Binding of FimR to the promoter region of mfa1 
The previous study has shown that the mechanism of FimR activation of the fimA gene involves a regulatory cascade (Nishikawa et al., 2004).
It was postulated that different mechanisms might be involved in FimR-mediated mfa1 expression, since expression regulation of mfa1 by FimR was not controlled as tightly as observed for fimA expression.
One possibility is that FimR modulates mfa1 expression by directly binding to the promoter region of mfa1.
To test this hypothesis, electrophoretic mobility shift assays were performed.
The mfa1 promoter (positioned from +18 to -138), fimA promoter (positioned from -22 to-190) (Xie & Lamont, 1999) and mfa1 coding DNA (positioned from +1253 to +1373) were generated by PCR with the 5' biotin-labeled primers (Table 1).
The recombinant FimR (rFimR) was expressed in pET expression system and purified from E. coli.
The rHGP44 protein, a binding domain of P. gingivalis gingipains (Xie et al., 2006), expressed in the same system and purified by the same procedures as rFimR was used as a control.
Cold competitor chase experiments with a 100-fold excess of unlabeled DNA probe as a specific competitor were also used to demonstrate the specificity of rFimR binding.
As shown in Fig. 3, the DNA fragment of the mfa1 promoter region was shifted in the presence of the rFimR.
Retarded mfa1 promoter-rFimR complex was detected with as little as 10 pmol muL-1 rFimR (Fig. 3).
As the concentration of rFimR increased, the retarded protein-DNA complex became evident, with complete loss of the mfa1 promoter DNA.
The unlabeled mfa1 promoter fragments effectively competed with the labeled fragment, suggesting a specific interaction between rFimA and the mfa1 promoter.
To investigate the role of phosphorylation of FimR in its binding to the mfa1 promoter region, EMSA experiments were also performed with the phosphorylated rFimR.
No significant difference was detected in the level of DNA binding between the phosphorylated rFimR and unphosphorylated rFimR (data not shown).
In agreement with a previous report (Nishikawa et al., 2004), rFimR did not bind to the fimA promoter region, suggesting that regulation of fimA expression by FimR is through a different mechanism.
Moreover, incubation of rHGP44 with mfa1 promoter fragment did not retard the DNA movement in polyacrylamide gel.
There was also no DNA shift detected when rFimR was incubated with the coding region of mfa1.
These data clearly show that FimR protein can bind specifically to the mfa1 promoter region, acting as an activator of mfa1 transcription.
EMSA experiments were also performed to examine whether the rFimR binds to the other promoter region identified by Park et al. (2006).
The biotin-labeled DNA fragment corresponding to this promoter region did not shift in the presence of the rFimA protein (data not shown), suggesting that only the promoter identified here is involved for mfa1 expression mediated by FimR.
