Non-canonical processes that shape the cell migration landscape

Abstract

Migration is a vital, intricate and multi-faceted process that involves the entire cell, entails the integration of multiple external cues and, at times, necessitates high-level coordination among fields of cells that can be physically attached or not, depending on the physiological setting. Recent advances have highlighted the essential role of cellular components that have not been traditionally considered when studying cell migration. This review details how much we recently learned by studying the role of intermediate filaments, the nucleus, extracellular vesicles and mitochondria during cell migration.

Introduction

The process of cell migration - whether amoeboid or mesenchymal, single cell or collective – is essential during embryonic development, tissue homeostasis in the adult, and in most disease settings. As such, cell migration receives a great deal of attention from biologists and biomedical scientists alike. Much consideration continues to be devoted to key determinants such as chemoattractant gradient sensing, actomyosin assembly, focal adhesion dynamics, and matrix remodeling [1]. Research focused on these important processes, and their regulation by intricate cellular pathways, are prevalent in the literature. This said, the growing power of genetically-tractable model systems along with the ever-expanding opportunity to ask probing questions using live imaging and high-throughput assays are revealing that, as could be expected, the complex and multifaceted process of cell migration requires contributions from cellular elements far beyond the usual suspects.

This review examines the role of hitherto unknown, unconventional, and therefore non-canonical contributors to cell migration. Specifically, we discuss recent advances made in uncovering and defining the contributions of four cellular components – the nucleus, extracellular vesicles, mitochondria, and intermediate filaments – to cell migration.

Cytoplasmic and nuclear intermediate filaments

Intermediate filaments (IFs) comprise a large (n=70) and diverse group of 10-nm filament-forming proteins that are expressed and function in a fashion that is tissue type- and differentiation-dependent, and are dynamically regulated depending on biological context. All nucleated cells in higher metazoans (eumetazoans except arthropods) feature at least two IF networks, with one located inside the nucleus and one residing in the cytoplasm. Each of these impact local mechanical properties and a surprisingly broad array of cellular processes and their associated signaling events [2–4].

A relevant example of the pleiotropic involvement of IFs in cell migration is provided by the type III IF vimentin in relation to epithelial-mesenchymal transformation (EMT). Vimentin is prominently expressed in the course of embryonic development. Its expression becomes restricted to cells of mesenchymal origin in the adult but is rapidly and broadly induced after injury to various tissues (e.g., muscle, CNS), and also occurs in disease states that range from cataracts to cancer [5]. Vimentin’s contributions to EMT range from the regulation of key upstream signaling events anchored by TGF-β1, ERK, Slug effectors [6–8] and even β-catenin [9] to impacting the organization and dynamics of F-actin [10–12], microtubules [13], and adhesion complexes [14]. In a recent study, Cheng et al. [8] showed that activation of TGF-β1 and Slug and the resulting impact on EMT and ECM deposition were compromised in the absence of vimentin in injured skin tissue in vivo, with an associated paracrine effect on keratinocyte activation and wound re-epithelialization. Together with previous studies [7,15], these findings support the existence of a feedback loop mechanisms whereby vimentin and Slug positively regulate one another.

A newly defined and prominent role for cytoplasmic IFs in the regulation of collective cell migration has recently come to the fore. The expression of a small group of keratin genes (the type II K6a and K6b, and type I K16 and K17) is robustly induced in wound-proximal keratinocytes following injury to skin and related complex epithelia and is maintained in cells actively migrating and engaged in wound re-epithelialization [16,17]. Interestingly, early studies of Krt6a/Krt6b null keratinocytes have shown that they migrate faster than their wildtype counterpart [18], correlating with increased activity and altered subcellular partitioning of Src kinase, a master regulator of cell migration, and altered regulation of several known Src substrates including focal adhesion proteins [19]. In a recent study, Wang and colleagues [20] built on these observations and showed that individual Krt6a/Krt6b null keratinocytes embedded in cell sheets show enhanced directionality but decreased cohesiveness correlating with lower desmoplakin levels and decreased cell-cell adhesion. Faster migration of Krt6a/Krt6b null skin keratinocytes is ECM-dependent and correlates with an enhanced rate of focal adhesion turnover, possibly the result of reduced myosin IIA protein levels. From this, one infers that K6a/K6b (along with partner K16) coordinate and integrate cell-cell and cell-matrix adhesion events to preserve the integrity of the keratinocyte sheet migrating into a wound site [20]. These findings significantly build on previous studies of skin keratinocytes completely null for all keratins [21,22] and are also reminiscent of a series of studies evidencing a related role for plectin in migrating skin keratinocytes [23]. Plectin, interestingly, is a versatile cytoskeletal crosslinker able to interact with all major classes of IFs and with F-actin, microtubules, components of cell-cell and cell-matrix adhesion complexes, the cytoplasmic side of the nuclear envelope, and mitochondria [24].

A related response takes place in the CNS, where injury gives rise to a phenomenon known as reactive gliosis, in which wound-proximal astrocytes are rapidly activated and mobilized to participate in tissue repair [25]. Mature astrocytes primarily express several isoforms of the type III glial fibrillary acidic proteins (GFAP) as their IF system [26], but after injury, they show increased GFAP expression in addition to inducing vimentin and the type IV nestin, a stem cell marker [27]. Together, the IF network comprising GFAP, vimentin, and nestin provides mechanical resilience and otherwise participates in several aspects of the glial and neuronal responses to various types of insults [26]. Recently DePascalis et al. [28] reported that the compound loss of GFAP, vimentin, and nestin restricts the collective migration of astrocytes in primary culture, secondary to major alterations in focal adhesion dynamics, the actin-driven treadmilling of adherens cell-cell junctions, and mechanical coupling to the acto-myosin system. Here again, plectin was again shown to be intimately involved [28]. While the overall influence of IF loss appears opposite in CNS astrocytes relative to skin keratinocytes, a common thread across these studies focused on collective cell migration (see also [29]) is emerging, whereby the composition and organization of IFs invariably and profoundly impact the regulation of F-actin, the acto-myosin system, cell-matrix adhesion, and the kinase-dependent regulation of cellular processes (see Table 1).

Table 1 –

Impact of manipulating IF gene expression on collective cell migration

Parameters	Type of IF deficiency
	Krt6a/Krt6b null Wang et al., J Cell Biol (2018)	GFAP/Vim/nestin null DePascalis et al., J Cell Biol (2018)	Krt8 knockdown (morpholino) Weber et al., Dev. Cell (2012) Sonavane et al. Development (2017)
Relevant cell types & biological context	Skin keratinocytes (mouse) Wound repair (skin)	Astrocytes Reactive gliosis post-CNS injury	Mesendoderm cells Developing Xenopus embryos
Migration phenotype	Faster cell migration; Increased migration speed and increased directionality	Slower cell migration; Decreased directionality and persistence of migration	Slower cell migration; Decreased directionality
Impact on Cell adhesion	Increased FA turnover; Decreased cell-cell adhesion; Decreased sheet cohesiveness	Greater tractions forces; Slower turnover of FAs; Slower actin-dependent retrograde flow of AJs	Altered regulation of FAs; Altered cell-cell adhesion in “follower” cells
Key molecular players	Src activity (activated) Desmoplakin Myosin IIA	Plectin Vinculin	C-Cadherin Plakoglobin Myosin II Rac1
Other elements of relevance	Phenotype is ECM-dependent; Properties are cell-autonomous; Nuclear-localized desmoplakin; K6 binds myosing IIA directly.	Centrosomes are reoriented; Reduced coupling of focal adhesions to acto-myosin; IFs impact the distribution of traction forces across the cell.	Pulling of C-cadherin at cell exterior induces cell polarization and directional migration via a keratin-plakoglobin complex; IFs impact the distribution of traction forces across migrating cells.

Abreviations: AJ, adherens junction; CNS, central nervous system; ECM, extracellular matrix; FA, focal adhesion; IF, Intermediate filaments.

Finally, the nuclear lamins, and lamin A/C in particular, represent an important determinant of the viscoelastic properties of the nucleus and the entire cell [30]. Higher levels of lamin A, which typically occur in differentiated cells, have been found to promote stiffness in a way that impedes migration in a restrictive 3D environment but enhances stress resistance and cell survival for a broad array of cell types [31,32]. Moreover, the maintenance of low levels of lamin A was recently found to be relevant in granulocytes undergoing chemokine-induced chemotaxis and cell squeezing and extravasation across endothelial barriers under physiological conditions [33].

Nucleus

The traditional understanding of how the nucleus affects cell migration falls into three categories: first by controlling cell polarity, second by acting as a mechano-sensing and force-transducing organelle and third by resisting deformation required for traction through confined spaces (as briefly discussed in the previous section).

The nucleus regulates the onset of cell polarity by controlling the retrograde actin flow between the nuclear envelope (NE) and the linker of nucleoskeleton and cytoskeleton (LINC) complex [34]. The LINC complex also plays an essential role in mechano-sensing during cell polarity by tethering the nucleus to the actin cytoskeleton through transmembrane actin-associated nuclear (TAN) lines – a process that is dependent on the ATP binding protein Torsin A and its activator lamina-associated polypeptide 1 (LAP1) [35]. Similarly, actin organization associated with the LINC complex controls cell migration by connecting the NE to focal adhesions (FAs), enabling transmission of mechanical signals to the nucleus [34]. In addition to F-actin and FAs, the LINC complex also physically connects the nucleoskeleton and NE to microtubules and IFs on the cytoplasmic side [36]. The LINC complex in conjunction with the nuclear lamina and microtubules (see previous section) also controls nuclear stiffness, which is in turn critical for confined cell migration. For example, enucleated rat embryonic fibroblast cytoplasts placed in PDGF gradients are able to polarize and respond to PDGF in 2D culture, but show migration defects in 3D collagen matrices, possibly due to the absence of mechano-sensing by the nucleus [37]. The perinuclear actin, in addition to its role in cell polarity initiation, seems to regulate this confined cell migration by controlling nuclear deformation in an Arp 2/3-dependent manner, probably through controlled rupture of the intranuclear lamina shell [38].

The physical role of the nucleus during cell migration, such as the ones related when discussing lamins, has been adeptly reviewed by others [39]. However, the interaction of the nucleus with traditional players of cell migration, such as FAs, is not just merely mechanical. These traditional components of the cell migration landscape also regulate cell migration non-canonically by regulating actin flow and the expression of motility-related genes (Figure 1). For example, FA components, such as FAK, translocate to the nucleus during cell de-adhesion, where they mediate p53 turnover [40] and upregulation of myocyte enhancer factor 2 (MEF2) during mechano-transduction in cardiomyocytes [41]. Nuclear FAK also promotes the transcription of chemokines, such as CCL5, during squamous cell carcinoma invasion, which in turn drives the recruitment of regulatory T cells [42]. Similarly, paxillin, another FA component, accumulates in the nucleus during high FA turnover [43], where it regulates androgen receptor-mediated transcription [44]. The presence of cytoskeletal regulators in the nucleus seems to be a common theme during cell migration. For example, nuclear actin accumulation occurring following F-actin depolymerization has been shown to increase gene expression of adhesion-related genes, thereby acting as a positive feedback mechanism for cell-matrix adhesion during cell migration [45]. However, whether this accumulation of nuclear actin is dependent on its filament-forming potential has been a matter of debate. Although F-actin polymers have not been detected by phalloidin, a natural compound that binds actin filaments with at least seven actin monomers, dynamic actin filaments were observed in fibroblasts using actin-chromobody-TagGFP tagged with an NLS sequence [46]. This suggests a three-way equilibrium between (i) G- and F-actin in the cytosol, (ii) G- and F-actin in the nucleus and (iii) nuclear and cytosolic G-actin (Figure 1). Any process that affects this tripartite equilibrium, such as an increase in F-actin during lamellipodia formation or increase in cytosolic G-actin during ATP depletion [47], will affect the actin equilibrium in the nucleus, thereby affecting the expression of migration-related genes [48,49].

Figure 1.

Summary of the role of the nucleus during cell migration.

(1) Nuclear deformation is one of the primary rate limiting steps of confined cell migration. The stress on the nuclear superstructure may be released momentarily by NE rupture, which is quickly repaired by the members of ESCRT III to maintain nuclear integrity.

(2) Nuclear translocation of FA components during high turnover events, such as FA disengagements, in the course of cell migration can act in a feed forward manner through the transcriptional regulation of chemokine genes.

(3) Monomeric actin released through the depolymerization of F-actin translocates into the nucleus through actin binding proteins such as MAL, Importin 9 and Cofilin. Monomeric G-actin also undergoes an alternate mode of polymerization within the nucleus by nucleating factors such as mDia1/2. Sequestration of monomeric actin within the nucleus directly affects the equilibrium of actin between the cytoplasm and the nucleus, which in turn affects actin polymerization during F-actin formation.

(4) G-actin mediates transcription of the actin gene itself as well as actin binding factors, therefore completing a feedback loop to maintain F-actin homeostasis in the cell.

New evidence also suggests an involvement of actin-binding proteins in nuclear actin assembly. Rho effectors, such as mDia1/2, regulate the transcriptional reprogramming of cytoskeletal and growth-promoting genes [50]. Unlike RhoA, which is thought to require actin-binding proteins for its nucleoplasmic shuttling, Rac1 regulates actin polymerization through the WAVE/Arp 2/3 complex and interacts with nucleoplasmin-1/B23 to shuttle in and out of the nucleus [51]. Interestingly, in this context, it was established that nuclear Rac1 acts as a nuclear membrane organizer by preferentially partitioning to nuclear membrane microdomains and not by directly regulating gene expression or nuclear actin polymerization [51].

It has become increasingly apparent that the NE undergoes substantial stress and rupture during migration in confined spaces [52]. Remarkably, recent evidence shows that NE integrity is restored in part by the translocation of membrane remodeling complexes, especially members of the ESCRT III family, to sites of NE damage [52]. In immune cells, the NE also acts as a signaling hub for the synthesis of eicosanoids. These proinflammatory modulators are potent chemoattractants for myeloid cells such as neutrophils and macrophages and are responsible for directing their migration [53]. Recent evidence from the study of wound mitigation in zebrafish also point to a direct link between NE swelling and eicosanoid synthesis, in a cPLA₂-dependent manner [54]. Eicosanoids are further discussed in the next section in the context of their synthesis from the NE and dissipation through extracellular vesicles.

Exosomes and Extracellular Vesicles

Extracellular vesicles (EVs) encompass exosomes and microvesicles, which respectively originate from the fusion of multivesicular bodies with the plasma membrane or are directly shed from the plasma membrane [55]. Initially thought to be cellular junk, EVs were recently rediscovered as key components in cellular communication through their ability to carry regulatory RNA. For example, EVs released by malignant breast cancer cells are able to transform non-malignant cells by transferring the miR-200 family of miRNAs, which regulate EMT through Zeb [56]. Similarly, miR-21-containing exosomes have been shown to promote the migration and invasion of recipient esophageal cancer cells by targeting programmed cell death 4 (PDCD4) and activating the c-Jun N-terminal kinase (JNK) pathway [57]. Recent evidence suggests that exosomes regulate cell migration not just through the transfer of specific regulatory RNAs (Figure 2) but also by the transfer of thousands of different proteins and RNAs participating in interconnected signaling networks, thereby phenocopying migratory properties of metastatic cells in non-metastatic ones [58].

Figure 2.

Summary of the role of extracellular vesicles during cell migration.

(1) MMP containing EVs are deposited at the invadopodia extensions, which help in cancer cell invasion by degrading the underlying extracellular matrices.

(2) Transformed malignant cells, fibroblasts or immune cells release miRNA-containing EVs, which upon internalization by recipient cells regulate the expression of migration related genes. In other cases, the internalized miRNA may itself transform a naïve cell into a transformed phenotype, e.g. initiation of EMT.

(3) EVs containing chemoattractants such as leukotrienes, prostaglandins and other lipid metabolites are released from activated cells, which either themselves form a gradient or release their chemoattractant cargo in a time-dependent manner to direct the migration of distant cells towards the site of EV release.

A role for EVs in regulating cell migration through RNA-independent mechanisms is increasingly coming into focus (Figure 2). For example, the protrusive activity and motility of breast cancer epithelial cells is enhanced by exosomes-containing Wnt-planar cell polarity pathway components secreted by cancer-associated fibroblasts [59]. Similarly, a gradient of CXCL13-containing exosomes released by lymphatic endothelial cells provides directional cues to dendritic cells [60]. In addition, it has been shown that surface-bound VEGF covering the outside of shedding microvesicles is required for β1-integrin turnover during astrocyte migration [61]. Exosomes can also synthesize and release chemoattractants that eventually form stable chemotactic gradients in the extracellular space. For example, Dictyostelium discoideum cells release cAMP from exosomes in an ABCC8 transporter-mediated process to regulate signal relay during their developmental program [62]. Similarly, chemotactic neutrophils release leukotriene B₄-containing exosomes that act on resting neutrophils to amplify inflammatory processes [53]. Other eicosanoids and leukotrienes have also been shown to be present in exosomes derived from macrophages and dendritic cells, which helps in the recruitment of neutrophils [63]. The packaging of sparsely soluble chemoactive lipids and fatty acids within exosomes to protect them from the extracellular milieu seems to be a common theme in chemotactic gradient formation. For example, prostaglandins and other arachidonic acid metabolites have been shown to be present in exosomes of RBL-2H3 rat basophilic cells, which act in an autocrine and paracrine fashion to induce cell migration in resting cells [64]. Similarly, exosomes from adipocytes, which contain enzymes responsible for fatty acid oxidation, induce cell migration in recipient melanoma cells by transfer to bioactive FAs [65].

Remarkably, EVs may play a more direct role in cell migration by modifying the extracellular matrix. Tumor-derived exosomes regulate extracellular matrix (ECM) remodeling through the deposition of cellular proteases as reported in pancreatic adenocarcinoma [66] or by depositing MMPs as shown using a spontaneous rat tumor model [67]. Deposition of MT1 metalloprotease-containing exosomes at the invadopodia of head and neck squamous cell carcinoma cells further exemplifies the role of EVs during invasion [68] (Figure 2). More importantly, exosomes also regulate ECM remodeling through the deposition and recycling of ECM components. Indeed, in migrating HT1080 cells, exosomes mediate the deposition of fibronectin in an autocrine manner to increase cellular persistence and speed [69].

While released EVs have been shown to confer increased migratory potential in recipient cells, few reports have implicated them in the suppression of migration. It has been shown that migration of endothelial cells can be inhibited by TAM-derived exosomes through the transfer of miR-146b-5p, which target the TRAF6/NF-kB/MMP2 pathway [70]. Similarly, miR-146a present in EVs-derived from atherogenic macrophages, accelerates the development of atherosclerosis by decreasing cell migration of naïve recipient macrophages [71]. Interestingly, tetraspanins, which have historically been used as exosome markers, have recently been shown to be associated with a reduction in tumor metastasis, owing to the ability of CD9, CD63 and CD82 to remodel host ECM by interacting with integrins and adhesion molecules [72]. Of note, other exosomal tetraspanins, such as CD151, have been implicated in tumor growth, suggesting that the balance between various classes of tetraspanins may govern metastatic outcome [73].

Mitochondria

Migration is known to require extensive cellular remodeling on a dynamic and cyclic fashion, a bioenergetically expensive process that is fueled to a significant degree by mitochondria. Indeed, exogenous pro-migration signals give rise to dynamic structural and spatial changes of the mitochondrial network within neurons [74]. These changes allow cells to synthesize ATP, regulate cytoplasmic calcium levels, elevate reactive oxygen species, and alter metabolic activity for a wide range of signaling purposes. We have known since the mid-1980’s that mitochondria do not traffic intracellularly through simple diffusion but rather utilize microtubules and F-actin fibers to facilitate anterograde and retrograde movement, respectively [75]. Efficient trafficking of mitochondria requires fission, to break the mitochondrial network into smaller bodies, and a series of motor and adaptor proteins to transport and dock small mitochondria to their intended destination [76,77]. It is now well appreciated that efficient cell migration requires coordinated and dynamic mitochondria positioning, particularly in the context of a spatiotemporal need for ATP production and calcium signaling (Figure 3).

Figure 3.

Summary of the role of mitochondria during cell migration.

(1) Upon fission, mitochondria are trafficked along microtubules and F-actin using motor proteins (i.e. KIF5) and adaptor proteins (i.e. Miro1).

(2) Mitochondria localize to areas of energy demand during cell migration in order to provide ATP needed for cytoskeletal polymerization, integrin adhesion, and focal adhesion turnover.

(3) Mitochondria are highly localized in the pseudopod of an invading or extravasating cell, where ATP levels are elevated compared to the main cell body.

(4) Plasma membrane calcium transporters replenish ER calcium reserves. The mitochondria can help maintain this influx of calcium by either (4.1) tethering to the ER to buffer calcium release or (4.2) trafficking to the cell periphery to directly buffer cytoplasmic calcium. Both require the mitochondrial calcium uniporter (MCU) to maintain the calcium homeostasis needed for cytoskeletal flexibility and focal adhesion turnover.

Studies focusing on mitochondria during the migration of cancer cells have led to several key findings. In particular, the reorganization of mitochondria plays a key role in regulating cancer cell metastasis and has become a target in cancer therapeutics. Both Zhao et al. [78] and Desai et al. [79] showed that mitochondria are strategically positioned between the nucleus and leading cell edge, respectively in breast cancer and epithelial cancer cells. Desai and colleagues further established that the intracellular trafficking of mitochondria promotes epithelial cancer cells’ directionality, velocity and persistence during migration [79]. Interestingly, this mitochondrial movement and the subsequent tumor invasion are regulated by docking of syntaphilin (SNPH), a protein that stabilizes mitochondria and prevents them from being trafficked (along microtubules) to the cell’s outer edges [80]. Decreased survival rate of breast cancer patients has been correlated with decreased levels of SNPH [80]. Importantly, mitochondrial transport also involves dynamin-1-like protein (Drp1)-dependent fission [80]. Seo et al. [81] observed that silencing SNPH in prostate adenocarcinoma PC3 cells enhances cell invasion, a process that is blocked with concomitant Drp1 knockdown. Preventing mitochondrial fission presumably hinders the ability of motor proteins to readily transport mitochondria to the cell periphery where ATP is needed [81]. Directing the mitochondrial network to a specific cellular region allows for the robust ATP production required to sustain processes such as focal adhesion turnover in both metastasizing tumor cells and non-tumor cells. When studying migrating SKOV-3 human ovarian adenocarcinoma cells, Cunniff et al. observed that mitochondria are actively trafficked to pseudopods where high levels of ATP were also measured. This trafficking was found to be dependent on microtubules. Moreover, the authors found that the local activation of AMP-activated protein kinase (AMPK) increased mitochondrial flux, ATP levels and cytoskeletal dynamics, suggesting that the regulation of cellular energy directly affects mitochondrial dynamics and cytoskeletal architecture [82].

Pro-inflammatory signals promote chemotaxis and invasion of immune cell effectors into infected tissues. This requires the formation of a uropod containing the microtubule-organizing center (MTOC) and adhesion zones, both of which have exhibited enhanced mitochondrial accumulation in primary and immortalized lymphocytes in vitro [83]. Using time-lapse confocal microscopy, Morlino et al. [84] showed that mitochondria localize to contact points between human T lymphoblasts and activated ombilical cord endothelial cells grown in co-culture conditions. These authors further showed that Miro-1, an adaptor protein that bridges microtubules and mitochondria, is required for mitochondria to localize first to the adhesion zone in response to TNFα and later to the uropod in a dynein/dynactin-dependent manner [84]. In mouse embryonic fibroblasts, Schuler et al. also found that the ATP:ADP ratio is significantly reduced at the periphery of cells deficient in Miro-1 as a result of mitochondria remaining in the perinuclear region. This ultimately decreased cellular migration, adhesion and focal adhesion stability [85]. Activation of mitochondrial ATP production and secretion has also been reported to occur in response to chemokine receptors activation in naïve T cells [85]. Remarkably, the released ATP works in an autocrine fashion by binding to P2X4 receptors and regulating pseudopod protrusion and migration in response to SDF-1a. Similarly, ATP release and autocrine purinergic signaling have also been shown to regulate neutrophil chemotaxis [86,87] through a differential mitochondrial impact on mTOR signaling at the front and rear end of these cells [88]. Together, these studies beautifully highlight how localized ATP production regulates intra- and extra-cellular signaling during cell migration.

Cell migration also relies on mitochondria to achieve the delicate state of calcium homeostasis needed to promote processes such as cytoskeletal dynamics and focal adhesion turnover (Figure 3). The mitochondria calcium uniport (MCU) - a calcium ion channel that transports cytosolic calcium into the mitochondria - has been reported to regulate exogenous calcium uptake. In breast cancer and HeLa cells, silencing or inhibition of the MCU decreases total cytoplasmic calcium levels. This leads to cytoskeletal stiffness by inhibiting Rho-family GTPases and focal adhesion stability via calpain inhibition, which collectively reduces cell motility [89]. This was also observed in B lymphocytes via a mitochondrial Na⁺-Ca²⁺ exchanger through calcium recycling with the ER in response to chemoattractants that induce F-actin polymerization [90]. It was further established that elevating MCU expression leads to increase mitochondrial calcium levels, resulting in a metabolic shift towards glycolysis [91]. This study showed both in vitro and in vivo that MCU overexpression promotes metastasis through enhance glycolysis. This so-called Warburg effect is a well-known promoter of tumorigenesis, and while it is not known whether calcium acts directly on glycolytic enzymes or if it damages the electron transport chain, mitochondrial calcium signaling is required for persistent aerobic respiration [91]. The role of the MCU and mitochondrial calcium homeostasis is still being investigated with respect to cellular migration. However, because the MCU has been shown to regulate breast cancer progression, it has become an attractive drug target [92]. Finally, other mitochondria-dependent processes, such as redox balance regulation, are being examined for their ability to regulate cellular migration. This was recently shown in vivo in zebrafish neutrophils, using CRISPR/Cas9 to knockout cytoplasmic superoxide dismutase 1 and mitochondrial superoxide dismutase 2. Both knockouts exhibited reduced neutrophil motility that could be rescued using reactive oxygen species scavengers [93]. With new evidence continuing to emerge on the role of mitochondria during migration for a variety of cell types, further investigation is required to understand the precise underlying mechanisms.

Perspectives

Irrespective of the modalities and associated significance, cell migration is a holistic endeavor that entails contributions from all components that make cells what they are. Here we reviewed recent advances that highlight hitherto underappreciated roles for intermediate filaments (Table 1), the nucleus (Figure 1), extracellular vesicles (Figure 2) and mitochondria (Figure 3) during migration. The contributions of these “unusual suspects” are key to signaling and communication between migrating cells, to the dynamic regulation of cellular stiffness and energy demands, and to the need for coordination among cellular elements and the cell’s relationship to its microenvironment. Continued efforts in this area will no doubt deepen our appreciation and understanding of the significance of these emerging effectors, and others, to the intricate and vital process of cell migration.