Parametric and Nonparametric Analyses of Repeated Ordinal Categorical Data

We compare two models for the analysis of repeated ordinal categorical data: the classical parametric model for means of scores assigned to the categories of the response variable and a nonparametric model based on relative effects derived from the marginal distribution functions of the response. An example in the field of Dentistry is used to illustrate and to compare the models. We also consider a simulation study to evaluate the type‐I error rates and the power of tests under both models in a balanced design setup. The simulation results suggest that both approaches behave similarly for equally spaced scores but may perform differently otherwise. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Differential binding of mannose-specific lectins to the carbohydrate chains of fibrinogen domains D and E

The interaction of fibrinogen with the mannose-specific lectins concanavalin A (ConA), its acetyl derivative (Ac-ConA) and Lens culinaris agglutinin (LcH) was studied. Both ConA and LcH interact specifically with individual fibrinogen B beta and gamma chains and with denatured fragments D and E. However, analysis of the binding data shows that four moles of Ac-ConA are bound per mole of fibrinogen with two sets of binding sites (Kd1 = 2.4 microM and Kd2 = 16.6 microM; n1 = n2 = 2) while only two moles of LcH are bound per mole of fibrinogen (Kd = 2.6 microM). Ultracentrifugation studies are also in agreement with the presence in the fibrinogen molecule of two and four binding sites for LcH and Ac-ConA, respectively. No aggregates of fibrinogen formed through LcH or Ac-ConA linkages are observed. The use of a crosslinking reagent and ultracentrifugal analysis of the lectin-fibrinogen fragments D1 and E complexes indicated that ConA, as well as Ac-ConA, interact with both fragments D and E while LcH interacts only with fragment D. Furthermore, the binding of ConA to both D and E domains in the intact fibrinogen molecule is clearly demonstrated by using a bifunctional reagent. The bivalent character of ConA tetramers may be misinterpreted as a lack of accessibility of the lectin to two of the four carbohydrate chains of fibrinogen. The differential binding of LcH and ConA to the carbohydrate chains of fibrinogen can be related to a different exposure of the oligosaccharide in D and E fragments and domains and to the different requirements of both lectins for their binding to glycoproteins.     

Effect of lectin-binding to fibrinogen D and E domains on coagulation and fibrinolysis

The effect on fibrinogen coagulation and fibrinolysis of the mannose-specific lectins concanavalin A, its acetyl derivative and Lens culinaris agglutinin was studied. Concanavalin A and acetyl-concanavalin A, which bind to the four carbohydrate chains of fibrinogen, and L. culinaris agglutinin, which only binds to the carbohydrate present in fibrinogen D domains, has the same effect on the coagulation rate: an inhibition at low lectin concentrations and an increase at high concentrations. On the other hand, L. culinaris agglutinin does not alter fibrin crosslinking while acetyl-concanavalin A produces a slight inhibition of both gamma-gamma and alpha-polymer formation. However, this effect is very small when compared with the clear inhibitory effect produced by concanavalin A. Concanavalin A and acetyl-concanavalin A have an inhibitory effect on the rate of fibrin clot lysis proportional to the lectin concentration. Nearly 100% inhibition was obtained when two lectin-binding sites were occupied by either concanavalin A or acetyl-concanavalin A. However, L. culinaris agglutinin has a clearly weaker effect and more than 50% inhibition was not observed. The comparative study of the effect of the three lectins on fibrinolysis as well as on the formation of fibrinogen aggregates suggests that the inhibitory effect of concanavalin A and acetyl-concanavalin A is primarily due to their binding to the carbohydrate chains of fibrinogen E domain.     

Analysis of polyadenylated RNA from brain synaptosomes and mitochondria

We have isolated RNA from sheep brain synaptosomes and mitochondria separated by an aqueous two-phase system composed of dextran and poly(ethylene glycol). RNA was fractionated through oligo(dT)-cellulose columns and analyzed by electrophoresis through agarose slab gels containing methylmercuric hydroxide and stained with ethidium bromide. The electrophoretic patterns of the poly(A)-containing RNA fraction from synaptosomes and mitochondria are very similar although some high molecular weight RNA species, clearly visible in the synaptosomal fraction, are scarcely detected in the mitochondrial preparations. The electrophoretic analysis of a cleaner RNA preparation from digitonin-treated free mitochondria (mitoplasts) showed that all the poly (A)-RNA species of the synaptosomal preparation are also present in mitoplast. These results strongly suggest that all the discrete poly(A)-RNA species identified in brain synaptosomes are of mitochondrial origin.     

Subcellular localization and glycoprotein nature of the invertase from the fission yeast Schizosaccharomyces pombe

The subcellular localization of the enzyme invertase in Schizosaccharomyces pombe cells, both repressed and derepressed for synthesis of the enzyme, was studied. Most of the invertase was found to be located outside the plasma membrane and only a small percentage was found to be associated to membranes. A substantial portion of the external enzyme remained firmly bound to cell-wall material.All of the invertase recovered in soluble form from cellular extracts reacted with concanavalin A and with the lectin from Bandeiraea simplicifolia seeds, indicating the presence in the enzyme of a carbohydrate moiety which probably contains terminal mannosyl (or structurally related) and galactosyl residues.The possibility of the presence of two different forms of invertase in S. pombe was considered. An intracellular, soluble form of invertase, devoid of carbohydrate, similar to the small invertase of the budding yeast Saccharomyces cerevisiae, was not found in S. pombe. However, the Michaelis constant for sucrose of the enzyme present in repressed cells was smaller than that of the invertase synthesized under derepressing conditions, although this difference could also be the result of a different pattern of glycosylation of the invertase synthesized under different growth conditions.     

Purification of plasma membranes from mouse parotid gland and membrane reorganization in response to isoproterenol

Two highly purified plasma membrane fractions have been obtained from mouse parotid glands by a combination of differential centrifugation and isopycnic centrifugation in discontinuous sucrose gradients. The membranes were characterized by enzymic, chemical and morphological criteria. The effect of isoproterenol, which induces parotid acinar cells to proliferate, upon sialic acid and five different enzyme activities located in the plasma membrane phosphodiesterase (EC 3.1.4.1), Mg2+-ATPase (EC 3.6.1.4), leucine aminopeptidase (EC 3.4.1.1), protein kinase (EC 2.7.1.37) and sialyltransferase (EC 2.4.99.1), were quantified along the cell cycle. Plasma membrane sialic acid content falls 30% within 30 min and remains depressed for at least 6 h with the major restoration towards normal levels occurring between 12 and 16 h later. In contrast multiple daily isoproterenol injections lead to a more than 2-fold elevation of sialic acid content. Sialyltransferase activity rises 2-fold by 12 h after isoproterenol treatment and then rapidly falls. This enzyme has a pH optimum of 6.5, requires a divalent cation for activity and is inhibited by Triton X-100. Other enzyme activities showed markedly different changes after isoproterenol stimulation, either increasing, decreasing or remaining unaltered. These continuous functional modifications suggest an active role of the plasma membrane in the control of the proliferative cycle.     

Tubulin and tubulin-colchicine complex bind to brain microsomal membrane in vitro

Brain tubulin was labeled in vitro by post-translational incorporation of [14C]-tyrosine or in vivo by intra-cranial injection of [3H]-leucine. The labeled protein was purified by ion-exchange chromatography. After incubating at 37 degrees C with a microsomal membrane preparation from rat brain, part of the labeled soluble tubulin became sedimentable at high-speed centrifugation. This was independent of the native configuration of tubulin, the state of tyrosination of the COOH-terminus, or the presence of 100 microM colchicine in the mixture. In addition, the double-labeled tubulin-colchicine complex obtained from the binding of [3H]-colchicine to [14C]-tyrosinated tubulin, bound to the membrane preparation to the same extent as [14C]-tyrosinated tubulin. The data show that either tubulin or the complex resulting from its binding to colchicine distributed between the soluble and the membrane fractions when mixed at 37 degrees C with a microsome preparation. Seemingly, the site for colchicine binding to tubulin needs not to be free for the protein-membrane association.