Dissecting the contribution of Kup and KimA to Enterococcus faecalis potassium homeostasis

Abstract

In previous studies, we highlighted the remarkable diversity of potassium transport systems in Enterococcus faecalis. This microorganism encodes five distinct potassium transport systems: the Kdp complex (KdpFABC and KdpED), the Ktr systems (KtrAB and KtrAD), and transporters from the Kup family, specifically Kup and KimA. Research focused on the E. faecalis JH2-2 strain, which, unlike E. faecalis V583, lacks the Kdp complex and exhibits an altered expression of the kup gene due to the insertion of the IS6770 sequence. Here, we characterize the kimA-kup locus of the JH2-2 strain, which harbors genes encoding a metal transport protein (MntE), the Opu osmoresistance system (OpuCA-CB-CC-CD), and a Ca²⁺-ATPase (CaaT). It was determined that the kimA gene is monocistronic and transcribed from a consensus vegetative promoter featuring conserved -10 (TAgcAT) and -35 (TTGACt) boxes. RT-qPCR analysis revealed that kimA expression was significantly upregulated under acidic stress. Furthermore, both kimA and kup expression are increased under alkaline conditions in the presence of NaCl. To dissect functional roles, growth parameters of mutant strains KimA⁺ (ΔktrAΔkup) and Kup⁺ (ΔktrAΔkimA) were analyzed. Kinetic assays demonstrated that the Kup⁺ strain exhibited higher potassium transport affinity than KimA⁺. Furthermore, differential responses to pH and osmotic stress highlighted distinct physiological contributions of KimA and Kup systems. KimA seems to be involved in K⁺ uptake at high concentrations (> 0.5 mM) while Kup played a dominant role at low potassium concentrations (> 0.2 mM) and under alkaline-osmotic stress. All these results delineate the functional divergence of the two transporters and our understanding of the ion homeostasis and stress adaptation mechanisms in E. faecalis.

Introduction

Several Gram-positive bacteria share the common characteristic of being opportunistic pathogens capable of colonizing diverse niches. Their remarkable ability to tolerate various stress conditions is a key factor in their success. These microorganisms have developed specific strategies that allow them to withstand adverse conditions such as high salinity, acidic environments, and temperature fluctuations. Additionally, they have adapted common mechanisms tailored to optimize their growth under different environmental stresses. By understanding the adaptive strategies that allow opportunistic pathogens to proliferate in challenging environments like food and hospital settings, we can implement more effective measures to control their presence¹.

Enterococcus faecalis is one of the most paradoxical groups within lactic acid bacteria (LAB). These microorganisms include natural commensals in the animal gut (including humans), bacteria of industrial significance used as probiotics, and multidrug-resistant hospital-acquired pathogens^2,3.

The ability of E. faecalis to adapt to fluctuating environments relies on the rapid movement of key ions, particularly protons, sodium, and potassium. Enterococcus exhibits remarkable adaptability due to the diversity of its H⁺ ATPase, Na⁺ ATPase, and K⁺ transport systems. These systems enable the bacteria to thrive under various environmental stresses^4,5,6.

Potassium is the predominant intracellular cation in all living cells and is critical for maintaining cellular turgor, membrane potential, and enzymatic activity. K⁺ homeostasis ensures osmotic balance, regulates cytoplasmic pH and supports processes such as protein synthesis and stress adaptation⁷. In the different environments in which E. faecalis can be isolated, potassium concentrations can vary from 0.1 to 10 mM, with an intracellular concentration reaching 400 mM bacterial cells can concentrate K⁺ more than 1000-fold through specific transport systems. Seminal studies on the biochemistry of ion homeostasis were carried out in the related strain E. hirae ATCC9790^4,5. They showed that at low pH the proton gradient is maintained by proton translocation through the active F₀F₁ ATPase. In addition, two Na⁺ ions and K⁺ are distributed at equilibrium concentration: K⁺ accumulates within the cell, while Na⁺ is exported by NapA. Under alkaline conditions, the F₀F₁ ATPase (atpIABCDFE operon) is inhibited. However, a V₀V₁-type sodium ATPase pumps Na⁺ out of the cell and K⁺ is accumulated by the Ktr system (KtrB, ntpJ gene), which is part of the ntpFIKECGABDJ operon in E. hirae^4,5.

Similar to E. faecalis, Staphylococcus aureus and Listeria monocytogenes are opportunistic pathogens. These bacteria possess various potassium transporters that aid them to accumulate K⁺ in the intracellular cytoplasm^{6,8,9,10,11,12,13}. This ability is essential for maintaining cellular functions under stressful conditions. In these microorganisms, transporters belonging to the SKT (Superfamily of K⁺ Transporters) superfamily, (including Ktr and Trk systems, and the Kdp complex), and the KUP family have been identified. These transporters play vital roles in osmoregulation, helping the cells adapt to high salinity, low water activity, and other challenging environmental conditions. However, the specific contribution of each transporter is not completely understood.

Recent studies have highlighted the critical role of cyclic di-adenosine monophosphate (c-di-AMP) in the fitness, antibiotic tolerance, and virulence of E. faecalis¹⁴. Mutations in the phosphodiesterases DhhP (dhhP) and GdpP (gdpP) of E. faecalis OG1RF, which degrade c-di-AMP, lead to its accumulation, enhanced colonization and multiplication within the host. Conversely, an E. faecalis strain lacking the cyclic diadenylate cyclase DAC enzyme (cdaA), resulting in c-di-AMP deficiency, exhibits nonpathogenic behavior in various animal models¹⁴.

Our previous studies have identified five distinct K⁺ transport systems in E. faecalis: KtrAB, KtrAD, the Kdp complex, KimA, and Kup. While the Ktr systems are ubiquitous in E. faecalis, the Kdp complex is less prevalent, and found in fewer than 40% of strains⁶. Notably, the Kdp system is predominantly associated with strains derived from clinical and gastrointestinal sources. The KdpDE two-component system regulates the expression of the kdpFABC operon, activating the Kdp system under low potassium conditions. Moreover, the transcriptional factor KdpD is a target of c-di-AMP, which negatively regulates its activity.

In our previous research, we identified two members of the KUP family in E. faecalis: Kup and KimA⁶. KUP proteins, part of the Amino acid-Polyamine-organoCation superfamily (APC superfamily TCDB 2.A.3), play diverse roles in bacteria and plants¹⁵. They have varying affinities for K⁺ and are essential under acidic conditions, functioning as K⁺/H⁺ symporters. The Escherichia coli Kup transporter was the first characterized member of this family^16,17. Additionally, in Firmicutes, the KupA and KupB transporters of Lactococcus lactis were found to be c-di-AMP targets¹⁸. KimA, initially characterized in B. subtilis, also binds c-di-AMP. Interestingly, initial studies found that c-di-AMP inhibits Kup in E. faecalis but not KimA, suggesting functional distinctions between these transporters⁶. These findings underscore the complexity of ion transport regulation in E. faecalis and support the hypothesis that distinct ion transporters perform specialized roles in maintaining cellular homeostasis and promoting survival under diverse circumstances, thereby contributing to the environmental adaptability of the organism. Further exploration of these transporters could provide valuable insights into how E. faecalis maintains its viability and flexibility under different stress conditions. More recently, we demonstrated that the IS6770 insertion sequence significantly impacts the expression of the kup gene, which encodes the Kup K⁺ uptake system in E. faecalis JH2-2¹⁹. This regulatory effect is not observed in the V583 strain, as IS6770 is not present in its genome.

Here, a transcriptional study of the kimA gene revealed its expression across diverse media and its regulation under different stress conditions. In M17G medium, kimA expression was significantly induced at low pH, while in LB and modified LB media (mLB), its expression increased under alkaline stress in the presence of NaCl.

Comparative growth kinetics of KimA⁺ (ΔktrAΔkup) and Kup⁺ (ΔktrAΔkimA) strains in potassium-limited semi-synthetic medium revealed divergent growth patterns. Under acidic or alkaline stress, the Kup⁺ strain displayed behavior similar to the wild type, whereas the KimA⁺ strain exhibited impaired growth, particularly under NaCl-induced stress.

Functionally, KimA serves as the primary potassium uptake system at elevated K⁺ concentrations, while Kup is more effective at lower K⁺ levels. Both systems are active at millimolar concentration of potassium, in contrast with the Ktr or Kdp complexes which are adapted to operate in the micromolar K⁺ range.

Results

Genomic context and mobile element characterization

Based on our previous study demonstrating the influence of IS6770 on kup gene expression in JH2-2 strain, we conducted a comparative analysis of the genomic region between the kimA and kup gene in both JH2-2 (IS6770-containing) and V583 (IS6770-free) strains¹⁹. In the V583 strain, the KimA gene is located approximately 13 kbp upstream of the kup gene. As illustrated in Fig. 1, kimA is predicted to be situated within a putative composite transposon in the strain JH2-2, flanked by two complete IS6770 sequences. This putative transposon is approximately 16.4 kbp in length and encodes seven membrane proteins including the MntE metal cation efflux pump. MntE is a divalent metal transporter that serves as an essential membrane protein protecting cells from toxic metal accumulation. This transporter, similar to FieF (YiiP) in E. coli²⁰, functions as an antiporter that exchanges toxic metal ions (such as Mn²⁺, Fe²⁺, Zn²⁺, and Cd²⁺) for protons, driven by the proton motive force. This mechanism contributes to the maintenance of intracellular metal homeostasis and provides protection against oxidative stress. Additionally, the KimA potassium transporter is involved in potassium homeostasis. The genes that encode the OpuC complex of osmolyte transporters (opuCA, opuCB, opuCC and opuCD) are also present. The OpuC ABC transporter found in other bacterial species is a highly efficient system for the import of a wide range of compatible solutes. Its broad substrate specificity, coupled with its ability to adapt to varying environmental conditions, ensures optimal solute uptake while minimizing energy expenditure^21,22. The caaT cation-transporting ATPase (P-type), is another gene in this region. Despite its similarity to SERCA1a (sarcoplasmic reticulum Ca²⁺-ATPase) and PMCA (plasma membrane Ca²⁺-ATPase), CaaT exhibits unique biochemical properties, including low Ca²⁺ affinity and high pH optimum. Moreover, LMCA1 (Listeria monocytogenes Ca²⁺-ATPase 1) functions as a Ca²⁺/H⁺ antiporter, potentially playing a role in Ca²⁺ homeostasis and stress response in L. monocytogenes²³. This genomic cluster highlights key players in ion homeostasis and osmolyte transport. This putative genetic element also contains a gene encoding a protein from the GnaT acyltransferase family (designated gnaT, Pfam00583), a gene encoding a protein from the glyoxalase family (designated glyF, Pfam00903) and a gene encoding an S-adenosylmethionine-tRNA ribosyltransferase-isomerase (designated queA, PfamPF02547).

Fig. 1

Schematic representation of the kimA and kup gene context in E. faecalis V583 and JH2-2 strains (a) The genes depicted by arrows encode the following proteins, the cation efflux protein (MntE, 389 residues), the KimA transporter (608 residues), the acyltransferase protein (GnaT, 154 residues), the glycine/betaine/carnitine/choline ABC transporter (OpuCD, 220; OpuCC, 307; OpuCB, 211; OpuCA, 397), the glyoxalase family protein (GlyF 296), the S-adenosylmethionine-RNA-t ribosyl transferase-isomerase (QueA, 343), the fragment of a putative transcriptional factor cro/CI (82 residues), the cation transporter (CaaT, 881 residues) and the transport Kup. The two copies of IS6770 (gray boxes containing an orange arrow) delimit the formation of a putative composite transposon in JH2-2 strain indicated by the dotted lines. (b) Locus tag of the genes described.

We performed an in silico analysis using BLASTn (NCBI) and the complete sequences of the described putative transposon and the IS6770 flanking regions adjacent to kimA and kup as probes. This analysis was performed against the genomes of Enterococcus and related lactic acid bacteria available in the NCBI genome database. However, no evidence of transposition events or rearrangements mediated by this putative mobile element was detected. This finding supports the hypothesis that this transposon represents a recent transposition event involving IS6770, particularly in strain JH2-2, in which we identified twelve intact copies of this insertion sequence¹⁹.

KimA promoter structure and transcriptional regulation

To define the DNA sequence required for kimA expression in E. faecalis, we analyzed the intergenic region upstream of kimA, starting from the 3’end of the mntE gene (Fig. 2A). This analysis encompassed approximately 350 nt between the mntE stop codon and the kimA start codon (Fig. 2A). Utilizing RNA folding prediction tools, we identified a putative rho-independent transcriptional terminator located 30 nt downstream of the mntE stop codon (Fig. 2B). This terminator structure consists of a stem-loop AAGCGAGAGccaAAAGGUCaaataGACUUUUaGCUCU CGCUU uuuuguuu with a predicted Gibbs free energy ΔGº = −22,9 kcal mol⁻¹, followed by an uracil-rich sequence (RNAfold²⁴, https://www.unafold.org/mfold/; ARNold program, http://rssf.i2bc.paris-saclay.fr/toolbox/). Furthermore, our analysis in this intergenic region aimed to identify the putative vegetative promoter regions driving for kimA gene expression; however, no evidence was found (https://services.healthtech.dtu.dk). In addition, we identified a putative rho-independent terminator with a ΔGº of −11.7 kcal mol⁻¹ located downstream of the kimA stop codon (Fig. 2B) (RNAfold, Zuker, ARNold program). To assess the promoter activity of the upstream region of kimA, we constructed a transcriptional fusion. A 574-bp fragment, located upstream of the kimA gene, was amplified and cloned into the fluorescent reporter vector pTLGR, resulting in the plasmid construct pKimA (Table 1 and Fig. 2B). Then, pKimA plasmid was introduced into E. faecalis JH2-2 via electroporation. This construct enabled visualization of transcriptional expression from the kimA gene through mCherry protein fluorescence (Fig. 2C). Strains were cultured in M17GA medium at an initial pH of 7.0 and incubated at 37 °C. The expression of the mCherry reporter was confirmed by fluorescence microscopy and quantified. Cells harboring the pKimA plasmid exhibited mCherry fluorescence with a normalized fluorescence value of 3.75 ± 0.57 RFU*/OD (Fig. 2D). In contrast, no fluorescence was observed in JH2-2 cells containing the empty pTLGR vector (negative control, Fig. 2C and 2D). Finally, we determined the precise location of the kimA transcriptional start site (TSS) using the Rapid Amplification of cDNA Ends (RACE) technique. Total RNA samples were extracted from E. faecalis JH2-2 cells grown in M17G medium (initial pH 7.0). Using the oligonucleotide Kim-1 Specific Primer (Table 2), we successfully identified the transcript initiation site, mapping the TSS to the adenine (A) nucleotide located 64 nt upstream from the initial ATG codon of the kimA gene. Based on this information, we predicted a consensus vegetative promoter, characterized by conserved −10 (TAgcAT) and −35 (TTGACt) boxes (Fig. 2A).

Fig. 2

Intergenic region of the MntE and KimA genes in E. faecalis JH2-2. (a) Nucleotide sequence of intergenic region, the transcription and translation directions for both genes are indicated by a horizontal arrow, gray for MntE and orange for KimA. The putative Shine-Dalgarno (SD) sites are highlighted by the corresponding nucleotides, and the putative −10 hexamers for kimA are framed in violet. The presence of a putative rho-independent termination region is marked in bold. The predicted stems and loops are underlined. In italics are shown the primers that amplify the 574 bp fragment (b) On the left, schematic representation of the predicted rho-independent type terminator for MntE. In the center pKimA plasmid, derived from pTLGR, is schematized, constructed from a 574 bp fragment upstream of kimA in E. faecalis JH2-2. The black arrow indicates that the transcriptional fusion is directed towards the expression of the mCherry fluorescent protein. On the right, a schematic representation of the predicted rho-independent terminator for KimA. (c) Optical microscopy images of visible light with red fluorescence microscope (mCherry) of E. faecalis JH2-2 cells containing the pKimA vector are shown. Increase 100x. Scale bar 10 μm. (d) Activity of the transcriptional fusion of pKimA at an initial pH of 7 in M17 GA. Specific fluorescence was calculated as the ratio between the detected fluorescence, normalized to bacterial biomass estimated from the OD₆₀₀ of the exponential growth culture. The average values of at least three independent experiments are presented, with error bars representing the standard deviations.

Table 1 Bacterial strains used in this study.

Table 2 Plasmid used in this study.

pH-dependent expression of kimA

To examine the transcriptional activity of the P_kimA promoter under acidic or alkaline growth conditions, we measured the expression levels of the fluorescent protein mCherry in E. faecalis JH2-2 cells harboring the pKimA plasmid or the empty vector pTLGR (Table 1). Cells were grown to exponential phase in M17G at an initial pH of 5.0 or M17GA at an initial pH of 9.0. As shown in Fig. 3A, cells containing the pKimA plasmid and grown in acidic medium (pH 5.0) exhibited significantly higher specific fluorescence (32.89 ± 2.75 RFU*/OD) compared to those grown in an alkaline medium (pH 9.0) (6.74 ± 1.15 RFU*/OD). This trend was also observed when comparing the specific fluorescence values obtained for growth in acidic and neutral conditions (compare Fig. 2D and 3A).

Fig. 3

Transcriptional analysis of kimA by mCherry reporter plasmid and RT-qPCR at different pH values. (a) Bacterial cultures were grown up to exponential phase in M17G at an initial of pH 5.0 and M17 GA to an initial of pH 9.0. Transcriptional fusion activity was determined by quantifying mCherry fluorescence for E. faecalis JH2-2 cells transformed with pTLGR and pKimA. Specific fluorescence was calculated as the ratio between normalized detected fluorescence and estimated bacterial biomass from the OD₆₀₀ of the exponential cultures. The mean values of at least three independent experiments are presented; the error bars represent the standard deviations. The asterisks indicate statistically significant differences determined by the Tukey Multiple Comparison Test (**** p < 0.0001). (b) Relative expression of the kimA gene in the E. faecalis JH2-2 strain under different pH conditions. E. faecalis JH2-2 grew to an exponential phase in M17G at pH 5.0 or M17GA to pH 9.0. Average values, normalized with respect to the gyrA gene, of at least three independent experiments are represented. Error bars represent standard deviations.

To corroborate these findings, we quantified kimA transcripts levels using RT-qPCR. Total RNA was extracted from exponential cultures of E. faecalis JH2-2 grown in M17G at an initial pH of 5.0 and in M17GA at initial pH 9.0. Specific oligonucleotides designed for the kimA gene using Primer3 software (Table  3) were employed for the qPCR reaction. The relative expression of kimA was calculated using a standard curve of genomic DNA and the normalizing gene gyrA, and plotted as a function of culture medium pH. As shown in Fig. 3B, when E. faecalis JH2-2 is grown in an acidic medium (pH 5.0), the expression of the kimA gene is induced, with the relative expression value being four times higher in acidic than at alkaline pH (0.21 ± 0.05 and 0.065 ± 0.002, respectively). All these results suggest that kimA gene expression in the M17G complex medium is induced at acidic pH compared to neutral or alkaline pH.

Response to osmotic stress: NaCl-induced expression of kup and kimA

To investigate the expression of kup family genes in strain JH2-2, focusing on the impact of osmotic stress (3% NaCl), we compared the expression of kimA and kup genes under conditions established in previously performed studies⁶. We utilized the plasmid pKimA and plasmid pKupL, which carry a 1769 nt fragment encompassing the P_kup promoter region and the full-length sequence of IS6770 from strain JH2-2 (Table 1). Gene expression was evaluated under alkaline conditions (pH 9.0) in LBG medium and its potassium-reduced variant, mLBG, with and without the addition of 3% NaCl⁶. Previous studies have indicated a growth phenotype in strains under these conditions⁶. Our results showed that when the cultures were grown in LBG medium pH 9.0, cells harboring the pKimA (5.32 ± 0.74 RFU*/OD) or pKupL (5.39 ± 1.17 RFU*/OD) plasmids, exhibited an increase in the mCherry fluorescence when 3% NaCl was added to the medium (41.2 ± 6.54 RFU*/OD and 39.84 ± 6.05 RFU*/OD, respectively).

To validate these observations, we extracted total RNA from E. faecalis JH2-2 cells grown in the same conditions LBG and LBG 3% NaCl at pH 9.0. The extracted RNA was used to perform RT-qPCR. Specific oligonucleotides for each gene were designed using Primer3 software (Table 3). The results showed a weak induction of both gene kup (from 1.08 ± 0.03 to 2.46 ± 0.50) and kimA (from 1.4 ± 0.09 to 2.59 ± 0.64) in response to the addition of 3% NaCl (Fig. 4B).

Fig. 4

Change in the relative expression of kimA and kup in E. faecalis JH2-2 induced by the presence of NaCl. (a and c) Bacterial cultures were grown to the exponential phase at an initial pH of 9.0 in LBG, LBG 3% NaCl, mLBG or mLBG 3% NaCl. Transcriptional fusion activity was assessed by quantifying mCherry fluorescence in E. faecalis JH2-2 cells transformed with pKimA. Specific fluorescence was calculated as the ratio between the normalized fluorescence detected and the estimated bacterial biomass, determined from the OD₆₀₀ of the exponential culture. The mean values from at least three independent experiments are shown, with error bars representing the standard deviations. Asterisks indicate statistically significant differences determined by the Tukey Multiple Comparison Test (**p < 0.01; ***p < 0.001; **** p < 0.0001). (b and d) Change in the relative expression of kimA and kup in E. faecalis JH2-2 induced by the decrease in K⁺ in the medium and the presence of NaCl. The average values, normalized to the gyrA gene, from at least three independent experiments are shown. The error bars represent the standard deviations. Asterisks indicate statistically significant differences determined by the Tukey Multiple Comparison Test (*p < 0.05).

Similarly, we analyzed the expression of the kimA and kup gene in the mLBG at pH 9.0 with and without the addition of 3% NaCl. As shown in Fig. 4, the addition of NaCl induces the expression of kimA and kup genes. Specifically, pKimA increased mCherry fluorescence from 4.25 ± 0.55 RFU*/OD to 16.83 ± 1.55 RFU*/OD, while pKup levels increased from 2.88 ± 0.47 RFU*/OD to 7.96 ± 0.89 RFU*/OD. Validation of these observations performed by RT-qPCR showed a slight increase in response to the addition of 3% NaCl. While LBG and mLBG media facilitated visualization of gene induction of kimA and kup in response to NaCl under alkaline conditions, they were less effective in revealing the impact of potassium concentration on gene expression. Upon measuring the residual potassium levels in mLBG, we found that it was higher than those reported by Kawano et al., 2001²⁸. Specifically, the potassium concentration in LB was 13.24 ± 0.70 mM, whereas mLB contained 6.54 ± 0.92 mM. Then, to precisely control potassium concentration, we opted for a semi-synthetic medium.

Functional characterization of Kup and KimA in potassium uptake

Additionally, to assess the differential responses of the potassium transporters KimA and Kup to variations in external K⁺ concentrations, E. faecalis JH2-2 mutant strains expressing only one of these transporters (Fig. 5A) were cultured in the presence of different KCl concentrations (Fig. 5B). We used Spizizen salts (MSSM) in which potassium was replaced by sodium salts; additionally, we supplemented it with a mixture of amino acids (see bacterial strains and cultures). The residual potassium level in this medium was 0.15 ± 0.06 mM. Then, the dependence of the growth rate on the external KCl concentration was analyzed using the Monod Eq ²⁹. This model reveals the half-saturation constant (K_s) and the maximum uptake rate (µ_max) for each evaluated strain. By growing E. faecalis Kup⁺ (ΔktrAΔkimA) and KimA⁺ (ΔktrAΔkup) strains in the presence of various KCl concentrations, we found that each strain required a distinct external K⁺ concentration to reach half-maximal growth rate: the Kup⁺ strain required 0.55 ± 0.08 mM, additionally the KimA⁺ strain required 2.18 ± 0.19 mM (Table 4). This increase in K_s indicates that the KimA⁺ strain requires higher amounts of external potassium to achieve maximum growth compared to the Kup⁺ strain under the same culture conditions. When comparing the µ_max values, it was observed that none of the double mutant strains evaluated reached the maximum growth rate of the wild-type strain (Table 4), indicating a marginally reduced uptake capacity. Specifically, the KimA⁺ strain displayed the slowest uptake rate among the strains (0.0027 min⁻¹). These differences highlight the varying efficiencies of their potassium transport systems. The wild type strain exhibits the highest affinity and uptake rate for potassium, which could confer a competitive advantage in environments with limited potassium availability. The potassium uptake kinetics of the KimA⁺ (ΔktrAΔkup) and Kup⁺ (ΔktrAΔkimA) mutants strains reveal significant differences in their efficiency and affinity for K⁺. The stark contrast observed between the KimA⁺ and Kup⁺ mutants underscores the essential role of these transport systems in maintaining potassium homeostasis. The Kup⁺ strain, with its intermediate K_s and relatively high µ_max, is better equipped to handle potassium fluctuations than the KimA⁺ strain. In contrast, the low K⁺ affinity and reduced uptake rate of the KimA⁺ strain may hinder its ability to succeed in potassium-limited environments, potentially affecting its survival and growth compared to the Kup⁺ strain and the wild-type strain.

Fig. 5

Potassium-Dependent Growth Profiles of E. faecalis Wild-Type and Mutant Strains (a) Schematic representation of the potassium transporters present in the wild-type strain and the different mutants (b) Potassium-dependent growth of different E. faecalis JH2-2 mutants was assessed. Growth curves were generated in a semi-defined medium deficient in K⁺ (MSSM supplemented with amino acids, vitamins, and CAA 0.1%), with different concentrations of KCl (0–10 mM). For each condition, μ_max was calculated and plotted against the KCl concentration in the medium. The wild-type E. faecalis JH2-2 strain (circle) was used as a control, Kup⁺ (triangle), KimA⁺ (square).

Table 3 Oligonucleotides used in this study.

Stress tolerance in Kup⁺ and KimA⁺ mutants

The influence of K⁺ concentration on the growth of strains containing only Kup or KimA transport under different stress conditions was further evaluated (Fig. 6). At an initial pH of 7.0 and 10 mM K⁺, the wild-type strain exhibited its maximum specific growth rate (μ_max). Under these optimal conditions, the Kup⁺ strain achieved a 90% of the μ_max determined for the wild-type strain, while KimA⁺ strain showed a significant reduction in μ_max (40%) (Fig. 6B). This suggests that a functional kup gene is crucial for optimal growth, even under mild stress. When the pH was lowered to 5.7, μ_max values decreased for all strains, indicating the negative impact of acidic conditions on growth. Notably, the KimA⁺ strain was particularly affected, showing inhibited growth even at a low K⁺ concentration of 0.5 mM (Fig. 6A). This highlights the importance of the kup gene product in coping with combined stress from low pH and limited potassium. Interestingly, the behavior of the strains diverged significantly at high pH levels. The wild-type strain maintained relatively stable μ_max values, suggesting a higher tolerance to alkaline stress. In contrast, the Kup⁺ and KimA⁺ strains displayed different levels of adaptation, with the Kup⁺ mutant showing a greater ability to grow (Fig. 6C). This underscores the intricate interplay between pH, potassium availability, and the specific role of kup and kimA genes in stress response mechanisms. To further investigate the interplay between potassium transport and stress tolerance, we examined the impact of NaCl supplementation on the growth of the wild-type, Kup⁺ and KimA⁺ strains under acidic, neutral, and alkaline conditions (Fig. 6 D, E and F).

Fig. 6

Growth Kinetics of E. faecalis Wild-Type and Mutant Strains in Response to pH and Osmotic Stress Growth rate of the double mutant strains Kup⁺ (light-blue triangle), KimA⁺ (orange square), and JH2-2 (light-green circle) at different initial pH values, 5.7, 7.0 and 8.0, respectively (a, b and c). Growth curves were performed in a semi-defined medium deficient in K⁺ (MSSM supplemented with amino acids, vitamins, and CAA 0.1%) and the addition of different concentrations of KCl (0.2, 0.5, 10, and 100 mM) was evaluated. For each condition, μ_max was calculated and plotted as a function of the KCl concentration added to the medium. The wild-type E. faecalis JH2-2 strain was used as a control. Growth curves of the double mutant strains Kup⁺, KimA⁺, and JH2-2 were performed in the MSSM medium at initial pH values of 5.7, 7.0, and 8.0, (d, e and f, respectively). The arrow in a, b and c graph indicates the condition under which 1% NaCl was added to perform the growth curves in D, E, and F.

In MSSM 10 mM KCl, the presence of 1% NaCl at an acidic pH of 5.7, none of the strains grew (Fig. 6E). This suggests that the combined stress of low pH and high salinity may be detrimental for all strains, regardless of their potassium transport capabilities. At neutral pH, in MSSM 10 mM KCl with 1% NaCl, the wild-type strain exhibited growth. However, both the Kup⁺ and KimA⁺ strains displayed a reduced capacity to reach their maximum specific growth rate (μ_max) (Fig. 6E). This implies that compromised potassium transport systems might hinder their ability to adapt to even moderate salt stress under neutral conditions. Interestingly, when the pH was adjusted to an alkaline value of 8.0 in the presence of 1% NaCl, the Kup⁺ strain displayed growth parameters comparable to the wild-type strain (Fig. 6F). This observation reinforces the distinct behavior of the Kup⁺ strain, suggesting a potential role for the kup gene product in facilitating adaptation to salt stress under alkaline conditions. Conversely, the KimA⁺ strain did not exhibit the same level of adaptation, highlighting the differential responses of these strains to salt stress depending on the environmental pH. These findings provide further insights into the complex relationship between potassium transport, stress tolerance, and environmental conditions. They suggest that in the Kup⁺ and KimA⁺ strains, the mutations not only affect potassium acquisition but also impact the ability to adapt to specific stress combinations.

Distribution of Kup and KimA in Firmicutes

To gain a broader understanding of the significance of the kimA-kup locus, we analyzed the presence of its members in complete genome sequences of both pathogenic and non-pathogenic microorganisms within the phylum Firmicutes. The selected microorganisms include species and strains relevant to the food industry, recognized for their probiotic properties or contributions to food sensory characteristics, as well as those of clinical origin in humans and animals, and strains of industrial interest.

As shown in Table 5, both Kup and KimA transporters are widespread among most examined microorganisms regardless of their origin (amino acid identity with the queries > 52% and 40%, respectively), highlighting their importance in the physiology of this group. Notably, Lactobacillus species typically possess Kup while lacking KimA, a pattern also observed in Vagococcus. Conversely, Streptococcus species demonstrate the absence of one or both transporters, suggesting reliance on alternative potassium uptake systems, such as the Ktr or Kdp complexes.

Table 4 K_s and µ_max values for the different double mutant strains derived from E. faecalis JH2-2, in a semi-defined medium with regulated K⁺ concentrations.

Regarding other protein-coding genes involved in ion/osmolite transport, OpuCC is widely distributed (Id% > 30), with notable absences in L. helveticus and O. oeni. Similarly, the CaaT transporter (Id% > 39) is common among the selected microorganisms but is absent in Staphylococcus species. Finally, the MntE protein (Id% > 30) is rarely found in the examined microorganisms but appears in related genera such as Enterococcus, Vagococcus, Carnobacterium, and Clostridium.

The distribution patterns of Kup, KimA, and associated ion transporters indicate both conserved and specialized functions in potassium and osmolyte homeostasis among Firmicutes. The differential presence of these systems among taxa reflects ecological adaptation and highlights potential targets for functional and evolutionary studies.

Discussion

E. faecalis can use a variety of transport systems to take up potassium under fluctuating environmental conditions. Potassium is known to contribute to the response to various stress conditions by modifying the proton motive force—both the pH gradient and the membrane potential—or by serving as the first line of defense in the osmotic stress response⁷. In these microorganisms, complex and sophisticated pumps (KdpFABC) controlled by two-component regulatory systems (KdpED) or membrane channels such as KtrB and KtrD, which share their cytoplasmic component, KtrA, are part of its arsenal of systems⁶. In this study, we investigated the transcriptional and biochemical regulation of potassium uptake systems in E. faecalis, focusing on the kimA and kup genes. Results are summarized in Fig. 7.

Fig. 7

Overview of the main ions/osmolytes transporters present in the locus kimA- kup in Enterococcus faecalis responding to diverse stress conditions. (a) Expression of the kimA and kup K⁺ transporters in E. faecalis is induced under stress conditions. In strain JH2-2, the IS6770 insertion sequence rompe la activation at low pH. KimA mediates K⁺ uptake at elevated K⁺ concentrations and exhibits, whereas Kup functions at lower K⁺ and exhibits higher-affinity c-di-AMP binding. (b) Under acidic stress, K⁺ uptake is coupled to proton extrusion via the F₀F₁-ATPase (encoded by the atp operon) and to Na⁺ extrusion by the NapA antiporter. This coupling helps maintain intracellular pH and ionic homeostasis during acid stress. (c) Under alkaline stress, K⁺ uptake is balanced by Na⁺ extrusion via a V-type ATPase (also encoded by the ntp operon), and the potentially assisted by a Ca²⁺-ATPase (caaT). (d) In alkaline media with elevated NaCl, the OpuC osmolyte uptake system also contributes to ionic balance and osmoprotection.

We found that these genes co-localise with genes potentially involved in the ion homeostasis response, including a divalent metal transporter (MntE) and a P-type ATPase (CaaT). MntE may be involved in the homeostasis of manganese, this gene is upregulated in E. faecalis by the addition of Mn²⁺ to the growth medium³⁰. CaaT from L. monocytogenes can remove Ca²⁺ from the cytosol in response to alkaline and salt stress²³, this gene is also reported as induced in transcriptomic studies performed in E. faecalis grown in the presence of NaCl (6,5%)³¹. Additionally, we identify a system homologous to OpuC, an ATP-dependent transporter belonging to the ATP-binding cassette transporter superfamily and involved in the glycine-betaine transport characterized within those induced in the osmotic stress response (Fig. 1 and 7)³¹. As shown in Table 5, CaaT and OpuC are distributed across different Firmicutes species, suggesting important functions in ion homeostasis and stress response. This integrated genomic arrangement highlights the sophisticated regulatory network that allows E. faecalis to thrive in diverse environments; characterization of these systems will improve our understanding of ion transport systems in bacteria (Fig. 7).

Our findings indicate that the kimA gene in E. faecalis strain JH2-2 is transcribed as a monocistronic unit. This transcription is likely controlled by a sigma-A type vegetative sigma factor, with proposed consensus promoter sequences at the −10 (TAgcAT) and −35 (TTGACt) regions. Expression of the kimA gene was up-regulated under conditions of low pH, alkaline pH, and salt stress. Bioinformatic analyses were performed to identify putative binding sites for alternative sigma factors, distinct from the canonical σ⁷⁰. This search did not reveal binding sites for known E. faecalis alternative sigma factors, including SigV (TGAAACN₁₆CGTC³²), SigN (TTGGCACN₅TTGCT³³), or SigH (RNAGGA(W₃)N_11–12R³⁴).

Previous research demonstrated that in strain JH2-2, the insertion sequence IS6770 impairs the low pH-dependent activation of the kup gene¹⁹. However, the current study shows that kimA induction is observed under a broader range of stress conditions, including alkaline pH and high salinity, in addition to acidic pH. This suggests the involvement of at least two distinct molecular regulatory mechanisms: one mediating expression specifically at acidic pH, and another responsible for induction under alkaline and NaCl stress conditions. Further investigations are warranted to elucidate the precise molecular mechanisms governing stress-induced kimA and kup expression in enterococcal cells.

The role of c-di-AMP is particularly critical, as it regulates ion homeostasis and stress adaptation by modulating the activity of potassium transport systems. In E. faecalis, c-di-AMP plays a pivotal role in potassium regulation, although functional redundancy among potassium transporters like kimA, kup, ktr, and kdp ensures robust adaptation under various stress conditions. Our experiments showed that the Kup⁺ strain displayed a higher affinity for potassium ions compared to the KimA⁺ strain, highlighting distinct functional roles. Kup appeared to be the primary transporter under stress conditions, showing wild-type-like behavior and significant regulation by c-di-AMP, which may mitigate potential ion toxicity. In contrast, KimA showed limited functionality under extreme conditions, suggesting that it serves as a supplementary system during non-stressful growth (Fig. 7).

In E. faecalis JH2-2, KimA operates under high extracellular potassium concentrations (> 0.5 mM) and is transcriptionally upregulated by three key stimuli: acidic pH, alkaline pH, and NaCl exposure. In contrast, the Kup system is similarly activated under these conditions but it shows a modified response to acidic pH due to the regulatory influence of the IS6770 insertion sequence, as previously reported¹⁹. Notably, Kup demonstrates optimal functionality at moderate potassium levels (0.2 mM), a concentration range frequently observed in animal-derived tissues and bodily fluids (e.g., blood and urine). This physiological relevance may explain the lack of discernible phenotypic differences in single or double mutants under in vivo conditions such as urine or blood, as reported by Acciarri et al.¹⁹.

Figure 7B-D illustrate different scenarios where kup and kimA function complementarily with other ion transporters. Under acidic environments (Fig. 7B): proton gradients are maintained through the coordinated activity of the F₀F₁ ATPase and the Na⁺/H⁺ antiporter NapA, which stabilize intracellular pH under acidic stress. Under alkaline environments (Fig. 7C): potassium uptake via KimA/Kup is accompanied by sodium extrusion mediated by the NtpAse and a putative Listeria-like calcium ATPase, suggesting a multifunctional ion-homeostasis strategy. Finally, under osmotic stress (Fig. 7D): the OpuC system, a glycine-betaine transporter, is implicated in osmoprotection under alkaline conditions, highlighting its potential role in mitigating osmotic stress.

In conclusion, the transcriptional regulation and functional characterization of kimA and kup provide new insights into potassium homeostasis in E. faecalis. These findings emphasize the adaptability of microorganisms and its dual role as a beneficial and opportunistic microbe. Further studies are required to clarify the molecular interactions that govern c-di-AMP signaling and its integration with other stress-response pathways. Gaining a deeper understanding of these mechanisms could lead to strategies to mitigate the pathogenic potential of E. faecalis while harnessing its beneficial properties.

Materials and methods

Bacterial strains and cultures

The bacterial strains used in this study are shown in Table 1. Briefly, Escherichia coli strain DH5-α, used as an intermediate host for cloning, was grown at 37 °C under aerobic conditions with vigorous shaking in Lysogeny Broth medium (LB). When appropriate, growth media were supplemented with erythromycin with final concentration of 250 μg/ml.

Table 5 Presence (+) and absence (-) of the kup and kimA genes in different strains selected from the firmicutes.

Enterococcus faecalis strains were routinely grown at 37 °C without shaking in M17 medium (Oxoid) containing 0.5% w/v glucose (M17G). For fluorescence experiments and the determination of transcriptional profiles, M17 were supplemented with either 0.5% (w/v) glucose (M17G) or 0.1% w/v glucose and 0.1% (w/v) arginine (M17GA) according to experimental needs. For the determination of transcriptional profiles in a low K⁺ medium, mLBG, a modified LBG medium in which yeast extract was reduced from 0.5 to 0.025%²⁸, was used. For growth experiments at regulated KCl concentrations, modified sodium Spizizen medium (MSSM)¹² supplemented with arginine 0.72 mM, glutamic acid 2.7 mM, isoleucine 1.6 mM, leucine 3.6 mM, tryptophan 0.05 mg/ml, valine 2.8 mM, 2.3 mM glycine, 0.84 mM histidine and 1 mM methionine, 2.5 mg/l biotin, 1 mg/l riboflavin, 1 mg/l pantothenic acid, 1 mg/l nicotinic acid, 1 mg/l, 5 mg/l pyridoxal hydrochloride, 0.05 mg/l folic acid, 0.1% w/v casamino acids was added to allow the growth of the JH2-2 strain. KCl and/or NaCl, were added to the growth media at the concentrations indicated in the figure legends. The pH was adjusted by adding HCl or NaOH, for the mLB and M17 media, while for the MSSM medium, by varying the amount of the salts in the phosphate buffer. When appropriate, cultures were supplemented with erythromycin with final concentration of 5 μg/ml.

DNA manipulation and transformation

For the genomic DNA isolation of E. faecalis, Microbial DNA Kit (Macherey–Nagel) was used. For the plasmid DNA from E. coli cells and DNA fragments purification, Wizard Plus Minipreps DNA purification system (Promega) and Wizard SV Gel and PCR Clean-Up System (Promega) were used, respectively. Treatment of DNA with restriction enzymes, T4 DNA ligase, and DNA polymerases was performed as recommended by the suppliers. DNA sequences were checked by sequencing in the World Meridian Venture Center, Macrogen sequencing service (Seoul, Korea). Transformation of E. coli cells was carried out using standard procedures³⁵. Electroporation of E. faecalis was performed according to the procedure of Dornan and Collins³⁶.

Construction of pTLGR derivatives

The pTLGR promoter probe vector³⁷ was used for generation of the recombinant plasmid listed in Table 2. The up-stream regions of the kimA gene from E. faecalis JH2-2 were amplified by PCR using the oligonucleotide pairs listed in Table 3. The fragments were digested with BamHI and ligated to the pTLGR vector previously digested with the same restriction enzyme. Fragment orientation was determined by PCR, and confirmed by sequencing.

Detection and quantification of mCherry in E. faecalis strains carrying pTLGR derivatives

Two different experiments were conducted: one using M17G or M17GA at different pH levels (5.0, 7.0, and 9.0), and the other using LBG or mLBG, both at pH 9.0. Overnight incubation media grown were diluted in 20 ml of fresh medium, with an initial pH corresponding to each case, to achieve an optical density at a wavelength of 600 nm (OD₆₀₀) of 0.1 and were grown until the middle of the exponential phase. The mCherry expression was measured as follows: cells from aliquots of the exponential cultures were sedimented by centrifugation (5000 × g for 10 min, at 4 °C), washed twice, and resuspended to the same original volume in PBS buffer (10 mM Na₂HPO₄, 1 mM KH₂PO₄, 140 mM NaCl, 3 mM KCl). mCherry fluorescence was measured at an excitation wavelength of 585 nm and an emission wavelength of 620 nm, in a Synergy HT equipment (BioTek, Vermont, USA). Background fluorescence of the control strain (harboring the pTLGR plasmid) was used to normalize the fluorescence signal. In addition, culture biomass was estimated by measuring the OD₆₀₀ and the pH of the medium at the time of aliquoting was recorded with a pre-calibrated pH meter (HI 9017 microprocessor pHmeter, HANNA instruments). Three independent trials were performed. The same suspensions, fixed in 1% agarose, were used for visible and fluorescent microscopy with a Nikon Eclipse E600 model microscope. Images were acquired with a Andor Clara model camera using the Nis-Elements software³⁸. Fiji software was used to superimpose and pseudo color images³⁸.

Isolation of RNA and transcript analysis

Total RNA was isolated from E. faecalis cells using the method described previously³⁹. Overnight cultures were grown in M17G pH 5.0, M17GA pH 9.0, LBG pH 9.0, LBG + 3% NaCl pH 9.0, mLBG pH 9.0 and mLBG + 3% NaCl pH 9.0, harvested at mid-exponential phase by centrifugation. The integrity of RNA was checked by agarose gel electrophoresis. Total RNA concentration was determined by UV spectrophotometry with a Bio-Tek Powerwave XS plate reader equipped with a Take3 plate adapter. Before the reverse transcription, RNA (8 μg of total RNA) were treated with 3 U of DNase (Promega) as described by the manufacturer. Then, 2 µg of total RNA was used to perform a Retro-Transcription (RT) step using Random Primers N6 (PB-L) and 200 U of Moloney murine leukemia virus reverse transcriptase (Promega) as described by the manufacturer. Quantitative real-time PCR (RT-qPCR) of cDNA was performed with the HOT FIREPol Eva Green qPCR Mix Plus (ROX) (Biocientífica, Solis BioDyne). Primers for real-time PCR were designed to have a length of about 19–23 bases and a Tm of about 60 °C. The length of the PCR products ranged between 99 and 100 bp (Table 3). Primer design software (Primer3) was used to select primer sequences. Relative transcript abundances were determined by a quantification method using a standard genomic DNA curve and the housekeeping gyrA gene as a reference.

Rapid Amplification of cDNA ends by polymerase chain reaction (5ʹ RACE PCR)

The identification of the kimA transcriptional start site was achieved using the method described by Acciarri et al.¹⁹. cDNA strand was synthesized from E. faecalis JH2-2 RNA using Kim-RT (Table 3) and M-MLV reverse transcriptase according to the manufacturer (Promega). After incubation at 70 °C to terminate the reaction, the cDNA was incubated at 37 °C with RNase H (Biolabs), then column purified (Sartorius Vivacom 500). A poly-A tail was added to the 3’ end of the cDNA using Terminal Deoxynucleotidyl Transferase (Promega) and then amplified using Kim-1 Specific Primer, QT_RACE and the Qo_RACE (Table 3). The nested product was isolated by agarose gel electrophoresis and extracted using a Wizard SV Gel and PCR clean-up System (Promega). The product was cloned into pGEM-T Easy Vector Systems (Promega) and the plasmids from two positive transformants were sequenced.

Potassium uptake and growth kinetics in E. faecalis JH2-2 mutants

Saturated cultures grown in M17G were used to inoculate fresh modified MSSM medium adjusted to the indicated pH and supplemented with amino acids, vitamins, 0.4% w/v casamino acids and 10 mM KCl to an initial OD₆₀₀ of 0.1. Once the cell culture reached the exponential growth phase, the cells were harvested and washed three times with fresh MSSM medium at the respective pH, without any supplementation. The cells were diluted to an initial OD600 of 0.1, in 200 µl final volume of MSSM medium adjusted to the indicated pH and supplemented with the amino acid mixture, the vitamin mixture, 0.1% w/v casamino acids and KCl at the indicated concentrations. Culture growth was monitored by measuring OD₆₀₀ at 20 min intervals for 16 h on a PowerWave™ XS microplate reader (BioTek, BioTek Instrument Inc., Vermont, USA). Maximum growth rates (μ_max) were calculated and plotted as a function of the KCl concentrations added to the medium. This allowed fitting to the Monod equation and calculation of maximum specific growth rate (µ_max) and the apparent K_s using the solver tool of Excel 2012 (Microsoft).

$$\mu =\frac{{\mu }_{max}\left[S\right]}{Ks+\left[S\right]}$$

μ is the growth rate of a considered microorganism, μmax is the maximum growth rate of this microorganism, [S] is the concentration of the limiting substrate S for growth, Ks is the"half-velocity constant"—the value of [S] when μ/μmax = 0.5. μmax and Ks are empirical (experimental) coefficients to the Monod Eq²⁹.

Experiments were repeated with three biological replicates. To analyze the effect of different stress conditions on the growth of the mutant strains, an inhibition index was calculated as the ratio between the μ_max under each stress condition evaluated and the maximum growth rate of the wild-type strain grown in minimal medium at pH 7.0 and supplemented with 10 mM KCl.

Data analysis

The data obtained from at least three independent experiments were subjected to one-way analysis of variance (ANOVA) by using R software version 4.3 Tukey´s test was selected to determine the significant differences between the variables at p ≤ 0.05.

Bioinformatic analysis

Selected Firmicutes whole-genomes and predicted protein sequences were downloaded from the NCBI nr database. Protein sequences KimA (WP_033601095.1), KupA (WP_002381294.1), OpuC (AAO80675.1), CaaT (pdb|6ZHH|A) and MntE (WP_002386999.1) were used as queries in local BLASTP searches using an E value of 10^–25 and > 75% of sequence coverage as cut-off values⁴⁰.