Probing of the structural stability of vimentin and desmin-type intermediate filaments with Ca2+-activated proteinase, thrombin and lysine-specific endoproteinase Lys-C

Intermediate filaments (IFs) reconstituted from purified, delipidated vimentin and desmin as well as respective protofilaments were subjected to degradation by Ca2+-activated neutral thiol proteinase, thrombin and lysine-specific endoproteinase Lys-C, respectively. The breakdown products were analyzed by SDS-polyacrylamide gel electrophoresis and negative stain electron microscopy. While Ca2+-activated proteinase and thrombin caused rapid and complete degradation of IFs with kinetics not significantly different from those of the degradation of protofilaments, lysine-specific endoproteinase did not exert any electron microscopically detectable effect on filament structure. Although both types of subunit proteins were truncated at their non-alpha-helical, C-terminal polypeptides by this proteinase, they were still able to assemble into 10 nm filaments. Closer electron microscopic inspection of IFs treated with Ca2+-activated proteinase revealed numerous ruptures along the filaments already at very early stages of digestion. SDS-polyacrylamide gel electrophoresis of the processed filaments in conjunction with previous biochemical characterizations of the breakdown of protofilaments by Ca2+-activated proteinase showed that these inhomogeneities primarily arose from degradation of the arginine-rich, non-alpha-helical N-termini of the filament proteins. These findings demonstrate that, although the N-terminus of vimentin and desmin is essential for filament stability, it is still highly susceptible to proteolytic attack in particular and very likely to posttranslational modification in general. Such structural modifications of the N-termini of IF proteins might exert great influences on the intracellular distribution and molecular organization of IFs in various physiological and pathological conditions.     

Efficient degradationin vitro of all intermediate filament subunit proteins by the Ca2+-activated neutral thiol proteinase from Ehrlich ascites tumor cells and porcine kidney

Vimentin, desmin, glial fibrillary acidic protein, neurofilament triplet proteins, and a mixture of cytokeratins were digested with Ca²⁺-activated neutral thiol proteinase isolated from Ehrlich ascites tumor (EAT) cells and porcine kidney. All intermediate filament proteins were degraded by the proteinase, although with different rates and Ca²⁺ optima. These results are in part at variance with our previous statement that the Ca²⁺-activated proteinase from EAT cells is specific for vimentin and desmin.     

Ca2(+)-calmodulin-dependent protein kinase II phosphorylates various types of non-epithelial intermediate filament proteins

We have investigated the actions of Ca2(+)-calmodulin (CaM)-dependent protein kinase II on various types of non-epithelial intermediate filament proteins, vimentin, desmin, glial fibrillary acidic protein (GFAP) and neurofilament triplet proteins. Most of these filament proteins could serve as substrates. The effects of phosphorylation on the filamentous structure of vimentin were investigated in sedimentation experiments and by using electron microscopy. The amount of unassembled vimentin increased linearly with increased phosphorylation. However, the extent of the effect of phosphorylation on the potential to polymerize was also affected by the MgCl2 concentration, under conditions for reassembly. The actions of Ca2(+)-CaM-dependent protein kinase II on non-epithelial intermediate filaments under physiological conditions are given attention.     

Cyclic AMP-modulated phosphorylation of intermediate filament proteins in cultured avian myogenic cells.

The intermediate filament proteins desmin and vimentin and the muscle tropomyosins were the major protein phosphate acceptors in 8-day-old myotubes incubated for 4 h in medium containing radiolabeled phosphate. The addition of isoproterenol or 8-bromo-cyclic AMP (BrcAMP) resulted in a two- to threefold increase in incorporation of 32PO4 into both desmin and vimentin, whereas no changes in the incorporation of 32PO4 into tropomyosin or other cellular proteins were observed. The BrcAMP- or hormonally induced increase in 32PO4 incorporation into desmin and vimentin was independent of protein synthesis and was not caused by stimulation of protein phosphate turnover. In addition, BrcAMP did not induce significant changes in the specific activity of the cellular ATP pool. These data suggest that the observed increase in 32PO4 incorporation represented an actual increase in phosphorylation of the intermediate filament proteins desmin and vimentin. Two-dimensional tryptic analysis of desmin from 8-day-old myotubes revealed five phosphopeptides of which two showed a 7- to 10-fold increase in 32PO4 incorporation in BrcAMP-treated myotubes. Four of the phosphopeptides identified in desmin labeled in vivo were also observed in desmin phosphorylated in vitro by bovine heart cAMP-dependent protein kinase. Although phosphorylation of desmin and vimentin was apparent in myogenic cells at all stages of differentiation, BrcAMP- and isoproterenol-induced increases in phosphorylation of these proteins were restricted to mature myotubes. These data strongly suggest that in vivo phosphorylation of the intermediate filament proteins desmin and vimentin is catalyzed by the cAMP-dependent protein kinases and that such phosphorylation may be regulated during muscle differentiation.     

The intermediate-filament proteins vimentin and desmin are phosphorylated in specific domains

Analysis of specific fragments of vimentin and desmin from 32P-labeled BHK-21 cells indicated that these intermediate-filament subunit proteins are phosphorylated in specific regions or domains. High performance liquid chromatography and sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis of lysine-specific protease-generated fragments demonstrated that both molecules were phosphorylated in their amino terminal or "head" domains. While this was the predominant site of phosphorylation for vimentin, additional phosphorylated fragments from desmin were observed. Chemical cleavage of [32P]desmin and subsequent examination of the phosphorylated peptides indicated that the major site of desmin phosphorylation was located within the "tail" domain. Analysis of vimentin and desmin from non-mitotic and mitotically selected cells indicated that the increased phosphorylation of intermediate-filament proteins observed during cell division occurs within the amino terminal domains of both molecules. These studies indicate that the increased phosphorylation of filament proteins during mitosis may involve the function of the amino terminal domain. In addition, filament proteins may be phosphorylated in a subunit-protein-specific manner which may reflect subunit-specific functions.     

Intermediate filament reconstitution in vitro. The role of phosphorylation on the assembly-disassembly of desmin

Desmin, the myogenic intermediate filament protein, is a phosphoprotein containing phosphoserine, in vivo. The role of phosphorylation on assembly-disassembly and organization of the desmin filament has remained obscure. We report here on a stable and purified system which enables a biochemical examination of desmin filament assembly and disassembly. Using this in vitro system, we carried out stoichiometrical phosphorylations by purified protein kinases. The extent of polymerization-depolymerization was estimated using procedures related to centrifugation and electron microscopy. The evidence we obtained suggests that disassembly of the desmin filament and inhibition of the NaCl-dependent polymerization of the soluble desmin can reversibly occur with either cAMP-dependent or Ca2+-activated, phospholipid-dependent desmin phosphorylation.     

Caspase proteolysis of desmin produces a dominant-negative inhibitor of intermediate filaments and promotes apoptosis

Caspase cleavage of key cytoskeletal proteins, including several intermediate filament proteins, triggers the dramatic disassembly of the cytoskeleton that characterizes apoptosis. Here we describe the muscle-specific intermediate filament protein desmin as a novel caspase substrate. Desmin is cleaved selectively at a conserved Asp residue in its L1-L2 linker domain (VEMD downward arrow M(264)) by caspase-6 in vitro and in myogenic cells undergoing apoptosis. We demonstrate that caspase cleavage of desmin at Asp(263) has important functional consequences, including the production of an amino-terminal cleavage product, N-desmin, which is unable to assemble into intermediate filaments, instead forming large intracellular aggregates. Moreover, N-desmin functions as a dominant-negative inhibitor of filament assembly, both for desmin and the structurally related intermediate filament protein vimentin. We also show that stable expression of a caspase cleavage-resistant desmin D263E mutant partially protects cells from tumor necrosis factor-alpha-induced apoptosis. Taken together, these results indicate that caspase proteolysis of desmin at Asp(263) produces a dominant-negative inhibitor of intermediate filaments and actively participates in the execution of apoptosis. In addition, these findings provide further evidence that the intermediate filament cytoskeleton has been targeted systematically for degradation during apoptosis.     

Balance between activities of Rho kinase and type 1 protein phosphatase modulates turnover of phosphorylation and dynamics of desmin/vimentin filaments

To analyze the cell cycle-dependent desmin phosphorylation by Rho kinase, we developed antibodies specifically recognizing the kinase-dependent phosphorylation of desmin at Thr-16, Thr-75, and Thr-76. With these antibodies, phosphorylation of desmin was observed specifically at the cleavage furrow in late mitotic Saos-2 cells. We then found that treatment of the interphase cells with calyculin A revealed phosphorylation at all the three sites of desmin. We also found that an antibody, which specifically recognizes vimentin phosphorylated at Ser-71 by Rho kinase, became immunoreactive after calyculin A treatment. This calyculin A-induced interphase phosphorylation of vimentin at Ser-71 was blocked by Rho kinase inhibitor or by expression of the dominant-negative Rho kinase. Taken together, our results indicate that Rho kinase is activated not only in mitotic cells but also interphase ones, and phosphorylates intermediate filament proteins, although the apparent phosphorylation level is diminished to an undetectable level due to the constitutive action of type 1 protein phosphatase. The balance between intermediate filament protein phosphorylation by Rho kinase and dephosphorylation by type 1 protein phosphatase may affect the continuous exchange of intermediate filament subunits between a soluble pool and polymerized intermediate filaments.     

A rapid method for the large scale purification of the intermediate filament protein vimentin by single-stranded DNA-cellulose affinity chromatography

A new method is described for the purification of the intermediate filament protein vimentin from Ehrlich ascites tumor cells using single-stranded DNA-cellulose affinity chromatography. The procedure is rapid and allows the large scale isolation of the protein. Partial characterization of vimentin shows that it has a molecular weight of 58000 and an apparent pI of 5.3. It can be degraded by the vimentin-specific, Ca²⁺-activated proteinase which results in the production of a characteristic set of degradation products. The vimentin also cross-reacts with the intermediate filament protein monoclonal antibody, α-IFA.     

Phosphorylation of vimentin in mitotically selected cells. In vitro cyclic AMP-independent kinase and calcium-stimulated phosphatase activities

The phosphorylation of the intermediate filament protein vimentin was examined under in vitro conditions. Cell cytosol and Triton-insoluble cytoskeleton preparations from nonmitotic and mitotically selected mouse L-929 cells exhibited vimentin kinase activity that is apparently cAMP and Ca2+ independent. The level of vimentin kinase activity was greater in preparations from mitotically selected cells than nonmitotic cells. Addition of Ca2+ to mitotic cytosol decreased net vimentin phosphorylation. Dephosphorylation experiments indicated that there is phosphatase activity in these preparations which is stimulated by addition of Ca2+. Fractionation of cytosol from nonmitotic cells on DEAE-Sephacel and phosphocellulose revealed a single major vimentin kinase activity (peak I). Fractionation of cytosol from mitotically selected cells yielded a similar activity (peak I) and an additional vimentin kinase activity (peak II) that was not found in nonmitotic preparations. Based on substrate specificity and lack of inhibition to characteristic inhibitors, the semipurified peak I and II vimentin kinase activities appear to be cAMP-independent enzymes that are distinct from casein kinases I and II. Phosphopeptide mapping studies indicated that both peak I and peak II vimentin kinases phosphorylate tryptic peptides in the NH2-terminal region of vimentin that are phosphorylated in intact cells. Electron microscopic examination of reconstituted vimentin filaments phosphorylated with both semipurified kinases indicated that phosphorylation induced filament disassembly. These experiments indicate that the increased phosphorylation of vimentin during mitosis may be catalyzed by a discrete cAMP-independent protein kinase. In addition, preparations from mitotic cells exhibited a Ca2+-stimulated phosphatase activity, suggesting that Ca2+ may play a regulatory role in vimentin dephosphorylation during mitosis.     

Localization of vimentin and desmin in BHK21/C13 cells and in baby hamster kidney

Specific antibodies against the intermediate filament protein subunits, desmin and vimentin, were used to characterize the fibroblastic tissue culture cell line BHK21/C13 and the cells comprising baby hamster kidney (BHK). The BHK21/C13 cells have previously been shown to contain desmin and vimentin by biochemical techniques. The results from double immunofluorescence analysis show that both immunologically distinct intermediate filament subunit proteins are expressed simultaneously within the same BHK21/C13 cell, and that the filamentous patterns are very similar, if not superimposable even in cells treated with colchicine. There are some cells that may contain vimentin only. Double immunofluorescence on cryostat sections of BHKs and preparations of dissociated kidney cells demonstrate that the cells, most likely smooth muscle, comprising the blood vessel walls contain vimentin and desmin simultaneously. The simultaneous expression of vimentin and desmin is not a phenomenon which is restricted to tissue culture cells. Thus, the simultaneous presence of these two intermediate filament proteins within the BHK21/C13 cell may not be the result of growth in tissue culture.     

Site-specific antibodies block kinase A phosphorylation of desmin in vitro and inhibit incorporation of myoblasts into myotubes.

Desmin and vimentin are two type III intermediate filament (IF) proteins, which can be phosphorylated in vitro by cAMP-dependent kinase (kinase A) and protein kinase C, and the in vitro phosphorylation of these proteins appears to favor the disassembled state. The sites of phosphorylation for desmin and vimentin have been mapped to their amino-terminal headpiece domains; in chicken smooth muscle desmin the most kinase A-reactive residues are ser-29 and ser-35. In this study we have examined the phosphorylation of desmin by the catalytic subunit of kinase A by using anti-peptide antibodies directed against residues 26-36. The antibodies, which we call anti-D26, recognize both native and denatured desmin and can discriminate between intact desmin and those derivatives that do not possess residues 26-36. Pre-incubation of desmin with affinity purified anti-D26 blocks total kinase A catalyzed incorporation of 32P into desmin by 75-80%. When antibody-treated IFs are subjected to phosphorylation, no filament break-down is observed after 3 hours. Thus anti-D26 antibodies block phosphorylation of IF in vitro. We have also explored the role of desmin phosphorylation in skeletal muscle cell differentiation using these antibodies. Quail embryo cells, induced to differentiate along the myogenic pathway by infection with avian SKV retroviruses expressing the ski oncogene, were microinjected with affinity purified anti-D26 at the mononucleated, myoblast stage. By 24 h post-injection, the vast majority of uninjected cells had fused into multinucleated myotubes, but all microinjected cells were arrested in the process of incorporating into myotubes and remained mononucleated. This observation suggests that kinase A phosphorylation-induced dynamic behavior of the desmin/vimentin IF cytoskeleton may be one of the many cytoskeletal restructuring events that must take place during myoblast fusion.     

The effect of limited proteolysis on GTP-dependent Ca2+ efflux and GTP-dependent fusion in rat liver microsomal vesicles.

1. Limited proteolytic digestion of rat liver microsomes (microsomal fractions) with trypsin (5 micrograms/ml), proteinase K (1.0 microgram/ml) and Pronase (20 micrograms/ml final concns.) resulted in abolition of GTP-dependent vesicle fusion. 2. Vesicle fusion could be partially restored to microsomes which had undergone limited tryptic digestion, by the addition of untreated microsomal vesicles. 3. GTP-dependent Ca2+ efflux from rat liver microsomes was also observed to be inhibited by limited proteolysis with trypsin and proteinase K. 4. Limited proteolysis of rat liver microsomes had no effect on subsequent GTP-dependent phosphorylation of polypeptides of Mr 17,000 and 38,000, and thus it is unlikely that the phosphorylation of these proteins is involved in GTP-dependent Ca2+ efflux and GTP-dependent vesicle fusion. 5. GTP binding by Gn proteins [proteins which bind GTP after transfer to nitrocellulose, as defined by Bhullar & Haslam (1986) Biochem. J. 245, 617-620] was inhibited by pre-treatment of microsomes with trypsin, proteinase K and Pronase at concentrations similar to those which abolished GTP-dependent Ca2+ efflux and vesicle fusion. 6. We suggest that one or more of the Gn proteins may be involved in the molecular mechanisms of GTP-dependent vesicle fusion and Ca2+ efflux in rat liver microsomes and that limited proteolytic digestion may be a useful tool in further investigation of these processes.     

Carcinogenesis: A comprehensive survey, 5. Modifiers of chemical carcinogenesis: An approach to the biochemical mechanism and cancer prevention: T. J. Slaga,ed. New York: Raven Press. 275 pp. $30.00

Electrophoretic and autoradiographic analyses of the incorporation of ³⁵S-methionine into newly synthesized proteins during myogenesis reveal that presumptive chicken myoblasts synthesize primarily one intermediate filament protein: vimentin. Desmin synthesis is initiated at the onset of fusion. Synthesis rates of both filament subunits increase during the first three days in culture, relative to the total protein synthesis rate. The observed increase in the rate of desmin synthesis (at least 10 fold) is significantly greater than that observed for vimentin, and is responsible for a net increase in the cellular desmin content relative to vimentin. Both filament subunits continue to be synthesized through at least 20 days in culture. Immunofluorescent staining using desmin- and vimentin-specific antisera supports the conclusion that desmin is synthesized only in fusing or multinucleate cells. These results indicate that the synthesis of the two filament subunits is not coordinately regulated during myogenesis. The distributions of desmin and vimentin in multinucleate chicken myotubes are indistinguishable, as determined by double immunofluorescence techniques. In early myotubes, both proteins are found in an intricate network of free cytoplasmic filaments. Later in myogenesis, several days after the appearance of α-actinin-containing Z line striations, both filament proteins become associated with the Z lines of newly assembled myofibrils, with a corresponding decrease in the number of cytoplasmic filaments. This transition corresponds to the time when the a-actinin-containing Z lines become aligned laterally. These data suggest that the two intermediate filament systems, desmin and vimentin, have an important role in the lateral organization and registration of myofibrils and that the synthesis of desmin and assembly of desmin-containing intermediate filaments during myogenesis is directly related to these functions. These results also indicate that the Z disc is assembled in at least two distinct steps during myogenesis.     

Evolution of homologous domains of cytoplasmic intermediate filament proteins and lamins

The earliest gene duplications in the evolution of the intermediate filament proteins created the ancestors of acidic keratins, basic keratins, nonepithelial intermediate filament proteins, and lamins. Biochemistry and function of cytoplasmic intermediate filaments differ greatly from those of lamins. Cytoplasmic intermediate filament proteins have a different cellular location than lamins, form different types of supramolecular structures, and are missing a protein segment found in lamins; but the data presented here indicate that the cytoplasmic intermediate filaments do not have a common ancestor separate from the ancestor of lamins. In the non-epithelial intermediate filament branch, the ancestor of neurofilament proteins and the common ancestor of desmin, vimentin, and glial fibrillary acidic protein (GFAP) diverged first. By evolutionary criteria, the intermediate filament protein recently discovered in neuronal cells does not belong to the neurofilament family but is more closely related to desmin, vimentin, and GFAP. Sequences of different sub-domains yield different evolutionary trees, possibly indicating existence of sub- domain-specific functions.      

Assembly of amino-terminally deleted desmin in vimentin-free cells

To study the role of the amino-terminal domain of the desmin subunit in intermediate filament (IF) formation, several deletions in the sequence encoding this domain were made. The deleted hamster desmin genes were fused to the RSV promoter. Expression of such constructs in vimentin- free MCF-7 cells as well as in vimentin-containing HeLa cells, resulted in the synthesis of mutant proteins of the expected size. Single- and double-label immunofluorescence assays of transfected cells showed that in the absence of vimentin, desmin subunits missing amino acids 4-13 are still capable of filament formation, although in addition to filaments large numbers of desmin dots are present. Mutant desmin subunits missing larger portions of their amino terminus cannot form filaments on their own. It may be concluded that the amino-terminal region comprising amino acids 7-17 contains residues indispensable for desmin filament formation in vivo. Furthermore it was shown that the endogenous vimentin IF network in HeLa cells masks the effects of mutant desmin on IF assembly. Intact and mutant desmin colocalized completely with endogenous vimentin in HeLa cells. Surprisingly, in these cells endogenous keratin also seemed to colocalize with endogenous vimentin, even if the endogenous vimentin filaments were disturbed after expression of some of the mutant desmin proteins. In MCF-7 cells some overlap between endogenous keratin and intact exogenous desmin filaments was also observed, but mutant desmin proteins did not affect the keratin IF structures. In the absence of vimentin networks (MCF-7 cells), the initiation of desmin filament formation seems to start on the preexisting keratin filaments. However, in the presence of vimentin (HeLa cells) a gradual integration of desmin in the preexisting vimentin filaments apparently takes place.     

In vitro translation of the intermediate filament proteins desmin and vimentin.

Polyadenylated ribonucleic acid (RNA) was isolated from chicken skeletal and smooth muscle and translated in a cell-free rabbit reticulocyte system. Both types of muscle tissue contain messenger RNAs that code for the intermediate filament proteins desmin and vimentin, and the relative concentrations of the two translation products reflect the prevalence of the two proteins in vivo. Desmin synthesis represents a greater proportion of the total protein synthesis from smooth muscle RNA than from skeletal muscle RNA, whereas the converse is true of vimentin synthesis. Fractionation of the RNA on formamide-containing sucrose gradients before translation indicates that the desmin messenger RNA is larger than the vimentin messenger RNA and contains an extensive noncoding segment. The desmin and vimentin messages code predominantly for the non-phosphorylated forms of desmin and vimentin. However, the ratio of phosphorylated to unphosphorylated forms of the proteins could be increased by adding cyclic adenosine monophosphate-dependent kinase activity to the translation mixtures. These results suggest that desmin and vimentin are each synthesized from a single messenger RNA species and that posttranslational phosphorylation generates the additional isoelectric variants of each which are observed in vivo.     

Cytoskeletal proteins at the cholinergic synapse: distribution of desmin, actin, fodrin, neurofilaments, and tubulin in Torpedo electric organ

Treatment of the electric organ of Torpedo marmorata with Triton X-100 in the presence of 2 mM MgCl2 generated a cytoskeletal fraction in which a 54 kDa polypeptide is a major constituent. This 54 kDa polypeptide accounted for about 8% of the cellular protein when total electric organ tissue was analyzed by two-dimensional gel electrophoresis. Immunoblotting experiments showed that this protein reacts with monoclonal antibodies to desmin, the major intermediate filament protein of avian and mammalian muscle tissue. Negative stain analysis revealed that filaments of about 10 nm diameter are the major structural elements of the electric organ cytoskeleton. In the presence of Ca2+ there was a rapid degradation of the desmin-like protein and intermediate filaments due to a Ca2+-activated protease. Some of the resulting fragments retained antigenic activity against the desmin antibodies. Immunoblotting of membrane fractions enriched in acetylcholine receptor revealed desmin in addition to some actin. A further cytoskeletal component was identified from biochemical and immunological properties as a homologue of the mammalian neurofilament L-polypeptide. Thus Torpedo expresses proteins homologous to the mammalian desmin and neurofilament L-protein which can be detected using immunological approaches. Immunofluorescence microscopy was used to map the location of various cytoskeletal proteins of the cholinergic synapse on paraffin sections and on en face preparations of membranes. Desmin staining was restricted to electrocytes and in en face preparations was seen associated with both the ventral receptor-containing membrane and with the non-innervated dorsal membrane. Antibodies to neurofilament L-protein stained only the axons and not the electrocytes. Staining for fodrin, a non-erythrocyte spectrin, resulted in submembraneous decoration of both the axons and the electrocytes. Axonal staining for neurofilaments and microtubules did not extend into the ends of the nerve terminal arborizations.     

Differential organization of desmin and vimentin in muscle is due to differences in their head domains

In most myogenic systems, synthesis of the intermediate filament (IF) protein vimentin precedes the synthesis of the muscle-specific IF protein desmin. In the dorsal myotome of the Xenopus embryo, however, there is no preexisting vimentin filament system and desmin's initial organization is quite different from that seen in vimentin-containing myocytes (Cary and Klymkowsky, 1994. Differentiation. In press.). To determine whether the organization of IFs in the Xenopus myotome reflects features unique to Xenopus or is due to specific properties of desmin, we used the injection of plasmid DNA to drive the synthesis of vimentin or desmin in myotomal cells. At low levels of accumulation, exogenous vimentin and desmin both enter into the endogenous desmin system of the myotomal cell. At higher levels exogenous vimentin forms longitudinal IF systems similar to those seen in vimentin-expressing myogenic systems and massive IF bundles. Exogenous desmin, on the other hand, formed a reticular IF meshwork and non-filamentous aggregates. In embryonic epithelial cells, both vimentin and desmin formed extended IF networks. Vimentin and desmin differ most dramatically in their NH2- terminal "head" regions. To determine whether the head region was responsible for the differences in the behavior of these two proteins, we constructed plasmids encoding chimeric proteins in which the head of one was attached to the body of the other. In muscle, the vimentin head- desmin body (VDD) polypeptide formed longitudinal IFs and massive IF bundles like vimentin. The desmin head-vimentin body (DVV) polypeptide, on the other hand, formed IF meshworks and non-filamentous structures like desmin. In embryonic epithelial cells DVV formed a discrete filament network while VDD did not. Based on the behavior of these chimeric proteins, we conclude that the head domains of vimentin and desmin are structurally distinct and not interchangeable, and that the head domain of desmin is largely responsible for desmin's muscle- specific behaviors.