Biphasic kinetics of the colchicine-tubulin interaction: role of amino acids surrounding the A ring of bound colchicine molecule

Isotypes of vertebrate tubulin have variable amino acid sequences, which are clustered at their C-terminal ends. Isotypes bind colchicine at different on-rates and affinity constants. The kinetics of colchicine binding to purified (unfractionated) brain tubulin have been reported to be biphasic under pseudo-first-order conditions. Experiments with individual isotypes established that the presence of beta(III) in the purified tubulin is responsible for the biphasic kinetics. Because the isotypes mainly differ at the C termini, the colchicine-binding kinetics of unfractionated tubulin and the beta(III) isotype, cleaved at the C termini, have been tested under pseudo-first-order conditions. Removal of the C termini made no difference to the nature of the kinetics. Sequence alignment of different beta isotypes of tubulin showed that besides the C-terminal region, there are differences in the main body as well. To establish whether these differences lie at the colchicine-binding site or not, homology modeling of all beta-tubulin isotypes was done. We found that the isotypes differed from each other in the amino acids located near the A ring of colchicine at the colchicine-binding site on beta tubulin. While the beta(III) isotype has two hydrophilic residues (serine(242) and threonine(317)), both beta(II) and beta(IV) have two hydrophobic residues (leucine(242) and alanine(317)). beta(II) has isoleucine at position 318, while beta(III) and beta(IV) have valine at that position. Thus, these alterations in the nature of the amino acids surrounding the colchicine site could be responsible for the different colchicine-binding kinetics of the different isotypes of tubulin.     

Kinetics of colchicine binding to purified beta-tubulin isotypes from bovine brain.

Tubulin, the constituent protein of microtubules, is an alpha beta heterodimer; both alpha and beta exist in several isotypic forms whose functional significance is not precisely known. The antimitotic alkaloid colchicine binds to mammalian brain tubulin in a biphasic manner under pseudo-first-order conditions in the presence of a large excess of colchicine (Garland, D. L. (1978) Biochemistry 17, 4266-4272). We have studied the kinetics of colchicine binding to purified beta-tubulin isotypes and find that each of the purified beta-tubulin isotypes binds colchicine in a monophasic manner. The apparent on-rate constants for the binding of colchicine to alpha beta II-, alpha beta III-, and alpha beta IV-tubulin dimers are respectively 132 +/- 5, 30 +/- 2, and 236 +/- 7 M-1 s-1. When the isotypes are mixed, the kinetics become biphasic. Scatchard analysis revealed that the isotypes differ significantly in their affinity constants (Ka) for binding colchicine. The affinity constants are 0.24 x 10(6), 0.12 x 10(6), and 3.31 x 10(6) M-1, respectively, for alpha beta II-, alpha beta III-, and alpha beta IV-tubulin dimers. Our results are in agreement with the hypothesis that the beta-subunit of tubulin plays a major role in the interaction of colchicine with tubulin. Our binding data raise the possibility that the tubulin isotypes might play important regulatory roles by interacting differently with other non-tubulin proteins in vivo, which in turn, may regulate microtubule-based functions in living cells.     

Increased microtubule assembly in bovine brain tubulin lacking the type III isotype of beta-tubulin

Tubulin, the major constituent protein of microtubules, is a heterodimer of alpha and beta subunits. Both alpha and beta exist in multiple isotypic forms. It is not clear if different isotypes perform different functions. In order to approach this question, we have made a monoclonal antibody specific for the beta III isotype of tubulin. This particular isotype is neuron-specific and appears to be phosphorylated near the C terminus. We have used immunoaffinity depletion chromatography to prepare tubulin lacking the beta III subunit. We find that removal of the beta III isotype results in a tubulin mixture able to assemble much more rapidly than is unfractionated tubulin when reconstituted with either of the two microtubule-associated proteins (MAPs), tau or MAP 2. Our results suggest that the different isotypes of tubulin differ from each other in their ability to polymerize into microtubules. We have also found that the anti-beta III antibody can stimulate microtubule assembly when reconstituted with tubulin and either tau or MAP 2. When reconstituted with tubulin lacking the beta III isotype, the antibody causes the tubulin to polymerize into a polymer that is a microtubule in the presence of MAP 2 and a ribbon in the presence of tau.     

A monoclonal antibody against the type II isotype of beta-tubulin. Preparation of isotypically altered tubulin

Mammalian brain tubulin consists of several isotypes of alpha and beta subunits that separate on polyacrylamide gels into three electrophoretic classes, designated alpha, beta 1, and beta 2. It has not been possible hitherto to resolve the different isotypes in a functional form. To this end, we have now isolated a monoclonal antibody, using as an immunogen a chemically synthesized peptide corresponding to the carboxyl-terminal sequence of the major tubulin isotype (type II) found in the beta 1-tubulin electrophoretic fraction. The antibody binds to beta 1 but not to alpha or beta 2. When pure tubulin from bovine brain is passed through an immunoaffinity column made from the anti-type II antibody, the tubulin that elutes in the unbound fraction is enriched greatly for the beta 2 electrophoretic variant. The tubulin that binds to the column appears to contain only alpha and beta 1, not beta 2. When these tubulin fractions are characterized by immunoblotting using the anti-type II antibody, the antibody binds only to the beta 1 band in the bound fraction, not to the beta 1 band in the unbound fraction. Using polyclonal antibodies generated against the carboxyl-termini of types I, III, and IV, we demonstrate that the beta 1 electrophoretic species is comprised of isotypes I, II, and IV, whereas the beta 2 variant is comprised exclusively of type III beta-tubulin. Further, we calculate that beta-tubulin in purified bovine brain tubulin is comprised of 3% type I, 58% type II, 25% type III, and 13% type IV tubulins.     

Kinetic studies on partially liganded species of carboxyhemoglobin: (alpha 1CO-beta 1CO)alpha 2 beta 2 or (alpha 2CO beta 2CO)alpha 1 beta 1

Kinetics of CO combination with and dissociation from isomer III, (alpha 1CO beta 1CO)alpha 2 beta 2 or alpha 1 beta 1 (alpha 2CO beta 2CO), and Hb Rothschild have been studied using the double mixing and microperoxidase methods. Isomer III was prepared in a manner so that it was the only reactive species in the reaction mixture. The biphasic reaction time course in both the "on" and "off" reactions of isomer III and the CO combination reaction of Hb Rothschild are attributed to slow relaxation between the fast and slow CO-reacting species in the two proteins: isomer III: l'f = 6 x 10(6) M-1 s-1, l'dimer = 1.7 x 10(6) M-1 s-1, l's = 2.2 x 10(5) M-1 s-1, lf = 0.15 s-1, ls = 0.01 s-1; Hb Rothschild: l'f = 2.8 x 10(6) M-1 s-1; l's = 2.7 x 10(5) M-1 s-1.     

The roles of cys124 and ser239 in the functional properties of human betaIII tubulin

Tubulin is the target for some very powerful anti-mitotic and anti-tumor drugs. The betaIII tubulin isotype is found in very few normal tissues, but is often found in tumors, where it has been implicated in resistance to anti-tumor drugs. The betaIII isotype occurs in fish, amphibians, birds and mammals and its unique features are highly conserved in evolution. One of these features is the replacement of cys239 by ser239. Cys239 is unusual in being highly sensitive to oxidation; in fact, oxidation of this residue inhibits microtubule assembly. The betaIII isotype also has a very unusual cys124, where other beta isotypes have ser/ala124. The striking conservation in betaIII of vertebrates strongly suggests that cys124 and ser239 play functional roles. We have prepared the C124S and S239C mutants of betaIII and tested their effects on the functional properties of tubulin. We have found that both the betaIII C124S and betaIII S239C mutants bind colchicine less well than does wild-type alphabetaIII, and also make transfected HeLa cells more resistant to colchicine. However, the double mutant, betaIII C124S/S239C, binds colchicine still less well than do either of the single mutants, but in contrast to the former, the double mutant increases the cells' sensitivity to colchicine. Our results indicate that the roles that these residues play in colchicine binding and microtubule integrity are far more complex than previously imagined and that the specific residues at which betaIII differs from the other isotypes act collectively to keep betaIII in a functional conformation.     

Association of thiocolchicine with tubulin

Thiocolchicine, a colchicine analog in which the C-10 methoxy is replaced with a thiomethyl moiety, was shown to bind with high affinity to the colchicine site on tubulin (Ka = 1.07 +/- 0.14 x 10(6) M-1 at 23 degrees C). Like colchicine, the association kinetics were biphasic, and the rate constants of both phases were temperature dependent. The rate constant of the fast phase of the association was 4 times greater than the rate constant for colchicine binding, and the activation energy was lower (19.1 +/- 1.8 kcal/mol). X-ray crystallographic analysis shows that thiocolchicine displays greater puckering of the tropone C ring than colchicine (Koerntgen, C. and Margulis, T. N. (1977) J. Pharm. Sci. 66, 1127-1131.). These results indicate that the conformation of the C ring may have little effect on the energetics of colchicinoids binding to tubulin.     

Identification by mass spectrometry of a new alpha-tubulin isotype expressed in human breast and lung carcinoma cell lines

The extensive C-terminal molecular heterogeneity of alpha- and beta-tubulin is a consequence of multiple isotypes, the products of distinct genes, that undergo several posttranslational modifications. These include polyglutamylation and polyglycylation of both subunits, reversible tyrosination and removal of the penultimate glutamate from alpha-tubulin, and phosphorylation of the beta III isotype. A mass spectrometry-based method has been developed for the analysis of the C-terminal diversity of tubulin from human cell lines. Total cell extracts are resolved by SDS--PAGE and transferred to nitrocellulose, and the region of the blot corresponding to tubulin (approximately 50 kDa) was excised and digested with CNBr to release the highly divergent C-terminal tubulin fragments. The masses of the human alpha- and beta-tubulin CNBr-derived C-terminal peptides are all in the 1500--4000 Da mass range and can be analyzed directly by MALDI-TOF mass spectrometry in the negative ion mode without significant interference from other released peptides. In this study, the tubulin isotype diversity in MDA-MB-231, a human breast carcinoma cell line, and A549, a human non-small lung cancer cell line, is reported. The major tubulin isotypes present in both cell lines are k-alpha 1 and beta 1. Importantly, we report a previously unknown alpha isotype present at significant levels in both cell lines. Moreover, the degree of posttranslational modifications to all isotypes was limited. Glu-tubulin, in which the C-terminal tyrosine of alpha-tubulin is removed, was not detected. In contrast to mammalian neuronal tubulin which exhibits extensive polyglutamylation, only low-level monoglutamylation of the k-alpha 1 and beta 1 isotypes was observed in these two human cell lines.     

Differential sorting of beta tubulin isotypes into colchicine-stable microtubules during neuronal and muscle differentiation of embryonal carcinoma cells.

Pluripotent P19 embryonal carcinoma (EC) cells were differentiated along the neuronal and muscle pathways. Comparisons of class I, II, III, and IV beta tubulin isotypes in total and colchicine-stable microtubule (MT) arrays from uncommitted EC, neuronal, and muscle cells were made by immunoblotting and by indirect immunofluorescence microscopy. In undifferentiated EC cells the relative amounts of these four isotypes are the same in both the total and stable MT populations. Subcellular sorting of beta tubulin isotypes was demonstrated in both neuronal and muscle differentiated cells. During neuronal differentiation, class II beta tubulin is preferentially incorporated into the colchicine-stable MTs while class III beta tubulin is preferentially found in the colchicine-labile MTs. The subcellular sorting of class II into stable MTs correlates with the increased staining of MAP 1B, and with the expression of MAP 2C and tau. Although muscle differentiated cells express class II beta tubulin, stable MTs in these cells do not preferentially incorporate this isotype but instead show increased incorporation of class IV beta tubulin. Muscle cells do not show high levels of MAP 1B and do not express MAP 2C or tau. These results are consistent with the hypothesis that a subcellular sorting of tubulin isotypes is the result of a complex interaction between tubulin isotypes and MT-associated proteins.     

Binding to tubulin of the colchicine analog 2-methoxy-5-(2', 3', 4'-trimethoxyphenyl)tropone. Thermodynamic and kinetic aspects

The thermodynamics and kinetics of the binding to tubulin of the colchicine analog 2-methoxy-5-(2', 3', 4'-trimethoxyphenyl) tropone (termed AC because it lacks the B-ring of colchicine) have been characterized by fluorescence techniques. The fluorescence of AC is weak in aqueous solution and is enhanced 250-fold upon binding to tubulin. The following thermodynamic values were obtained for the interaction at 37 degrees C: K = 3.5 X 10(5) M-1; delta G0 = -7.9 kcal/mol; delta H0 = -6.8 kcal/mol; delta S0 = 3.6 entropy units. The AC-tubulin complex is 1-2 kcal/mol less stable than the colchicine-tubulin complex. The change in fluorescence of AC was employed to measure the kinetics of the association process, and quenching of protein fluorescence was used to measure both association and dissociation. The association process, like that of colchicine, could be resolved into a major fast phase and a minor slow phase. The apparent second order rate constant for the fast phase was found to be 5.2 X 10(4) M-1 S-1 at 37 degrees C, and the activation energy was 13 kcal/mol. This activation energy is 7-11 kcal/mol less than that for the binding of colchicine to tubulin. The difference in activation energies can most easily be rationalized by a mechanism involving a tubulin-induced conformational change in the ligand ( Detrich , H. W., III, Williams, R. C., Jr., Macdonald, T. L., Wilson, L., and Puett , D. (1981) Biochemistry 20, 5999-6005). Such a change would be expected to have a small activation energy in AC because it possesses a freely rotating single bond in place of the B-ring of colchicine.     

Class II tubulin, the major brain beta tubulin isotype is polyglutamylated on glutamic acid residue 435.

Protein sequencing shows that porcine brain tubulin retains the N-terminal sequences of alpha and beta tubulin after a mild treatment with subtilisin. C-terminal peptides released by subtilisin were purified and characterized by automated Edman degradation and mass spectrometry. We confirm the polyglutamylation of alpha tubulin on glutamic acid residue 445 reported by others and show in addition that class II beta tubulin, the major beta tubulin isotype of adult brain, is also polyglutamylated. The substitution is restricted to glutamic acid residue 435. Thus all major tubulin isotypes of adult brain are subjected to polyglutamylation.     

Comparison of beta-tubulin mRNA and protein levels in 12 human cancer cell lines

Antimitotic drugs are chemotherapeutic agents that bind tubulin and microtubules. Resistance to these drugs is a major clinical problem. One hypothesis is that the cellular composition of tubulin isotypes may predict the sensitivity of a tumor to antimitotics. Reliable and sensitive methods for measuring tubulin isotype levels in cells and tissues are needed to address this hypothesis. Quantitative measurements of tubulin isotypes have frequently relied upon inferring protein amounts from mRNA levels. To determine whether this approach is justified, protein and mRNA levels of beta-tubulin isotypes from 12 human cancer cell lines were measured. This work focused on only beta-tubulin isotypes because we had readily available monoclonal antibodies for quantitative immunoblots. The percentage of beta-tubulin isotype classes I, II, III, and IVa + IVb mRNA and protein were compared. For beta-tubulin class I that comprises >50% of the beta-tubulin protein in 10 of the 12 cell lines, there was good agreement between mRNA and protein percentages. Agreement between mRNA and protein was also found for beta-tubulin class III. For beta-tubulin classes IVa + IVb, we observed higher protein levels compared to mRNA levels.Beta-tubulin class II protein was found in only four cell lines and in very low abundance. We conclude that quantitative Western blotting is a reliable method for measuring tubulin isotype levels in human cancer cell lines. Inferring protein amounts from mRNA levels should be done with caution, since the correspondence is not one-to-one for all tubulin isotypes.     

The reaction of choline acetyltransferase with sulfhydryl reagents. Methoxycarbonyl-CoA disulfide as an active site-directed reagent.

The reaction of choline acetyltransferase with methoxycarbonyl alkyl disulfides leads to a progressive loss in enzyme activity as the size of the alkyl group increases from methyl to n-butyl. Reaction with 5,5'-dithiobis(2-nitrobenzoic acid) (DTNB) or methoxycarbonyl coenzyme A (CoA) disulfide, leads to a total loss of enzyme activity. DTNB inactivation is biphasic (k1 = approximately 9 x 10(2) M-1 s-1, k2 = approximately 6 x 10(1) M-1 s-1) with the slow phase being diminished by acetyl-CoA. Methoxycarbonyl-CoA disulfide inactivation is also biphasic (k1 = approximately 2.1 x 10(3) M-1 s-1, k2 = approximately 6 x 10(1) M-1 s-1), with the rapid phase being diminished in the presence of acetyl-CoA. Inactivation by methoxycarbonyl methyl disulfide, ethyl disulfide, or hydroxyethyl disulfide, or by methyl methanethiosulfonate is not biphasic. Pretreatment of the enzyme with methyl methanethiosulfonate, which leads to a 25% loss in enzyme activity, abolishes the fast phase of DTNB inactivation, the slow phase of methoxycarbonyl-CoA disulfide inactivation, and any further inactivation by methoxycarbonyl ethyl disulfide. These results are interpreted to suggest that choline acetyltransferase contains two classes of reactive sulfhydryl groups, neither of which are required for enzyme activity.     

During Drosophila spermatogenesis beta 1, beta 2 and beta 3 tubulin isotypes are cell-type specifically expressed but have the potential to coassemble into the axoneme of transgenic flies

alpha and beta Tubulins exist in a number of different isotypes with distinct expression patterns during development. We have shown by immunofluorescent staining that beta 1, beta 2 and beta 3 tubulins are distributed very specifically in the testes of Drosophila. beta 3 Tubulin is present exclusively in cytoplasmic microtubules of cells somatic in origin, while the beta 1 isotype is localized in the somatic cells and in early germ cells of both the microtubules of the cytoskeleton as well as in the mitotic spindle. In contrast, beta 2 tubulin is present in all microtubular arrays (cytoskeleton, meiotic spindles, axoneme) of germ cells from meiotic prophase onward, though not detectable in somatic cells. Thus, a switch of beta tubulin isotypes from beta 1 to beta 2 occurs during male germ cell differentiation. This switch is also observed in the distantly related species Drosophila hydei. By fusing beta 1 or beta 3 amino acid coding regions to the control region of the beta 2 tubulin gene and performing germ line transformation experiments, we have examined the copolymerization properties of the different tubulin isotypes. Neither beta 1 nor beta 3 are detectable in the axoneme in the wild-type situation. Analysis of transgenic flies carrying beta 2-beta 1 fusion genes or beta 2-beta 3 fusion genes revealed that both beta 1 and beta 3 tubulin isotypes have the potential to co-incorporate with beta 2 tubulin into microtubules of the sperm axoneme. Male flies homozygous for the fusion genes (beta 2-beta 1 or beta 2-beta 3) remain fertile, despite the mixture of beta tubulin isotypes in the axoneme.     

Intrinsically slow dynamic instability of HeLa cell microtubules in vitro

The dynamic behavior of mammalian microtubules has been extensively studied, both in living cells and with microtubules assembled from purified brain tubulin. To understand the intrinsic dynamic behavior of mammalian nonneural microtubules, we purified tubulin from cultured HeLa cells. We find that HeLa cell microtubules exhibit remarkably slow dynamic instability, spending most of their time in an attenuated state. The tempered dynamics contrast sharply with the dynamics of microtubules prepared from purified bovine brain tubulin under similar conditions. In accord with their minimal dynamic instability, assembled HeLa cell microtubules displayed a slow treadmilling rate and a low guanosine-5'-triphosphate hydrolysis rate at steady state. We find that unlike brain tubulin, which consists of a heterogeneous mixture of beta-tubulin isotypes (beta(II), beta(III), and beta(IV) and a low level of beta(I)), HeLa cell tubulin consists of beta(I) tubulin ( approximately 80%) and a minor amount of beta(IV) tubulin ( approximately 20%). The slow dynamic behavior of HeLa cell microtubules in vitro differs strikingly from the dynamic behavior of microtubules in living cultured mammalian cells, supporting the idea that accessory factors create the robust dynamics that occur in cells.     

Differential utilization of beta-tubulin isotypes in differentiating neurites

beta-Tubulin is encoded in vertebrate genomes by a family of six to seven functional genes that produce six different polypeptide isotypes. We now document that although rat PC-12 cells express five of these isotypes, only two (classes II and III) accumulate significantly as a consequence of nerve growth factor-stimulated neurite outgrowth. In contrast to previous efforts that have failed to detect in vivo distinctions among different beta-tubulin isotypes, we demonstrate using immunoblotting with isotype-specific antibodies that three beta-tubulin polypeptides (classes I, II, and IV) are used preferentially for assembly of neurite microtubules (with approximately 70% of types I and II assembled but only approximately 50% of type III in polymer). Immunofluorescence localization shows that an additional isotype (V) is partially excluded from neurites. Distinctions in in vivo localization of the neuron-specific, class III isotype have also been directly observed using immunofluorescence and immunogold electron microscopy. The sum of these efforts documents that some in vivo functional differences between tubulin isotypes do exist.     

Expression of cold-adapted beta-tubulins confer cold-tolerance to human cellular microtubules

Isolated microtubule proteins from the cold-adapted fish, Atlantic cod (Gadus morhua), assemble at temperatures between 8 and 30 degrees C, while avian and mammalian microtubules normally do not assemble at temperatures below 20 degrees C. Tubulin, the main component in microtubules, is expressed as many isotypes. Microtubules with different isotype composition have been shown to have different dynamic properties in vitro. Our hypothesis was that cold-tolerance of microtubules is caused by tubulin isotypes that differ in the primary sequence compared to mammalian tubulins. Here we show that transfection of human HepG2 cells with cod beta-tubulin induced cold-adaptation of the endogenous microtubules. Incorporation of one single tubulin isotype can induce cold-tolerance to cold-intolerant microtubules. Three cod beta-tubulin isotypes were tested and two of these (beta1 and beta2) transferred cold-tolerance to HepG2 microtubules, thus not all cod beta-tubulins were able to confer cold-stability.     

Presence of the betaII isotype of tubulin in the nuclei of cultured mesangial cells from rat kidney

Tubulin has generally been considered to be a cytosolic protein whose only function is to form microtubules. This assumption is supported by a great deal of evidence derived from immunohistochemical studies using antibodies directed against whole tubulin or its component polypeptides alpha- and beta-tubulin. We have re-examined the intracellular distribution of tubulin using monoclonal antibodies specific for the betaI, betaII, betaIII, and betaIV isotypes of beta-tubulin. Our test system is the cultured rat kidney mesangial cell. We have found that betaIII is absent from these cells and that beta1 and betaIV are present in microtubules throughout the cytosol. In contrast, betaII is present largely in the nuclei. Immunoblotting of purified nuclear extracts shows that the betaII-reactive antigen co-migrates with beta-tubulin. Extraction of the cytosol and chromatin suggests that betaII is concentrated in the nucleoli and also in a reticulated network in the rest of the nucleoplasm. An antibody to tyrosinated alpha-tubulin shows that alpha is also present in the nucleoli. Treatment of the cells with fluorescent colchicine shows an accumulation of colchicine in the nucleoli. Finally, fluorescently labeled alphabetaII-tubulin dimers, when microinjected into the cells, enter the nuclei and are concentrated in the nucleoli. These results suggest that the betaII isotype of tubulin is present as an alphabetaII dimer in the nuclei of cultured mesangial cells and suggest the possibility that different tubulin isotypes may have specific functions within the cell.     

Subtilisin cleavage of tubulin heterodimers and polymers.

Native pig brain tubulin in heterodimer or polymer form was subjected to limited proteolysis by subtilisin, which is known to cleave at accessible sites within the last 50 amino acids of the highly variable carboxyl-termini of the alpha and beta subunits. Heterodimeric tubulin or tubulin polymerized in the presence of 4 M glycerol or taxol was used in these experiments. Digested tubulin was purified by cycles of polymerization and depolymerization, ammonium sulfate precipitation, or ion-exchange chromatography in the absence or presence of nonionic detergent; however, smaller cleaved products of about 34,000 to 40,000 MW remained associated with the major cleaved subunits, alpha' and beta', under all purification conditions. In order to determine the effect of subtilisin cleavage on tubulin heterogeneity, purified native or subtilisin-cleaved tubulin was subjected to isoelectric focusing, followed by SDS-PAGE. The total number of isotypes was reduced from 17-22 for native alpha,beta tubulin to 7-9 for subtilisin-cleaved alpha',beta' tubulin. When tubulin heterodimers were cleaved, a single major beta' isotype was evident; however, when tubulin polymerized in 4 M glycerol was cleaved, two major beta' isotypes were found. Monoclonal antibodies that recognize a beta carboxyl-terminal peptide, residues 410-430, reacted with both major beta' isotypes, indicating that subtilisin cleavage occurred within the last 20 of the 450 amino acids. In order to establish whether this difference was in fact associated with polymer or heterodimer forms of tubulin, digestion was carried out in the presence of taxol, which stabilizes tubulin polymers. A single major beta' isotype different from the cleaved heterodimer, but coincident with one of the bands of the cleaved glycerol-induced polymers, was found when taxol-treated tubulin was digested. This result suggests the presence of more than one subtilisin site in the beta subunit, near residues 430-435, with different accessibility to the enzyme in the heterodimer and polymer form.