Divalent cation signaling in immune cells

Abstract

Divalent cations of two alkaline earth metals Ca²⁺ and Mg²⁺ and the transition metal Zn²⁺ play vital roles in the immune system, and several immune disorders are associated with disturbances of their function. Until recently, only Ca²⁺ was considered to serve as a second messenger. However, signaling roles for Mg²⁺ and Zn²⁺ have been recently described, leading to a reevaluation of their role as potential second messengers. Here we review the roles of these cations as second messengers in light of recent advances in Ca²⁺, Mg²⁺ and Zn²⁺ signaling in the immune system. Developing a better understanding of these signaling cations may lead to new therapeutic strategies for immune disorders.

Cytosolic free ionic pool and cellular homeostasis

In eukaryotic cells, divalent cations exist in two principal states; one tightly bound to proteins or other negatively-charged macromolecules such as mono- or polyphosphates, and a second ionized state involved in dynamic chemical equilibria. The “bound” pool represents the majority of the intracellular cation and plays a number of roles as a vital structural and functional co-factor through strong electrostatic interactions. On the other hand, the ionized fraction (0.05% to 10%) of each cation remains “free” in the cytosol or sequestered into organelles, such as the endoplasmic reticulum (ER) or mitochondria. The signaling ability of an ion relies on the high chemical activity of the intracellular “free” pool, and its ability to be rapidly modulated without affecting the total cellular amount of the ion (Box 1: Notion of ion signaling). This modulation results from mobilization of the cation across the plasma membrane (PM) or from intracellular stores to increase the cytosolic concentration to generate transient binding complexes between the cation and proteins or other macromolecules. There are numerous technical considerations that affect the accurate measurement of intracellular concentrations divalent cations. Whereas the total amount (bound and free) can be quantified by destructive biophysical analysis, the assessment of the free pool is achieved mainly by the use of chemical indicators [1].

Box 1. Notion of ion signaling.

The second messenger concept is well defined, however its application to ion signaling requires some adjustments. Indeed, ions signal through variation of their intracellular concentrations via different transport mechanisms, which is different from the “de novo” production of a second messenger by an enzyme. Also, the size and rapidity of movement could mean that the effective “range” of signaling cations is quite broad in the cell. The concept of ion signaling relies on several fundamental features: (1) a cytosolic resting free pool present in unstimulated cells that increases in response to an extracellular stimulus, such as the engagement of a cell surface receptor to become a mobilized free pool through mechanisms supporting its (2) homeostasis and (3) mobilization of the ion from the extracellular milieu, internal stores, or a bound depot; and (4) the mobilized free pool needs to alter one or more cellular processes at physiological level.

1. Cytosolic free pool. In eukaryotic cells, divalent cations are mostly bound with protein or other bioactive molecules and play essential structural and functional roles. For example Zn²⁺ is associated with up to 10% of all cellular proteins including over 300 enzymes and more than 2,000 transcription factors. Similarly, Mg²⁺ is associated with more than 300 enzymes as well as nucleotides, nucleic acids, and other negatively charged macromolecules. Given the essential structural and functional roles of these ions, signaling functions require the existence of a pool that can be modulated without affecting those functions. This pool is the cytosolic free pool which is considered free because it is in ionized form and able to bind to potential effectors. The difference between the bound and free forms is that the association constant for the divalent cation is much lower for the former.

2. Cellular homeostasis. To fulfill its purpose without affecting the total amount of the cation, the cytosolic free pool generally represents a small fraction of the total intracellular amount. To support the signaling functions of the cytosolic free pool, homeostatic regulation maintains the cytosolic free pool low at resting state by extrusion to the outside of the cell, sequestration in intracellular pools (stores) or immobilization by binding to cytosolic binding partners (bound).

3. Mobilization. Extracellular, or intracellular (not represented in the figure), stimuli trigger the mobilization of the cytosolic free pool via release from intracellular stores or bound pool, or by transport from the extracellular environment.

4. Effectors. Finally, the mobilized pool modulates cellular functions via tipping the equilibrium to the cation-bound form of specific effector molecules generally with the comparatively high association constant that is near the concentration achieved by the mobilized free pool and above the resting free pool concentration.

The ability for a cation to be mobilized in the cell, especially across the plasma membrane, depends on two driving forces, the chemical and the electric gradient [2]. The chemical gradient corresponds to the net difference of concentrations between the extracellular (or reservoir) environment and the cytosol. For example, Ca²⁺ and Zn²⁺ intracellular free concentrations, ([Ca²⁺]_i and [Zn²⁺]_i, respectively, are maintained at approximately 10⁴-fold lower than the physiological extracellular concentration, thus generating a large chemical gradient for their mobilization into the cytosol (Table 1). On the other hand, there is a much smaller difference (< 2-fold) between cytosolic free Mg²⁺ ([Mg²⁺]_i) and extracellular free Mg²⁺ ([Mg²⁺]_o) level leading to the conventional wisdom that it is a poor candidate for a second messenger (Table 1) [1, 2]. However, because non-excitable cells, such as immune cells, harbor a negative membrane potential ≈ −70mV, if intracellular free Mg²⁺ was at electrical equilibrium, its resting concentration should be 50 mM [3, 4]. Nevertheless, [Mg²⁺]_i ranges from 0.2 to 0.5 mM, showing that intracellular free Mg²⁺ is regulated and maintained at a lower concentration. This creates an electrochemical gradient of 100 to 250-fold for Mg²⁺, which is sufficient to allow rapid mobilization across the plasma membrane (PM) (Table 1, see Regulation of cytosolic free Mg²⁺ section) [3, 5, 6]. Indeed, the physiological function of rapid Mg²⁺ fluxes in T lymphocytes has been recently demonstrated and led to interesting insights into novel immunoregulatory mechanisms [7]. Table 1 summarizes extracellular and intracellular free concentrations of Ca²⁺, Mg²⁺ and Zn²⁺ as well as their physicochemical properties.

Table 1.

Ca²⁺, Mg²⁺ and Zn²⁺ characteristics

Ion	[x]e	% Free	Free [x]i	E/I gradienta,b	Ion radius (pm)	H/Ic	References
Ca2+	1–2 mM	0.05%	100 nM	10,000	100	70	[11]
Mg2+	1 mM	5–10%	0.2–0.5 mM	2–5b	72	210	[3, 5, 6]
Zn2+	15 µM	5–10%	0.4 nM	20,000	74	196	[122, 123]

^bExtracellular (E)/intracellular (I) concentration gradient,

^bThe electrochemical E/I gradient for Mg²⁺ is 100–250,

^cH/I, hydrated to ionized volume ratio

Here, we discuss the criteria for ion signaling and report recent advances on Ca²⁺, Zn²⁺ and Mg²⁺ mobilization and signaling in immune cells as well as their importance in human disease pathophysiology and treatment.

Homeostasis and mobilization of divalent cations in immune cells

Mobilizing Ca²⁺

In the basal state, cytosolic free Ca²⁺ concentration ([Ca²⁺]_i) is maintained at 100 nM by active extrusion of the Ca²⁺ from the cytosol. This occurs either through the PM by the Ca²⁺-ATPase (PMCA) and Ca²⁺/Na⁺ exchangers (NCX), or by deposition in ER or mitochondrial stores by the sarcoplasmic/ER Ca²⁺-ATPase (SERCA) or the mitochondrial Ca²⁺ uniporter (MCU), respectively (Figure 1a, Table 2) [8–10].

Figure 1. Divalent cations transporters.

(a) Ca²⁺ transport mechanisms in immune cells. Ca²⁺ homeostasis regulation involves the PMCA, SERCA ATPases as well as the MCU and Na⁺/Ca²⁺ exchangers (NCX). SOCE is induced by released of Ca²⁺ from the ER by the IP₃R. STIM senses the store depletion and triggers the opening of the Ca²⁺ CRAC channel, ORAI. SOCE is modulated by two K⁺ permeable channels, K_Ca 1.3 and K_v 3.1 regulated by intracellular Ca²⁺ and depolarization respectively. These two channels help sustain SOCE generation by inducing hyperpolarization of the cell. Additionally, the Na⁺ permeable channel TRPM4 is gated by intracellular Ca²⁺ and reduces the Ca²⁺ mobilization. (b) Mg²⁺ transport mechanisms in immune cells. Mg²⁺ transporters include TRPM7, which have an intracellular Ser/Thr kinase domain, MAGT1, and SLC41A1/2 expressed on the PM and Mrs2 on the inner membrane of the mitochondria. (c) Zn²⁺ transport mechanisms in immune cells. Zn²⁺ transporters of the ZnT and ZIP families are expressed in various cellular compartments. Zn²⁺ homeostasis is completed by MT. Abbreviations: ChR, Chemokine receptor; DAG, Diacylglycerol; FcR, Fc receptor; TPC, Two pore channel.

Table 2.

Ca²⁺, Mg²⁺ and Zn²⁺ channels

Channel	Selectivity	Loc.	Stimuli	Regulation	Expression	Functions	References
Calcium
SOCE and related
ORAI1	Ca2+	PM	TCR, BCR, TLR, FcR, Chemokine Rc	SOCE, inhibited by Ca2+ and ROS	Ubiquitous	Cell proliferation, Cytokines production, Degranulation, Phagocytosis, Chemotaxis, Treg development	[124]
ORAI3	Ca2+	PM	TCR	SOCE, inhibited by Ca2+	Upregulated in activated T cells	Cell proliferation, Cytokines production	[19]
TRPC1/6	Ca2+/Na+	PM	TLR, FcR, chemokine Rc	SOCE, ARC	Mast cells, Neutro. (B, T cells ?)	Cytokine production, Degranulation	[125, 126]
Cav (1.2/1.3/1.4)	Ca2+	PM	Voltage–independent	Inhibited by STIM	T cells	SOCE modulation ?, T cell homeostasis	[110, 127]
			Voltage-dependent	SOCE (STIM independent)	DC, Macro.	Cytokine production, Chemotaxis, MHC II upregulation	[54]
SOCE modulators
KCa3.1	K+	PM	TCR, BCR	Increased [Ca2+]i	TH1, TH2, TCM, B cells	Hyperpolarization (sustains SOCE)	[128]
Kv1.3	K+	PM	TCR, BCR	Depolarization	Naïve T cells, TEM, TH17	Hyperpolarization (sustains SOCE)	[128]
TRPM4	Na+	PM	TCR, BCR	Increased [Ca2+]i	Ubiquitous	Depolarization (inhibits SOCE)	[129]
Non SOCE
Cav (1.2/1.3/1.4)	Ca2+	PM	Voltage–independent	Inhibited by STIM	T cells	SOCE modulation ?, T cell homeostasis	[110, 127]
			Voltage-dependent	SOCE (STIM independent)	DC, Macro.	Cytokine production, Chemotaxis, MHC II upregulation	[54]
P2X (1/4/7)	Ca2+/Na+	PM	Extracellular ATP/UTP	Inhibited by Mg2+	T, B, NK cells, Neutro.	Pro-inflammatory Cytokines production, Chemotaxis, Proliferation	[117, 118, 120, 121]
TRPV1/2	Ca2+/Na+	PM	Heat (fever?), low pH, mechanical stress		Mono., Macro.	Degranulation, Phagocytosis, Cytokines production	[107, 108]
TRPC3/6	Ca2+/Na+	PM	PLC activation (DAG), PIP2		T, B, NK cells, Neutro.	Chemotaxis, Degranulation	[130, 131]
TRPM2	Ca2+/Na+	PM Lys.	H2O2, NAADP, cADPR		T, B, Neutro., Mast cells, DC	Cytokine production, Degranulation	[101]
Magnesium
TRPM6	Mg2+>Ca2+	PM		Inhibited by [Mg2+]i	Gut, Kidney, Hematopoietic (not T cells)	Unknown in immune cells	[3]
TRPM7	Mg2+>Ca2+	PM	Unknown (BCR, TCR?) PIP2(?)	Inhibited by [Mg2+]i	Ubiquitous	T cell development, T and B cells proliferation, Cytokine production	[36]
MAGT1	Mg2+	PM	TCR	Unknown	Ubiquitous	CD4 T cells development, SOCE induction in T cells, NKG2D expression	[7, 32]
Zinc
Importers
ZIP1	Zn2+	PM		Unknown			[38]
ZIP2	Zn2+	PM	Unknown		Upregulated in asthma and sepsis	Unknown	[38, 92, 94]
ZIP3	Zn2+>others	PM		Unknown	CD34+ progenitors	T cell development	[38, 132]
ZIP4	Zn2+	PM		Unknown	Intestine	AE	[38]
ZIP6	Zn2+	PM		TCR (?)	Ubiquitous, T cells	INFγ production in T cells	[43–45]
ZIP8	Zn2+	Lys.		TCR (?)	Ubiquitous, T cells, Mono.	INFγ production in T cells	[43–45]
ZIP14	Zn2+	PM		LPS			[46, 47]
Exporters
ZnT1	Zn2+	PM		Unknown	Upregulated in T cells	Unknown	[37]
ZnT4	Zn2+	PM		Unknown	Upregulated in T cells	Unknown	[37]
ZnT5	Zn2+	Endo., PM (?)	FcεR stimulation	Unknown	Upregulated in activated mast cells	Recruitment of PKC to the PM (?)	[37, 68]

Abbreviations: ARC, arachidonate-regulated Ca²⁺ channel, Endo., endosome; Loc., localization; Lys., lysosome; Mito., mitochondria; PA, Prader-Willi/Angelman syndrome; Ves., vesicles

Many immune functions are triggered by receptor-mediated acute elevations of cytosolic free Ca²⁺. The main mechanism leading to this elevation is the store-operated Ca²⁺ entry (SOCE), which involves mobilization across the PM triggered by the release of Ca²⁺ stored in the ER. SOCE is induced by a wide range of immune receptors including the T cell receptor (TCR), the B cell receptor (BCR), Fc receptors (FcR), Toll-like receptors (TLR) and chemokine receptors, among others. Receptor engagement activates various isoforms of phospholipase C (PLC), leading to the generation of the second messengers inositol 1, 4, 5-triphosphate (IP₃) and diacylglycerol (DAG). IP₃ activates Ca²⁺ release through the IP₃ receptor channels (IP₃R) on the ER. The depletion of Ca²⁺ from the ER stores is sensed by the N-terminal region of the stromal interaction molecule (STIM). This results in STIM oligomerization and translocation to a region of the ER proximal to the PM. The cytosolic C-terminal region of STIM then recruits and activates the prototypical Ca²⁺-release activated Ca²⁺ (CRAC) channel, ORAI, which promotes Ca²⁺ influx through the PM [11]. Our understanding of SOCE has increased greatly since the discovery of STIM and ORAI and recent updates are reviewed in [11].

Genetic deficiencies of ORAI1 and STIM1 in mice and humans lead to a drastic reductions of SOCE in immune cells including T cells, B cells, NK cells and mast cells and impaired T cell and mast cell activation (see Ca²⁺ and SOCE deficiency section [12–14]. In addition, there are two STIM isoforms, STIM1 and STIM2, that differ in their relative abundance and activation thresholds [15]. STIM1 activation requires higher Ca²⁺ store depletion than STIM2 [15]. In the mouse, deletion of both STIM isoforms in T cells profoundly decreases SOCE, greater than either one alone, and causes a lymphoproliferative disorder associated with a selective reduction of regulatory T cells (T_regs) [16]. Similarly, in B cells, the loss of both STIMs diminishes interleukin 10 (IL-10) production by regulatory B cells [17]. In both cell types, STIM2 deficiency alone leads to a milder SOCE decrease than STIM1 deficiency [15, 16]. However, STIM2 deficient T cells and B cells exhibit lower sustained Ca²⁺ mobilization and reduced nuclear translocation of the transcription factor, nuclear factor of activated T-cells (NFAT) [15, 16]. Additionally, recent studies suggest that both isoforms are required for optimal cytotoxic function of CD8⁺ T cells [15, 18]. The specific role of STIM2 in immune cells remains to be elucidated. Similarly, the three isoforms of ORAI are ubiquitously although differentially expressed in distinct cell types [15]. The ORAI1 knockdown (KD) in mouse T cells completely abrogates SOCE, whereas an ORAI2 KD has no effect implying that the latter may have no role in T cells [15]. However, other studies detected residual SOCE in ORAI1 deficient murine mast cells and T cells, suggesting ORAI2 participates in SOCE in these cells [13]. ORAI3 is induced after T cell activation and, unlike ORAI1 and ORAI2, is not inhibited by reactive oxygen species (ROS) [19]. Given that differentiation of naïve T cells into effector cells is associated with desensitization to ROS, ORAI3 might mediate SOCE desensitization following T cell activation [11, 19]. In addition to SOCE, store-independent or non-SOCE are also potentially important, although less established, mechanisms, for Ca²⁺ mobilization in immune cells and are discussed in Box 2.

Box 2. Non-SOCE Ca²⁺ mobilization: new directions and controversies.

In addition to SOCE, immune cells several non-SOCE, or store-independent, Ca²⁺ mobilization mechanisms that are triggered by specific extracellular or intracellular ligands in immune cells (summarized in Figure 3 and Table 2) [95]. However, the physiological importance of some of these mechanisms remain to be established.

The superfamily of the transient receptor potential (TRP) channels encompass 28 members divided in six subfamilies based on their sequence homology [100]. The role of certain TRPs for Ca²⁺ mobilization in immune cells is relatively well established. For example, TRP melastatin 2 (TRPM2) is regulated by various intracellular ligands related to oxidative stress, such as cyclic ADP-ribose (cADPR), hydrogen peroxide (H₂O₂), nicotinic acid adenine dinucleotide phosphate (NAADP), and adenosine monophosphate (AMP) [101]. In vitro stimulation of T cell and B cell lines by ADPR show that TRPM2 triggers Ca²⁺ mobilization [101]. In neutrophils and macrophages, TRPM2 is expressed in the PM and promotes the production of various inflammatory cytokines in response to ROS, LPS and cellular acidification. Whereas, in DC, TRPM2 is expressed in the lysosomes and seems to play a role in the chemotaxis and maturation [102–104]. Another interesting example is the members of the vanilloid TRP family, TRPV1 and TRPV2 that are both expressed in immune cells [105, 106]. These channels associated with nociception are activated by noxious heat (>43°C), low pH, lipidic ligands such as prostaglandins, and mechanical stress suggesting an interesting intersection between nociception and inflammation [107, 108]. TRPV2 deficiency in macrophages is linked to impaired phagocytosis, whereas TRPV1 may have a role in sepsis but this remains unclear [107, 109]. TRPV2 has also has been associated with mast cell degranulation, however, in these studies the stimulus used were not physiological and the relevance of these findings requires further investigation [108].

In excitable cells, such as muscle cells and neurons, L-type voltage-dependent Ca²⁺ channels (Ca_v1) and ryanodine receptor channels (RyR) are responsible for SOCE after depolarization. These are also expressed in immune cells, including T cells, B cells and DC. In macrophages and DCs, Ca_v activation is voltage-dependent and causes SOCE by promoting Ca²⁺ release from the ER through a direct interaction with the channel RyR₁ after LPS stimulation [54]. On the other hand, the role of Ca_v in T cells is more controversial [110]. Whereas, T cells expressed all Ca_v1 channel isoforms (Ca_v1.1, Ca_v1.2, Ca_v1.3 and Ca_v1.4) along with their regulatory subunits β3 and β4, their activation in T cells is not induced by depolarization [54, 111–113]. Recent studies established roles for Ca_v1 in thymic development and naïve T cell homeostasis in mouse [114]. However, incomplete characterization of the biophysical properties of Ca_v1 in T cells, causes their role in T cell function to be debated. In addition, two recent studies report inhibition of Ca_v channels by STIM in T cells, suggesting a potential reciprocal regulation of the ORAI1 and Ca_v channels by STIM that needs to be further investigated [115, 116].

Purinergic P2X receptors are non-selective Ca²⁺ channels activated by extracellular ATP/UTP. During immune responses, the release of ATP by damaged or dying cells is considered to be an important inflammatory “danger” signal. P2X4 and P2X7 strongly influence chemotaxis and pro-inflammatory cytokine production in DCs and macrophages [117, 118]. In T cells and B cells, P2X7-dependent Ca²⁺ influx induces proliferation and cytokine production [119]. Additionally, activation of T cells and neutrophils stimulates P2X7 via autocrine release of ATP [120, 121]. Although, all these findings are supported by in vivo studies, the biophysical characteristics of P2X in immune cells is only partly understood.

Figure Box 2: SOCE and non-SOCE Ca²⁺ mobilization mechanisms in immune cells. In addition to the SOCE induced by released of Ca²⁺ from the ER by the IP₃R, ryanodine receptors (RyR) channels can also release Ca²⁺ from the ER. STIM senses the store depletion and triggers the opening of the Ca²⁺ CRAC channel, ORAI and potentially TRPC1. L-type “voltage-dependent” Ca²⁺ (Ca_v) channels induce voltage-dependent (+) and independent Ca²⁺ mobilization via direct interaction with RyR and are potentially inhibited by STIM. Non-SOCE mechanisms include, purinergic receptor channels (P2X), transient receptor potential vanilloid 1 and 2 (TRPV1/2) channel, TRPC3/6 and TRPM2. ADPR, ADP-ribose; NAADP, nicotinic acid adenine dinucleotide phosphate; PIP₂, phosphoinositides diphosphate; ROS, radical oxygen species; Temp, temperature.

Regulation of cytosolic free Mg²⁺ in immune cells

The homeostatic regulation of cytosolic free Mg²⁺ is less well understood than Ca²⁺. It involves Mg²⁺ transporters, channels and exchangers, such as the Mg²⁺/Na⁺ exchanger (Figure 1b, Table 2). Another major participant is the melastatin transient receptor potential 7 (TRPM7) channel [3]. TRPM7 is a ubiquitously expressed, non-selective Mg²⁺ channel that also conducts Ca²⁺, Zn²⁺ and even Na⁺ in absence of divalent cations [3, 20, 21]. TRPM7 has a serine/threonine kinase domain in its cytosolic tail [22]. Whereas the kinase domain is not essential, its activity can modulate the gating of TRPM7 [22]. The critical role of TRPM7 for Mg²⁺ homeostasis in immune cells is illustrated by decreased free Mg²⁺ and cell cycle arrest in TRPM7 deficient B cell lines and impaired T cell development in T cell-specific TRPM7 conditional knockout (KO) mice [22–24]. Other regulators of free Mg²⁺ homeostasis in immune cells include members of the solute carrier (SLC) family SLC41A1 and SLC41A2. These transporters were identified by homology to the bacterial Mg²⁺ transporter, MgtE, and are expressed relatively ubiquitously [20, 25]. The role of SLC41A1/2 in free Mg²⁺ homeostasis was uncovered by ectopic expression in TRPM7-deficient B cell line that can rescue the decreased intracellular Mg²⁺ levels and defective proliferation [26, 27].

Finally, MAGT1 is a key Mg²⁺ transporter in immune signaling [30, 31]. Unlike TRPM7 and SLC41A1/2, MAGT1 is a uniquely Mg²⁺-selective channel that is ubiquitously expressed but with higher levels in immune and epithelial cells [28, 29]. Since the early 90s, elevation of intracellular free Mg²⁺ has been observed after stimulation of T cells with lectins in some studies, but not others [5, 30, 31]. The functional importance of Mg²⁺ mobilization remained unknown for twenty years until the discovery of a new primary immunodeficiency named "X-linked immunodeficiency with Mg²⁺ defect, Epstein Barr Virus (EBV) infection, and neoplasia" (XMEN) disease [7, 32, 33]. This disease is due to a genetic deficiency of MAGT1, which we found to be essential for a TCR-gated Mg²⁺ flux (see divalent cations in pathophysiology section) [7]. The MAGT1-dependent Mg²⁺ flux is required for the optimal activation of PLC-γ1, IP₃ generation, protein kinase C theta (PKC θ) phosphorylation, and Ca²⁺ mobilization via SOCE (Figure 2b) [7]. The molecular mechanism of TCR gating of MAGT1 is currently unknown. MAGT1 deficiency also leads to decreased cytosolic free Mg²⁺ and decreased Mg²⁺ uptake in T cells and B cells [7, 32, 34]. Moreover, oral Mg²⁺ supplementation restores the level of intracellular free Mg²⁺ in XMEN patients [32]. The channel or transporter that restores Mg²⁺ homeostasis in immune cells after supplementation remains under investigation. Conversely, the Mg²⁺ flux cannot be restored in XMEN T cells by high external Mg²⁺ and apparently requires the presence of the MAGT1 protein [32]. Recently, a study using drosophila MAGT1 expressed in human neuroblastoma SH-SY5Y cell line showed MAGT1-dependent currents induced by the protein kinase C (PKC) inducer phorbol-12-myristate-13-acetate (PMA) via phosphorylation by PKC [35]. The role of this interesting observation in immunity has yet to be established. In B cells, the phosphorylation of serine 1164 of PLC-γ2 by the kinase domain of TRPM7 is dependent on extracellular Mg²⁺ and participates in Ca²⁺ mobilization (via SOCE) in response to BCR stimulation [36]. However, in this study, the authors did not assess Mg²⁺ mobilization in response to BCR stimulation and there is no evidence of Mg²⁺ mobilization by B cell activation (Figure 2c) [36].

Figure 2. Divalent cations effectors and functions in immune cells.

(a) A broad range of functions can be rapidly deployed through Ca²⁺ signals. One major Ca²⁺ function is the regulation of gene expression involved in cell activation, differentiation and cytokine production. Ca²⁺ controls the activation of NFAT via CaM-CN, CREB and myocyte enhancer factor 2 (MEF2) via CaMK and CaM and NF-κB via PKC and CaMKII. Ca²⁺ controls exocytosis through Ca²⁺-sensitive proteins such as synaptotagmins (Syt) and Munc13. S100 proteins are Ca²⁺-binding proteins involved in the cytoskeleton organization during migration, lysosome-phagosome fusion during phagocytosis. In neutrophils, Ca²⁺ controls ROS production by NOX2 via PKC and S100 proteins (S100A8/A9). (b) In T and B cells, Mg²⁺ is regulating Ca²⁺ mobilization via SOCE, by modulating PLC-γ activation. In T cells, the TCR-induced MAGT1-dependent Mg²⁺ flux is required to activate PLC-γ1. In B cell, TRPM7 kinase domain participates in PLC-γ2 activation together with Bruton’s tyrosine kinase (BTK) in a Mg²⁺-dependent manner. (c) Zn²⁺ is involved in the recruitment of kinases such as PKC or LCK. Zn²⁺ enhances MAPK activation through inhibition of the MAPKP. Zn²⁺ controls gene expression through NF-κB and through inhibition of CN (which promotes CREB activation). In innate immune cells, Zn²⁺ binding-proteins such as MT and S100 proteins (S100A8/A9) are playing anti-microbial roles by chelating Zn²⁺. Zn²⁺ inhibits IL-6 signaling by inhibiting the activation of STAT3. Finally, Zn²⁺ regulates the cAMP pathway through the inhibition of AC or PDE.

Different waves of Zn²⁺ mobilization

Cytosolic free Zn²⁺ is maintained at low levels by tight regulation of the expression and subcellular localization of two families of Zn²⁺ transporters controlling cytosolic import and export. The zinc transporter (ZnT, SLC30) family consists of 10 transporters that reduce cytosolic Zn²⁺ levels, whereas the fourteen importers of the Zrt1-, Irt1-like Protein (ZIP, SLC39) family increase it (Figure 1c) [37, 38]. Zn²⁺ homeostasis is also regulated by intracellular Zn²⁺-binding proteins, including metallothioneins (MT) that can sequester up to 15% of the cellular Zn²⁺ and release it during oxidative stress [39]. The role of MT in controlling Zn²⁺ homeostasis appears to be important in innate immune cells. In macrophages, MT deficiency impairs cytokine production and bactericidal activity after lipopolysaccharide (LPS) stimulation due to Zn²⁺-deficiency [40, 41].

Two types of Zn²⁺ mobilization have been described with different timing. First, in T cells, monocytes, neutrophils and mast cells, fast transcription-independent Zn²⁺ mobilization has been observed within minutes after the stimulation of activating receptors using fluorescent Zn²⁺-sensitive probes [42]. How these Zn²⁺ fluxes are triggered is mostly unknown. In human T cells, co-incubation with IL-2 or antigen-loaded dendritic cells (DCs) induces a rapid Zn²⁺ elevation that requires ZIP6 on the PM and ZIP8 on the lysosomal membrane, respectively [43–45]. In monocytes/macrophages, LPS stimulation induces quick release of Zn²⁺ from intracellular stores [46, 47]. The intracellular stores and the role of MT involved in Zn²⁺ release remain to be determined. In mast cells, FcεRI stimulation induces Zn²⁺ release from the ER termed “Zn²⁺ waves”. The mechanisms governing these “Zn²⁺ waves” are not well understood and their biological function remains unclear [48]. The second type of Zn²⁺ mobilization occurs several hours after various stimuli and requires changes in channel expression. For example, activation of T and B cells leads to a sustained cytosolic free Zn²⁺ elevation due to the down-regulation of several Zn²⁺ exporters (ZnT1, ZnT4, ZnT5, ZnT6 and ZnT7) and the up-regulation of several importers (ZIP6, ZIP10 and ZIP8) over long time course [43, 49]. Silencing ZIP6 and ZIP8 in human T cells prevents the elevation of cytosolic free Zn²⁺ and impairs proliferation and cytokine production during T cell activation [43–45]. The physiological role of these transporters in B cells is unknown. On the other hand, stimulation of DCs with LPS decreases intracellular Zn²⁺ over time via the down-regulation of ZIP6 and ZIP10 and the up-regulation of ZnT1, ZnT4 and ZnT6 [50]. Importantly, reduced cytosolic free Zn²⁺ is essential for major histocompatibility complex class II (MHC II) up-regulation as well as DC maturation and phagocytic function [50]. The precise molecular mechanism of these effects is unknown and it is not clear whether this is modulation of a cofactor or a signaling function per se.

Over the past few years, investigation of Ca²⁺, Mg²⁺ and Zn²⁺ signaling has provided many new insights in their regulatory roles in immune cells. The identification of ORAI as the CRAC channel provided the long-sought molecular validation of SOCE. However, there are still unanswered questions about the respective roles of STIM and ORAI isoforms in immune cells. Additionally, non-SOCE mechanisms may be important in immune functions even if we don’t fully understand how they work at this time (Box 2). Recently, the discovery of inducible Mg²⁺ and Zn²⁺ fluxes in immune cells has revealed the signaling function of these cations but has raised many new questions about their regulation. For example, in T cells, the gating mechanisms and compartmentalization of MAGT1 (Mg²⁺) and ZIP6 (Zn²⁺) requires further investigation in immune cells. Furthermore, new specific chelators and fluorescent probes must be developed to study Mg²⁺ and Zn²⁺ mobilization with the clarity and elegance achieved in the Ca²⁺ field. Mobilizing divalent cations is only the first step for ion signaling, in the next section we discuss how changes in the concentration of cytosolic free Ca²⁺, Mg²⁺ and Zn²⁺ control immune signaling pathways and effector functions.

Divalent cations functions in immune cells

Ca²⁺, a central second messenger

Ca²⁺ signaling triggers cell activation and differentiation principally through gene expression and the rapid deployment of effector functions (Figure 2a). Ca²⁺ controls the activity of several transcription factors. It stimulates NFAT via the Ca²⁺-sensitive protein calmodulin (CaM)-dependent activation of the phosphatase calcineurin (CN) (Figure 2a). Ca²⁺ also activates the cAMP response element-binding protein (CREB) and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) through CaM kinases (CaMK) and PKC (Figure 2a) [51]. For cytolytic effector functions, Ca²⁺ plays an important role in immune exocytosis. In mast cells, NK cells and cytotoxic T cells (CTL), increased free Ca²⁺ is required for degranulation of following FcεRI and FcγRIII (CD16) and TCR stimulation, respectively [13, 52, 53]. Similarly, the up-regulation of MHC II molecules and synaptotagmin VII, a Ca²⁺ sensor involved in vesicle trafficking, during DC maturation depends on Ca²⁺ mobilization [54, 55]. Ca²⁺ chelation studies in mouse and human phagocytes have established that phagocytosis is mostly Ca²⁺-independent [56]. However, the Ca²⁺ requirement for the fusion of the phagosome with the lysosome is still debated [56, 57]. Moreover, in neutrophils, ROS production associated with antimicrobial functions relies on the Ca²⁺-dependent activation of NADPH oxidase 2 (NOX2). The mechanism of activation of NOX2 is not well defined, but apparently involves PKC, phosphatidylinositide 3-kinases (PI3K) and sphingosine kinase (SK) and the Ca²⁺/Zn²⁺-binding protein calprotectin (S100A8/A9) [56, 58, 59]. Thus, a broad range of immune functions can be rapidly deployed through the extraordinary speed of free Ca²⁺ moving inside and throughout the cell.

Emerging role for Mg²⁺ in immune cells

Beyond the essential role of Mg²⁺ as a cofactor for ATP, enzymes and other processes in all cells, special signaling functions have been recently documented in immune cells. Early studies documented that Mg²⁺ is important for T and B cell proliferation, but it was unclear whether this was due to a cofactor role, a signaling role, or both [7, 27, 60, 61]. As mention earlier, TRPM7-deficient B cell lines exhibit reduced cytosolic free Mg²⁺ and defective proliferation, which is restored by increasing extracellular Mg²⁺ and/or overexpressing other Mg²⁺ transporters [27, 61]. Whether TRPM7 deficiency affects total, mainly bound, Mg²⁺ in B cell lines was not determined [22]. Both TRPM7 and MAGT1 appear to play a role in T cell differentiation [3, 7]. Germline TRPM7 KO mice are embryonic lethal, whereas T cell-specific TRPM7 KO mice are viable but exhibit a blockade of thymic T cell development at the double negative stage [24]. XMEN patients (MAGT1-deficient), have reduced numbers of CD4 T cells especially the naïve CD4⁺/CD45RO⁻/CD27⁺/CD31⁺ cells, suggesting diminished thymic output possibly due to impaired positive selection [7, 33]. As mentioned previously, in mature T cells, TCR stimulation induces a MAGT1-dependent Mg²⁺ flux required for optimal activation of PLC-γ1 and Ca²⁺ mobilization via SOCE [7]. There appears to be no such requirement for B cell activation through BCR signaling [7].

Interestingly, we found that decreased cytosolic free Mg²⁺ in MAGT1 deficiency causes the loss of expression of NKG2D, an essential cytotoxicity activating receptor on NK and CD8+ T cells [32]. NKG2D recognizes specific nonclassical MHC class I-like ligands that are upregulated in virally-infected and transformed cells, including EBV-infected B cells [32]. Biochemical analysis in lymphocytes from XMEN patients revealed incompletely glycosylated NKG2D proteins that are retained intracellularly in reduced abundance [32]. It is not clear whether the failure of N-glycosylation is the cause of NKG2D deficiency. However, MAGT1 shares 60% amino acid identity with TUSC3, an ortholog of a subunit of the yeast N-glycosylation enzyme complex, raising the question of whether MAGT1 may participate in these processes [20, 29]. Strikingly, exogenous Mg²⁺ supplementation restores the level of cytosolic free Mg²⁺ and the expression of fully glycosylated NKG2D, independently of MAGT1, indicating that the full maturation and glycosylation of NKG2D is strictly dependent on the level of cellular free Mg²⁺ [32].

Zn²⁺ more than a trace element in immune cells

Dietary Zn²⁺ deficiency decreases T cell activation, shifts T_H1 responses toward T_H2 responses, reduces the cytotoxic function of NK and NKT cells, and impairs cytokine production by mast cells and NK cells [62]. Interestingly, Zn²⁺ supplementation restores those functions, but also induces chronic inflammatory responses and suppresses T_H17 development in the elderly [63–65]. T_H17 cell suppression appears to be caused by the direct Zn²⁺-dependent inhibition of IL-6-induced phosphorylation of signal transducer and activator of transcription 3 (STAT3) by janus kinases (JAKs) [64]. At the cellular level, several studies using the metal chelator tetrakis-(2-Pyridylmethyl)ethylenediamine (TPEN) and the Zn²⁺ ionophore pyrithione established that Zn²⁺ mobilization regulates specific signaling pathways [66, 67]. Those pathways involve tyrosine kinases, PKCs, mitogen-activated protein kinases (MAPK), cAMP and JAK/STAT. In T cells, the role of Zn²⁺ in leukocyte C-terminal Src kinase (LCK) activation is multiple: it facilitates recruitment to the cytoplasmic tail of CD4 and CD8, stabilizes homodimerization of the enzyme, and enhances its activation by reducing the recruitment of the inhibitory phosphatase Src homology region 2 domain-containing phosphatase-1 (SHP-1) to the TCR complex [66, 67]. Reduced SHP-1 recruitment has been directly tied to Zn²⁺ mobilization in T cells, but whether Zn²⁺ plays a structural or signaling role in LCK activation is unclear [44]. In mast cells, PKC recruitment to the PM induced by FcεRI stimulation depends on Zn²⁺ mobilization and ZnT5 expression, but the mechanism has not been defined [68]. Zn²⁺ enhances mitogen-activated protein kinases (MAPK) used by phagocytes to eliminate pathogens by inhibiting MAPK phosphatases (MAPKP) [45, 46]. Also, cAMP and cyclic guanosine monophosphate (cGMP) are important second messengers produced by adenylate cyclase (AC) and guanylate cyclase (GC), respectively; and degraded by phosphodiesterases (PDE). In monocytes, the levels of cytosolic free Zn²⁺ modulates the cAMP/cGMP pathway by two different mechanisms. First, Zn²⁺ promotes cGMP accumulation by limiting the production of cAMP via direct inhibition of AC [69, 70]. This blocks monocyte differentiation without a requisite decrease of intracellular Zn²⁺ [69, 70]. Second, high free Zn²⁺ concentrations (in the low nM range) promote cAMP and cGMP accumulation by inhibiting PDEs [69, 70]. During T cell activation, inhibition of CN by Zn²⁺ released from the lysosomes promotes the activation of NF-κB and IFNγ production [43]. In IL-27-induced T_reg cells, IL-10 production requires Zn²⁺ and is inhibited by MT [71]. Lastly, Zn²⁺ sequestration by calprotectin (S100A8/A9), expressed in monocytes and neutrophils is important to suppress Candida albicans growth [72]. Thus, Zn²⁺ fluctuations affect many immunological functions.

Whereas many immune functions controlled by Ca²⁺ are well established, our knowledge of the effectors of Zn²⁺ and Mg²⁺ is not as well developed. Given their essential roles as structural supports and co-factors in many cell functions, it is important to further characterize the effectors involved in Mg²⁺ and Zn²⁺ signaling. Additionally, Ca²⁺, Mg²⁺ and Zn²⁺ signaling pathways are intertwined and each cation can potentially modulate the signaling of the others.

Agonist and antagonist interactions in divalent cations signaling

Not just acting in parallel, Ca²⁺, Mg²⁺ and Zn²⁺ signaling pathways can intersect by direct competition for the same effectors or by cross-modulation. Classically, Mg²⁺ is considered to be a Ca²⁺ antagonist. This concept stems from their physico-chemical properties as alkali metals carrying divalent positive charges. Ca²⁺ has a larger atomic radius, but Mg²⁺ has a higher affinity for electronegative oxygen, which leads to a greater hydrated radius (Table 1). These properties allow Mg²⁺ to compete with Ca²⁺ for certain binding sites. However, the propensity of Mg²⁺ to be hydrated prevents the conformational changes associated with Ca²⁺ binding which antagonizes Ca²⁺ signaling [73]. For example, Mg²⁺ competes with Ca²⁺ for the binding to the EF-hand sites of CaM and causes the protein to remain in a closed, inactivated conformation [74]. Mg²⁺ can also antagonize Ca²⁺ signaling by blocking Ca²⁺ mobilization. For example, in vitro studies using HEK cells engineered to express various P2X channels, show that extracellular Mg²⁺ inhibits Ca²⁺ currents induced by ATP [75]. Similar observations were made on endogenous P2X7 channels using the monocyte-like cell line THP1 [76]. Other Ca²⁺ channels were shown to be inhibited by extracellular Mg²⁺, such as TRPV3 in mouse keratinocytes, Ca_v1.2 in mouse myocytes [77, 78].

In T cells, Mg²⁺-dependent activation of PLC-γ1 following TCR stimulation participates in Ca²⁺ mobilization, which defines a new regulatory role for Mg²⁺ signaling in the upstream control of Ca²⁺ signaling [7, 32]. Unlike the direct competition mentioned earlier, this “pro-Ca²⁺ signaling” role of Mg²⁺ does not involve common binding sites, but a convergence of two independent signaling pathways, in which Mg²⁺ increases PLC-γ1 activation leading to a more robust generation of IP₃. This example of coordinated Mg²⁺ and Ca²⁺ signaling opens the door to further investigations and identification of the Mg²⁺ effector(s) that control PLC-γ1 activation [7]. In B cells, the phosphorylation of PLC-γ2 in a Mg²⁺-dependent manner by the kinase domain of TRPM7 might illustrate another convergence of Mg²⁺ and Ca²⁺ signaling [36].

Similar to Mg²⁺, Zn²⁺ can inhibit Ca²⁺ signaling via inhibition of CN or S100 proteins [43, 79, 80]. However, unlike Mg²⁺, Zn²⁺ does not compete with Ca²⁺ for the same binding site, but rather binds different sites on the same protein. For example, the inhibitory effect of Zn²⁺ on CaM does involves sites other than the Ca²⁺-binding EF-hand sites [74]. Zn²⁺ can also promote Ca²⁺ signaling by enhancing the recruitment of Ca²⁺-sensitive effectors, such as the recruitment of PKC to the membrane where the localized Ca²⁺ mobilization can induce activation [66, 68]. Lastly, some proteins have both Ca²⁺-dependent and Zn²⁺-dependent functions involving different binding sites, such as calprotectin (S100A8/A9 dimers) which is a Ca²⁺ effector for the phagocytosis as well as a Zn²⁺-binding protein involved in anti-bacterial function of monocytes [79, 80]. Conversely, Ca²⁺ signaling can promote Zn²⁺ mobilization in mast cells, wherein Zn²⁺ waves elicited in response to FcεRI stimulation are dependent on Ca²⁺ mobilization [48]. Moreover, Ca²⁺ and Zn²⁺ signaling coordinately regulate complex cellular processes in ways that are not completely understood. For example, during DC maturation, both Ca²⁺ elevation and Zn²⁺ diminution are required for up-regulation of MHC II and cytokine production [50, 54, 55].

Divalent cations in physiopathology

The study of human primary immunodeficiencies (PIDs) often yield novel insights into how the immune system is regulated. The understanding of the signaling roles of divalent cations is no exception. For example, the CRAC channel was identified by screening patients with severe combined immunodeficiency (SCID) associated with loss of SOCE [81]. The case is even more striking for Mg²⁺, because XMEN disease reveals its unexpected signaling functions [7, 32].

Ca²⁺ and SOCE deficiency

In humans, deleterious mutations in ORAI1 and STIM1 cause a loss of SOCE in immune cells, SCID and muscle abnormalities [81]. Both ORAI1- and STIM1-deficient patients exhibit recurrent life-threatening viral, bacterial and fungal infections, and require hematopoietic stem cell transplantation (HSCT) early in life. ORAI1 and STIM1 deficiency do not alter the numbers of innate, B or T cells, except for a reduction in the T_reg population. However, defective SOCE severely compromises mature T cell activation, proliferation, and cytokine production, although it has a much milder effect on B cell function. Other features of SOCE-related immunodeficiency are lymphoproliferative disorders and autoimmunity. More comprehensive description of human SOCE deficiency is provided in [14].

Mg²⁺deficiency in immunodeficiency, autoimmunity and inflammation

In XMEN disease, “loss of function” mutations in MAGT1 result in essentially complete loss of the MAGT1 mRNA and protein [7, 33]. For unknown reasons, XMEN patients have modestly decreased numbers of CD4⁺ T cells and inversion of the CD4/CD ratio [7, 33]. As mentioned above, MAGT1 affects Mg²⁺ in two key respects : (1) the loss of a T cell receptor (TCR)-induced transient Mg²⁺ influx required for optimal T cell activation and (2) a chronic decrease of cytosolic free Mg²⁺ [7, 32]. The latter defect is almost entirely responsible for the deficient expression of the NKG2D receptor on NK and CD8 T cells which is an important cytolysis activating receptor against virally-infected or transformed cells (see Emerging role for Mg²⁺ in immune cells section) [7, 32]. The most severe and life-threatening complication of XMEN disease is uncontrolled EBV infection with increased susceptibility to EBV-driven lymphoproliferative disorders [7, 32, 33]. The fact that EBV induces the specific ligands for NKG2D might account for the severity of EBV infection when NKG2D is deficient [32]. Since increasing the cytosolic free Mg²⁺ to normal levels rescues the cell surface expression of NKG2D, Mg²⁺ supplementation may be an effective therapeutic strategy for XMEN patients and could be important for immune suppression of EBV and avoiding EBV-induced lymphoma [32].

In rodents, Mg²⁺ deficiency induced by low Mg²⁺ diet leads to decreased antibody production, thymic involution and chronic inflammation [82, 83]. Similarly, in humans, Mg²⁺ deficiency promotes a chronic inflammatory state that is associated with metabolic syndrome and type 2 diabetes mellitus (T2DM) [84]. Mg²⁺ supplementation inhibits IL-1β, TNFα and IL-6 secretion by mouse macrophages in vitro and in vivo [85, 86]. This latter observation fits with the concept of Mg²⁺ - Ca²⁺ antagonism, because reduced extracellular Mg²⁺ sensitizes immune cells to Ca²⁺ signaling leading to chronic inflammation. The precise relationship between the blood level (magnesemia) and the cytosolic level of Mg²⁺ is not well understood however free basal Mg²⁺ equilibrates more effectively with external Mg²⁺ then with the large amounts of intracellular bound Mg²⁺ [87, 88]. Thus, the full immunological impact of hypomagnesemia may be greater than we currently understand. Importantly, more than 60% of the US population does not meet the estimated daily requirement for Mg²⁺. The impact of the Mg²⁺ dietary deficiency on cytosolic free Mg²⁺, susceptibility to viral infections, and long-term health requires further investigation [89, 90].

Zn²⁺ deficiency and aging

A vital role for Zn²⁺ in immunity has been known for over 50 years, mostly through studies of dietary Zn²⁺ deficiency. The loss of expression of ZIP4 in acrodermatitis enteropathica (AE), a Zn²⁺ malabsorption syndrome, leads to immunodeficiency associated with lymphopenia and thymus atrophy [91]. Additionally, leukocytes from asthmatic children and pulmonary tuberculosis patients over-express ZIP2, and monocytes from septic patients up-regulate ZIP8 suggesting there is abnormal cytosolic free Zn²⁺ in these pathological settings [92–94].

As mentioned earlier, the study of PID has clarified signaling roles for Ca²⁺, Mg²⁺ and Zn²⁺. Nonetheless, these conditions also raise numerous questions about the cellular homeostasis of these ions. Mg²⁺ and Zn²⁺ deficiency can be compensated to a certain extent by external supplementation, but the mechanisms remains unknown. Unlike Zn²⁺ deficiency, Mg²⁺ deficiency can be masked by the lack of correlation between serum and cellular levels of Mg²⁺, which could mask the potential role of Mg²⁺ in pathological settings.

Concluding remarks: ion signaling as therapeutic target

Recent advances in understanding the role of divalent cations in immune cells have uncovered new signaling functions for Mg²⁺ and Zn²⁺ and have unveiled new vistas for future investigation. Many questions about Mg²⁺ and Zn²⁺ signaling are still unanswered. In fact, there are a plethora of apparent Mg²⁺ transport and channel proteins whose role in human physiology has not yet been determined [20]. For example, amongst the twenty-two Mg²⁺ channels known, sixteen are expressed in T cells, but except the few mentioned is this review, little is known about their potential role in immunity. Moreover, Mg²⁺ and Zn²⁺ mobilization has been observed in a variety of cell types, however the molecular mechanisms regulating these cation changes and the molecular consequences remain elusive [95]. Technical limitations such as probe sensitivity, low electrogenicity, and other considerations prevent the precise study of the regulated mobilization of these cations. Another aspect that begs additional experimentation is the understanding of the various effectors of divalent cations signaling and the network(s) that they generate. Beyond enhancing our fundamental scientific understanding, answering these questions has important medical implications especially, but not only, for immune disorders associated with divalent cations signaling deficiency, such as XMEN disease, SOCE deficiency or even dietary Zn²⁺ deficiency. Thirty years ago, the discovery of the immunosuppressant abilities of the Ca²⁺ signaling inhibitors such as the CN inhibitor cyclosporine A, revolutionized organ transplantation [96]. Thus, uncovering new Mg²⁺ and Zn²⁺ sensitive targets or functions could identify new therapeutic targets for immunomodulation or affecting the pathophysiology of other diseases. Moreover, the promising results of Mg²⁺ and Zn²⁺ supplementation in various pathological settings should encourage the development of new ion supplementation strategies [32, 86, 97–99].

Highlights.

• Ca²⁺, Mg²⁺ and Zn²⁺ are signaling cations with interconnected signal networks

• Cation signaling is essential for immunity and defects underlie immune diseases

• Mechanisms of Mg²⁺ and Zn²⁺ mobilization and effectors require further definition

• Modulation of ion signaling may lead to novel therapies for immune diseases