Effects of myosin and heavy meromyosin on actin-related gelation of HeLa cell extracts.

The gelation induced by warming (to 25 degrees C) the 100,000 g supernatant fraction (extract) of HeLa cells lysed in a buffer containing sucrose, ATP, DTE, EGTA, imidazole, and Triton X-100 was studied in the presence of myosin and heavy meromyosin (HMM). Myosin mixed with extract induces shrinkage of the gel, but jelled extract or myosin alone does not shrink. In the concentration range, 0.14-1.04 mg/ml of myosin, the degree of shrinkage is roughly proportional to the concentration of myosin. Supplementa MgCl2 also promotes shrinkage. HMM (0.4-0.8 mg/ml) can inhibit gel formation by extract in tubes or floated on a sucrose cushion. Gel electrophoresis of gels shrunken by added myosin or electrophoresis of the proteins which can be sedimented from extract after incubation in the presence of HMM indicate that both myosin and HMM interfere with the changes in sedimentability of the high molecular weight protein (HMWP) thought to participate (together with actin) in gel formation in HeLa cell extracts (R. R. Weihing, 1976. J. Cell Biol. 71:303-307). These results, together with previous results showing that actin is present and that HMWP is enriched in the plasma membrane fraction of HeLa cells (R. R. Weihing, 1976. Cold Spring Harbor Conf. Cell Proliferation. 3:671-684), point to the possibility of dynamic changes in the interactions of HMWP or myosin with actin in processes of movement occurring at the cell surface.

Cytochalasin B inhibits actin-related gelation of HeLa cell extracts.

When the 100,000 g supernatant fraction (extract) of HeLa cells lysed in a buffer containing sucrose, ATP, DTE, EGTA, imidazole, and Triton X-100 is incubated at 25 degrees C, it gels, and actin and a HMWP are progressively enriched in the extract and in gel isolated from extract. CB (greater than or equal to 0.25 muM) inhibits gelation and specifically lowers the concentrations of actin and the HMWP in the fraction which sediments at 100,000 g after incubation. These results indicate that actin and HMWP are partly disaggregated by cytochalasin treatment, and thus that their aggregation is related gelation. Inasmuch as previous results showed that actin is present and HMWP is enriched in the plasma membrane fraction of HeLa cells, the results also point to a possible relation between plasma membrane-associated gel and in vivo effects of CB.

Actin associated with membranes from 3T3 mouse fibroblast and HeLa cells.

A protein component of membranes isolated from 3T3 mouse fibroblasts and HeLa cells has been identified as actin by peptide mapping. Extensive but apparently not total coincidence was found between the peptide maps of these two nonmuscle membrane-associated actins compared to chick skeletal muscle actin. Between 2 and 4 percent of the total membrane protein appears in the actin band on sodium dodecyl sulfate polyacrylamide gels of 3T3 membranes while about 4 percent of the membrane protein appears as the actin band from HeLa membranes. These values represent approximately the same proportion of actin to total protein found in the cell homogenates. Treatment of intact cells with levels of cytochalasin B sufficient to cause pronounced morphological changes did not change the amount of actin associated with the membrane in either 3T3 or HeLa cells. However, incubation of isolated membranes under conditions favoring conversion of actin from filamentous to monomeric form resulted in dissociation of approximately 80 and 60 percent of the actin from 3T3 and HeLa membranes, respectively. Thus, approximately 20 percent of 3T3 membrane actin and 40 percent of HeLa membrane actin remained associated with the membrane even under actin depolymerizing conditions.

In vitro polymerization of microtubules from HeLa cells.

Although the purification of microtubules from brain by alternate cycles of polymerization and depolymerization in vitro has become routine, the application of this method to non-neural cultured cells has been less successful. Previous investigations have suggested that it was necessary to use substrate-grown cells and 4 M glycerol to obtain microtubules from cultured cells. We have developed a method for preparing microtubules from HeLa cells in spinner cultures without the use of glycerol. Microtubules can be readily carried through two complete cycles of polymerization at 37 degrees C and depolymerization at 4 degrees C in vitro. The microtubules obtained are morphologically similar to brain microtubules in electron micrographs, and the tubulin subunits have mobilities similar to those of brain tubulins on polyacrylamide gels. Typical yields in the second polymerization pellet are about 1 mg protein/ml of packed cells or 2.5-3.0% of the total protein in the soluble cell extract. The major nontubulin protein present after two cycles of polymerization and depolymerization has an apparent mol wt of 68,000 daltons. If glycerol is used during polymerization, this band is virtually absent.

Microtubule-associated proteins of HeLa cells: heat stability of the 200,000 mol wt HeLa MAPs and detection of the presence of MAP-2 in HeLa cell extracts and cycled microtubules.

One of the major groups of microtubule-associated proteins (MAPs) found associated with the microtubules isolated from HeLa cells has a molecular weight of just over 200,000. Previous work has demonstrated that these heLa MAPs are similar in several properties to MAP-2, one of the major MAPs of mammalian neural microtubules, although the two types of proteins are immunologically distinct. The 200,000 mol wt HeLa MAPs have now been found to remain soluble after incubation in a boiling water bath and to retain the ability to promote tubulin polymerization after this treatment, two unusual properties also shown by neural MAP-2. This property of heat stability has allowed the development of a simplified procedure for purification of the 200,000 HeLa MAPs and has provided a means for detection of these proteins, even in crude cell extracts. These studies have also led to the detection of a protein in crude extracts of HeLa cells and in cycled HeLa microtubules which has been identified as MAP-2 on the basis of (a) comigration with calf brain MAP-2 on SDS PAGE, (b) presence in purified microtubules, (c) heat stability, and (d) reaction with two types of antibodies prepared against neural high molecular weight-MAPs, one of these a monoclonal antibody against hog brain MAP-2, although present in HeLa cells, is at all stages of microtubule purification a relatively minor component in comparison to the 200,000 HeLa MAP's.

Increased visualization of microtubules by an improved fixation procedure.

We have found that when a buffer utilized for in vitro polymerization of microtubules, i.e., 1 mM guanosine triphosphate, 1 mM MgSO4, 2 mM ethylene glycol bis(beta-aminoethyl ether)-N, N'-tetraacetic acid 100 mM piperazine-N,N'-bis(2-ethanesulfonic acid), pH 6.9 polymerization mix, was used in the glutaraldehyde prefixation regimen instead of classical fixative buffers, i.e., isotonic cacodylate or phosphate buffer, the following features were observed in thin-sections of the cytoplasm of interphase HeLa cells: (a) a greater than 2-fold increase in total microtubule contour length, (b) a 2-fold increase in a number of microtubules greater than or equal to 1 mu long, (c) an enhanced association of microtubules with cytoplasmic organelles, and (d) an increased clustering of 100 A filaments located in a perinuclear region of the cell. Furthermore, we found that after we incubated purified chick brain microtubules on a Sephadex G-25 column pre-equilibrated with polymerization mix, cacodylate or phosphate buffer at 37 degrees C, and then eluted the microtubules at 37 degrees C, the exposure to cacodylate or phosphate buffer caused extensive depolymerization, but exposure to polymerization mix buffer allowed reisolation of highly polymerized microtubules. Our results imply that prefixation with cacodylate or phosphate buffered glutaraldenyde destabilizes microtubules leading to the decreased visualization of microtubules.

The filamins: properties and functions.

The filamins are a group of homologous proteins defined by their high native molecular weight (500,000), their amino acid compositions, their cross-reactivity to antibodies to heterologous filamins, their localization to actin networks and bundles in situ, and their ability to cross-link actin filaments in vitro into three-dimensional networks and bundles. Native filamins contain two subunits (relative mass = 250 000). Each subunit carries at least one actin-binding site and formation of bivalent dimers is therefore believed to explain filamin's ability to cross-link actin filaments. Formation of networks in vitro (corresponding to formation of macroscopic gels) has been analyzed using the theory of Flory. As predicted, a sharp transition to gel (at the critical gelation concentration of filamin) is observed when actin is mixed with increasing concentrations of filamin and the critical gelation concentration is found to vary inversely with the length of actin filaments. However, the measured values of the critical gelation concentration are all higher (2- to 14-fold) than predicted by the theory and the prediction that the critical concentration varies directly with the actin concentration was verified with only one of two techniques used. Filamin's length (160-190 nm) and flexibility (1000-fold greater than actin filaments) may make it especially well fitted to cross-link actin filaments into three-dimensional networks when present in low molar ratios (1:200 to 1:50) relative to actin. At higher molar ratios (greater than 1:20) it also cross-links actin filaments into bundles. Assuming that filamin actually helps organize supramolecular structures inside cells (not yet tested directly), then its concentration relative to actin may help determine whether networks or bundles are formed. Other factors that may influence its localization and function inside cells include competition with other actin-binding proteins (such as myosin and tropomyosin) for binding sites on actin and phosphorylation, which may alter its ability to bind to actin.

The cytoskeleton and plasma membrane.

The major cytoskeletal elements (microfilaments, microtubules, and 10-nm filaments) are frequently found attached to or near the plasma membrane in arrays which can sometimes be shown experimentally to be related to cell form and movement. Ultrastructural investigations show that attachment is direct, by amorphous electron-dense material, by cell-cell junctions, or by cell-substrate attachment sites, but the chemistry of attachment is poorly understood. The structural and functional polarity of the attached elements has been defined for some microfilaments attached to plasma membrane, but this important parameter has been investigated only slightly for microtubules and not at all for 10-nm filaments attached to membranes. Assemblies of cytoskeletal elements and plasma membrane evidently are structurally stable enough to be observable by electron microscopy and to survive isolation by the usual biochemical techniques, but observations of living cells show that many assemblies of cytoskeleton and plasma membrane undergo rearrangement and interconversion. The biochemical basis and physiological meaning of many of these changes are poorly understood.

Purification of a HeLa cell high molecular weight action binding protein and its identification in HeLa cell plasma membrane ghosts and intact HeLa cells.

The high molecular weight protein (HMWP) which was previously observed to be a major component of the actin based gels formed by incubating cytoplasmic extracts of HeLa cells at 25 degrees C [Weihing, R. R. (1977) J. Cell Biol. 75, 95-103] has now been purified by gel filtration of 0.6 M KCl extracts of precipitated gels. A few hundred micrograms of HMWP, which is about 90% pure, can be isolated from 4 X 10(9) cells. HMWP can gel muscle actin and cross-link it into filament bundles. Its subunit molecular weight is 250 0000, its Stokes radius is 125 A, and its sedimentation coefficient is 9 S. A native molecular weight of 480 000 was calculated by using the latter two parameters, and therefore the native molecule is a dimer. Its amino acid analysis is nearly indistinguishable from that of macrophage actin binding protein and of mammalian and avian filamins. All of these findings indicate that HMWP is homologous to the latter proteins. However, HeLa cell HMWP and avian filamin must differ in their primary sequences because their partial peptide maps are distinct and because an antiserum against HMWP reacts only weakly with filamin. For studies on the intracellular location of HMWP, a goat antiserum against purified HMWP was prepared and characterized and then used to localize HMWP in suspension grown cells. The technique of immunoblotting revealed that the antiserum reacted virtually exclusively with the high molecular weight polypeptide that comigrates with HMWP in cell lysates and in ZnCl2-stabilized plasma membrane ghosts prepared from HeLa cells [Gruenstein, E., Rich, A., & Weihing, R. R. (1975) J. Cell Biol. 64, 223-234] and that it did not react with rabbit myosin heavy chain, microtubule proteins (MAPS and tubulin) from HeLa cells and calf brain, or the proteins of human erythrocyte ghosts including spectrin. Suspension-grown cells which were stained with the antiserum by the technique of indirect immunofluorescence showed bright fluorescence at the rim of the cells and less intense generalized fluorescence. If preimmune serum or immune serum treated with HMWP was substituted for the immune serum, then staining at the rim was not observed, but the generalized fluorescence was only slightly reduced; unpermeabilized cells were not stained. These results indicate that HMWP is a component of the cortical cytoplasm of HeLa cells. Possible functions of cortical HMWP are discussed briefly.

Actin-binding and dimerization domains of HeLa cell filamin.

HeLa cell filamin is a linear, bivalent, homodimer of high molecular weight subunits (Mr 250,000 that may cross-link actin filaments in vivo into supramolecular structures such as networks and bundles. We used millimolar Ca protease from chicken breast muscle to cleave the subunit into smaller fragments that we mapped with respect to the overall structure of the dimer. The enzyme cleaves HeLa filamin into a larger (Mr 192,000) and a smaller (Mr 104,000) fragment; the smaller fragment is the precursor of a still smaller (Mr 92,000) fragment. Only the larger fragment bound to actin in a cosedimentation test, suggesting that it contains the actin-binding region of the subunit. Digestion of HeLa filamin that had been cross-linked with dimethyl adipimidate produced a good yield of the Mr 192,000 fragment but a poor yield of the Mr 104,000/92,000 fragments. Since native filamins are head-to-head dimers, it was expected that cross-linking would proceed most readily at the dimerization site and, therefore, it appears that the Mr 192,000 fragment is cleaved from cross-linked filamin because it is distal to the dimerization region, while the Mr 104,000/92,000 fragments are absent because they lie at the dimerization region and were cross-linked to a form that was not identifiable by sodium dodecyl sulfate electrophoresis.

Adenovirus binds to rat brain microtubules in vitro.

We have found by negative staining electron microscopy that when similar concentrations of adenovirus and reovirus (viruses of about the same diameter, 75 to 80 nm, and density, 1.34 to 1.36 g/cm3) were incubated with a carbon support film containing microtubules, 72% of adenovirus on the grid, but only 32% (equivalent to random association) of reovirus, were associated with microtubules. Similar concentrations of both larger and smaller particles, such as polystyrene latex spheres and coliphage f2, also exhibited a low degree of interaction, viz., 17 to 37%, with microtubules. Moreover, 90% of microtubule-associated adenovirus binds to within +/- 4 nm of the edge of microtubules, but lower fractions (again equivalent to a random association) of the other particles bind to the edge of the microtubules. The mechanism behind this phenomenon, which we denote as "edge binding," is presently obscure; however, it provides us with a second, albeit empirical, method to distinguish between the microtubular association of adenovirus and other particles. We found that edge binding of adenovirus also occurred when adenovirus was initially placed on the carbon support film and then incubated with microtubules and when adenovirus and microtubules were mixed prior to placement on the support. In contrast, reovirus or the other particles prepared by similar techniques exhibited a random amount of edge binding. The binding of adenovirus appears to involve the hexon capsomers of the virion since (i) high resolution electron micrographs showed that the edge of the virus was in contact with the edge of the microtubules, and (ii) adenovirions briefly treated with formamide to remove pentons and fibers bind as efficiently as intact virions. Core structures, which were obtained by further formamide degradation of the virion, do not associate with microtubules. These observations support the hypothesis of Dales and Chardonnet (1973) that the transport of adenovirions within infected cells is mediated by interaction with microtubules.

Striated paracrystals that contain HMWP, the homolog of actin-binding protein and filamin from HeLa cells.

HMWP (high molecular weight protein), a high molecular weight actin binding protein, was previously isolated from HeLa cells; its physical properties, amino acid composition, and intracellular localization indicated its homology with actin-binding protein and filamin [Weihing, 1982, 1983]. We now report the identification of HMWP in striated paracrystals. Purified HMWP is incubated at 25 degrees C and subjected to negative staining with uranyl acetate. Examination by electron microscopy reveals long, striated paracrystals formed from filaments a few nanometers in diameter that lie parallel to the long axis of the paracrystal. At intervals of about 200 nm, the filaments are crossed by granular aggregates, accounting for the striated appearance. Treatment of the paracrystals with an affinity-purified antibody to HMWP decorates the filaments; such decorations are not observed if nonimmune goat IgG or phosphate-buffered saline are substituted for the antibody. Electron microscopic and electrophoretic analysis of paracrystals sedimented onto grids by centrifugation at 864 g reveals that the grids are covered with paracrystals and the major polypeptide present on grids centrifuged in parallel is HMWP. Taken together, these data indicate that the filaments of the paracrystals contain elongated molecules of HMWP. Additional experiments are needed to decide if the paracrystals form by self-association between HMWP molecules or by association with one or more of the minor polypeptides that remain in the purified HMWP.

Purification and reconstitution of HeLa cell microtubules.

Microtubules from suspension cultures of HeLa cells have been purified by carrying them through four complete cycles of polymerization at 37 degrees C and depolymerization at 4 degrees C. These microtubules show, in addition to the major alpha- and beta-tubulin components, major proteins with molecular weights of 201 000-206 000 (comprising 4.5% of the total protein), proteins with molecular weights of 97 000, 100 000, 104 000, and 114 000 (together comprising approximately 2% of the total protein), and minor components with molecular weights of 68 000 and 151 000. HeLa microtubules have also been reconstituted from purified HeLa tubulin and proteins from HeLa microtubules separated from tubulin by DEAE-cellulose column chromatography. Experiments on the fractionation and reconstitution of both two- and four-cycle microtubules suggest that the 201 000-206 000-dalton proteins are incorporated into microtubules and promote tubulin polymerization. Microtubules formed by fractionationand reconstitution of two-cycle microtubules also contain several other proteins with molecular weights of 132 000, 146 000, 151 000, 160 000, and 284 000, although these are not present in microtubules carried through four assembly-disassembly cycles. Evidence is also presented which shows that a 68 000-dalton protein which is a prominent component of HeLa microtubules after two polymerization-depolymerization cycles does not stoichiometrically copurify with tubulin through repeated assembly--disassembly cycles and does not stimulate tubulin polymerization. On the other hand, the sedimentation of this 68 000-dalton protein is apparently influenced by the presence of polymerized microtubules, suggesting that this protein may be a component of a system whjich interacts weakly with microtubules. Finally, evidence is presented suggesting that two-cycle microtubules contain a proteolytic activity that can digest the 201 000-206 000-dalton proteins.

Binding of adenovirus to microtubules. II. Depletion of high-molecular-weight microtubule-associated protein content reduces specificity of in vitro binding.

A specific in vitro association between adenovirus and pruified rat brain microtubules has been previously demonstrated (R. B. Luftig and R. R. Weihing, 1975). When examined by negative-staining electron microscopy, approximately 90% of the virus associated with microtubules was edge bound, i.e., associated within +/-4 nm of the microtubule edge. Similar results are now found for the association of adenovirus with purified chick brain microtubules. When the content of the high-molecular-weight proteins (MAPs) normally present as projections on the surface of microtubules is depleted by fractionation of cold-depolymerized microtubules on agarose A-15M columns or by brief treatment of polymerized microtubules with trypsin, the percentage of edge-bound microtubule-associated viruses is reduced to a level close to that found for particles such as reovirus, coliphage f2, or polystyrene latex spheres, which randomly associate with microtubules (54 to 64% for column-fractionated microtubules; 45 to 68% for trypsin-treated microtubules). Counts of adenovirus particles specifically bound to microtubules, corrected for variations in microtubule and virus concentrations, gave values 2.5 to 3.5 times higher for unfractionated microtubules than for microtubule-associated protein-depleted microtubules. These results are consistent with the suggestion that the specific association between adenovirus and microtubules is mediated by microtubule-associated proteins.

Prostatic marker-negative amphicrine carcinoma of the prostate.

It has been shown that prostatic adenocarcinoma differentiation correlates with prostatic-specific marker and neuroendocrine expression; that is, the more undifferentiated the tumor, the less it expresses prostatic markers and the more neuroendocrine cells are found in it. Complete absence of prostatic markers together with marked neuroendocrine expression has been associated with small cell morphology. This report describes a case of a metastatic, prostatic marker-negative, non-small cell prostatic adenocarcinoma with a prominent neuroendocrine component. The architecturally well-organized luminal-exocrine cells appeared ultrastructurally undifferentiated, however, displaying an almost empty cytoplasm. This contrasted with the prostatic marker-positive control cases of prostatic carcinoma, which contained relatively numerous cytoplasmic vacuoles. The neuroendocrine cells could be identified by light microscopy as eosinophilic cells. The number of the latter cells was markedly increased in the metastatic foci compared with the primary tumor. Light microscopically and ultrastructurally, the eosinophilic cells in this case differed from the Paneth-like cells described in prostatic carcinoma in previous reports. This case provides support for the general concept of multidirectional differentiation in human epithelial cancers and in particular for the association of poor tumor differentiation with neuroendocrine expression in prostatic carcinoma. In addition, in contrast with previous reports describing absence of basement membrane in metastatic foci of prostatic carcinoma, in the current case well-formed basal laminae were identified.

Hepatobiliary and pancreatic mucinous cystadenocarcinomas with mesenchymal stroma: analysis of estrogen receptors/progesterone receptors and expression of tumor-associated antigens.

Hepatobiliary and pancreatic mucinous cystadenocarcinomas with mesenchymal stroma are relatively rare neoplasms that occur preponderantly in women, suggesting a role for unidentified sex-specific factor(s) in the pathogenesis of these tumors. We used paraffin tissue immunohistochemical analysis with an appropriate panel of monoclonal antibodies to look for estrogen and progesterone receptors in two cases of hepatobiliary mucinous cystadenocarcinoma with mesenchymal stroma and one case of pancreatic mucinous cystadenocarcinoma. In all three of these cases, the nuclei of tumor stroma and, in the hepatic tumors, the nuclei of tumor epithelium, reacted with both antibodies. These data strongly suggest that a relationship to hormonal functions exists for these tumors. Because of the rarity of these tumors we also investigated the expression of a variety of oncoprotein antigens, epithelial antigens, and cytoskeletal antigens. The oncoprotein antigens, p53 and c-erbB-2, were focally expressed in hepatic and pancreatic tumor epithelium; bcl-2 was focally expressed in hepatic tumor epithelium. Keratin was strongly expressed in most epithelial cells. In addition, epithelial membrane antigen, carcinoembryonic antigen, and chromogranin were focally expressed in epithelial cells. Actin and vimentin were strongly expressed in most stromal cells but not in epithelial cells, and desmin expression was similar but less widespread.

Reovirus serotypes 1 and 3 differ in their in vitro association with microtubules.

Utilizing negative-stain electron microscopy in which similar concentrations of reovirus types 1 and 3 are incubated with a carbon support film containing chick brain, rabbit brain, or HeLa cell microtubules, 81% of the type 1 and 56% of type 3 exhibited an association with the apparent "edge" of the microtubule. This implies that there is a high level of specific affinity for type 1 but not for type 3 to microtubules, since it has previously been determined that only 50% of randomly associated particles would be associated with the edge. The high edge binding of reovirus type 1 is virtually independent of the origin of microtubule, or of whether microtubules or virus has been initially adhered to the support film. On the other hand, reovirus type 1-specific antiserum reduced the edge binding or reovirus type 1 to 45%, whereas type 3 specific antiserum caused no less (within the variability of the assay) of the edge binding of reovirus type 1 to microtubules (76% edge bound). High edge binding of reovirus type 1 to microtubules is correlated with the presence of type 1 or sigma 1 polypeptide. This minor outer capsid polypeptide is encoded in the S1 double-stranded RNA segment and is the viral hemagglutinin and neutralization antigen. Recombinant reovirus clones containing the S1 double-stranded RNA segment of type 1 (80 and 802) show about 85% edge binding, as compared to a value of 42% for clones and the S1 gene of type 3 (204. Electron microscopy of purified reovirus types 1 and 3 by negative staining reveals that type 1 and 802 capsomers are distinctly visualized, whereas those of type 3 and 204 appear diffuse. Thus, the greater in vitro binding of type 1 to microtubules may reflect an increased accessibility of certain of its outer capsomers, and thereby, sigma 1 polypeptides to microtubules. Examination of its outer sections of reovirus type 1- and 3-infected cells at 24 to 48 h postinfection at 31 degrees C showed that about eight times as many viral factoris in type 1-infected cells exhibited an extensive association of virus particles with microtubules, as compared to viral factories of type 3-infected cells. Thus, both in vivo and in vitro there appears to be a greater specificity for the association of reovirus type 1 particles with microtubules, as compared to reovirus type 3 particles.