Mammalian Plakins, Giant Cytolinkers: Versatile Biological Functions and Roles in Cancer

Cancer is a highly lethal disease that is characterized by aberrant cell proliferation, migration, and adhesion, which are closely related to the dynamic changes of cytoskeletons and cytoskeletal-adhesion. These will further result in cell invasion and metastasis. Plakins are a family of giant cytolinkers that connect cytoskeletal elements with each other and to junctional complexes. With various isoforms composed of different domain structures, mammalian plakins are broadly expressed in numerous tissues. They play critical roles in many cellular processes, including cell proliferation, migration, adhesion, and signaling transduction. As these cellular processes are key steps in cancer development, mammalian plakins have in recent years attracted more and more attention for their potential roles in cancer. Current evidence shows the importance of mammalian plakins in various human cancers and demonstrates mammalian plakins as potential biomarkers for cancer. Here, we introduce the basic characteristics of mammalian plakins, review the recent advances in understanding their biological functions, and highlight their roles in human cancers, based on studies performed by us and others. This will provide researchers with a comprehensive understanding of mammalian plakins, new insights into the development of cancer, and novel targets for cancer diagnosis and therapy.


Introduction
The cytoskeleton, as the important cell scaffold, is a highly organized and interconnected network of filaments consisting of microfilaments (F-actin), microtubules (MTs), and intermediate filaments (IFs). The cytoskeleton is dynamically organized and plays central roles in determining cell shape and numerous cellular functions, including cell polarization, adhesion, migration, proliferation, and intracellular trafficking. All these cellular functions are closely associated with cancer development and progression. Thus, the correct organization of cytoskeleton is critical for cell function. Recently, numerous studies have demonstrated that plakins are versatile giant cytoskeletal crosslinkers and play key roles in regulating cytoskeletal organization and diverse cellular functions [1,2].
Desmoplakin was identified as a critical component of desmosomes in epidermal cells and has two alternatively spliced isoforms, desmoplakin I and desmoplakin II [33,34]. Desmoplakin I and desmoplakin II both comprise an N-terminal region consisting of a plakin domain and a CCR domain, as well as a C-terminal region containing three plakin repeat domains (PRDs) (Figure 1a). However, desmoplakin II contains a shorter CCR domain compared with desmoplakin I [35] (Figure 1a). As one important component of desmosomes that provide strong adhesion between cells, desmoplakin I and II are present in all epithelia, while desmoplakin II is not detected in cardiac muscle [36] (Table 1).
Envoplakin and periplakin were first found to be components of the cornified envelope in the terminal differentiation of epidermal keratinocytes [37]. Further studies demonstrate that envoplakin and periplakin colocalize with desmoplakin at desmosomes, as well as on keratin filaments in the human epidermis [38,39]. Envoplakin is homologous to desmoplakin, BPAG1 and plectin, while periplakin is also one component of desmosomes. Envoplakin and periplakin both consist of the N-terminal plakin domain, central CCR and C-terminal L subdomain, while envoplakin has an additional C type PRD in its C-terminus ( Figure 1a). Through the CCR, the envoplakin and periplakin can form heterodimers [39,40]. Without isoforms being reported, envoplakin and periplakin mainly exist in keratinizing and nonkeratinizing stratified squamous epithelia, but not in most simple epithelia or nonepithelial cells [38,39] (Table 1).
Epiplakin was originally discovered as a 450 kD human epidermal autoantigen that reacted with the serum from a patient with a subepidermal blistering disease [41]. Further molecular cloning of the epiplakin protein reveals that it has a molecular mass of 552 kD [42]. Epiplakin is an unusual plakin due to its unique domain structure that contains only 13 B-type PRDs but lacks an N-terminal plakin domain and CCR domain [42] (Figure 1a). Like other plakins, epiplakin is broadly expressed in many tissues [42][43][44] (Table 1).

Domain Structure of Plakins
The mammalian plakins, some of which have isoforms, are composed of different combinations of varied domains that mainly include the ABD, plakin domain, spectrin repeats, CCR, PRD, EF-hand, and growth arrest specific 2 (GAS2)-related protein (GAR) domain ( Figure 1). Specifically, BPAG1 and MACF1 are also grouped as "spectraplakins" because they have both spectrin repeats and a plakin domain [45]. Each domain has its specific function for conferring plakins with the ability to interact with F-actin, MTs, and IFs, as well as cellular junctional proteins and signaling molecules (Figure 1b).

Actin-Binding Domain (ABD)
ABDs are found in BPAG1, MACF1, and plectin, and they are located at the N-terminal (Figure 1a). Two calponin homology (CH) domains, CH1 and CH2, are arranged in tandem to form the ABD (Figure 1b). Although the CH1 domain can bind actin by itself and CH2 also has a weaker binding affinity, the tandem organization pattern of CH1 and CH2 greatly increases the binding affinity for F-actin [46][47][48]. Further crystal structure analysis of the F-actin-binding domain of MACF1 reveals that the ABD adopts a closed conformation with extensive intramolecular contacts between the CH1 and CH2 domains [49]. Moreover, there is a key tyrosine-phosphorylation site within the ABD of MACF1, and the phosphorylation of ABD by Src/focal adhesion kinase (FAK) is essential for the F-actin-binding of MACF1 [49].
Thus, the ABD confers plakins with a capacity to bind F-actin and to regulate cytoskeletal dynamics. For example, the ABD of MACF1 mediates the interaction between MACF1 and F-actin, and coordinates the MTs/F-actin dynamics to regulate directed cell migration [50]. Although the tandem organization of two CH domains exerts high binding affinity for F-actin, some isoforms of BPAG1 and MACF1 produced by alternative splicing have only the CH2 domain or no CH domain [13,20,51,52] (Figure 1a). The specific function of this alternative splicing is still unknown and requires further study.

Plakin Domain
The plakin domain is one specific feature of the plakin family, and all plakins except epiplakin possess a plakin domain characterized by a high α-helical content. Both protein structure analysis and crystallography determination have demonstrated that the plakin domain consists of two pairs of spectrin repeats and an inserted Src-Homology 3 (SH3) domain [53,54] (Figure 1b).
The plakin domain provides an important platform for protein-protein interactions and makes plakins interact with components of specialized cell junctions, such as hemidesmosomes (HDs, which are specific cell-matrix junctions) and desmosomes (cell-cell junctions). The plakin domain facilitates the interaction of BPAG1 and plectin with HD components, BPAG2 and β4 integrin, in order to strengthen the stable adhesion of epithelial cells to the extracellular matrix [55][56][57]. Moreover, the plakin domain of desmoplakin has been shown to interact with desmosomes proteins, such as plakoglobin, plakophilin, and cadherins [58].

Spectrin Repeats
Spectrin repeats present one main domain of the spectraplakins BPAG1 and MACF1 (Figure 1a). Each spectrin repeat is a 106-122 amino acid segment that folds into three α-helices. These α-helices form an antiparallel triple-helical coiled coil [59,60]. These spectrin repeats compose an extended rod-like structure, which is located between the N-and C-terminal functional domains and acts as a spacer region to confer proteins with flexibility.

Coiled-Coil Rod (CCR)
The CCR domain exists in most mammalian plakins, including BPAG1e, plectin, desmoplakin, envoplakin, and periplakin [1]. The CCR domain is composed of heptad repeats and generally presented between the plakin domain and the PRD (Figure 1a).
Evidence shows that the CCR domain mediates the dimerization of plakins [63]. However, the function of the CCR domain has not been extensively studied and requires further investigation.

Plakin Repeat Domain (PRD)
PRD is a distinctive hallmark of the plakin family. It is a globular domain composed of 4.5 copies of plakin-repeat motif with each repeat motif containing 38 amino acids [64] (Figure 1b). PRDs are classified as three types, A, B, and C [35,64], and are found at the C-terminus of most plakins and at the internal region of some spectraplakin isoforms. However, epiplakin is solely composed of the B-type PRD ( Figure 1a).
PRDs are responsible for binding to various IFs, as demonstrated in BPAG1 [65], plectin [66], desmoplakin [64], envoplakin [67], and epiplakin [68]. Choi et al. solved the crystal structures of Band C-type PRDs of desmoplakin and found a conserved basic groove on the domain, which could serve as an IF-binding site [64]. Fogl et al. further reported the mechanism by which envoplakin bound to IFs by solving the envoplakin PRD crystal structure [67]. They found that a basic groove on the PRD could accommodate acidic patches within IF proteins. In addition, the residues within the groove mediated the targeting of envoplakin to vimentin and keratin IFs [67]. These studies demonstrate that PRDs confer plakins with the ability to link IFs. However, the PRDs in MACF1b have been shown to target MACF1b to the Golgi complex [23]. This suggests that the specific role of different PRD types (A, B, and C) and their different combinations may require further study.

EF-Hand and GAS2-Related Protein (GAR) Domain
Like spectrin repeats, the EF-hand and GAR domain is present in spectraplakins BPAG1 and MACF1 ( Figure 1a). There are two EF-hand motifs followed by a GAR domain at the C-terminus of the spectraplakins [69]. The GAR domain is named after the growth arrest specific 2 (Gas2) and contains 57 amino acids. The crystal structure analysis of the hACF7 EF1-EF2-GAR module reveals that the EF1-EF2 and GAR domains are connected by a flexible linker [70].
The EF1-EF2 domain functions in calcium-binding, and each EF-hand motif presents as an EF β-scaffold with two bound Ca 2+ ions that straddle an N-terminal α helix [70]. The GAR domain is responsible for binding to and stabilizing MTs [21,70]. While the EF1-EF2 domain is not sufficient for MT binding, the GAR domain enhances EF1-EF2-MT engagement [70]. Although the EF-hand motifs of MACF1 have no effect on MACF1 interacting with MTs [21], the EF-hand motifs of BPAG1 have been shown to play an important role in BPAG1-binding MTs [71]. By adopting the eukaryotic expression vector pFLAG-MACF-GAR-GSR that encoded a FLAG-tagged MACF1 C-terminal protein without the EF-hand motifs to transfect the COS-7 cells, Sun et al. found that the overexpressed MACF1 GAR-GSR proteins colocalized perfectly with the MT network, which was similar to the result obtained with the entire MACF1 C-terminal protein (containing EF-hand motifs). This finding indicates that the EF-hand motifs of MACF1 have no effect on the interaction between MACF1 and MTs [21]. However, Kapur et al. found that mutations of the EF-hands abolished the switch between the MT plus end interaction and the MT lattice binding of BPAG1n4, suggesting the importance of EF-hand motifs in BPAG1-binding MTs [71].

Mammalian Plakins in Cell Migration
Cell migration is a multistep process consisting mainly of four coordinated steps: (1) protrusion of a cell's leading edge; (2) adhesion to the substrate; (3) contraction of the cell body; and (4) the turnover of focal adhesions (FAs) and retraction. Each step contains the dynamic changes of the F-actin and MTs cytoskeleton. Both F-actin and MTs are polymer structures that can remodel themselves to facilitate cell migration. In cells, actin exists either in monomeric (G-actin) or polymeric forms (F-actin). F-actin is comprised of double helical polymers that are formed by G-actin. MTs consist of α/β-tubulin heterodimers that selfassociate into polymers. F-actin and MTs show structural polarity with two ends, a plus end and a minus end. They undergo fast assembly and disassembly dynamics during cell migration [72,73]. Evidence shows that F-actin plays a key role in stabilizing and guiding MTs on their trek from the microtubule organizing center (MTOC) to the cell periphery [74,75]. Moreover, Salmon et al. have reported that MTs grow along F-actin bundles that guide the movement and organization of MTs during cell migration [76]. They suggest that the guidance of MTs along F-actin bundles may be mediated by some MTs/F-actin crosslinking factor [76].
Plakins, as important crosslinkers for cytoskeleton, show a critical role in regulating cell migration by orchestrating the organization and dynamics of F-actin, MTs, and IFs, as well as the interaction between the cytoskeleton and junctional complexes ( Figure 2). The knockout of BPAG1 inhibits the epidermal cell migration by severing the attachment between keratin IFs and HDs [77]. Moreover, the knockdown of BPAG1a and BPAG1b causes a specific decrease in the directness of cell migration in myoblasts possibly by increasing dispersal of the Golgi apparatus, which influences MT orientation toward the cell's leading edge [78]. BPAG1e deficiency leads to a loss of cell polarity and an aberrant cell migration in keratinocytes, through the reduction of the β4 integrin-mediated activity of Rac1 and cofilin [79]. In addition, a BPAG1e deficiency causes the rearrangement of the actin cytoskeleton and the changes in the focal contact protein distribution, which work together with Rac1 and cofilin to form lamellipodia [79]. However, the keratinocytes isolated from patients carrying homozygous nonsense mutations in the BPAG1e gene display increased cell migration, along with abnormal levels of keratin-14, and β4 and β1 integrins [80]. Further BPAG1e knockdown studies of normal keratinocytes in vitro failed to recapitulate the above changes in vivo, suggesting that the incomplete depletion of BPAG1e is not sufficient to induce the in vivo phenotype [80]. These studies suggest that BPAG1 is needed to regulate cell migration. Moreover, it is demonstrated that different BPAG1 isoforms exert either promotion or inhibition effects on cell migration possibly via the following mechanisms: (1) mediating the connections between keratin IFs and HDs; (2) maintaining the Golgi apparatus structure that influences MT orientation toward the leading edge of the cell; (3) mediating β4-integrin-mediated activation of Rac1 and cofilin; and (4) controlling adhesion stability.
MACF1 is critical in regulating cell migration because it coordinates the dynamics of F-actin, MTs, and FAs and mediates GSK-3β signaling [81][82][83]. By binding to both F-actin and MTs, MACF1 guides MTs along F-actin to target FAs, thus regulating epidermal cell migration [81]. A loss of MACF1 suppresses cell migration by impairing the targeting of MTs to FAs and by stabilizing FA-actin networks in epidermal cells [81]. An MACF1 absence also inhibits cell migration by impairing the association of MACF1 with GSK-3β signaling in both skin stem cells and neurons [82,83]. An abnormal interneuron migration was also observed in MACF1 conditional knockout mice [84]. In addition, MACF1 regulates cell migration by being recruited by some molecules, such as ErbB2 and engulfment and motility (ELMO) [85,86].
Plectin shows the capacity to bind to F-actin, IFs, and MTs, acting as a regulator of the cytoskeleton dynamics [87][88][89]. Plectin deficiency impairs stress fiber and FA dynamics by increasing the number of the actin stress fibers and FAs [87], inhibits vimentin network dynamics and the stepwise formation of stable IFs [88], and decreases MT dynamics [89]. Additionally, plectin mediates the crosstalk between vimentin and actin networks, and plectin deficiency destroys the shape of adherens junctions and tight junctions [90]. By regulating these cytoskeletal dynamics and cell adhesions, plectin regulates cell migration in a cell-type-dependent manner. A loss of plectin decreases cell migration in lymphocytes [91] but increases cell migration in keratinocytes [92]. Moreover, plectin deficiency increases cell migration by activating Rac1 and FAK in normal human hepatic cells [93]. This indicates plectin's cell-type-dependent effects on cell migration, and this effect requires further verification in more cell types.
Furthermore, periplakin and epiplakin have also been demonstrated to regulate epithelial cell migration by controlling cytoskeleton reorganization or the expression of adhesion and cytoskeletal proteins [94,95]. Moreover, periplakin has been shown to regulate epithelial migration by interacting with plectin [96], suggesting that crosstalk exists between different mammalian plakins and that they may work together to regulate cell migration.

Mammalian Plakins in Cell Proliferation
Evidence has shown the involvement of mammalian plakins in cell proliferation, especially in cancer cells (this is discussed in further detail in Section 4), although extensive studies are needed.
Our studies show that MACF1 regulates cell proliferation of osteoblasts. By adopting a stable MACF1-knockdown MC3T3-E1 osteoblastic cell line, we found that the knockdown of MACF1 significantly inhibited cell proliferation by disrupting the normal organization of F-actin and MTs and by inducing the cell cycle arrest at the S phase [97]. In normal cells, F-actin was distributed at the cell periphery or passed through the cytoplasm as rope-like structures, and MTs radiated from the MTOC to the cell membrane and presented as tortuous filaments. Meanwhile, in the MACF1-knockdown cells, the F-actin filaments were mainly localized at the cell periphery, and MTs became straighter filaments with a loss of the MTOC. The disorganization of F-actin and MTs may cause defective cytokinesis that results in the cell cycle arrest at the S phase, which leads to decreased cell proliferation in MACF1-knockdown cells [97]. We further determined the effect of MACF1 on cell proliferation by transfecting the MC3T3-E1 osteoblastic cells and MACF1-knockdown cells with MACF1 overexpression plasmid, respectively. The results showed that the overexpression of MACF1 increased cell proliferation in MC3T3-E1 cells and partially rescued the decrease of cell proliferation in MACF1-knockdown cells [98]. Thus, our findings indicate the promotion effect of MACF1 on cell proliferation. However, Fuchs' group reported that a lack of MACF1 did not cause a significant decrease in cell proliferation nor defective mitosis in epidermal and endodermal cells [50,81,82]. These different findings may be due to the different cell types, suggesting that more studies of different cell types are needed.
Plectin has been indirectly shown to regulate cell proliferation by forming complexes with integrin β4 [99]. Jeon et al. found that the reduction of the integrin β4/plectin complex induced by glucosamine promoted cell proliferation in mouse embryonic stem cells. They further revealed that the decrease of the β4/plectin complex was due to the increase of protein kinase C (PKC) phosphorylation, as well as specificity protein 1 (Sp1) and Snail1 glycosylation [99]. Although little is known about the role of plectin in cell proliferation, this finding paves a way for further studies.
More recently, Kokado et al. reported that the knockdown of epiplakin in corneal epithelial cells significantly accelerated cell proliferation, suggesting a negative regulatory role of epiplakin on cell proliferation [95].
The above findings on MACF1, plectin, and epiplakin illustrate the function of mammalian plakins in cell proliferation.

Mammalian Plakins in Cell Signaling
Besides crosslinking F-actin, MTs, IFs, and junctional complexes, plakins also interact with signaling proteins and possess phosphorylation sites, thus participating in cell signaling transduction to regulate cellular functions, such as cell migration, cell proliferation, and cell differentiation.
BPAG1e has been demonstrated to be a scaffold for Rac1/cofilin signal transduction [79]. The knockdown of BPAG1e dramatically decreased the amount of Rac1 associating with β4 integrin, as well as the activity of both Rac1 and cofilin [79]. By determining the activation of the Rac1/cofilin signal, BPAG1e is essential for the efficient regulation of keratinocyte polarity and migration.
MACF1 possesses phosphorylation sites for GSK-3β at the C-terminus and mediates GSK-3β signaling [82,83]. Moreover, MACF1 is essential for Wnt/β-catenin signaling transduction by interacting with the Axin complex that contained Axin, GSK-3β, β-catenin, and APC, and by translocating the complex from cytoplasm to the cell membrane [4]. Upon Wnt stimulation, MACF1 forms a complex in the cytosol with Axin, GSK-3β, β-catenin, and APC, and translocates the Axin complex (except for APC) to the cell membrane, where they interact with low-density lipoprotein receptor 5/6 (LRP5/6). Subsequently, GSK-3β is inactivated, and β-catenin is released and translocated into the nucleus to activate target genes. However, the loss of MACF1 impairs the Axin complex translocation, inhibits the nuclear translocation of β-catenin, and thus suppresses the downstream T-cell factor (TCF)/β-catenin-dependent transcriptional activation [4]. Our studies further reveal that MACF1 promotes osteoblast differentiation and bone formation by activating β-catenin/TCF1-dependent osteoblastic genes expression [100,101]. These findings indicate the importance of MACF1 in mediating cell signaling transduction.
Like MACF1, plectin also has phosphorylation sites in its C-terminal region and serves as a target for protein kinases, such as p34 cdc2 kinase and protein kinase A (PKA) [102,103]. The phosphorylation of plectin by PKA or mitogen-activated protein kinase-interacting kinase 2 (MNK2) attenuates the interaction between plectin and IFs, thus remodeling IF networks and the cell function [104]. More recently, Matsubara et al. found that plectin was phosphorylated by tyrosine kinase Src to regulate the actin ring formation and osteoclast differentiation [105]. Moreover, plectin modulates several signaling pathways, such as ERK1/2 signaling and adenosine 5 -monophosphate-activated protein kinase (AMPK) signaling [5,106]. These indicate the key role of plectin in cell signaling, where it either acts as a substrate for kinases or modulates the signaling by itself.
Desmoplakin and periplakin also function in cell signaling transduction. Similar to MACF1 and plectin, desmoplakin participates in cell signaling through either being post-translationally modified by phosphorylation or by directly regulating the signaling pathway. Desmoplakin is phosphorylated by glycogen synthase kinase 3 (GSK3) to control desmoplakin-cytoskeleton dynamics, and it also activates AKT-Wnt/β-catenin signaling [107,108]. Periplakin has been shown to act as a scaffold for the modulation of protein kinase B (PKB)-dependent signaling [6]. In addition, it has been identified as the first protein to modify the melanin-concentrating hormone-1 (MCH-1) receptor-mediated signaling transduction [109]. By binding to the MCH-1 receptor, periplakin reduces the capacity of the receptor to activate G proteins and hence initiate signal transduction [109].
All of the above findings show the importance of mammalian plakins in cell signaling transduction. They are similar in that they either act as substrates for post-translational modification or regulate signaling pathways by themselves.

The Role of Mammalian Plakins in Cancer
With versatility in biological functions closely related to cancer, the role of mammalian plakins in various cancers has recently been shown, as discussed below.

Bullous Pemphigoid Antigen 1 (BPAG1) in Cancer
BPAG1 has recently been linked to cancer. The decrease of BPAG1 was observed in invasive breast cancer cells, nasopharyngeal carcinoma cells, and metastatic prostate cancer [110][111][112]. Since BPAG1 is a major component of HD that mediates the attachment of epithelial cells to the basement membrane, the decrease of BPAG1 may facilitate cancer cell invasion and metastasis. However, Chaudhari et al. found that BPAG1e positively regulated the cell motility of oral squamous cell carcinoma (OSCC)-derived cells [113]. The knockdown of both BPAG1e and plectin, which are two HD linker proteins, inhibits the OSCC-derived cell motility, invasion, and tumorigenicity, possibly by upregulating the N-Myc downstream regulated gene 1 (NDRG1) [114]. These inconsistent findings indicate that BPAG1 may function differently in different cancers or that there are cancer-specific BPAG1 isoforms. The discovery that the alternative splicing of the dystonin (DST) gene (encoding for BPAG1) occurs in head and neck squamous cell carcinoma (HNSCC) provides a basis for the existence of cancer-specific BPAG1 isoforms [115]. Based on current findings (Table 2), more studies are needed to unravel the role of BPAG1 in cancer, as well as related mechanisms.

Microtubule Actin Crosslinking Factor 1 (MACF1) in Cancer
MACF1 has recently received increasing attention for its role in various cancers (Table 2). It was identified as a cancer biomarker candidate for lung cancer by an integrated genome-scale co-expression network [116]. The upregulation of MACF1 was detected in lung adenocarcinoma with a connection to metastasis and migration [116]. Further experiments directly validated the high level of MACF1 in several lung cancers, and the knockdown of MACF1 inhibited the reproductivity of solid tumors [117]. MACF1 was also linked to breast cancer by a genome-wide DNA methylation profiles analysis and was speculated to regulate cell motility [118]. Del Valle et al. further reported the different levels of MACF1 in stromal cells in different tumor sites of breast cancer patients [119]. Moreover, the mutation of MACF1 in renal cell carcinoma [120], endometrial cancer [121], and colon cancer [122] has been detected by whole-exome sequencing and targeted gene sequencing. It has also been shown that the MACF1 mutation was correlated with Wnt/β-catenin signaling in cancer [120][121][122]. While most findings indirectly show the role of MACF1 in cancers, Quick's group was the first to provide direct evidence of MACF1's function in glioblastoma [123]. They found that MACF1 was predominately expressed in glioblastomas but not in either normal brain tissue or lower-grade brain tumors. Further suppression of MACF1 in glioblastoma cells inhibited cell proliferation and migration associated with reduced Axin1 and β-catenin, which are Wnt signaling mediators [123]. As MACF1 is critical in mediating the Wnt signaling pathway [4], it is suggested that MACF1 may function in cancers through Wnt signaling. In addition, MACF1 has also been shown to be a target of microRNA in both hepatocellular carcinoma and malignant tumors [124,125]. All these findings demonstrate that MACF1 is a promising target for cancer diagnosis and therapy [126]. Colon cancer High expression, promotes cancer cell migration, invasion and metastasis. Be targeted to podosome-like adhesions in an isoform-specific manner [130] Bladder cancer High expression, promotes cancer cell migration and metastasis. Anchor invadopodia to IF and stabilizing invadopodia [131] OSCC High expression, promotes cancer cell migration, invasion and tumorigenicity. Upregulate NDRG1 [114] Liver cancer Low expression, inhibits cancer cell migration. Unclear [132] Desmoplakin OSCC Downregulation. Unclear [133] Oropharyngeal cancer Downregulation. Unclear [134] Lung cancer Downregulation, suppresses cell proliferation, migration and invasion, increases cancer sensitivity to anticancer drug-induced apoptosis. Inhibits Wnt/β-catenin signaling [135] Envoplakin Skin cancer Promotes skin cancer development. Unclear [136] Periplakin Pharyngeal squamous cancer Promotion effect on cell movement and attachment. Activates the PI3K/Akt axis [137] Urinary bladder cancer Downregulation. Unclear [138] Esophageal squamous cancer Downregulation, inhibits cell migration. Mislocation of periplakin from cell-cell boundaries to cytoplasm; DNA hypermethylation [139,140] Colon cancer Downregulation, inhibits cell proliferation, migration, and invasion, induces G1/G0 cell cycle arrest.

Plectin in Cancer
Plectin has been identified as an excellent biomarker for pancreatic intraductal papillary mucinous neoplasms (IPMNs) and as a novel biomarker for pancreatic cancer [127,128]. High levels of plectin have also been observed in HNSCC [129], colon carcinoma cells [130], and high-metastatic bladder cancer cells [131], showing promotion effects on cancer cell migration, invasion, and metastasis. The suppression of plectin inhibits cell migration and invasion in both HNSCC and colon carcinoma cells [129,130]. In HNSCC, plectin has been shown to promote the migration and invasion of HNSCC cells through activation of ERK1/2 [129], while in colon carcinoma cells plectin is targeted to podosome-like adhesions to regulate cell migration and invasion in an isoform-specific manner [130]. Moreover, in high-metastatic bladder cancer, plectin plays a key role in anchoring invadopodia to IF and in stabilizing invadopodia, which is critical for cancer cell invasion [131]. This all demonstrates plectin's promotion effect on cancer cell migration and cancer progression. Chaudhari et al. also found that the inhibition of plectin reduced cell migration, invasion, and tumorigenicity in OSCC-derived cells [114]. However, a recent study showed a low expression of plectin in liver cancer cells, accompanied by a high cell migration rate, suggesting an inhibitory effect of plectin on liver cancer cell migration [132]. As plectin regulates cell migration in a cell-type-dependent manner, as discussed above (Section 3.1), it may also regulate cancer cell migration in this way. In any case, plectin is a highly reliable biomarker and a potential target for various cancer diagnoses and therapies ( Table 2).

Desmoplakin in Cancer
Desmoplakin has been shown to be a tumor suppressor with decreased expression observed in several cancers. Decreased desmoplakin has been detected in oral cancer, including OSCC and metastatic oropharyngeal cancer, and has been shown to be a useful marker for evaluating changes in tissue morphology, as well as the risk of distant metastasis formation [133,134]. Interestingly, the desmoplakin II isoform may play a specific function in oral carcinogenesis because its expression is only detected in tumors associated with a poor clinical outcome [134]. Furthermore, Yang et al. provide direct evidence that desmoplakin functions as a tumor suppressor [135]. They detected a downregulation of desmoplakin in lung cancer and found that an overexpression of desmoplakin inhibited lung cancer cell proliferation, migration, and invasion and that it increased the sensitivity of lung cancer cells to anticancer-drug-induced apoptosis through the inhibition of Wnt/β-catenin signaling [135]. All of this shows desmoplakin's role as a tumor suppressor, with potential implications for clinical applications.

Other Mammalian Plakins in Cancer
More recently, envoplakin and periplakin, functioning as epidermal barrier components, have also attracted attention in cancer development (Table 2). Cipolat et al. found that mice that were triply deficient in envoplakin, periplakin, and involucrin (EPI−/− mice) were highly resistant to skin tumor development [136], indirectly suggesting a positive role of envoplakin and periplakin in skin cancer development. Tonoike et al. also showed that periplakin knockdown reduced cell movement and attachment via a suppression of the phosphatidylinositol 3 kinase (PI3K)/Akt axis in pharyngeal squamous cancer cells [137]. Interestingly, a downregulation of periplakin was observed in urinary bladder cancer [138], esophageal squamous cancer [139,140], and colon carcinoma [141], pointing to periplakin's role as a cancer suppressor. The cancer-suppressive role of periplakin was further confirmed by Li et al. [141]. They reported that an overexpression of periplakin reduced cancer cell migration and invasion, induced a G1/G0 cell cycle arrest, and inhibited colon cancer cell proliferation by suppressing the activation of ERK1/2 and the expression of proliferating cell nuclear antigen (PCNA). Meanwhile, a periplakin knockdown showed contrary effects [141]. Moreover, Otsubo et al. indicated that the decreased periplakin was due to DNA hypermethylation, which contributes to the metastatic phenotype of cancer [140]. Additionally, the dislocation of periplakin from the cell-cell boundaries to the cytoplasm may promote cell invasion and cause tumor development [139]. All of this points to periplakin being a cancer suppressor for numerous cancer types, although there are some contrary findings. Compared to other plakins, epiplakin's function and role in cancer are still largely unknown. As a plakin that is solely composed of PRD, and that is therefore unusual, epiplakin is only found to be downregulated in pancreatic cancer [43] (Table 2). Its function and mechanisms in cancer require further investigation.
These reports demonstrate the critical role of mammalian plakins in various human cancers (Table 2). Moreover, we investigate the expression levels of mammalian plakins in a broader set of cancers by using the Oncomine, a web-based microarray database (http://www.oncomine.org). As shown in Figure 3 [110,112] (Table 2). Similar to BPAG1, in most of the datasets, MACF1 mRNA levels are significantly decreased in cancer as opposed to normal tissues ( Figure 3). Interestingly, contrary to the findings of Bidkhori et al. [116,117] (Table 2), 10 unique analyses show that the MACF1 mRNA levels are significantly reduced in lung cancer (Figure 3). This difference may require further study. Downexpressions of EVPL and PPL can also be found in the majority of datasets (26 for EVPL and 51 for PPL) across the cancer types. For PLEC and EPPK1, 19 datasets show increased expression in cancers, while only 8 datasets and 7 datasets demonstrate decreased expression of PLEC and EPPK1 in cancers, respectively ( Figure 3). DSP displays significant overexpression in cancers in 26 datasets and downexpression in 24 datasets. Overexpression and downexpression of mammalian plakins occur across the multiple cancer types, indicated by Oncomine data. However, the expression trend of mammalian plakins in some cancer type is contrary to the literature reports (Table 2, Figure 3). These differences require further determination.

Mammalian Plakins as Potential Targets for Cancer Therapy
The above findings demonstrate the importance of mammalian plakins in cancer, suggesting them as potential targets for cancer therapy. Several evidences show the potential of mammalian plakins as promising targets for cancer therapy, though there is no drug or pharmacological agent that specifically targets them at present. Lorch et al. have reported that treatment with the proteasome inhibitor bortezomib in OSCC results in significant increased desmoplakin and reduced cell migration [142]. As the downregulation of desmoplakin has been found in OSCC [133] (Table 2), the finding of Lorch et al. suggests desmoplakin as a potential target for cancer therapy. More recently, Quick, by using the Research Collaboratory for Structural Bioinformatics (RCSB) protein data bank, has identified several anticancer drugs that interact with BPAG1, MACF1, and plectin [143]. Ibrutinib is one common ligand for both BPAG1 and MACF1, and is used clinically for the treatment of lymphoma and leukemia [144,145]. PP121, showing antitumorigenic effects on esophageal cancer and anaplastic thyroid carcinoma by perturbing PI3K and mammalian target of rapamycin (mTOR) kinase activities [146,147], is identified as an MACF1 ligand. Moreover, plectin is determined to be a substrate for sparsomycin, pactamycin, amicoumacin A, and omacetaxine mepesuccinate, all of which show antitumorigenic effects in leukemia, HNSCC, breast cancer, or lung cancer [148][149][150][151].
Therefore, mammalian plakins play key roles in various human cancers and could be novel biomarkers and promising targets for cancer diagnosis and therapy.

Conclusions and Perspectives
Cancer is one of the most lethal diseases in the world. It is characterized by an uncontrolled cell proliferation, migration, and adhesion, which are closely related to the dynamic changes of the cytoskeleton and cytoskeletal-adhesion. Furthermore, the uncontrolled cell proliferation, migration, and adhesion will cause cell invasion and metastasis, which make cancer malignant and lethal. Unraveling the underlying mechanisms of aberrant cell proliferation, migration, and adhesion is therefore important not only for understanding cancer development but also for providing new targets for cancer diagnosis and therapy.
Plakins are a family of giant cytolinkers that play important roles in cell proliferation, migration, and signaling transduction by modulating the dynamics of F-actin, MTs, IFs, and FAs and by regulating signaling pathways. There are seven mammalian plakins: BPAG1, MACF1, plectin, desmoplakin, envoplakin, periplakin, and epiplakin. They are broadly expressed in numerous tissues. Most of them (BPAG1, MACF1, plectin, and desmoplakin) have isoforms, and all of them have multidomain structures that enable them to connect with different cytoskeletal and signaling molecules. Because they possess various isoforms, structures, and tissue distributions, mammalian plakins are critical in maintaining normal cell proliferation, migration, and signaling transduction. In regulating these cellular processes, mammalian plakins either coordinate the cytoskeletal-adhesion dynamics (e.g., cytoskeleton-HDs) or adjust signaling pathways, such as GSK-3β signaling, Wnt/β-catenin signaling, Rac1/cofilin signaling, ERK1/2 signaling, and AMPK signaling.
Mammalian plakins show their significance in human cancers by showing versatility in their regulation of cell proliferation, migration, and signaling transduction, key steps for cancer development, and by showing potential to be targets for anticancer drugs. It has been demonstrated that the altered expression or mutation of mammalian plakins is closely related to numerous cancers, such as breast cancer, prostate cancer, oral cancer, lung cancer, endometrial cancer, colon cancer, glioblastoma, lung cancer, liver cancer, skin cancer, pancreatic cancer, and bladder cancer. Moreover, the different expression levels of mammalian plakins in many cancer types have been indicated by the Oncomine database. Mammalian plakins participate in cancer development and progression by regulating cell proliferation, migration, and invasion. To study the mechanisms, the altered cytoskeleton dynamics, along with certain signaling pathways such as Wnt/β-catenin, ERK1/2, and the PI3K/Akt axis have been shown to be involved. However, the exact mechanisms are still largely unknown. This is due to the fact that the available literature on the direct roles and associated mechanisms of plakins in human cancers is still scanty. Moreover, there are inconsistent findings on the role of some plakins in different cancers. Thus, more work needs to be conducted to answer the following questions: (1) What is the exact function of specific plakins in specific cancers? (2) What is the mechanism of specific plakins in specific cancer development and progression? (3) As most plakins have many different isoforms, is there a cancer-specific plakin isoform? (4) As our studies show the crucial function of MACF1 in osteoblasts and bone, are plakins involved in bone cancer? Answering these questions will increase the appeal of mammalian plakins to cancer research. Furthermore, the potential of mammalian plakins to be targets for cancer therapy makes them more attractive in future cancer research.
In summary, mammalian plakins, as giant cytolinkers that crosslink the three main cytoskeletal elements and junctional complexes, show versatility in regulating cell proliferation, migration, and signaling transduction. The alteration of mammalian plakins results in aberrant cell proliferation, migration, and signaling transduction, which causes cancer (Figure 4). Given their versatile biological functions, their critical roles in numerous cancers, and their potential to be targets for cancer therapy, mammalian plakins will therefore be novel biomarkers and promising targets for cancer diagnosis and therapy ( Figure 4).