Contribution of the H- and L-chains and of the binding site to the idiotypic specificities of mouse anti-GAT antibodies.

The contribution of the H- and L-chains to the structure of the main idiotype of anti-poly (Glu60-Ala30-Tyr10) (GAT) antibodies has been studied. This idiotype has been previously divided into four types of specificity: (1) the highly conserved idiotypic specificity (h.c. GAT) is expressed by anti-GAT antibodies from the guinea-pig, rat and mice; (2) the public specificity (p. GAT) is expressed in an identical form by all anti-GAT antibodies from all strains of mice tested and by all hybridoma products (HP) with anti-GAT activity; (3) the strain-restricted specificity (s.r. GAT-1) is only expressed by anti-GAT antibodies from strains with Ig-1a, Ig-1c and Ig-1c allotypic markers; and finally (4) the individual specificity i1-GAT defined on HP G5 is also expressed by most of the hybridoma protein with anti-poly (Glu50-Tyr50) (GT) activity. In this paper we show that h.c.GAT, p.GAT and i1-GAT require the interaction of H- and L-chains to be expressed: (1) isolated H- and L-chains from HP G5 did not express these specificities; and (2) recombinant molecules composed of H- and L-chains from HP with anti-GAT activity and an irrelevant myeloma protein (MOPC21) never expressed h.c.GAT, p.GAT and i1-GAT. We next investigated the relationship between the GAT binding site and the p.GAT, h.c.GAT and s.r.GAT-1 idiotypic specificities. GAT and GT were not able to inhibit the binding to s.r.GAT-1 while they inhibit the idiotypic binding to p.GAT and h.c.GAT. A GAT fragment of mol. wt 3000 was also shown to inhibit the binding of p.GAT and h.c.GAT to the appropriate sera. Thus p.GAT and h.c.GAT are very close to the GAT combining site while s.r.GAT-1 represents an idiotypic specificity located outside the GAT binding site.

Effect of physical exercise on lipoprotein(a) and low-density lipoprotein modifications in type 1 and type 2 diabetic patients.

To evaluate the effect of physical exercise on blood pressure, the lipid profile, lipoprotein(a) (Lp(a)), and low-density lipoprotein (LDL) modifications in untrained diabetics, 27 diabetic patients (14 type 1 and 13 type 2) under acceptable and stable glycemic control were studied before and after a supervised 3-month physical exercise program. Anthropometric parameters, insulin requirements, blood pressure, the lipid profile, Lp(a), LDL composition, size, and susceptibility to oxidation, and the proportion of electronegative LDL (LDL(-)) were measured. After 3 months of physical exercise, physical fitness improved (maximal O2 consumption [VO2max], 29.6 +/- 6.8 v 33.0 +/- 8.4 mL/kg/min, P < .01). The body mass index (BMI) did not change, but the waist circumference (83.2 +/- 11.8 to 81.4 +/- 11.2 cm, P < .05) decreased significantly. An increase in the subscapular to triceps skinfold ratio (0.91 +/- 0.37 v 1.12 +/- 0.47 cm, P < .01) and midarm muscle circumference ([MMC], 23.1 +/- 3.4 v 24.4 +/- 3.7 cm, P < .001) were observed after exercise. Insulin requirements (0.40 +/- 0.18 v 0.31 +/- 0.19 U/kg/d, P < .05) and diastolic blood pressure (80.2 +/- 10 v 73.8 +/- 5 mm Hg, P < .01) decreased in type 2 diabetic patients. High-density lipoprotein cholesterol (HDL-C) increased in type 1 patients (1.48 +/- 0.45 v1.66 +/- 0.6 mmol/L, P < .05), while LDL cholesterol (LDL-C) decreased in type 2 patients (3.6 +/- 1.0 v3.4 +/- 0.9 mmol/L, P < .01). Although Lp(a) levels did not vary in the whole group, a significant decrease was noted in patients with baseline Lp(a) above 300 mg/L (mean decrease, -13%). A relationship between baseline Lp(a) and the change in Lp(a) (r = -.718, P < .0001) was also observed. After the exercise program, 3 of 4 patients with LDL phenotype B changed to LDL phenotype A, and the proportion of LDL(-) tended to decrease (16.5% +/- 7.4% v 14.0% +/- 5.1%, P = .06). No changes were observed for LDL composition or susceptibility to oxidation. In addition to its known beneficial effects on the classic cardiovascular risk factors, regular physical exercise may reduce the risk of cardiovascular disease in diabetic patients by reducing Lp(a) levels in those with elevated Lp(a) and producing favorable qualitative LDL modifications.

Development of binary classification of structural chromosome aberrations for a diverse set of organic compounds from molecular structure.

Classification models are generated to predict in vitro cytogenetic results for a diverse set of 383 organic compounds. Both k-nearest neighbor and support vector machine models are developed. They are based on calculated molecular structure descriptors. Endpoints used are the labels clastogenic or nonclastogenic according to an in vitro chromosomal aberration assay with Chinese hamster lung cells. Compounds that were tested with both a 24 and 48 h exposure are included. Each compound is represented by calculated molecular structure descriptors encoding the topological, electronic, geometrical, or polar surface area aspects of the structure. Subsets of informative descriptors are identified with genetic algorithm feature selection coupled to the appropriate classification algorithm. The overall classification success rate for a k-nearest neighbor classifier built with just six topological descriptors is 81.2% for the training set and 86.5% for an external prediction set. The overall classification success rate for a three-descriptor support vector machine model is 99.7% for the training set, 92.1% for the cross-validation set, and 83.8% for an external prediction set.

Linear regression and computational neural network prediction of tetrahymena acute toxicity for aromatic compounds from molecular structure.

A quantitative structure toxicity relationship (QSTR) has been derived for a diverse set of 448 industrially important aromatic solvents. Toxicity was expressed as the 50% growth impairment concentration (ICG(50)) for the ciliated protozoa Tetrahymena and spans the range -1.46 to 3.36 log units. Molecular descriptors that encode topological, geometrical, electronic, and hybrid geometrical-electronic structural features were calculated for each compound. Subsets of molecular descriptors were selected via a simulated annealing technique and a genetic algorithm. From this reduced pool of descriptors, multiple linear regression models and nonlinear models using computational neural networks (CNNs) were derived and then used to predict the ICG(50) values for an external set of representative compounds. An average of 10 nonlinear CNN models with 11-5-1 architecture was found to best describe the system with root-mean-square errors of 0.28, 0.29, and 0.34 log units for the training, cross validation, and prediction sets, respectively.