4). The LacZ activity levels of cells recovered between 12 and 48 h were low, but increased markedly after 4 days, indicating that a certain incubation period was required for the induction. This delayed expression of LacZ activities was not observed by the constitutively lacZ-expressing

strain, 17616cox::lacZ (Nishiyama et al., 2010), which showed similar levels of LacZ activities at 12 h and 14 days after inoculation in the soil (data not shown). To determine whether the andA operon is essential to survive or grow in soil, the 17616ΔandAc was tested for its ability to proliferate and survive in the soil. The 17616ΔandAc and 17616cox::lacZ cells form white and blue colonies, respectively, on X-gal-containing agar plate. The 1 : 1 mixture of the Ixazomib two kinds of cells was inoculated into the soil sample, the cells PLX 4720 were recovered after various intervals, the CFUs of each type of cells g−1 of soil were counted (Fig. 5a), and the proportion of white colonies to the total (i.e. white plus blue) ones was also calculated (Fig. 5b). During the first 15 days, the CFUs of 17616ΔandAc remained at a low level, whereas the CFUs of 17616cox::lacZ increased. In Fig. 5b, the mutant cell ratio declined during the first week and reached a

steady low level, clearly showing that andA is necessary in the soil environment. We also tested two deletion mutants of ATCC 17616, 17616ΔpdyP and 17616Δsdh. Each mutant carried a chromosomal deletion of a genomic locus that was induced in the soil environment (Nishiyama et al., 2010). Although these mutants were originally included to investigate the

role of the deleted genomic locus in the soil, they showed no decreased CFUs and fitness, and they are controls here (as the results for the two mutants are essentially the same, the result for 17616Δsdh is not shown). Our present study clarified that the andAcAdAbAa gene cluster, predicted to encode anthranilate dioxygenase in B. multivorans ATCC 17616, is indeed involved in the catabolism of tryptophan and anthranilate, and that this gene cluster is under the control of two transcriptional regulators, AndR and Fur, in both the laboratory and soil environments. We RANTES also showed that this cluster plays a pivotal role in the proliferation in the soil environment. We showed that the andA operon is regulated by Fur, which is an iron-responsive transcriptional regulator. The anthranilate dioxygenase belongs to a class of dioxygenases, which require a [2Fe-2S] cluster in its active site (Batie et al., 1991), and it is not surprising that the iron-regulatory scheme operates on the andA operon. However, the effects of the iron-chelating agent and the disruption of fur gene on the transcriptional activity of andA operon were not remarkable (only at the level of twofold change) in B.