05; however, Rgp/cell and Kgp/ext in Fig. 3a were statistically different; P<0.01), but significantly low in 83K7 (8–36% of those of 83K5; P<0.01). These results show that the function of Sov is affected by the subtle find protocol structural difference between 83K6 and 83K7 (the C-terminals

are –Phe–His–His–His–His–His–His and –Phe–Arg–His–His–His–His–His–His), but not by the dramatic structural difference between wild-type W83 and 83K6 (the C-terminals are –Phe–Arg–Phe–Asn–Leu–Thr–Gln and –Phe–His–His–His–His–His–His). We suspect that a steric effect or the length of the C-terminal portion of Sov influences its function. Nevertheless, the expression of histidine-tagged Sov in 83K7 (Fig. 4a, lane 2) and 83K6 (lane 3) was similar to that in 83K5 (lane 1), suggesting that the hypoactivity of Sov renders the primary defect in the gingipain activity of 83K7. Finally, we clarified whether the C-terminal portion of Sov locates to the extracellular milieu. We investigated see more the effect of anti-histidine-tag IgG on the secretion of Arg-gingipains

by 83K5 and 83K6, both of which express histidine-tagged Sov. As a polar-effect control, we used 83K4, which carries the erm cassette like 83K5, but expresses Sov (Saiki & Konishi, 2007). The Arg-gingipain activities in the extracellular fractions were comparable among 83K4, 83K5, and 83K6 (P<0.05). As shown in Fig. 4b, the secretion of Arg-gingipains by 83K5 and 83K6 cells was significantly reduced (decreased to 84% and 79% of that by 83K4; P<0.01) by rabbit anti-histidine-tag IgG (50 μg mL−1). By contrast, the secretion of Arg-gingipains by 83K4, 83K5, and 83K6 cells was slightly affected by rabbit anti-histidine-tagged IgG (5 μg mL−1; P<0.05) or bovine IgG (50 μg mL−1; P<0.05). Although the inhibition

by anti-histidine-tag IgG was weaker than by anti-Sov32-177:2408-2499 antiserum acetylcholine (Fig. 1c), the results showed that the C-terminal portion of Sov may protrude into the extracellular milieu and may be involved in the modulation of Sov function. Sov contains a putative signal sequence, suggesting that Sov is a secreted protein (Saiki & Konishi, 2007). However, Sov shows no other conserved structural feature. Our investigation provides evidence that Sov is localized to the outer membrane and possibly participates in the secretion of gingipains. Nelson et al. (2007) reported that Flavobacterium johnsoniae SprA, a homologue of Sov, is likely an outer membrane protein. SprA is required for the gliding motility of F. johnsoniae (Nelson et al., 2007); however, the function of SprA has not yet been determined. In P. gingivalis, which is nonmotile, Sov appears to play a role in protein secretion. Perhaps F. johnsoniae SprA functions in the secretion of proteins required for gliding motility. We found that a five-residue section (Phe2495–Gln2499) in the C-terminal portion of Sov is essential for its function; this section may protrude into the extracellular milieu.