Supplemental Results

The lack of P∼RR468 detection in phosphotransfer assays is due to HK853-catalyzed P∼RR468 dephosphorylation. When P∼HK853 (either the cytoplasmic portion or the full-length form) is mixed with an equimolar amount of RR468, the 32P label of the HK is lost in <15 secs. However, no concomitant labeling of RR468 is observed (Casino et al., 2007). The lack of P~RR468 accumulation did not result from intrinsic instability of P~RR468, since P∼RR468, which was prepared chemically with acetylphosphate, was dephosphorylated (monitored by the change of mobility of the protein on native electrophoresis) with a half-life of ∼20 hours at 4ºC. (Figure S7A). Even at 80ºC, the living temperature of T. maritima, dephosphorylation took longer than 15 secs (t1/2 of ∼2 min, Figure S7B). This latter value resembles the half-lives of the phosphorylated forms of mesophilic RRs at the normal living temperatures of the organisms hosting them (Stock et al., 2000). We then tested whether the lack of RR468 labeling upon phosphoryl transfer from HK853 was due to HK853-triggered P∼RR468 dephosphorylation (phosphatase activity of HK853). Incubation at 21ºC of P~RR468 with a 10-fold lower (in molar terms) HK853CP concentration triggered rapid dephosphorylation of the RR since no detectable band corresponding to the P~RR468 was observed after native electrophoresis. In contrast, virtually all of the P~RR468 remained phosphorylated in the absence of the HK (Figure S7C). Thus, the phosphatase activity of the HK accounts for the lack of RR468 labeling in the phosphoryl transfer experiments. This phosphatase activity involves the His260 residue, since ∼45 % of the P∼RR468 remained phosphorylated at the end of the experiment when His260 was mutated to alanine. Supplemental Experimental Procedures PhoRCP cloning and purification. The coding DNA sequence for the predicted cytoplasmic catalytic portion (residues 316-554) of Staphylococcus aureus PhoR (UniprotKB accession numbe