B-scan ultrasonic diagnosis and HPL level measurement) as well as studies into in vitro stimulation of foetopancreatic insulin secretion have shown purely mechanical, surgical abortion, sometimes following prostaglandin cervical gel priming, to be indispensable a condition for successful in vitro elucidation of the aetiopathogenesis of foetal diabetes. Extra-amniotic prostaglandin induction of abortion has failed to yield any relevant information, in that context.3 different methods of legal abortion were administered to 23 patients who had requested termination of pregnancy between weeks 11-14 of pregnancy. In vitro studies into human fetal pancreas were taken into consideration when the choice was made. Carbohydrate tolerance was normal in all cases (50 g oGTT). Timing of intrauterine fetal death, in the context of the 3 above methods (B-scan ultrasonic diagnosis and HPL level measurement), as well as studies into in vitro stimulation of fetopancreatic insulin secretion, have shown purely mechanical, surgical abortion, sometimes following prostaglandin (PG) cervical gel priming, to be indispensable as a condition for successful in vitro elucidation of the etiopathogenesis of fetal diabetes. Extraamniotic PG induction of abortion has failed to yield any relevant information in that context. (author's)

A sensitive radioimmunoassay (RIA) for canine C-peptide (CCP) was established using synthetic CCP, a specific antiserum, and rabbit anti-guinea pig serum. Radioiodination was performed according to a modified chloramine-T method. Tracer preparations were used for long as 6 weeks after iodination. The standard curve ranges from 0.028 to 3.0 nmol/l. The intra-assay coefficient of variation (CV) was 3-5% and the inter-assay CV was 6-9% in the optimal range between 0.3 and 0.8 nmol/l. The average recovery of CCP added to plasma samples was 100.6% (n = 9). Canine insulin, porcine proinsulin, bovine proinsulin, and human C-peptide exhibited no cross-reactivity. The mean fasting plasma CCP concentration was 0.089 +/- 0.021 nmol/l in normal dogs and -0.005 +/- 0.007 nmol/l (mean +/- SEM) in diabetic dogs, respectively.

A reproducible method for preparation of 125I-monoiodoinsulin with fully biological activity was developed. Monoiodoinsulin has been prepared from a heterogeneous 125I-iodination mixture by anion exchange chromatography on DEAE-Sephadex A-25 without using any gradient elution technique. The specific radioactivity of 125I-monoiodoinsulin was calculated to 14.3 +/- 0.8 TBq/g, i.e. an iodine content of 1.04 +/- 0.06 atoms per molecule of insulin. Monoiodoinsulin was indistinguishable from native insulin with respect to binding to guinea pig anti-insulin serum, and to insulin receptors of isolated rat adipocytes. The biological potency (96.5 +/- 7.5 per cent of the immunoreactive insulin activity) determined by the conversion of [14C1]-D-glucose to 14CO2 in vitro by rat fat cells was not significantly different from that of native insulin.

In 19 patients aged from 12 to 24 years (average age 17.0 years) with Turner's syndrome the influence of the hormonal substitution with mestranol and chlormadinone acetate on both basal and arginine stimulated HGH secretion was investigated. This test was performed before, during and after treatment. HGH serum levels were determined by RIA. Both the basal values and the increase of HGH secretion after arginine stimulation were used for evaluation of the test results. Under estrogen-gestagen treatment in 15 of the 19 patients hypersomatotropic basal levels were determined. The basal HGH levels before and during as well as during and after treatment are significant different (p less than 0,01). The responsibility of the somatotrophs to the arginine stimulation was individually different and did not show any uniform trend.

We investigated four insulin-specific hybridoma antibodies with respect to their kinetic properties as well as the binding behaviour of some combinations of them. From equilibrium binding data all but one antibodies were shown to bear homogeneous binding sites. They revealed homogeneity of binding sites also by kinetic experiments, thus with high probability being monoclonal. At 0 degree C, two of them showed discrepancies between kinetic and steady state binding data in as much as, at steady state, the measured bound-to-free ratio of tracer insulin was 3-4 times lower than calculated from kinetic data. Thus a simple bimolecular reaction mechanism could possibly not be applicable. Mixing two monoclonal insulin antibodies, neither cooperative nor additive binding to the insulin molecule could be observed but only competitive effects. Especially, no positive cooperativity between two or more antibodies could be detected, which would be able to account for the higher affinity usually observed for polyclonal vs. monoclonal antibodies.

We investigated the plasma insulin binding patterns of 15 insulin-treated juvenile diabetics with a very low residual B-cell function. Determination of the insulin binding was performed after removal of therapeutic insulin. 125I-monoiodoinsulin was used and insulin binding was measured over a large range of insulin concentration. Insulin binding parameters were evaluated by Scatchard analysis of the binding data. Significantly elevated insulin binding was detected when the diabetic control was stable. These patients had a lower insulin requirement and low equilibrium dissociation constants of the high affinity antibodies. Furthermore, a strong negative correlation between the duration of insulin treatment and insulin binding could be demonstrated resulting from both decreasing antibody affinities and maximum binding capacities. We discuss the stabilizing effect of insulin antibodies on the metabolic character assuming that dissociating insulin-antibody complexes mimic a "basal insulin secretion" similar to pancreatic B-cells with residual functional capacity.

On the basis of the estimation of the plasma glucose and the insulin concentrations in 10-minute intervals during 24 hours by the mathematical analysis (multiple regression) over relations between the behaviour of plasma glucose and insulin concentration insulin secretion rates independent of plasma glucose can be calculated. In non-diabetics a circadian rhythm of the disposal of insulin - independent of food - is established. Thus the secretion rate increases in the early morning hours still before before the first meal and is maintained in different size up to nearly 18 o'clock. Then it established itself to the level of the nocturnal basal rate. Adequate changes of the insulin need during night are proved in diabetics without endogenic residual insulin secretion by means of the artificial B-cell Biostator. The circadian rhythm of the disposal of insulin take place parallel with adequate changes of other biological parameters and is of importance for the practice of the conventional insulin therapy as well as in the future use of non-measuring value-regulated insulin infusions (open-loop system).