c-di-AMP, a likely master regulator of bacterial K+ homeostasis machinery, activates a K+ exporter

Tatiana B Cereija; João P L Guerra; João M P Jorge; João H Morais-Cabral

doi:10.1073/pnas.2020653118

. 2021 Mar 31;118(14):e2020653118. doi: 10.1073/pnas.2020653118

c-di-AMP, a likely master regulator of bacterial K⁺ homeostasis machinery, activates a K⁺ exporter

Tatiana B Cereija ^a,^b, João P L Guerra ^a,^b,¹, João M P Jorge ^a,^b, João H Morais-Cabral ^a,^b,²

PMCID: PMC8040634 PMID: 33790011

Significance

The importance of dinucleotides in the control of bacterial physiology is still being uncovered. One of these molecules is c-di-AMP, thought to control the mechanisms of import and export of potassium ions in many bacteria, including pathogens. Maintaining a balanced internal concentration of potassium is crucial for bacterial viability, even during infection. However, there has been a lack of data showing that c-di-AMP effectively changes the activity of potassium transporters. Here, we demonstrate that c-di-AMP increases the activity levels of a potassium exporter and that pH increases the sensitivity of the transporter to c-di-AMP. Our quantitative data support, therefore, the central role of c-di-AMP in the control of the potassium homeostasis machinery in bacteria.

Keywords: KhtT, KhtTU, B. subtilis, RCK domain, K⁺/H⁺ antiporter

Abstract

bis-(3′,5′)-cyclic diadenosine monophosphate (c-di-AMP) is a second messenger with roles in virulence, cell wall and biofilm formation, and surveillance of DNA integrity in many bacterial species, including pathogens. Strikingly, it has also been proposed to coordinate the activity of the components of K⁺ homeostasis machinery, inhibiting K⁺ import, and activating K⁺ export. However, there is a lack of quantitative evidence supporting the direct functional impact of c-di-AMP on K⁺ transporters. To gain a detailed understanding of the role of c-di-AMP on the activity of a component of the K⁺ homeostasis machinery in B. subtilis, we have characterized the impact of c-di-AMP on the functional, biochemical, and physiological properties of KhtTU, a K⁺/H⁺ antiporter composed of the membrane protein KhtU and the cytosolic protein KhtT. We have confirmed c-di-AMP binding to KhtT and determined the crystal structure of this complex. We have characterized in vitro the functional properties of KhtTU and KhtU alone and quantified the impact of c-di-AMP and of pH on their activity, demonstrating that c-di-AMP activates KhtTU and that pH increases its sensitivity to this nucleotide. Based on our functional and structural data, we were able to propose a mechanism for the activation of KhtTU by c-di-AMP. In addition, we have analyzed the impact of KhtTU in its native bacterium, providing a physiological context for the regulatory function of c-di-AMP and pH. Overall, we provide unique information that supports the proposal that c-di-AMP is a master regulator of K+ homeostasis machinery.

bis-(3′,5′)-cyclic diadenosine monophosphate (c-di-AMP) is part of an expanding family of cyclic dinucleotides, which includes c-di-GMP and cGAMP, that function as important bacterial signaling molecules and are sensed by the innate immune response during infection (1–3). c-di-AMP is present in many bacteria, including Bacillus subtilis and pathogens such as Bacillus anthracis, Staphylococcus aureus, Streptococcus pneumoniae, Streptococcus pyogenes, and Listeria monocytogenes (4–8). In contrast, Escherichia coli lacks the enzymes that synthesize c-di-AMP. This dinucleotide has roles in central metabolism (9), cell wall (10, 11) and biofilm formation (12, 13), DNA integrity (14), and virulence (15). c-di-AMP has also been proposed as a master regulator of the K⁺ homeostasis machinery (16–19). In bacteria and archaea, high intracellular K⁺ is not just crucial for the activity and stability of the ribosome and setting the membrane electrical potential (20, 21), but it also has roles in pH homeostasis, determining internal pressure, and in the adaptation to extracellular osmotic changes (22–24). For example, upon hyperosmotic challenge, the intracellular concentration of K⁺ in B. subtilis will rise from 300 mM to 600 or 700 mM, during the first phase of the adaptation mechanism, decreasing back to 300 mM in the second phase (25). It is therefore crucial that the activity of dedicated K⁺ importers and exporters is tightly controlled.

The proposal that c-di-AMP has a central role in the regulation of the K⁺ homeostasis machinery in many bacteria, inactivating K⁺ import and activating K⁺ export (8), has resulted from a series of experimental observations. The nucleotide concentration increases in parallel with intracellular K⁺ (17, 26) and elevated c-di-AMP results in osmosensitivity (27, 28). Phenotypical changes are observed when some K⁺ transporters are coexpressed in E. coli mutant strains with diadenylate cyclases, which catalyze synthesis of c-di-AMP, suggesting that c-di-AMP affects the function of the transporter (29, 30). In addition, c-di-AMP binds to a riboswitch, regulating the transcription levels of some K⁺ transport proteins (31) and to regulatory domains or proteins of known and putative K⁺ transporters (29, 30, 32–36). However, there is little or no quantitative evidence demonstrating and characterizing a direct functional impact of c-di-AMP on the activity of K⁺ transporters, and it is not known how c-di-AMP modulation is coordinated with other signaling cues.

To understand the regulatory impact of c-di-AMP on a K⁺ transporter protein, we analyzed the mechanism of regulation of the KhtTU K⁺/H⁺ antiporter from B. subtilis. This protein complex is composed by KhtU, a membrane protein that belongs to the CPA2 superfamily of cation/H⁺ antiporters, and KhtT, a cytoplasmic protein that is proposed to bind c-di-AMP at its C-terminal RCK_C domain (30, 32–35, 37–40). The genes of the two proteins are organized in an operon that also includes KhtS, which is thought to associate with KhtU but has unknown function and is absent from many other bacterial species with KhtTU orthologs (37). Using an E. coli phenotype complementation assay, it was suggested that KhtU alone is active and involved in the export of K⁺, while the complex KhtTU is inactive (37). Using an in vitro fluorescence-based transport assay with inside-out (everted) vesicles, it was demonstrated that KhtTU transports H⁺ in response to the external addition of K⁺, consistent with K⁺/H⁺ antiport activity, and that KhtTU is selective for K⁺ over Na⁺ (38). In addition, it was shown that KhtTU activity is pH dependent, with no activity at pH 7.5 and maximum activity at pH 9.0.

Here, we have performed a structural, biochemical, and functional characterization of KhtT, KhtU, and of their complex KhtTU and demonstrate that c-di-AMP directly activates KhtTU in a concentration-dependent manner by binding to KhtT and that pH modulates nucleotide activation. Overall, our results strongly support the model in which c-di-AMP controls the activity of K⁺ transporters in many bacterial species.

Results

Structure of KhtT in Complex with c-di-AMP.

It has been proposed using a ligand overlay assay that the cytosolic protein of the KhtTU complex, KhtT, binds c-di-AMP (30). We performed isothermal titration calorimetry (ITC) experiments at pH 8.0 with c-di-AMP and KhtT and determined that the dissociation constant (K_D) is ∼130 nM with a stoichiometry of ∼0.5, corresponding to 1 c-di-AMP bound per KhtT dimer (Fig. 1 and Table 1). In contrast, similar titrations with c-di-GMP did not show any heat changes at two temperatures (25 °C and 15 °C). These results demonstrate that KhtT binds c-di-AMP tightly and specifically.

Fig. 1. — Characterization of the interaction of c-di-AMP with KhtT using ITC. (A) Titration of c-di-AMP (152 µM) into KhtT (18 µM). (B) Titration of KhtT (175 µM) into c-di-AMP (6 µM). (C) Titration of c-di-GMP (308 µM) into KhtT (35 µM). *Top* show the raw titration heat values plotted as a function of time, and *Bottom* show normalized integration heat values of injectant plotted as a function of molar ratio. Experiments were performed at pH 8.0 and 25 °C. Curves shown in *Bottom* correspond to a single binding site model. See also *SI Appendix*, Fig. S8.

Table 1.

Thermodynamic parameters for the interaction of KhtT with c-di-AMP determined at 25 °C

Parameters	pH 8.0	pH 7.5
N	0.52 ± 0.04	0.53 ± 0.002
K_D (nM)	126.1 ± 74.0	42.6 ± 2.3
ΔH (cal · mol⁻¹)	−1.8x10⁴ ± 6.2x10³	−3.4x10⁴ ± 9.5x10²
ΔS (cal · mol⁻¹)	−29.4 ± 19.2	−79.7 ± 3.14
n	4	3

Open in a new tab

Mean ± SD. N, stoichiometry of binding; K_D, dissociation constant; n, number of replicates.

To gain insights into the interaction between KhtT and its ligand, we determined the crystal structure of KhtT:c-di-AMP complex at 1.85 Å (Fig. 2A and SI Appendix, Table S1). The structure reveals a dimer organized in two domains. The C-terminal domain (C-domain) stretches from residue 77 to the C terminus, adopts the typical fold of C-domains of RCK proteins (also known as RCK_C domains), and binds c-di-AMP in the dimer interface. The N-terminal domain (N-domain) runs from the N terminus to residue 67 and has an unusual fold, comprised by a twisted five antiparallel β-sheet followed by an α-helix. Unlike the C-domain, it appears that the dimeric arrangement of the N-domain is an integral part of its fold since the buried surface area in the N-domain dimer is extensive (∼1,360 Å²), while in the C-domain, only ∼390 Å² are buried in the dimer interface. The N- and C- domains are connected by a nine–amino acid linker.

We also solved the KhtT:c-di-AMP structure from a different crystal, with diffraction to 3.14 Å (Fig. 2B and SI Appendix, Table S1). This structure shows the same dimeric arrangement, but the N- and C-domains have an altered relative position. Superposition of the N-domains of the two structures shows little alterations in that domain and reveal that the C-domain dimer undergoes a rigid-body torsion with extension of the linker region and conversion of its first β-strand into an α-helix (SI Appendix, Fig. S1), demonstrating an impressive flexibility in the arrangement between the N- and C-domains of KhtT.

In both crystal forms, c-di-AMP binds at the interface between the C-domains, adopting a “U”-shaped conformation (Fig. 2C). Its position is stabilized by contacts between the adenine groups and apolar residues (Leu104, Val109, Val117, Gly132, and Ala133), as well as hydrogen bonds with Pro131 and Val117, in each subunit. At the mouth of the binding site, the two Asp108 residues and one Gln111 (from chain B) contact the ribose groups, while the other Gln111 is recruited into the coordination of a calcium ion mediating a crystal contact (SI Appendix, Fig. S2). Strikingly, Arg110 from chain B stretches into the inner part of the “U”-shaped c-di-AMP and contacts the two phosphate groups in c-di-AMP, one of them directly and the other through bridging by a water molecule that is positioned across the cyclic phosphate group. A second water molecule also bridges the phosphate and Ala133 from molecule B. This disposition is observed in three out of four dimers present in the asymmetric unit of the high-resolution structure. In the fourth dimer, the side chains of Arg110 from both subunits point into the inner part of the c-di-AMP “U,” but one of the arginine side chains displays partial occupancy, showing also density for an alternate conformation away from the nucleotide (SI Appendix, Fig. S3). In the low-resolution structure, the c-di-AMP molecule adopts the same disposition in the dimer interface, but it is harder to define the details of the binding sites in the two dimers present in the asymmetric unit. Only one of the four Arg110 could be modeled, and it displays a similar position to those found in the high-resolution structure.

The contacts established by Arg110 with the c-di-AMP phosphate, the adoption of a very similar disposition by an arginine in the binding site of the RCK_C domain of KtrA from S. aureus (34), and its conservation in other KhtT proteins (SI Appendix, Fig. S4), together with thermal shift and ITC analysis of the KhtT_R110A mutant (SI Appendix, Fig. S5 A–C), establish that the arginine has a crucial role in the binding of the nucleotide.

KhtTU from B. subtilis Is Activated by c-di-AMP.

We tested the functional impact of c-di-AMP on KhtTU activity using a previously described fluorescence-based flux assay with everted vesicles (38). Everted vesicles or inside-out vesicles are formed during lysis of bacterial cell membranes, according to a well-established protocol (41), and present the cytoplasmic face of the cell membrane toward the bath solution. We generated KhtTU-vesicles from the E. coli strain KNabc (which lacks NhaA, NhaB, and ChaA cation/H⁺ antiporters) expressing the KhtTU complex. In the assay (Fig. 3A), lactate is first added to the bath solution to promote accumulation of H⁺ in the vesicle lumen and generate a H⁺ gradient, quenching the fluorescence of 9-amino-6-chloro-2-methoxyacridine (ACMA) (a H⁺-gradient sensitive dye). Efflux of H⁺ is initiated by adding K⁺ to the bath solution, causing an increase in fluorescence (dequenching), consistent with H⁺ efflux coupled to K⁺ uptake mediated by KhtTU.

Fig. 3. — KhtT is activated by c-di-AMP. (A) Representative fluorescence traces of lactate-dependent quenching and salt-induced (50 mM KCl) dequenching of ACMA for KhtTU-vesicles with (15 µM) and without nucleotides at pH 8.5. (B) Fold change in rate constant determined from fitting single exponentials to dequenching curves for wild-type KhtTU and mutant KhtT_R110AU. Mean ± SD are shown for three to four experiments performed at pH 8.5 and with 50 mM KCl. (C–E) Normalized fluorescence curves for KhtTU-vesicles with increasing c-di-AMP concentrations (shown by color intensity) in the presence of 50 mM KCl at pH 7.5 (C), 8.0 (D), and 8.5 (E). Leak activity, in the absence of KCl and c-di-AMP, is shown by dequenching curve marked by asterisk. (F) Plot of rate constants versus c-di-AMP concentration determined from titrations at different pHs. Mean ± SD are shown for at least four independent experiments from two everted-vesicle preparations, excepting the three lowest c-di-AMP concentrations at pH 8.0, for which three measurements were performed using a single batch of vesicles. Data were fitted with a Hill equation and Hill coefficient (n) determined from the fits K_1/2 = 0.70 ± 0.08 µM and n = 1.76 ± 0.24 (pH 8.5), K_1/2 = 3.58 ± 0.49 µM and n = 2.18 ± 0.44 (pH 8.0), and K_1/2 = 43.4 ± 2.7 µM and n = 1.1 ± 0.1 (pH 7.5). See also *SI Appendix*, Figs. S5–S7.

We performed the assay in the presence of signaling nucleotides (added prior to H⁺ charging) and saw no change in the dequenching curve with pApA, c-di-GMP, ppGpp, or pppGpp relative to no nucleotide addition. However, with c-di-AMP, we observed an increase in fluorescence dequenching, strongly indicating that this nucleotide activates KhtTU (Fig. 3A). In contrast, no changes in dequenching were observed with everted vesicles prepared from cells transformed with empty plasmid (SI Appendix, Fig. S6). To confirm that activation by c-di-AMP resulted from binding to the site shown in the KhtT crystal structure, we replaced the binding site Arg110 for an alanine, produced KhtT_R110AU-vesicles, and measured activity. As expected, this mutant could not be activated by c-di-AMP (Fig. 3B).

Quantification of c-di-AMP activation was done with KhtTU-vesicles titrated at pH 7.5, 8.0, and 8.5 with 50 mM KCl. At all pH values, c-di-AMP was shown to promote the activation of KhtTU in a concentration-dependent manner (Fig. 3 C–E). The rate of dequenching at different concentrations of c-di-AMP was determined by fitting a single exponential to the experimental curves (SI Appendix, Fig. S7). Rate constants were plotted as a function of c-di-AMP concentration, and a Hill equation was fitted to these data (Fig. 3F). At pH 7.5, the K_1/2 of activation by c-di-AMP was ∼40 µM with a Hill coefficient of ∼1. Lower concentrations of c-di-AMP were required to activate KhtTU at higher pH, with K_1/2 = 3.6 ± 0.5 µM (Hill coefficient = ∼2) at pH 8.0 and K_1/2 = 0.70 ± 0.08 µM (Hill coefficient = ∼2) at pH 8.5. Importantly, the 10-fold increase in c-di-AMP sensitivity of KhtTU between pH 7.5 and pH 8.0 is not explained by an increase in the binding affinity of nucleotide for KhtT. Calorimetry titrations of c-di-AMP into KhtT at pH 7.5 showed a K_D ∼40 nM relative to K_D ∼130 nM at pH 8.0 (Table 1 and SI Appendix, Fig. S8). This threefold change in K_D is well below the 10-fold change necessary to indicate a molecularly significant alteration in the ligand binding energy, but if meaningful, it would suggest a decrease in binding affinity at higher pH. The lower constants of dissociation for c-di-AMP binding to KhtT alone relative to the K_1/2 of activation in the KhtTU complex are expected. The energy required for activation of KhtTU is imparted by the chemical energy released from the interactions established by the nucleotide in its binding site, resulting in higher half concentrations of nucleotide for complex activation than for binding to the regulatory protein alone.

In addition, both basal and maximal activities also increased with higher pH. Basal rates (without c-di-AMP) were ∼0.002 s⁻¹ at pH 7.5, ∼0.003 s⁻¹ at pH 8.0, and ∼0.004 s⁻¹ at pH 8.5, and the maximal rates (with saturating concentrations of c-di-AMP) were ∼0.009 s⁻¹ and ∼0.01 s⁻¹ at pH 8.0 and pH 8.5, respectively. We estimated that the maximal rate at pH 7.5 is ∼0.007 s⁻¹ (see section below).

Importantly, these results demonstrate that KhtTU is activated by both c-di-AMP and pH, becoming more active and more sensitive to c-di-AMP at higher pH.

Mechanisms of Activation by c-di-AMP and pH.

We gained clues about the underlying mechanism of activation of the KhtTU complex by both c-di-AMP and pH while setting up the everted vesicle assay. We noticed that the flux properties of vesicles prepared at pH 8.0 differed from those prepared at pH 7.0. In particular, when tested in identical assay conditions, vesicles prepared at pH 8.0 displayed faster and more pronounced dequenching curves and did not respond to activation by c-di-AMP (Fig. 4 A and B). Importantly, addition of recombinant KhtT to the external bath of the vesicles reduced the rate of ACMA dequenching from ∼0.007 s⁻¹ to ∼0.002 s⁻¹ at pH 7.5 and ∼0.01 s⁻¹ to ∼0.004 s⁻¹ at pH 8.5 and resulted in recovery of c-di-AMP sensitivity (Fig. 4C). This demonstrated that KhtT was being washed away from KhtU at high pH, giving rise to KhtU-vesicles that did not respond to c-di-AMP and that it is possible to reassemble the complex by adding the regulatory protein. The binding affinity of KhtT for KhtU at pH 7.5 was estimated from a functional titration of KhtU-vesicles with KhtT, displaying a K_1/2 ∼0.5 µM (SI Appendix, Fig. S5D). The mutant KhtT_R110A behaved similarly. It also disassembled from KhtU at high pH, associated and inhibited KhtU activity as efficiently at pH 7.5, and had similar basal activity (SI Appendix, Fig. S5 D and E).

Fig. 4. — KhtT–KhtU interaction. (A) Representative fluorescence traces for vesicles prepared at pH 7.0 (light gray) or pH 8.0 (dark gray) and assayed under the same experimental conditions at pH 7.5. (B) Representative fluorescence traces for vesicles prepared at pH 8.0 and assayed at pH 7.5 (pink) and pH 8.5 (blue) in the presence (dark pink/blue) or absence (light pink/blue) of 25 µM c-di-AMP. (C) Plot of the rate constants measured from dequenching traces of KhtU-vesicles incubated with increasing concentrations of full-length KhtT at pH 7.5 (light gray) and 8.5 (dark gray), with and without c-di-AMP. All assays in A, B, and C were performed with 50 mM KCl. (D) Representative fluorescence traces for KhtU-vesicles assayed at pH 8.5 and with added KCl, NaCl, choline chloride, or salt mixtures. (E) Plot of rate constants measured from dequenching traces of KhtU-vesicles incubated with 10 or 100 mM KCl at pH 7.5 (light gray) or 8.5 (dark gray). (F) Western blot of supernatant (S) and pellet (P) fractions from ultracentrifugation of KhtU-vesicles+KhtT_His after incubation with c-di-AMP, c-di-GMP, or assay buffer (control) at pH 7.5 and pH 8.0. Blot was probed with α anti-His. (G) Plot of rate constants measured from dequenching traces of KhtU-vesicles incubated with increasing concentrations of KhtT N-domain at pH 7.5 (light gray) and 8.5 (dark gray). Mean ± SD of three to four experiments shown in C, E, and G. (H) Western blot of the pellet fractions from ultracentrifugation of empty or KhtU-vesicles after incubation with N- or C-domain of KhtT (N_His-KhtT and C_His-KhtT, respectively) or full-length KhtT_His. Total inputs (mixture of KhtU-containing vesicles with the KhtT variants) were loaded as positive controls. Total input of KhtT_His is threefold diluted compared to the other inputs, and pellet fractions containing KhtT_His are sixfold diluted compared to the other pellet samples. KhtU-vesicles alone were loaded as negative control. Transfer efficiency into PVDF membrane of KhtT variants varied. See also *SI Appendix*, Figs. S9 and S10.

We can now compare the functional properties of KhtU alone with those of KhtTU and understand better the mechanisms of activation by c-di-AMP and pH. Strikingly, and although some variation in flux rates was observed between vesicles prepared from different cell culture batches, the experiments above reveal a parallel in the flux activity of KhtTU- and KhtU-vesicles. Flux rates for KhtU increase with pH, and in particular, the maximal flux rate measured for KhtTU-vesicles incubated with saturating concentrations of c-di-AMP (see previous section) is identical (∼0.01 s⁻¹) to that measured from KhtU-vesicles without KhtT added (Fig. 4C). This leads us to propose that the fully active state of KhtTU is functionally similar to that of KhtU. Based on this, we estimated the maximum flux rate (0.007 s⁻¹) for KhtTU-vesicles with saturating amounts of c-di-AMP at pH 7.5 (which we could not measure due to the large nucleotide concentrations required) from measurements of KhtU-vesicles at the same pH.

We also wondered what happens with cation selectivity in KhtU, another functional property previously described for KhtTU (38). Addition of 10 mM KCl to KhtU-vesicles (at pH 8.5) caused dequenching that is markedly different from no addition or addition of 10 mM choline salt (an organic cation that will not permeate through KhtU). Adding 100 mM KCl intensified dequenching further (Fig. 4D). In contrast, addition of 10 mM NaCl resembles no addition, while addition of 100 mM NaCl or choline chloride results in further quenching of fluorescence, probably due to a change in vesicle volume caused by increased external osmolarity that is not compensated by KhtU-mediated ion influx. We also tested a mixture of 10 mM KCl and 90 mM NaCl. In these conditions, the fluorescence curve resembles that with 10 mM KCl alone. Experiments at pH 7.5 showed similar results but with less pronounced dequenching curves. Determination of dequenching rates showed that at pH 7.5, there is an increase in ion flux when the K⁺ concentration rises from 10 to 100 mM, with rates changing from ∼0.003 to ∼0.007 s⁻¹. In contrast, at pH 8.5 with 10 mM KCl, flux is already at maximum since there is no change in the dequenching rate with 100 mM KCl, indicating higher affinity of KhtU for K⁺ at pH 8.5 (Fig. 4E). Overall, these results demonstrate that except for c-di-AMP sensitivity, KhtU alone displays similar functional properties to KhtTU (38). It is very selective for K⁺ over Na⁺ and has similar pH dependence, with higher activity at higher pH values.

The falling apart of the KhtTU complex during preparation of everted vesicles at high pH and the similar flux rate measured for KhtU-vesicles and KhtTU-vesicles with saturating concentrations of c-di-AMP raised the possibility that the mechanism of activation by c-d-AMP and pH involves disruption of the interaction between KhtT and KhtU.

To explore this hypothesis, we incubated C-terminal hexahistidine-tagged KhtT (KhtT_His) with KhtU-vesicles at pH 7.0 and isolated the KhtU-vesicles, with associated KhtT_His, by ultracentrifugation. We incubated these vesicles with 200 µM c-di-AMP or c-di-GMP (negative control), a concentration that is five- to 200-fold above K_1/2 of activation for c-di-AMP, at pH 7.5 or pH 8.0 for 10 min. The vesicles were then subjected to a second ultracentrifugation after which free and vesicle-bound KhtT_His was detected by Western blot, in the supernatant and pellet respectively. In contrast to the c-di-GMP control, c-di-AMP reduced KhtT_His associated with the pelleted vesicles and increased the amount of free KhtT_His in the supernatant at both pHs tested (Fig. 4F). In parallel, we showed that everted vesicles prepared from cells coexpressing KhtT_His and KhtU are still activated by c-di-AMP (SI Appendix, Fig. S9A). The Western blot results appear to strongly support the hypothesis that c-di-AMP activation involves detachment of the KhtT:c-di-AMP complex from KhtU. However, comparing the levels of KhtT associated with KhtU-vesicles and vesicles prepared with empty plasmid (empty-vesicles) (SI Appendix, Fig. S9B) after incubation with c-di-AMP clearly shows that not all KhtT detaches from the vesicles despite the relatively long incubation time and the saturating concentration of nucleotide. Moreover, there is little or no difference between experiments performed at pH 7.5 and 8.0 (Fig. 4F), contrasting with what was observed with the preparation of KhtU-vesicles at high pH. It is therefore more likely that c-di-AMP and pH lead to structural alterations that weaken the KhtT–KhtU interaction, resulting in the activation of KhtU-mediated K⁺/H⁺ antiport. In experimental situations that include long incubations and repeated washes the complex falls apart, as seen with vesicles prepared at pH 8.0.

In contrast to the function of the C-domain in nucleotide binding, the function of the N-domain is less clear. We wondered if its role is to mediate the interaction with KhtU and to act as a conduit for the conformational change occurring in the C-domain upon binding/unbinding of the nucleotide. Incubation of different KhtT protein variants (full-length KhtT, N-KhtT [corresponding to the N-domain], and C-KhtT [corresponding to C-domain]) with empty-vesicles or KhtU-vesicles followed by ultracentrifugation revealed the levels of protein associated with KhtU (Fig. 4H). A simple quantification was performed by determining the ratio $(I_{K h t U} - I_{e m p t y}) / I_{i n p u t}$ , in which I_KhtU is the intensity of the signal associated with KhtU-vesicles, I_empty corresponds to background signal associated with the empty-vesicles, and I_input is the signal in the input lanes and accounts for differences is blot transfer efficiency. The ratios were 0.14 and 0.22 for N-KhtT from two independent experiments and 1.21 and 1.04 for KhtT, indicating that the N-domain forms interactions with KhtU but not as efficiently as full-length KhtT. In contrast, the ratio for C-KhtT is very low, 0.01 from a single experiment in which a band was detectable, indicating that C-domain alone is unlikely to bind KhtU. Moreover, unlike full-length KhtT, supplementation of truncated N-domain to KhtU-vesicles up to 10 µM has little or no effect in the dequenching rates at either pH (Fig. 4G).

In addition, in the absence of apo-KhtT crystals, we crystallized the truncated N-domain as we reasoned that truncation of the C-domain could release the constraints imposed by the nucleotide-binding domain and reveal a conformational change in the N-domain of KhtT. We solved the structure of the N-domain of KhtT (N-KhtT protein) (up to Gly68) at different pHs (6.5, 7.5, and 8.5) and in different space groups. Besides local changes in the organization of some loops, we did not observe any major alterations in the overall structure relative to the structure of the full-length KhtT. All the structures could be nicely superposed (SI Appendix, Fig. S10), revealing considerable stability and rigidity of the N-domain, independent of pH and crystal packing.

We conclude that the N-domain alone associates weakly with KhtU, does not inhibit KhtU activity, and that its structure is not affected by the presence of the C-domain or likely by binding of c-di-AMP. We propose therefore that the function of the N-domain is to mediate some of the contacts established between KhtT and KhtU, while the C-domain binds c-di-AMP and undergoes a nucleotide-dependent conformational change that weakens the interaction between the KhtT and KhtU, resulting in the activation of the antiporter.

Role of KhtTU in K⁺ and pH Homeostasis.

We have established that KhtTU is regulated by both c-di-AMP and pH. To understand how these modulatory factors fit with the physiological role of KhtTU, we analyzed the impact of this K⁺ transporter on growth phenotypes in its native bacterium.

The ∆khtU strain (in which the khtU gene was replaced for an antibiotic cassette in the B. subtilis strain 168) was obtained from the Bacillus Genetic Stock Center. It revealed no obvious growth phenotype in a diverse set of growth conditions, with both low and high K⁺ concentrations, relative to the wild-type strain (example shown in SI Appendix, Fig. S11A). In contrast, expression of KhtU in the ∆khtU strain using the bacitracin-inducible plasmid pBS0E in Spizizen Minimal Medium (SMM) at pH ∼7.0 and containing only 2 mM KCl caused a strong growth impairment relative to a strain with the empty vector (Fig. 5A and SI Appendix, Fig. S11 B and C). The growth defect was partially complemented by the presence of 100 mM KCl in the growth media (Fig. 5B). Higher concentrations did not improve the outcome (SI Appendix, Fig. S11D). This is consistent with the proposed role of KhtU as a K⁺ exporter (37), in which unregulated expression results in a toxic effect through a decrease in intracellular K⁺. On the other hand, when we coexpressed KhtU and its regulatory subunit KhtT in the ∆khtU strain, we saw a recovery in bacterial growth relative to expression of KhtU alone, both at 2 mM and 100 mM K⁺ (Fig. 5), indicating that KhtT inhibits the function of KhtU, in accordance with our in vitro data and previous findings (37).

Fig. 5. — Impact of KhtU and KhtTU in the growth of *B. subtilis*. Optical density of cell cultures with strains Δ*khtU* (pBS0E), Δ*khtU* (pBS0E-*khtU*), and Δ*khtU* (pBS0E-*khtTU*) growing in SMM, pH ∼7.0, with 2 mM (A) or 100 mM (B) KCl. Plots show mean ± SD of triplicate growths from a single experiment. Similar results were obtained with experiments performed in different days. The empty plasmid control strain (Δ*khtU* [pBS0E]) was tested with 100 mM KCl in a single experiment. See also *SI Appendix*, Fig. S11.

In vitro experiments have shown that KhtU transports H⁺. Therefore, we analyzed the impact of KhtU and KhtTU expression on pH homeostasis by comparing cell growth in yeast extract-peptone-dextrose (YPD) medium buffered with Bis-Tris propane at pH 7.0 and 9.0. The composition of this medium does not include any added monovalent cation. While all strains grew similarly well at pH 7.0, they did not grow at pH 9.0 over a period of 12 h (Fig. 6 A–C). In contrast, supplementation of the medium at pH 9.0 with 100 mM KCl allowed cells expressing KhtU to grow after a ∼6 h lag period but not the strain with empty plasmid (Fig. 6D). Coexpression of KhtT and KhtU also resulted in cell growth at pH 9.0 with 100 mM KCl but appeared to display a slightly longer lag period. Interestingly, cells coexpressing KhtU and KhtT_R110A mutant (KhtT_R110AU) did not grow, in agreement with the inhibition of KhtU activity by the KhtT mutant that is insensitive to c-di-AMP activation. KhtT_R110A expressed as well or slightly better than wild-type KhtT as inferred from peptide mass spectrometry analysis of cell extracts from ∆khtU strains transformed with empty pBS0E, pBS0E-KhtTU, and pBS0E-KhtT_R110AU (SI Appendix, Fig. S12 and Table S3). Relative KhtT and KhtT_R110A abundance was estimated as ∼4.1 × 10⁸ and ∼7.9 × 10⁸, respectively, after normalization by total peptide amount. In addition, both protein variants were found among the top 10 most abundant proteins of similar molecular size while the strain with empty plasmid showed much lower levels of native KhtT (position 150 in the abundance ranking). Importantly, we have also shown above that the mutant KhtT_R110A inhibits KhtU as efficiently as the wild-type protein (SI Appendix, Fig. S5D) and that at pH 8.0, 8.5, and 9.0 displays slightly higher thermal stability (SI Appendix, Fig. S5 B and F). Altogether, these results demonstrate that the KhtU antiporter can contribute for bacterial adaptation to alkaline conditions and that c-di-AMP has a role in this process.

Fig. 6. — Impact of the expression of KhtTU in the growth of *B. subtilis* at pH 7.0 and 9.0. Optical density of cell cultures with strains Δ*khtU* (pBS0E), Δ*khtU* (pBS0E-*khtU*), Δ*khtU* (pBS0E-*khtTU*), and Δ*khtU* (pBS0E-*khtT*_R110AU) growing in YPD at pH 7.0 (A and B) or pH 9.0 (C and D) without salt supplementation (A and C) or with 100 mM KCl added (B and D). Plots show mean ± SD of triplicate growths from a single experiment. Similar results were obtained with experiments performed in different days. See also *SI Appendix*, Fig. S12 and Table S3.

Discussion

With this work, we have established that the second messenger c-di-AMP directly affects the function of a K⁺ transporter, in this case, as an activator of the K⁺ exporter KhtTU. We have quantified this effect, defined the molecular details of the interaction between c-di-AMP and the cytosolic protein KhtT, and validated the role of KhtTU and of its membrane protein component KhtU in a physiological context. Overall, these data provide strong support to the proposal that c-di-AMP has a central role in the regulation of the K⁺ homeostasis machinery in B. subtilis and in other bacterial species, including pathogens.

Importantly, our data revealed that KhtTU activation by c-di-AMP is highly dependent on pH. A pH increase from 7.5 to 8.5 shifts the K_1/2 of c-di-AMP activation by ∼60-fold, from ∼40 μM at pH 7.5 to less than 1 μM at pH 8.5. This shows that to understand the role of c-di-AMP in the modulation of the bacterial intracellular K⁺ concentration, we need to consider what other factors regulate the function of the K⁺ transporters in the cell and how these factors affect c-di-AMP regulation.

We have also gained good insights into the mechanism of activation of KhtTU by c-di-AMP and pH. In particular, we have found that KhtT is responsible for tight binding of the nucleotide and that, in contrast, KhtU is c-di-AMP insensitive but is activated by high pH, with its flux rate increasing from 0.007 s⁻¹ at pH 7.5 to 0.1 s⁻¹ at pH 8.5. Strikingly, this pH-dependence is mirrored in the KhtTU complex, in which the basal (in the absence of c-di-AMP) and maximal (with saturating concentrations of c-di-AMP) flux rates also increase when pH changes from 7.5 to 8.5. The pH dependence shown by KhtU, including the apparent increase in the affinity for K⁺ at high pH, closely resembles the properties of other proteins from the Cation-Proton-Antiporter (CPA) superfamily such as the well-studied Na⁺/H⁺ antiporter NhaA from E. coli, which lacks a cytosolic subunit (42, 43). Our results also suggest that c-di-AMP binding weakens the KhtT–KhtU interaction. Therefore, we propose a model in which KhtU functions as a “typical” cation/H⁺ antiporter of the CPA superfamily and apo-KhtT acts as a “loose brake” that slows the transport cycle of the antiporter but does not impede it, so that in KhtTU, we observe basal pH-dependent flux even in the absence of c-di-AMP. Upon c-di-AMP binding, there is a conformational change in the C-domain of KhtT that weakens the interaction with KhtU, allowing the membrane protein to undergo its transport cycle at maximum rate. Interestingly, this mechanism resembles that proposed for c-di-AMP regulation of the Trk channel SPD_0076 from S. pneumonia (44). Binding of c-di-AMP to the RCK regulatory protein, CapB, results in the disassembly of the functional complex formed by SPD_076 and CapB and inactivation of K⁺ uptake.

Strikingly, our data showed that expression of KhtU and KhtTU (independently of the original promoter) in the native bacterium B. subtilis affects growth in a K⁺- and pH-dependent way, providing a physiological context for the regulation of KhtTU activity. In particular, expression of KhtU and KhtTU prove to be fundamental for bacterial growth at pH 9.0 when K⁺ is available in the media. It is not straightforward to understand this gain-of-function phenotype since addition of K⁺ to the external medium should decrease the K⁺ gradient that drives H⁺ uptake in a K⁺/H⁺ antiporter. Nevertheless, these results demonstrate that KhtTU can contribute to alkaline adaptation. This gain of function at high pH allowed us to demonstrate that c-di-AMP has a role in the regulation of KhtTU in a physiological context since coexpression of KhtU with a KhtT mutant that is insensitive to c-di-AMP (KhtT_R110A) generated a strain that did not adapt to alkaline conditions.

It has been determined that the average concentration of c-di-AMP in B. subtilis varies between 1 and 5 μM (45) and shown that c-di-AMP concentration rises as intracellular K⁺ increases (17, 26). Therefore, the high-activation K_1/2 determined for KhtTU (∼40 μM) at a “normal” intracellular pH (pH ∼7.5) fits nicely with the idea that with rising concentrations of K⁺ and consequently, rising c-di-AMP, KhtTU becomes more active and transports K⁺ out of the cell. Reduction in intracellular K⁺ concentration is reflected in lower c-di-AMP availability, reducing KhtTU activity back to basal levels. In addition, when intracellular pH becomes too high, both basal and maximal activity of KhtTU and its sensitivity to c-di-AMP increase (K_1/2 < 1 μM), requiring lower amounts of c-di-AMP to fully activate KhtTU. These properties transform KhtTU into a H⁺ importer in alkaline conditions, providing support for a role of KhtTU in pH homeostasis. It is worthwhile pointing out that our in vivo results with the KhtT_R110A mutant indicate that in alkaline conditions, c-di-AMP signaling is still important for the activity of KhtTU, suggesting that besides its role as a master regulator of K⁺ homeostasis, c-di-AMP also has a role in pH homeostasis.

Materials and Methods

Extended materials and methods are described in SI Appendix.

Expression and Purification of KhtT Protein Variants.

Full-length KhtT, KhtT_His, KhtT_R110A, N-KhtT, N_His-KhtT, and C_His-KhtT were expressed in E. coli BL21 (DE3) or B834 (DE3) (for production of selenomethionine-substituted protein) after induction with IPTG at 37 °C. Proteins were purified by affinity chromatography in Ni²⁺-beads and size-exclusion chromatography.

ITC and Thermal Shift Assay.

Recombinant KhtT and KhtT_R110A were titrated with c-di-AMP (or c-di-GMP for KhtT) in a MicroCal VP-ITC instrument (GE Healthcare). Data were visualized and analyzed with Origin 7 Software using a single binding site model. Melting temperatures of KhtT and KhtT_R110A, in presence and in absence of c-di-AMP or c-di-GMP, were determined by Sypro Orange thermal shift assay.

Structure Determination.

Diffraction data were collected at beamlines XALOC of ALBA and PROXIMA-1 of SOLEIL. X-ray diffraction data collection and processing statistics are summarized in SI Appendix, Table S1. The high-resolution structure of KhtT:c-di-AMP complex was solved by single-wavelength anomalous diffraction (SAD), using the anomalous signal of selenium and non-crystallographic symmetry (NCS) averaging. The low-resolution KhtT and its N-domain structures were solved by molecular replacement. Refinement statistics are summarized in SI Appendix, Table S1. Refined coordinates and structure factors were deposited at the Protein Data Bank (https://www.rcsb.org/) (46): 7AGV, high-resolution KhtT structure; 7AHM, low-resolution KhtT structure; 7AGW, N-domain of KhtT at pH 6.5; 7AHT, N-domain of KhtT at pH 7.5; and 7AGY, N-domain of KhtT at pH 8.5.

Preparation of Everted Membrane Vesicles.

Everted membrane vesicles were prepared as previously described (47) from KNabc E. coli strain transformed with pQE60-khtTU (for KhtTU- or KhtU-vesicles), pQE60-khtT_HisU, pQE60-khtT_R110AU, or pQE60 (for empty vesicles) at pH 7.0 (for preparation of KhtTU-, KhtT_R110AU-, KhtT_HisU-, or empty-vesicles) or pH 8.0 (for preparation of KhtU-vesicles).

Antiport Assays.

Fluorescence-based flux assays were performed at room temperature in a FluoroMax-4 spectrofluorometer (Horiba) with excitation at 410 nm and emission at 480 nm. Everted vesicles were mixed with 2 µM ACMA. Assays were initiated with an addition of 4 mM sodium lactate to promote a respiration-generated ΔpH followed by 50 mM KCl (unless otherwise noted) in different conditions.

KhtT Interaction with KhtU.

Everted membrane vesicles produced were incubated with KhtT, N-KhtT, or C-KhtT with C-terminal hexahistidine tags, followed by ultracentrifugations and analysis of supernatant and pellets by Western blot using an antipolyhistidine-tag antibody and horseradish peroxidase (HRP) conjugated secondary antibody. Detection was done using ECL Prime Western Blotting System (GE Healthcare).

B. subtilis Culture Conditions.

For K⁺ homeostasis experiments, B. subtilis strain ΔkhtU transformed with pBS0E, pBS0E-khtU, pBS0E-khtTU, or pBS0E-khtT_R110AU were grown in SMM supplemented with antibiotics and bacitracin (protein expression inducer). All growth medium conditions included 150 mM of monovalent cations (mixture of Na⁺ and K⁺), and pH was ∼7.0. For experiments at different pH, strains were grown in YPD medium, buffered at pH 7.0 or pH 9.0 with and without 100 mM KCl, and supplemented with antibiotics and bacitracin. Expression levels of KhtT and KhtT_R110A in ΔkhtU with pBS0E, pBS0E-khtTU, or pBS0E-khtT_R110AU were estimated by peptide mass spectrometry of cell extracts from cultures grown in YPD medium at pH 7.0, since the ΔkhtU (pBS0E-khtT_R110AU) strain does not grow at pH 9.0.

Supplementary Material

Supplementary File

pnas.2020653118.sapp.pdf^{(2.7MB, pdf)}

Acknowledgments

We acknowledge the SOLEIL and ALBA synchrotrons for access and thank their staff for help with data collection. Support from the Biochemical and Biophysical Technologies, Cell Culture and Genotyping, and X-ray Crystallography scientific platforms of i3S (Porto, Portugal) is also acknowledged. Mass spectrometry analysis was performed by Hugo Osório at the i3S Proteomics Scientific Platform. This work had support from the Portuguese Mass Spectrometry Network, integrated in the National Roadmap of Research Infrastructures of Strategic Relevance (ROTEIRO/0028/2013; LISBOA-01-0145-FEDER-022125). Work was supported by Fundo Europeu de Desenvolvimento Regional (FEDER) funds through the COMPETE 2020 Operational Programme for Competitiveness and Internationalization (POCI), Portugal 2020, and by Portuguese funds through Fundação para a Ciência e a Tecnologia (FCT)/Ministério da Ciência, Tecnologia e Ensino Superior in the framework of the project POCI-01–0145-FEDER-029863(PTDC/BIA-BQM/29863/2017) and by Fundação Luso-Americana para o Desenvolvimento through the FLAD Life Science 2020 award “Bacterial K⁺ transporters are potential antimicrobial targets: mechanisms of transport and regulation.”

Footnotes

The authors declare no competing interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at https://www.pnas.org/lookup/suppl/doi:10.1073/pnas.2020653118/-/DCSupplemental.

Data Availability

Structure coordinates and integrated diffraction data have been deposited in the Protein Data Bank (https://www.rcsb.org/) (7AGV, 7AHM, 7AGW, 7AGY, and 7AHT).

References

1.Zaver S. A., Woodward J. J., Cyclic dinucleotides at the forefront of innate immunity. Curr. Opin. Cell Biol. 63, 49–56 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
2.Krasteva P. V., Sondermann H., Versatile modes of cellular regulation via cyclic dinucleotides. Nat. Chem. Biol. 13, 350–359 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
3.da Purificação A. D., de Azevedo N. M., de Araujo G. G., de Souza R. F., Guzzo G. C. R., The world of cyclic Dinucleotides in bacterial behavior. Molecules 25, 2462 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Commichau F. M., Heidemann J. L., Ficner R., Stülke J., Making and breaking of an essential poison: The cyclases and phosphodiesterases that produce and degrade the essential second messenger cyclic di-AMP in bacteria. J. Bacteriol. 201, e00462-18 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Huynh T. N., Woodward J. J., Too much of a good thing: Regulated depletion of c-di-AMP in the bacterial cytoplasm. Curr. Opin. Microbiol. 30, 22–29 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
6.Fahmi T., Port G. C., Cho K. H., c-di-AMP: An essential molecule in the signaling pathways that regulate the viability and virulence of gram-positive bacteria. Genes (Basel) 8, 197 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Commichau F. M., Dickmanns A., Gundlach J., Ficner R., Stülke J., A jack of all trades: The multiple roles of the unique essential second messenger cyclic di-AMP. Mol. Microbiol. 97, 189–204 (2015). [DOI] [PubMed] [Google Scholar]
8.Stülke J., Krüger L., Cyclic di-AMP signaling in bacteria. Annu. Rev. Microbiol. 74, 159–179 (2020). [DOI] [PubMed] [Google Scholar]
9.Sureka K., et al., The cyclic dinucleotide c-di-AMP is an allosteric regulator of metabolic enzyme function. Cell 158, 1389–1401 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Luo Y., Helmann J. D., Analysis of the role of Bacillus subtilis σ(M) in β-lactam resistance reveals an essential role for c-di-AMP in peptidoglycan homeostasis. Mol. Microbiol. 83, 623–639 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]
11.Corrigan R. M., Abbott J. C., Burhenne H., Kaever V., Gründling A., c-di-AMP is a new second messenger in Staphylococcus aureus with a role in controlling cell size and envelope stress. PLoS Pathog. 7, e1002217 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Peng X., Zhang Y., Bai G., Zhou X., Wu H., Cyclic di-AMP mediates biofilm formation. Mol. Microbiol. 99, 945–959 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Gundlach J., Rath H., Herzberg C., Mäder U., Stülke J., Second messenger signaling in Bacillus subtilis: Accumulation of cyclic di-AMP inhibits biofilm formation. Front. Microbiol. 7, 804 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
14.Bejerano-Sagie M., et al., A checkpoint protein that scans the chromosome for damage at the start of sporulation in Bacillus subtilis. Cell 125, 679–690 (2006). [DOI] [PubMed] [Google Scholar]
15.Cho K. H., Kang S. O., Streptococcus pyogenes c-di-AMP phosphodiesterase, GdpP, influences SpeB processing and virulence. PLoS One 8, e69425 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Gundlach J., Commichau F. M., Stülke J., Perspective of ions and messengers: An intricate link between potassium, glutamate, and cyclic di-AMP. Curr. Genet. 64, 191–195 (2018). [DOI] [PubMed] [Google Scholar]
17.Gundlach J., et al., Control of potassium homeostasis is an essential function of the second messenger cyclic di-AMP in Bacillus subtilis. Sci. Signal. 10, eaal3011 (2017). [DOI] [PubMed] [Google Scholar]
18.Gundlach J., et al., Adaptation of Bacillus subtilis to life at extreme potassium limitation. mBio 8, e00861-17 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
19.Zarrella T. M., Metzger D. W., Bai G., Stress suppressor screening leads to detection of regulation of cyclic di-AMP homeostasis by a Trk family effector protein in Streptococcus pneumoniae. J. Bacteriol. 200, e00045-18 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
20.Rozov A., et al., Importance of potassium ions for ribosome structure and function revealed by long-wavelength X-ray diffraction. Nat. Commun. 10, 2519 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
21.Bakker E. P., Mangerich W. E., Interconversion of components of the bacterial proton motive force by electrogenic potassium transport. J. Bacteriol. 147, 820–826 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]
22.Kroll R. G., Booth I. R., The role of potassium transport in the generation of a pH gradient in Escherichia coli. Biochem. J. 198, 691–698 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]
23.Meury J., Robin A., Monnier-Champeix P., Turgor-controlled K+ fluxes and their pathways in Escherichia coli. Eur. J. Biochem. 151, 613–619 (1985). [DOI] [PubMed] [Google Scholar]
24.Epstein W., Schultz S. G., Cation transport in Escherichia coli: V. Regulation of cation content. J. Gen. Physiol. 49, 221–234 (1965). [DOI] [PMC free article] [PubMed] [Google Scholar]
25.Whatmore A. M., Chudek J. A., Reed R. H., The effects of osmotic upshock on the intracellular solute pools of Bacillus subtilis. J. Gen. Microbiol. 136, 2527–2535 (1990). [DOI] [PubMed] [Google Scholar]
26.Pham H. T., et al., Enhanced uptake of potassium or glycine betaine or export of cyclic-di-AMP restores osmoresistance in a high cyclic-di-AMP Lactococcus lactis mutant. PLoS Genet. 14, e1007574 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
27.Devaux L., et al., Cyclic di-AMP regulation of osmotic homeostasis is essential in Group B Streptococcus. PLoS Genet. 14, e1007342 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
28.Zeden M. S., et al., Cyclic di-adenosine monophosphate (c-di-AMP) is required for osmotic regulation in Staphylococcus aureus but dispensable for viability in anaerobic conditions. J. Biol. Chem. 293, 3180–3200 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
29.Quintana I. M., et al., The KupA and KupB proteins of Lactococcus lactis IL1403 are novel c-di-AMP receptor proteins responsible for potassium uptake. J. Bacteriol. 201, e00028-19 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
30.Gundlach J., et al., Sustained sensing in potassium homeostasis: Cyclic di-AMP controls potassium uptake by KimA at the levels of expression and activity. J. Biol. Chem. 294, 9605–9614 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Nelson J. W., et al., Riboswitches in eubacteria sense the second messenger c-di-AMP. Nat. Chem. Biol. 9, 834–839 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Chin K. H., et al., Structural insights into the distinct binding mode of cyclic di-AMP with SaCpaA_RCK. Biochemistry 54, 4936–4951 (2015). [DOI] [PubMed] [Google Scholar]
33.Rocha R., Teixeira-Duarte C. M., Jorge J. M. P., Morais-Cabral J. H., Characterization of the molecular properties of KtrC, a second RCK domain that regulates a Ktr channel in Bacillus subtilis. J. Struct. Biol. 205, 34–43 (2019). [DOI] [PubMed] [Google Scholar]
34.Kim H., et al., Structural studies of potassium transport protein KtrA regulator of conductance of K+ (RCK) C domain in complex with cyclic diadenosine monophosphate (c-di-AMP). J. Biol. Chem. 290, 16393–16402 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Corrigan R. M., et al., Systematic identification of conserved bacterial c-di-AMP receptor proteins. Proc. Natl. Acad. Sci. U.S.A. 110, 9084–9089 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Moscoso J. A., et al., Binding of cyclic Di-AMP to the Staphylococcus aureus sensor kinase KdpD occurs via the universal stress protein domain and downregulates the expression of the Kdp potassium transporter. J. Bacteriol. 198, 98–110 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
37.Fujisawa M., Wada Y., Ito M., Modulation of the K+ efflux activity of Bacillus subtilis YhaU by YhaT and the C-terminal region of YhaS. FEMS Microbiol. Lett. 231, 211–217 (2004). [DOI] [PubMed] [Google Scholar]
38.Fujisawa M., Ito M., Krulwich T. A., Three two-component transporters with channel-like properties have monovalent cation/proton antiport activity. Proc. Natl. Acad. Sci. U.S.A. 104, 13289–13294 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
39.Guo Q., et al., Using a phenotype microarray and transcriptome analysis to elucidate multi-drug resistance regulated by the PhoR/PhoP two-component system in Bacillus subtilis strain NCD-2. Microbiol. Res. 239, 126557 (2020). [DOI] [PubMed] [Google Scholar]
40.Chandrangsu P., Dusi R., Hamilton C. J., Helmann J. D., Methylglyoxal resistance in Bacillus subtilis: Contributions of bacillithiol-dependent and independent pathways. Mol. Microbiol. 91, 706–715 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
41.Rosen B. P., Ion extrusion systems in Escherichia coli. Methods Enzymol. 125, 328–336 (1986). [DOI] [PubMed] [Google Scholar]
42.Taglicht D., Padan E., Schuldiner S., Overproduction and purification of a functional Na+/H+ antiporter coded by nhaA (ant) from Escherichia coli. J. Biol. Chem. 266, 11289–11294 (1991). [PubMed] [Google Scholar]
43.Padan E., Functional and structural dynamics of NhaA, a prototype for Na(+) and H(+) antiporters, which are responsible for Na(+) and H(+) homeostasis in cells. Biochim. Biophys. Acta 1837, 1047–1062 (2014). [DOI] [PubMed] [Google Scholar]
44.Bai Y., et al., Cyclic di-AMP impairs potassium uptake mediated by a cyclic di-AMP binding protein in Streptococcus pneumoniae. J. Bacteriol. 196, 614–623 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Oppenheimer-Shaanan Y., Wexselblatt E., Katzhendler J., Yavin E., Ben-Yehuda S., c-di-AMP reports DNA integrity during sporulation in Bacillus subtilis. EMBO Rep. 12, 594–601 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Berman H. M., et al., The protein data bank. Nucleic Acids Res. 28, 235–242 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]
47.Rosen B. P., Tsuchiya T., Preparation of everted membrane vesicles from Escherichia coli for the measurement of calcium transport. Methods Enzymol. 56, 233–241 (1979). [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary File

pnas.2020653118.sapp.pdf^{(2.7MB, pdf)}

Data Availability Statement

Structure coordinates and integrated diffraction data have been deposited in the Protein Data Bank (https://www.rcsb.org/) (7AGV, 7AHM, 7AGW, 7AGY, and 7AHT).

[r1] 1.Zaver S. A., Woodward J. J., Cyclic dinucleotides at the forefront of innate immunity. Curr. Opin. Cell Biol. 63, 49–56 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r2] 2.Krasteva P. V., Sondermann H., Versatile modes of cellular regulation via cyclic dinucleotides. Nat. Chem. Biol. 13, 350–359 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r3] 3.da Purificação A. D., de Azevedo N. M., de Araujo G. G., de Souza R. F., Guzzo G. C. R., The world of cyclic Dinucleotides in bacterial behavior. Molecules 25, 2462 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r4] 4.Commichau F. M., Heidemann J. L., Ficner R., Stülke J., Making and breaking of an essential poison: The cyclases and phosphodiesterases that produce and degrade the essential second messenger cyclic di-AMP in bacteria. J. Bacteriol. 201, e00462-18 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r5] 5.Huynh T. N., Woodward J. J., Too much of a good thing: Regulated depletion of c-di-AMP in the bacterial cytoplasm. Curr. Opin. Microbiol. 30, 22–29 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r6] 6.Fahmi T., Port G. C., Cho K. H., c-di-AMP: An essential molecule in the signaling pathways that regulate the viability and virulence of gram-positive bacteria. Genes (Basel) 8, 197 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r7] 7.Commichau F. M., Dickmanns A., Gundlach J., Ficner R., Stülke J., A jack of all trades: The multiple roles of the unique essential second messenger cyclic di-AMP. Mol. Microbiol. 97, 189–204 (2015). [DOI] [PubMed] [Google Scholar]

[r8] 8.Stülke J., Krüger L., Cyclic di-AMP signaling in bacteria. Annu. Rev. Microbiol. 74, 159–179 (2020). [DOI] [PubMed] [Google Scholar]

[r9] 9.Sureka K., et al., The cyclic dinucleotide c-di-AMP is an allosteric regulator of metabolic enzyme function. Cell 158, 1389–1401 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r10] 10.Luo Y., Helmann J. D., Analysis of the role of Bacillus subtilis σ(M) in β-lactam resistance reveals an essential role for c-di-AMP in peptidoglycan homeostasis. Mol. Microbiol. 83, 623–639 (2012). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r11] 11.Corrigan R. M., Abbott J. C., Burhenne H., Kaever V., Gründling A., c-di-AMP is a new second messenger in Staphylococcus aureus with a role in controlling cell size and envelope stress. PLoS Pathog. 7, e1002217 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r12] 12.Peng X., Zhang Y., Bai G., Zhou X., Wu H., Cyclic di-AMP mediates biofilm formation. Mol. Microbiol. 99, 945–959 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r13] 13.Gundlach J., Rath H., Herzberg C., Mäder U., Stülke J., Second messenger signaling in Bacillus subtilis: Accumulation of cyclic di-AMP inhibits biofilm formation. Front. Microbiol. 7, 804 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r14] 14.Bejerano-Sagie M., et al., A checkpoint protein that scans the chromosome for damage at the start of sporulation in Bacillus subtilis. Cell 125, 679–690 (2006). [DOI] [PubMed] [Google Scholar]

[r15] 15.Cho K. H., Kang S. O., Streptococcus pyogenes c-di-AMP phosphodiesterase, GdpP, influences SpeB processing and virulence. PLoS One 8, e69425 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r16] 16.Gundlach J., Commichau F. M., Stülke J., Perspective of ions and messengers: An intricate link between potassium, glutamate, and cyclic di-AMP. Curr. Genet. 64, 191–195 (2018). [DOI] [PubMed] [Google Scholar]

[r17] 17.Gundlach J., et al., Control of potassium homeostasis is an essential function of the second messenger cyclic di-AMP in Bacillus subtilis. Sci. Signal. 10, eaal3011 (2017). [DOI] [PubMed] [Google Scholar]

[r18] 18.Gundlach J., et al., Adaptation of Bacillus subtilis to life at extreme potassium limitation. mBio 8, e00861-17 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r19] 19.Zarrella T. M., Metzger D. W., Bai G., Stress suppressor screening leads to detection of regulation of cyclic di-AMP homeostasis by a Trk family effector protein in Streptococcus pneumoniae. J. Bacteriol. 200, e00045-18 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r20] 20.Rozov A., et al., Importance of potassium ions for ribosome structure and function revealed by long-wavelength X-ray diffraction. Nat. Commun. 10, 2519 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r21] 21.Bakker E. P., Mangerich W. E., Interconversion of components of the bacterial proton motive force by electrogenic potassium transport. J. Bacteriol. 147, 820–826 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r22] 22.Kroll R. G., Booth I. R., The role of potassium transport in the generation of a pH gradient in Escherichia coli. Biochem. J. 198, 691–698 (1981). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r23] 23.Meury J., Robin A., Monnier-Champeix P., Turgor-controlled K+ fluxes and their pathways in Escherichia coli. Eur. J. Biochem. 151, 613–619 (1985). [DOI] [PubMed] [Google Scholar]

[r24] 24.Epstein W., Schultz S. G., Cation transport in Escherichia coli: V. Regulation of cation content. J. Gen. Physiol. 49, 221–234 (1965). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r25] 25.Whatmore A. M., Chudek J. A., Reed R. H., The effects of osmotic upshock on the intracellular solute pools of Bacillus subtilis. J. Gen. Microbiol. 136, 2527–2535 (1990). [DOI] [PubMed] [Google Scholar]

[r26] 26.Pham H. T., et al., Enhanced uptake of potassium or glycine betaine or export of cyclic-di-AMP restores osmoresistance in a high cyclic-di-AMP Lactococcus lactis mutant. PLoS Genet. 14, e1007574 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r27] 27.Devaux L., et al., Cyclic di-AMP regulation of osmotic homeostasis is essential in Group B Streptococcus. PLoS Genet. 14, e1007342 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r28] 28.Zeden M. S., et al., Cyclic di-adenosine monophosphate (c-di-AMP) is required for osmotic regulation in Staphylococcus aureus but dispensable for viability in anaerobic conditions. J. Biol. Chem. 293, 3180–3200 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r29] 29.Quintana I. M., et al., The KupA and KupB proteins of Lactococcus lactis IL1403 are novel c-di-AMP receptor proteins responsible for potassium uptake. J. Bacteriol. 201, e00028-19 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r30] 30.Gundlach J., et al., Sustained sensing in potassium homeostasis: Cyclic di-AMP controls potassium uptake by KimA at the levels of expression and activity. J. Biol. Chem. 294, 9605–9614 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r31] 31.Nelson J. W., et al., Riboswitches in eubacteria sense the second messenger c-di-AMP. Nat. Chem. Biol. 9, 834–839 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r32] 32.Chin K. H., et al., Structural insights into the distinct binding mode of cyclic di-AMP with SaCpaA_RCK. Biochemistry 54, 4936–4951 (2015). [DOI] [PubMed] [Google Scholar]

[r33] 33.Rocha R., Teixeira-Duarte C. M., Jorge J. M. P., Morais-Cabral J. H., Characterization of the molecular properties of KtrC, a second RCK domain that regulates a Ktr channel in Bacillus subtilis. J. Struct. Biol. 205, 34–43 (2019). [DOI] [PubMed] [Google Scholar]

[r34] 34.Kim H., et al., Structural studies of potassium transport protein KtrA regulator of conductance of K+ (RCK) C domain in complex with cyclic diadenosine monophosphate (c-di-AMP). J. Biol. Chem. 290, 16393–16402 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r35] 35.Corrigan R. M., et al., Systematic identification of conserved bacterial c-di-AMP receptor proteins. Proc. Natl. Acad. Sci. U.S.A. 110, 9084–9089 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r36] 36.Moscoso J. A., et al., Binding of cyclic Di-AMP to the Staphylococcus aureus sensor kinase KdpD occurs via the universal stress protein domain and downregulates the expression of the Kdp potassium transporter. J. Bacteriol. 198, 98–110 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r37] 37.Fujisawa M., Wada Y., Ito M., Modulation of the K+ efflux activity of Bacillus subtilis YhaU by YhaT and the C-terminal region of YhaS. FEMS Microbiol. Lett. 231, 211–217 (2004). [DOI] [PubMed] [Google Scholar]

[r38] 38.Fujisawa M., Ito M., Krulwich T. A., Three two-component transporters with channel-like properties have monovalent cation/proton antiport activity. Proc. Natl. Acad. Sci. U.S.A. 104, 13289–13294 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r39] 39.Guo Q., et al., Using a phenotype microarray and transcriptome analysis to elucidate multi-drug resistance regulated by the PhoR/PhoP two-component system in Bacillus subtilis strain NCD-2. Microbiol. Res. 239, 126557 (2020). [DOI] [PubMed] [Google Scholar]

[r40] 40.Chandrangsu P., Dusi R., Hamilton C. J., Helmann J. D., Methylglyoxal resistance in Bacillus subtilis: Contributions of bacillithiol-dependent and independent pathways. Mol. Microbiol. 91, 706–715 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r41] 41.Rosen B. P., Ion extrusion systems in Escherichia coli. Methods Enzymol. 125, 328–336 (1986). [DOI] [PubMed] [Google Scholar]

[r42] 42.Taglicht D., Padan E., Schuldiner S., Overproduction and purification of a functional Na+/H+ antiporter coded by nhaA (ant) from Escherichia coli. J. Biol. Chem. 266, 11289–11294 (1991). [PubMed] [Google Scholar]

[r43] 43.Padan E., Functional and structural dynamics of NhaA, a prototype for Na(+) and H(+) antiporters, which are responsible for Na(+) and H(+) homeostasis in cells. Biochim. Biophys. Acta 1837, 1047–1062 (2014). [DOI] [PubMed] [Google Scholar]

[r44] 44.Bai Y., et al., Cyclic di-AMP impairs potassium uptake mediated by a cyclic di-AMP binding protein in Streptococcus pneumoniae. J. Bacteriol. 196, 614–623 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r45] 45.Oppenheimer-Shaanan Y., Wexselblatt E., Katzhendler J., Yavin E., Ben-Yehuda S., c-di-AMP reports DNA integrity during sporulation in Bacillus subtilis. EMBO Rep. 12, 594–601 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r46] 46.Berman H. M., et al., The protein data bank. Nucleic Acids Res. 28, 235–242 (2000). [DOI] [PMC free article] [PubMed] [Google Scholar]

[r47] 47.Rosen B. P., Tsuchiya T., Preparation of everted membrane vesicles from Escherichia coli for the measurement of calcium transport. Methods Enzymol. 56, 233–241 (1979). [DOI] [PubMed] [Google Scholar]

PERMALINK

c-di-AMP, a likely master regulator of bacterial K+ homeostasis machinery, activates a K+ exporter

Tatiana B Cereija

João P L Guerra

João M P Jorge

João H Morais-Cabral

Significance

Abstract

Results

Structure of KhtT in Complex with c-di-AMP.

Fig. 1.

Table 1.

Fig. 2.

KhtTU from B. subtilis Is Activated by c-di-AMP.

Fig. 3.