For many years, our lab has investigated the role of energy-dependent proteolysis in regulation of gene expression in bacteria. The ATP-dependent cytoplasmic proteases, akin to the eukaryotic proteasome, contain ATPase domains or subunits that recognize substrates and unfold them, feeding them to the proteolytic domains. Bacteria contain multiple ATP-dependent proteases; five of them have been characterized in E. coli. Abnormal or misfolded proteins are degraded by these proteases. In addition to this quality control role, the proteases degrade proteins that are naturally unstable; for these proteins, degradation is likely to play an important biological role. Such protease substrates fall into two general classes: proteins that are always degraded, so that regulation of their abundance depends primarily on changes in synthesis, and proteins that show regulated proteolysis. In all cases, identifying how the substrate is recognized by the protease and how recognition is affected by growth conditions is important in understanding how regulation is carried out. In the past, our lab showed that the Lon ATP-dependent protease regulated capsular polysaccharide synthesis and cell division by degrading the RcsA and SulA proteins, discovered and characterized the two-component Clp proteases, ClpAP and ClpXP, and investigated the roles of these proteases in vivo and in vitro. In recent years, our focus has been on the regulated degradation of the RpoS sigma factor, a subunit of RNA polymerase that directs the polymerase to specific promoters. RpoS is important for cells to switch to a stationary or stress response gene expression program, and the cell regulates RpoS accumulation in a variety of ways, including at the level of translation via small RNA activators of translation, and by regulated proteolysis. We have been studying this proteolysis, one of the best examples of regulated protein turnover in E. coli. RpoS is rapidly degraded during active growth, in a process that requires the energy-dependent ClpXP protease and the adaptor protein RssB, a phosphorylatable protein that presents RpoS to the protease. RpoS becomes stable after various stress or starvation treatments; the mode of stabilization was a mystery until recent work from our lab. A genetic screen for regulators of RpoS degradation led to discovery of a small, previously uncharacterized protein, now named IraP (inhibitor of RssB activity after phosphate starvation). Mutants of iraP abolish the stabilization of RpoS after phosphate starvation. IraP blocks RpoS turnover in a purified in vitro system, and directly interacts with RssB. In E. coli, phosphate starvation is sensed by an increase in the levels of the small molecule ppGpp, and the iraP promoter is positively regulated by ppGpp. Two other small proteins also stabilize RpoS in a purified in vitro system, IraM, and IraD. These proteins are not similar in predicted structure to IraP. IraM is made in response to magnesium starvation, dependent on the PhoP and PhoQ regulators; IraD is important after DNA damage. The anti-adaptors define a new level of regulatory control, interacting with the RssB adaptor protein and blocking its ability to act; environmental signals regulate RpoS turnover by regulating expression of different anti-adaptors. RssB structure and function have been further investigated using a range of approaches: 1) a genetic selection to identify mutations in RssB resistant to a specific anti-adaptor, 2) a bacterial two-hybrid system to investigate the interaction of wild-type and mutant derivatives of RssB and its domains with the anti-adaptors and other components of the system, and 3) collaborative in vitro studies of RpoS degradation with the mutant proteins and the antiadaptors. The results of these studies demonstrate that both IraP and IraD interact with the N-terminal domain of RssB. This domain of RssB is a member of the widespread response regulator family. Although both anti-adaptors interact with this conserved domain, they do not interact in the same fashion. Thus, mutations in RssB that abolish interaction with IraP retain interactions with IraD. Collaborative studies on the structure of IraP reveal that it is a unique protein with similarity to B-Zip dimers. The interaction of IraP with RssB may mimic how RpoS interacts with RssB. IraD, on the other hand, appears to interact with the inactive form of RssB, possibly blocking activation. IraM interacts with C-terminal domain of RssB; this domain has homology to an inactive PP2C phosphatase domain. One class of anti-adaptor mutations activates RssB, bypassing the stimulatory effect of phosphorylation. These mutants provide new insight into how RssB works and how regulatory proteins can disrupt the function of the conserved domains that make up RssB. We are further defining how RssB interacts with ClpX, the ATPase subunit of the ClpXP protease. The N-terminal domain of ClpX, known to interact with some other adaptors and substrates, interacts with the RssB C-terminus. Continued dissection of this system is providing insight into how this process is balanced in the cell. Other anti-adaptors are likely to exist, based on a variety of results, including the observation that the transcriptional regulator AppY stabilizes RpoS in the absence of all three known anti-adaptors. Other mutations that stabilize RpoS work through the known anti-adaptors. For instance, deletion of the global repressor, H-NS, also stabilize RpoS. Expression of IraD and IraM is repressed by H-NS, and in the absence of these two anti-adaptors, much but not all of the stabilizing effect of H-NS is lost. Mutations in aceE, a component of central metabolism, also leads to stabilization of RpoS, dependent on IraP and IraD. aceE mutants have induced levels of iraP and iraD transcription, apparently dependent on the signaling molecule ppGpp. The bacterial two-hybrid system has also been used to screen a library of E. coli proteins for other proteins that interact with RssB. Interacting proteins may be additional anti-adaptors, giving us a sense of how broad this family of regulators is; alternatively, we may find other substrates for RssB, which has previously been believed to be specific to RpoS. A number of interesting regulatory proteins have been identified and found to affect RpoS turnover, and thus may be acting as anti-adaptors or competing substrates, both of which are of significant interest. One of them, AnmK, has been studied in some detail. AnmK is an enzyme involved in recycling of peptidoglycan. In addition to interacting with RssB in the bacterial two-hybrid assay, it interacts in a pull-down assay. Overexpression of AnmK stabilizes RpoS, consistent with it acting either as a competitive inhibitor or an anti-adaptor; deletion of anmK may have an effect on stability of RpoS under particular stress conditions. Further study of these interacting proteins is in process. Overall, our proteolysis studies continue to provide novel insights into regulatory mechanisms used by bacteria.
| Hamdallah, Issam; Torok, Nadia; Bischof, Katarina M et al. (2018) Experimental Evolution of Escherichia coli K-12 at High pH and with RpoS Induction. Appl Environ Microbiol 84: |
| Gottesman, Susan (2017) Introduction. Annu Rev Microbiol 71:i-ii |
| Gottesman, Susan (2017) Stress Reduction, Bacterial Style. J Bacteriol 199: |
| Tripathi, Arti; Gottesman, Susan (2016) Cell biology: Phosphate on, rubbish out. Nature 539:38-39 |
| Battesti, Aurelia; Majdalani, Nadim; Gottesman, Susan (2015) Stress sigma factor RpoS degradation and translation are sensitive to the state of central metabolism. Proc Natl Acad Sci U S A 112:5159-64 |
| Battesti, Aurelia; Hoskins, Joel R; Tong, Song et al. (2013) Anti-adaptors provide multiple modes for regulation of the RssB adaptor protein. Genes Dev 27:2722-35 |
| Battesti, Aurelia; Gottesman, Susan (2013) Roles of adaptor proteins in regulation of bacterial proteolysis. Curr Opin Microbiol 16:140-7 |
| Battesti, A; Tsegaye, Y M; Packer, D G et al. (2012) H-NS regulation of IraD and IraM antiadaptors for control of RpoS degradation. J Bacteriol 194:2470-8 |
| Battesti, Aurelia; Majdalani, Nadim; Gottesman, Susan (2011) The RpoS-mediated general stress response in Escherichia coli. Annu Rev Microbiol 65:189-213 |
