Effect of a fermented milk combining Lactobacillus acidophilus Cl1285 and Lactobacillus casei in the prevention of antibiotic-associated diarrhea: a randomized, double-blind, placebo-controlled trial.

BACKGROUND: Antibiotic-associated diarrhea is an important problem in hospitalized patients. The use of probiotics is gaining interest in the scientific community as a potential measure to prevent this complication. The main objective of the present study was to assess the efficacy and safety of a fermented milk combining Lactobacillus acidophilus and Lactobacillus casei that is widely available in Canada, in the prevention of antibiotic-associated diarrhea. METHODS: In this double-blind, randomized study, hospitalized patients were randomly assigned to receive either a lactobacilli-fermented milk or a placebo on a daily basis. RESULTS: Among 89 randomized patients, antibiotic-associated diarrhea occurred in seven of 44 patients (15.9%) in the lactobacilli group and in 16 of 45 patients (35.6%) in the placebo group (OR 0.34, 95% CI 0.125 to 0.944; P=0.05). The median hospitalization duration was eight days in the lactobacilli group, compared with 10 days in the placebo group (P=0.09). Overall, the lactobacilli-fermented milk was well tolerated. CONCLUSION: The daily administration of a lactobacilli-fermented milk was safe and effective in the prevention of antibiotic-associated diarrhea in hospitalized patients.

A 91-day feeding study in rats with caprenin.

Caprenin, a randomized triglyceride primarily comprising caprylic (C8:0), capric (C10:0), and behenic (C22:0) acids, was administered in a semi-purified diet to weanling Sprague-Dawley rats (25/sex/group) at dose levels of 5.23, 10.23 or 15.00% (w/w) for 91 days. Corn oil was added at 8.96, 5.91 and 3.00%, respectively, to provide essential fatty acids and digestible fat calories. Corn oil alone (12.14%) and a blend of medium-chain triglyceride (MCT) oil plus corn oil (11.21 and 3.13%, respectively) served as controls. All diets were formulated to provide about 4000 kcal/kg of diet and 26.8% of digestible calories from fat by assuming that corn oil, MCT oil, and caprenin provided 9, 7 and 5 kcal/g, respectively. Survival, clinical signs, body weight, feed consumption, feed efficiency, organ weights, organ-to-body-weight ratios, organ-to-brain-weight ratios, haematological values and clinical chemistry parameters were evaluated in all groups. Histopathology of a full complement of tissues was evaluated in the corn oil and MCT oil control groups as well as the high-dose caprenin group. Additional rats (n = 5/sex/group) were included in the study to determine whether there was marked storage of C22:0 in heart, liver or perirenal fat at the end of the 91-day feeding period. No significant differences in body weight gain were measured with the balanced caloric diets, although feed conversion efficiency was reduced in the high-dose caprenin group. No adverse effects from the ingestion of caprenin were detected, nor were significant amounts of C22:0 present in the fat extracted from the selected fat depot sites. These results establish a no-observable-adverse-effect level (NOAEL) of more than 15% (w/w) caprenin in the diet (or more than 83% of total dietary fat), which is equal to a mean exposure level of more than 13.2 g/kg/day for male rats and more than 14.6 g/kg/day for female rats.

Decreased in vivo survival of hydrogen peroxide-damaged baboon red blood cells.

In this study we attempt to establish the consequence of in vitro hydrogen peroxide (H2O2)-induced membrane damage as manifested by spectrin-hemoglobin (Sp-Hb) complex formation and decreased red blood cell (RBC) deformability to in vivo RBC survival in baboons. After exposure to 135 to 581 mumols/L H2O2 and reduction with dithiothreitol (DTE), baboon RBCs were infused into the animal, and the fraction of cells remaining in circulation after 24 hours and the lifespan of surviving cells were quantitated. In a dose-dependent fashion, a positive correlation was observed between in vitro membrane alterations and the 24-hour in vivo survival. While 12% of the control cells were removed from circulation in 24 hours, 23% were removed after treatment with 339 mumols/L H2O2, and 36% following exposure to 581 mumols/L H2O2. Pretreatment with carbon monoxide before exposure with H2O2 increased the survival of oxidized RBCs. RBCs not removed from circulation in the first 24 hours had a normal lifespan. Moreover, by selectively isolating biotin-labeled, peroxide-treated cells that survived the first 24-hour posttransfusion period, a significant decrease in Sp-Hb crosslinking was observed in these cells. These results suggest that a subpopulation of cells sensitive to oxidation were removed during the first 24 hours. To identify this population, the survival of density-fractionated RBCs exposed to oxidant stress was quantitated. No differences in either the 24-hour survival or RBC life span were observed between untreated low-density (MCHC less than or equal to 32g/dL) and high-density cells (MCHC greater than or equal to 37g/dL). However, striking differences were noted after treatment with 339 mumols/L H2O2, with the 24-hour survival of high-density cells showing a marked decrease compared with low-density cells. These data support our hypothesis that during peroxidative membrane damage, Hb oxidation initiates a sequence of events resulting in skeletal changes that lead to membrane alterations and, eventually, in vivo destruction, and that the dense, dehydrated cells are more susceptible to oxidant damage.

The relationship between in vivo generated hemoglobin skeletal protein complex and increased red cell membrane rigidity.

In vitro induced oxidative damage to normal human RBCs has previously been shown to result in increased membrane rigidity as a consequence of the generation of a protein complex between hemoglobin and spectrin. In order to determine if in vivo generated hemoglobin-spectrin complexes may play a role in increased membrane rigidity of certain pathologic red cells, we measured both these parameters in membranes prepared from hereditary xerocytosis (Hx), sickle cell disease (Sc), and red cells from thalassemia minor (beta thal). Membranes were prepared from density-fractionated red cells, and membrane deformability was measured using an ektacytometer. Hemoglobin-spectrin complex was determined by sodium dodecyl sulfate (SDS)-polyacrylamide gel analysis, as well as by Western blot analysis using a monoclonal antibody against the beta-subunit of hemoglobin. For these three types of pathologic red cells, progressive cellular dehydration was associated with increased membrane rigidity and increased content of hemoglobin-spectrin complex. Moreover, the increase in membrane rigidity appeared to be directly related to the quantity of hemoglobin-spectrin complex associated with the membrane. Our findings imply that hemoglobin-spectrin complex is generated in vivo, and this in turn results in increased membrane rigidity of certain pathologic red cells. The data further suggest that oxidative crosslinking may play an important role in the pathophysiology of certain red cell disorders.

The role of membrane protein sulfhydryl groups in hydrogen peroxide-mediated membrane damage in human erythrocytes.

The formation of spectrin-hemoglobin complex following treatment of red cells with hydrogen peroxide (H2O2) has previously been shown to be associated with alterations in cell shape, decreased membrane deformability and increased recognition of modified cells by anti-IgM immunoglobulin in a phagocytic assay by monocytes. Prior treatment with carbon monoxide completely inhibited the H2O2-associated membrane changes, indicating a role for oxidized hemoglobin in the complex formation. Also, in a cell-free system, blockage of sulfhydryl (SH) groups on purified spectrin by N-ethylmaleimide significantly reduced the complex formation, suggesting a role for SH groups of spectrin in crosslinking process. The present study was undertaken to examine the role of SH blockade by N-ethylmaleimide on intact red cells undergoing oxidative damage. Pretreatment of erythrocytes with N-ethylmaleimide at concentrations ranging from 0.1 to 0.2 mM resulted in decreased lipid peroxidation and spectrin hemoglobin crosslinking. Moreover, pretreatment with N-ethylmaleimide resulted in less marked alterations in cell shape and membrane deformability as well as reduced recognition of peroxidized cells by antiglobulin serum. N-Ethylmaleimide treatment had no effect on methemoglobin formation. Studies with 14C-labeled N-ethylmaleimide showed that over 50% of N-ethylmaleimide was incorporated into spectrin. Pretreatment of cells with higher concentrations of N-ethylmaleimide (over 0.2 mM) was associated with membrane dysfunction independent of H2O2. These results imply that blocking of reactive SH groups leads to reduced interaction of spectrin with oxidized globin. These data, along with our prior observations, indicate that SH groups on spectrin play an important role in hemoglobin oxidation-induced formation of spectrin-hemoglobin complex and the resultant deleterious effects on membrane properties.

Crosslinking of isolated cytoskeletal proteins with hemoglobin: a possible damage inflicted to the red cell membrane.

Crosslinking of isolated red cell membrane cytoskeletal proteins and hemoglobin mediated by H2O2 was studied. The products of spectrin and hemoglobin interaction were demonstrated electrophoretically to be high-molecular-weight polypeptides crosslinked by nondisulfide covalent bonds. The molecular weight of the protein bands correlated with various combinations of spectrin and hemoglobin chains and the relative amount of the different products was dependent on the molar ratio of the interacting proteins. Free hemin caused spectrin crosslinking as well, but globin in the absence of hemin was inactive. Since the H2O2-mediated reaction resulted in reduction of the spectrin tryptophan fluorescence, the latter was used to monitor the reaction progress under various conditions. Both oxyhemoglobin and methemoglobin were found to be most efficient, whereas cyanmethemoglobin and hemichrome were relatively inactive. Analysis of the data implied that tryptophan oxidation as well as spectrin conformational changes follow an iron-induced crosslinking of the interacting proteins. Actin, the second major protein in the red cell cytoskeleton, behaved similarly to spectrin. The intrinsic fluorescence intensity of both G- and F-actin was decreased upon addition of H2O2 to the mixture of hemoglobin and each of the actin forms. SDS-polyacrylamide gel electrophoresis revealed that G-actin crosslinked one or two hemoglobin chains. F-actin-hemoglobin interaction induced by H2O2 produced very high aggregates that could not penetrate the gel. It is suggested that crosslinking of cytoskeletal proteins in red cells containing membrane-associated hemoglobin provides a rationale for the loss of membrane flexibility.

Antiglobulin serum mediated phagocytosis of normal senescent and oxidized RBC: role of anti-IgM immunoglobulins in phagocytic recognition.

Fresh human monocytes usually do not recognize normal RBCs; however, in our newly developed assay antiglobulin-opsonized normal RBCs were phagocytized. Both anti-IgG and anti-IgM fractions present in the antiglobulin serum were involved but the major opsonin was anti-IgM. The anti-IgM opsonized mainly senescent RBCs and therefore could be used to discriminate young from senescent RBCs. The antiglobulin serum and monospecific anti-IgM increased opsonization of in vitro oxidized and desialylated RBCs, whereas trypsin-treatment of RBCs decreased phagocytosis. The material removed by trypsin from the RBCs surface inhibited the antiglobulin and monospecific anti-IgM phagocytic assay supporting the view that membrane associated elements crossreacted with anti-IgM. These results suggest that both internal cellular events and external removal of sialic acid play a role in the emergence of non-IgG covered epitopes on the surface of senescent and oxidized erythrocytes.

The effect of temperature on single-column ion-chromatography of metal ions.

Temperature is shown to be a valuable parameter for optimizing single-column ion-chromatographic separations of metal ions. With a perchloric acid eluent, retention times of bivalent metal ions decrease with an increase in system temperature, but with doubly protonated amines as eluents the retention times increase with an increase in temperature. Consequently, increasing the system temperature improves separations of bivalent metal ions when p-phenylenediamine dilhydrochloride is used as the eluent. Both conductivity detection and post-column reaction followed by spectrophotometric detection are suitable detection methods at above-ambient temperatures.

Demonstration of haemoglobin associated with isolated, purified spectrin from senescent human red cells.

Employing a direct and sensitive radioimmunoassay (RIA) we have confirmed the presence of haemoglobin associated with isolated, purified spectrin from senescent red cells. Haemoglobin associated with spectrin occurs in the highest amount in cells with an MCHC greater than 36 g/dl and is approximately 3% of the total spectrin extract. Spectrin from the young cells had the least haemoglobin, while an intermediate amount was found in unfractionated, whole red cells. The RIA results were in close approximation with estimation of the haemoglobin-spectrin complex obtained by carefully integrating the Coomassie blue stain profiles from 4% SDS PAGE in densitometric scans from isolated spectrin.

Effect of hydrogen peroxide exposure on normal human erythrocyte deformability, morphology, surface characteristics, and spectrin-hemoglobin cross-linking.

To further define the conditions for forming spectrin-hemoglobin cross-linking in human erythrocyte membranes and to examine its possible effects on membrane function, we incubated normal human erythrocytes for up to 3 h in concentrations of H2O2, varying from 45 to 180 microM, in an azide phosphate buffer, pH 7.4. The chemical changes observed indicated that methemoglobin formation occurred early and at a low concentration (45 microM). Morphologic changes characterized by increased echinocyte formation occurred in a dose-dependent fashion. In addition, decreased cell deformability commensurate with increased membrane rigidity was found. Finally, an increase in cell recognition as determined by monocyte phagocytosis and adherence in vitro, as well as decreased phosphatidylcholine accessibility to bee venom phospholipase A2, was found in H2O2-treated erythrocytes compared with controls. Both of these latter changes were closely correlated with the extent of spectrin-hemoglobin cross-linking. In addition to these protein-mediated interactions, lipid peroxidation also occurred after H2O2 exposure, as shown by generation of fluorescent amino propene derivatives. The addition of the antioxidant, butylated hydroxytoluene, decreased the fluorescent derivatives, but did not prevent the effects on membrane function. This suggests that lipid peroxidation, though present, was not necessary for the membrane changes found. In contrast, spectrin-hemoglobin aggregation and the alterations in membrane function were completely prevented by prior exposure of the erythrocytes to carbon monoxide.

Separation and determination of aluminium by single-column ion-chromatography.

A quick, reliable method for the determination of Al(III) in the presence of other metal ions is presented. A Chromatographic system consisting of a low-capacity cation-exchange column, an eluent of diprotonated p-phenylenediamine, and a conductivity detector was used to measure the retention times for various cations. During the course of this work, it was found that Al(III) was eluted later than most bivalent metal ions but earlier than other tervalent metal ions. Therefore the concentration of eluent was adjusted so that an early sharp peak was obtained for Al(III) and the bivalent metal ions were eluted as a group. Through analysis of an NBS standard, as well as of solutions containing Al(III) and other metal ions, the method was shown to be precise, accurate and rapid for determination of Al(III) without interference from common bivalent metal ions.

Availability of the inner surface of the red cell membrane to antibodies in digitonin ghosts. Studies on D-negative cells with anti-D.

Studies were done to determine whether antibodies can detect antigens on the inner red cell stromal membrane. When albumin, anti-A, anti-D, or IgG were combined with varying volumes of intact O, Rh-negative red cells (RBCs), these proteins diluted, on the average, to 76 percent of the total volume, which was almost identical to the volume of the supernatant fluid (mean 77%). In contrast, when the protein markers were combined with packed stroma prepared from O, Rh-negative RBCs by digitonin, they diluted, on the average, to 94 percent of the total reaction volume rather than to the volume of the supernatant (mean 70%). Similar dilution was observed in the fluid volume harvested from stromal suspensions by ultracentrifugation, indicating that the protein markers occupied the total fluid volume of the reaction mixture (inside and outside the stroma). Nonspecific adsorption of the protein markers to stroma did not occur since their dilution was unaffected by doubling the stromal volume and since they could be totally recovered in the fluid harvested by ultracentrifugation. These data indicate that antibodies easily traverse the membrane of RBC digitonin stroma but not the membrane of intact RBCs. Therefore, antibodies may be used to detect antigenic determinants on the inner stromal membrane. When anti-D was incubated with Rh-negative stroma, we did not observe consumption of this antibody. Thus, our data did not indicate that the D antigen is present on the cytoplasmic membrane of Rh-negative RBCs.

Modulation of mononuclear phagocyte cytotoxicity by alpha-tocopherol (vitamin E).

The in vitro effect of colloidal suspensions of alpha-tocopherol (alpha-T) on phorbolmyristate-acetate (PMA)-induced monocyte cytotoxicity and on antibody-dependent monocyte cytotoxicity (ADCC) was studied. We observed that 1) in the presence of alpha-T, the inhibition was twice as high in the PMA-induced assay than in ADCC; 2) monocytes preincubated with alpha-T were inhibitory in both assays but much less in ADCC, and 3) target erythrocytes preincubated with alpha-T decreased the cytotoxicity in the PMA-induced assay only. Since alpha-T preincubated monocytes showed a decreased release of H2O2 but not of O2-, we concluded that one of the mechanisms by which alpha-T decreased cytotoxicity could be decreased release of H2O2. Whereas the role of H2O2 was documented in the PMA-induced cytotoxicity, in ADCC non-oxidative injury seems more important. This is supported by 1) lesser inhibition of the assay with alpha-T preincubated monocytes; 2) lack of protection with alpha-T preincubated erythrocytes, and 3) mild inhibition with protease inhibitor.

Properties and characterization of vesicles released by young and old human red cells.

We have presently demonstrated morphologic differences between young and senescent red cells following 18 h of metabolic depletion in vitro. Young and old red cells both form echinocytes, whereas only young cells demonstrated myelin forms or microspheres. Furthermore, vesicles were released in greater quantities into the cell-free supernatant from young cells. Isolated vesicles from both young and old red cells contained lipids, intrinsic membrane proteins (especially band 3), and haemoglobin, and they were also enriched in acetylcholinesterase (AChE). Young cells produced more vesicles than old cells but the composition of the low density vesicles was similar except that haemoglobin-spectrin complex was found exclusively in vesicles from young cells. Oxidation of young red cells prior to metabolic depletion prevented both myelin formation and vesicle release.

Increased sensitivity of isolated alpha subunits of normal human hemoglobin to oxidative damage and crosslinkage with spectrin.

After peroxidation, alpha and beta subunits of normal human hemoglobin demonstrated a significant differential reactivity in their ability to form methemoglobin subunits and irreducible crosslinkages with spectrin. The alpha subunits formed crosslinks with spectrin in the absence of exogenous hydrogen peroxide, whereas both the beta subunit and the intact hemoglobin molecule required a minimum of 40 and 4 microM peroxide, respectively, in order to form these crosslinks. Changes in the amount of methemoglobin occurred at much higher concentrations of hydrogen peroxide for the beta hemoglobin subunit (100 microM H2O2) than for the alpha subunit (0.1 microM H2O2). A possible explanation for the existing reactivity between each of the two hemoglobin subunits and spectrin is considered and discussed. In our notation, alpha subunit = alpha SH and the beta subunit = beta SH.

Spectrin-haemoglobin crosslinkages associated with in vitro oxidant hypersensitivity in pathologic and artificially dehydrated red cells.

Protein crosslinkages are apparent at 215 000-250 000 daltons in electrophoregrams of membranes from hydrogen peroxide treated erythrocytes (British Journal of Haematology, 48, 435, 1981). Hydrogen peroxide is also capable of inducing crosslinkages of identical molecular weights in stage I and II (red) ghosts and in a mixture of purified spectrin and haemoglobin, but not in white ghosts or in either spectrin or haemoglobin alone. Autoradiographic studies using 14C-methaemoglobin and 32P-spectrin confirm the involvement of spectrin and haemoglobin in this reaction. The alpha chains of both proteins are more reactive than the corresponding beta chains: 3 times more reactive in the case of spectrin and 10 times more reactive in haemoglobin. The reaction is almost totally inhibited by NaCN and partially inhibited by N-ethylmaleimide. Direct addition of malondialdehyde to a spectrin-haemoglobin mixture does not result in protein crosslinkage. Metabolic depletion (40 h), in vivo ageing and sucrose dehydration of fresh, normal cells enhance the reaction considerably, whereas in vitro rehydration of xerocytes normalizes their H2O2 sensitivity.

Evaluation of methods using adherence to substrate and density gradient for the isolation of human monocytes.

A new method for monocyte isolation based on cell adherence to gelatin-coated plastic and dislodgement of the adhering monocytes at low temperature was compared with 5 other methods based on cell adherence to substrate and density gradient separation. All methods produced yields of monocytes ranging approximately between 50-70% with about the same degree of purity (less than 90%) except for the method using Percoll density gradient centrifugation where the purity of monocytes was about 80%. When lidocaine at different concentrations was used for cell dislodgement or Percoll density gradient for separation, phagocytosis, Fc receptor function and cytotoxicity were adversely affected, unlike in methods using EDTA or low temperature for dislodgement of the adherent cells. In monocyte chemotaxis assays the rate of migration was affected but not the number of migrating cells for all the isolation procedures investigated. Cell spreading function was apparently well maintained only when gelatin coated plastic was used for adherence and low temperature for cell dislodgement. These data indicate that the newly described method, similar to methods using EDTA for cell dislodgement, yielded relatively intact monocytes but unlike the latter method with better preserved cell spreading. Thus, this method can be considered for standardization to obtain pure monocyte populations from peripheral blood which then can be submitted for comprehensive biochemical and physiologic studies.