pGpp, a new analog? We have, with others, discovered an enzyme that synthesizes pGpp from GMP by pyrophosphate transfer from ATP. This route is parallel to (p)ppGpp synthesis (ref 1). It is well known that (p)ppGpp is synthesized and hydrolyzed in Gram positives by a large bifunctional protein homolog of enzymes present in Gram negative enterics such as E. coli. Within the past decade Gram positives were found generally to also harbor one or two much smaller single domain synthetases, termed RelP or RelQ. Each of the small proteins also synthesize (p)ppGpp by pyrophosphate transfer. The functions of RelP and RelQ-like enzymes are being slowly revealed. Unlike their multidomain counterparts, the activities of these enzymes do not appear strikingly regulated by stress. The RelQ enzyme from Enterococcus faecalis (RelQEfa) has been found to be an enzymatic source of pGpp synthesis, a feature shared to variable extents with both of the RelP and RelQ homologs from Streptococcus mutans but not with the larger bifunctional proteins. Documenting the enzymatic source of pGpp is an important finding because pGpp could arise instead as a degradation product of (p)ppGpp due to sequential removal of the 5-terminal phosphates, as might occur with ubiquitous Nudix enzymes. The hydrolysis of pGpp occurs by a diesterase, releasing pyrophosphate as for (p)ppGpp- see below. A preliminary comparison of the pGpp in vitro regulatory properties with (p)ppGpp has been made. This survey reveals that pGpp is as potent an inhibitor as ppGpp for several enzymes in the pathway for GTP synthesis. This inhibition is a crucial ppGpp function in Gram positives. In contrast, pGpp is a less potent autocatalytic synthetase activator than ppGpp for multidomain and some single domain synthetases. Defining the physiological roles for pGpp and perhaps unique regulatory features of single domain synthetases awaits quantitative cellular studies. It is notable that in 1979, pGpp was tentatively identified as a radioactive spot on thin layer autoradiograms of 32P-labelled extracts of amino acid starved B.subtilis but this was not pursued further. (p)ppApp, a new analog? Initial studies that led to a focus on (p)ppApp have been presented in our earlier annual reports. These involved use of single domain hydrolases, called MESH (Metazoan E. coli SpoT Homologs), that were discovered to exist in flies and humans in another laboratory. These enzymes are stand alone hydrolases with structural similarities to known (p)ppGpp hydrolase domains that are present in bifunctional synthetases. All these hydrolases remove an intact pyrophosphate residues from (p)ppGpp and are Mn++ dependent. A pivotal discovery made by Dr. K. Potrykus was that the MESH hydrolases hydrolyze (p)ppApp as well as (p)ppGpp. A concerted attempt by Dr.Tamara James to exploit the known crystal structures of hydrolases with differing substrate specificities to derive a protein that could preferentially hydrolyze (p)ppApp as not successful. Three successive generations of predicted MESH enzyme mutants that might have selective hydrolysis activity towards (p)ppApp were instead found to lose both activities in parallel. Dr. Potrykus returned to Gdansk and formed a group that searched bacterial genomes for MESH homologs. This led to discovery of a MESH homologs that hydrolyzes (p)ppApp but not (p)ppGpp. The search now is for its cellular targets. ppGpp coupled transcription-dependent DNA repair. There have been several earlier reports suggesting that maintenance of genomic integrity falls within the domain of (p)ppGpp-centric regulation. The first clue to the involvement of the nucleotide excision repair mechanism was finding that ppGppdeficient (ppGpp0) cells were much more sensitive than cells with slightly elevated basal ppGpp when exposed to UV or genotoxic bulky DNA adducts. Similar exposures but with wild type cells revealed a transient accumulation of ppGpp after toxic agent exposures. The ppGpp levels increased far above basal levels by 15 min and thereafter decayed with a first order half-life of about 10 min. This stress induces much more ppGpp than pppGpp, which suggests the signal is not overt amino acid limitation. After UV induction ppGpp-dependent repair of thymine dimers were localized predominately on the transcribed strand of LacZ, implicating transcription coupled repair. One mechanism for this pathway involves Mfd bound to the leading surface of RNAP that positions the stalled enzyme near the bulky DNA lesion and allows recruitment of initial proteins for nucleotide excision repair. Binding of UvrD helicase backs up RNAP, exposes the DNA damage to allow entry of additional nucleotide excision repair enzymes. Studies of epistasis in ppGpp0 cells with deletions of either mfd or uvrD reveal additive effects of mfd but not uvrD. This suggests ppGpp0 and uvrD phenotypes effect the same mechanism and the backtracking role for uvrD is likely the same as for ppGpp. Accordingly deleting greAB abolishes factor competition with UvrD backtracking and renders ppGpp0 cells DNA damage resistant. Conversely DksA often acts to amplify ppGpp effects and itis consistent that deleting dksA makes wild type cells more damage sensitive while overexpressing DksA made them more resistant. The sensitizing effect of the dksA was reversed by also deleting greA, as predicted. These deductions could be verified with in vitro transcription experiments. Adding UvrD potentiated backtracking, visualized on gels as enhanced pausing in single round runoff transcripts, which are further enhanced by adding ppGpp +/- DksA, but not by adding DksA alone. Others report a modest opening of the DNA clamp structure (sometimes called crab claw) with ppGpp added to RNAP. It is proposed that this favors backtracking and can explain how ppGpp couples transcription to DNA repair. The proposal was verified in vitro with toe-print like studies using exonuclease III RNAP together with assembled stable elongation complexes; ppGpp was found to shift the complexes back to a pretranslocated state. Consistent in vivo footprinting changes were found using a plasmid bearing a Lac repressor bound to lac operator +/- inducing ppGpp and overexpressing or deleting uvrD. Finally effects of ppGpp were mimicked by a RNA polymerase mutant, RpoB*35 (H1244Q), that gives a phenocopy of ppGpp effects. It is noted that the transient feature of this DNA repair model would minimize ill effects of co-directional collisions of replication forks with lesion repair complexes. Overall it is argued DNA damage from bulky adducts leads to transient ppGpp accumulation that loosens the grip of RNA polymerase on DNA, which together with UvrD fosters backtracking to facilitate transcription coupled nucleotide excision repair.
Showing the most recent 10 out of 15 publications
