Onward and [K⁺]Upward: a New Potassium Importer under the Spell of Cyclic di-AMP

Cyclic di-AMP (c-di-AMP) is a second messenger which plays a major role in osmotic homeostasis in bacteria. In work by Quintana et al. (I.

ABSTRACT

Cyclic di-AMP (c-di-AMP) is a second messenger which plays a major role in osmotic homeostasis in bacteria. In work by Quintana et al. (I. M. Quintana, J. Gibhardt, A. Turdiev, E. Hammer, et al., J Bacteriol 201:e00028-19, 2019, https://doi.org/10.1128/jb.00028-19), two Kup homologs from Lactococcus lactis were identified as high-affinity K⁺ importers whose activities are inhibited by direct binding of c-di-AMP. The results broaden the scope of K⁺ level regulation by c-di-AMP, with Kup homologs found in a number of pathogenic, commensal, and industrial bacteria.

COMMENTARY

Research on the second messenger cyclic di-AMP (c-di-AMP) has recently passed the 10-year mark following its discovery in a crystal structure of the DNA scanning protein DisA (1). A significant amount of work exploring how it stimulates the host immune system and why it is essential in many bacteria under normal growth conditions has occurred (2, 3). Levels of this nucleotide in the cell are controlled through regulated synthesis and degradation by diadenylate cyclase (DAC) and phosphodiesterase (PDE) enzymes, respectively. In most bacteria, one DAC is present (CdaA/DacA), while one or two c-di-AMP PDEs can exist (GdpP or PgpH) (2). A number of c-di-AMP binding receptors have been identified (4–9), which have provided some insight into its physiological role. Most characterized protein receptors regulate potassium (K⁺) or compatible solute (e.g., carnitine and glycine betaine) levels (2). The c-di-AMP binding ydaO riboswitch also regulates K⁺ and compatible solute transporters, amino acid transporters, and cell wall-modifying enzymes (9). A commonly observed phenotype of bacteria with elevated c-di-AMP levels is osmosensitivity (5, 10–14), and mutants with low or no c-di-AMP can resume growth in complex media with high osmolarity or chemically defined media with low levels of K⁺ (15–17). Together, these findings have provided evidence for an osmotic homeostasis role of c-di-AMP via its regulation of water movement (osmosis) and control of cell turgor (18).

In work by Quintana et al. (19), a new component of the c-di-AMP signaling pathway has been uncovered. From Lactococcus lactis, a major cheese starter culture and model lactic acid bacterium, a selection of 8 potential c-di-AMP binding candidate proteins were expressed in Escherichia coli. Two homologous K⁺ uptake (Kup) proteins of the Kup/HAK/KT family (PF02705) were identified as the strongest binders of radiolabeled c-di-AMP using the differential radial capillary action of ligand assay (DRaCALA). Other tested proteins showed low but significant binding, including the oligopeptide binding protein OppA, which was also identified in a screen for c-di-AMP binding receptors from Listeria monocytogenes cell lysates (4). L. monocytogenes OppA did not bind c-di-AMP when expressed from E. coli, however, and it was suggested that it interacts with another protein that directly binds c-di-AMP or that posttranslational modifications are needed for c-di-AMP binding (4). It remains to be verified if OppA and other proteins tested by Quintana et al. (19) interact with c-di-AMP and, if so, whether the interaction is direct or indirect. Proteins being membrane bound or embedded presents a challenge, since these proteins typically are expressed poorly in E. coli.

Next, to determine if L. lactis KupA and KupB are functional K⁺ transporters, an E. coli strain devoid of its primary K⁺ importer genes (TrkH, TrkG, Kup, and KdpABC genes) and unable to grow in K⁺ levels below 10 mM was utilized (20). The introduction of genes encoding either L. lactis KupA or KupB was found to rescue growth in media with very low K⁺ concentrations (≤0.025 mM). Rescue of growth in low K⁺ concentrations was also observed upon introduction of the high-affinity K⁺ importer genes ktrAB from Bacillus subtilis but not with the putative low-affinity transporter genes ktrCD from L. monocytogenes. E. coli provides an excellent host to analyze the impact of c-di-AMP on specific targets, since it does not naturally contain c-di-AMP. It was found that expression of the c-di-AMP synthesis gene cdaA from L. monocytogenes in a K⁺ transport-deficient E. coli strain with either Kup protein inhibited growth under low K⁺ conditions, demonstrating that c-di-AMP inhibits Kup activity. A similar strategy was employed by Bai et al., where expression of the c-di-AMP synthase DisA in a K⁺ transporter-deficient E. coli strain containing the Streptococcus pneumoniae TrkH/CabP K⁺ transporter prevented growth in low-K⁺ media (6). However, differently from Kup, which is a single-c-di-AMP binding K⁺ transport protein, the cytosolic CabP-gating component of the K⁺ transporter complex binds c-di-AMP, which inhibits its interaction with potentially multiple-membrane-embedded TrkH K⁺ channels (6, 21).

The Kup proteins are clearly distinct from other known c-di-AMP-regulated K⁺ transporters and do not contain any known c-di-AMP binding domains. The Ktr/Trk family of c-di-AMP binding K⁺ transporters bind c-di-AMP via a regulator of the conductance of K⁺ (RCK_C or TrkA_C) domain (PF02080) (5, 6). The two other previously known K⁺ transporters are under the control of c-di-AMP at a transcriptional level. Expression of the KdpFABC K⁺ transporter is regulated by the two-component sensor kinase response regulator KdpDE proteins (22). KdpD binds c-di-AMP via its universal stress protein (USP) domain, which inhibits kdpFABC expression and, ultimately, K⁺ uptake (22). KimA is the other known c-di-AMP-regulated K⁺ transporter which is located downstream of the ydaO family riboswitch (17). c-di-AMP binding to the riboswitch reduces the expression of kimA, leading to lower K⁺ transport (17). In some bacteria, ydaO riboswitches are also found upstream of trk, ktr, and kdp genes (9). Similarly, ydaO family riboswitches have been identified upstream of kup genes in several Proteobacteria (Pelobacter, Geobacter, Syntrophobacter, and Desulfovibrio spp.) and Verrucomicrobia (Opitutus spp.) (9). Therefore, multiple modes of c-di-AMP regulation of at least 3 different K⁺ transporters in bacteria are likely. This allows for rapid responses by modulating transporter activity or slower responses through adjustments in gene expression in preparation for environmental osmotic challenges. Whether or not different threshold c-di-AMP levels in the cell are necessary to trigger this differential regulation remains to be determined.

Kup family proteins are conserved across phyla, being present in a wide range of bacteria (Firmicutes, Proteobacteria, Actinobacteria, and Bacteriodetes), as well as eukaryotes, including yeast and plants. Since many of these Kup-carrying organisms do not produce c-di-AMP, alternative modes of regulation must exist. Most work on Kup has been performed in plants, where it is found typically in many copies (e.g., 27 copies in rice) and has been shown to play an important role in K⁺ acquisition and resistance to osmotic stress (23). In plants, kup expression is regulated by external K⁺ levels and osmolarity changes, with posttranslational regulation occurring through phosphorylation of the C-terminal intracellular domain (24). While the 12-transmembrane section of Kup is relatively well conserved (39% amino acid identity between L. lactis IL403 KupA and E. coli Kup), the intracellular C-terminal ∼200-amino-acid intracellular domain is less well conserved (21% amino acid identity between L. lactis IL403 KupA and E. coli Kup). The role of the C-terminal domain is not known; however, the deletion of it in E. coli Kup resulted in reduced, but not absent, K⁺ transport activity (25). Recently, in a suppressor screen for osmoresistance in a high-c-di-AMP gdpP mutant of L. lactis MG1363, several mutations in the KupB C-terminal domain were identified which increased K⁺ uptake (26), suggesting that this domain plays a regulatory role. Whether or not these gain-of-function amino acid changes simply elevate KupB activity or hinder c-di-AMP-mediated inhibition remains to be determined. Structural analysis and localization of the c-di-AMP binding site in Kup will perhaps provide an answer.

As noted by Quintana et al., Kup proteins are present in all L. lactis strains, while Kdp and Ktr systems are found in only a subset of strains. We analyzed 149 annotated genomes of L. lactis present in the National Center for Biotechnology Information (NCBI) database and also observed that all strains contain at least one kup gene. Some strains contain a frameshift mutation or transposon insertion element inactivating one of their kup genes, while other strains have additional kup genes located in distal locations within mobile genetic elements. The reason for multiple high-affinity Kup proteins in L. lactis is not known; however, one possibility is that they exhibit differences in c-di-AMP binding affinity allowing for finely tuned K⁺ uptake control over a broad range of intracellular c-di-AMP levels. Although binding affinities were not determined, Quintana et al. observed that KupA bound to c-di-AMP more strongly than KupB in all DRaCALA experiments (19), which is in agreement with this hypothesis. The kdp system was the next most common K⁺ importer and was found in 61% of L. lactis genomes, while the Ktr/Trk family of transporters was less common, again with only 39% of strains containing at least one representative. Interestingly, the Kdp system was much more common in L. lactis subsp. lactis (90% of strains) than in L. lactis subsp. cremoris (3% of strains). Grouping of subspecies of L. lactis is done using phenotypic methods; however, strains of L. lactis subsp. cremoris, which are genetically related, can also have “L. lactis subsp. lactis phenotypes.” Interestingly, one of the L. lactis subsp. lactis phenotypes is higher osmotolerance (tolerance to 3% NaCl). However, this is not solely due to Kdp, since L. lactis subsp. lactis strain IL1403, which is salt resistant, lacks this transporter, and several kdp-lacking L. lactis subsp. cremoris strains are also osmoresistant (27). Recent analyses using genotype-phenotype matching approaches with 43 L. lactis genomes have been unable to a identify specific gene(s), based on a presence/absence analysis, that is responsible for differences in osmoresistance between strains (28).

Another c-di-AMP receptor identified recently in L. lactis is the transcriptional regulator BusR (26). BusR represses expression of the compatible solute glycine betaine transporter BusAA-AB (29), and intracellular levels of glycine betaine have been found to strongly correlate with the salt tolerance level of lactococcal strains (27). BusR has been found to bind to the busAA-AB promoter in response to ionic strength (KCl concentration) in vitro, even in the absence of c-di-AMP (30). Quintana et al. (19) demonstrated that c-di-AMP regulates K⁺ uptake by Kup, which may also have a follow-on effect of modulating the intracellular ionic strength and, thus, BusR repression of busAA-AB. Therefore, c-di-AMP-mediated regulation of glycine betaine uptake may be both indirect, through regulation of Kup activity, and direct via binding to BusR.

One area in which only limited progress has been made in c-di-AMP signaling research is around the signal(s) that feeds into the c-di-AMP system and understanding how environmental changes are sensed by the DAC and PDE enzymes. With many c-di-AMP receptors being involved in K⁺ and compatible solute import, the most obvious signal would be related to the level of these osmolytes in the cell or their effects of cell turgor pressure. Indeed, several studies have indicated that K⁺ levels positively correlate with c-di-AMP levels. Growth of B. subtilis in K⁺-abundant medium led to a 2-fold increase in c-di-AMP due to higher CdaA expression (17), while inactivation of the Trk gating component CabP in S. pneumoniae resulted in lower c-di-AMP levels, most likely due to lower K⁺ uptake (13). Gain-of-function mutations in KupB leading to higher intracellular K⁺ levels in L. lactis resulted in elevated c-di-AMP levels (26). These results combined with the findings of Quintana et al. (19) provide additional support for the existence of a negative-feedback loop whereby imported K⁺ stimulates higher c-di-AMP levels which in turn inhibit K⁺ import. How K⁺ levels are monitored by the DAC or PDE enzymes is not known and if this is direct sensing of [K⁺] or indirect sensing possibly due to the expected increase in cell turgor. Higher c-di-AMP levels were found in a L. lactis busR mutant, which has higher intracellular glycine betaine levels (26), which suggests that the signals feeding into the c-di-AMP system are broader than [K⁺].

c-di-AMP-mediated regulation of Kup is likely to occur in other Gram-positive bacteria containing this second messenger, including food fermentation and commensal bacteria (Lactobacillus and Bifidobacterium spp.) and pathogens (Streptococcus pyogenes, Enterococcus faecalis, and Clostridium perfringens). In contrast to L. lactis, however, these bacteria contain only a single copy of kup. While screening efforts have identified c-di-AMP binding proteins in bacteria, including Mycobacterium smegmatis, Staphylococcus aureus, S. pneumoniae, and L. monocytogenes (4–7, 31), the work of Quintana et al. (19) demonstrates that there are almost certainly more players in the c-di-AMP signaling pathway yet to be identified in unexplored bacteria. More broadly, the recent identification of many bacterial enzymes capable of synthesizing a diverse range of both purine- and pyrimidine-containing cyclic dinucleotides and a cyclic trinucleotide (32) is likely to significantly extend knowledge of the influence that nucleotide second messengers have in bacterial physiological processes and host-microbe interactions.

The views expressed in this article do not necessarily reflect the views of the journal or of ASM.