Electrochemical regulation of cell polarity and the cytoskeleton

Authors


  • Monitoring Editor: Pekka Lappalainen

Abstract

Cell polarity plays a key role in regulating cell–cell communication, tissue architecture, and development. Both internal and external cues participate in directing polarity and feedback onto each other for robust polarization. One poorly appreciated layer of polarity regulation comes from electrochemical signals spatially organized at the level of the cell or the tissue. These signals which include ion fluxes, membrane potential gradients, or even steady electric fields, emerge from the polarized activation of specific ion transporters, and may guide polarity in wound-healing, development or regeneration. How a given electrochemical cue may influence cytoskeletal elements and cell polarity remains unclear. Here, we review recent progress highlighting the role of electrochemical signals in cell and tissue spatial organization, and elucidating the mechanisms for how such signals may regulate cytoskeletal assembly for cell polarity. © 2012 Wiley Periodicals, Inc.

Electrochemical signals have been widely studied for their role in regulating cell physiology. Proper ion transport through the membrane is a prerequisite for cell survival and defects in this process have been implicated in numerous diseases [Prevarskaya et al., 2010; Hollenhorst et al., 2011; Webb et al., 2011]. Electrochemical aspects of cell biology have also been deeply characterized in neurons, where waves of ion influx and efflux and consequent membrane potential changes allow for the propagation of action potentials. One less understood aspect of cellular electrochemistry stands in the control that electrochemical cues may have on the spatial organization of cells and tissues.

Most cells in our body are surrounded by organized electrical signals. These have relatively small magnitude and depict near steady-state organization, and could serve as guiding cues for spatially regulating cell and tissue behavior [Jaffe and Nuccitelli, 1977; McCaig et al., 2005]. Endogenous electrical currents, fields, and membrane potential gradients have been mapped around epithelial monolayers, developing embryos, and regenerating organisms [Hotary and Robinson, 1992; Szatkowski et al., 2000; Reid et al., 2007; Levin, 2009; Kucerova et al., 2011; Reid and Zhao, 2011]. They have been proposed to participate in directing cell migration, division, and proliferation in these processes [McCaig et al., 2005; Zhao et al., 2006; Levin, 2009]. Organized ionic currents have also been detected around large polarized single cells such as developing eggs or pollen tubes and suggested to help establishing a spatial order for polarized division or growth [Robinson and Jaffe, 1975; Weisenseel et al., 1975; Kline et al., 1983; Schreurs and Harold, 1988]. These currents may be generated by the activation of ion transporters at defined locations around cells or tissues [Feijo et al., 1999; Levin et al., 2002; McCaig et al., 2005; Minc and Chang, 2010]. Application of exogenous electric fields, similar to those measured in vivo can direct polarity in many different cell types, ranging from bacteria, fungi to neutrophils [McCaig et al., 2005].

Very little is known on how this electrical patterning arises in the first place, and how it is transduced to influence cytoskeletal elements for cell polarity. The lack of molecular descriptions has impaired progress in this area and has kept the field outside of mainstream research for many years. In here, we review novel insights into the molecular and biophysical mechanisms linking electrochemical cues such as membrane potential, pH, or external electric fields, to the regulation of cytoskeletal elements and polarity proteins.

Membrane Potential and Tissue Architecture

The membrane potential of a cell results from gradients of charges segregated across the insulating plasma membrane. Membrane potential is dynamically regulated by ion channels and pumps, which function in exporting and importing anions and cations through the membrane. Values of resting membrane potentials may vary largely between different cell types, ranging from −10 to −150 mV [Levin, 2012]. Some cells globally modify their membrane potential to perform specific functions or during different periods of their life cycle, for instance, during egg fertilization or cell differentiation [Blackiston et al., 2009; Wessel and Wong, 2009]. Metastatic cancer cells usually display depolarized (reduced) membrane potential which has been associated with metastatic potential [Binggeli and Weinstein, 1985; Binggeli et al., 1994]. Membrane potential may feedback on ion transport, intracellular pH, or membrane surface charges at the membrane inner and outer leaflet. It has long been suggested as a cue regulating tissue patterning [Jaffe and Nuccitelli, 1977]. One proposed view is that tissue-scale membrane potential gradients could yield electrophoresis of morphogens through gap junctions or other cell–cell connections [Woodruff and Telfer, 1980; Bohrmann and Gutzeit, 1987; Levin et al., 2002; Esser et al., 2006]. Another one is that membrane potential could indirectly influence downstream cytoplasmic factors organization or even gene transcription [Levin, 2012]. Some recent works which revisit these concepts within the modern knowledge of cell and developmental genetics begin to provide functional evidence for the role of membrane potential, in tissue patterning in embryos and animals.

Developmental fidelity relies on proper localized gene transcription in the embryo. In the Xenopus embryo, a group of cells in the anterior neural field specify the activation of eye field transcription factors (EFTF) in the neighboring tissue for eye development (Fig. 1A). Membrane potential may have important inputs in this activation. At stage 17/18, these neural anterior cells have striking hyperpolarized membrane potential as compared to the neighboring tissue cells, that can be qualitatively visualized by sensitive fluorescent imaging of different membrane potential dyes [Pai et al., 2012]. Perturbation of this hyperpolarization pattern, achieved by expressing and locally activating exogenous anionic or cationic channels, yields major defects in eye development or the formation of an ectopic eye in the gut and other caudal areas. In situ visualization of expression patterns, show that membrane potential values correlate with specific gene transcription, and may thus contribute to proper developmental patterning in this tissue.

Figure 1.

Influence of membrane potential on embryonic development and tissue architecture. A: Embryonic eye development in Xenopus laevis involves the bilateral patterning of EFTFs by two clusters of hyperpolarized cells in the anterior neural field at stage 18 before the formation of eye primordial (stage 22). Disruption of the membrane potential signal results in diminished or lost EFTF signal, leading to disrupted eye development. B: Planarian tissue regeneration. Upon amputation of both head and tail, wild-type fragments generate a membrane potential gradient with the head blastema depolarized and the tail blastema less depolarized. H,K-ATPase inhibition with SCH treatment results in relative hyperpolarization and consequent regeneration of a neither tail-nor head structure. On the other hand, ivermectin (IVM) treatment leads to relative depolarization and consequent head formation. C: Zebrafish pigment pattern formation. The wild-type (WT) fish is characterized by a regular segregation of blue and yellow stripes on the skins. Pigment cells in these stripes are called melanophores (M) and xanthophores (X), respectively. Stripes may be generated by the repulsive behavior of M contacting X, which is in turn regulated by a contact-dependent membrane depolarization of M. In the homozygous kir7.1 mutant (jaguar), contact-dependent depolarization is malfunctioning, resulting in a lack of repulsive behavior, X-M mis-segregation and defects in the stripe pattern.

Similar evidence for the role of membrane potential in tissue patterning has been put forward in regeneration processes [Adams et al., 2007; Levin, 2009; Beane et al., 2011]. Planarians are flat worms which serve as model organisms to study regeneration of tissues, because they exhibit an extraordinary ability to regenerate lost body parts. Regeneration involves a specific timing of events, starting from sensing the tissue cut, regenerating specific cells, and growing back tissues in a proper manner. Hints on the importance of bioelectrical signals on regeneration mechanisms in similar organisms had long been suggested [Marsh and Beams, 1957]. In a recent study, Beane et al. measured membrane potential gradients generated upon amputation of either the head or the tail of a planarian [Beane et al., 2011]. When the head is amputated, cells in the border of the cut, called the head blastemas depict a strongly depolarized membrane, while tail blastemas are typically hyperpolarized (Fig. 1B). These membrane polarization patterns may depend on the differential expression of an H+-K+ ATPase pump in these two different parts of the animal. To assess the functional role for membrane voltage as a potential regulator of anterior/posterior polarity, the authors express exogenous ion channels that alter membrane potential values to these border blastema cells. Strikingly, hyperpolarizing tail blastemas yields regeneration of a second head and a bi-headed organism, while depolarization of the head blastemas yields regeneration of a tissue that is neither a tail nor a head (Fig. 1B). These results suggest that membrane potential may act as an upstream signal that allows the organism to discern its global head to tail polarity.

Dose-dependent values of membrane potentials may impose differential expression of specific sets of genes. Several mechanisms have been suggested to transduce membrane potential into specific gene expression patterns. One proposed view is that membrane potential values are transduced via voltage-sensitive calcium channels that indirectly regulate gene expression [Weick et al., 2003]. Another one comes from the identification of voltage sensitive phosphatases that may trigger specific signaling pathways from membrane potential—dependent conformational changes [Villalba-Galea, 2012]. Other examples include membrane-potential dependent affinity between specific ligands and G-protein type receptors [Ben-Chaim et al., 2006], or even movement of signaling molecules through cellular gap junctions [Levin, 2012]. As these different mechanisms may be triggered at different threshold of membrane potential changes, it is likely to expect different dominant mechanisms in different tissues or situations. How specific is the signal-transduction downstream of such electrical cue and how membrane potential patterns are established in tissues in the first place, remain interesting open questions.

Novel work in zebrafish tissues brings an interesting point of view for how cellular membrane potential may help to pattern a tissue or an organism [Inaba et al., 2012]. Zebrafish animals have pigmented skin stripes of alternating blue and gold (for males) and blue and silver (for females) that run along their body (Fig. 1C). Each stripe is typically composed of a pigment cell type. The melanophores compose the gold stripe and the xanthophores are in the blue stripes. Mechanisms for how these two cell types stay apart to regulate stripes patterning are lacking, but several fish mutants that depict defect in stripe patterns have been identified [Maderspacher and Nusslein-Volhard, 2003; Iwashita et al., 2006]. One such mutant, called Jaguar has a mutation in a gene encoding an inward potassium rectifying channel, Kir7.1. This channel regulates the resting membrane potential by promoting the entry of potassium in the cytoplasm, and is expressed in the melanophores cells of the fish skin. Dynamic tracking of the melanophores membrane potential visualized with fluorescent dyes reveals a Kir7.1-dependent membrane depolarization when these cells contact a xanthophore. This membrane depolarization underlies a contact inhibition process that yields a polarity switch to drive migration away from the xanthophore. Thus, cellular membrane potential values are sensitive to cell–cell contacts and may regulate cell polarity and consequent large-scale tissue patterning. In this situation, it remains to be established whether cellular membrane potential modulation triggering this polarity switch is global or local at the subcellular level.

Electrochemical Cues Regulating Cell Polarity

Other specific electrochemical cues have been involved in organizing tissues and embryos. Large scale ion fluxes originating from the differential expression of ion transporters are found around tissues and cells and regulate the intracellular content of certain ions such as calcium, potassium, and even protons. These intracellular gradients could modulate the activity of intracellular polarity factors [Robinson and Jaffe, 1975; Levin et al., 2002]. Although data correlating these current patterns with polarity abound, mechanisms for achieving and sensing intracellular ion gradients are still poorly defined. Proton currents and consequent pH gradients are particularly interesting, because many cytoskeletal components are thought to be tightly regulated by pH values [Bachewich and Heath, 1997; Denker and Barber, 2002b].

Migrating cells establish a polarization axis by specifying a zone of actin polymerization at the leading edge. This process is regulated by the polarized recruitment and activation of polarity factors such as the conserved Rho-type small GTPase cdc42 [Etienne-Manneville, 2004]. At the leading edge, activated GTP-bound cdc42, recruits regulators of actin nucleation such as formin, N-Wasp, and components of the Arp2/3 complex, for progression of the leading edge and migration [Frantz et al., 2008]. Local activation of cdc42 is mediated by Guanine exchanging factors (GEFs) that are recruited to the cortex through interactions with specific proteins or lipids such as phosphatidylinositol 4,5-bisphosphate (PIP2). In migrating fibroblast, an interesting link between pH regulation, cdc42 activation, and polarity establishment has been put forward throughout subsequent studies (Fig. 2B) [Hooley et al., 1996; Denker and Barber, 2002a; Frantz et al., 2007]. In this system, the conserved sodium/proton exchanger NHE1 promotes fibroblast polarity and migration [Denker and Barber, 2002a]. This transporter, through its regulatory effect on cortical pH, may function in stabilizing at the leading edge the binding of a cdc42 GEF to PIP2 at the plasma membrane (Fig. 2B) [Frantz et al., 2007]. Interestingly, NHE1 recruitment to the leading edge depends on cdc42 activation [Hooley et al., 1996], and thus NHE1, pH regulation and polarity activation work within a positive feedback loop regulating polarity axis establishment in fibroblast migration.

Figure 2.

Electrochemical cues regulating cell polarity. A: Tumor cell invasion. During invadopodia maturation, the sodium/proton exchanger NHE1 is recruited to the invadopodium site by cortactin phosphorylation and generates local pH increase. This change in pH triggers cortactin–cofilin unbinding, de novo actin polymerization and invadopodia progression. B: Fibroblast cell migration. NHE1 is localized along the leading edge of a migrating cell, generating a local basic pH that stabilizes GEF interaction with PIPs in the membrane. GEF recruitment mediates cdc42 activation and actin polymerization. Active cdc42 in turn recruits NHE1 to the leading edge, amplifying a positive feedback loop that reinforces polarized migration. C: Drosophila epithelial cell polarity. Proton transporters play a role in regulating the asymmetric localization of PCP core proteins to polarized epithelia in the plane of a tissue. The sodium-proton exchanger Nhe2 regulates the recruitment of Dsh to the plasma membrane, through its effect on pH and consequent electrostatic interaction of the positively charged Dsh DEP domain with negatively charged acidic phospholipid groups in the membrane.

Similar components have been involved in polarized cancer cell invasion. Invasion, which relies on breakage of the extracellular matrix through a specific polarized structure called invadopodia is a hallmark of cancer metastasis [Weaver, 2006]. Formation of a polarized invadopodia depends on the specific timing of the recruitment of actin polymerization regulators. Early steps of polarization include cortactin phosphorylation and detachment of cofilin to generate free barbed ends filaments and de novo actin polymerization followed by invadopodia progression [Oser et al., 2009]. In a recent study, Magalhaes et al. provide evidence that these early steps are associated with the recruitment of NHE1 and local pH gradients [Magalhaes et al., 2011]. In particular, they demonstrate that cortactin phosphorylation leads to the local recruitment of NHE1, which in turn yields local alkalization in the future invadopodia area (Fig. 2A). This local pH increase may trigger unbinding of cofilin from cortactin, by reducing the binding strength between the two proteins, promoting actin polymerization and invadopodia protrusion. Thus, pH regulation through NHE1 may work in reinforcing the definition of a polarity axis and in catalyzing actin polymerization for progression of this subcellular structure. pH deregulation through the sodium/proton exchangers has been implicated in numerous cases of metastasis [Cardone et al., 2005]. pH may directly regulate actin-binding proteins including cofilin, profilin, or talin in vitro and in vivo [McLachlan et al., 2007; Frantz et al., 2008; Srivastava et al., 2008], and thus a similar mechanism may be at play in other actin-based polarization systems.

pH regulation and protein domain charges have also been implicated in planar cell polarity (PCP) pathway regulation in Drosophila epithelial tissues [Simons et al., 2009]. This conserved signaling pathway plays a key role in setting epithelial tissue polarity and architecture in different organisms [Goodrich and Strutt, 2011]. In the Drosophila wing and eye, the pathway serves to define asymmetric recruitment of specific factors within the plane of the tissue. This polarity relies on the recruitment of the transmembrane protein Frizzled (Fz) at cell–cell contact along the tissue axis. Fz binding to Dishevelled (Dsh) activates the recruitment of cytoskeleton elements for cell polarity, growth, and division [Segalen and Bellaiche, 2009; Goodrich and Strutt, 2011]. Fz binding to Dsh requires the targeting of Dsh to the plasma membrane. By performing a genome-wide RNAi screen in drosophila cells, Simons et al. identified the sodium/proton exchanger Nhe2 as a key regulator of Dsh targeting to the membrane. This protein shares homology with the human NHE family, and more particularly with hNHE3. They propose that pH values regulated by Nhe2 influence head phospholipids protonation level (whose pKa are close to neutral) impacting the binding of the polybasic stretch of Dsh DEP (Dishevelled, Egl-10, Pleckstrin) to the plasma membrane inner leaflet. Thus Nhe2 and electrochemical cues regulate Dsh targeting to the membrane, which allows this protein to interact with Fz. In a subsequent study, the same group showed that Fz localization relies on pH regulation, through the action of a V-ATPase proton pump, that extrude protons form organelles and cells by consuming ATP energy [Hermle et al., 2010]. These studies thus demonstrate the importance of proton transporters, pH regulation, and protein charges for the proper stabilization of a polarity axis in a tissue context [Hermle et al., 2011].

Other recent evidences suggest that membrane inner leaflet charges of certain lipids such as phosphatidylserine, may participate in recruiting or stabilizing polarity components in the budding yeast S. cerevisiae [Yeung et al., 2008; Fairn et al., 2011; Das et al., 2012]. In the fission yeast, S. pombe, the conserved proton ATPase pump pma1p, that regulates membrane potential and intracellular pH in fungi and plants [Sanders et al., 1981], has been shown to regulate polarized growth and cell morphogenesis [Minc and Chang, 2010]. Transcellular pH gradients exist in large polarized cells such as pollen tubes or fucoid eggs [Gibbon and Kropf, 1994; Feijo et al., 1999], and in fission yeast (our unpublished results). These pH gradients are likely to be cortical or submembranous, since rapid diffusion of protons may impair formation of cytoplasmic gradients. At the cortex, the fast proton extrusion rate by ATPase pumps or ion exchangers coupled with sharp subcellular localization of the channel could explain gradient formation but remains to be explored quantitatively. All together, these data at the single cell and tissue level highlight the importance of electrochemical regulation for polarity. It is likely that pH or membrane potential affect the membrane targeting of different factors or binding strength between proteins.

Yet, these studies suggest some specificity in the polarity pathways activated by the electrochemical cue. Hierarchy in molecular effectors regulating cytoskeletal assembly and polarity pathways coupled with positive feedback loops may allow cells to interpret and use electrochemical cues for specifying and maintaining a polarity axis.

Orienting Cell Polarity to Exogenous Electric Fields

Striking evidences highlighting the electrical aspects regulating polarity, come from electrotactic experiments. In this assay, cells are exposed to small exogenous electric fields (EFs) similar to those found in vivo and orient their polarization with respect to the EF direction [McCaig and Dover, 1991; Rajnicek et al., 1994; Zhao et al., 2006]. EFs can direct division, migration, or polarized growth, and appear as unique universal signals to orient polarity. Cells, ranging from bacteria, fungi to mammalian cells, orient polarity to external EFs. Puzzlingly, different cell types polarize to different directions, some to the anode (positive electrode of the EF), others to the cathode and some even to a direction perpendicular to the EF (Fig. 3). Effects of endogenous EFs on cell polarity are thought to be important for wound healing and development. These effects are used in clinical contexts, for instance in healing therapy and nerve repair [McCaig et al., 2005]. How might cells sense EF signals and reorganize the cytoskeleton and polarity machinery is not well understood. It is generally accepted that the EFs only affect events close to the plasma membrane [Jaffe, 1977]. EFs may influence ion transport and/or membrane potential locally around the cell [Jaffe and Nuccitelli, 1977]. Transmembrane proteins have also been shown to move along the membrane in the presence of EFs through putative electrophoresis mechanisms, potentially leading to the local recruitment or stabilization of downstream polarity factors [Poo and Robinson, 1977; Poo, 1981]. These EF effects may reflect the natural electrochemical organization of ion currents or membrane domains around a single cell. Recent genetic characterization of such effects have begun to shed light on molecular mechanisms regulating polarity by external EFs and their relevance to natural polarity regulation.

Figure 3.

Directing polarity with exogenous electric fields. A: EF-induced orientation of the division plane: Mitotic corneal epithelial cells position their division plane perpendicular to the EF direction. B: WT amoeba D. discoideum migrates to the cathode of an applied EF. Mutant cells deficient in the PI3K and cGMP pathway migrate to the anode. C: Polarized tip growth in WT fission yeast S. pombe cells reorients perpendicular to the EF, but pma1-1 mutants with impaired proton pumping activity reorient to the anode. The pathogen yeast C. albicans grows hyphae to the cathode of the EF, a process that depends on the voltage gated Ca2+ channel CaCch1.

Connections between EF signals and downstream cytoskeletal regulators, including the small GTPase cdc42, the Rho/Rac pathways, integrin signaling and phosphatidylinositol (PIP) signaling, have been suggested in different cellular systems [Pullar et al., 2006; Rajnicek et al., 2006; Zhao et al., 2006; Sato et al., 2009; Minc and Chang, 2010] (Fig. 4). During mammalian wound healing, neutrophils and keratinocytes wound-directed migration depends on PI3K and PTEN which, respectively, positively and negatively regulate PIP2 homeostasis [Zhao et al., 2002, 2006]. Wound-healing relies in part on endogenous EFs generated in the wound, and can be inhibited or accelerated by exogenous application of EFs pointing toward or away from the wound, respectively [Zhao et al., 2006]. In this electrotactic assay, exogenous EFs induce high and sustained phosphorylation of key signaling kinases including, ERK, p38, Src, and Akt. Mutants lacking the catalytic γ-subunit of the PI3K depict impaired activation of Akt and reduced phosphorylation of Src, p38, and ERK and display reduced electrotactic migration of neutrophils and keratinocytes and defective wound closure. Conversely, PTEN deletion enhances EF induced Akt and Src phosphorylation and directional migration, and accelerates wound healing. Thus, PIP signaling regulates electrotactic migration of cells in the wounded tissue and proper healing [Zhao, 2009]. It is interesting to note that PIP signaling is also regulating chemotaxis in these cell types. The downstream machinery required for directional migration is thus likely to be similar regardless of the nature of the spatial cue in this situation.

Figure 4.

Pathways regulating electrotactic responses in different cell types. A: D. discoideum migration toward the cathode of an exogenous EF involves at least two different pathways: The PI3 kinase and PTEN pathways that lead to a polarized distribution of PIP (in yellow); The cyclic GMP pathway (in red). Upstream EF sensors are not well characterized but Ca2+ signaling may be important in this process (in blue). Downstream effectors linking these pathways to actin polymerization may include the protein kinases PkbA and PakA. B: In mammalian cells, EF induced migration has been proposed to rely on different pathways. PIP signaling may be important for keratinocytes and neutrophils during wound responses (in yellow). Integrin signaling has been involved in keratinocytes electrotaxis (in green). ROS signaling may be important in cancer cell migration (in cyan). Upstream regulators of EF sensing may include NHE1 (in red), and calcium signaling, although links to downstream effectors are still lacking. Akt and PakA may link PIP and ROS signaling to actin polymerization machinery. C: In S. pombe cells, reorientation of polarized growth perpendicular to the EF depends on the formin for3p and the small GTPase cdc42p that both regulate polarized actin polymerization. The proton ATPase pump pma1p which regulates pH and is located on the side of the cell may establish putative transcellular proton currents and cortical pH gradients that transduce EF effects to for3p and cdc42p regulation.

The amoebae Dictyostelium discoideum has been instrumental in dissecting molecular mechanisms of directional cell migration [Devreotes and Zigmond, 1988; Parent and Devreotes, 1996; Chen et al., 1996, 1997; Kim, et al. 1997]. When exposed to homogeneous concentrations of cyclic adenosine monophosphate (cAMP) these cells migrate in random directions. In the presence of small EFs, they orient their migration to the cathode of the EF within minutes of EF applications. This EF response is independent of upstream chemotactic receptors [Song et al., 2002]. Downstream signaling modules regulating directional cell migration for instance during chemotaxis include PIP and intracellular cyclic guanosine monophosphate (cGMP) signaling. These effectors promote actin polymerization at the leading edge for migration [Veltman and Van Haastert, 2006; Veltman et al., 2008]. In a recent work, Sato et al. tested the role of these signaling modules in electrotaxis [Sato et al., 2009]. Intracellular cGMP is produced mainly by two enzymes, soluble guanylyl cyclase (sGC) and guanylyl cyclase A (GCA). Mutants lacking the sGC and GCA (gca/sgc) and mutants lacking the cGMP-binding protein C (gbpC-), which display reduced levels of cGMP, exhibited attenuated cathodal electrotactic migration. Similar phenotypes were obtained when PIP signaling was repressed through PI3-kinase inhibition. Strikingly, when both PIP2 synthesis and cGMP pathways were knocked down, cells migrated to the opposite direction, to the anode of the EF. These results suggest the existence of parallel pathways participating in regulating electrotaxis and put forward the existence of a third pathway promoting anodal migration [Sato et al., 2009]. sGC may have structural and spatial competing functions in this regulation, as expression of the C terminal catalytic domain of sCG which produces cGMP, can rescue the cathodal migration defects in gca/sgccells, while expression of the N terminal domain of sCG which binds the actin cytoskeleton at the leading edge promotes anodal migration. These studies support the role of PIP signaling for electrotaxis in another organism, and provide detailed genetic characterization of the molecular mechanisms involved. Cross talk between EFs and polarity in these systems have been proposed to be mediated by calcium transport and membrane potential [Onuma and Hui, 1988; Pullar and Isseroff, 2005; Shanley et al., 2006; Gao et al., 2011] yet the details of this transduction remain to be established (Fig. 4B).

Reactive oxygen species (ROS) signaling may also mediate electrotactic responses. HT-1080 fibrosarcoma cells display EF-induced migration to the anode of the field and concomitant Superoxide (O2) production [Li et al., 2012]. This O2 overproduction may induce the phosphorylation of ERK and downstream signaling components including p38 and AKT for directed migration. Cells overexpressing superoxide dismutase which transform O2 into hydrogen peroxide (H2O2) display impaired ERK1/2, p38, and AKT phosphorylation and impaired EF anodal migration. ROS signaling has been implicated in wound response to direct leukocytes migration to the wound and could thus represent another regulation layer mediating electrotactic response in this context [Niethammer et al., 2009].

Fungal cells and yeasts are model systems to dissect molecular mechanisms underlying polarity. These nonmigrating cells exhibit polarized growth which involves similar regulatory modules and conserved effectors as higher eukaryotes [Chang and Peter, 2003]. EFs may be present in natural fungal habitat and some fungi and molds have been suggested to target wounds by following ion currents and EFs [van West et al., 2002]. Ion transporters and membrane potential regulators are also widely shared between fungi and higher organisms. Most fungal cells display strong electrotropism (EF directed polarized growth) [Harold et al., 1985; McGillivray and Gow, 1986] (Fig. 3). During filamentous growth, the pathogen Candida albicans grows, for instance, its hyphae to the cathode of an exogenous EF [Crombie et al., 1990]. This orientation depends on Ca2+ transport mediated by the voltage gated Ca2+ channel CaCch1p [Brand et al., 2007]. Fission yeast cells which serve as excellent systems to study polarity [Chang and Martin, 2009], also display electrotactic behavior [Minc and Chang, 2010]. These cells which normally grow into a perfect rod-shape, reorient their growth axis to a direction perpendicular to an applied EF, creating cells with a bent morphology [Minc and Chang, 2010]. Candidate genetic screen for polarity regulators and membrane transporters suggest that this response depends on the formin for3p and the small GTPase cdc42p which regulate aspects of actin polymerization and polarity [Martin et al., 2007; Minc et al., 2009] and on pma1p, a conserved proton ATPase regulating intracellular pH. Surprisingly, mutants in these genes still orient to the EF but to the wrong direction, toward the anode of the EF. These results suggest a key role of pH regulation for proper actin polymerization in electrotactic responses and natural polarity regulation [Minc and Chang, 2010].

All together these genetic dissections in different organisms suggest downstream transduction of EF effects by small GTPase, lipid signaling, and actin regulation factors. The cross talk between EF and these polarity modules remains to be clearly defined, although some transporters and ions have been specifically identified in these different systems. Calcium and pH regulation may play key roles in mediating EF effects into polarized reorganization of cytoskeletal regulators. Future experiments dissecting the molecular basis of exogenous EF effects are likely to generate novel fundamental understanding in general mechanisms regulating these effects.

Genetic studies in Dictyostelium and yeast begin to reveal why different cells may polarize to different directions, and suggest that multiple layer of polarity regulation may be spatially reorganized by EFs, with one predominant mode. When the dominant layer is turned off through mutation, the secondary mode takes over and directs polarity to another direction. These EF experiments are likely to reflect physiological events, in wound healing, neuron organization, and development. They may also begin to reveal the natural electrochemical regulation of polarity and cytoskeletal elements. In that view, the EF effect may bias or exacerbate an existing electrical organization, which yields the polarized reorientation in the EF. A specific cytoskeletal regulator may for instance naturally bind to portions of the plasma membrane with specific charges [Fairn et al., 2011; Das et al., 2012], and the EF-induced perturbation on the membrane potential and membrane charges would induce the relocation of this element to redirect polarity. It has long been a puzzle to understand how such small EFs which perturb only 1% of resting membrane potential could orient polarity in such a striking manner [Jaffe and Nuccitelli, 1977]. Positive feedbacks regulating polarization modules, cytoskeleton, and ion transport may begin to bring answers to these long lasting questions.

Many decades after initial suggestions for a role of “bioelectricity” in orchestrating cells and tissue behavior [Piccolino, 2000], novel functional data begin to unravel mechanisms for their importance in development, tissue architecture, and single cell polarity. Molecular mechanisms regulating and transducing electrochemical signals into cytoskeleton regulation are beginning to emerge, and future work should allow to draw more generic models, and to assess specificity in different cell-types and situations. Genetically encoded fluorescent sensors allowing to measure pH [Miesenbock et al., 1998] and recently developed sensors for quantitatively estimating membrane potential shall bring complementary information of electrochemical patterning in cells, tissues, and embryos [Kralj et al., 2011, 2012]. Optogenetics systems may also serve as powerful tools to locally control electrochemical cues at the level of tissues and cells [Fenno et al., 2011]. These optical tools shall bring important numbers and orders of magnitude that are still missing in the field. Other quantitative approaches, such as systems genetics, microfabrication, and modeling [Minc and Chang, 2010; Li and Lin, 2011], will likely help moving forward in dissecting the crosstalk between electrochemical and biochemical signaling for cell polarity.

Acknowledgements

The authors acknowledge Matthieu Piel for careful reading of the manuscript. N.M. acknowledges financial support from the Agence Nationale de la Recherche (ANR) “retour post-doctorants” grant ANR-10PDOC-003-01 and a European FP7-People-CIG grant. D.B. is supported by an Institut Curie PhD fellowship.

Ancillary