Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Apr 4;114(14):3726-3731.
doi: 10.1073/pnas.1617868114. Epub 2017 Mar 20.

Negative Allosteric Regulation of Enterococcus faecalis Small Alarmone Synthetase RelQ by Single-Stranded RNA

Affiliations
Free PMC article

Negative Allosteric Regulation of Enterococcus faecalis Small Alarmone Synthetase RelQ by Single-Stranded RNA

Jelena Beljantseva et al. Proc Natl Acad Sci U S A. .
Free PMC article

Abstract

The alarmone nucleotides guanosine pentaphosphate (pppGpp) and tetraphosphate (ppGpp), collectively referred to as (p)ppGpp, are key regulators of bacterial growth, stress adaptation, pathogenicity, and antibiotic tolerance. We show that the tetrameric small alarmone synthetase (SAS) RelQ from the Gram-positive pathogen Enterococcus faecalis is a sequence-specific RNA-binding protein. RelQ's enzymatic and RNA binding activities are subject to intricate allosteric regulation. (p)ppGpp synthesis is potently inhibited by the binding of single-stranded RNA. Conversely, RelQ's enzymatic activity destabilizes the RelQ:RNA complex. pppGpp, an allosteric activator of the enzyme, counteracts the effect of RNA. Tetramerization of RelQ is essential for this regulatory mechanism, because both RNA binding and enzymatic activity are abolished by deletion of the SAS-specific C-terminal helix 5α. The interplay of pppGpp binding, (p)ppGpp synthesis, and RNA binding unites two archetypal regulatory paradigms within a single protein. The mechanism is likely a prevalent but previously unappreciated regulatory switch used by the widely distributed bacterial SAS enzymes.

Keywords: (p)ppGpp; RNA–protein interaction; allosteric regulation; nucleotide signaling; stringent response.

Conflict of interest statement

The authors declare no conflict of interest.

Figures

Fig. 1.
Fig. 1.
mRNA is a potent inhibitor of ppGpp synthesis by E. faecalis RelQ. (A) 3H ppGpp synthesis activity of 250 nM E. faecalis RelQ (62.5 nM tetrameric RelQ) in the presence (gray bars) and absence (empty bars) of 100 μM ppGpp, as well as starved ribosomal complexes or individual components thereof. Note that ppGpp is a strong activator of RelQ’s enzymatic activity and mitigates the inhibition by starved ribosomal complexes or mRNA(MF). (B and C) Single-stranded mRNA inhibits RelQ’s activity in a sequence-specific manner, and this inhibition is countered by ppGpp. Titrations were performed with increasing concentrations of either single-stranded (empty circles) or double-stranded (filled circles) RNA, in the absence (black circles) or presence (red circles) of 100 μM ppGpp. All reaction mixtures contained 250 nM (62.5 nM tetramer) E. faecalis RelQ, 300 μM 3H GDP, and 1 mM ATP. Titration data were fitted with the 4PL model. Error bars represent SDs of the turnover estimates determined by linear regression. Each experiment was performed at least three times.
Fig. S1.
Fig. S1.
Synthetic activity of 100 nM E. coli RelA activated by starved ribosomal complexes is unaffected by up to 10 μM mRNA(MF). Starved ribosomal complexes were assembled from 0.5 μM E. coli 70S, 2 μM deacylated tRNAPhe and tRNAMeti, and 2 μM model mRNA encoding the Met-Phe (MF) dipeptide (5′-GGCAAGGAGGUAAAAAUGUUCAAA-3′) in Hepes:Polymix buffer at 37 °C; 300 μM 3H GDP and 1 mM ATP served as substrates for the 3H ppGpp synthesis. Error bars represent SDs of the turnover estimates determined by linear regression. Each experiment was performed at least three times.
Fig. S2.
Fig. S2.
Double-stranded DNA ineffectively inhibits E. faecilis RelQ. (A) ssDNA(MF) is a poor inhibitor of RelQ. Titrations were performed with increasing concentrations of either single-stranded (empty circles) or double-stranded (filled circles) DNA(MF), in the absence (black circles) or presence (red circles) of 100 μM ppGpp. All reaction mixtures contained 250 nM (62.5 nM tetramer) E. faecalis RelQ, 300 μM 3H GDP, and 1 mM ATP. Experimental data were fitted using the 4PL model (Hill equation). (B) EMSA analysis detected a negligible amount of complex formation between 0.15 μM dsDNA(MF) or dsRNA(MF) and increasing concentrations of RelQ.
Fig. S3.
Fig. S3.
Enzymatic activity of RelQ is unaffected by inorganic polyphosphate. Experiments were performed with 0.5 mg/mL BSA, 0.15 µM RelQ (37.5 nM tetramer), 300 µM 3H GDP, 100 µM ppGpp, 1 mM ATP, and increasing concentrations of long-chain (p700) polyphosphate. Error bars represent SDs of the turnover estimates by linear regression. The experiment was performed twice.
Fig. 2.
Fig. 2.
Sequence specificity of RelQ inhibition by RNA. Here 24-nt-long mRNA(MF) (red) and its complementary antisense RNA (blue) were used as a positive and negative controls, respectively. Based on the two RNAs, we generated chimeras (A), cut-backs (B), and point mutants (C). We also reconstituted the inhibitory activity by adding two GG elements to otherwise inactive poly(A) RNA (D). To calculate the RelQ activity, the turnover rate (3H ppGpp synthesized per RelQ per minute) in the presence of RNA was divided by that in the absence of RNA. All reaction mixtures contained 100 nM RNA, 250 nM (62.5 nM tetramer) E. faecalis RelQ, 300 μM 3H GDP, and 1 mM ATP. Error bars represent SDs of the turnover estimates determined by linear regression. Each experiment was performed at least three times.
Fig. 3.
Fig. 3.
mRNA and pppGpp have a destabilizing effect on each other’s binding to RelQ. (A) Complex formation between 0.15 μM mRNA(MF) and increasing concentrations of E. faecalis RelQ was monitored by EMSA in the absence (empty circles) and presence (filled circles) of 100 μM ppGpp. (B) EMSA analysis of complex between 0.15 μM mRNA(MF) and 2 μM RelQ in the presence of increasing pppGpp concentrations. (C) Complex formation of increasing concentrations of RelQ with 50 nM 32P-labeled ATP, ppGpp, or pppGpp monitored by DRaCALA. (D) 32P pppGpp is displaced from 20 μM RelQ by increasing concentrations of mRNA(MF), as monitored by DRaCALA. Error bars represent SDs of the mean. Each experiment was performed at least three times.
Fig. S4.
Fig. S4.
EMSA analysis of E. coli RelA (A) and E. faecilis RelQ (C and D) interactions with mRNA(MF). (A) E. coli RelA does not bind mRNA(MF) in either the presence or absence of 100 μM ppGpp. (B) Increasing ppGpp concentrations up to 1 mM moderately destabilizes complex formation between 0.15 μM mRNA(MF) and 2 μM RelQ. Experimental data were fitted using the 4PL model (Hill equation). (C) The addition of ATP and GDP, but not of any other combinations of nucleotides, disrupts RelQ:mRNA(MF) complex formation. All tested nucleotides were added to a final concentration of 1 mM. (D) Omission of the additional 1 M KCl during purification of catalytically inactive D82G RelQ results in a smearing pattern. EMSA analysis was performed in the presence of 0.15 μM mRNA(MF). The positions of RelQ:mRNA and its supershifted complex in the presence of substrates are indicated with empty and filled red triangles, respectively.
Fig. S5.
Fig. S5.
DRaCALA analysis of RelQ interactions with 32P ATP, 32P ppGpp, and 32P pppGpp. (A) ppGpp, but not ATP, is displaced from RelQ by increasing concentrations of mRNA(MF). (B) Catalytically inactive RelQ mutant D82G (EF2671 locus numbering) efficiently binds pppGpp, but not ATP. Complex formation is monitored using 50 nM 32P-labeled ATP, ppGpp, or pppGpp and increasing concentrations of RelQ. IC50, EC50, and the Hill coefficient (nH) were calculated using the 4PL model (Hill equation). Error bars represent SDs of the mean. Each experiment was performed at least three times.
Fig. 4.
Fig. 4.
RelQ binding to mRNA and ppGpp synthesis are mutually exclusive. (A) Although 1 mM ppGpp, ATP, or GDP alone does not affect the stability of the RelQ:mRNA(MF) complex, a combination of ATP and GDP has a strong destabilizing effect in both the presence and absence of 100 μM ppGpp. The positions of RelQ:mRNA (open red triangles) and supershifted complex in the presence of substrates (filled red triangles) are indicated to the left. (B) Increasing GDP substrate concentration in the presence of 1 mM ATP progressively destabilizes the RelQ:mRNA(MF) complex. (C) Addition of 100 μM ppGpp to RelQ both increases its catalytic efficiency (Vmax) and relaxes the positive substrate cooperativity, as shown by a decrease in the Hill constant, nH. (D) Enzymatically inactive RelQ mutant D82G (EF2671 locus numbering) forms the complex with mRNA(MF) as efficiently as the WT protein, whereas the addition of 1 mM ATP and GDP does not destabilize the complex in either the presence or absence of 100 μM ppGpp. Error bars represent SDs of the mean. Each experiment was performed at least three times.
Fig. S6.
Fig. S6.
Excess of RelQ over mRNA masks the inhibitory effect of mRNA on 3H ppGpp. In contrast to enzymatic assays in which mRNA(MF) is added in excess over RelQ [0.2 μM RelQ vs. 2 μM mRNA(MF)] (A), in EMSA experiments RelQ is added to the reaction mixture in excess over mRNA(MF) [2 μM RelQ vs. 0.15 μM mRNA(MF)] (B). This results in inefficient inhibition of RelQ’s enzymatic activity under EMSA conditions and accumulation of in situ-produced ppGpp.
Fig. 5.
Fig. 5.
The combination of GDP as a RelQ substrate and pppGpp as an allosteric regulator provides the best protective effect against RelQ inhibition by mRNA(MF). (A) The combination of the preferred substrate (GDP) and the best binding allosteric regulator (pppGpp) provides the strongest protective effect against mRNA(MF). All reaction mixtures contained 250 nM (62.5 nM tetramer) E. faecalis RelQ, 1 mM ATP, 1 μM mRNA(MF), 300 μM 3H GDP/GTP, and increasing concentrations of ppGpp/pppGpp. Error bars represent SDs of the mean. Each experiment was performed at least three times. (B) EMSA analysis of complex formation between the WT and enzymatically inactive D82G mutant RelQ and 0.15 μM mRNA(MF) in the presence of 1 mM substrates GTP and ATP and 100 μM allosteric regulator pppGpp.
Fig. S7.
Fig. S7.
Addition of 100 μM pppGpp to RelQ moderately increases its affinity to GTP substrate. All reaction mixtures contained 250 nM (62.5 nM tetramer) E. faecalis RelQ, 1 mM ATP, and increasing concentrations of 3H GTP in the presence or absence of 100 μM pppGpp. Maximum velocity (Vmax), the substrate concentration at which enzyme achieves 0.5 Vmax (K0.5), and the Hill coefficient (nH) were calculated using the 4PL model (Hill equation). Error bars represent SDs of the mean. Each experiment was performed at least three times.
Fig. 6.
Fig. 6.
An intact tetrameric structure is essential for ppGpp synthesis by RelQ. (A) C-terminal helix 5α (amino acids 174–234 in E. faecalis RelQ; EF2671 locus numbering) is highlighted in yellow in this homology model of E. faecalis RelQ based on the SAS1 tetramer of B. subtilis (18). Helix 5α, which forms contacts in tetrameric RelQ, is SAS-specific, i.e., absent in ribosome-associated RSHs such as RelA. Sequence alignment is shown in Fig. S10. Two allosteric pppGpp molecules are intercalated in the central cleft. (B–D) Deletion of the C-terminal α5 helix results in Δα5 RelQ, which is monomeric as shown by analytical gel filtration (B), enzymatically inactive in the presence or absence of 100 μM ppGpp (C), and unable to bind mRNA (D). Error bars represent SDs of the mean. Each experiment was performed at least three times.
Fig. S8.
Fig. S8.
Enzymatically inactive D82G RelQ does not affect the synthetic activity of WT RelQ. (A) When the D82G (EF2671 locus numbering) mutant is purified following the standard protocol for purification of WT RelQ (19), the resulting protein is contaminated with RNA and exerts a strong inhibitory activity when added in excess over 0.5 μM WT RelQ. (B) Addition of an extra 1 M KCl to buffers used for purification mitigates the RNA contamination. RNA-free D82G RelQ does not inhibit the synthetic activity of 0.5 μM WT RelQ in either the presence or absence of externally added 100 μM ppGpp. Error bars represent SDs of the turnover estimates by linear regression. The experiment was performed twice.
Fig. S9.
Fig. S9.
Tetrameric structure of RelQ is not affected by substrates and/or mRNA. Here 10-μL samples of WT (black and brown traces) or D82G RelQ (EF2671 locus numbering, red traces) at a concentration of 18.5 μM were analyzed on a Superdex 200 Increase 5/150 GL analytical gel filtration column, either alone (A) or supplemented with 1 mM ATP and GDP (B), or with 60 μM mRNA(MF) along with 1 mM ATP and GDP (C).
Fig. S10.
Fig. S10.
The 3D structure of E. faecalis RelQ and alignment with the ppGpp synthesis domain of E. coli RelA. (A) Secondary structure elements (α-helixes and β-sheets) of B. subtilis SAS1 (9) are indicated as spirals and arrows, respectively. Insertions in RelQ are highlighted in green, deletions are highlighted in red, and the C-terminal helix α5 is in yellow. Amino acid D82 (EF2671 locus numbering) is indicated with an orange asterisk. The first 17 characters are unaligned. The figure was generated using ESPript (34). (B) A homology model of the RelQ monomer based on B. subtilis SAS1 (18). Green indicates insertions, and red indicates the boundary of the deletion relative to RelA. Catalytically important amino acid D82 (EF2671 locus numbering) is highlighted in orange.
Fig. S11.
Fig. S11.
RelQ:mRNA interaction as a regulatory switch. RelQ’s enzymatic activity is potently inhibited by association with the single-stranded RNA in a sequence-specific manner with a putative consensus of GGAGG. Association of the primary allosteric regulator pppGpp or, to a lesser extent, the secondary allosteric regulator ppGpp strongly counteracts the inhibition by RNA and destabilizes the RNA:RelQ complex. The protective effect is especially strong when the primary allosteric regulator pppGpp synergizes with the preferred substrate, GDP. Both the enzymatic activity and mRNA binding can serve as cellular effectors, acting via intracellular concentration of the alarmone and direct interaction with mRNA, respectively.

Similar articles

See all similar articles

Cited by 12 articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback