S. Typhimurium and E. coli harboring pRST98 formed thicker BF in vitro, compared with the isogenic strains not carrying pRST98. It was also observed that S. Typhimurium x3306 and x3337 had similar abilities to form BFs, which is inconsistent with the study of Teodo´ sio JS et al. We speculated the different plasmids and BF-producing systems may contribute to this inconsistency. We noticed that E. coli K12W1485/pRST98 had a weak ability to form BFs compared with Salmonella strains harboring pRST98. This heterogeneity in BF formation may arise because the synthesis of extracellular polymeric substances in Salmonella outcompetes that in E. coli in medium, as reported by Rong Wang et al. Regarding the heterogeneity in the promotion of BFs by conjugative plasmids, Røder HL et al. proposed that the different genetic backgrounds of the plasmidharboring hosts may account for different BF formation when the same plasmid was used. Our previous study demonstrated that in different genera, the conjugal transfer conditions of the pRST98 plasmid were different in vitro or in mice, and the resistance markers encoded by the same plasmid varied in different strains, which showed the diversity and complexity of the gene expression from the plasmid. Thus, the effects of BF formation by different plasmids in various hosts may demand specific analysis. In animal experiments, a tumor bearing mouse model was used to study the effects of pRST98 on BF formation in S. Typhimurium, which was used as a surrogate of S. Typhi because no animal model is available for S. Typhi infection. In the tumor-bearing mouse model, x3337lux/pRST98 was found preferentially in tumors with a considerably larger amount than x3337lux. The observation that solid tumors are treatable via bacterial infection was made previously. Colonization of bacteria on solid tumors could cause growth retardation or even the complete elimination of the tumors. pRST98 promoting host bacterial BF formation may have a therapeutic potential in fighting against tumors. Furthermore, our invasion study in vitro proved that bacteria in BFs showed a lower invasion ability compared with the corresponding planktonic form, which is consistent with the finding by Katja Crull et al. that BF-forming bacteria did not invade intracellularly in vivo after they established BFs. The intracellular invasion by Salmonella may be due to the differential expression of invasive genes on Salmonella pathogenicity island 1 induced by BF formation. Another animal model, a mouse urethral catheter model, was established to study the effects of pRST98 in E. coli on BF formation in vivo. E. coli K12W1485/pRST98 was found to form only discrete patchy BFs at 3 d post-implantation, while E. coli K12W1485 was not detected in tubes until 5 d post-implantation. E. coli K12W1485/pRST98 developed denser BFs at 5 d post-implantation, in line with bacterial titers recovered from established BFs on tubes. No histological changes were observed in the livers and kidneys of either group. When the implantation with tubes pre-incubated with E. coli was extended to 8 d or beyond, more BKM120 severe inflammation was observed. Significantly, S. Typhimurium x3337lux/pRST98 caused more severe inflammation in organs than x3337lux did. A similar phenomenon was observed for E. coli K12W1485/pRST98 and K12W1485. These results indicate that pRST98 aggravates the infection by promoting BF formation. Recently Rong Wang and Victoria J. Savage et al. demonstrated that the BF increases horizontal transfer of multi-resistant conjugative plasmids to plasmid-free bacteria compared to planktonic bacteria. Therefore, it seems that conjugative plasmids facilitate BF formation, and vice versa.