Contribution of intermediate filaments to cell stiffness, stiffening, and growth

It has been shown previously thatintermediate filament (IF) gels in vitro exhibit stiffening athigh-applied stress, and it was suggested that this stiffening propertyof IFs might be important for maintaining cell integrity at largedeformations (Janmey PA, Evtenever V, Traub P, and Schliwa M, JCell Biol 113: 155-160, 1991). In this study, the contribution ofIFs to cell mechanical behavior was investigated by measuring cellstiffness in response to applied stress in adherent wild-type andvimentin-deficient fibroblasts using magnetic twisting cytometry. Itwas found that vimentin-deficient cells were less stiff andexhibited less stiffening than wild-type cells, except at the lowestapplied stress (10 dyn/cm²) where the difference in thestiffness was not significant. Similar results were obtained frommeasurements on wild-type fibroblasts and endothelial cells aftervimentin IFs were disrupted by acrylamide. If, however, cells wereplated over an extended period of time (16 h), they exhibited asignificantly greater stiffness before than after acrylamide, even atthe lowest applied stress. A possible reason could be that theinitially slack IFs became fully extended due to a high degree of cellspreading and thus contributed to the transmission of mechanical stressacross the cell. Taken together, these findings were consistent withthe notion that IFs play important roles in the mechanical propertiesof the cell during large deformation. The experimental data also showedthat depleting or disrupting IFs reduced, but did not entirely abolish,cell stiffening. This residual stiffening might be attributed to theeffect of geometrical realignment of cytoskeletal filaments in thedirection of applied load. It was also found that vimentin-deficientcells exhibited a slower rate of proliferation and DNA synthesis thanwild-type cells. This could be a direct consequence of the absence ofthe intracellular IFs that may be necessary for efficient mediation ofmechanical signals within the cell. Taken together, results of thisstudy suggest that IFs play important roles in the mechanical properties of cells and in cell growth.

     

Human epithelial cell intermediate filaments: isolation, purification, and characterization

Intermediate filaments (IF) isolated from human epithelial cells (HeLa) can be disassembled in 8 M urea and reassembled in phosphate-buffered solutions containing greater than 0.1 mg/ml IF protein. Eight proteins were associated with HeLa IF after several disassembly-reassembly cycles as determined by sodium dodecyl sulfate gel electrophoresis (SDS PAGE). A rabbit antiserum directed against HeLa IF contained antibodies to most of these proteins. The immunofluorescence pattern that was seen in HeLa cells with this antiserum is complex. It consisted of a juxtanuclear accumulation of IF protein and a weblike array of cytoplasmic fibers extending to the cell border. Following preadsorption with individual HeLa IF proteins, the immunofluorescence pattern in HeLa cells was altered to suggest the presence of at least two distinct IF networks. The amino acid composition and alpha-helix content (approximately 38%) of HeLa IF proteins was similar to the values obtained for other IF proteins. One-dimensional peptide maps show extensive homology between the major HeLa IF protein of 55,000-mol- wt and a similar 55,000-mol-wt protein obtained from hamster fibroblasts (BHK-21). HeLa 55,000-mol-wt homopolymer IF assembled under conditions similar to those required for BHK-21 55,000-mol-wt homopolymers. Several other proteins present in HeLa IF preparations may be keratin-like structural proteins. The results obtained in these studies indicate that the major HeLa IF protein is the same major IF structural protein found in fibroblasts. Ultrastructural studies of HeLa cells revealed two distinct IF organizational stages including bundles and loose arrays. In addition, in vitro reconstituted HeLa IF also exhibited these two organizational states.     

Vimentin filaments and centrosomes: Are they associated?

HeLa cells were examined by immunofluorescence using anti-vimentin and anti-centrosphere anti-bodies, and by transmission electron microscopy (TEM), after vimentin redistribution induced by the action of nocodazole or taxol. A redistribution of vimentin bundles in the centriolar area was observed after nocodazole treatment, although no direct interaction between centrioles and vimentin filaments could be detected. After taxol treatment, the juxtanuclear accumulation of vimentin filaments and the centrioles were rarely observed in the same area. Our results do not support the concept of a direct association between centrioles and vimentin filaments.     

The function of intermediate filaments in cell shape and cytoskeletal integrity

This study describes the development and use of a specific method for disassembling intermediate filament (IF) networks in living cells. It takes advantage of the disruptive effects of mimetic peptides derived from the amino acid sequence of the helix initiation 1A domain of IF protein chains. The results demonstrate that at 1:1 molar ratios, these peptides disassemble vimentin IF into small oligomeric complexes and monomers within 30 min at room temperature in vitro. Upon microinjection into cultured fibroblasts, these same peptides induce the rapid disassembly of IF networks. The disassembly process is accompanied by a dramatic alteration in cell shape and the destabilization of microtubule and actin-stress fiber networks. These changes in cell shape and IF assembly states are reversible. The results are discussed with respect to the roles of IF in cell shape and the maintenance of the integrity and mechanical properties of the cytoplasm, as well as the stability of the other major cytoskeletal systems.     

Monoclonal antibodies to intermediate filaments in chick muscle cell cultures

Two monoclonal antibodies, FIFI and PHIL, have been prepared using detergent-washed myogenic cells as immunogen. On Western blots of total protein extracts of muscle cells, both antibodies bind to vimentin (52 kD) and its degradation products (major band at 42 kD), but do not bind to mouse proteins or to actin (42 kD). Specificity for a determinant common to vimentin and desmin was confirmed by 2-D gel electrophoresis of muscle cell extracts and purified desmin. Western blots with FIFI reveal particularly well the extreme sensitivity of intermediate filaments (IFs) to proteolysis, which was preventable in brain tissue only by boiling in 1% SDS, although it could be reduced in both brain and muscle by less extreme methods. Western blots suggest a large increase in IF content of differentiating myoblast cell cultures at the time of cell fusion and an increase of at least 4-fold is confirmed by a quantitative immunoassay using a direct ELISA method. Immunofluorescence microscopy shows that this increase is due to the appearance of high concentrations of the intermediate filament antigen at the ends of early myotubes, preceding the appearance of cross-striations in myofibrils. Furthermore, whereas the polar filaments detected by FIFI run right to the ends of the early myotubes and only sparingly penetrate the central area, cross-striated myofibrils (as detected by the monoclonal antibody, SAM) run the length of the myotube but do not reach the ends. Colcemid and colchicine cause the vimentin filaments in fibroblasts to collapse into perinuclear rings or caps, but do not have this effect on the polar fluorescence in early myotubes. Heat shock (2 h at 45 degrees C) has a similar differential effect. The results suggest that early in muscle differentiation intermediate filament proteins accumulate rapidly at myotube ends, where they are organized differently from those in fibroblasts.     

Microtubules and filaments in developing axons and optic stalk cells

Fibrous structures have been studied in the developing optic nerve of chick embryos. The first ganglion cell axons (3-day embryos) were of moderate size, with both neurofilaments and microtubules. Subsequently (4- and 5-day embryos), very small axons were also present. In thesc embryos and in the 4-day hatched chick, the density of microtubules fell within the same range for all but the very small axons, which tended to have more microtubules per unit area. Filaments similar to those previously thought to represent neurofilaments in other parts of the embryonic nervous system were present in the early optic stalk cells, calling into question the reliability of identifying early nerve cells on the basis of 'neurofilaments'.     

Coalignment of microtubules, cytokeratin intermediate filaments, and collagen fibrils in a collagen-secreting cell system

The distribution of microtubules and intermediate filaments in the collagen-secreting scleroblasts of the goldfish scale was investigated by immunofluorescence and electron microscopy. Many of the microtubules and cytokeratin type intermediate filaments formed bundles that were aligned with the underlying, parallel collagen fibrils. The intermediate filament bundles were evenly spaced and located adjacent to the basal plasma membrane. The microtubules, on the other hand, were located further away from the membrane, although many were found very close to the intermediate filament bundles. No detectable change was observed in scleroblast microtubules when cells on scales were treated with colchicine or cooled (greater than or equal to 0 degrees C) for up to 1 h. Cells had to be cooled overnight before the microtubules were affected. The final number and length of the microtubules in the cell depended only on the final steady-state temperature and not the temperature history of the scale cell, and steady state was reached more slowly at colder temperatures. The microtubules but not the intermediate filaments rapidly (within 5 min) and reversibly depolymerized when cells were chilled to -2 approximately -4 degrees C. When chilled cells were warmed, the microtubules polymerized back, within 15 min at room temperature, to the same pattern of parallel coalignment with the underlying collagen. They appeared to repolymerize via two different pathways: (1) a radial growth outwards from the microtubule organizing center followed by a progressive realignment with the underlying collagen and (2) a gradual and simultaneous polymerization along cold-stable, antitubulin staining fibers. These fibers were also aligned with the collagen fibrils and may be related to the aligned intermediate filaments.     

Role of p120 Ras-GAP in directed cell movement

We have used cell lines deficient in p120 Ras GTPase activating protein (Ras-GAP) to investigate the roles of Ras-GAP and the associated p190 Rho-GAP (p190) in cell polarity and cell migration. Cell wounding assays showed that Ras-GAP-deficient cells were incapable of establishing complete cell polarity and migration into the wound. Stimulation of mutant cells with growth factor rescued defects in cell spreading, Golgi apparatus fragmentation, and polarized vesicular transport and partially rescued migration in a Ras-dependent manner. However, for directional movement, the turnover of stress fibers and focal adhesions to produce an elongate morphology was dependent on the constitutive association between Ras-GAP and p190, independent of Ras regulation. Disruption of the phosphotyrosine-mediated Ras-GAP/p190 complex by microinjecting synthetic peptides derived from p190 sequences in wild-type cells caused a suppression of actin filament reorientation and migration. From these observations we suggest that although Ras-GAP is not directly required for motility per se, it is important for cell polarization by regulating actin stress fiber and focal adhesion reorientation when complexed with 190. This observation suggests a specific function for Ras-GAP separate from Ras regulation in cell motility.     

Microtubules and actin filaments are not critically involved in the biogenesis of epithelial cell surface polarity

We have studied the role of microtubules and actin filaments in the biogenesis of epithelial cell surface polarity, using influenza hemagglutinin and vesicular stomatitis G protein as model apical and basolateral proteins in infected Madin-Darby canine kidney cells. Addition of colchicine or nocodazole to confluent monolayers at concentrations sufficient to completely disassemble microtubules did not affect the asymmetric budding of influenza or vesicular stomatitis virus and only slightly reduced the typical asymmetric surface distribution of their envelope proteins, despite extensive cytoplasmic redistribution of the Golgi apparatus. Alteration of microtubular function by taxol or dissociation of actin filaments by cytochalasin D also failed to have a significant effect. Furthermore, neither colchicine nor cytochalasin D pretreatment blocked the ability of subconfluent Madin-Darby canine kidney cells to sustain polarized budding of influenza virus a few hours after attachment to the substrate. Our results indicate that domain-specific microtubule or actin filament "tracks" are not responsible for the vectorial delivery of apically or basolaterally directed transport vesicles. In conjunction with currently available evidence, they are compatible with a model in which receptors in the cytoplasmic aspect of apical or basolateral regions provide vectoriality to the transport of vesicles carrying plasma membrane proteins to their final surface localization.     

Apoptosis and keratin intermediate filaments

Intermediate filament (IF) proteins utilize central alpha-helical domains to generate polymeric fibers intermediate in size between actin microfilaments and microtubules. The regions flanking the central structural domains have diverged greatly to permit IF proteins to adopt specialized functions. Keratins represent the largest two groups of IF proteins. Most keratins serve structural functions in hair or epidermis. Intracellular epidermal keratins also provide strength to epithelial sheets. The intracellular type I keratins and other IF proteins are cleaved by caspases during apoptosis to ensure the disposal of the relatively insoluble cellular components. However, recent studies have also revealed an unexpected protective role for keratin 8 during TNF and Fas mediated apoptosis. Evidence for possible functions of keratins both upstream and downstream of apoptotic signaling are considered.     

Overexpression of cofilin stimulates bundling of actin filaments, membrane ruffling, and cell movement in Dictyostelium

Cofilin is a low molecular weight actin-modulating protein whose structure and function are conserved among eucaryotes. Cofilin exhibits in vitro both a monomeric actin-sequestering activity and a filamentous actin-severing activity. To investigate in vivo functions of cofilin, cofilin was overexpressed in Dictyostelium discoideum cells. An increase in the content of D. discoideum cofilin (d-cofilin) by sevenfold induced a co-overproduction of actin by threefold. In cells over-expressing d-cofilin, the amount of filamentous actin but not that of monomeric actin was increased. Overexpressed d-cofilin co-sedimented with actin filaments, suggesting that the sequestering activity of d- cofilin is weak in vivo. The overexpression of d-cofilin increased actin bundles just beneath ruffling membranes where d-cofilin was co- localized. The overexpression of d-cofilin also stimulated cell movement as well as membrane ruffling. We have demonstrated in vitro that d-cofilin transformed latticework of actin filaments cross-linked by alpha-actinin into bundles probably by severing the filaments. D. discoideum cofilin may sever actin filaments in vivo and induce bundling of the filaments in the presence of cross-linking proteins so as to generate contractile systems involved in membrane ruffling and cell movement.     

Connections of intermediate filaments with the nuclear lamina and the cell periphery

We investigated the relationship between intermediate filaments (IFs) and other detergent- and nuclease-resistant filamentous structures of cultured liver epithelial cells (T51B cell line) using whole mount unembedded preparations which were sequentially extracted with Triton X-100 and nucleases. Immunogold labelling and stereoscopic observation facilitated the examination of each filamentous structure and their three-dimensional relationships to each other. After solubilizing phospholipid, nucleic acid and soluble cellular protein, the resulting cytoskeleton preparation consisted of a network of cytokeratin and vimentin IFs linked by 3 nm filaments. The IFs were anchored to and determined the position of the nuclear lamina filaments (NLF) network and the centrioles. The NLF was composed of the nuclear lamina filaments measuring 3-6 nm in diameter which radiated from and anchored to the skeleton of the nuclear pores. The IFs located in the nuclear region appeared to be interwoven with the NLF. At the cell surface, the IFs seemed to be attached to the putative actin filament network. They formed a focally interrupted plexus-like structure at the cell periphery. Fragments of vimentin filaments were found among the filamentous network located at the cell surface, and some filaments terminated blindly there.     

Microtubules,protoplasts and plant cell shape

Indirect immunofluorescence has been used to study the function of cytoplasmic microtubules in controlling the shape of elongated carrot cells in culture. Using a purified wall-degrading preparation, the elongated cells are converted to spherical protoplasts and the transverse hoops of bundled microtubules are disorganised but not depolymerised in the process. Since microtubules remain attached to fragments of protoplast membrane adhering to coverslips and are still seen to be organised laterally in bundles, it would appear that re-orientation of the transverse bundles is due to loss of cell wall and not to the cleavage of microtubule bridges. After 24 h treatment in 10^-3 M colchicine, microtubules are depolymerised in elongated cells but, at this time, the cells retain their elongated shape. This suggests that wall which was organised in the presence of transverse microtubule bundles can retain asymmetric shape for short periods in the absence of those tubules. However, after longer periods of time the cells become spherical in colchicine. Neither wall nor tubules therefore exert individual control on continued cellular elongation and so we emphasize the fundamental nature of wall/microtubule interactions in shape control. It is concluded that the observations are best explained by a model in which hooped bundles of microtubules—which are directly or indirectly associated with molecules involved with cellulose biosynthesis at the cell surface—act as an essential template or scaffolding for the orientated deposition of cellulose.     

A periodic ultrastructure in intermediate filaments

Intermediate sized filaments reconstituted in vitro from purified desmin, epidermal keratin and the M_r 68,000 protein of neurofilaments were examined after high resolution metal shadowing. The filaments demonstrate a marked longitudinal periodicity of about 21 nm. This is the first procedure that allows detection of a periodic substructure in these filaments using the electron microscope.