The assembly of tubulin into membranes

Plasma membrane tubulin

The association of tubulin with the plasma membrane comprises multiple levels of penetration into the bilayer: from integral membrane protein, to attachment via palmitoylation, to surface binding, and to microtubules attached by linker proteins to proteins in the membrane. Here we discuss the soundness and weaknesses of the chemical and biochemical evidence marshaled to support these associations, as well as the mechanisms by which tubulin or microtubules may regulate functions at the plasma membrane.

Diabetic rat serum has an increased capacity to inhibit brain microtubule formation in vitro

The assembly of pig brain microtubule proteins was measured in vitro in the presence of serum from control rats and rats that had been rendered diabetic with 50 mg/kg streptozotocin 14 d previously. Control serum inhibited total microtubule assembly and increased the lag time before assembly commenced. Serum from diabetic animals was significantly more potent in both respects. The effect on lag time was reproduced in a predominantly albumin-containing fraction of serum that had been fractionated by affinity chromatography. Glycosylation of rat albumin in vitro led to an increase in its ability to increase polymerization lag time, but the concentration of albumin required was greater than that found in the serum fractions. The results indicate that diabetic serum contains factors that can adversely affect microtubule formation and that part of this effect may be caused by the presence of glycosylated albumin. This phenomenon may underlie some of the complications associated with diabetes.

Kainate reversibly aggregates brain tubulin in vitro

The effect of kainate, an heterocyclic analogue of glutamate on tubulin polymerization was studied. We demonstrate that kainate induces a dose-dependent aggregation of rat brain tubulin either purified by DEAE-Sephadex A-50 or in the presence of MAPs into a mesh-like structure. Such polymer is cold-, CaCl2- and colchicine-insensitive. Removal of kainate from the incubation medium yields free tubulin competent for polymerizing into normal microtubules in the presence of GTP.

Colchicine-induced alteration of hormone-stimulated cyclic AMP synthesis in pig skin (epidermis)

Effects of colchicine on the epidermal adenylate cyclase systems were investigated. When pig skin (epidermis) was incubated in RPMI 1640 medium without the addition of serum, the beta-adrenergic adenylate cyclase response (epinephrine-induced cyclic AMP accumulations) gradually decreased, whereas adenosine and histamine responses remained high or increased during the long-term (up to 48 h) incubation period. The addition of colchicine (1 mumol/liter) in the incubation medium resulted in an increase in the beta-adrenergic responsiveness and a decrease in adenosine and histamine responsivenesses. The effects of colchicine were both time- and concentration-dependent; they could be observed after 9-12 h incubation, and the maximal effect was obtained at a concentration of 0.1 mumol/liter. Similar effects were observed by the addition of another microtubule-disruptive agent, vinblastine. On the other hand, cytochalasin B, which affects the microfilament system, apparently decreased the beta-adrenergic response and increased adenosine and histamine responses during the long-term incubation period. The addition of serum in the incubation medium resulted in essentially the same effect as that of colchicine; in the presence of serum, colchicine-treated skin responded much more markedly to epinephrine (and much less to adenosine and histamine) than the control skin after 24- and 48-h incubation. Previously we reported that hydrocortisone has similar potentiating effects on the beta-adrenergic system of epidermis. The comparison of the effects of both compounds revealed that colchicine had a stronger effect than hydrocortisone, and furthermore, the simultaneous addition of both compounds (colchicine and hydrocortisone) in the incubation medium resulted in the more marked increase of beta-adrenergic response than the single addition of each chemical. Our overall results, coupled with the finding that hydrocortisone has no toxic effects on the adenosine- or histamine-adenylate cyclase system of epidermis, suggest that colchicine affects epidermal adenylate cyclase systems probably through a mechanism that is independent of glucocorticoid (hydrocortisone) effect.

Morphological alterations in dorsal root ganglion neurons and supporting cells of organotypic mouse spinal cord-ganglion cultures exposed to taxol

Explants of 14-day fetal mouse spinal cord with attached dorsal root ganglia, which had become differentiated over 2-3 weeks in culture, were exposed to 1-2 microM taxol for up to 6 days. The culture medium was supplemented with nerve growth factor (300 units/ml) during exposure to the drug. By 3-6 days in taxol, unusually numerous microtubules were seen in peripheral perikaryal and proximal neuritic regions of ganglion neurons. Microtubules also engirdled massive aggregations of pleomorphic vesicular/cisternal elements in many neurons. These aggregates were visible as unusual 'clear' spheroidal regions in the living cells, and were often as large as the nuclei. Some of the elements comprising these striking vesicular/cisternal accumulations appeared to be portions of disrupted Golgi complexes normally polarized around the cytocentrum, as well as hypertrophied smooth endoplasmic reticulum formations. In other neuronal areas, Golgi complexes and other organelles were altered or disrupted to lesser degrees. Ordered microtubular arrays occurred along endoplasmic reticulum cisternae both in neuron somata and neurites. Over time, a plethora of microtubules assembled throughout the perikarya in various orientations apparently unrelated to microtubule organizing centers. Unlike the effects of other plant alkaloids that interact with tubulin, there was no discernible increase in filaments, although their distribution appeared altered. Concentric ordered microtubular-macromolecular lamellated complexes were seen only in neurites. Neuronal nuclei were misshapen, often displaced, and displayed fine structure reminiscent of chromatolysis. Satellite and Schwann cells contained atypically abundant microtubules, abnormal cisternae, disrupted Golgi complexes, and increased lysosomes. Some nuclei displayed abnormal chromatin, and in rare cases even microtubules. We suggest that taxol alters the distribution, integrity, and/or organization of organelle systems in dorsal root ganglion cells by engendering unusually abundant microtubules in abnormal groupings and aberrant locations in these cells.

Subcellular localization of tubulin in chick retina

Tubulin was measured through [3H]colchicine-binding in membrane and soluble components of chick retinal subcellular fractions. Total tubulin content was concentrated in the synaptosomal and rod outer segment fractions. Although in total retinal homogenate only 20% of total tubulin was associated to the membrane, in synaptosomes and photoreceptor outer segments, up to 50% of tubulin was bound to the membrane fraction. Results raise the possibility of tubulin participation in transmembrane phenomena which are common to transmitter release and photoexcitation.

Reconstitution of ciliary membranes containing tubulin

Membranes from the gill cilia of the mollusc Aequipecten irradians may be solubilized readily with Nonidet P-40. When the detergent is removed from the solution by adsorption to polystyrene beads, the proteins of the extract remain soluble. However, when the solution is frozen and thawed, nearly all of the proteins reassociate to form membrane vesicles, recruiting lipids from the medium. The membranes equilibrate as a narrow band (d = 1.167 g/cm3) upon sucrose density gradient centrifugation. The lipid composition of reconstituted membranes (1:2 cholesterol:phospholipids) closely resembles that of the original extract, as does the protein content (45%). Ciliary calmodulin is the major extract protein that does not associate with the reconstituted membrane, even in the presence of 1 mM calcium ions, suggesting that it is a soluble matrix component. The major protein of reconstituted vesicles is membrane tubulin, shown previously to differ hydrophobically from axonemal tubulin. The tubulin is tightly associated with the membrane since extraction with 1 mM iodide or thiocyanate leaves a vesicle fraction whose protein composition and bouyant density are unchanged. Subjecting the detergent-free membrane extract to a freeze-thaw cycle in the presence of elasmobranch brain tubulin or forming membranes by warming the extract in the presence of polymerization-competent tubulin yields a membrane fraction with little incorporated brain tubulin. This suggests that ciliary membrane tubulin specifically associates with lipids, whereas brain tubulin preferentially forms microtubules.

The ATPase activity in brain microtubule preparations is membrane-associated

Microtubule protein prepared from bovine brain by a temperature-dependent assembly-disassembly procedure contained Mg2+- or CA2+-stimulated ATPase activity. However, activity decreased with successive cycles of assembly-disassembly such tht 15% of the Mg2+-stimulated and 31% of the Ca2+-stimulated activity of the second-cycle material remained after seven cycles. Microtubule preparations purified by three cycles of assembly-disassembly contained many membrane fragments and vesicles which were absent in microtubule preparations cycled eight times. Histochemistry and electron microscopy revealed that much of the activity is associated with the vesicles. Vesicles with an accumulation of led phosphate deposits (indication of ATPase activity) were observed in high-speed pellets (150,000 g, 60 min) of microtubule-associated proteins. Most of the activity in the microtubule-associated protein preparations, but only a fraction of the total protein is pelleted. 53-78% of the ATPase activity, but only 6% of the total protein, is recovered in a microtubule-associated protein fraction eluted from phosphocellulose with 0.17 M NaCl. Polypeptides resolved on SDS polyacrylamide gradient gels have estimated molecular weights of 30,000-76,000. Electron micrographs of this material revealed short filaments, vesicles, and small ring-like structures. None of the inhibitors of possible contaminating ATPases affected the ATPase activity.

Protein synthesis and transport in the regenerating goldfish visual system

The nature of the proteins synthesized in the goldfish retina and axonally transported to the tectum during optic nerve regeneration has been examined. Electrophoretic analysis of labeled soluble retinal proteins by fluorography verified our previous observation of a greatly enhanced synthesis of the microtubule subunits. In addition, labeling of a tubulin-like protein in the retinal particulate fraction was also increased during regeneration. Like soluble tubulin, the particulate material had an apparent MW of 53-55K and could be tyrosylated in the presence of cycloheximide and [3H]tyrosine. Comparison of post-crush and normal retinal proteins by two-dimensional gel electrophoresis also revealed a marked enhancement in the labeling of two acidic 68-70K proteins. Analysis of proteins slowly transported to the optic tectum revealed changes following nerve crush similar to those observed in the retina, with enhanced labeling of both soluble and particulate tubulin and of 68-70K polypeptides. the most striking change in the profile of rapidly transported protein was the appearance of a labeled 45k protein which was barely detectable in control fish.

Association between endocrine pancreatic secretory granules and in-vitro-assembled microtubules is dependent upon microtubule-associated proteins

By use of dark-field light microscopy, secretory granules isolated from the anglerfish endocrine pancreas were observed to attach to and release from microtubules assembled in vitro from brain homogenates. Secretory granules only bound to microtubules assembled in the presence of microtubule-associated proteins (MAPs) and not to microtubules assembled from purified tubulin. The addition of a MAP fraction to purified tubulin restored secretory granule binding. The secretory granules were released from MAP-containing microtubules by the addition of Mg-ATP but not by other nucleotides. The number of secretory granules bound to MAP-containing microtubules was increased in the presence of cyclic AMP. In addition to the associations of secretory granules with microtubules, MAP-containing microtubules also associated with each other. These laterally associated microtubules were dispersed by the addition of Mg-ATP. Electron micrographs confirmed that the associations between MAP-containing microtubules and secretory granules as well as the associations of microtubules with one another were mediated by the high molecular weight MAPs known to project from the surface of in-vitro-assembled microtubules.

The characterization of tubulin in CNS membrane fractions

Rough endoplasmic reticulum (RER), smooth endoplasmic reticulum (SER), and a plasma membrane (PM) fraction enriched in synaptic membranes were isolated from rat forebrain. The proteins in these membrane fractions were analyzed by two-dimensional gel electrophoresis (2DGE) in the isoelectric range of 5.1 to 6.0 by a modification of the O'Farrell procedure. Proteins were detected by Coomassie Brilliant Blue staining of the electrophoretograms. The results of these analyses were compared with 2DGE analysis of cytosol proteins, with particular attention given to tubulin subunits and actin. The RER contained one major protein (53K 5.4) in the beta-tubulin region with a molecular weight of 53,000 and an isoelectric point of 5.4. The SER contained at least two major proteins in the alpha-tubulin region; one with a migration identical to 53K 5.4 and other proteins with slightly higher apparent molecular weights and more acidic isoelectric points (54K, 5.4 to 5.3), identical to cytoplasmic beta-tubulin. The PM fraction also contained multiple overlapping proteins (54K, 5.4 to 5.3) in the beta-tubulin area and a trace amount of the 53K 5.4 protein. The proteins in the beta-tubulin region were removed from the 2DGE electrophoretogram and digested by Staphylococcus aureus V8 protease, and the peptides separated on one-dimensional polyacrylamide gels. The peptide patterns of 53K 5.4 protein from RER and SER were almost identical and differed significantly from the cytoplasmic beta-tubulin pattern; however, the peptide maps of the PM and SER beta-tubulin region were identical to the cytoplasmic beta-tubulin. The 2DGE analysis of RER did not contain proteins in the region of cytoplasmic alpha-tubulin. SER and PM contained proteins in the alpha-tubulin region with a similar, but not identical, peptide analysis to cytoplasmic alpha-tubulin. Significant amounts of actin were detected in 2DGE analysis of SER and PM, and the peptide analysis of the actin was identical to the cytoplasmic actin analysis. The RER fraction contained only trace amounts of actin. The cytosol and all membrane fractions contained a protein (68K 5.6) found among microtubule-associated proteins, as judged by molecular weight and isoelectric point. Several proteins present in all membrane fractions (61K 5.1 and 58K 5.1) bound to concanavalin A agarose.

Stimulation by tubulin of an adenylate cyclase from murine plasmacytoma

Different brain tubulin preparations were shown to stimulate membrane-bound adenylate cyclase from the murine plasmacytoma MOPC 173. Purified tubulin devoid of microtubule-associated proteins and of nucleoside-diphosphate kinase activity was responsible for this stimulation. Activation of the basal adenylate cyclase activity occurred in less than 2 min at 32 degrees C and was amplified by a 4 degree C preincubation of tubulin with plasma membranes. Tubulin affected the Km and the V of the enzyme and was shown to be associated with the membrane during the activation phenomenon. Tubulin was more active on the basal adenylate cyclase activity than that stimulated by fluoride or guanosine 5'-[beta, gamma-imido]triphosphate. GTP has no effect on the tubulin-stimulated enzyme.