When the σS levels in pgsA3ΔcpxA and pgsA3ΔcpxR double mutants were examined, the high level of σS in pgsA3 mutant cells was found to be considerably reduced (Fig. 2d), indicating that the activation of the Cpx system is one important cause for the high level of σS. In order to clarify how the system affects σS, we examined the activities of clpP′-lacZ and clpX′-lacZ in the double mutants. The activities of these transcriptional fusions recovered after disruption of the Cpx system from the very low levels in pgsA3 mTOR inhibitor mutant cells as expected, although not completely (Fig. 2b). These results indicate that the activated Cpx system increases

σS levels by contributing to the repression of clpPX in pgsA3 mutant cells. Does the PI3K Inhibitor Library cost Cpx system repress clpPX through rpoE, rpoH, and rpoD, and to what extent does it control their levels? Microarray analyses suggested that in pgsA mutant cells, the expressions of rpoE and rpoH genes and the genes under the control of σE and σH are reduced, but the level of σD is not (Nagahama et al., 2007). In fact, real-time PCR analysis of the mRNA levels of these sigma factors indicated that the mRNAs of rpoE and rpoH in pgsA3 mutant cells were reduced to 1/40 and 1/18 of those in pgsA+ cells,

respectively, whereas the mRNA level of rpoD was almost the same as in wild-type cells, supporting the results of the microarray analyses (Fig. 4). Further examination of the mRNA levels in ΔcpxR pgsA double mutant cells indicated that the low level of rpoE mRNA recovered to 1/10, but that rpoH mRNA was not much changed (to 1/14) (Fig. 4). These results indicate that the repression of rpoE in the filipin pgsA3 mutant cells can be partly attributed to the activated Cpx system (see Fig. 5). There must be an unknown component in the repression of rpoE, independent of the Cpx system. The results also imply that the repression of rpoH is independent of the

Cpx system. The transcription of clpPX from the σD promoter may also be repressed in the presence of increased σS, which probably contributes to the repression by competing with σD for RNA polymerase core enzyme, because σS has a high affinity for the core enzyme (Maeda et al., 2000). To elucidate the intriguing roles that these sigma factors play in the repression of clpPX in pgsA mutant cells, further analysis of the cellular levels of the sigma factors and of each promoter of clpPX will be required. Figure 5 summarizes our ideas about the regulatory pathways that lead to σS accumulation in pgsA mutant cells. In order to fully understand the molecular mechanism of the accumulation of σS in pgsA mutant cells, detailed examination of the signal transduction systems that respond to acidic phospholipid deficiency will also be necessary. We thank Drs Robert Simons, Michele Garsha, Christophe Merlin, Kouji Busujima, Koichi Inoue, and Hiroshi Matsuzaki for the gifts of bacterial strains and suggestions, and Dr Kan Tanaka for the antiserum against σS.