Impact Of Low Levels Of Tilmicosin On Human Colonic Microflora In Chemostat Models

Tilmicosin has been widely used as a new macrolide agent specifically in animal feed as a growth promoter. The determination of tilmicosin ADI is according to microbiology NOAEL data. However, the NOAEL data has a limitation to detect the number of total anaerobes, total enterobacteria and the proportion of resistant enterobacteria. There were no data whether low levels of tilmicosin have adverse effects on human intestinal microflora or not under a long-term administration. These adverse events may include changes in bacterial populations, modification of the metabolic activity of flora, selection of resistant bacteria, and perturbation of the barrier effect. It is necessary to conduct a more systematic study on the effect of low levels of tilmicosin on human intestinal microflora, so as to assess the safety of tilmicosin present ADI in China.In this study, we adopted a chemostat model for human stools to study the effect of low levels of tilmicosin on human microflora. Whilst the growth of bacteria reached steady state in each chemostat model, three chemostats were exposed to tilmicosin 0.436, 4.36, and 43.6μg/mL respectively for 7 days and one served as a no-drug control. We detected the changes in SCFA concentration, population of dominant bacteria and proportion of resistant facultative anaerobes before and during administration. The extent of changes in intestinal microflora of chemostat models was determined through a linear regression method. Each chemostat model was tested by three successive daily challenges of 108CFU/mL Salmonella typhimurium delivered in 1mL/d. The colonization resistance (CR) of each community was determined through the growth curves of Salmonella typhimurium that cultivated in the different chemostat model.Populations of Bacteroides fragilis and Enterococci were reduced and then increased in 43.6μg/mL tilmicosin group. Nevertheless, the variance of the numbers of Escherichia coli and Bifidobacteria were augmented, ranging from 6.61×105-6.31×106 CFU/mL and 1.02×106-1.23×107 CFU/mL to 3.02×105-1.62×107 CFU/mL and 5.13×105-1.35×107 CFU/mL respectively. This showed that the balance of microflora community disturbed significantly. In 4.36μg/mL tilmicosin group, the number of Bacteroides fragilis decreased 370 times, while the population of Enterococci reduced and then increased. On the other hand, the population of Escherichia coli and Bifidobacteria fell below 95% confidence line. There was no effect on the population of Enterococci and Escherichia coli in lowest-dose level (0.436μg/mL) chemostat model. Bifidobacteria level transiently declined within 24h after administration and the population of Bacteroides fragilis variation beyond the 95% confidence line.There was no effect on the susceptibility of Escherichia coli in the three groups (0.436,4.36 and 43.6μg/mL). However, the susceptibility of Enterococci declined significantly in two groups (0.436 and 43.6μg/mL). The proportion of Enterococci resistant to 16μg/mL tilmicosin has been apparently increased. There was no alteration on the molar concentration of acetic acid and butyric acid in the four groups. Nevertheless, the concentration of propionic acid has no change in the three groups (0,0.436 and 4.36μg/mL), but with a transient decline in 43.6μg/mL group.The number of Salmonella typhimurium was enumerated through cultural plates count with 4μg/mL ciprofloxacin. Results showed that there was no significant difference between the experimental groups compared with control group. The colonization barrier of human colonic flora in chemostat models was not disrupted after tilmicosin administration.Enterococcus faecalis strains isolated from four chemostat groups at 12d,17d and 20d, were used to detect esp, cylA, asa and gelE. After that broth dilution antimicrobial susceptibility test was adopted to determine the isolated strains'susceptibility for tilmicosin, tylosin and erythromycin, meanwhile ermB gene was detected by PCR. Finally the expression of virulence genes was determined through relative quantitative real-time PCR, then the relationship between resistance and the expression of virulence genes was investigated. Results showed that only 27 strains held gelE, the other virulence genes were not detected. Simultaneously, the results showed that cross resistance to the three kinds of macrolides emerged in these strains that isolated from high-dose level group at 17d and 20d, and accompanied with the detection of resistant gene ermB. It was found that the expression level of gelE was significantly increased 200 times in the strains isolated from high-dose level group at 20d. However the expression level of gelE pre-administration,17d strains and low and middle dose of 20d revealed no significant differences. In conclusion, three levels of tilmicosin have different effects on human colonic microflora in chemostat models. The population of Bacteroides fragilis was suppressed significantly in 4.36μg/mL group, and the proportion of resistant Enterococci was markedly influenced in 0.436μg/mL. These indicated that not only the present ADI of tilmicosin (40μg/kgbw) in our country has potential hazards on human health, but also the ADI (4μg/kgbw) in EU has adverse impact on human colonic microflora. It required that a lower ADI of tilmicosin should be established to keep human health security. On the other hand, we found that resistant strains would be induced or increased in high-dose group, and the expression of virulence genes of these strains would increase in accordance to exposure time. It suggested that, when studying the effect of low levels of antimicrobial drug on human colonic microflora, besides these factors that were recommended above, monitoring the expression of virulence genes of the opportunistic pathogen was also important. It helps us to understand the impacts of antimicrobial drugs on human colonic microflora more clearly. Thereby we could use the antimicrobial drug with more reasonability and safety.