Fascin is a monomeric actin-bundling protein that was first isolated from sea urchin coelomocytes, later isolated as 55-kDa actin-bundling protein from HeLa cells (see for review, Jayo and Parsons 2010). It has a molecular weight of 55–58,000 kDa and makes extensive actin bundles in vitro with uniform polarity. The cloning of sea urchin fascin by Bryan et al. in 1993 revealed that fascin is homologous to the Drosophila singed gene product (Bryan et al. 1993). In 1994, Duh et al. and Holthuis et al. isolated human and Xenopus homologs of fascin, respectively (Duh et al. 1994; Holthuis et al. 1994). The sequence analyses revealed that fascin proteins form a unique family of actin-bundling proteins, sharing no apparent homology with nonfascin actin-bundling proteins including alpha-actinin, villin, and fimbrin.
Vertebrates have three fascin genes (fascin-1 through −3): fascin-1 shows widespread expression in a variety of tissues, whereas expression of fascin-2 is restricted to retina and hair cell stereocilia, whereas fascin-3 is restricted to testis. Fascin is absent in Dictyostelium discoideum, C. elegans, or yeasts. As expected from the actin-bundling activity, fascin is mainly localized in filopodia of mammalian cultured cells, microvilli of sea urchin eggs, and drosophila bristles. Consistent with highly motile structure of filopodia, fascin is shown to promote cell motility and metastasis of tumor cells. Intriguingly, a fascin homologue is found in microvilli and filopodia of choanoflagellates. Because this unicellular organism is considered as the last ancestor of multicellular animals, fascin1 is suggested to be an ancestral component for filopodial assembly machinery (Sebe-Pedros et al. 2013).
Structure and Function of Fascin
X-ray structural analyses revealed that fascin has a β-trefoil structure (Sedeh et al. 2010), showing structural similarity with other β-trefoil proteins including fibroblast growth factors, interleukin-1, Kunitz soybean trypsin inhibitors, ricin-like toxins, plant agglutinins, and hisactophilin. Significance of the similarity is not clear because they are not related to each other. Among them, only hisactophilin, 17-kDa protein isolated from Dictyostelium, is an actin-binding protein. The protein, however, does not bundle actin filaments, rather it is reported to function as a link between actin and the cortex.
Fascin1 plays an important role in the formation of filopodia in mammalian cells. In vitro, fascin1 makes parallel actin bundles with uniform polarity, which is consistent with its localization in filopodia. Upregulation of fascin1 induces membrane protrusions and increases cell motility of epithelial cells, as well as colonic epithelial and carcinoma cells. Conversely, Fascin1 knock down has been reported to block filopodia assembly of B16F1 mouse melanoma cells, as well as colon carcinoma cells and mature antigen-presenting dendritic cells (DCs) (Ross et al. 1998).
In addition to binding to actin filaments, fascin1 has been recently reported to bind directly to microtubules (Villari et al. 2015). Fascin1-microtubule binding occurred independently of fascin1-actin binding. The association was shown to increase the dynamics of focal adhesions, as well as that of microtubules, thereby controlling cell motility. This regulation of focal adhesion dynamics may be controlled via FAK because fascin1 was found to bind to FAK.
Analyses of Drosophila singed mutations indicate that fascin1 is involved in female sterility, in addition to the gnarled bristle phenotype. In Drosophila oogenesis, each developing oocyte is surrounded by and connected to 15 nurse cells via intercellular bridges. Nurse cell cytoplasmic contents flow into the oocyte along actin filaments traversing these cytoplasmic bridges. A singed allele affects the microfilament structure required for this nurse cell cytoplasmic flow, resulting in female sterility (Cant et al. 1994). Interestingly, fascin1 is involved in prostaglandin-mediated actin remodeling (Groen et al. 2012). Prostaglandin is synthesized in Drosophila by peroxidase (Pxt), a cyclooxygenase (COX)-like enzyme. Like singed mutation, pxt mutation caused a defect in actin remodeling in nurse cells, resulting in female sterility. Furthermore, pxt mutation enhanced the singed phenotype of female sterility, while overexpression of fascin suppressed the pxt phenotype. These results suggest that fascin is a downstream target of prostaglandin signaling.
Analyses of fascin1 knockout mice revealed that fascin1 is not an essential protein for mouse development (Yamakita et al. 2009). Fascin1 KO mice are viable and fertile with no apparent developmental defects though they are less favorable for neonatal survival. Brain anatomy is grossly normal except that fascin1 KO brain shows larger lateral ventricle and lacks posterior extension of anterior commissure neuron. Perhaps, the lack of fascin1 is compensated by other actin-binding proteins in the case of mouse development.
Regulation of Fascin
PKC phosphorylates fascin1 at a conserved site of Ser39 (human fascin1 sequence) and reduces its actin-bundling activity (Ono et al. 1997). Cell-matrix adhesion changes fascin1 phosphorylation, depending on the types of extracellular matrix (Adams et al. 1999). When C2C12 myoblasts were adhered to fibronectin, PKC-dependent phosphorylation of fascin was observed, resulting in diffuse localization of fascin1. This phosphorylation event was not observed with adhesion to thrombospondin-1 or to laminin-1.
The binding of fascin1 to PKC depends both on the phosphorylation state of fascin1 and an activation state of PKC (Hashimoto et al. 2007): While active PKC binds to fascin1, a Ser39 phosphomimetic mutant fascin1 binds to both active and inactive PKC. PKC-fascin interaction appears to be required for Rac-dependent motility of human colon carcinoma cells. Interestingly, expression of either Ser39 phosphomimetic mutant or unphosphorylatable mutant of fascin1 did not increase metastatic tumor development, suggesting that cycling of phosphorylation and dephosphorylation of fascin1 is required for fascin1-mediated enhanced metastasis.
Fascin has been demonstrated to bind directly to an active form (GTP-bound) of Rab35, a small GTPase, involved in endocytic recycling (Zhang et al. 2009). This binding allows Rab35 to target fascin1 to the plasma membrane, thereby generating filopodia: A constitutively active mutant of Rab35 generated many filopodia, while the dominant negative mutant blocked membrane protrusions. Rab35 has been shown to recruit both Rac1 and Cdc42 to the plasma membrane (Shim et al. 2010). Thus, the fascin-Rab35 binding would recruit fascin to the plasma membranes where Cdc42 and Rac1 coordinate with fascin to generate filopodia and lamellipodia, respectively.
The association between fascin1 and the membrane receptors including p75 neurotrophin receptor and protocadherin α may also control the targeting of fascin1 to the plasma membranes. For example, NGF treatment increased the association between fascin and p75, which was found to be essential for NGF-dependent migration of melanoma. Furthermore, the expression of a Ser39 phosphomimetic mutant of fascin1 abrogated NGF-induced migration, suggesting dephosphorylation of fascin1 is critical for invasiveness of melanoma cells (Shonukan et al. 2003).
Fascin1 is shown to bind to LC3 both in vitro and in vivo (Matsumura et al. 2013). LC3 is a critical component of autophagosomes. GST-pull down assays revealed that fascin1 directly binds to LC3. Proximity Ligation Assays showed that LC3 and fascin1 are in close contact inside cells. While the significance of LC3-fascin1 association has not been determined, this association may be important for xenophagic activity in antigen-presenting dendritic cells (DCs).
It appears that fascin1 is phosphorylated at another site (Ser289 in Drosophila, corresponding to Ser274 in human) (Zanet et al. 2012). Unlike the Ser39A mutant, a S289A mutant lost both actin-bundling and actin-binding activities. However, the expression of the S289A mutant is able to rescue female sterility, one of the phenotypes caused by singed mutation. This mutant was not found along cytoplasmic actin bundles but at the tips of bundles. The authors suggest that the association of this mutant at the distal end of filopodia stabilizes bundle formation by an actin-independent mechanism.
A number of laboratories have reported that fascin-1 expression is increased in many human carcinomas including T- or B-cell lymphoma, breast, pancreatic, lung, colorectal, gastric, and esophageal carcinomas and that fascin1 upregulation is correlated in many cases with metastasis and cancer grade. Meta-analyses revealed that fascin1 is clearly associated with increased risk of mortality with breast, colorectal, and esophageal carcinomas (Hashimoto et al. 2005). The association of fascin1 upregulation and cancer risk is consistent with the results that forced expression of fascin1-increased metastatic activity of many types of cancer cells in vivo. Thus, fascin1 is not only an excellent marker for cancer diagnosis but also a target of cancer therapy.
Fascin1 has also been reported to play a role in tumor self-seeding. It was previously believed that circulating tumor cells are naturally destined for metastasis and may not return to the original tumor site. Brunhuber et al. have shown that circulating tumor cells do return to the origin and proliferate there (called tumor self-seeding) (Brunhuber et al. 2008). Tumor self-seeding is critical for tumor development because self-seeded tumor cells increase tumor size, and enhance angiogenesis, as well as stromal cell recruitment. Attraction of circulating tumor cells to the origin is controlled by chemokines of IL-6 and IL-8, as well as fascin1 and matrix metalloprotease-1. The requirement of fascin1 and IL-6 is intriguing because IL-6 is known to increase fascin1 expression, as well as cell motility (Li et al. 2010). Furthermore, fascin1 has been reported to increase secretion of IL-6 and TNFα in macrophage cell lines (RAW264.7 and THP-1) by enhancing PKC-mediated translational activity (Kim et al. 2011). It is thus possible that fascin1 and IL-6 form a positive feedback for enhancing cancer cell migration.
The almost ubiquitous upregulation of fascin1 in a variety of cancer cells has prompted many researchers to examine which transcriptional regulators activate fascin1 expression in cancer cells. Several transcriptional regulators including NF-κB, Stat3, TGF-β/Smad4, slug, CREB, AhR, and AP1 have been reported to activate the promoter of the fascin1 gene. It appears that activation of the fascin1 promoter depends on cell types and physiological conditions.
Roles of Fascin1 in Antigen-Presenting Dendritic Cell (DC) Physiology
Fascin1 expression is induced to a great extent upon maturation of DCs, whereas no fascin1 expression is detected in immature DCs. DCs play central roles in innate and acquired immunity. When DCs encounter pathogens, they undergo terminal differentiation called maturation, which changes their functions from antigen sampling to antigen presentation. During maturation, DCs show massive alterations in their morphology by producing numerous dorsal ruffling, as well as high-speed motility toward chemokines, CCL19/CCL21.
Analyses of bone marrow-derived DCs (BM-DCs) isolated from fascin1 KO mice revealed the roles of fascin1 in the development of dendritic cell morphology, as well as motility (see for review, Yamashiro 2012). Fascin1–KO BM-DCs were thinner and more widespread with fewer and smaller dorsal ruffling membranes than wild-type counterparts. As expected from these morphologic alterations, fascin-1 KO BM-DCs exhibited less dynamic membrane protrusion and retraction. Fascin1 was also found to promote chemotaxis toward CCL19 both in vivo and in vitro.
Analyses of fascin1 KO BM-DCs also revealed that fascin1 is critical for podosome disassembly. Podosomes are specialized cell-to-matrix adhesion structure, which are found in many immunological cells including immature BM-DCs and macrophages. Maturation of wild-type DCs resulted in the loss of podosomes. However, fascin1 KO BM-DCs failed to disassemble podosomes. Introduction of fascin1 resulted in podosome disassembly in mouse BM-DCs, as well as in THP-1 cells, indicating that fascin1 is directly involved in podosome disassembly (Yamakita et al. 2011).
The loss of podosomes upon DC maturation is a key event for the changes in migration patterns of DCs upon migration. Immature DCs require podosomes to attach to the extracellular matrix so that they can move around the peripheral tissues, which is advantageous for sampling of foreign and host antigens. On the other hand, mature DCs need to immigrate from the peripheral to draining lymph nodes as quickly as possible in order to transfer pathogen information to naïve T-cells. The loss of podosomes would result in low adhesion to the substrate, leading to “high speed” migration shown by mature DCs.
The role of fascin1 in podosome disassembly in DCs, however, is not consistent with two other reports that fascin1 favors the assembly of podosomes in other types of cells: Fascin1 knock down suppressed podosome assembly in smooth muscle cells and destabilizes invadopodia (podosome-like structure) of melanoma cells. While the reason for the discrepancy is currently unknown, it is possible that extremely high expression observed in DCs would inhibit podosome assembly, while fascin1 expression at moderate levels in smooth muscle cells or melanoma cells may favor podosome assembly (Quintavalle et al. 2010, Li et al. 2010).
Fascin1 has been reported to be essential for the assembly of the immunological synapse (IS) between DCs and T-cells in an allogeneic combination of DC and T-cells (Al-Alwan et al. 2001). Fascin1 and actin are both localized to the IS in a DC side. Fascin1 downregulation was reported to inhibit IS assembly. It is important to perform the follow-up experiment whether fascin1 is also required for DC-T-cell interactions in antigen-dependent, syngenic interaction between DCs and T-cells.
Fascin1 has been shown to confer resistance to Listeria monocytogenes (Lm) in DCs (Matsumura et al. 2013). Lm is a foodborne pathogen that can cause serious infections in immunocompromised individuals and pregnant women. Upon entry into DCs and macrophages via phagocytosis, the bacterium lyses the primary vacuole with listeriolysin O (LLO), a pore-forming cytolysin, and escapes into the cytoplasm for proliferation. Interestingly, DCs have been shown to be more resistant to Lm than macrophages. The resistance of DCs may be critical for eradication of Lm because DCs need to survive for priming naive cytotoxic T-lymphocytes that are specific to Lm antigens.
It was found that DCs with high expression of fascin1 cleared Lm while fascin1 KO DCs were heavily and uniformly infected, suggesting that fascin1 increases resistance to Lm infection in DCs (Matsumura et al. 2013). Fascin1 appears to increase the resistance to Lm infection in two ways. One is that fascin1 increases phagolysosomal fusion to kill phagosome-encapsulated Lm. Wild-type DCs showed lower phagosomal pH than fascin1 KO DCs. Because phagosomal acidification facilitates phagolysosomal fusion, the lower pH of phagosomes in wild-type DCs indicates more efficient killing of Lm by phagosome-lysosome fusion.
The second way is that fascin1 appears to promote xenophagy to kill cytoplasmic Lm. Cytoplasmic Lm that are escaped from phagosomes are captured by autophagosomes and subsequently fused with lysosomes, resulting in eradication of Lm (called xenophagy). The autophagosome-lysosome fusion events can be monitored by expressing LC3 tandemly labeled with EGFP and mCherry. As LC3 is a component of autophagosomes, tandemly labeled LC3 lose EGFP signals when fused with lysosomes (due to lower pH). It was found that LC3 in wild-type DCs lost EGFP signals more rapidly than did LC3 in fascin1 KO DCs. These results suggest that cytoplasmic Lm encapsulated with autophagosomes are more rapidly eradicated in wild-type DCs.
Fascin is able to assemble bundles of actin filaments with uniform polarity. The assembly of unidirectional bundles is consistent with the function of fascin in filopodia assembly, as well as with its localization to filopodia. This notion is well supported by the manipulation of fascin1 expression: Knock down of fascin1 inhibits both filopodia assembly and cell motility while overexpression of fascin1 promotes these two activities. It should be noted, however, that fascin1 is not essential for filopodia assembly because neuron and DCs from fascin1 KO mice are able to form filopodia.
Several outstanding questions remain to be answered. Fascin1 promotes phagosomal acidification in DCs. What is the molecular mechanism? Also how does fascin1 promote xenophagy? Another outstanding question is the role of fascin1 in IL-6 signaling. What is the physiological significance of fascin1-mediated increase in IL-6 secretion? How are fascin1 and IL-6 required for tumor self-seeding? Future studies should direct toward answering these critical questions.
- Sedeh RS, Fedorov AA, Fedorov EV, Ono S, Matsumura F, Almo SC, et al. Structure, evolutionary conservation, and conformational dynamics of Homo sapiens fascin-1, an F-actin crosslinking protein. J Mol Biol. 2010;400:589–604. doi: 10.1016/j.jmb.2010.04.043. S0022-2836(10)00431-6 [pii].PubMedCrossRefGoogle Scholar