Resistance or susceptibility of bifidobacteria to lysozyme and growth of bifidobacteria in human milk were tested. Susceptible bifidobacterial strains stopped their growth almost immediately after the addition of lysozyme (400 microg/ml), moderately susceptible strains exhibited reduced growth rate, and growth curves of resistant strains were not affected. Strains of human origin were more resistant to lysozyme than animal strains. While strains of B. bifidum grew well in human milk samples, the growth B. animalis strains was inhibited after inoculation to human milk. The resistance to lysozyme seems to be a promising criterion for the selection of new probiotic bifidobacterial strains.

Twenty-five bifidobacteria were isolated from feces of calves. Isolates were identified, and their functional properties and antimicrobial activity were determined. From 10 strains with suitable properties rifampicin-resistant mutants (RRBs) were prepared and mixture of RRBs was administered to 2-d-old calves. These strains were identified by sequencing as Bifidobacterium animalis ssp. animalis (6 strains), B. thermophilum (2 strains), B. choerinum (1 strain) and B. longum ssp. suis (1 strain). The control group was without probiotic treatment. Survival ability of administered bifidobacteria was monitored in fecal samples by cultivation on modified TPY agar supplemented with mupirocin, acetic acid, and rifampicin. Administered bifidobacteria survived in gastrointestinal tract of calves for at least 60 d. Other bacteria were also determined after cultivation using fluorescence in situ hybridization (FISH). Bifidobacteria and lactobacilli dominated in fecal microflora. Significantly lower amounts of E. coli and higher amounts of bifidobacteria and total anaerobes were found in the treated group relative to the control group.

Twenty-six bifidobacteria were isolated from faecal samples of lambs. The isolates were identified, functional properties (survival ability at low pH and bile conditions) and antimicrobial activities against potential pathogens were determined. From the isolates with suitable properties (13 strains) rifampicinresistant mutants were prepared by gradient plate techniques. This property enabled us to differentiate the administered organism from wild strains because resistance to rifampicin is rare among bifidobacteria. Rifampicin-resistant bifidobacteria (RRBifs) were administered to 3-days-old lambs in two trials. In the first trial the strain B. ruminantium L29 was applied to 3 lambs and was detected in faecal samples at high counts (6 log CFU/g on average) for one week. In the second trial 3 lambs received a "cocktail" of 12 strains and RRBifs survived in the intestinal tract at counts of about 6 log CFU/g for 25 days. The control group without probiotic treatment consisted of 6 animals. In both treated groups RRBifs dominated among bifidobacteria after their administration. Total bifidobacterial counts (5.64-7.32 log CFU/g) were significantly higher (P < 0.05) in treated groups compared to 2.31-2.85 log CFU/g detected in the control group during the first month of lamb life. Lactobacilli counts were also significantly higher (P < 0.05) in treated groups compared to the control. The administered bifidobacteria did not affect any other monitored bacterial groups. On the basis of in vitro test results, suitable probiotic bifidobacterial strains for lambs were chosen. Some of them survived for 30 days in the gastrointestinal tract of treated lambs, but no tested strain was able to colonise the lamb's tract permanently. The administration of bifidobacterial "cocktail" and consequent identification of the best survived strain seems to be an effective method for selection of potential probiotics.

The presence of raffinose series oligosaccharides (RSO) was determined by an enzymic method in three commercially available chicken feed mixtures. All feed mixtures contained RSO at a concentration of 2.1-2.2%. Soya meal was identified as the exclusive source of RSO. Subsequently, the bifidogenic effect of stachyose (main soya bean RSO) was also assigned on the growth of poultry intestinal bifidobacteria. Bifidobacteria were counted in chicken intestinal tract using cultivation and FISH methods. Four out of 6 bifidobacterial strains tested grew significantly better on stachyose than on glucose. It can be thus concluded that chicken feed mixtures naturally contain prebiotic oligosaccharides in the form of RSO in higher levels (>2%) compared with the concentration (usually up to 1%) recommended for artificially added prebiotics. Our results therefore indicate that there is no reason for the supplementation of chicken feed mixtures with prebiotics with bifidogenic properties.

The "cocktail" of human origin rifampicin-resistant bifidobacteria (RRBs) and RRBs from commercial products was administrated to 9 volunteers aged from 22 to 46 years and the survival ability in gastrointestinal tract of these strains was determined. Bifidobacteria represented 0-8 % of total anaerobes detected in gastrointestinal tract of volunteers before the administration of probiotic strains. After the administration of probiotics, bifidobacterial counts increased to 16 % of total bacterial counts. RRBs formed 9-44 % of total counts of bifidobacteria. Then, the counts of RRBs decreased at day 7 after administration, and they were not detected after 14 d. In our study, suitable probiotic bifidobacterial strains for human should be chosen on the basic of in vitro test but the results showed that no strain was able to colonize human tract permanently.