Improving solubility of catalytic domain of human beta-1,4-galactosyltransferase 1 through rationally designed amino acid replacements.

beta-1,4-galactosyltransferase 1 (beta4gal-T1, EC 2.4.1.38) transfers galactose from UDP-galactose to free N-acetyl-D-glucosamine or bound N-acetyl-D-glucosamine-R. Soluble beta4gal-T1, purified from human milk has been refractory to structural studies by X-ray or NMR. In a previous study (Malissard et al. 1996, Eur. J. Biochem. 239, 340-348) we produced in the yeast Saccaromyces cerevisiae an N-deglycosylated form of soluble beta4gal-T1 that was much more homogeneous than the human enzyme, as it displayed only two isoforms when analysed by IEF as compared to 13 isoforms for the native beta4gal-T1. The propensity of recombinant beta4gal-T1 to aggregate at concentrations > 1 mg.mL(-1) prevented structural and biophysical studies. In an attempt to produce a beta4gal-T1 form suitable for structural studies, we combined site-directed mutagenesis and heterologous expression in Escherichia coli. We produced a mutated form of the catalytic domain of beta4gal-T1 (sfbeta4gal-T1mut) in which seven mutations were introduced at nonconserved sites (A155E, N160K, M163T, A168T, T242N, N255D and A259T). Sfbeta4gal-T1mut was shown to be much more soluble than beta4gal-T1 expressed in S. cerevisiae (8.5 mg.mL(-1) vs. 1 mg.mL(-1)). Catalytic activity and kinetic parameters of sfbeta4gal-T1mut produced in E. coli were shown not to differ to any significant extent from those of the native enzyme.     

Detection of terminal N-linked N-acetylglucosamine residues in the Golgi apparatus using galactosyltransferase and endoglucosaminidase F/peptide N-glycosidase F: adaptation of a biochemical approach to electron microscopy

Purified human milk beta-N-acetylglucosaminide beta 1, 4 galactosyltransferase (EC 2.4.1.38) was used to galactosylate N-acetylglucosamine (GlcNAc) residues present in ultra-thin sections of Lowicryl K4M-embedded rat and pig liver. Both endogenous galactose and galactosylated transferase products could be revealed by Ricinus communis lectin I-gold complexes (RcL I-g15). Without galactosyltransferase (GT) treatment, labeling for galactose (gal) was limited to the trans region of rat and pig hepatocyte Golgi apparatus. After exposure to GT, additional labeling was found over cis Golgi apparatus cisternae. RcL I-g15 labeling was sensitive to a purified preparation of endoglucosaminidase F/peptide N-glycosidase F (at pH 9). This indicates that endogenous gal and gal transferred by GT to terminal GlcNAc residues are present N-linked oligosaccharides. The RcL I-g15 labeling produced by GT was insensitive to extensive washing with solutions containing either EDTA and urea or SDS and 2-mercaptoethanol or 0.1 M GlcNAc. Substrate inhibition studies showed that 50 mM GlcNAc specifically inhibited the additional RcL I-g15 labeling produced by GT. The use of purified glycosyltransferases therefore appears to allow specific detection of oligosaccharide substrates and their high resolution localization in thin sections by electron microscopy.     

Affinity labeling of the carbohydrate binding site of the lectin discoidin I using a photoactivatable radioiodinated monosaccharide

N-(4-Azidosalicyl)galactosamine (GalNASA), a photoactivatable, radioiodinatable analogue of N-acetylgalactosamine (GalNAc), has been prepared and characterized. We have used this reagent for labeling of the carbohydrate binding site of discoidin I, an endogenous lectin produced by Dictyostelium discoideum. GalNASA behaved as a ligand for discoidin I, as judged by its ability to compete in an assay measuring the carbohydrate binding activity of discoidin I. In this assay, it exhibited a Ki,app of 800 microM, comparable to that of GalNAc. The Ki,app of GalNASA decreased to 40 microM upon prior photolysis with ultraviolet light. In contrast, N-(4-azidosalicyl)ethanolamine produced no inhibition of carbohydrate binding regardless of photolysis. Covalent labeling of discoidin I with 125I-GalNASA was entirely dependent upon ultraviolet light. A portion of the labeling, representing 40-60% of the total, was sensitive to reagents which were known to inhibit carbohydrate binding by discoidin I, including GalNAc, asialofetuin, and ethylenediaminetetraacetic acid. N-Acetylglucosamine, which is not a ligand of discoidin I, was without effect. As a control, no carbohydrate-sensitive labeling was observed upon incubation of 125I-GalNASA with bovine serum albumin. The carbohydrate-sensitive fraction of discoidin I photolabeling with 125I-GalNASA exhibited a Kd of 15-40 microM, in agreement with the Ki,app of prephotolyzed GalNASA observed in the carbohydrate binding assay. Some labeling occurred if 125I-GalNASA was photolyzed prior to incubation with discoidin I, suggesting the involvement of long-lived species in the labeling reaction. Partial proteolytic digestion of photolabeled discoidin I revealed specific fragments whose labeling was completely blocked by GalNAc.(ABSTRACT TRUNCATED AT 250 WORDS)     

Reclustering of scattered Golgi elements occurs along microtubules

Depolymerization of the interphase microtubules by nocodazole results in the scattering and apparent fragmentation of the Golgi apparatus in Vero fibroblast cells. Upon removal of the drug, the interphase microtubules repolymerize, and the scattered Golgi elements move back to the region around the microtubule-organizing center (MTOC) within 40 to 60 min. Using a fluorescent lipid analogue (C6-NBD-ceramide) as a vital stain for the scattered Golgi elements, their relocation was visualized by video-enhanced fluorescence microscopy in Vero cells maintained at 20 degrees C. The NBD-labeled structures were identified as Golgi elements by their colocalization with galactosyltransferase in the fixed cells. During reclustering, NBD-labeled Golgi elements were observed to move by discontinuous saltations towards the MTOC with velocities of 0.1 to 0.4 micron/s. Paths along which Golgi elements moved were super-imposable on microtubules visualized by indirect immunofluorescence. Neither the collapse of intermediate filaments caused by microinjection of antibodies to vimentin nor the disruption of microfilaments by cytochalasin D had an effect on the reclustering of Golgi elements or the positioning of the Golgi apparatus. These data show that scattered Golgi elements move along microtubules back to the region around the MTOC, while neither intact intermediate filaments nor microfilaments are involved.     

Different numbers of maternal and paternal siblings of cystic fibrosis patients

Summary When 458 parents of children suffering from cystic fibrosis (CF) from all over the German Democratic Republic were interviewed to determine the number of their siblings, it was found that the maternal families had a total of 1369 children and the paternal, 1220. While the fathers of CF patients tended to originate from families with one or two children, more mothers than fathers came from families with three to twelve children (P=0.01). The average number of children in the maternal families was 2.99; in the paternal families, only 2.66. To rule out any methodological errors, sibs of mothers and fathers of various control groups were studied. We found that the number of siblings in these groups was balanced. The differences in our findings are probably due to CF heterozygosity. The underlying mechanism is unknown.     

Modification of the zona-free hamster ova bioassay of boar sperm fertility and correlation with in vivo fertility

These studies were designed to evaluate the ability of the zona-free hamster ova bioassay to detect differences in fertility of boar sperm. In the first study, sperm from two previously infertile boars were compared to sperm from seven previously fertile boars. The percentage of zona-free hamster ova penetrated by sperm from the previously infertile boars was significantly lower than the percentage of ova penetrated by sperm from previously fertile boars (18% of ova penetrated vs. 83%, P < .001). In the 14 ejaculates from the previously infertile boars that had ejaculate motilities of 50% or greater, the percentage of zona-free hamster ova penetrated continued to be lower than in ejaculates from the fertile boars. One of the two previously infertile boars consistently had a normal semen analysis. The only two observed manifestations of his reduced fertility were his zero conception rate and the limited ability of his sperm to penetrate zona-free hamster ova. In the second study, females were inseminated with equal numbers of sperm from two previously fertile males and the paternity of offspring determined at birth. The experiment was replicated with four combinations of six boars. A high correlation was observed between the percentage of offspring sired and the ability to penetrate zona-free hamster ova (R = .89). Neither morphology nor the ability of the sperm to undergo an acrosome reaction during in vitro incubation was correlated with fertility in the competitive mating situation. These results suggest the zona-free hamster ova bioassay can improve the in vitro fertility assessment of fresh boar semen.