Intermediate Filaments as Signaling Platforms

See allHide authors and affiliations

Science's STKE  19 Dec 2006:
Vol. 2006, Issue 366, pp. pe53
DOI: 10.1126/stke.3662006pe53


Intermediate filaments (IFs) are cytoskeletal structures that are crucial for maintaining the structural and mechanical integrity of cells and tissues. Intriguingly, a wide range of previously unknown nonmechanical roles for the IF cytoskeleton are emerging: Recent studies have linked IFs to the integration of signals related to the determination of cell size, the regulation of cell migration and cell survival, and the buffering of the effects of stress-activated kinases. The characteristic structural features and expression patterns of the different members of this diverse family of highly abundant proteins make them well suited to act as cell- and tissue-specific modifiers and organizers of signaling.

Intermediate filaments (IFs) constitute an extensive cytoskeletal network that extends from the nuclear envelope throughout the cytoplasm of eukaryotic cells. IFs are composed of more than 70 different IF proteins, which are expressed in a tissue- and differentiation state–specific manner [reviewed in (1, 2)]. IF proteins are divided into five different types based on their characteristic molecular features (Table 1): Type I to IV IF proteins form cytoplasmic IFs, whereas type V IF proteins, the lamins, constitute a filamentous network inside the nuclear membrane [for detailed review, see (1)]. Although the IF proteins share a well-conserved α-helical rod domain, which is crucial for polymer assembly, their N- and C-termini show remarkable sequence variability. These variable domains, which include a broad range of potential interaction motifs, are likely to underlie the wide-ranging biochemical and physiological properties of members of this protein family (1).

Table 1.

Classification of intermediate filaments.

IFs were originally regarded as relatively static cytoskeletal structures, the main function of which was mechanical support (1). However, this view of IFs is rapidly changing. IF networks are highly dynamic, showing continuous exchange between polymer and soluble subunits [reviewed in (2)]; they interact with other cytoskeletal systems [reviewed in (3)]; and they are actively regulated by phosphorylation and other posttranslational modifications (2). More surprisingly, based on results from IF-related diseases and from different animal models [knockout (KO) mice and transgenic mice expressing functional mutations], IF proteins are involved in a whole range of metabolic, signaling, and regulatory processes that are by no means directly related to their mechanical functions. Rather, these functions appear to depend on the ability of IFs to sequester or act as scaffolds for various molecules involved in cell signaling.

IFs as Buffers of Cellular Stress Signals

Many of the emerging signaling interactions relate to cell stress tolerance and cell survival. In this respect, IFs, including keratins 8 and 18 (K8 and K18) of simple epithelia, are phosphorylated during different forms of cell stress [for review, see (2)]. A clue to the physiological role of stress-mediated IF phosphorylation involved data showing that the K18 Ser52→Ala52 (S52A) mutation, which eliminates phosphorylation on this residue, predisposes transgenic mice to hepatotoxic injury, indicating that keratin phosphorylation could have a protective effect (4). This paradigm of a protective role of keratin phosphorylation was recently extended to a human liver disease–associated mutation, K8 Gly61→Cys61 (G61C) (5). Transgenic mice overexpressing K8 with this mutation were more prone to liver injury and apoptosis than were wild-type mice (5). Moreover, the G61C mutation inhibited K8 phosphorylation at Ser73, which is targeted by the stress-activated protein kinases, such as p38 kinase, JNK (c-Jun N-terminal kinase), and p42 (5). Because the susceptibility to injury of mice overexpressing a Ser73→Ala73 (S73A) mutant was the same as that of mice overexpressing the G61C mutant, the authors postulated that keratins protect tissues from injuries by serving as a "phosphate sponge" that absorbs stress-induced kinase activity (Fig. 1). The authors proposed that the decrease in keratin phosphorylation in the G61C mutant was due to lack of accessibility of the kinase to Ser73 (5). The amino acid sequence surrounding Ser73 in K8 resembles phosphorylation motifs found in other type-II keratins (K4, K5, and K6) that also become phosphorylated during stress (6), indicating that this type of phosphate buffering could be a general phenomenon in type-II keratins (5).

Fig. 1.

Schematic presentation of intermediate filament interactions with different signaling entities. By acting as stress kinase buffers and through their positioning and scaffolding of signaling molecules, the intermediate filaments act as signal integrators and determinants of key cellular processes. The tissue-specific expression and distribution of these high-abundance proteins make them suitable for cell- and tissue-specific signal functions. The cell in the background displays the vimentin intermediate filament network in a fibroblast transfected with vimentin–green fluorescent protein. [Confocal image courtesy of Mikko Nieminen]

Defense Through Signaling Interactions Involving Survival Pathways

The cytoprotective roles of IFs also stem from their capacity to interact with signaling pathways involved in determining cell survival (Fig. 1). For example, the IF protein nestin forms a scaffold with cyclin-dependent kinase 5 (Cdk5) (7), thereby regulating the apoptosis-inducing activity of Cdk5 during oxidant-induced cell death (8). It is tempting to speculate that this function could be related to a cytoprotective role of nestin in asymmetric cell division of neuronal stem cells, during which one daughter cell survives and the other undergoes apoptosis. The surviving daughter cell has high nestin content, in contrast to the apoptotic cell, which is low in nestin content but has large amounts of the proapoptotic protein Par-4 (9). In an analogous way, K8 sequesters the proapoptotic JNK after death-receptor stimulation (10). The interaction with keratin filaments may keep JNK from phosphorylating proapoptotic nuclear transcription factor targets, because c-Jun phosphorylation was clearly reduced under these conditions (10).

IFs may also transport stress-induced signaling molecules. The recently described ability of nonfilamentous IF precursors or short filaments to travel along microtubules with the help of kinesin and dynein [for review, see (11)] is employed in a surprising way upon nerve injury, when soluble vimentin subunits enable the transport of phosphorylated extracellular signal–regulated kinase 1 and 2 (ERK1/2) from the site of axonal lesion to the nerve cell body (12). Phosphorylated ERK1/2 (pERK) is linked to importin β and thereby to dynein-mediated retrograde transport by means of vimentin, the abundance of which specifically increases in injured neurons (12). The use of vimentin as a long-distance messenger provides an explanation for the rapid and specific increase in vimentin abundance observed under severe forms of stress, such as nerve injury (12).

IFs on the Watchtower over Death Receptors

Many recently identified IF-interacting proteins are apoptotic signaling determinants. For example, the IFs formed by K8 and 18 interact with tumor necrosis factor receptor 2 (TNFR2) and thereby modulate JNK signaling and nuclear factor κB (NF-κB) activation (13). Moreover, K8 and 18 filaments regulate Fas receptor targeting to the cell surface, maintaining receptor density at an appropriate level and determining whether Fas receptor signaling is directed toward the caspase machinery or toward survival signaling provided by the mitogen-activated kinases ERK1/2 (14, 15). This death receptor switch to ERK signaling is crucial for determining the outcome of death receptor stimulation [for review, see (16)]. In addition, death receptor–linked proteins associate with IFs. K8 and 18 filaments bind to the death receptor–associated caspase-8 inhibitor protein c-FLIP (cellular FLICE-like inhibitory protein), an interaction that is crucial for Fas-induced antiapoptotic ERK signaling in hepatocytes (15). K18 sequesters TRADD (TNFR1-associated DEATH domain protein), thereby controlling its interactions with TNFR1 in response to apoptotic stimuli (17). Hair follicle K17 modulates TNF-α signaling, probably through similar TRADD scaffolding (18). Consequently, K17-null skin keratinocytes in primary culture are more sensitive to TNF-α stimulation than are wild-type keratinocytes (18), indicating that K17 and TNF-α act as interdependent regulators of hair follicle keratinocyte cycling.

Signal Integration Through IF Association with 14-3-3 Proteins

14-3-3 proteins are a family of highly conserved and abundant cellular proteins that play key roles in the regulation of central physiological pathways through their interactions with a number of different signaling molecules (19). 14-3-3 proteins play an important role in regulating the G2-to-M phase checkpoint through their interaction with the tyrosine phosphatase Cdc25. An IF-mediated link to this function of 14-3-3 proteins, which bind in a cell cycle–dependent manner to a specific binding domain on phosphorylated K18 (20), was uncovered in studies showing that mitotic progression was disturbed in K8 KO hepatocytes, which completely lack the hepatic K8/18 network (21), as well as in hepatocytes from a transgenic mouse overexpressing keratin 18 with a mutated 14-3-3 binding site Ser33→Ala33 (S33A) (22). Because 14-3-3ζ displayed nuclear sequestration both in K8 KO liver cells (21) and in K18 S33A transgenic hepatocytes (22), as compared to its normally rather diffuse cytoplasmic organization, it was postulated that IFs were necessary to control the subcellular localization of 14-3-3 and prevent uncontrolled and potentially harmful interactions between 14-3-3 and Cdc25 in the nucleus (21). Outside of the nucleus, phosphorylated vimentin interacts in vivo with 14-3-3, thereby displacing the association of Raf with 14-3-3 (23). Intriguingly, IF-14-3-3 interactions were recently also implicated in determining cell size. 14-3-3δ, an epithelial-specific isoform, associates with K17 in a phosphorylation-dependent manner through a specific interaction domain (24). Mouse skin keratinocytes lacking K17 are smaller in size than wild-type keratinocytes and showed depressed protein translation and delayed wound closure (24). These defects turned out to be due to disturbed sequestration of 14-3-3δ in the cytoplasm, where it is normally needed to stimulate the mTOR (mammalian target of rapamycin) signaling pathway (Fig. 1), which plays an important role in the control of protein synthesis and thereby cell growth (24).

IFs as Mediators of Motility

Focal contacts or focal adhesions are cell-matrix adhesion sites that assist the movements of cells on a substratum. Although focal contacts are considered to anchor actin filaments, in several endothelial cell types, vimentin has also been reported to accumulate on adhesion sites (25) at focal contact–resembling structures, called vimentin-associated matrix adhesions (VMAs). Because VMAs assemble in actively migrating cells, it has been suggested that they support motility (25). The binding of vimentin at adhesion sites can be either indirect, mediated through linker proteins such as plectin (26), or direct, with vimentin binding to integrins (27). The architecture of focal contacts is disturbed in vimentin KO fibroblasts (28), suggesting that vimentin might participate in regulating focal contacts. In support of a role for vimentin at the focal contacts, a more recent study demonstrated that the vimentin cytoskeleton regulates focal contact size and helps to stabilize cell-matrix adhesions in endothelial cells (29), indicating that vimentin could regulate migration by affecting the structural integrity of adhesion sites. Moreover, the transcellular migration of lymphocytes through endothelial cell barriers is severely impaired in vimentin KO cells, an effect that is related to the disturbed distribution of adhesion molecules on the vimentin-deficient cells [integrin β1 in lymphocytes and both intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) in endothelial cells (30)]. In agreement with the paradigm of the latter study, depicting vimentin as an important organizer of integrin, vimentin has indeed been shown to participate in the protein kinase C–ε–mediated trafficking of integrins to the plasma membrane, thereby modulating cell motility that is dependent on the cycling of adhesion site proteins (31).

More support for a role for IFs in migration comes from injury studies. Studies with KO animals have shown that the formation of glial scars is impaired in GFAP (glial fibrillary acidic protein)–vimentin double-KO mice (32) as a consequence of defects in cell motility (33). Wound healing is seriously impaired in vimentin KO animals because of delayed migration of fibroblasts (34). K6, K16, and K17 become expressed upon epidermal injuries at the wound edge (35). The healing in the above-mentioned situations is dependent on rapid mobilization and increased migration, indicating an involvement of the reexpressed IF proteins in cell movements.


The broad-ranging, previously unknown functions of IF proteins all relate to protein-protein interactions and the capacity of IFs to actively sequester, position, or act as scaffolds for signaling molecules, including stress-activated kinases. The latter capacity also makes them highly suitable for their recently identified task of acting as a phosphate sponge in stressed cells. The relatively large quantity of IFs in most cells makes them highly efficient both as stress buffers and signaling scaffolds, even for high-abundance signaling molecules, such as 14-3-3.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
View Abstract

Stay Connected to Science Signaling

Navigate This Article