The relationship between cell differentiation/tumorisation and plasma membrane glycoproteins was approached using peanut agglutinin (PNA) a lectin specific for the Gal-beta(1,3)GalNAc sequence and a homologous cell system consisted of normal rat hepatocytes (HyC) and a poorly differentiated hepatoma (ZHC). This work is focused on the molecular nature of PNA receptors. PNA bound strongly to ZHC, but bound very weakly, if at all to hepatocytes. After sialidase treatment this binding was slightly enhanced in ZHC and HyC. The total number of binding sites on ZHC was 9.6 x 10(6)/cell and 1.2 X 10(7)/cell before and after sialidase treatment respectively. In contrast, this number could not be calculated on HyC, even after sialidase treatment. The PNA receptors were isolated and identified from ZHC using affinity chromatography on immobilized PNA and lectin overlay. Two bands were revealed after SDS-PAGE of PNA receptors: a major one with a relative molecular mass of 160 kDa and a minor one of 110 kDa. The latter disappeared after sialidase treatment of ZHC suggesting the possibility that these two bands could be less and more sialylated forms of the PNA receptors, respectively. In contrast no PNA receptors could be detected on HyC. These PNA receptors could be considered O-linked glycoproteins containing the Gal-beta(1,3)GalNAc disaccharide because: i) PNA carbohydrate specificity toward this disaccharide found in this glycoprotein type; ii) their carbohydrate composition with Gal and GalNAc but not man residues; iii) their sensitivity to alkaline treatment; and iv) strong inhibition of PNA binding to ZHC with the Gal-beta(1,3)GalNAc structure. The absence of PNA receptors on HyC appeared to be related to the absence of this glycoprotein containing the disaccharide but not to the change or failure of glycosylation of the polypeptide chain of PNA receptors. The relationship between the presence of PNA receptors and differentiation/tumorisation phenomena as well as the mechanism that induced the expression of these receptors are discussed.

Calmodulin levels are elevated three- to fourfold in the dividing cells, resulting from the lectin-induced stimulation of fresh human lymphocytes. This increase in calmodulin appears to be related mainly to progression into S phase and supports the hypothesis that calmodulin might be crucial in regulating the progression of lymphoblasts through their division cycle. Calmodulin levels are higher in a lymphoid cell line derived from human acute lymphoblastic leukemia blood cells than in a lymphoid cell line derived from normal human blood cells, suggesting that calmodulin could be an important mediator of the leukemogenetic process.

Rice lectin (Oryza sativa, var. Balilla 28) was purified from defatted embryos by aqueous acid extraction at pH 1.3 followed by ammonium sulfate precipitation between 2 and 4 m, affinity chromatography on agarose-p-aminophenyl-β-d-N-acetylglucosamine,and gel filtration on AcA 54. The homogeneity of the lectin was checked by polyacrylamide gel electrophoresis, gel filtration, and immunodiffusion. The amino acid analysis revealed a high half-cystine content (9%) and a low aromatic and hydrophobic amino acid content. The lectin contained neither neutral carbohydrates nor amino sugars. The isoelectric point was estimated to be 8.1. The molecular weight of rice lectin was estimated to be 38,000. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis under reducing conditions showed two polypeptides with Mr 19,000 and 15,000. The circular dichroism spectrum of rice lectin in far ultraviolet was characterized by a positive maximum at 228 nm and a negative band at 203 nm suggesting the presence of a β-pleated sheet and the absence of α-helix. Rice lectin had no human blood group specificity and agglutinated rabbit erythrocytes more efficiently than erythrocytes from other animal species. Furthermore, agglutination was enhanced by trypsin treatment of erythrocytes. The erythroagglutinating activity was very high since the minimal concentration needed to agglutinate erythrocytes was 0.05 μg/ml. Although [methyl-3H]thymidine incorporation was stimulated in human lymphocytes, rice lectin could not be considered as a mitogenic lectin since it stimulated neither blast transformation nor lymphocyte proliferation. The saccharide specificity of rice lectin was related to N-acetylglucosamine and its oligomers: N,N′,N″-triacetylchitotriose was the most powerful inhibitor. Furthermore, the N-acetylneuraminic acid was not a specific rice lectin determinant. Finally, the double immunodiffusion method revealed a cross-reactivity between rice lectin and wheat germ agglutinin, indicating that these lectins were closely antigenically related. The analogies and differences between biological and immunological properties of rice lectin and wheat germ agglutinin are discussed and the possibility of their evolution from a common ancestor is put forward.

The molecular properties of the haemagglutinin of Ricinus communis (RCA I or RCA 120) were evaluated by analytical ultracentrifugation, light-scattering, c.d. and fluorescence. The native molecule had a fairly expanded structure (f/f0 = 1.43) and dissociated into two subunits of equal size in 6 M-guanidinium chloride. This native structure was stable in alkali (up to pH 11) and resistant to thermal denaturation at neutrality. A pH-triggered change in the haemagglutinin conformation was observed and characterized by analytical ultracentrifugation, c.d. and fluorescence between pH 7 and 4.5, the range in which its affinity for galactosides decreased [Yamasaki, Absar & Funatsu (1985) Biochim, Biophys. Acta 828, 155-161]. These results are discussed in relation to those reported in the literature for other lectins and more especially ricin, for which a pH-dependent conformation transition has been observed in the same range of low pH.

An in vitro study of the behaviour of a human acute lymphoblastoid leukemia cell line (REH) towards the action of a mitogenic lectin of Robinia pseudoacacia was carried out. The results were compared with those a reference cell line (LHN13) established from normal human lymphocytes. In both cell lines, the lectin induces agglutination (measured by counting the number of aggregates as well as the number of cells in each aggregate) and decrease of growth (measured by counting the number of cells and the incorporation of tritiated thymidine into TCA-precipitable material per 10(6) cells). The agglutination and the decrease of growth are produced at the doses of 0.5 and 1 microgram/ml of culture medium and after 4 h of exposure of cells to the lectin, respectively. These effects increase progressively with higher doses of lectin and continues throughout the culture. However, the REH line is less sensitive than the LHN13 line to the effects of lectin. Both agglutination and growth decrease of REH as well as LHN13 cell lines by the lectin are reversible; this is confirmed by the fact that the monospecific anti-Robinia lectin serum suppresses these effects.

A lectin was purified from rice embryos by aqueous acid extraction of crude embryo powder, followed by ammonium sulfate precipitation, affinity chromatography on agarose p-aminophenyl-beta-D-N-acetylglucosamine and gel-filtration on AcA 54. Its homogeneity was checked by polyacrylamide gel electrophoresis, gel-filtration and immunological methods. The hemagglutinating activity of the purified rice lectin was 0.02 micrograms/ml. This lectin labelled with [14C] acetic anhydride was shown to interact in vitro with different bacteria isolated from the rhizosphere of rice. The most efficient binding was obtained with Beijerinckia V.. The affinity constant Ka was (1.04 +/- 0.30) X 10(7) M-1 and each bacterium contained 1660 +/- 150 lectin receptor sites. In contrast, no interaction between bacteria isolated from the rhizosphere of maize or E. coli K 12 and rice lectin was evidenced.

After treatment with trypsin, chick embryo fibroblasts exhibited an age‐related difference in their capacity to readhere to the substratum, since 8‐day‐cells readhered more rapidly than 16‐day‐cells. Treatment with tunicamycin altered embryo cell readhesion to the substratum in varying degrees, depending on the duration of drug treatment and of readhesion assay. The effect of tunicamycin was not toxic and was totally reversible with time after its removal. These results indicated that embryo cell readhesion involved trypsin‐sensitive cell surface glycoproteins. During embryo development, the glycosylation of cell surface glycoproteins altered markedly. The ratio of N‐linked to O‐linked glycan chains dropped from 80/20 in 8‐day‐cells to 55/45 in 16‐day‐cells, indicating that the relative labelling of O‐linked glycan chains increased during embryo development. This result was confirmed by alkaline treatment of radiolabelled glycan chains, and by the fact that tunicamycin treatment reduced 14C‐glucosamine incorporation by greater than or equal to 80% in 8‐day‐cells but only 60% in 16‐day‐cells. Marked changes were observed during embryo development in the structure of the N‐linked glycan chains; concanavalin A‐Sepharose chromatography showed that these changes concerned the glycopeptides containing complex type carbohydrate chains. The ratio of tri‐ plus tetra‐antennary chains to bi‐antennary chains increased about 2.5‐fold between the 8th and 16th day of development. A correlation was noted between embryo cell readhesion to the substratum and N‐glycosylation of cell surface glycoproteins. The N‐linked glycoconjugates played a crucial part in cell readhesion. The possible role of O‐linked structures in such readhesion is discussed.

Qualitative variations in the glycoconjugates which make up the lectin receptor sites on the membranes of leukemic lymphocytes, compared with those of normal cells, have been studied by the use of three tritiated lectins: Robinia pseudoacacia lectin, Concanavalin A and Ricinus communis (var. Sanquineus) agglutinin (RCA 120). The binding specificity of these lectins has been demonstrated using specific determinants: alpha-methylmannoside and galactose for Concanavalin A and Ricinus communis agglutinin respectively. For the Robinia lectin this specificity was determined by saturation of the receptor sites with the unlabeled Robinia lectin before the addition of isotopically labeled Robinia lectin. The results show a decrease in the number of receptor sites on the leukemia cells, especially in chronic lymphoid leukemia, relative to that on normal cells. The apparent affinity constants of leukemic cells in all cases remain higher than those of normal cells.

The interaction between Concanavalin A and chick embryo fibroblasts was studied. Cells from younger and older embryos had the same number of lectin receptor sites per cell at 4°, 21° and 37°C but the affinity constants increased with increasing temperature. Analysis of the binding data according to Scatchard showed that the apparent changes in binding as a function of temperature might be related to thermodynamic properties. The lectin binding sites on the cell surface proved homogeneous with regard to binding properties.

Cytochalasin B (CB) has been used as a tool to ascertain whether the increase in the rate of 3‐ O ‐methylglucose (3‐O‐MeG) uptake between the 8th and the 16th day of development in chick embryo fibroblasts could be attributed to an increase in the number of hexose transport carriers. There was a 2—3‐fold difference in glucose‐specific CB binding between the 8‐ and the 16‐day cells, a difference which is comparable to the previously reported differences in rates of 3‐O‐MeG uptake. We therefore suggest that glucose‐specific CB binding represents binding to the 3‐O‐MeG carrier and that the increase in the rate of 3‐O‐MeG uptake from the 8th and the 16th day of development is probably due to an increase in the number of hexose carriers.

Peanut agglutinin (PNA) has been shown to be insoluble at low temperatures. This cryoinsolubility has been studied by means of absorption spectroscopy, fluorescence, circular dichroism, and analytical ultracentrifugation. It was found to be dependent on pH, temperature, and protein concentration. No effects on dimer-tetramer equilibrium could be determined nor any conformational changes provoked by exposure of the PNA preparation to low temperatures. The dimer half-molecule apparently does not precipitate. The cryoinsolubility was partially reversible and totally inhibited in the presence of galactosides, the specific ligands of PNA. Their efficacy as inhibitors of cryoinsolubility was related to their affinity for the lectin. The effects of neutral salts and particularly inhibition of the insolubility by strongly chaotropic salts indicate that charge-charge interactions are of little importance and that hydrogen bonds and/or van der Waals interactions are most probably responsible for the formation of the cryoprecipitate.

Two neuraminidase activities have been found in normal human liver, one soluble and the other particulate and essentially bound to the lysosomes. With storage of the liver at −80°C, no loss of either type of activity was noted for up to a year. KCl extracts of the soluble enzyme were stable for 15 days when stored in liquid nitrogen. For the sediments (particulate or lysosomal), almost 100% of the initial activity was recovered after up to a month's storage at −20°C. With neuramine-lactose as substrate, the pH optimum was 4.0 for the soluble, particulate and lysosomal enzymes. The Km values were 8.0 × 10−3M for the soluble neuraminidase and 16.66 × 10−3M and 16.95 × 10−3M for the light and heavy lysosomal neuraminidases, respectively. The results suggest that normal human liver contains a single neuraminidase which may exist in either soluble form or incorporated in a membrane structure, depending on the condition of donor. The differences in behaviour could be explained in terms of “allotopic” properties.

The ability of fibroblasts from 8- and 16-day-old chick embryos to adhere to a substratum was altered by trypsin treatment. The consequences of this treatment were investigated on cell re-adhesion to the substratum and cell morphology in relation to the regeneration of cell surface glycoproteins as estimated by the incorporation of [3H]leucine and [14C]glucosamine.

With regard to the toxic effects of Ricinus lectin, neuraminidase-treated hepatoma cells have been found to be the most sensitive, and untreated hepatoma cells the least. Cells treated with neuraminidase and galactose oxidase exhibited an intermediate sensitivity. At 37 degrees C, the number of Ricinus lectin molecules bound to untreated, neuraminidase-treated and neuraminidase and galactose oxidase-treated cells required to being about 30% toxicity within 2 h was 15 . 10(5), 7.5 . 10(5) and 11.5 . 10(5) molecules/cell, respectively. This difference was rather small and suggests that the additional binding sites exposed following enzyme treatment were as efficient in mediating lectin toxicity as those present before enzyme treatment. Positive cooperativity was observed during Ricinus lectin binding to enzyme-treated cells at 37 degrees C and the apparent association constant increased with the increase of binding site occupancy. The binding sites on enzyme-treated cells appeared to be homogeneous since under different physical conditions (4 degrees C) the shape of the Scatchard plot could be altered in such a way as to produce a single line of slope. In contrast to enzyme-treated cells, untreated cells did not exhibit a positive cooperative process either at 37 degrees C or at 4 degrees C. We found that the toxicity of Ricinus lectin paralleled the irreversible specific binding of lectin, suggesting that only this was able to mediate the toxic effect. Our results are discussed in terms of the possible entry into the cells of Ricinus lectin and this occurs more rapidly in enzyme-treated than in untreated cells. This difference agrees with the sequence of events proposed: (i) Binding of Ricinus lectin; (ii) Clustering of lectin binding sites; and (iii) Endocytosis.

The glycoconjugates on Toxoplasma gondii membrane have been investigated in several ways, usually used for the study of cell surfaces: binding of |14C| lectins, Concanavalin A (Con A) and Wheat germ agglutinin (WGA) to parasites; parasites membrane labelling, using galactose oxidase/tritiated sodium borohydride (NaB|3H4|); determination of sialic acid amount, sensitive to the neuraminidase. Toxoplasma gondii binds Con A but did not bind significantly WGA. The affinity constant for Con A binding was on the same order of magnitude as for other cell systems, but the number of sites per parasite was low (3×104 sites per parasite). In contrast, with other cell systems, no surface incorporation of tritium was obtained using galactose oxidase/NaB|3H4|. Moreover the determination of sialic acid, sensitive to the neuraminidase gave a negative result. These results shown that Toxoplasma gondii surface do not contain all sugars, usually found in oligosaccharidic chains of membrane glycoconjugates and suggest the presence of low amount of surface glycoconjugates.

The proliferative capacity of Chick embryo fibroblasts was modified by Con A treatment. Con A decreased the growth of fibroblasts from young embryos (8 days), whereas the lectin stimulated the growth of fibroblasts from older embryos (16 days). This differential effect of Con A did not result from changes in cellular permeability to thymidine, but rather from Con A induced modifications of hexose transport. Changes in hexose transport would cause, as a response, parallel modifications in glycolysis and hence energy charge, which would alter proliferative capacity.

Les protéines de surface de lymphocytes humains normaux sont étudiées après hydrolyse trypsique douce des cellules. La liaison de ces composants membranaires à l'agglutinine de Ricinus sanguineus (poids moléculaire: 120.000 daltons) marqués au [14C] a été étudiée par chromatographie sur colonne de Sephadex G 200. Les quantités de lectine liée et non liée ont été calculées à partir de plusieurs gel-filtrations réalisées à concentration constante de récepteurs et concentration croissante de lectine. Ces valeurs ont été représentées selon la méthode de Scatchard. La constante d'affinité du système récepteur-lectine ainsi déterminée est 3,5 × 106 M−1. Comparativement, la liaison lectine-cellule entière a été étudiée. Dans ce système, la constante d'affinité apparente obtenue est 6 × 106 M−1. Connaissant le nombre de sites de liaisons de la lectine pour ses ligands, une évaluation du nombre de sites récepteurs en solution et sur les cellules entières peut être effectuée. Le rendement en récepteurs de l'hydrolyse trypsique peut ainsi être calculé.