Global Regulation of Gene Expression by (p)ppGpp

Eunice Kennedy Shriver National Institute Of Child Health & Human Development

Investigators

Linked publications & trials

Abstract

During the past few years, for E. coli our model bacterium it was found that an abrupt starvation for nutrients leads to elevating levels of (p)ppGpp about 100-fold to approximate GTP. This leads surprisingly to instability of tRNA, rRNA and ribosomes, a reversal of a longstanding view. It also clashes with an otherwise convincing hypothesis that (p)ppGpp controls growth physiology by mainly by inhibiting synthesis of rRNA, tRNA and ribosome formation. A review of the impact of this new information has been submitted for review that proposes fundamentally different consequences between (p)ppGpp basal levels and the high levels that accompanying severe stress (Fernandez-Coll and Cashel submitted to Frontiers in Microbiology). Another long-standing hypothesis was threatened when we found cells deleted for all (p)ppGpp displayed the same high chromosomal DNA initiation rates at low and high growth rates. Using RNA polymerase mutants that phenocopied the presence of (p)ppGpp in its absence led to demonstrating that (p)ppGpp is necessary and sufficient for the initiation of chromosomal DNA synthesis. The prevailing hypothesis was that the supercoiling effects of ppGpp inhibition of transcription of seven rRNA operons was transmitted through the chromosome to inhibit ori region activity. We tested mutants with none of the rRNA genes on the chromosome that would be lethal except for a single rRNA operon present on a high copy plasmid, instead of seven. We found wild type regulation, which voided the old hypothesis. This paper (DOI:10.1128/mBio.03223-19 was awarded as Editors pick by mBio physiology and molecular biology. A random mutant GreA suppressor library suggests new functions in old motifs. A manuscript for Nature Scientific Reports, accepted on Sept 9th 2020, and is now in press. This work reflects the four laboratories Dr. Fernandez-Coll has worked in over the years. It describes a random mutant GreA library isolated as survivors of toxic greA overexpression. Several of these residues acquire new functions that are in unexpected places in otherwise well-studied structural domains. This is a classical genetic approach to new understanding whose power is no assumptions are required. Methodologically, this work uses a RNAP pull-down assays that Dr Fernandez-Coll developed; these allow estimating competition among RNAP secondary channel binding proteins that we have studied before in cells by changing ratios. A continuing collaboration with Dr. K. Potrykus, a former postdoc, has led to appreciating (p)ppApp as a likely new alarmone. She discovered at NIH that a eukaryotic (p)ppGpp hydrolase called Mesh-1 (Sun et al. 2010) also hydrolyzed (p)ppApp with a TLC assay. Our lab then developed a fast real time nonradioactive coupled assay for all four polyphosphorylated substrates Potrykus et al. 2020 (Frontiers in Microbiology, under review. The Potrykus lab earlier showed in vitro that (p)ppApp stimulates transcription of a model E. coli promoter that (p)ppGpp inhibited along with a unique RNAP binding site for (p)ppApp conserved in B. subtilis (Bruhn-Olezewska B. et al 2018). This was followed by their finding a strain called Methanobacterium extorquen (Mex) with a Mesh-1 like sequence. The pure protein was found to hydrolyze (p)ppApp but not (p)ppGpp, the first of its kind. One M. ext source for (p)ppApp was found to be a RSH catalytic fragment to synthesize (p)ppApp and (p)ppGpp in vitro without ribosomes; this work also detected pppApp extracted along with ppGpp from cells of E. coli and B. subtilis which rules out their existence as an in vitro artifact (Sobala M et al 2019). A Swedish lab this year induced individual members of members of T-A clusters identified by bioinformatics to sometimes find small hydrolases and synthetases that also had degenerate substrate specificities; in this case (p)ppApp is viewed as a toxin (Jimmy et al 2020) because when a Pseudomonad injects a very active (p)ppApp synthetase, tas-1, into any neighboring cell, it is toxic (Ahmed T et al 2019. We suggest that so far there is no rigorous evidence that (p)ppApp instead could have alarmone function because searches of function due to low, nonlethal concentrations have not yet been carried out. The diversity of ppGpp regulatory effects in any one bacterial species is amplified further by bacterial diversity. We focus on E. coli as a model enteric organism where most regulation is thought due to ppGpp binding to two sites on RNA polymerase. Our past work has contributed to structural and regulatory features of both sites. Binding leads to activation or inhibition of nearly 1/3 of all E coli genes. Almost imperceptible perturbations of amino acid, lipid or energy metabolism are sensed to provoke (p)ppGpp accumulation. This can occur by stimulating synthesis, limiting hydrolysis or both. Two E. coli proteins act as sensors. A RelA synthetase binds to an empty codon-specific ribosomal acceptor sites together with uncharged tRNA complex to activate synthesis if the metabolic yield of any AA is defective. A bifunctional synthetase-hydrolase SpoT protein senses deficiencies of energy, lipid synthesis and/or physical stress. Either way, the ensuing ppGpp quickly alerts cells to subtle as well as severe stress and then triggers an astonishing array of regulatory functions that promote bacterial survival and optimize growth. These responses also protect pathogens from host cell defenses to enhance pathogenicity. Fundamental evidence that pathogenicity is enhanced by (p)ppGpp is pervasive as for antibiotic tolerance, resistance and persistence. The profligate misuse of antibiotics and the dearth of new drugs has made fundamental studies of new regulators like ppApp acutely more relevant if we are not to enter an era (recently predicted as 2040) when all antibiotics are ineffective.