Selection of a nucleation-promoting element following chemical modification of tubulin

Following a 16-h incubation with a large excess of 2,5-hexanedione (2,5-HD) while in the assembled state, bovine brain tubulin contained a powerful nucleating component, the presence of which lowered the dissociation rate from 83 s-1 for untreated tubulin to 13 s-1 for 2,5-HD-treated tubulin. This nucleating component could be selectively concentrated by sequential stringent (conditions of low temperature and low tubulin concentration) cycles of assembly and disassembly. In 2-(N-morpholino)ethanesulfonic acid buffer without glycerol, the critical concentration of assembly of untreated tubulin (2.4 mg/mL) was 19 times higher than that of 2,5-HD-treated tubulin subjected to three sequential stringent cycles of assembly and disassembly (0.13 mg/mL). This highly nucleating 2,5-HD-treated tubulin preparation could both copolymerize with untreated tubulin and seed subcritical concentration assembly of untreated tubulin. Experiments to define the assembly-altering component have identified structural alterations to the alpha-tubulin monomer. While the alpha-tubulin subunit of native untreated tubulin dimer contained no chymotryptic cleavage sites, the native 2,5-HD-treated alpha-tubulin subunit was cleaved by chymotrypsin to yield a 37-kDa C-terminal fragment.     

Identification of proteins binding the native tubulin dimer

Microtubules play an essential role in eukaryotic cells, where they perform a wide variety of functions. In this paper, we describe the characterization of proteins associated to tubulin dimer in its native form, using affinity chromatography and mass spectrometry. We used an immunoaffinity column with coupled-monoclonal antibody directed against the alpha-tubulin C-terminus. Tubulin was first loaded onto the column, then interphase and mitotic cell lysates were chromatographed. Tubulin-binding proteins were eluted using a peptide mimicking the alpha-tubulin C-terminus. Elution fractions were analyzed by SDS-PAGE, and a total of 14 proteins were identified with high confidence by mass spectrometry. These proteins could be grouped in four classes: known tubulin-binding proteins, one microtubule-associated protein, heat shock proteins, and proteins that were not shown previously to bind tubulin dimer or microtubules.     

Posttranslational tyrosination/detyrosination of tubulin

Tubulin can be posttranslationally modified at the carboxyl terminus of the alpha-subunit by the addition or release of a tyrosine residue. These reactions involve two enzymes, tubulin: tyrosine ligase and tubulin carboxypeptidase. The tyrosine incorporation reaction has been described mainly in nervous tissue but it has also been found in a great variety of tissues and different species. Molecular aspects of the reactions catalyzed by these enzymes are at present well known, especially the reaction carried out by the ligase. Several lines of evidence indicate that assembled tubulin is the preferred substrate of the carboxypeptidase, whereas nonassembled tubulin is preferred by the ligase. Apparently this posttranslational modification does not affect the capacity of tubulin to form microtubules but it generates microtubules with different degrees of tyrosination. Variation in the content of the carboxyterminal tyrosine of alpha-tubulin as well as changes in the activity of the ligase and the carboxypeptidase are manifested during development. Changes in the cellular microtubular network modify the turnover of the carboxyterminal tyrosine of alpha-tubulin. Different subsets of microtubules with different degrees of tyrosination have been detected in interphase cells and during the mitotic cycle. Data from biochemical, immunological, and genetic studies have been compiled in this review; these are presented, with pertinent comments, with the hope of facilitating the comprehension of this particular aspect of the microtubule field.     

The interaction between subunits in the tubulin dimer.

Limited proteolysis and chemical cross-linking techniques have been used to study the interaction between alpha- and beta-tubulin subunits. Trypsin digestion of tubulin dimer resulted in the cleavage of the alpha-subunit into two fragments, whereas chymotrypsin cleaved the beta-subunit into two distinct fragments. All of these fragments have been mapped on the tubulin subunits by further proteolysis with formic acid. Cross-linking of trypsin- and chymotrypsin-cleaved subunits has been performed with two different cross-linker agents of different cross-linking distance. The addition of formaldehyde resulted in the cross-linking of the alpha-tubulin N-terminal fragment with beta-tubulin C-terminal domain. The same result was obtained when methyl 4-mercaptobutyrimidate was used.     

The B-ring substituent at C-7 of colchicine and the alpha-C-terminus of tubulin communicate through the "tail-body" interaction

The carboxy terminals of alphabeta-tubulins are flexible regions rich in acidic amino acid residues that play an inhibitory role in the polymerization of tubulin to microtubules. We have shown that the binding of colchicine and its B-ring analogs (with C-7 substituents) to tubulin are pH sensitive and have high activation energies. Under identical conditions, the binding of analogs without C-7 substituents is pH independent and has lower activation energy. Beta-C-terminus-truncated tubulin (alphabeta(s)) shows similar pH sensitivity and activation energy to native tubulin (alphabeta). Removal of the C-termini of both subunits of tubulin (alpha(s)beta(s)) or the binding of a basic peptide P2 to the negatively charged alpha-C-terminus of tubulin causes a colchicine-tubulin interaction independent of pH with a low activation energy. Tubulin dimer structure shows that the C-terminal alpha-tail is too far from the colchicine binding site to interact directly with the bound colchicine. Therefore, it is likely that the interaction of the alpha-C-terminus with the main body of tubulin indirectly affects the colchicine-tubulin interaction via conformational changes in the main body. We therefore conclude that in the presence of tail-body interaction, a B-ring substituent makes contact with the alpha-tubulin and induces significant conformational changes in alpha-tubulin.     

Pig sperm tail tubulin. Its extraction and characterization

Axonemal tubulin extracted from pig sperm tails has been characterized by one- and two-dimensional electrophoresis and by one-dimensional peptide mapping. The electrophoretic mobilities of its subunits after reduction and carboxymethylation were similar to those of the major subunits of pig brain tubulin. Sperm tail tubulin subunits also had roughly the same isoelectric points as pig brain tubulin subunits, except that they appeared to have a relatively larger tailing effect. The proteolytic cleavage pattern of the pig sperm tail beta-tubulin closely resembled those of both the tunicate (Ciona intestinalis) sperm beta-tubulin and pig brain beta-tubulin. The peptide pattern of pig sperm tail alpha-tubulin, however, was more similar to that of tunicate sperm tail alpha-tubulin than to that of pig brain alpha-tubulin. This supports the hypothesis put forward in a previous investigation [1] that functionally similar tubulins from taxonomically distant species can be more related than functionally dissimilar tubulins from the same species.     

Tyrosination of microtubules and non-assembled tubulin in brain slices

Brain slices were used to examine comparatively the incorporation of [14C]tyrosine into the C terminus of alpha-tubulin of the microtubule and non-assembled tubulin pools. We found that the incorporation of [14C]tyrosine from 5 min up to 60 min of incubation was higher in microtubules than in non-assembled tubulin. The possibility that this result was due to the activity of tubulin carboxypeptidase or tubulin:tyrosine ligase during the in vitro isolation of tubulin was discarded. We also found that tubulin:tyrosine ligase was mainly associated with microtubules when brain slices were homogenized under microtubule-preserving conditions. Conversely the enzyme behaved as a soluble entity when homogenization was performed under conditions that do not preserve microtubules. In addition, soluble tubulin:tyrosine ligase did not become sedimentable when in vitro conditions were changed to induce the formation of microtubules. The results presented in this work indicate the possibility that, in vivo, microtubules and not tubulin dimers are the major substrate for tubulin:tyrosine ligase. This is in contrast with previous findings from in vitro experiments, which showed a preference of the ligase for non-assembled tubulin.     

Glycolytic enzyme–tubulin interactions: Role of tubulin carboxy terminals

Tubulin and microtubules were modified with the protease, subtilisin. The modification reduced the length of α-or β-tubulin by cleaving a peptide fragment from the C-terminals. Generation of α′β′-tubulin, which is cleaved at both the α- and β-subunit terminals, and αβ′-tubulin, which is cleaved at the β′-subunit C-terminal, have already been reported. In this work an isotype, α′β-tubulin, was produced. The three modified tubulin isotypes were compared for their ability to interact with glycolytic enzymes. Cleavage of α led to a poorer interaction when tested via affinity chromatography. Tubulin also inhibits the activity of aldolase and glyceraldehyde 3-phosphate dehydrogenase. When the α-subunit C-terminal was intact, inhibition was greatest. These results imply that the C-terminal of the tubulin α-subunit is subunit is responsible for interactions with glycolytic enzymes.     

Modification of tubulin by tyrosylation in cells and extracts and its effect on assembly in vitro

A post-translational modification of tubulin with potential regulatory significance has been revealed by the discovery of an enzyme (tubulin-tyrosine ligase) in brain extracts which can add a tyrosine residue to the alpha chain, apparently through peptide bond linkage to a C-terminal glutamate. We have investigated whether this modification also occurs in vivo, and whether it alters the extent to which tubulin can assemble in vitro. Cytoplasmic tubulin purified from bovine brain by cycles of assembly was shown to be partially tyrosylated. Carboxypeptidase A digestion of isolated alpha chains liberated about 0.3 equivalent of tyrosine. Brief digestion of native tubulin increased the proportion of alpha chains which could be tyrosylated by ligase, from 25 to 45%. The tubulin assembled to the same extent before and after carboxypeptidase treatment. When tubulin was purified after introducing labeled tyrosine with ligase, the labeled species assembled in the same proportion as unlabeled. Thus tubulin can be incorporated into microbubules in vitro with or without C-terminal tyrosine. An apparent resolution of alpha chain into two components by hydroxylapatite chromatography was shown not to be due to the presence or absence of C-terminal tyrosine. Tubulin-tyrosine ligase was found in extracts of every rat tissue examined, but was not detected in sea urchin eggs before or after fertilization, in Tetrahymena cells or cilia, or in yeast. Cultured neuroblastoma cells fixed tyrosine into tubulin alpha chains under conditions where protein synthesis was inhibited; this in vivo fixation appeared to be into an insoluble moiety of tubulin. Incidental to these studies, a new assay utilizing an enamine substrate for carboxypeptidase was investigated.     

OP18/stathmin binds near the C-terminus of tubulin and facilitates GTP binding.

It is has been previously suggested that the protein Op18/stathmin may interact with tubulin via the alpha-tubulin subunit [Larsson, N., Marklund, U., Melander Gradin, H., Brattsand, G. & Gullberg, M. (1997) Mol. Cell. Biol. 17, 5530-5539]. In this study we have used limited proteolysis and cross-linking analysis to localize further the stathmin-binding site on alpha-tubulin. Our results indicate that such a binding site is in a region close to the C-terminus of the molecule comprising residues 307 to the subtilisin-cleavage site on the alpha-tubulin subunit. Based on a recent model of the structure of tubulin [Nogales, E., Wolf, S.G. & Dowing, D.H. (1998) Nature (London) 391, 199-203], we found that this region contained the same areas that may be involved in longitudinal contacts of alpha-tubulin subunits within the microtubule. We also observed that the binding of stathmin to tubulin can modulate the binding of GTP to tubulin, as a consequence of a conformational change in the beta-tubulin subunit that occurs upon interaction of stathmin with tubulin.     

Regulation of cytoplasmic tubulin carboxypeptidase activity during neural and muscle differentiation: characterization using a microtubule-based assay.

A cycle of posttranslational modification of alpha-tubulin has previously been described in higher eukaryotes, in which a C-terminal tyrosine residue is removed and replaced by two complementary cytoplasmic enzymes. The activity of the detyrosinating enzyme, tubulin carboxypeptidase (TCP), and its potential for regulating the level of detyrosinated (Glu) subunits in microtubules (MTs) is of great interest, since TCP catalyzes the primary modification of tubulin and since the level of Glu alpha-tubulin in MTs increases during a variety of differentiative and morphogenetic events. As a first step in examining the role of TCP in cellular morphogenesis, it was necessary to develop an assay for TCP with sufficient sensitivity and specificity to detect TCP activity during these events. Unlike previously described assays for TCP, ours makes use of the affinity TCP exhibits for MTs. NGF-induced neurite outgrowth in PC-12 cells was accompanied by a moderate (approximately 2-fold) increase in TCP activity, while myogenesis of L6 cells resulted in an almost insignificant decrease in activity. Measurements of TCP activity during differentiation were correlated with the level of extract Tyr tubulin, which increased (approximately 37%) during neurite outgrowth and was unchanged during myogenic differentiation. Our results suggest that TCP activity is regulated relative to its substrate, Tyr tubulin, and that changes in MT dynamics, rather than enzymatic activities, are the primary determinants of MT posttranslational modification state during differentiation. In addition, the assay we have devised for TCP and the characterization of TCP during differentiation may allow the future delineation of the mechanism(s) of regulation of TCP and the role this enzyme plays in modulating MT function during differentiation.     

Calcium binding properties of the beta tubulin subunit from chicken erythrocytes

Taxol-stabilised erythrocyte microtubules assembled less readily than similarly prepared brain microtubules on adding 10(-4) M-10(-3) M concentrations of calcium at 2 degrees C. Scatchard plot analyses of the high affinity calcium binding sites showed that the erythrocyte tubulin contained only 0.9 high affinity binding sites per dimer compared to 1.4 binding sites per dimer for brain tubulin. Association constants, however, for calcium binding to both erythrocyte and brain tubulin were similar (3.0 x 10(-6) M and 2.1 x 10(-6) M). The beta-tubulin subunit appeared to be responsible for the lower calcium binding ability of erythrocyte tubulin as shown by a gel overlay assay with 45Ca. Strains-all, a dye that stains many calcium binding proteins blue, did not stain erythrocyte beta-tubulin or its chymotryptic C-terminal fragment blue as was the case for brain beta-tubulin and its chymotryptic C-terminal fragment. We suggest that the lower calcium binding ability of erythrocyte beta-tubulin may be implicated in the differential behaviour of erythrocyte microtubules.     

Localization of a highly divergent mammalian testicular alpha tubulin that is not detectable in brain.

Sequence analysis of a mouse testicular alpha-tubulin partial cDNA, pRD alpha TT1, reveals an isotype that differs from both the somatic and the predominant testicular alpha tubulins at approximately 30% of the 212 amino acid residues determined. Although this mouse testicular cDNA retains the highly conserved sequence, Glu-Gly-Glu-Glu, found in the carboxyl termini of many alpha tubulins, the protein extends substantially beyond this sequence and does not terminate with a C-terminal tyrosine. Using rabbit antiserum prepared to a novel synthetic peptide predicted from this mouse testis alpha-tubulin cDNA, we have have detected by immunoblot and indirect immunofluorescence an antigenic epitope present in testicular alpha tubulin that is not detectable in brain alpha tubulins. We find that the antiserum specifically binds to the manchettes and meiotic spindles of the mouse testis but not with neural fibers or tubulin extracts of the adult mouse brain. These results demonstrate that at least one of the multiple alpha-tubulin isotypes of the mammalian testis is expressed and used in male germ cells but not in the brain.     

Tubulins in Trichomonas vaginalis: molecular characterization of alpha-tubulin genes, posttranslational modifications, and homology modeling of the tubulin dimer

We have isolated and analysed an alpha-tubulin-encoding gene (atub1) in an early-diverging eukaryote, Trichomonas vaginalis. The complete atub1 open reading frame included 1.356 bp encoding a polypeptide of 452 amino-acyl residues. A second alpha-tubulin gene (atub2) was amplified by PCR using primers derived from consensus alpha-tubulin amino acid sequences. Both T. vaginalis alpha-tubulin sequences showed high identity to those described in other parabasalids (94.4%-97.3%), and exhibited a high degree of similarity to sequences from Metazoa (such as pig brain) and diplomonads (such as Giardia). Despite large evolutionary distances previously observed between trichomonads and mammals, the three-dimensional model of the T. vaginalis tubulin dimer was very similar to that of pig brain. Possible correlations between alpha-tubulin sequences and posttranslational modifications (PTMs) were examined. Our observations corroborated previous data obtained in T. vaginalis using specific anti-PTMs antibodies. As described in the related species Tritrichomonas mobilensis, microtubules are likely acetylated, non-tyrosinated, glutamylated, and non-glycylated in T. vaginalis. Evolutionary considerations concerning the time of appearance of these tubulin PTMs are also discussed since trichomonads are potentially one of the earliest diverging eukaryotic lineages.     

Preferential action of a brain detyrosinolating carboxypeptidase on polymerized tubulin

A carboxypeptidase purified from brain catalyzes the release of COOH-terminal tyrosine without further digesting tubulin. It is distinct from previously described carboxypeptidases, and appears to have specificity for tubulin as it is not inhibited by peptides and proteins with COOH-terminal tyrosine, and because, unlike carboxypeptidase A (which by removing tyrosine from aldolase causes its inactivation), this enzyme does not decrease aldolase activity. The enzyme detyrosinolates both self-assembly-competent (cycle-purified) and -incompetent (phosphocellulose-purified) tubulin. However, under assembly conditions the rate was 2-3-fold higher for competent tubulin. Preincubation of assembly-competent tubulin with podophyllotoxin or colchicine resulted in a parallel concentration-dependent inhibition of tubulin polymerization and detyrosinolation. Similarly, when incompetent tubulin was induced to polymerize by preincubation with purified microtubule-associated protein 2 (an assembly-promoting protein) or taxol, the initial rate of its detyrosinolation increased 3-5-fold, and this increase was blocked if podophyllotoxin was also added along with microtubule-associated protein 2 or taxol during the preincubation. Oligomers induced by adding vinblastine to incompetent tubulin were also detyrosinolated more rapidly, and the stimulation was abolished by maytansine, which has been shown to disperse the vinblastine-induced oligomers. When polymerized and subunit fractions were separated after a steady state mixture had been partially digested with the carboxypeptidase, the former was found to have lost 2-3 times more COOH-terminal tyrosine. Although both polymer and monomer can be detyrosinolated by the enzyme, polymeric and oligomeric forms are the preferred substrates. Carboxypeptidase appeared to release tyrosine at the same rate from populations of short and long microtubules.     

In vivo incorporation of [14C]tyrosine into the C-terminal position of the α subunit of tubulin

In vitro incorporation of [¹⁴C]tyrosine into the C-terminal position of the α subunit of tubulin was not affected by 4 mm cycloheximide. This inhibitor of protein synthesis was used for in vivo experiments. The in vivo incorporation of [¹⁴C]tyrosine into soluble brain protein of cycloheximide-treated rats was 10% of that of untreated rats. Treatment with vinblastine sulfate of the soluble brain protein showed that the incorporation of [¹⁴C]tyrosine into tubulin was higher in cycloheximide-treated than in untreated rats with respect to the incorporation into the total soluble protein. In the case of cycloheximide-treated rats, about 60% of the radioactivity incorporated into protein was released by the action of carboxypeptidase A, whereas 10% was liberated from the protein of untreated rats. The radioactive compound released by the action of carboxypeptidase A was identified as [¹⁴C]tyrosine. The α and β subunits of tubulin from animals that received [¹⁴C]tyrosine were separated by polyacrylamide gel electrophoresis. The radiosactivity ratio of $\text{α}\text{β}$ subunits of tubulin from cycloheximide-treated rats was threefold higher than that of untreated rats. When a mixture of [¹⁴C]amino acids was injected, the radioactivity ratio of $\text{α}\text{β}$ subunits of tubulin was similar for cycloheximide-treated and untreated rats. The results reported are consistent with the assumption that the α subunit of tubulin can be tyrosinated in vivo.     

Posttranslational modifications of tubulin in cultured mouse brain neurons and astroglia

Posttranslational modifications of tubulin were analyzed in mouse brain neurons and glia developing in culture. Purified tubulin was resolved by isoelectric focusing. After 3 weeks of culture, neurons were shown to express a high degree of tubulin heterogeneity (8 alpha and 10 beta isoforms), similar to that found in the brain at the same developmental stage. Astroglial tubulin exhibits a less complex pattern consisting of 4 alpha and 4 beta isoforms. After incubation of neuronal and glial cells with 3H-acetate in the presence of cycloheximide, a major posttranslational label was found associated with alpha-tubulin and a minor one with beta-tubulin. The acetate-labeled isotubulins of neurons were resolved by isoelectric focusing into as many as 6 alpha and 7 beta isoforms, while those of astroglia were resolved into only 2 alpha and 2 beta isoforms. The same alpha isoforms were also shown to react with a monoclonal antibody recognizing selectively the acetylated form(s) of alpha-tubulin. Whether acetate-labeling of alpha-tubulin in these cells corresponds to the acetylation of Lys40, as reported for Chlamydomonas reinhardtii, is discussed according to very recent data obtained by protein sequence analysis. Tubulin phosphorylation was analyzed by incubation of cell cultures with 32PO4. No phosphorylation of alpha-tubulin isoforms was detected. A single beta-tubulin isoform (beta'2), expressed only in neurons, was found to be phosphorylated. This isoform is similar to that previously identified in differentiated mouse neuroblastoma cells.     

Coordination of posttranslational modifications of bovine brain alpha-tubulin. Polyglycylation of delta2 tubulin

Microtubules participate in a large number of intracellular events including cell division, intracellular transport and secretion, axonal transport, and maintenance of cell morphology. They are composed of tubulin, a heterodimeric protein, consisting of two similar polypeptides alpha and beta. In mammalian cells, both alpha- and beta-tubulin occur as seven to eight different genetic variants, which also undergo numerous posttranslational modifications that include tyrosination-detyrosination and deglutamylation, phosphorylation, acetylation, polyglutamylation, and polyglycylation. Tyrosination-detyrosination is one of the major posttranslational modifications in which the C-terminal tyrosine residue in alpha-tubulin is added or removed reversibly. Although this modification does not alter the assembly activity of tubulin in vitro, these two forms of tubulin have been found to be distributed differently in vivo and are also correlated with microtubule stability (Gunderson, G. G., Kalnoski, M. H., and Bulinski, J. C. (1984) Cell 38, 779-789). Thus, the question arises as to whether these two forms of tubulin differ in any other modifications. In an effort to answer this question, the tyrosinated and the nontyrosinated forms of the alpha1/2 isoform have been purified from brain tubulin by immunoaffinity chromatography. matrix-assisted laser desorption/ionization-time of flight mass spectrometric analysis of the C-terminal peptide revealed that the tyrosinated form is polyglutamylated with one to four Glu residues, while the Delta2 tubulin is polyglycylated with one to three Gly residues. These results indicate that posttranslational modifications of tubulin are correlated with each other and that polyglutamylation and polyglycylation of tubulin may have important roles in regulating microtubule assembly, stability, and function in vivo.     

Detyrosination of tubulin regulates the interaction of intermediate filaments with microtubules in vivo via a kinesin-dependent mechanism

Posttranslationally modified forms of tubulin accumulate in the subset of stabilized microtubules (MTs) in cells but are not themselves involved in generating MT stability. We showed previously that stabilized, detyrosinated (Glu) MTs function to localize vimentin intermediate filaments (IFs) in fibroblasts. To determine whether tubulin detyrosination or MT stability is the critical element in the preferential association of IFs with Glu MTs, we microinjected nonpolymerizable Glu tubulin into cells. If detyrosination is critical, then soluble Glu tubulin should be a competitive inhibitor of the IF-MT interaction. Before microinjection, Glu tubulin was rendered nonpolymerizable and nontyrosinatable by treatment with iodoacetamide (IAA). Microinjected IAA-Glu tubulin disrupted the interaction of IFs with MTs, as assayed by the collapse of IFs to a perinuclear location, and had no detectable effect on the array of Glu or tyrosinated MTs in cells. Conversely, neither IAA-tyrosinated tubulin nor untreated Glu tubulin, which assembled into MTs, caused collapse of IFs when microinjected. The epitope on Glu tubulin responsible for interfering with the Glu MT-IF interaction was mapped by microinjecting tubulin fragments of alpha-tubulin. The 14-kDa C-terminal fragment of Glu tubulin (alpha-C Glu) induced IF collapse, whereas the 36-kDa N-terminal fragment of alpha-tubulin did not alter the IF array. The epitope required more than the detyrosination site at the C terminus, because a short peptide (a 7-mer) mimicking the C terminus of Glu tubulin did not disrupt the IF distribution. We previously showed that kinesin may mediate the interaction of Glu MTs and IFs. In this study we found that kinesin binding to MTs in vitro was inhibited by the same reagents (i.e., IAA-Glu tubulin and alpha-C Glu) that disrupted the IF-Glu MT interaction in vivo. These results demonstrate for the first time that tubulin detyrosination functions as a signal for the recruitment of IFs to MTs via a mechanism that is likely to involve kinesin.     

Microtubules in the fission yeast Schizosaccharomyces pombe contain only the tyrosinated form of alpha-tubulin.

The state of tubulin tyrosination in the fission yeast Schizosaccharomyces pombe was investigated using a combination of indirect immunofluorescence microscopy and Western blotting. Antibodies specific for the tyrosinated form of alpha-tubulin stained all microtubule arrays in wild type cells and recognised the two alpha-tubulin polypeptides in Western blots of cell extracts enriched for tubulin by DEAE-Sephadex chromatography. Antisera that specifically recognised the detyrosinated, glu, form, on the other hand, gave consistently negative results, both in cells undergoing rapid exponential growth and in those allowed to accumulate in stationary phase. Neither the "ageing" of microtubules, by arresting cells at different points (late G1 or G2/M) in the cell division cycle, nor stabilising them, using D2O, lead to any detectable tubulin detryrosination. These results suggest that S. pombe lacks the carboxypeptidase that carries out the tubulin detyrosination reaction. This is the first report of an organism that possesses the correct C-terminal alpha-tubulin sequence yet fails to carry out this post-translational modification. The implication of this novel finding for the biological role of these events is discussed.