Peroxidative crosslinking of myosins.

The effect of myoglobin, free hemin and H2O2 on myosins from heart and skeletal muscle was studied. SDS-gel electrophoresis revealed that each agent caused intermolecular thiol crosslinking of both myosins dissociable by excess of beta-mercaptoethanol. In the simultaneous presence of H2O2 and myoglobin or H2O2 and free hemin, myosin formed covalent aggregates undissociable by beta-mercaptoethanol and therefore assessed to formation of non S-S inter molecular covalent bonds. The latter aggregates are suggested to result from pairing of myosin radicals formed by the H2O2 induced ferryl iron state in myoglobin, free hemin or hemo-myosin.     

Detection of disulfide bonds in bovine brain tubulin and their role in protein folding and microtubule assembly in vitro: a novel disulfide detection approach

Cysteine residues in tubulin are actively involved in regulating ligand interactions and microtubule formation both in vivo and in vitro. These cysteine residues are sensitive reporters in determining the conformation of tubulin. Although some of the cysteines are critical in modulating drug binding and microtubule assembly, it is not clear how many of these normally exist as disulfides. The controversy regarding the disulfide bonds led us to develop a disulfide detection assay to reexamine the presence of the disulfide linkages in purified alphabeta tubulin and explore their possible biological functions in vitro. The accessible cysteine residues in alphabeta tubulin were alkylated with an excess of iodoacetamide to prevent artifactual generation of disulfide linkages in tubulin. After removal of excess iodoacetamide, tubulin was unfolded in 8 M urea. Half of the unfolded tubulin was treated with dithiothreitol to reduce any disulfide bonds present. The aliquots were then treated with iodo[(14)C]acetamide and the incorporation of radioactivity was measured. We also used the same approach to detect the disulfide linkages in the tubulin in a whole-cell extract. We found in both cases that the samples which were not treated with dithiothreitol had little or no incorporation of iodo[(14)C]acetamide, while the others that were treated with dithiothreitol had significant amounts of (14)C incorporation into tubulin. Moreover, the reduction of the disulfide linkages in tubulin resulted in inhibition of microtubule assembly (29-54%) and markedly affected refolding of the tubulin from both an intermediate and a completely unfolded state. All these data therefore suggest that tubulin has intrachain disulfide bonds in the alpha- and beta-subunits and that these disulfides assist in correct refolding of tubulin from the intermediate unfolded state or help to recover the hydrophobic domains from the completely unfolded state. These disulfides also regulate microtubule assembly and the stability of tubulin in vitro. Our results suggest that tubulin disulfides may play a role in tubulin folding and that thiol-disulfide exchange in tubulin could be a key regulator in microtubule assembly and dynamics of tubulin in vivo.     

Structural investigation of radiation-induced aggregates of ribonuclease

Following irradiation of bovine pancreatic ribonuclease in aqueous solution with 60Co gamma-rays protein aggregates are formed. The nature of the bonds linking these radiation-induced aggregates together has been investigated by chromatographic and electrophoretic methods. Thin-layer gel filtration and polyacrylamide gel electrophoresis, both in the presence of sodium dodecyl sulphate, demonstrated the existence of covalent crosslinks between the aggregates. However, non-covalent crosslinking also plays a role in the radiolysis of ribonuclease. Thin-layer gel filtration with and without 6 M urea and 2 per cent beta-mercaptoethanol added to the gel, revealed that only part of the covalent bonds between the aggregates consisted of disulphide linkages. By separation of the reduced aggregates by thin-layer gel filtration and electrophoresis, both with SDS, this finding was substantiated. Densitometric measurements indicated for example that the percentage of covalently linked dimers held together by disulphide bridges amounted to about 40-45 per cent, whereas the remaining 55-60 per cent of the dimers must be linked by other covalent bonds. The existence of covalent crosslinks other than disulphide bonds was also confirmed by isoelectric focusing. By this method definite differences were established between the proteolytic hydrolysates of the reduced aggregates and the reduced monomer of gamma-irradiated ribonuclease.     

Xanthine oxidase-catalyzed crosslinking of cell membrane proteins

Isolated erythrocyte membranes exposed to protease-free xanthine oxidase plus xanthine and ferric iron undergo lipid peroxidation and protein crosslinking (appearance of high molecular weight aggregates on sodium dodecyl sulfate (SDS) gel electrophoresis). Spectrin is more susceptible to crosslinking than the other polypeptides. Thiol-reducible bonds (disulfides) as well as nonreducible bonds are generated, the former type relatively rapidly (detected within 10-20 min) and the latter type more slowly (usually detected after 1 h). Reducible crosslinking is inhibited by catalase, but not by superoxide dismutase, desferrioxamine, butylated hydroxyltoluene, and mannitol; whereas nonreducible crosslinking, like free radical lipid peroxidation, is inhibited by all of these agents except mannitol. Zinc(II) also inhibits lipid peroxidation, but stimulates disulfide bond formation to the virtual exclusion of all other crosslinking. Our results indicate that disulfide formation is dependent on H2O2, but not O2- or iron. However, O2-, H2O2, and iron are all required for lipid peroxidation and nondisulfide crosslinking, suggesting the intermediacy of OH generated via the iron-catalyzed Haber-Weiss reaction. The possible role of malonaldehyde (MDA, a by-product of lipid peroxidation) in the latter type of crosslinking was examined. Solubilized samples of xanthine/xanthine oxidase-treated membranes showed a strong visible fluorescence (emission maximum 450 nm; excitation 390 nm). This resembled the fluorescence of membranes treated with authentic MDA, which forms conjugated imine linkages between amino groups. Fluorescence scanning of SDS gels from MDA-treated membranes showed a strong signal coincident with crosslinked proteins and also one in the low molecular weight, nonprotein region, suggestive of aminolipid conjugates. Similar scanning on xanthine/xanthine oxidase-reacted membranes indicated that all fluorescence is associated with the lipid fraction. Thus, nonreducible protein crosslinks in this system do not appear to be of the MDA-derived, Schiff base type.     

Characterization of a macromolecular aggregate of ovine pituitary corticotropin-beta-lipotropin common precursor

A macromolecular aggregate of corticotropin-beta-lipotropin common precursor had been observed in ovine pituitary preparations as an excluded fraction of Sephadex G-200 gel filtration. This fraction could not penetrate a 10% gel during sodium dodecyl sulfate-polyacrylamide gel electrophoresis, when 2-mercaptoethanol or other disulfide-cleaving agents were not present in the buffer used to solubilize the protein preparation prior to the electrophoresis. On a 4.6% gel (acrylamide:bisacrylamide, 20:1), the material migrated as a diffuse band to a position between those of beta-galactosidase (Mr 130 000) and myosin (Mr 200 000). Both observations were consistent with an apparent Mr greatly in excess of that of the corticotropin-beta-lipotropin common precursor reported by many investigators. Neither 5% SDS nor 1% Triton X-100 could dissociate the macromolecular aggregate, but 2-mercaptoethanol and urea, either alone or in combination, were able to dissociate it to two main protein components, one of which was identified as corticotropin-beta-lipotropin with an apparent Mr of 34 000. The fact that urea alone could dissociate this macromolecular aggregate led us to believe that it might be a non-covalent aggregate and that 2-mercaptoethanol probably did not achieve the dissociation through the cleavage of an interchain disulfide bond but by bringing about conformational changes as a result of reduction of intrachain disulfide bonds so that aggregation became unfavorable. Moreover, the dissociation by urea or by 2-mercaptoethanol was found to be irreversible. The origin of the macromolecular aggregate of corticotropin-beta-lipotropin common precursor remains obscure.     

Specific photoaffinity crosslinking of [125I]cholecystokinin to pancreatic plasma membranes. Evidence for a disulfide-linked Mr 76 000 peptide in cholecystokinin receptors

In rat pancreatic plasma membranes, preincubated with [125I]cholecystokinin-33 (CCK-33) and washed free of unbound tracer, the irradiation by UV light induced the irreversible binding of radioactivity to high molecular weight peptides as shown by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate (SDS) and autoradiography. This was not observed when the membranes were preincubated in the simultaneous presence of [125I]CCK-33 and of either an excess of unlabelled CCK-8 or of guanosine 5'-(beta, gamma-imido)-triphosphate. The radioactivity was mostly crosslinked with a Mr 96,000 peptide and peptide species of Mr greater than 200,000, after SDS solubilization in the absence of beta-mercaptoethanol. Peptide reduction with beta-mercaptoethanol converted the high molecular weight radioactive species into a Mr 76,000 peptide that contained as much as 65% of the radioactivity crosslinked. The Mr 76,000 peptide appears, therefore, to be a disulfide-linked constituent of rat pancreatic cholecystokinin receptors.     

Characterization of oligomers of tubulin by two-dimensional native electrophoresis

We and others [Lee et al. (1973) J. Biol. Chem. 248, 7253-7262; Kravit et al. (1982) J. Cell Biol. 95, 344a; Kravit et al. (1984) J. Cell Biol. 99, 188-198] have observed oligomers of tubulin by native polyacrylamide gel electrophoresis (PAGE), even when they were not evident in sedimentation velocity or gel-exclusion chromatography experiments under comparable conditions. Aggregates of tubulin are also seen on native starch gels. Tubulins purified from calf brain, sea urchin egg (Strongylocentrotus purpuratus), and antarctic fish brain (Pagothenia borchgrevinki) give rise to similar distributions of aggregates. Unlike microtubules, these oligomers are relatively insensitive to temperature (5-25 degrees C), pH (6.1-8.8), the absence of excess GTP and/or Mg+2, stoichiometric concentrations of colchicine, and a variety of electrophoresis buffers. These aggregates, once formed during electrophoresis, associate and dissociate slowly. Depending upon the incubation conditions, they give rise to kinetically controlled distributions that appear in two-dimensional native PAGE as a square array of discrete polymeric species. The fastest migrating species (monomers) are often observed to reequilibrate preferentially into the second band. The second band reequilibrates into the fourth, the third band into the sixth, the fourth into the eighth, etc. (The assignment of molecular weights to these species by Ferguson analysis is tentative due to their slow reequilibration.) Thus, a feature of the reequilibration is that association occurs more rapidly than dissociation and each species is occasionally observed to "dimerize." This behavior is suggestive of irreversible aggregation (possibly crosslinking) or of the formation of slowly dissociating aggregates. Although they may be related to the protofilaments of microtubules, these oligomers appear to be another example of nonmicrotubular, polymorphic aggregates of tubulin.     

Analysis of tubulin isoforms by two-dimensional gel electrophoresis using SDS-polyacrylamide gel electrophoresis in the first dimension.

A two-dimensional electrophoretic system using SDS-polyacrylamide gel electrophoresis (SDS-PAGE) in the first dimension and isoelectric focusing (IEF) in the second dimension was devised. In spite of its simplicity, this method could show a markedly high resolution for tubulin isoforms and moreover could classify them into alpha- or beta-tubulin as a two-dimensional profile. With this method, seven alpha- and four beta-tubulin isoforms could be detected within axoneme from Tetrahymena cilia. Moreover this method could also resolve tubulin isoforms from the rabbit brain. These results indicate that the present two-dimensional gel electrophoresis is a useful tool for the electrophoretic analysis of tubulin isoforms from various sources.     

Inhibition of plant cell proteolytic activities that degrade tubulin

The requirement for proteinase inhibitors during the chromatographic isolation of tubulin from cultured cells of rose (Rosa sp. cv. Paul's scarlet) was examined by NadodecylSO4-polyacrylamide gel electrophoresis, electron microscopy and immunoblotting. Tubulin fractions isolated in the absence of proteinase inhibitors showed substoichiometric ratios of alpha-subunit to beta-subunit, and low molecular weight polypeptides, one (approximately 32 Kd) of which coassembled with polymers. Electron microscopy revealed polymorphic structures, including C- and S-shaped ribbons and free protofilaments. Immunoblotting experiments with IgGs to the individual alpha- and beta-subunits showed that some of the low molecular weight polypeptides were fragments of proteolytically degraded subunits. The use of low micromolar concentrations of the synthetic proteinase inhibitors leupeptin hemisulfate and pepstatin A protected tubulin from endogenous proteolytic activities during the isolation procedure and resulted in increased tubulin purity.     

A membrane-cytoskeletal complex containing Na+,K+-ATPase, ankyrin, and fodrin in Madin-Darby canine kidney (MDCK) cells: implications for the biogenesis of epithelial cell polarity

In polarized Madin-Darby canine kidney (MDCK) epithelial cells, ankyrin, and the alpha- and beta-subunits of fodrin are components of the basolateral membrane-cytoskeleton and are colocalized with the Na+,K+-ATPase, a marker protein of the basolateral plasma membrane. Recently, we showed with purified proteins that the Na+,K+-ATPase is competent to bind ankyrin with high affinity and specificity (Nelson, W. J., and P. J. Veshnock. 1987. Nature (Lond.). 328:533-536). In the present study we have sought biochemical evidence for interactions between these proteins in MDCK cells. Proteins were solubilized from MDCK cells with an isotonic buffer containing Triton X-100 and fractionated rapidly in sucrose density gradients. Complexes of cosedimenting proteins were detected by analysis of sucrose gradient fractions in nondenaturing polyacrylamide gels. The results showed that ankyrin and fodrin cosedimented in sucrose gradient. Analysis of the proteins from the sucrose gradient in nondenaturing polyacrylamide gels revealed two distinct ankyrin:fodrin complexes that differed in their relative electrophoretic mobilities; both complexes had electrophoretic mobilities slower than that of purified spectrin heterotetramers. Parallel analysis of the distribution of solubilized Na+,K+-ATPase in sucrose gradients showed that there was a significant overlap with the distribution of ankyrin and fodrin. Analysis by nondenaturing polyacrylamide gel electrophoresis showed that the alpha- and beta-subunits of the Na+,K+-ATPase colocalized with the slower migrating of the two ankyrin:fodrin complexes. The faster migrating ankyrin:fodrin complex did not contain Na+,K+-ATPase. These results indicate strongly that the Na+,K+-ATPase, ankyrin, and fodrin are coextracted from whole MDCK cells as a protein complex. We suggest that the solubilized complex containing these proteins reflects the interaction of the Na+,K+-ATPase, ankyrin, and fodrin in the cell. This interaction may play an important role in the spatial organization of the Na+,K+-ATPase to the basolateral plasma membrane in polarized epithelial cells.     

Isolation and characterization of the subunits of bovine follitropin.

Highly purified bovine follitropin was dissociated into its alpha- and beta-subunits after treatment with 1 M-propionic acid. The dissociated subunits were fractionated by chromatography on DEAE-cellulose and further purified by gel filtration on Sephadex G-100. The isolated alpha- and beta-subunits were biologically inactive, but their recombinants regenerated 80% of the follitropin activity. The alpha-subunit of bovine follitropin recombined with the beta-subunits of bovine lutropin and thyrotropin to regenerate 70% of lutropin and 50% of thyrotropin activities respectively. The beta-subunit of bovine follitropin recombined with the alpha-subunit of either bovine lutropin or thyrotropin to regenerate about 75% of follitropin activity. Recombinations were monitored by specific radioligand-receptor assays and polyacrylamide-gel electrophoresis. The elution volumes of the alpha- and beta-subunits of bovine follitropin after gel filtration on Sephadex G-100 were almost identical. The amino acid composition of bovine follitropin-alpha was low in histidine, arginine, isoleucine and leucine, but relatively high in lysine, threonine and glutamic acid. The bovine follitropin-beta contained one methionine residue and low amounts of histidine and phenylalanine, but relatively high in aspartic acid, threonine and glutamic acid. The N-terminal residues of the alpha- and beta-subunits of bovine follitropin were identified to be phenylalanine and glycine respectively.     

Assembly of the sea urchin extraembryonic hyaline layer; Ca2+ and Mg2+ act independently and at different sites on the pathway leading to hyalin-gel formation

We have studied the interactions of Ca2+ with the sea urchin extraembryonic coat protein hyalin. As reported previously, Ca2+ alone was ineffective in inducing hyalin-gel (large aggregate) formation. This reaction required the additional presence of Mg2+ and NaCl. However, the results of tryptic digestion and nondenaturing agarose gel electrophoresis experiments demonstrated that Ca2+ could induce hyalin self-association into small aggregates in the absence of Mg2+ and NaCl. Magnesium did not modulate the interactions of Ca2+ with hyalin. In addition, Mg2+ had minimal effects on the conformation of hyalin. These results have been incorporated into a model delineating the pathway leading to hyalin-gel formation.     

Conformational states of the insulin receptor

Insulin binding to the alpha-subunit of the purified insulin receptor changed the interaction between beta-subunits. This conformational change was demonstrated after labeling the receptor's beta-subunit by autophosphorylation in the absence of insulin, and then crosslinking the subunits to each other with bis (sulfosuccinimidyl) suberate. The convalent oligomers were resolved by reduction and denaturing gel electrophoresis. Insulin increased the rate of crosslinking, especially the formation of beta-beta dimers. These results support a conformational change following insulin binding, and may reflect the insulin-induced activation of autophosphorylation.     

Reducer driven baric denaturation and oligomerisation of whey proteins

The influence of high pressure on alpha-lactalbumin (ALA)/beta-lactoglobulin (BLG) mixtures of various compositions was studied at pH 8.5 by gel-permeation chromatography and sodium dodecyl sulphate (SDS) gel electrophoresis without 2-mercaptoethanol. High-molecular protein disulfide oligomers formed after denaturation by the pressure of 10 kbar (1000 MPa) if the weight fraction of BLG (W(BLG)(0)) in the protein mixture exceeded 0.2. The maximum yield of these oligomers of order 80-85% is observed at W(BLG)(0)>/=0.4. Conversions of both proteins in the oligomers are roughly the same. The estimates of the oligomerisation yield obtained by the gel-permeation chromatography and SDS gel electrophoresis agree well. This indicates that the formation of intermolecular disulfide bonds is necessary for the oligomerisation. Thus, the oligomerisation of pressure denatured ALA and BLG is driven by the thiol<-->disulfide exchange rendered possible by the vigorous baric denaturation and the exposure of the free thiol group of BLG, which acts as an initiator of disulfide bridges scrambling.     

Studies on factor VIII-related protein. I. Ultrastructural and electrophoretic heterogeneity of human factor VIII-related protein.

Human factor VIII-related protein precipitates with specific heterologous anti-bodies directed against purified factor VIII and supports ristocetin-induced aggregation of washed platelets. We purified human factor VIII from cryoprecipitate by subsequent gel filtration on crosslinked large-pore agarose. Factor VIII-related protein appeared as a large aggregate following electrophoresis on 3% polyacrylamide gels in the presence of sodium dodecyl sulfate (SDS). The same material was separated into multiple bands (molecular weight in excess of several millions) following electrophoresis on SDS-1% agarose gels. After complete disulfide reduction of factor VIII-related protein and electrophoresis on SDS-5% polyacrylamide gels a single subunit chain (Mr approximately equal to 200 000) was revealed. Analysis of this protein, in its non-reduced state, by negative contrast electron microscopy showed filaments of markedly variable size. The calculated molecular weight of such filaments ranged from about 0.6.10(6) to 20.10(6). We conclude that size heterogeneity is an essential feature of human factor VIII-related protein.     

Disulfide bond formation in the regulation of eIF-2 alpha kinase by heme

The inhibition of the autophosphorylation of the heme-regulated eukaryotic initiation factor (eIF)-2 alpha kinase (HRI) by hemin is very similar to that produced by thiol oxidation by diamide. The results obtained from the analysis of sodium dodecyl sulfate-polyacrylamide gel electrophoresis of unphosphorylated and phosphorylated HRI under reducing and nonreducing conditions indicate that hemin promotes disulfide formation in HRI. Hemin-promoted disulfide formation in HRI occurs under quasi-physiological conditions, i.e. 30 degrees C, 10 min at hemin concentrations of 5-10 microM. Under nondenaturing conditions, unphosphorylated HRI, phosphorylated HRI, hemin-treated unphosphorylated HRI, and hemin-treated prephosphorylated HRI are all eluted identically on Sephacryl S-300 column chromatography with an apparent molecular mass of 290,000 daltons. It appears, therefore, that the disulfide formation promoted by hemin occurs within the unit of 290,000 daltons. In addition, hemin treatment of phosphorylated HRI results in the appearance of a disulfide-linked form of higher molecular mass when analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis under nonreducing conditions. A similar high molecular mass form is observed when HRI is treated with 1,6-bismaleimidohexane, a double sulfhydryl cross-linker agent, and the autophosphorylation of HRI and the phosphorylation of eIF-2 alpha by HRI are greatly diminished; these effects are similar to the effects of hemin on HRI. We conclude that disulfide formation by hemin provides a likely mechanism by which hemin prevents the activation and inhibits the activity of HRI.     

Carbonylation of ER chaperone proteins in aged mouse liver

Progressive accumulation of oxidative damage to macromolecules in aged tissues is thought to contribute to the decline in tissue function characteristic of the aged phenotype. Mitochondria are a major intracellular source of reactive oxygen species (ROS); however, other organelles are also endogenous sources of oxyradicals and oxidants, which can damage macromolecules. We, therefore, sought to examine the relationship between aging and oxidative damage to ER resident proteins, which exist in a strongly oxidizing environment necessary for disulfide bond formation. In these studies, we have fractionated young and aged liver homogenates, resolved the proteins by 2D gel electrophoresis, assayed for oxidative damage as indicated by protein carbonylation, and identified BiP/Grp78, protein disulfide isomerase (PDI), and calreticulin as exhibiting an age-associated increase in oxidative damage. Increased carbonylation of these key proteins in aged liver suggests an age-associated impairment in protein folding, disulfide crosslinking, and glycosylation in the aged mouse liver.     

Electron spin resonance study of the role of nitrosyl-heme in the activation of guanylate cyclase by nitrosoguanidine and related agonists.

One major component of lens plasma membrane is a glycoprotein that SDS-polyacrylamide gel electrophoresis shows to possess an apparent molecular weight of 26,000. When this protein is solubilized in low ionic strength buffers containing SDS, and heated to 100° for 1 to 3 min prior to electrophoresis, conversion into high molecular weight aggregate results. The heat lability of this protein is greatly enhanced if it solubilized and heated in buffers containing 0.1 M NaCl. At this ionic strength, incubation for 3 h at 38° results in conversion of 20% of the protein into high melecular weight aggregates. Most other membrane proteins isolated from lens membrane are insensitive to heat treatment. It is concluded that temperature and ionic strength must be recorded and controlled carefully when using SDS-polyacrylamide gel electrophoresis to study this membrane protein.     

Comparative analysis of protein aggregates by blue native electrophoresis and subsequent sodium dodecyl sulfate-polyacrylamide gel electrophoresis in a three-dimensional geometry gel

We describe the comparative analysis of protein aggregates by combining blue native electrophoresis and subsequent sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using a 3-D geometry gel for simultaneous processing of many samples. The first native electrophoresis step, separating the aggregates, is carried out for a series of samples in parallel lanes within a slab gel. This gel is then placed on the top surface of a cylindrical, 3-D geometry gel for the second denaturing electrophoresis step, separating the proteins composing the aggregates. The samples migrate parallel to the vertical axis of the gel cylinder. Data are acquired online by photodetection of laser-induced fluorescence during electrophoresis. For this purpose, the samples are fluorescently labeled within the slab gel after the first separation step. A 3-D geometry gel separates the equivalent of many conventional SDS slab gels represented by vertical layers in the 3-D gel body. In this way, many samples are analyzed in the same gel under identical conditions, improving comparability and resolution and making the process considerably more efficient. This novel technique allowed the identification of several aggregate classes of recombinant proteins expressed in bacteria. We observed that proteins preferentially bind to homolog polypeptides, but also seem to form a trapping mesh co-aggregating with other proteins. The aggregation pattern revealed by this technique supplements data obtained from standard two-dimensional gel electrophoresis analysis. We expect interesting applications, for instance in aggregate monitoring of clinical samples. It should be feasible to quickly gain a diagnostic picture during amyloid-related neurodegenerative disease development or to observe drug effects on protein aggregation.     

Interaction of vinblastine with calf brain tubulin: effects of magnesium ions

The effects of magnesium ions on the binding of the anticancer drug vinblastine to calf brain tubulin were investigated by a batch gel equilibration method. Magnesium ions at 1 mM strongly enhanced the binding of the first vinblastine molecule to each tubulin dimer without affecting either the drug affinity toward the rest of the binding site or the total stoichiometry of the vinblastine binding to tubulin. Sedimentation velocity studies indicated that magnesium ions can enhance strongly the vinblastine-induced tubulin self-association and suggested that the drug-induced self-association still proceeds through the isodesmic indefinite mechanism in the presence of the divalent cation. In PG buffer (0.01 M NaPi, 10(-4) M GTP, pH 7.0) containing more than 2.5 mM MgCl2, vinblastine induced tubulin to form large amorphous aggregates. The aggregate formation was rapid and took place at a drug stoichiometry between 0.7 and 1.0 mol of vinblastine per mole of tubulin dimers. Increasing the solution ionic strength decreased the rate of aggregate formation. Between an ionic strength of 0.05 and 0.1, the self-association led to the formation of paracrystalline aggregates instead of the amorphous ones. The results indicated that the binding of only the first vinblastine molecule to each tubulin dimer is linked to the self-association of the protein. They also confirmed our previously proposed rationale for the disagreement among the vinblastine-tubulin binding constants reported in the literature in terms of the different magnesium ion concentrations and ionic strength of the buffers used in the various studies.