IonoBiology: The functional dynamics of the intracellular metallome, with lessons from bacteria

Abstract

Metal ions are essential for life and represent the second most abundant constituent (after water) of any living cell. While the biological importance of inorganic ions has been appreciated for over a century, we are far from a comprehensive understanding of the functional roles that ions play in cells and organisms. In particular, recent advances are challenging the traditional view that cells maintain constant levels of ion concentrations (ion homeostasis). In fact, the ionic composition (metallome) of cells appears to be purposefully dynamic. The scientific journey that started over 60 years ago with the seminal work by Hodgkin and Huxley on action potentials in neurons is far from reaching its end. New evidence is uncovering how changes in ionic composition regulate unexpected cellular functions and physiology, especially in bacteria, thereby hinting at the evolutionary origins of the dynamic metallome. It is an exciting time for this field of biology, which we discuss and refer to here as IonoBiology.

The vast majority of molecular interactions that occur within any living cell have remained obscure (Ross, 2016). How can this be? Most molecular interactions that have been studied to date, and on which our current understanding of biology is primarily based on, are covalent interactions. Covalent bonds result from sharing of electrons and are thus relatively strong interactions that persist over time, such as those that bind amino acids together to make proteins, or the phosphodiester bonds that help form the DNA’s double helix. Their strength and persistence make covalent bonds highly suitable for experimental measurements. However, intermolecular interactions in cells are non-covalent, including ionic and van der Waals interactions, dipole intermolecular forces, and hydrogen bonds. The sheer number of non-covalent interactions at the atomic and molecular scale thus hint at a complexity and novel regulatory mechanisms that have remained obscured. To deeply understand cellular regulation, we must therefore overcome experimental challenges and characterize the weak and transient non-covalent interactions that pervade the interior of cells (Figure 1A).

Figure 1.

The metallome composition and its dynamics in bacterial cells

A Cartoon of a bacteria, illustrating the crowded interior of cells. Cells regulate the concentration of ions (colored circle) creating a difference in ionic concentration between the interior and the exterior of the cell. This difference in ionic concentration generates an electrochemical potential (V) across the membrane. Ion channels can passively conduct specific ions driven by the electrochemical potential. In contrast, ion transporters can conduct ions even against electrochemical gradients through active transport by consuming ATP and/or co-transporting with protons or other ions.

B Pie chart representing the percentage of molecules per cell in bacteria, with ions the most abundant component. Molecules of water are not considered in this representation, which corresponds to 70% of the total cellular content. The values come from measurements in Escherichia coli and were obtained from (Milo and Phillips, 2015). The specific values are: ions 71%, lipids 24%, proteins 4%, lipopolysaccharides 1%, other components such as DNA represent less than 1%.

C Average distribution of metal ions in bacterial cells, with potassium and magnesium the two most abundant metal ions in any living cell. The values used to generate the pie chart are for Escherichia coli and were obtained from (Milo and Phillips, 2015). The specific percentages are: potassium 83%, magnesium 12%, sodium 3%, calcium 1%, other ions such as zinc and iron represent less than 1%.

D Cartoons representing the dynamics of the metallome in a given cell (top) and how this affects its membrane potential (bottom) over time. The membrane potential can for example become hyperpolarized (more negative) by a decrease in the concentration of a specific ion. Conversely, the membrane potential can also become depolarized by the increase of the metallome content.

Electric diagram of the cell membrane. The membrane potential of a cell can be considered as a capacitor (C_m), and each ion channel can be considered as a resistor (g_i), with the Nernst potential of the given ion (V_i). This example shows the case of a cation (X) accumulated in the interior of the cell. The leak term refers to other ions and charged molecules contained in the cell that also contributes to the membrane potential.

Cations are Essential for Life

Why are metal cations so abundant in all cells? The most fundamental biological macromolecules in cells, such as DNA, RNA, and proteins, all carry numerous charges on their surface. The majority of these charges are negative, primarily due to the many phosphate atoms in DNA and RNA molecules. Consequently, the structural stability and function of these macromolecules require counterions, in particular positively charged metal ions, to provide charge balance (Box 1, see Matsarskaia et al., 2020 for a comprehensive review). While there are many different ions in the cell, including numerous nitrogen and carbon-based organic ions such as ammonium, glutamate, and amino acids, our discussion here primarily focuses on inorganic metal ions, including potassium, magnesium, calcium, sodium, iron, manganese, zinc, and copper (Milo and Phillips, 2015). In addition to being the second most abundant component of any living cell, after water molecules (Figure 1B), these inorganic ions cannot be created or destroyed by cells. Consequently, cells can only regulate their ionic content and concentration by importing or exporting ions from their extracellular environment. Textbooks refer to the global ionic composition of cells as the “metallome” (Figure 1C) (Silva and Williams, 2016, 2013). Despite the high abundance of ions, our understanding of the metallome has lagged far behind that of other cellular components, primarily due to technical difficulties in measuring such weak and short-lived charge-based interactions. Consequently, how biology uses those numerous interactions to accomplish cellular tasks still remains to be elucidated. Here we review recent insights into the functional dynamics of inorganic ions, specifically focusing on examples from bacterial systems. Finally, we also discuss research strategies that may help shed further light on this emerging field of IonoBiology.

Box 1: A tour of biomolecules that require cations for stability, form, and function.

DNA

How the chromosome is condensed while maintaining the ability to dynamically control transcription, replication, and segregation has remained a puzzle. While protein-DNA interactions certainly play an important role (Morgan et al., 1987; Shen and Landick, 2019; Teif and Bohinc, 2011), there is also emerging evidence that ion-DNA interactions are similarly important for the structure of the chromosome (Carrivain et al., 2012; Koltover et al., 2000). Not surprisingly, given the high density of negative charges on DNA, cations such as potassium, magnesium, as well as zinc and sodium are required for chromosomal condensation in cells (Shen and Landick, 2019; Vries, 2010) (Box 1A). In fact, ions are proposed to impact all levels (spatial scales) of DNA organization in cells, including higher level structures such as chromatin (Allahverdi et al., 2015; Kilic et al., 2018). The condensation state of the chromosome can impact transcription, which requires physical access for polymerases and other transcription factors to DNA sequences (Kuhlman and Cox, 2012; Shen and Landick, 2019). Thus, changes in the metallome could play a direct role in regulating transcription.

RNA

Similar to DNA, cations play a role in facilitating the compaction of RNA as it begins to fold, thereby facilitating not only the folding process but also stabilizing the final 3D structure, in either a diffuse or site-specific manner (Bowman et al., 2012; Draper, 2008; Lipfert et al., 2014; Nguyen et al., 2016). Magnesium is particularly important, playing a role in the folding of most large RNA structures, such as tRNAs (Box 1B), as well as a catalytic center for many ribozymes (Bowman et al., 2012; Draper, 2008; Zheng et al., 2015).

Proteins

Ion-protein interactions have been of interest since Hofmeister first discovered the hierarchy of ions in protein solubility in the 19^th century (Hofmeister, 1888; Jungwirth and Cremer, 2014). It is now appreciated that ions play a major role in protein structure and conformational motion, with monovalent ions often interacting transiently with the protein surface, while multivalent ions tend to bind more specifically (Friedman, 2011; Hagedoorn, 2015). Ions often play a key role in protein-protein or protein-ligand interactions (Box 1C). For example, metal ions stabilize many protein-ligand interactions (Strasser et al., 2015; Zheng et al., 2008).

Enzymes

Approximately 35-60% of enzymes need a metal ion to function properly, including several enzymes involved in key biological processes such as photosynthesis, respiration, and nitrogen fixation (Andreini et al., 2008; Yannone et al., 2012). Metal ions can play a variety of roles in enzymes. For multivalent cations, these include activating the substrate, electrostatically stabilizing intermediates, or acting as the catalytic center for a redox reaction (Box 1D). Magnesium, zinc, calcium, and manganese most commonly perform the former two roles, while iron and copper are the most common redox centers. Monovalent cations, due to their lower charge density, tend to interact more transiently with enzymes and their substrates, yet they can still promote enzymatic activity, either through direct coordination or through indirect conformational changes (Gohara and Cera, 2016; Page and Cera, 2006).

Ribosomes

Composed of both protein and RNA, ribosomes contain many negatively charged atoms, such as phosphates. Thus, it is not surprising that crystallographic studies have revealed that each ribosome contains more than 400 bound magnesium and potassium ions (Auffinger et al., 2021; Rozov et al., 2019) (Box 1E). These magnesium and potassium ions play important roles in stabilizing both tertiary and quaternary structure (Hori et al., 2021), with low and high ion concentrations impacting the function of ribosomes (Akanuma et al., 2014, 2018; Gavrilova et al., 1966; Gesteland, 1966; Klein et al., 2004; Näslund and Hultin, 1971; Nierhaus, 2014). It has thus been argued that cells need to maintain a range of ion concentrations within the cell to ensure ribosome function; however, different ionic species can also compete for binding sites on the ribosome, leading to toxicity (Zitomer and Flaks, 1972).

Riboswitches

Riboswitches are RNA molecules that control gene expression in bacteria, as well as in some plants and fungi. These molecules have the ability to form – or switch to – a specific structure that is coordinated by the interaction with the metabolite. Many riboswitches require not only the metabolite to fold properly but also the coordination with multivalent ions (Edwards and Ferré-D’Amaré, 2006; Serganov et al., 2009; Sherwood and Henkin, 2016) (Box 1F). For example, several studies have shown that magnesium can directly coordinate with the metabolite, allowing the riboswitch to recognize its target (Musiari et al., 2014). Also, some riboswitches directly detect the cellular concentration of a specific cation, such as magnesium or manganese (Dambach et al., 2015; Wedekind et al., 2017), and consequently modulate the expression of ion transporters.

ATP

Given the importance of cations in neutralizing the phosphates of nucleic acids, it is perhaps not surprising that ATP, the principal cellular energy source, is predominantly associated with magnesium in solution (Gout et al., 2014; Storer and Cornish-Bowden, 1976) (Box 1G). The connection extends far beyond a transient interaction; it is believed magnesium allows conformation flexibility of the triphosphate, making ATP functionally active. Indeed, magnesium coordination is pivotal to the transition state in ATP synthase (Ko et al., 1999), and nearly all phosphorylating enzymes use Mg-ATP as their substrate, including RNA polymerase and DNA helicase (Bojovschi et al., 2012; Plattner and Verkhratsky, 2016).

A Magnesium ions (red) act as counterions in the DNA structure (PDB ID 4R4D; Mandal et al., 2014) and are necessary for its replication.

B tRNAPhe structure from PDB ID 3L0U (Byrne et al., 2010). The zoomed area shows the interaction of magnesium (red) and potassium (yellow) ions with specific RNA regions, facilitating the folding of the structure.

C Metalloproteinase inhibitor and its ligand (yellow) from PDB ID 1G4K (Dunten et al., 2001). The zoomed area shows the interaction of a zinc ion (green) with the protein and its ligand, promoting the stabilization of the connection.

D Protein cytochrome b562 structure from PDB ID 1LM3 (Springs et al., 2002). The catalytic heme redox center (pink) contains a central iron ion (green).

E Structure of the glycine riboswitch, which regulates gene expression in response to glycine presence, PDB ID 3OWI (Huang et al., 2010). The recognition of the glycine (yellow) is facilitated by magnesium ions (red).

F Magnesium stabilizes the ATP (adenosine triphosphate) structure, generating the ATP-Mg2+ active complex. ATP-Mg2+ is necessary for several enzymes and is the principal source of energy for the cell.

G The ribosomal structure (PDB ID 6QNQ; Rozov et al., 2019) contains approximately 400 metal ions of magnesium (red), potassium (yellow), and zinc (green) to maintain the structure and necessary for its protein production activity.

All molecular graphics (except ATP) were created with UCSF Chimera (Pettersen et al., 2004). Marvin was used for drawing and displaying the ATP-Mg structure, Marvin 21.1, 2020, ChemAxon (http://www.chemaxon.com).

Ion-content modulation and membrane potential dynamics in cells

Cells preserve an electrochemical potential across their membrane by maintaining a charge imbalance with their external environment (Figure 1A). In both prokaryotic and eukaryotic cells, this membrane potential is critical for energy production via ATP synthesis, and plays crucial roles in many cellular processes such as motility (Guragain et al., 2013; Miller and Koshland, 1977; Nakamura and Minamino, 2019) and regulating the transport of various molecules across the membrane (Poolman and Konings, 1993; Zorova et al., 2018). The membrane potential has also been found to regulate fundamental cellular processes in bacteria such as cell division, antibiotic resistance, and cell-to-cell communication (Benarroch and Asally, 2020; Bruni et al., 2020; Damper and Epstein, 1981; Hudson et al., 2020; Prindle et al., 2015; Strahl and Hamoen, 2010). Because of the significance of these processes, it is widely assumed that most cells, and in particular bacteria, maintain a near-constant membrane potential, primarily by tightly regulating ion homeostasis (with the notable exception of excitable cells like neurons and cardiac cells). Recently, there has been growing interest in membrane potential dynamics in non-excitable cells (Badou et al., 2013; Benarroch and Asally, 2020; Kaestner et al., 2018). Specifically, single-cell resolution measurements in bacteria have revealed that the membrane potential, and thus the metallome, can be highly dynamic, with cells experiencing dramatic depolarization and hyperpolarization events, lasting from milliseconds to hours (Figure 1D, and Figure 2, C and G) (Kralj et al., 2011; Lee et al., 2019; Prindle et al., 2015; Sirec et al., 2019). Growing evidence thus suggests that bacteria utilize membrane potential dynamics and changes in specific metal ion concentrations to regulate key cellular processes.

Figure 2.

Examples of membrane potential dynamics driven by ion flux in bacterial cells.

A Single-cell measurements show highly dynamic metallome and membrane potential. Panels B and C were adapted from Bruni et al. This study shows that changes in calcium content in Escherichia coli cells cause changes in membrane potential.

B Fluorescence microscopy images show E. coli cells expressing the CaPR sensor. This construct contains GCaMP6f fluorescent calcium sensor (top image) and PROPS voltage-sensitive fluorescent sensor (PRoteorhodopsin Optical Proton Sensor, bottom image). Scale bar of 5 μm.

C The single-cell time traces show the fluorescence signal of calcium content (in blue) and membrane potential (in red) over time. The results indicate that transient changes in calcium content drive changes in membrane potential which last milliseconds. The y-axis is ΔF/F = 100% per 100 mV.

D Bacillus subtilis biofilms present long-range action potentials propagated through potassium flux.

E Fluorescence time-lapse images of a B. subtilis biofilm over the course of six hours showing the global oscillations in potassium efflux. Extracellular potassium changes are represented by the extracellular potassium dye APG-4 (a.u.). Scale bar of 50 μm.

F The corresponding images of panel E showing the global oscillations in membrane potential. Membrane potential is represented by the Nernstian voltage dye Thioflavin-T (ThT, a.u.).

G Membrane potential oscillations in B. subtilis biofilms (ThT, a.u., blue line) are caused by oscillations in potassium flux (APG-4, a.u., yellow line). From Prindle et al. (2015).

Since cells can neither create or destroy metal ions, they use ion channels and transporters to regulate their membrane potential and ionic content (Figure 1A). Ion channels are membrane proteins that passively and often selectively conduct ions across the cell membrane according to their electrochemical potential, also known as the Nernst potential. In contrast, ion pumps (transporters) require energy, typically in the form of ATP hydrolysis, to pump specific ions against their concentration gradients. Experimental manipulation of ion channels and transporters makes it possible to not only manipulate the membrane potential, but also probe the functional roles of a dynamic metallome in bacteria. Such experimental approaches thus provide an important tool for IonoBiology and set the stage for fundamental biological discoveries beyond bacteria.

The Hodgkin-Huxley model

The relation between the flux of ions across its membrane and the membrane potential of a cell (Figure 1D and E) can be described mathematically using the formalism introduced by Alan Hodgkin and Andrew Huxley in the early 1950s. Originally developed to describe action potentials in neurons (Hodgkin and Huxley, 1952), for which it earned the 1963 Nobel Prize in Physiology, it was later extended to cardiac cells (Noble et al., 2012) and more recently to bacteria (Lee et al., 2019; Martinez-Corral et al., 2019; Prindle et al., 2015; Yang et al., 2020). The model describes the time dependence of the electric potential difference across the cell membrane using an electric circuit analogy (Figure 1E). According to this description, the cell membrane acts as a capacitor that separates various inorganic ions, whose concentrations differ between the interior and exterior of the cell. Embedded within the membrane, selective ion channels act as resistors that enable the flux of specific ions into and out of the cell. Charge balance requires that the currents across the resistors (ionic currents, I_ion) and the capacitor (capacitive current, I_cap) cancel each other out, so that

$$I_{cap}=C_m\frac{dV}{dt}=-{\sum }_{ions}{I}_{ion}=-{\sum }_i{g}_i(V-V_i)-g_{leak}(V-V_{leak})$$ (1)

where the subindex i represents all the ions relevant to the problem under study (Na⁺ and K⁺ in the original work of Hodgkin and Huxley, and later extended to many other cations). The leak term represents the rest of ion fluxes and other biological processes that determine the steady-state value of the membrane potential, including ion pumps. The coefficients g_i represent the conductances of the various ion channels, which can be modulated by the membrane potential itself (in voltage-gated channels) or by other biological processes, such as cellular stress in the case of bacterial biofilms (Prindle et al., 2015). This modulation is usually made explicit through gating variables (originally referred to by H.H. as n, m, etc.), which follow their own dynamical equations that depend on the modulating factor (V itself, or stress, for instance) (Hodgkin and Huxley, 1952).

The way in which ion concentrations affect the dynamics of the membrane potential V in Equation 1 is through the Nernst potentials V_i. For a given ion, V_i corresponds to the potential at which the corresponding net ion flux would stop, due to a balance between the effects of the concentration gradient (between the cell’s interior and exterior) and the electric field. Its value is given by

$$V_i=\frac{RT}{zF}\text{ln}\frac{X_i^{ext}}{X_i^{int}}$$ (2)

where $E_i^{ext}$ and $E_i^{int}$ are the extracellular and intracellular cation concentrations, respectively, R is the gas constant, T is the temperature, z is the ion charge, and F is Faraday’s constant.

The traditional view is that during an action potential in neurons, the change in the intracellular ionic concentration (and thus the Nernst potential itself) can be considered negligible, because of the relatively large volume of eukaryotic cells. However, the small size of bacteria compared to eukaryotic cells means that equivalent changes in membrane potential yield larger fractional changes in their ion concentration. For example, a 100 mV depolarization of a typical eukaryotic cell with an ~100 mM ion concentration (volume of ~1 pL, and surface area of 1600 μm²) leads to the net transfer of only a small fraction of its ion across the membrane (~10⁻⁴). The same 100 mV depolarization in a cell the size of a typical bacterial cell (~1 fL, and 6 μm²), because of its larger surface area to volume ratio, leads to the transfer of an order of magnitude larger fraction of its charged molecules (~10⁻³) (Benarroch and Asally, 2020). As mentioned before, these calculations assume identical resting ion concentrations and membrane potential changes, which may not be true in biologically relevant cases. Thus, it is possible that changes in the bacterial cell membrane potential can be even larger and cause appreciably changes in their metallome. Recent studies have provided evidence for striking changes in membrane potential of bacteria that in turn can lead to nontrivial interactions between the ion content of bacterial cells and their physiological state (Lee et al., 2019; Luder et al., 2021; Prindle et al., 2015; Xu et al., 2020). Presumably, these large changes in the metallome of bacteria have facilitated their experimental observation, enabling the pursuit of how such ionic changes may serve functional roles, and thereby setting the stage for IonoBiology from a bacterial perspective as discussed below.

Functional consequences of the dynamic metallome

As discussed in Box 1, various metal ions play key roles in the structure and function of macromolecules, thus it should not be surprising that bacteria can control the activity of specific molecules by modulating their metallome. For example, we recently uncovered that changes in magnesium ion content increase bacterial tolerance against aminoglycoside antibiotics that target ribosomes (Figure 3, A-C) (Lee et al., 2019). In contrast to the commonly accepted view that only non-growing, or dormant bacteria can have high tolerance to antibiotics, we showed that actively growing bacteria with higher magnesium content can also survive in the presence of antibiotics that target ribosomes. These findings further emphasize the importance of the hundreds of structural magnesium ions that are an integral part of the ribosome complex (Box 1E). Our findings are also in line with previous results from other groups (Akanuma et al., 2014), which indicate that increasing the cellular magnesium content stabilizes the ribosomal complex. From a fundamental perspective, these observations demonstrate how changes in the ionic composition of cells affects the ancient and fundamental molecular process of protein synthesis. Interestingly, other metal ions, such as iron have also been suggested to promote the appearance of antibiotic-resistant bacterial strains (Méhi et al., 2014). From a biomedical perspective, these studies suggest ways to combat antibiotic resistance through targeting ion flux and the metallome of bacteria.

Figure 3.

IonoBiology; examples of functional dynamics of the bacterial ionic content.

A Ions play an important role in structural stability and function of different macromolecules. For example, ribosomes contain hundreds of bounded magnesium ions (see also Box 1G).

B The increase in magnesium content in B. subtilis cells increases the survival upon ribosomal stress (Akanuma et al., 2014; Lee et al., 2019).

C The addition of magnesium in agar plates increases the number of colonies in the presence of Spectinomycin, a ribosome targeting antibiotic. From Lee et al. (2019).

D Ions can act as signaling cues in bacterial cells.

E B. subtilis biofilms present oscillations in potassium efflux which attracts motile cells (Humphries et al., 2017).

F Filmstrip showing the edge of a biofilm and its membrane potential oscillations (ThT in cyan) due to potassium signaling. The oscillations in potassium efflux periodically attract motile cells labeled red here using the genetically encoded Phyp-mKate2 inducible promoter expressing a red fluorescent protein. The periodic attraction of motile cells ultimately leads to their incorporation into the biofilm. Signaling is species-independent. Scale bar of 50 μm. From Humphries et al. (2017).

G Bacterial cells use the flux ions as a response to different types of stresses.

H Potassium efflux through the YugO channel is essential for the complete development of B. subtilis biofilms on agar surfaces (Lundberg et al., 2013; Prindle et al., 2015).

I Images show wild-type (WT, left image) and ΔyugO mutant (right image) biofilms that were grown in MSgg medium for 48h at 30°C. Scale bar of 5 mm.

Beyond modulating the activity of a specific macromolecule, ions can also act in intracellular signal transduction in bacteria. For several decades, it has been suggested that calcium plays an important role in various bacterial functions, including chemotaxis (Liu et al., 2020; Tisa and Adler, 1992), motility (Gode-Potratz et al., 2010; Guragain et al., 2013; Liu et al., 2020), virulence (Broder et al., 2016), and cell differentiation (Raeymaekers et al., 2002). However, understanding how calcium regulates such processes has often proven challenging due to a scarcity of research into the function of prokaryotic calcium binding proteins and the challenge of detecting intracellular calcium concentrations or dynamics (Domínguez, 2004; Domínguez et al., 2015; Norris et al., 1996). Recently, mechanically-induced voltage depolarization in single Escherichia coli cells has been shown to cause calcium-mediated intracellular signaling. By employing a novel voltage-sensitive fluorescent protein, Kralj et al. were able to measure the bacterial membrane potential of single cells (Figure 2, A-C), revealing depolarization on a timescale of milliseconds (Kralj et al., 2011). Later, a genetically encoded calcium sensor was used to demonstrate that the depolarization events induce calcium influx. Furthermore, the voltage-induced calcium influx events increased on agarose pads, suggesting ion flux is mechanically induced. While a specific molecular mechanism for this mechano-sensation remains elusive, the calcium influx seems to modulate protein expression levels in cells, increasing the expression of the virulence translation regulator Hfq (Bruni et al., 2017). Several pathways involved in these calcium transients were also recently identified by high-throughput screening (Luder et al., 2021). Therefore, it appears that bacteria employ voltage-gated calcium channels similarly to neurons or muscle cells to regulate protein expression in the cell.

Another example of ion-mediated signaling in bacteria is through iron. Iron is only weakly soluble and is toxic at high concentrations, yet it is required in small amounts by cells for survival. Bacteria, and particularly pathogenic bacteria, have developed chelating proteins, called siderophores, to scavenge and help transport iron into the cell. Because many hosts actively limit the amount of iron available to pathogens, iron limitation can be a useful signal to promote virulence. Indeed, many bacteria contain an iron-sensing transcriptional control protein called Fur that contributes to virulence (Ratledge and Dover, 2000). Other bacteria also physiologically respond to Heme- or siderophores bound to iron (Beare et al., 2003; Skaar, 2010). Zinc and manganese have similar functions in promoting virulence (Ma et al., 2015; Porcheron et al., 2013).

Finally, ions can mediate intercellular signaling between bacteria cells. Many bacteria are able to form multicellular communities called biofilms. While these biofilms provide resistance to antibiotics, these crowded communities also limit nutrient access to cells in the biofilm interior. The first example of the functional role of potassium ion channels in bacteria was published by our group, showing that B. subtilis bacteria within biofilms generate action potentials under this nutritional stress (Beagle and Lockless, 2020; Prindle et al., 2015) (Figure 2, D-G). This discovery was motivated by our preceding study (Liu et al., 2015) which revealed long-range cooperation among bacteria that reside within the biofilm interior and periphery. We identified a periodic halting of biofilm growth so that nutrients can diffuse towards the starved cells located in the biofilm interior. By using a microfluidic-based technique that allowed us to temporally modulate the composition of the growth media, we determined that the charged metabolites glutamate and ammonium were involved in this cooperation. However, our measurements suggested the cooperation occurred too quickly to be mediated by passive diffusion of metabolites. Instead, we later demonstrated that bacteria communicate across long distances (over cm compared to the ~1μm long cell, with the signal traveling ~100 μm/min) via ion channel-generated action potentials. Specifically, glutamate starvation results in the periodic opening of potassium ion channels, presumably through gating domains of bacterial ion channels which are sensitive to the cell’s metabolic activity (Cao et al., 2013; Krüger et al., 2020; Zhang et al., 2020). The released potassium, in turn, triggers depolarization of neighboring cells, causing them to halt growth and release their potassium as well. This domino-effect results in a global wave of extracellular potassium propagating through the biofilm community (Figure 2, E and F), which is driven by single-cell action potentials. These bacterial action potentials conform precisely to the mathematical framework developed by Hodgkin and Huxley. Accordingly, biofilms provide a new prokaryotic paradigm for complex population-level behavior that emerges from electrochemical communication among millions of cells. Interestingly, this electrochemical communication can extend beyond the community facilitating the communication between B. subtilis biofilms (Liu et al., 2017), and even attracting bacteria of a different species to a pre-existing biofilm (Humphries et al., 2017) (Figure 3, D-F). Thus, the formation and composition of mixed-species bacterial communities could be regulated through this electrochemical signaling mediated by ion channels.

The role of ion channels in action potentials has traditionally been studied in the context of neurons and other animal-centric tissues such as muscles (Jan and Jan, 1989; Thorneloe and Nelson, 2005). However, the evolution of ion channels long precedes that of eukaryotic and multicellular organisms (Martinac et al., 2008) and is conserved across the domains of life (Jan and Jan, 1992). In fact, bacterial and mammalian ion channels exhibit numerous structural similarities, which has led structural biologists to utilize bacterial ion channels to better understand their mammalian counterparts. Our results revealed that bacterial and mammalian ion channels are not only structurally similar, but also seem to share a similar functional role in generating action potentials for the purpose of communication (Prindle et al., 2015). The paradigm of ion channel mediated signaling among cells through propagation of action potentials is thus no longer viewed to be exclusive to neurons and cardiac cells. This provokes the question of what other functions bacteria may have up their sleeves that are commonly attributed to “higher” organisms, especially given the recent demonstration that membrane potential-based memory can be encoded in single bacteria within biofilms (Yang et al., 2020).

The Path Forward

Weak charge interactions in ion-rich environments such as the cytoplasm are an inherently interdisciplinary phenomenon at the interface of physics and chemistry. These fields have traditionally not considered such interactions in the crowded and heterogeneous environment that represents the interior of a cell. In addition, computational power is now beginning to approach the level required to realistically simulate interactions among biomolecules within heterogeneous and ion-rich conditions at biologically relevant time scales. Measuring how ionic interactions at the molecular level drive processes at the cellular or multicellular level also requires experimental techniques, such as microscopy, to be multiplexed and multiscaled. The study of IonoBiology thus requires the integration of biology, physics, chemistry, computer science, and imaging, as summarized in Box 2.

Box 2: Interdisciplinary fields for IonoBiology.

Chemistry

Synthetic organic and physical chemistry can be used to mimic the cell’s crowded ionic interior. For example, ionic hydrogels and other polymer chemistry approaches can be used to deconstruct how changes in ionic concentration affect RNA, lead to conformational changes in proteins, and in general drive macromolecular interactions (Fels et al., 2009; Kowacz and Pollack, 2020). Novel chemical sensors can also be developed to precisely measure the local environment in such systems, including changes in pH, the concentration of specific ions, and viscosity (Clark et al., 1999; Shamsipur et al., 2019; Xie and Bakker, 2015). Developing in vitro tools and techniques will in turn facilitate the ultimate goal of applying such approaches in living cells.

Physics-based frameworks

Physics-based models provide an ideal framework to model ionic interactions across spatial and temporal scales (Bedrov et al., 2019). Condensed matter physics provides a suitable framework to better understand macroscale materials, such as ionic gels, which approximate the cytoplasmic interior of the cell (Kowacz and Pollack, 2020). In addition, existing theories can characterize the thermodynamics and phase equilibria of solutions (Brangwynne et al., 2015; Ditlev et al., 2018). Typically, these fields currently focus on solutions of simple particles and thus would need to be expanded to include the dynamics and heterogeneity inherent in cellular systems. Finally, nonlinear dynamics can be used to better understand how these molecular changes drive physiological changes at the single-cell or tissue level (Ferrell, 2012).

Computer simulations

Molecular simulations will also play an important role in studying short-lived and weak ionic interactions. While it may take many years before we can experimentally measure ionic interactions and dynamics with molecular resolution in living cells, currently available computer simulations can provide many insights (Bedrov et al., 2019; Friedman, 2011). It is possible to use fundamental laws of physics to study ionic interactions and dynamics with atomic resolution. However, current large-scale molecular dynamics simulations are commonly limited to the nanosecond or millisecond time scale. Such limitations should be overcome with advances in computational power. In the meantime, it is possible to use coarse-graining approaches to expand the time scale of computer simulations, while keeping sufficient spatial resolution to assess the effects of ions on biomolecules and their interactions (Campitelli et al., 2020; Singh and Li, 2019).

Electrical engineering

Advances in electrical engineering make it possible to design new electronic interfaces with cells. New nano-electrode approaches are providing unprecedented interfaces with cells at the nanometer scale (Lubrano et al., 2020; Stratford et al., 2019). These devices provide the ability to apply local, time-controlled electrical fields that can span across the frequency spectrum of electromagnetic waves. They can also inform the development of bioelectronic systems to assess the effects of electromagnetic radiation on cells. This work may also identify new electronics-based approaches to precisely control and manipulate cells and tissues. These technological tools will be extremely helpful in perturbing the dynamic metallome of cells and gain deeper insight into the functional roles of ions and ionic interactions.

In vivo measurements

Ultimately, direct measurements of the dynamic metallome and ionic interactions are needed. The challenges of measuring non-covalent interactions necessitate a combination of the various technological advances mentioned above. In addition, the need for in vivo imaging of changes at the molecular, cellular, and multi-cellular level will likely involve integrating these techniques with super-resolution and multiplexed imaging approaches. Structural biology has already provided a wealth of information about how ions effect macromolecular structure (see Box 1, Matsarskaia et al., 2020) as well as how ions are transported across the cell membrane (Stautz et al., 2021). Finding novel ways to acquire and interpret data from Nuclear Magnetic Resonance, immunoprecipitation, and other standard techniques, may also allow us to infer details of these ionic interactions, particularly in combination with the modeling methods described above (Bonomi et al., 2017).

For large interdisciplinary research endeavors, such as the one discussed here, it is important to define what may constitute success. Here we propose two quantifiable benchmarks of success for IonoBiology, focused respectively on knowledge generation and the identification of new fundamental principles of biological regulation. Such measures will provide a helpful outlook to establish IonoBiology as a worthy endeavor and hopefully recruit young scientists to further define and shape this field.

(i). Knowledge generation.

The scientific community would benefit from establishing a catalog of the functions of all the ions that are present in cells. Specifically, it is important to establish the roles of ions in the global coordination of cellular functions, and whether ions can compete with and substitute for each other functionally. This goal requires a precise measurement of all ionic species, focused on metal cations, with single-cell resolution to identify cell-to-cell heterogeneities that can reveal the range and impact of changes in the metallome of individual cells. This benchmark goes beyond simply mapping the language that we intend to decode; it will broaden our understanding of why essential ions are necessary and how cells might modulate their ionic composition to control cellular and physiological functions.

(ii). Novel principles of biological regulation.

We propose that the long-term objective of IonoBiology should be to determine and understand how ion fluxes and electrostatic interactions in cells regulate fundamental biological processes. Ion-based signaling has been suggested to provide an important mechanism for organizing multicellular systems, from biofilms to embryos (Cervera et al., 2020). A focal point of future efforts could be to establish how changes in ionic strength and composition affect bacterial cellular decision-making, with an emphasis on community development, stress responses and antibiotic tolerance (Figure 3). This effort will directly benefit from the above-mentioned characterization and cataloging of the prokaryotic metallome dynamics. Given the vast number of ionic interactions within any given cell, it is very likely that these efforts will uncover a new layer of molecular regulation of fundamental biological processes. In particular, bacterial cells may modulate their cytoplasmic ion composition to globally regulate transcription and translation. It is likely that the “central dogma of biology” as formulated by Sir Francis Crick decades ago, is globally controlled by changes in the ionic composition and strength of the cytoplasm. Furthermore, ionic content and interactions may play a role in bacterial cell fate decisions, such as entry into, and exit from dormancy. Such discoveries would provide a new paradigm for understanding and controlling the regulation of fundamental stress responses in bacteria. We hope that this article, which is part review and part perspective, will be helpful for the larger scientific community to become interested in and get involved in the fascinating and promising area of IonoBiology.