TRPM7, the cytoskeleton and neuronal death

Abstract

Ischemic stroke is one of the leading causes of disability and death in the world. Elucidation of the underlying mechanisms associated with neuronal death during this detrimental process has been of significant interest in the field of research. One principle component vital to the maintenance of cellular integrity is the cytoskeleton. Studies suggest that abnormalities at the level of this fundamental structure are directly linked to adverse effects on cellular well-being, including cell death. In recent years, evidence has also emerged regarding an imperative role for the transient receptor potential (TRP) family member TRPM7 in the mediation of excitotoxic-independent neuronal demise. In this review, we will elaborate on the current knowledge and unique properties associated with the functioning of this structure. In addition, we will deliberate the involvement of distinct mechanistic pathways during TRPM7-dependent cell death, including modifications at the level of the cytoskeleton.

Introduction

For a fundamental process that is an inevitable part of the day-to-day function of living organisms, the precise classification of cell death has been the subject of much debate.¹ Initial strategies relied principally on morphological characteristics (such as cell rounding, alterations in cellular and organelle volume, and nuclear fragmentation), largely due to limitations in the availability of assays employing suitable biochemical markers.² Recently, the Nomenclature Committee on Cell Death (NCCD) released an updated set of guidelines that incorporates current advancements in the literature as well as modern developments in the use of biochemical tests such as the monitoring of caspase and reactive oxygen species activity.³

In recent years, evidence has emerged implicating dysfunction of the cellular cytoskeleton in a variety of neurodegenerative diseases including amyotrophic lateral sclerosis, Alzheimer disease and Parkinson disease.⁴ The cytoskeleton is composed of filamentous organizations of microfilaments (or actin filaments), intermediate filaments (which can comprise of one of several proteins including vimentin, keratin, lamin or neurofilaments) and cylindrical microtubules (consisting of α and β-tubulin subunits), as well as multiple regulatory binding partners. The cytoskeleton has been implicated in a variety of critical cellular processes, including maintenance of basic architecture and integrity, organelle and vesicular trafficking, motility and cell division.^5-12 Hence, this protein-based framework is vital to the preservation of cellular well-being, playing a key role in growth, differentiation and proliferation. More recently, cytoskeletal proteins have been shown to be active participants in signal transduction and signal regulation downstream of a variety of receptors and channels.^5-12 Therefore, it is no surprise that abnormalities at the level of the cytoskeleton have been implicated in certain forms of cell death.^5,13-18 Indeed, evidence suggests that cytoskeletal integrity is intrinsically associated with key mediators of cell death including intracellular calcium (Ca²⁺) regulation.^19-21 Ca²⁺ is a vital second messenger for the propagation of information and has been a mainstay in theories explaining cytotoxicity.^22-28 The abnormal build-up of cellular Ca²⁺ has also been demonstrated as a major factor during the initiation of ischemic death.^22-24

Research into the molecular mechanisms of brain neuronal death has been a source of promise and disappointment in stroke research over the last two decades.^29-33 Several key players including NMDA (N-Methyl-D-aspartate) and AMPA (α-amino-3-hydroxy-5-methyl-4- isoxazole-propionic acid) glutamatergic receptors have been identified with the spotlight given to the role of Ca²⁺ influx in neuronal death.^29,34-40 Unfortunately, many of these studies have concentrated on the individual effects of neuronal channels and proteins, rather than examining the cell death process as whole. To date, neuronal cell demise remains poorly defined and viable therapeutic strategies for treating neurodegeneration following ischemic brain injury are still lacking. This review takes a new look at neuronal cell death including roles for both excitotoxic and non-excitotoxic processes (see Table 1). We will highlight recent insights into the function of the TRPM7 cation channel which has been implicated in ischemia-related neuronal death,^41-47 and elaborate on the multiple mechanistic pathways that TRPM7 may utilize in contributing to the cell death process, including regulation of the cellular cytoskeleton.

Table 1. Prominent structures implicated in the mediation of excitotoxic and non-excitotoxic neuronal death.

Excitotoxic components involved in ischemic neuronal death
Name	Known regulators	Key findings
NMDA receptors	Glutamate	NMDA receptor antagonists block glutamate neurotoxicity (Choi, 1987)
AMPA receptors	Glutamate	AMPA receptor antagonists diminished loss of ischemic neuronal demise (Buchan et al., 1991)
EAATs	Sodium electrochemical gradient	Antisense oligonucleotide-mediated disruption of glutamate transporter synthesis produced enhanced glutamate levels and neurodegeneration (Rothstein et al., 1996)
Non-excitotoxic structures involved in ischemic neuronal death
Name	Known regulators	Key findings
ASICs	Acidosis	Cerebroventricular injections of ASIC inhibitors or ASIC knockout provided a protective effect on neurons from ischemic insult (Xiong et al., 2004)
Gap junction channels	Calcium and pH levels	Gap junction hemichannels expressed in neurons were activated following OGD (Thompson et al., 2006)
TRP channels	TRPM2: Hydrogen peroxide TRPM7: See Table 2	TRPM2: Hydrogen peroxide induced neuronal death in cultures was averted in the presence of TRPM2-siRNA (Kaneko et al., 2006) TRPM7: See Table 2

Abbreviations: NMDA, N-Methyl-D-aspartate; AMPA, α-amino-3-hydroxy-5-methyl-4- isoxazole-propionic acid; EAATs, excitatory amino acid transporters; ASICs, acid-sensing ion channels; OGD, oxygen glucose deprivation; TRPM2, transient receptor potential melastatin 2; TRPM7, transient receptor potential melastatin 7; siRNA, small interfering RNA

Understanding Excitotoxicity: The Role of Glutamatergic Receptors

To better appreciate TRPM7 activity and the ischemic process as a whole, it is necessary to first discuss the phenomenon of excitotoxicity. Glutamate is the major excitatory neurotransmitter in the CNS and plays a fundamental role in critical processes from neuronal development and survival to sensory perception and the ability to learn new tasks.⁴⁸ Glutamate signaling via ionotropic receptors, in particular, the NMDA and AMPA subtypes, have received considerable attention for their roles in synaptic function.^49-51 Ca²⁺ influx via these structures can instigate numerous kinase signaling pathways, including protein kinase C (PKC) and calcium/calmodulin-dependent kinase II (CaMKII).⁴⁹ These cascades act to stimulate overall AMPA receptor activity by means of phosphorylation and fortifying their relative numbers at the postsynaptic membrane, bolstering synaptic function and plasticity.^25,26 In pathological situations, inordinate levels of glutamate release can exceedingly stimulate both NMDA and AMPA receptors to allow an abnormally high level of Ca²⁺ ion influx. This elevated ion flow results in the activation of several enzyme cascades such as phospholipases, endonucleases and proteases that are detrimental to cellular well-being in a process that is referred to as “excitotoxicity,” where glutamate acts as an “excitotoxin.”^25-28 The ensuing injury to critical structures such as the membrane and cytoskeleton can finally lead to cellular death. Another possible mechanism for excitotoxicity may result from abnormalities at the level of excitatory amino-acid transporters (EAATs). These structures play an essential role in maintaining extracellular concentrations of glutamate within normal parameters via membrane transport to prevent detrimental damage as a consequence of receptor overstimulation.⁵² Disruption of EAAT production has been shown to induce excitotoxic degeneration under both in vivo and in vitro settings.⁵³ Excitotoxicity has been demonstrated to occur in acute neuronal damage during stroke and brain or spinal cord trauma, as well as neurodegenerative ailments such as multiple sclerosis, amyotrophic lateral sclerosis, Alzheimer disease, Parkinson disease and Huntington's disease.^26-30

Non-Excitotoxic Cell Death

The association of excitotoxicity with pathological situations led to the logical supposition of targeting glutamatergic receptors as a therapeutic countermeasure during ischemia. Unfortunately, anti-excitotoxic therapy (AET) has not lived up to expectations, with the presence of serious side-effects (including increased early mortality rates) and other detrimental consequences that result from blocking physiological NMDA receptor activity.^54-57 Therefore, researchers sought to explore glutamate-independent mechanisms as alternative therapeutic targets (see Table 1). Ongoing research shows that Ca²⁺ signaling plays a central role in neuronal death in the absence of excitotoxic sources. Examination of the underlying mechanisms associated with ischemic-related acidosis revealed the activation of Ca²⁺ permeable acid-sensing ion channels (ASICs).^58-60 Inhibiting the function of these ASICs via pharmacological or genetic manipulations resulted in robust protection of neurons from ischemic damage. Additionally, the exploration of intercellular gap junction hemichannels during anoxic processes has also yielded intriguing results.^61-64 In interneurons, these channels are aggregates of proteins named connexins that permit communication between opposing cell membranes. Knockout mice specifically lacking connexin 32 demonstrated an increased susceptibility to induced ischemia in comparison to wild-type animals.⁶¹ Moreover, neuronally expressed gap junction channels comprising of pannexin proteins were shown to open in response to oxygen glucose deprivation (OGD), a well known model of experimental ischemia.⁶³ Other primary candidates for the mediation of non-excitotoxic mechanisms of neuronal death have been members of the transient receptor potential (TRP) family of channels.^41,65-67 Henceforth, this review will focus on the discussion of recent developments and potential mechanisms associated with the actions of one particular member of this family, i.e., TRPM7.

Transient Receptor Potential (TRP) Channels and TRPM7

In comparison to other families of ionic channels, TRP channels stand out in respect to their varied modes of activation and diverse functionalities.^68-74 The need for activity-inducing regulatory mechanisms vary greatly among different TRP channels, ranging from not being required to the specific necessity of factors such as calcium concentration state, pH levels, oxygenation status, osmolarity and mechanosensitivity. To date, around 33 subunits belonging to this superfamily have been identified that have been further classified into two groups subdivided into seven smaller families relative to homology. These include TRPC (canonical), TRPM (melastatin), TRPV (vanilloid), TRPA (ankyrin), TRPML (mucolipin), TRPN (no mechanoreceptor potential C) and TRPP (polycystin). In contrast to some TRP members (TRPA, TRPC, TRPN and TRPV), TRPM channels do not possess numerous N-terminal ankyrin repeats, instead comprising of four analogous domains termed the TRPM homology regions. These regions have been speculated to be involved in channel formation and membrane transport mediation.⁷⁵

TRPM7 is one of eight members of the TRPM subfamily that was first discovered over a decade ago. This structure has been uniquely coined a “chanzyme,” being distinctive among proteins of its class (including most belonging to the same superfamily) in that it possesses both cationic ion channel activity as well as an inherent protein kinase domain.⁷⁶ Interestingly, TRPM7 was first identified and cloned in 2001 by multiple studies under diverse settings. One report investigating phospholipase C interacting partners employed yeast two-hybrid screening to initially determine and then clone TRPM7 via polymerase chain reaction (PCR) using mouse brain cDNA.⁷⁷ In exploring novel putative members of the α kinase class of proteins through a combination of cloning and sequencing, another group utilized mouse melanoma and kidney cDNA libraries to ascertain large regions sharing homology with the TRP family that were expressed over a wide range of tissues.⁷⁸ A further study aiming to demonstrate previously unidentified cation channels in hematopoietic cells provided functional evidence of characteristic TRPM7 activity following cloning.⁷⁹ Therefore, the diverse nature of these initial studies only served to demonstrate a taste of the varied functionality that would eventually be linked with TRPM7. To date, these broad range of activities have also encompassed the proliferation of retinoblastoma,⁸⁰ head and neck carcinoma cell lines,⁸¹ and gastric cancer cells.^82,83 TRPM7 subunits comprise of 1863 amino acid residues with an expected molecular weight of around 212kDa, with the mouse gene being over 90% homologous to human TRPM7.

The Distinct Composition of TRPM7

Structurally, the TRPM7 channel is made up of six membrane-traversing domains (a characteristic shared by multiple voltage-dependent channels) along with a pore-forming loop (P-loop) located between the last two regions (see Fig. 1). Channel function was demonstrated to be intrinsically linked with Mg²⁺ and Mg-ATP levels,⁷⁹ and was proposed to be necessary for the homeostasis of divalent cations (including Ca²⁺ and Mg²⁺) within the cell.⁸⁴ In keeping with these key characteristics, TRPM7 currents detected in cell lines were referred to as Magnesium-Nucleotide-regulated Metal (MagNuM) currents.^79,85 Alongside the channel region, the TRPM7 membrane domains are flanked by lengthy intracellular N- and C-terminals that are thought to represent areas of critical function and activities. Evidence from other TRP family members has suggested that the N-terminal may play a role in channel efficacy, where expression of either a curtailed version of this region or isolated N-terminals disrupted normal channel function.^86,87 TRPM7 has also been previously localized to synaptic vesicle membranes of sympathetic neurons, where it can complex with proteins such as synapsin I and synaptotagmin I.^88,89 Furthermore, immunoprecipitation experiments revealed a direct interaction between the TRPM7 N-terminal and snapin,⁸⁸ a protein that is vital for neurotransmitter release through mediation of soluble N-ethylmaleimide-sensitive factor activating protein receptor (SNARE) activity.⁹⁰

Figure 1. The unique structural composition of TRPM7. TRPM7 is composed of six membrane-spanning domains (blue), with a pore-forming loop (P, in green) located between domains 5 and 6. Intracellularly, TRPM7 comprises of substantial terminal areas that are suggested to be important for its function. The N-terminal (N) is proposed to contribute to neurotransmitter release via binding interactions with proteins such as snapin (Snapin 87–326, in gray). Identified structures with putative roles on the C-terminal (C) include the TRP binding domain (TRP, in orange), the coiled-coiled region (C-C, in yellow) and the α kinase domain (Alpha kinase domain, in purple). The probable interplay between these structural areas in relation to TRPM7 activity has been the subject of continued debate and analysis.

With regards to the TRPM7 C-terminal, multiple inherent components have been identified that have been predicted to contribute to cellular activity. Among these, a compact TRP binding domain (or TRP box, an area also observed in the TRPC and TRPV subfamilies) is purported to be located in close proximity or even within the sixth transmembrane domain. Functionally, the TRP box has been contended to act as a mediator that regulates between the open and closed states of the channel based on thermoreceptive studies of TRPM8 in the presence of phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P₂].⁹¹ Distal to the TRP box, a coiled-coil domain is also present on the C-terminal. These are ubiquitous structural motifs made up of several α helices that recur systematically in what is known as a heptad repeat. Examination of the TRPM7 crystal structure at high resolution has revealed a symmetrical four-stranded coiled-coil domain that is likely to be involved in the formation of antiparallel tetrameric assemblies.⁵⁷ The furthest intracellular region recognized on the TRPM7 C-terminal with a highly anticipated level of importance is an atypical α kinase domain. Characteristically, kinases belonging to eukaryotic systems share similar homology in their catalytic domains (as in the case of serine/threonine and tyrosine kinases) and are collectively called conventional protein kinases (CPKs).^92,93 However, a smaller number of distinct kinases have been identified^94-96 that do not possess analogous sequencing with CPKs. Members of this unique grouping include phosphatidylinositol 3-kinase related kinases and α kinases, and hence have been termed atypical protein kinases (APKs). The α kinase subgroup was named after the fact that the targeted residues of these proteins are situated largely within α helices⁹⁷ as opposed to the loopy and irregular structures associated with CPK phosphorylation sites.^93,98,99 Interestingly, irrespective of the sequencing differences observed, a number of similarities between α kinases and CPKs were ascertained during structural alignment investigations. This included the presence of two-lobed folding, uniformity in the conservation of structural components involving subdomains 1–8 and shared positioning of fundamental catalytic residues.⁹³

The TRPM7 α kinase region has been demonstrated by cloning studies in various laboratories.^77-79 Interactions at the level of this domain can include the phospholipase C enzyme group,¹⁰⁰ the motor protein myosin IIA,^101,102 the calcium and phospholipid-associated protein annexin 1,¹⁰³ and the translational determinant eukaryotic elongation factor 2.¹⁰⁴ The most likely candidate to mediate phosphotransferase activity within the kinase domain is Mg²⁺.^105-107 It has been postulated that a “linked uptake and sensor” mechanism may explain this modulation whereby Mg²⁺ flow through an open TRPM7 channel could activate the catalytic activity of its kinase region as required via a regulatory feedback loop.⁶⁹ Initially, it was proposed that TRPM7 kinase activity may be a necessary prerequisite for channel function, where the use of kinase mutants exhibited a decrease in current level and requirement for intracellular adenosine triphosphate (ATP) during whole-cell studies.⁷⁷ However, later investigations contradicted these findings. One report found that while TRPM7 could functionally regulate Mg²⁺ homeostasis to affect channel activity through contribution of the kinase domain, this action was independent of autophosphorylation.¹⁰⁵ These results were echoed in another study where site-specific disruption of either TRPM7 self-phosphorylation or catalytic activity did not significantly modify measurements of Ca²⁺ influx or whole-cell current-voltage relationships.¹⁰⁷ Therefore, the involvement of phosphotransferase activity in the regulation of the TRPM7 channel remains open to debate.

TRPM7 in Neuronal Cell Death

The use of AET was initially thought to be a promising approach to combat the detrimental effects of glutamate-mediated excitotoxicity.^54-57 However, favorable outcomes in relation to the prevention of neuronal cell death through the utilization of glutamate-receptor inhibitors were limited to a 1 h time-frame following the initiation of ischemic insult during experimentation.⁴¹ A widely recognized model to mimic stroke and ischemic conditions has been through the use of OGD.^31,108 This is usually accomplished by subjecting primary neuronal cultures to a temperature regulated and controlled environment via an anaerobic chamber.

We utilized this methodology in order to divulge the causative factor(s) behind the persistence of cell demise in spite of blocked excitotoxicity. Toward achieving this goal, several factors were kept in mind. Primarily, the perseverance of Ca²⁺ signaling during the process despite the presence of MK-801, CNQX and Nimodipine (inhibitors for NMDA, AMPA and L-type calcium channels, respectively) suggested an atypical source of ion flow. Second, patch-clamp recordings under these conditions detected the existence of a cation current whose current-voltage relationship was sensitive to relative Ca²⁺ levels and demonstrated a non-selective cation conductance in response to polyvalent cations such as gadolinium (Gd³⁺) and zinc (Zn²⁺).⁴¹ This current was accordingly termed I_OGD, and demonstrated a dependence on reactive oxygen species (ROS) production and the nitric oxide signaling pathway during antioxidant and nitric oxide synthase (NOS) inhibitor studies respectively. Given the above evidence, the TRP family of proteins were considered as leading contenders for I_OGD mediation, with TRPM7 being at the forefront. Expression of TRPM7 in HEK-293 cells displayed outwardly rectifying current-voltage curves during electrophysiology, a trait also shared by I_OGD. Taking advantage of known TRPM7 modulators PtdIns(4,5)P₂ (activator) and Gd³⁺ (blocker) in addition to small interfering RNA (siRNA) inhibition as experimental tools, the plausible role of this channel in anoxic neuronal death was further investigated in culture systems. Critically, blockage of TRPM7 function abolished key I_OGD-associated attributes, including calcium uptake, generation of ROS and most importantly, cell death.^41,109,110 These results were consistent with previous studies that suggested the involvement of TRPM7 in cell viability during overexpression experiments.^79,84

Another study focused on TRPM7 functionality in the CA1 region of the hippocampus,⁴² an area known to be vital for learning and memory.^49-51 Here, reduction in the levels of extracellular Ca²⁺ and Mg²⁺ presented a cation current in neurons with similar electrical characteristics previously assigned to TRPM7. Boosting channel function in this manner was seen to provoke death in HEK-293 cells, while lowering TRPM7 expression diminished current activity induced by attenuation of divalent cation concentrations.⁴² This data bolstered the notion of TRPM7 acting as a putative sensor of cation levels under certain settings. Subsequently, these findings were extended to examine the possible advantageous effects of in vivo neuronal TRPM7 suppression in the same area.⁴³ The use of viral vector injections bearing TRPM7-specific shRNA into the hippocampus did not alter the normal well-being of adult rodents, with neuronal structure and electrical activity being preserved. Ischemia was subsequently induced in these animals via occlusion of the common carotid and vertebral arteries (also known as 4-vessel occlusion). Following this insult, TRPM7 shRNA possessing cells exhibited an increased resistance to cell death and demonstrated enhanced recovery from the detrimental effects of ischemia in relation to synaptic plasticity, contextual fear conditioning and Morris Water Maze-based spatial navigation.⁴³ Another pathway implicated in neuronal demise is through lipopolysaccharides (LPS). A recent report revealed that either LPS or hydrogen peroxide treatment of differentiated PC12 neurons resulted in an augmentation of TRPM7 expression levels.⁴⁴ Moreover, siRNA inhibition of TRPM7 in primary hippocampal neuronal cultures and PC12 neurons safeguarded these cells against deterioration. Similarly, TRPM7 involvement during oxidant insult was also observed in mouse cortical neurons, where channel inhibition via 2-aminoethoxydiphenyl borate or siRNA offered protection from hydrogen peroxide-mediated damage.⁴⁵

While investigations into channel-mediated Ca²⁺ overload have been at the forefront of cell death research, studies have also suggested that TRPM7 is capable of inflicting injury via homeostasis of other cations. One report demonstrated that Zn²⁺-influenced damage caused by either direct application of the ion or OGD required the actions of TRPM7 in mouse neuronal cultures.⁴⁶ Another group verified that neuronal levels of Mg²⁺ were indeed elevated by means of a TRPM7-dependent mechanism following OGD or chemically-induced ischemia in rat hippocampal cultures.⁴⁷

Though the several studies above have demonstrated a clear role for TRPM7 during ischemic cell death, data concerning the precise signaling pathways engaged in this process has been sparse, including the possible involvement of cytoskeletal components.

TRPM7 and the Cytoskeleton

The sizeable composition of the cellular cytoskeleton along with its inherent capacity to sequester a diverse network of integral components has greatly contributed to its role as a region of cellular enrichment.^5,6,8,10 In addition to the probable functional roles outlined previously, a significant amount of research has been devoted to the possible role of TRPM7 in regulation of the cytoskeleton. Based on earlier work that implicated myosin heavy chain (MHC) α kinases^111-113 in mediating Dictyostelium amoebae cytoskeletal components, one group explored the possibility of the TRPM7 kinase domain performing a similar role in mammals.¹⁰¹ Through utilization of the N1E-115 neuroblastoma cell line, the authors found that expressed TRPM7 promoted cell spreading and cell-matrix adhesion through channel rather than kinase domain activity. Confocal imaging revealed a scattered distribution of the contractility-inducing motor protein myosin IIA and enhanced podosome formation in TRPM7 expressing vs. control cells.¹⁰¹ Podosomes are dynamic actin-enriched sites of adhesion that can interact with the extracellular matrix.¹¹⁴ In this case, podosome development through TRPM7 stimulation via the actomyosin relaxing agent bradykinin was dependent on the kinase region by means of myosin II suppression. The authors also discovered that TRPM7 could directly complex with both β-actin and myosin IIA, and cause phosphorylation of the latter at the COOH-terminus.^101,102 Furthermore, the same group determined that such substrate identification and subsequent phosphorylation was greatly enhanced in the presence of TRPM7 autophosphorylation.¹¹⁵

Another group employed the use of inducible TRPM7 overexpression in HEK-293 cell lines to study its potential ability in regulating members of the calpain protease family during cell adhesion.¹¹⁶ In this experimental set-up, an increase in cellular rounding and loss of adhesive properties were observed within a day. In contrast, a generated cell line with reduced endogenous expression of TRPM7 demonstrated the opposite results, with an enhancement in cell adherence. Through the use of TRPM7 mutants, the above findings were attributed to channel rather than kinase function, with cell rounding decreasing appreciably in cell lines completely lacking or deficient in channel activity.¹¹⁶ Further investigations with short hairpin RNA (shRNA) and the calpain inhibitor ALLM confirmed m-calpain as a key mediator of these TRPM7-dependent morphological alterations.¹¹⁶ M-calpain is a non-lysosomal protease that is responsive to millimolar concentrations of cytosolic Ca²⁺ and has been previously implicated in the destabilization of cellular adhesion.¹¹⁷ Moreover, this TRPM7-associated activation of m-calpain was correlated with the actions of the p38 MAPK (mitogen-activated protein kinase) and JNK (c-Jun N-terminal kinase) signaling cascades in response to stress-related increases in nitric oxide and ROS.¹¹⁸

Recently, the ability of TRPM7 to regulate Mg²⁺ homeostasis has also been suggested to play a possible modulatory role during its cytoskeletal effects.¹¹⁹ Fibroblast cell lines lacking TRPM7 exhibited notable alterations in cell morphology, with modification of focal adhesions, diminished actin fibers and elevated cortical levels of actin and myosin IIA.¹¹⁹ These effects were reversible by adenoviral expression of TRPM7 constructs with intact channel activity (suggesting a lack of functional involvement of the kinase domain during this process). In addition, the presence of a selective Mg²⁺ transporter SLC41A2 (solute carrier family 41 member 2) was also able to rescue the observed deficits and restore cytoskeletal integrity.¹¹⁹ Another intriguing perspective on TRPM7-related cell death may involve its presence in vascular smooth muscle cells (VSMCs). Here, it has been suggested that TRPM7 is intrinsically involved in responding to blood vessel wall damage during shear stress.¹²⁰ The authors of this study propose that shear stress results in a rapid upregulation of TRPM7 activity at the plasma membrane via the actions of cytoskeletal components and motor proteins. Therefore, TRPM7 may also have an effect on cellular viability by means of the vascular network in relation to acute changes in blood flow.

Recent data from our laboratory (Bent et al., publication under review) has also revealed an integral role for TRPM7-dependent cytoskeletal regulation in the mediation of ischemic neuronal cell death. Specifically, OGD-influenced instigation of TRPM7 resulted in the activation of the key actin-binding protein cofilin in rat cortical cultures. Cofilin is affiliated with the ADF/cofilin (actin depolymerization factor) family and has been associated with reorganization of the actin network, a process vital to filament recycling and turnover.^121-124 The utilization of Gd³⁺ and siRNA also indicated that the stimulation of this protein was correlated with a TRPM7-dependent inactivation of LIMK1 (Lin-11, Isl-1 and Mec-3 kinase 1), a known suppressor of cofilin actuation.^125-127 More interestingly, RNA silencing of cofilin significantly reduced neuronal demise following the detrimental effects of OGD, reflecting the importance of this protein in cellular viability under these circumstances. These results offer a rare glimpse into the largely unknown signaling pathways initiated exclusively by TRPM7 activation.

Taken together, the above studies provide compelling evidence for TRPM7-guided governance of cytoskeletal components under a diverse range of experimental settings.

TRPM7-Dependent Ischemic Cell Death: Lingering Questions and Future Directions

The discovery of the unique composition of TRPM7 has been somewhat of a double-edged sword in the analysis of this structure. On one front, it has offered the exciting possibility of exploring distinct and previously unknown mechanisms due to the rare bifunctional nature of this protein.^{77,78,104,106,107} Furthermore, the ubiquitous distribution of TRPM7 has alluded to its possible role in a myriad of processes in various regions of the body. On the other hand, the dual presence of both a channel pore and a kinase domain in close proximity has greatly complicated the delineation of specific functional roles for each region.

While data accumulated from our and other laboratories clearly reflect the significance of TRPM7 activity during pathological situations, it is also likely that this structure may display considerable influence over neuronal physiology. It has been observed that activation or overactivation of NMDA receptors can lead to physiological (induction of synaptic plasticity)^49,50 or pathophysiological (excitotoxicity)^25-28 consequences respectively. In a similar manner, the hospitable bifunctional composition of TRPM7 may also be associated with basic but essential neuronal activities in relation to relevant stimuli or cellular status. In particular, a detailed examination of the possible role of TRPM7 in common forms of synaptic plasticity (such as long-term potentiation and long-term depression)^49,50 and associated behavioral tasks is lacking. One report recently implicated TRPM7 channel activity in sensory neurons to preferentially modulate touch-induced escape in zebrafish, potentially through central synapse neurotransmitter release.¹²⁸ Furthermore, while global knockout mice for TRPM7 have shown to be embryonically lethal,¹²⁹ specific interference of this factor in the neural crest impeded derivation of dorsal root ganglion sensory neurons,¹³⁰ indicating a clear developmental role.

We and others have provided conclusive evidence of an essential mediatory role for TRPM7 during anoxic neuronal death (see Table 2). This has been demonstrated through the use of at least three different experimental models of ischemia (OGD, 4-vessel occlusion and chemically-induced) encompassing both in vitro and in vivo settings, indicating the ubiquitous nature of TRPM7 involvement.^41-47 However, a number of important questions still linger in this regard. While TRPM7 channel activity has proven essential for neuronal demise, a defined role for the kinase region has yet to be determined. This includes the possibilities that the kinase domain may coordinate/facilitate channel function during cell death, or alternatively participate in the process in an independent manner. The resourceful use of various TRPM7 channel and kinase mutants during past studies in cell lines may provide a viable means to approach questions regarding region-specific roles in neurons.^{101,102,115,116,118,119} Another important factor to consider when assessing ischemic-dependent TRPM7 activation is the putative role of environmental pH. It is widely accepted that ischemia and its ensuing cellular damage is associated with a reduction in pH levels.¹³¹ Since a number of studies have linked TRPM7 functionality to pH status,^132-134 it can be assumed that decreasing pH may contribute to the enhancement of channel activity under anoxic conditions. However, specific data examining this relationship in the context of neuronal cell death has been lacking.

Table 2. Notable TRPM7-related findings during neuronal studies examining ischemia.

Neuronal preparations	Ischemic or stress model utilized	TRPM7 modulators employed	Major results
Rat cortical cultures	OGD	Gd3+ (inhibitor), TRPM7 siRNA	Suppression of TRPM7 presented decreased Ca2+ uptake, reduction in ROS levels, and averted cell death (Aarts et al., 2003)
Mouse hippocampal cultures	Attenuation of divalent cation concentration	Reduction in Ca2+ levels (activator), TRPM7 shRNA	Diminishing Ca2+ levels augmented TRPM7 current in neurons and increased cell death in HEK-293 cells. Current was abolished in the presence of TRPM7 shRNA (Wei et al., 2007)
Rat hippocampal sections, cultures and acute slices	4-vessel occlusion	In vivo intrahippocampal stereotaxic injections of recombinant serotype 1 adeno-associated virus containing TRPM7 shRNA	TRPM7 shRNA attenuated neuronal cell death and improved recovery from ischemic damage in relation to synaptic plasticity, Morris Water Maze performance and contextual fear memory (Sun et al., 2009)
Rat hippocampal cultures and neurons differentiated from pheochromocytoma PC12 cells	LPS or hydrogen peroxide	TRPM7 siRNA	Elevated TRPM7 expression was observed following either LPS or hydrogen peroxide administration. Both primary and PC12 derived neurons were protected from LPS-mediated cell death in the presence of TRPM7 siRNA (Nunez-Villena et al., 2011)
Mouse cortical cultures	Hydrogen peroxide	2-aminoethoxydiphenyl borate (inhibitor), TRPM7 siRNA	Application of either 2-aminoethoxydiphenyl borate or TRPM7 siRNA safeguarded neurons from hydrogen peroxide-induced damage (Coombes et al., 2011)
Mouse cortical cultures	Zn2+ treatment and OGD	Gd3+ (inhibitor), 2-aminoethoxydiphenyl borate (inhibitor) or TRPM7 siRNA	Utilization of TRPM7 inhibitors or siRNA silencing protected neurons from Zn2+-mediated injury under direct ion administration or OGD (Inoue et al., 2010)
Rat hippocampal cultures	Acute chemical ischemia (KCN administration) or OGD	Gd3+ (inhibitor), 2-aminoethoxydiphenyl borate (inhibitor) or TRPM7 shRNA	OGD or chemically-induced increases in neuronal Mg2+ levels were inhibited by TRPM7 blockers or shRNA (Zhang et al., 2011)

Abbreviations: TRPM7, transient receptor potential melastatin 7; OGD, oxygen glucose deprivation; Gd³⁺, gadolinium; Ca²⁺, calcium; siRNA, small interfering RNA; shRNA, short hairpin RNA; LPS, lipopolysaccharide; Zn²⁺, zinc; Mg²⁺, magnesium; KCN, potassium thiocyanate

In addition to elucidating the factors behind TRPM7 activation during ischemia, another pending issue remains the clarification of the underlying mechanisms that this chanzyme may instigate. In this case, the observed elevation of Zn²⁺ or Mg²⁺ following OGD in primary cultures^46,47 suggests the involvement of distinct downstream signaling pathways from those normally associated with NMDA-receptor mediated excitotoxicity and Ca²⁺ overload.²⁸ The demonstrated ability of TRPM7 to interact with cytoskeletal structures during cell line studies may also provide another viable avenue for exploration. While results pertaining to the nature of TRPM7 involvement in cellular adhesion have been conflicting (likely due to variances in relative expression levels and type of cell lines utilized),^101,116 a comprehensive assessment of this phenomenon during ischemic neuronal death would be of significant interest. In this regard, recent reports have revealed the breakdown of Neural Cell Adhesion Molecule (NCAM) as a functional consequence of injury caused by middle cerebral artery occlusion.^135,136 A noteworthy finding in the N1E-115 neuroblastoma cell line was the involvement of TRPM7 in the development of podosomes. While these actin-based protrusions have been suggested to contribute toward tissue invasiveness during malignancy,¹³⁷ they are also thought to be mediators of cell adhesion.¹¹⁴ The relatively recent discovery that synaptically-localized podosomes can also exist in neural structures may be indicative of their potential roles in neuronal systems.¹³⁸ In this case, podosomes appeared as dynamic adhesive structures that possessed a multitude of prominent proteins (including actin, dynamin, myosin IIA, talin, vinculin and paxillin) and contributed functionally to postsynaptic maturation. With regards to the neuronal cytoskeleton, our findings that cofilin is a crucial effector of TRPM7-related signaling suggests a vital interplay between these two factors for the realization of anoxic cell death. Further studies in this area would be imperative to shed light on the TRPM7/cytoskeleton rapport. For these purposes, the utilization of preexisting knockout mice for the cofilin regulator LIMK and its upstream mediators may prove fruitful.^126,139,140 Interestingly, modifications in the level of intracellular Ca²⁺ ions are observed during both TRPM7-mediated cell injury^41-47 and disruption of the cytoskeleton^19-21 respectively. Therefore, a more detailed examination of factors associated with this second messenger pathway may yield additional evidence in the intriguing relationship between TRPM7 and the cytoskeleton. In consideration of the numerous processes involved during cerebral ischemia, the ubiquitous nature of TRPM7 (for example its presence in VSMCs) may also suggest broader and as yet unspecified roles for its functionality beyond synaptic capabilities. Hence, investigating the contribution of non-neuronal sources of TRPM7 (such as the cerebral vasculature) to these conditions may be of considerable interest.

Recent studies have also yielded promising results in relation to the development of pharmacological modulators for this chanzyme. TRPM7 activity was shown to be enhanced in the presence of phospholipase C-coupled receptor agonists such as bradykinin, lysophosphatidic acid and thrombin under physiological settings.¹⁴¹ In addition to cation blockers, reports have identified 5-lipoxygenase inhibitors (such as AA861, NDGA and MK886) as robust attenuators of TRPM7 current.^83,142 Administration of AA861 in HEK-293 cells also prevented the occurrence of death from the actions of apoptotic agents (such as staurosporine) to an extent comparable to TRPM7 knock-down.¹⁴² Another study demonstrated that the SK (small conductance calcium-activated potassium) channel regulator NS8593 was an effective suppressor of TRPM7 current in a variety of different cell lines and preparations.¹⁴³ Furthermore, an intriguing mediation of TRPM7 activity was exhibited by the protease inhibitor nafamostat mesylate in accordance with variations in the extracellular concentrations of Ca²⁺ and Mg²⁺.¹⁴⁴ Research dedicated to the discovery of such drug classes represents an exciting endeavor in the continuing attempts to discern more specific TRPM7 inhibitors.¹⁴⁵ It is our hope that ongoing explorations into the underlying mechanisms and pharmacology associated with this unique chanzyme will result in the development of effective countermeasures that may be utilized during clinical trials of ischemic therapy.

Glossary

Abbreviations:

ADF

actin depolymerization factor

AET

anti-excitotoxic therapy

AMPA

alpha-amino-3-hydroxy-5-methyl-4- isoxazole-propionic acid

APKs

atypical protein kinases

ATP

adenosine triphosphate

ASICs

acid sensing ion channels

Ca²⁺

calcium

CaMKII

calcium/calmodulin-dependent kinase II

CPKs

conventional protein kinases

EAATs

excitatory amino acid transporters

Gd³⁺

gadolinium

JNK

c-Jun N-terminal kinase

KCN

potassium thiocyanate

LIMK1

Lin-11, Isl-1 and Mec-3 kinase 1

LPS

lipopolysaccharides

MagNuM

magnesium-nucleotide-regulated metal

MAPK

mitogen-activated protein kinase

Mg²⁺

magnesium

Mg-ATP

magnesium-adenosine triphosphate

MHC

myosin heavy chain

NCAM

neural cell adhesion molecule

NCCD

Nomenclature Committee on Cell Death

NMDA

N-Methyl-D-aspartate

NOS

nitric oxide synthase

OGD

oxygen-glucose deprivation

PCR

polymerase chain reaction

PtdIns(4,5)P₂

phosphatidylinositol 4,5-bisphosphate

PKC

protein kinase C

ROS

reactive oxygen species

shRNA

short hairpin RNA

siRNA

small interfering RNA

SK

small conductance calcium-activated potassium

SLC41A2

solute carrier family 41 member 2

SNARE

soluble N-ethylmaleimide-sensitive factor activating protein receptor

TRP

transient receptor potential

TRPA

transient receptor potential ankyrin

TRPC

transient receptor potential canonical

TRPM

transient receptor potential melastatin

TRPM7

transient receptor potential melastatin 7

TRPM8

transient receptor potential melastatin 8

TRPML

transient receptor potential mucolipin

TRPN

transient receptor potential no mechanoreceptor potential C

TRPP

transient receptor potential polycystin

TRPV

transient receptor potential vanilloid

VSMCs

vascular smooth muscle cells

Zn²⁺

zinc