Search

Great advances in the field of axonal transport have been made in the past year, including the identification of new molecular motors associated with microtubules and actin. In addition, studies on the mechanisms of bidirectional fast axonal transport have clarified new aspects of this process, such as the isolation of a kinesin-binding protein, kinectin, and the finding that phosphorylation regulates kinesin's dissociation from membranous organelles. New approaches to studying slow transport of cytoskeletal proteins have provided further evidence that the axonal cytoskeleton in mammalian systems is largely stationary, although a dynamic exchange occurs between polymers and a small pool of moving subunits.

The ultrastructure of detergent-resistant cytoskeletons in the noncortical cytoplasm of sea urchin eggs was studied by quick-freeze, deep-etch electron microscopy. Two different cytoskeletal organizations were identified in the detergent-treated sea urchin eggs. They were distinguished by the presence or the absence of long actin filaments and probably correspond to the cortex and the noncortical cytoplasm, respectively. The non-cortical cytoplasm was composed of a complex network (designated here as the ground network) of filaments 6 to 13 nm in diameter, that interconnected aggregates of small globular materials, yolk granules and a meshwork of uniform filaments (8-9 nm in diameter). The 6 to 13 nm filaments comprising the ground network were branched and associated with filaments of the same or other sizes, resulting in the formation of an extremely complex network. The meshwork of 8-9 nm filaments was homogeneous in composition and constitutes a novel structure which has not been previously described. The 8-9 nm filaments were connected to one another at their ends, forming a meshwork of polygons. Meshworks, ranging up to 3 microns in diameter, were distributed throughout the non-cortical cytoplasm of the egg. Similar cytoplasmic structures were also observed in fertilized eggs.

We have studied the organization of the cytoskeleton in both unstimulated and stimulated cultured chromaffin cells, as well as its relationship with their secretory process by exocytosis. We found the spatial heterogeneity in the intensity of cortical rhodamine-phalloidin staining within a cell. The overall staining pattern or intensity was minimally altered after stimulation, although dopamine-beta-hydroxylase (DBH) antigen, a marker for the chromaffin granule membrane, was exposed preferentially on the plasma membrane areas with lower intensity of rhodamine-phalloidin staining. Using the quick-freeze, deep-etch technique, we found the heterogeneity in the organization of cortical cytoskeletal networks--some regions have actin filament bundles running parallel to the plasma membrane interspersed between granules and the plasma membrane, while others have few actin filaments beneath the plasma membrane before stimulation. Actin filaments were rarely observed in the inner cytoplasm. We did not observe the overall change in its organization after stimulation. Double-label immunogold EM using anti-DBH antibody and anti-actin antibody combined with statistical analysis showed that (1) DBH was exposed on the plasma membrane preferentially where actin was sparse after stimulation (significant at less than 0.1%), although (2) regions having sparse actin were not always the sites for DBH exposure, and (3) the cortical actin zone was sometimes disrupted at the DBH-exposed sites after stimulation. The present data suggested that (1) secretion is related to heterogeneous organization of cortical cytoskeleton after stimulation and (2) massive synchronized reorganization of the cytoskeleton in the whole cell is not necessary for secretion, although small changes of the cytoskeleton might occur under local regulation at each exocytotic site at the moment of the release.

Using the smallest subunit (NF-L) of a neurofilament and a glial fibrillary acidic protein, the subunit arrangement in intermediate filaments was studied by low-angle rotary shadowing. NF-L formed a pair of 70 to 80 nm rods in a low ionic strength solution at pH 6.8. Two 70 to 80 nm rods appeared to associate in an antiparallel manner with an overlap of about 55 nm, almost the same length as the alpha-helix-rich central rod domain of intermediate filament proteins. The overlap extended for three-beaded segments, present at 22 nm intervals along the pairs of rods. The observations that (1) 70 to 80 nm rods were a predominant structure in a low ionic strength solution at pH 8.5, (2) the molecular weights of the rod and the pair were measured by sedimentation equilibrium as 190,000 and 37,000 respectively, and (3) the rods formed from the trypsin-digested NF-L had a length of about 47 nm, indicated that the 70 to 80 nm rod is the four-chain complex and the pair of rods is the eight-chain complex. Similar structures were observed with glial fibrillary acidic protein, indicating that these oligomeric structures are common to other intermediate filament proteins. NF-L assembled into short intermediate-sized filaments upon dialysis against a low-salt solution containing 1 to 2 mM-MgCl2 at 4 degrees C. The majority of these short filaments possessed four or five-beaded segments, suggesting that the pair of rods were arranged in a half-staggered fashion in neurofilaments. On the basis of these observations, we propose the following model for the intermediate filament subunit arrangement. (1) The four-chain complex is the 70 to 80 nm rod, in which two coiled-coil molecules align in parallel and in register. (2) Two four-chain complexes form the eight-chain complex by associating in an antiparallel fashion with the overlap of the entire central rod domain. (3) The eight-chain complex is the building block of the intermediate filament. The eight-chain complexes are arranged in a half-staggered fashion within the intermediate filament.

Reassembly of the neurofilament (NF) in vitro was studied by low-angle rotary shadowing electron microscopy. Various intermediate stages of the reassembly were reconstructed from the smallest molecular mass subunit (NF-L) under controlled reassembly conditions. NF-L in 6 M-urea took the form of spherical particles with a diameter of about 12 nm. NF-L aggregated into rodlets of 70 to 80 nm long in a low-salt solution at alkaline pH. By reducing the pH of the dialyzing solution to 6.6, a pair of rods was formed by association side-by-side. Increasing the temperature of low-salt solutions from 4 degrees C to 35 degrees C did not produce intermediate-sized filaments. The addition of Mg2+ to the dialyzing solution resulted in the formation of short intermediate-sized filaments even at 4 degrees C. Further dialysis of the short intermediate-sized filaments against reassembly solution containing both NaCl and MgCl2 at 37 degrees C failed to elongate them into longer filaments, suggesting that annealing does not contribute to the elongation of neurofilaments. Different roles for Mg+ and NaCl in neurofilament reassembly were indicated. While Mg2+ strengthened the lateral association between 70 to 80 nm rods, NaCl appeared to promote the end-to-end association of filaments preferentially. Longer filaments were formed by increasing the NaCl concentration. By dialyzing NF-L against a buffer containing 50 mM-NaCl in the absence of Mg2+, unraveled filaments were formed. The many unraveled filaments were composed of four 8 nm wide filaments, which have been called the subfilament or the protofibril. Time-course experiments of the reassembly were performed in the absence of Mg2+, in which condition the rate of neurofilament reassembly appeared to be reduced. Star-like clusters, about four protofibrils joined together at one end, were suggested to be the initial stage of the intermediate-sized filament formation. The following two-step elongation mechanism of neurofilaments was deduced from these results. The pairs of rods were added to the ends of the protofibrils of neurofilaments, and after all four protofibrils were elongated they were then packed into neurofilaments. Distribution of larger molecular mass subunits, NF-M and NF-H, was studied. Addition of NF-M or NF-H to NF-L did not change the assembly properties of neurofilaments. Unraveled filaments reconstituted from NF-L plus either NF-M or NF-H indicated that NF-M and NF-H are incorporated evenly into each protofibril.

The cytoskeleton that supports microvilli in intestinal epithelial cells was visualized by the quick-freeze, deep-etch, rotary-replication technique (Heuser and Salpeter. 1979. J. Cell Biol. 82: 150). Before quick freezing, cells were exposed to detergents or broken open physically to clear away the granular material in their cytoplasm that would otherwise obscure the view. After such extraction, cells still displayed a characteristic organization of cytoskeletal filaments in their interiors. Platinum replicas of these cytoskeletons had sufficient resolution to allow us to identify the filament types present, and to determine their characteristic patterns of interaction. The most important new finding was that the apical "terminal web" in these cells, which supports the microvilli via their core bundles of actin filaments, does not itself contain very much actin but instead is comprised largely of narrow strands that interconnect adjacent actin bundles with one another and with the underlying base of intermediate filaments. These strands are slightly thinner than actin, do not display actin's 53A periodicity, and do not decorate with myosin subfragment S1. On the contrary, two lines of evidence suggested that these strands, could include myosin molecules. First, other investigators have shown that myosin is present in the terminal web (Mooseker et al. 1978. J. Cell Biol. 79: 444-453), yet we could find no thick filaments in this area. Second, we found that the strands were removed completely in the process of decorating the core filament bundles with the myosin subfragment S1, suggesting that they had been competitively displaced by exogenous myosin. We conclude that myosin may play a structural role in these cells, via its cross-linking distribution, in addition to whatever role it plays in microvillar motility.

The elaborate cross-connections among membranous organelles (MO), microtubules (MT), and neurofilaments (NF) were demonstrated in unifixed axons by the quick-freeze, deep-etch, and rotary-shadowing method. They were categorized into three groups: NF-associated cross-linker, MT-associated cross-bridges, and long cross-links in the subaxolemmal space. Other methods were also employed to make sure that the observed cross-connections in the unfixed axons were not a result of artifactual condensation or precipitation of soluble components or salt during deep-etching. Axolemma were permeablized either chemically (0.1% saponin) or physically (gentle homogenization), to allow egress of their soluble components from the axon; or else the axons were washed with distilled water after fixation. After physical rupture of the axolemma or saponin treatment, most of the MO remained intact. MT were stabilized by adding taxol in the incubation medium. Axons prepared by these methods contained many longitudinally oriented NF connected to each other by numerous fine cross-linkers (4-6 nm in diameter, 20-50 nm in length). Two specialized regions were apparent within the axons: one composed of fascicles of MT linked with each other by fine cross-bridges; the other was in the subaxolemmal space and consisted of actinlike filaments and a network of long cross-links (50-150 nm) which connected axolemma and actinlike filaments with NF and MT. F-actin was localized to the subaxolemmal space by the nitrobenzooxadiazol phallacidin method. MO were located mainly in these two specialized regions and were intimately associated with MT via fine short (10-20 nm in length) cross-bridges. Cross-links from NF to MO and MT were also common. All these cross-connections were observed after chemical extraction or physical rupture of the axon; however, these procedures removed granular materials which were attached to the filaments in the fresh unextracted axons. The cross-connections were also found in the axons washed with distilled water after fixation. I conclude that the cross- connections are real structures while the granular material is composed of soluble material, probably protein in nature.

Microtubules (MT) and neurofilaments (NF) are linked by frequent crossbridges in situ. In order to answer the question of what makes these crossbridges, we performed the immunogold procedure on rat spinal cord motor neurons using an affinity-purified polyclonal antibody against rat brain MAP2 and gold-labeled anti-rabbit IgG goat IgG. A quick-freeze, deep-etch technique (QF-DE) in conjunction with decoration with anti-MAP2 antibody and ferritin-labeled second antibody was also used. In motor neuron dendrites crossbridges were clearly displayed between MTs and NFs by QF-DE. These crossbridges were revealed in thin sections as fuzzy filamentous structures between MT and NF. Gold particles studded the fuzzy structures associated with MT. Many such structures connected MTs to NFs. Furthermore, antibody complexes containing ferritin were localized on the crossbridges between MTs and NFs by the QF-DE study. In addition, we performed reconstitution experiments. We isolated 70 kDa (L) protein of neurofilaments from calf spinal cords and assembled L to form neurofilaments in vitro. MAP2 bound these neurofilaments according to both SDS-PAGE and QF-DE electron microscopy of the pellets of suspensions containing L proteins and MAP2. When we added tubulin to this suspension and polymerized it in the presence of taxol, neurofilaments were crosslinked with microtubules by MAP2 crossbridges. Hence, from these 2 approaches we concluded that MAP2 is a component of crossbridges between MTs and NFs in the neuronal cytoskeleton in vivo and in vitro.

At the terminal web of chicken intestinal epithelial cell, the actin bundles are cross-linked by a fine filamentous network of actin-associated cross-linkers. Myosin, fodrin, and TW 260/240 have been identified as major components of the cross-linkers. We studied the development of the cross-linkers by quick-freeze, deep-etch electron microscopy, and the expression of cross-linker proteins (myosin, fodrin 240, and TW 260) by immunofluorescence and immunoblotting analysis during the embryogenesis. Microvilli start to form at 5-7 days, and the rootlets begin to elongate at 10 days. At an early stage of the development of the terminal web (13 days), fodrin 240 and a small amount of myosin are expressed, and a few actin-associated cross-linkers are present between the rootlets. However, TW 260 is not expressed at this stage. At an intermediate stage (19 days), the amount of myosin increases, and TW 260 begins to be expressed. The number of cross-linkers associated with the unit length of the rootlets is 24/microns. At the final stage of the terminal web formation (2 days after hatching), the amount of fodrin 240, myosin, and TW 260 is similar to the adult level, and the number of the actin-associated cross-linkers per unit length of the rootlet is 27/microns (approximately 85% of the adult). These results suggest that the synthesis of cross-linker proteins may be intricately regulated to achieve the desired density of cross-linkages at each developmental stage: at early and intermediate stages, sufficient and not an excess of cross-linkages are formed; and at a final stage, a higher complexity of cross-linkages is achieved. In addition, there is a differential expression of the components of the actin-associated cross-linkers: myosin and fodrin could be early components of the cross-linkers involved in the basic stabilization of the terminal web structure, whereas TW 260/240 becomes incorporated later, possibly involved in the stabilization preparatory to the rapid elongation of microvilli, which occurs after the formation of the terminal web.

The cytoskeletal architecture of the postsynaptic cytoplasm in the cerebellar cortex of mice and rats was observed by quick-freeze, deep-etch electron microscopy. The postsynaptic cytoplasm was mainly filled with a network of actin filaments (approximately 8 nm in width). The tips of the actin filaments were closely associated with the true inner side of the postsynaptic membranes. However, the organization of the actin filaments was distinct depending on the types of synapses. In axosomatic synapses the actin filaments tended to run randomly and form a network while in the postsynaptic spine, such as Purkinje cell dendritic spines, the actin filaments were mainly arranged parallel to the stalk of the spines. Only a few actin filaments were found in the postsynaptic cytoplasm of some axodendritic synapses such as mossy fiber-granule cell synapses. In most cases a mesh of fine strands (approximately 6 nm in width) and granular substances was observed just underneath the postsynaptic membranes which also associated with actin filaments. The arrangement of actin filaments in the spine does not support the possibility of constriction of spines as a basis for long-term depression (LTD).

We have studied cytoskeletal architectures of isolated mitotic apparatus from sea urchin eggs using quick-freeze, deep-etch electron microscopy. This method revealed the existence of an extensive three-dimensional network of straight and branching crossbridges between spindle microtubules. The surface of the spindle microtubules was almost entirely covered with hexagonally packed, small, round button-like structures which were very uniform in shape and size (approximately 8 nm in diameter), and these microtubule buttons frequently provided bases for crossbridges between adjacent microtubules. These structures were removed from the surface of microtubules by high salt (0.6 M NaCl) extraction. Microtubule-associated proteins (MAPs) and microtubules isolated from mitotic spindles which were mainly composed of a large amount of 75-kD protein and some high molecular mass (250 kD, 245 kD) proteins were polymerized in vitro and examined by quick-freeze, deep-etch electron microscopy. The surfaces of microtubules were entirely covered with the same hexagonally packed round buttons, the arrangement of which is intimately related to that of tubulin dimers. Short crossbridges and some longer crossbridges were also observed. High salt treatment (0.6 M NaCl) extracted both 75-kD protein and high molecular weight proteins and removed microtubule buttons and most of crossbridges from the surface of microtubules. Considering the relatively high amount of 75-kD protein among MAPs isolated from mitotic spindles, it is concluded that these microtubule buttons probably consist of 75-kD MAP and that some of the crossbridges in vivo could belong to MAPs. Another kind of granule, larger in size (11-26 nm in diameter), was also on occasion associated with the surface of microtubules of mitotic spindles. A fine sidearm sometimes connected the larger granule to adjacent microtubules. Localization of cytoplasmic dynein ATPase in the mitotic spindle was investigated by electron microscopic immunocytochemistry with a monoclonal antibody (D57) against sea urchin sperm flagellar 21S dynein and colloidal gold-labeled second antibody. Immunogold particles were closely associated with spindle microtubules. 76% of these were within 50 nm and 55% were within 20 nm from the surface of the microtubules. These gold particles were sporadically found on both polar and kinetochore microtubules of half-spindles at both metaphase and anaphase. They localized also on the microtubules between sister chromatids in late anaphase. These data indicate that cytoplasmic dynein is attached to the microtubules in sea urchin mitotic spindles.(ABSTRACT TRUNCATED AT 400 WORDS)

Cytoskeletons underneath the postsynaptic membrane of neuromuscular junctions were studied by using a quick-freeze deep-etch method and immunoelectron microscopy of ultrathin frozen sections. In a quick-freeze deep-etched replica of fresh, unfixed muscles, 8.9 +/- 1.5-nm particles were present on the true postsynaptic membrane surface. Underneath this receptor-rich postsynaptic membrane, networks of fine filaments were observed. These cytoskeletal networks were more clearly observed in extracted samples. In these samples, diameters of the filaments which formed networks were measured. In the platinum replica, three kinds of filament were recognized--12 nm, 9 nm, and 7 nm in diameter. The 12-nm filament seemed to correspond to the intermediate filament. The other two filaments formed meshworks between intermediate filaments and plasma membrane. In ultrathin frozen sections vinculin label was localized just beneath the plasma membrane. Thirty-six percent of the label was within 18 nm from the cytoplasmic side of the plasma membrane and 50% was within 30 nm. Taking the size of the vinculin molecule into account, it was concluded that vinculin is localized just beneath the plasma membrane and might play some role in anchoring filaments which formed meshworks underneath the plasma membrane.

Carboxy-terminal tail domains of larger molecular mass subunits (NF-M and NF-H) of neurofilaments (NFs), which are the highly phosphorylated moieties, were observed as thin flexible filaments projecting from NF core filaments by rotary shadowing (Hisanaga and Hirokawa, 1988). Dephosphorylation of NFs has been suspected to affect the structures and the functions of the carboxy-terminal tail projections. We report here the effects of the dephosphorylation on the structure of NFs studied by electron microscopy. (1) The structures of carboxy-terminal tail projections after dephosphorylation were compared with those of the control NFs by low-angle rotary shadowing. This was examined with 2 samples; the isolated neurofilaments and the short filaments assembled from NF-H. Both the dephosphorylated NFs and the short filaments showed many projections laterally extending from core filaments similar to those observed in the control samples. (2) With respect to the structure of NF in physiological solution, the density of NFs in the precipitates was examined by thin-section electron microscopy. No difference in the density was noted between control and dephosphorylated NFs. (3) The ability to form cross-bridges in vitro was examined by quick-freeze, deep-etch electron microscopy. The structure and frequency of cross-bridges appeared to be similar in both control and dephosphorylated NFs. (4) Phosphate determination revealed that about 90% of the phosphate groups of NF-H subunit were removed by treatment with E. coli alkaline phosphatase. These results indicated that the dephosphorylation of NF did not affect the structure and the ability to form cross-bridges of the carboxy-terminal tail projections in vitro.

Replicas of the apical surface of hair cells of the inner ear (vestibular organ) were examined after quick freezing and rotary shadowing. With this technique we illustrate two previously undescribed ways in which the actin filaments in the stereocilia and in the cuticular plate are attached to the plasma membrane. First, in each stereocilium there are threadlike connectors running from the actin filament bundle to the limiting membrane. Second, many of the actin filaments in the cuticular plate are connected to the apical cell membrane by tiny branched connecting units like a "crow's foot." Where these "feet" contact the membrane there is a small swelling. These branched "feet" extend mainly from the ends of the actin filaments but some connect the lateral surfaces of the actin filaments as well. Actin filaments in the cuticular plate are also connected to each other by finer filaments, 3 nm in thickness and 74 +/- 14 nm in length. Interestingly, these 3-nm filaments (which measure 4 nm in replicas) connect actin filaments not only of the same polarity but of opposite polarities as documented by examining replicas of the cuticular plate which had been decorated with subfragment 1 (S1) of myosin. At the apicolateral margins of the cell we find two populations of actin filaments, one just beneath the tight junction as a network, the other at the level of the zonula adherens as a ring. The latter which is quite substantial is composed of actin filaments that run parallel to each other; adjacent filaments often show opposite polarities, as evidenced by S1 decoration. The filaments making up this ring are connected together by the 3-nm connectors. Because of the polarity of the filaments this ring may be a "contractile" ring; the implications of this is discussed.

The structure of the peripheral domains of neurofilaments (NFs) was revealed by rotary shadowing electron microscopy. NFs were isolated from bovine spinal cords by Sepharose CL-4B gel filtration and examined by low angle rotary shadowing. The peripheral domains appeared as thin, flexible, filamentous structures projecting from the intermediate filament core, with a constant density along their entire length. The average length of the projections was approximately 85 nm and the width about 4 nm. These projections appeared from regularly distributed sites, at 22 nm spacing, which seemed to correspond to the typical repeat of the alpha-helix-rich rod domain of the core filament. The density of the projections was found to be 4.1 (+/- 0.6) per 22 nm. We performed reconstitution experiments using purified NF polypeptides to confirm that the projection was indeed the NF peripheral domain. Individual components of the NF triplet, i.e. NF-L, NF-M and NF-H, were purified by DE-52 and Mono-Q anion exchange chromatographies in the presence of 6 M-urea and were assembled in various combinations into filaments. Reassembled filaments were somewhat more slender than the isolated NFs and exhibited a distinct 22 nm axial periodicity. While prominent projections were not observed in the filaments assembled from NF-L alone, reconstructed filaments containing NF-L plus either NF-M or NF-H revealed many projections. The average length of the projections in the filaments reconstructed from NF-L and NF-H was about 63 nm. The projections of reconstructed filaments from NF-L and NF-M were about 55 nm in length. The difference in the lengths of the projections might reflect the difference in the length of the carboxy-terminal tail domain between NF-M and NF-H. The results are interpreted to show that the carboxy-terminal tail domains of NFs project in a regular pattern from the core filament, which is consistent with a half-staggered organization of the tetrameric subunits.

Calcium plays a vital part in the regulation of cytoskeletal structure and contractility in the brush border of intestinal epithelial cells. An increased Ca2+ concentration causes a rapid but reversible solation of microvillar core filaments, which is mediated by an actin-'severing' protein of the core (Mr = 95 000), referred to as MV-95 K or villin. Results of recent experiments on the Ca2+-dependent interaction of MV-95 K with actin are summarized, and various functions for this actin-severing protein in the intestinal epithelial cell are discussed. Calcium also regulates contractility in the brush border, as shown by studies of isolated brush borders in vitro from chicken intestine. In the presence of Ca2+ (greater than 1 microM) and ATP (at 37 degrees C) isolated brush borders dramatically contract, over 1-4 min, via an isometric contraction of the terminal web region, similar to that observed by R. Rodewald on neonatal rat brush borders. This contraction is mediated, at least in part, by contraction of the circumferential bundle of actin filaments that are associated with the zonula adherens and may also involve myosin-mediated contractions between adjacent microvillar rootlets. Analysis of Ca2+-dependent phosphorylation of brush border proteins during terminal web contraction demonstrates a simultaneous phosphorylation of the regulatory light chain of brush border myosin. Like contraction, the brush border myosin kinase is activated by free Ca2+ (greater than 1 microM) and is inhibited by trifluoperazine, an inhibitor of calmodulin function. These results demonstrate that the machinery required for both production and regulation of force are integral components of the brush border cytostructure.

Terminal webs prepared from mouse intestinal epithelial cells were examined by the quick-freeze, deep-etch, and rotary-replication method. The microvilli of these cells contain actin filaments that extend into the terminal web in compact bundles. Within the terminal web these bundles remain compact; few filaments are separated from the bundles and fewer still bend towards the lateral margins of the cell. Decoration with subfragment 1 (S1) of myosin confirmed that relatively few actin filaments travel horizontally in the web. Instead, between actin bundles there are complicated networks of the fibrils. Here we present two lines of evidence which suggest that myosin is one of the major cross-linkers in the terminal web. First, when brush borders are exposed to 1 mM ATP in 0.3 M KCl, they lose their normal ability to bind antimyosin antibodies as judged by immunofluorescence, and they lose the thin fibrils normally found in deep-etch replicas. Correspondingly, myosin is released into the supernatant as judged by SDS gel electrophoresis. Second, electron microscope immunocytochemistry with antimyosin antibodies followed by ferritin-conjugated second antibodies leads to ferritin deposition mainly on the fibrils at the basal part of rootlets. Deep-etching also reveals that the actin filament bundles are connected to intermediate filaments by another population of cross-linkers that are not extracted by ATP in 0.3 M KCl. From these results we conclude that myosin in the intestinal cell may not only be involved in a short range sliding-filament type of motility, but may also play a purely structural role as a long range cross-linker between microvillar rootlets.

Axons from rats treated with the neurotoxic agent beta,beta'-iminodipropionitrile (IDPN) were examined by quick-freeze, deep-etch electron microscopy. Microtubules formed bundles in the central region of the axons, whereas neurofilaments were segregated to the periphery. Most membrane-bounded organelles, presumably including those involved in rapid axonal transport, were associated with the microtubule domain. The high resolution provided by quick-freeze, deep-etch electron microscopy revealed that the microtubules were coated with an extensive network of fine strands that served both to cross-link the microtubules and to interconnect them with the membrane-bounded organelles. The strands were decorated with granular materials and were irregular in dimension. They appeared either singly or as an extensive anastomosing network in fresh axons. The microtubule-associated strands were observed in fresh, saponin-extracted, or aldehyde-fixed tissue. To explore further the identity of the microtubule-associated strands, microtubules purified from brain tissue and containing the high molecular weight microtubule-associated proteins MAP 1 and MAP 2 were examined by quick-freeze, deep-etch electron microscopy. The purified microtubules were connected by a network of strands quite similar in appearance to those observed in the IDPN axons. Control microtubule preparations consisting only of tubulin and lacking the MAPs were devoid of associated strands. To learn which of the MAPs were present in the microtubule bundles in the axon, sections of axons from IDPN-treated rats were examined by immunofluorescence microscopy using antibodies to MAP 1A, MAP 1B, MAP 2, and tubulin. Anti-MAP 2 staining was only marginally detectable in the IDPN-treated axons, consistent with earlier observations. Anti-MAP 1A and anti-MAP 1B brightly stained the IDPN-treated axons, with the staining exclusively limited to the microtubule domains. Furthermore, thin section-immunoelectron microscopy using colloidal gold-labeled second antibodies revealed that both anti-MAP 1A and anti-MAP 1B stained fuzzy filamentous structures between microtubules. In view of earlier work indicating that rapid transport is associated with the microtubule domain in the IDPN-treated axon, it now appears that MAP 1A and MAP 1B may play a role in this process. We believe that MAP 1A and MAP 1B are major components of the microtubule-associated fibrillar matrix in the axon.

We studied the cytoskeletal reorganization of saponized human platelets after stimulation by using the quick-freeze deep-etch technique, and examined the localization of myosin in thrombin-treated platelets by immunocytochemistry at the electron microscopic level. In unstimulated saponized platelets we observed cross-bridges between: adjoining microtubules, adjoining actin filaments, microtubules and actin filaments, and actin filaments and plasma membranes. After activation with 1 U/ml thrombin for 3 min, massive arrays of actin filaments with mixed polarity were found in the cytoplasm. Two types of cross-bridges between actin filaments were observed: short cross-bridges (11 +/- 2 nm), just like those observed in the resting platelets, and longer ones (22 +/- 3 nm). Actin filaments were linked with the plasma membrane via fine short filaments and sometimes ended on the membrane. Actin filaments and microtubules frequently ran close to the membrane organelles. We also found that actin filaments were associated by end-on attachments with some organelles. Decoration with subfragment 1 of myosin revealed that all the actin filaments associated end-on with the membrane pointed away in their polarity. Immunocytochemical study revealed that myosin was present in the saponin-extracted cytoskeleton after activation and that myosin was localized on the filamentous network. The results suggest that myosin forms a gel with actin filaments in activated platelets. Close associations between actin filaments and organelles in activated platelets suggests that contraction of this actomyosin gel could bring about the observed centralization of organelles.

In order to understand the cytoskeletal architecture at the terminal web of the ciliated cell, we examined chicken tracheal epithelium by quick-freeze deep-etch (QFDE) electron microscopy combined with immunocytochemistry of fodrin. At the terminal web, the cilia ended into the basal bodies and then to the rootlets. The rootlets were composed of several filaments and globular structures attached regularly to them. Decoration with myosin subfragment 1 (S1) revealed that some actin filaments ran parallel to the apical plasma membrane between the basal bodies, and other population traveled perpendicularly or obliquely, i.e., along the rootlets. Some actin filaments were connected to the surface of the basal bodies and the basal feet. Among the basal bodies and the rootlets there existed three kinds of fine crossbridges, which were not decorated with S1. In the deeper part of the terminal web, intermediate filaments were observed between the rootlets and were sometimes crosslinked with the rootlets. Immunocytochemistry combined with the QFDE method revealed that fodrin was a component of fine crossbridges associated with the basal bodies. We concluded that an extensive crosslinker system among the basal bodies and the rootlets along with networks of actin and intermediate filaments formed a structural basis for the effective beating of cilia.