Cross-talk between calcium and protein kinase A in the regulation of cell migration

Calcium (Ca²⁺) and the cAMP-dependent protein kinase (PKA) are pleiotropic cellular regulators and both exert powerful, diverse effects on cytoskeletal dynamics, cell adhesion and cell migration. Localization, by A-kinase anchoring proteins (AKAPs), of PKA activity to the protrusive leading edge, integrins, and other regulators of cytoskeletal dynamics has emerged as an important facet of its role in cell migration. Additional recent work has firmly established the importance of Ca²⁺ influx through mechanosensitive transient receptor potential (TRP) channels and through store-operated Ca²⁺ entry (SOCE) in cell migration. Finally, there is considerable evidence showing that these mechanisms of Ca²⁺ influx and PKA regulation intersect – and often interact – and thus may work in concert to translate complex extracellular cues into the intracellular biochemical anisotropy required for directional cell migration.

Introduction

Successful cell migration requires cells to interpret their extracellular environment through diverse receptors in order to regulate and rearrange distinct cytoskeletal and adhesive events in subcellular space. Thus, the cell migration machinery must be regulated by signaling intermediates that can be activated by diverse stimuli and can exert control over a large number of downstream targets – all with temporal and spatial specificity. In this regard, two systems – Ca²⁺ influx and signaling through the cAMP-dependent protein kinase (PKA) – are perfectly suited [1-3]. However, despite intense study of Ca²⁺ and PKA, and despite – or perhaps because of – their myriad possible connections to cell migration, our understanding of how these signals control cell motility is far from complete.

Some of the first formal descriptions of Ca²⁺ patterns in migrating cells established the enduring tenet that [Ca²⁺]_i during migration is highly variable in both time and subcellular space (e.g.[4,5]). Since then, the mechanisms controlling this variability have been a major research focus. Recent work has highlighted the importance of Ca²⁺ influx from two distinct sources – namely, mechanosensitive channels and store-operated Ca²⁺ entry (SOCE) – in the regulation of migration.

Over a similar period, the role of PKA as both a positive and negative regulator of cell motility was established in a number of systems [3]. A critical advance in understanding this regulation came from incorporating the importance of AKAP-mediated localization in specifying the effects of PKA [6] and demonstrating that PKA is spatially regulated in motile cells [7••], providing a mechanism through which its promiscuous activity can be brought to bear specifically on processes in migration. Recent work has confirmed and expanded this observation, bringing specific migration- or cytoskeleton-associated AKAPs to the fore and providing mechanisms for focusing PKA activity even more tightly on discrete cytoskeletal events.

Finally, a wealth of literature demonstrates the substantial cross-talk and coincidence between Ca²⁺ and cAMP/PKA [8•-11••], and recent observations further support the hypothesis that these pathways may act in concert to achieve the specificity and diversity of controls required for the regulation of cell migration.

Ca²⁺ in migration I – It's a stretch, but there are flickers of hope

Recent work investigating Ca²⁺ in migration has capitalized on our growing understanding of the importance of mechanical forces in cell migration [12] by demonstrating important roles for mechanosensitive stretch-activated Ca²⁺ channels (SACCs) during migration. Motile cells generate a number of cellular forces [12] which inevitably impinge on the shape and tension of the plasma membrane. Given that nearly every type of cell expresses some form of mechanosensitive or stretch-activated ion channel [13,14•], one would hypothesize that such channels might regulate and/or be regulated by cell migration. Indeed, the importance of SACCs in migration was definitively demonstrated by Lee and co-workers who reported that tension-dependent, SACC-mediated Ca²⁺ transients are required for trailing edge release and retraction [15••]. Conversely, a role for SACCs in the leading edge was established later by studies in which local application of Gd³⁺, an effective (but not exclusive) inhibitor of SACC-mediated Ca²⁺ influx, to the leading (but not trailing) edge resulted in global inhibition of traction forces and migration [16•]. In neither case, however, was the identity of the SACC known.

A wealth of recent evidence [17-27] suggests that excellent candidates for such SACCs include members of the transient receptor potential (TRP) family [28]. A prominent, recent example is a report describing that TRPM7 ([23••]; Figure 1), a unique SACC which also contains an active α-kinase domain [28], mediates generation of transient microdomains of elevated cytosolic Ca²⁺ - coined ‘Ca²⁺ flickers’ – that occur within leading edge lamellae in an actomyosin contractility-dependent manner. The importance of TRPM7 in migration is further underscored by work showing that TRPM7 interacts with F-actin and myosin IIA, and phosphorylates the latter, in an agonist-dependent manner [17], and that TRPM7-mediated Ca²⁺ entry regulates cell-substrate adhesion through localized modulation of m-calpain activity [18].

Figure 1.

Connections between Ca²⁺ flickers and PKA. During cell migration, highly localized Ca²⁺ transients – ‘flickers’ – are mediated by TRPM7 through a mechanism that requires cellular tension through integrins (αβ), focal adhesions (FA), and actomyosin contractility. Ca²⁺ entering through TRPM7 stimulates Ca²⁺-induced Ca²⁺ release (CICR) from the endoplasmic reticulum (ER) through the IP₃ receptor (IP₃R), which increases flicker amplitude. Local Ca²⁺ may directly activate nearby Ca²⁺-sensitive adenylyl cyclases (AC) or indirectly activate others, e.g. via protein kinase C. Ca²⁺ can also exert both positive and negative effects on AKAP function. PKA is known to enhance or activate TRPM7 currents and our unpublished observations suggest that PKA can directly phosphorylate the channel (see Figure 3). PKA can also phosphorylate and enhance the activity of the IP₃R, which contributes to flicker amplitude. Finally, many focal adhesion proteins and cytoskeletal regulatory proteins (X,Y,Z) are regulated by PKA, and this may also represent a route for cross-talk between PKA and TRPM7 activity. ECF, extracellular fluid; Cyto, cytoplasm.

Other TRP channels also have connections to the cytoskeleton and cell migration [29]. TRPC1 is recruited to the leading edge of glioma cells migrating in response to EGF, and inhibition or silencing of TRPC function inhibits EGF-induced chemotaxis [24]. At least one function of TRPC1 appears to be the generation of a gradient of Ca²⁺ in leading edge lamellipodia (in contrast to the punctuate leading edge flickers generated by TRPM7) required for chemotactic cell polarity [21]. Expression of TRPV1 enhances motility in some systems [20] but exerts a negative role in others [27], while the closely-related TRPV4 is required for cytoskeletal remodeling and reorientation in endothelial cells in response to cyclic stretch [22] – an effect quite likely dependent on TRPV4's functional and physical association with β1 integrins [22,26] and F-actin [19]. Even in these detailed reports, however, the exact molecular mechanisms coupling TRP and other mechanosensitive channels to cellular tension are not well understood [14•], and thus represent an important avenue for future study.

Ca²⁺ in migration II – What's in store?

As mentioned above, TRPM7-mediated flickers do not act alone, but rather occur on the backdrop of a diffuse gradient of Ca²⁺ that increases from the front to the rear of the cell [23••]. Also, while flicker activity was completely eliminated by RNAi against TRPM7, flicker amplitude (but not probability) was also partially blunted by inhibition of the function or expression of the inositol (3,4,5) trisphosphate receptor (IP₃R) [23••]. This suggests that Ca²⁺ flickers are locally triggered by tension-dependent activation of TRPM7, but are amplified by local IP₃R-mediated store release ([23••]; Figure 1). In broad support of this, Ca²⁺ influx through TRPM7 was shown to promote Ca²⁺-induced Ca²⁺ release through the ryanodine receptor during migration [25].

The connection between Ca²⁺ stores and cell migration is further highlighted by recent work focusing on store-operated Ca²⁺ entry (SOCE) mediated by STIM1 and Orai1. In SOCE, receptor-mediated activation of phospholipases C leads to IP₃-mediated depletion of Ca²⁺ from the ER, which in turn stimulates Ca²⁺ influx through the plasma membrane ([30] and Figure 2). Groundbreaking work (reviewed in [30]) identified STIM1 as the ER-localized Ca²⁺ sensor which, upon store depletion, interacts with and activates Orai1 at the plasma membrane to allow influx of extracellular Ca²⁺. Recent work elegantly demonstrates that STIM1, Orai1, and SOCE play critical roles in the migration of a number of cell types. Specifically, genetic inhibition of Orai1 or STIM1 or selective pharmacological inhibition of SOCE inhibits the in vitro migration of breast cancer cells [31••], smooth muscle cells [32•], neutrophils [33•], and endothelial cells [34•], and also inhibits breast tumor metastasis [31••] and angiogenesis [34•] in vivo. Of note, there is strong evidence demonstrating that, in addition to Orai1, STIM1 can activate TRPC1 and possibly other TRPC isoforms in response to store-depletion [30], and it may be this axis of SOCE that regulates some programs of chemotaxis (see above and [24]).

Figure 2.

Connection between store-operated Ca²⁺ entry (SOCE) and PKA. SOCE, and its principal mediators STIM1 and Orai1, play increasingly important roles in the migration of a number of cell types. During SOCE, signals generated by receptors (e.g. G-protein coupled receptors (GPCR) or receptor tyrosine kinases (RTK)) at the plasma membrane activate phospholipases C (PLC) to generate IP₃, which causes the release of Ca^2+- from intracellular stores via the IP₃R. Store depletion activates STIM1, which re-localizes within the ER membrane to allow interaction with and activation of the plasma membrane Ca²⁺ channel Orai1. Store-depletion has also been shown to stimulate adenylyl cyclase (AC) and activate PKA. Reciprocally, PKA may regulate the SOCE pathway at a number of levels, e.g. through phosphorylation and modulation of many PLC-coupled receptors, the IP₃R, phospholamban associated with the sarco-/endoplasmic reticulum Ca²⁺-ATPase (SERCA), and possibly Orai1 itself (see Figure 3).

Whither goest the Ca²⁺ in migration?

Currently, it is unknown whether Stim1/Orai1/SOCE exert a global or localized effect in migrating cells. In that regard, an intriguing hypothesis is that, in a cell undergoing chemotaxis, the higher relative concentration of extracellular stimulus at the leading edge would produce a localized or graded concentration of intracellular IP₃, which would in turn elicit localized store depletion and SOCE within the leading edge. This would result in the generation of leading edge Ca²⁺ signals with kinetics, localization, and perhaps targets that would be distinct from Ca²⁺ flickers. The idea of differential regulation of SOCE or its components at the leading edge is supported by preliminary evidence from our laboratory that phosphorylation of Orai1 is differentially regulated in cell bodies versus chemotactic pseudopodia (Figure 3).

Figure 3.

New observations of PKA phosphorylation of migration-associated Ca²⁺ channels. (a) Schematic depicting separation and isolation of cell bodies (CB) from chemotactic pseudopodia (PD) (see [7••] and references therein). COS7 cells were transfected with pEYFP-Orai1 then cultured as depicted in to form pseudopodia in response to a gradient of EGF. (b) YFP-tagged Orai1 was immunoprecipitated from isolated cell bodies and pseudopodia and analyzed by SDS-PAGE and immunoblotting against Orai1 and the phosphorylated PKA consensus site (RRXpS/T). (c) Schematic for back-phosphorylation of TRPM7 by PKA. HEK293 cells were transfected with pcDNA4TO-FLAG-GFP-TRPM7 then treated with either (to inhibit PKA and promote accessibility of PKA sites) or forskolin and IBMX (F+I, to increase cAMP, activate PKA, & drive full phosphorylation of PKA sites). FLAG-GFP-TRPM7 was then immunoprecipitated and subject to in vitro phosphorylation with recombinant PKA + ³²P-γ-ATP with or without prior dephosphorylation with calf intestinal phosphatase (CIP). (d) Reaction products were analyzed by SDS-PAGE, immunoblotting, and autoradiography.

This begs the question – what specific purposes are served by distinct Ca²⁺ signals such as SOCE or flickers during cell migration? Flickers appear to be critical for cells to reorient towards an oblique gradient of growth factor [23••] and thus resemble the filopodial Ca²⁺ transients that steer growth cones during axon guidance [35]. And while SOCE has been recognized for decades, its specific physiological effects are largely unknown [36]. One possibility is that SOCE may regulate migration through highly localized control of focal adhesions. Nanometer-scale targeting of microtubules to focal adhesions promotes focal adhesion disassembly, putatively through localized delivery of one or more “relaxing signals” [37,38]. Importantly, localized Ca²⁺ waves can trigger disassembly of focal adhesions [39] and STIM1 has been shown to track to microtubule plus ends in an EB1-dependent manner [40]. These observations support a hypothesis in which microtubules target STIM1 to focal adhesions and thus promote localized SOCE and focal adhesion turnover. The question of what functions are served by distinct types of Ca²⁺ signals during cell migration will ultimately be answered by a better understanding of the distinct effectors regulated by these signals.

PKA in cell migration – Still in the lead

PKA has numerous cytoskeleton- and migration-associated targets [3] – indeed, various subunits of PKA have recently been identified as part of a myosin II-responsive focal adhesion proteome [41], placing PKA in immediate proximity to many known and potential targets. Given this, and the fact that PKA can exert both negative and positive effects on cell migration [7••], it is not surprising that the role of PKA in migration is critically dependent on its localization through AKAPs. In all systems in which it has been directly examined, PKA activity is enriched in the leading edge [7••,42,43] and AKAP-mediated localization is required for leading edge formation and/or maintenance and for efficient cell migration [7••,42-45]. The question of which AKAP(s) are responsible for localizing PKA during cell migration is not resolved. The most direct investigation so far demonstrated that leading edge gradients of PKA activity in breast and colon carcinoma cells are mediated by AKAP-Lbc, a PKA scaffold and Rho-specific guanine nucleotide exchange factor [43]. However, several other AKAPs are also physically or functionally associated with cytoskeletal or adhesive elements [6]. The Wiskott-Aldrich Syndrome protein family verprolin-homologous protein 1 (WAVE1) was one of the first and most direct examples of a cytoskeleton-associated AKAP [46] and recent work suggests that WAVE2 may serve a similar function [45]. Similarly, an early report of the interaction between IQGAP1 and AKAP79 [47] is complemented by recent work demonstrating a ternary complex between IQGAP2, AKAP220, and PKA, in which PKA-mediated phosphorylation of IQGAP2 enhances its interaction with active Rac [48]. Finally, non-canonical mechanisms of PKA anchoring, e.g. via α4 integrins [44••] or actin filaments [49], may also play an important role during migration.

As mentioned above, the role for PKA in migration is certainly not always a positive one. A recent, elegant example of this is the ability of PKA to act as a critical switch during vascular sprouting by suppressing endothelial cell polarity and migration while up-regulating cadherin-mediated cell-cell adhesion [50•]. This occurs through indirect inhibition of Src activity via PKA-mediated phosphorylation and activation of the C-terminal Src kinase (Csk). This inhibition of Src is in contrast to a parallel ability of PKA to directly phosphorylate and activate Src and related kinases [51,52]. The differential regulation of Src by PKA under different cellular circumstances highlights both the breadth of PKA's effects and the fundamental requirement for mechanisms that specify these effects.

Ca²⁺ and cAMP/PKA: Connections

The collaboration between Ca²⁺ and cAMP/PKA signaling in regulating cell function has been appreciated for decades [8•] and has since been expertly reviewed [9-11••]. Indeed, there are numerous, elegant examples of how closely connected the two signals can be in both time and space (e.g. [53]; reviewed in [11••]). The number of levels at which Ca²⁺ and PKA can regulate one another, combined with the strong influence each signal has over migration-associated processes, makes it quite likely that cross-talk between Ca²⁺ and PKA will play an important role in regulating cell migration. This notion is underscored by recent, compelling reports of Ca²⁺/PKA cross-talk – including unpublished observations from our laboratory – related to both TRPM7 and SOCE.

The most direct route connecting Ca²⁺ to PKA is through Ca²⁺-mediated regulation of AC activity and thus cAMP production. This effect is governed by the isoform of AC expressed; ACs 1, 3, 5, 6, 8, and 9 are Ca²⁺-regulated, while the remaining isoforms are insensitive ([54]; see also [53] and references therein). Specifically, AC1 and AC8 are activated by Ca²⁺ via calmodulin, while the other isoforms are inhibited by Ca²⁺, either directly or through phospho-regulation. It should be noted, however, that even the classically ‘insensitive’ isoforms (i.e. ACs2, 4, and 7) may still be indirectly regulated by Ca²⁺ via phosphorylation by PKC [54]. Ca²⁺ may also regulate PKA signaling by affecting AKAP function. Ca²⁺/CaM binding inhibits the interaction of two AKAPs (AKAP79 and gravin/SSeCKS) with membrane phospholipids [55,56]. Conversely, Ca²⁺ promotes the binding of IQGAP2 to AKAP220, thereby facilitating PKA-mediated phosphorylation of IQGAP2 [48]. These observations support a mechanism whereby local increases of Ca²⁺ may exclude PKA or, alternatively, displace specific AKAP-containing complexes while retaining or enhancing others – an efficient way for Ca²⁺ to regulate the specificity of PKA signaling. In addition, the anchoring of PKA to IQGAPs [47,48], coupled with the ability of IQGAPs to serve as Ca²⁺/calmodulin-dependent regulators of cytoskeletal and microtubule dynamics [57], bring IQGAPs to the fore as examples of cytoskeletal effectors at the nexus of PKA and Ca²⁺ signals.

A recent and elegant study showed that depletion of ER Ca²⁺ can elicit cAMP synthesis and PKA activation in a STIM1-dependent manner ([58••]; Figure 2). Also, SOCE efficiently activates AC8 [53,59] and induces PKA, STIM1, and AC8 to co-localize at the plasma membrane and cofractionate in lipid rafts [59]. Conversely, PKA can potentiate SOCE in a variety of cells [10], although the mechanism for this is unclear. Given the hypothesis outlined earlier, suggesting a mechanism of focal adhesion turnover through microtubule-mediated targeting of Stim1, it is intriguing to note that PKA has also been shown to associate with microtubule plus ends [60] and may phosphorylate Orai1 (Figure 3). These considerations, combined with the growing recognition of the importance of STIM1, Orai1, and SOCE in cell migration [31•• ,32•,33•, 34•], make this an area rife with potential future studies.

PKA is implicated in the regulation of a large number of mechanosensitive TRP channels, including those implicated in migration, e.g. V1 [61], V4 [62], and, notably, M7 [63]. For TRPV1, this regulation is critically dependent on AKAP-mediated anchoring of PKA to the channel [64] and there is evidence to suggest a similar requirement for other TRP channels [62]. Furthermore, in all cases, PKA signaling has a positive or sensitizing effect on channel function, and for at least TRPV1 and V4, this requires direct phosphorylation of the channel [61,62]. Takezawa et al [63] showed that the inhibition and stimulation of TRPM7 by muscarinic and beta-adrenergic receptors, respectively, required modulation of cAMP levels and that βAR-mediated stimulation also required PKA activity. While clearly demonstrating that PKA activity positively regulates TRPM7, that report did not demonstrate direct phosphorylation of the channel. However, TRPM7 contains at least four strong consensus sites for PKA phosphorylation and preliminary evidence from our laboratory strongly suggests that TRPM7 can be directly phosphorylated by PKA (Figure 3). Given this, and the ability of mechanical tension to elevate cAMP and activate PKA [65], it is tempting to hypothesize a role for PKA in the mechanosensitive activation of some TRP channels associated with cell migration.

Finally, in addition to increasing [Ca²⁺]_i, PKA can also contribute to clearance of cytosolic Ca²⁺ – and thus termination of Ca²⁺ signaling – through regulation of the activity of plasma membrane Ca²⁺-ATPases (PMCAs) via direct phosphorylation and, in some cells, the sarco-/ endoplasmic reticulum Ca²⁺-ATPase (SERCA), via phosphorylation of phospholamban [10].

CONCLUSIONS

Beyond the Ca²⁺ events discussed above, it should be noted that PKA interacts with numerous other factors that control or interpret Ca²⁺ flux and that have been implicated in cell migration [10,66,67]. The hypothesis – or perhaps ‘prediction’ - being proffered here is that, based on the strong influence of Ca²⁺ and PKA on cell migration and the staggering complexity and ubiquity of cross-talk between these signals, collaboration between Ca²⁺ and PKA will play a fundamentally important role in the spatial and temporal regulation of cell migration. This hypothesis prompts two important questions. First, is the regulation of Ca²⁺ influx – an activity for which PKA is to the manner born – a major target for localized PKA in migrating cells? Conversely, is the spatial regulation of PKA activity – with its numerous targets within the cell migration machinery – a major target for specific Ca²⁺ signals? While much more experimentation is needed to address these questions, the investigation of the interplay between Ca²⁺ and PKA during cell migration will inevitably provide important mechanistic insights into the regulation of this complex, fundamental cellular process.