Release of tyrosine from tyrosinated tubulin. Some common factors that affect this process and the assembly of tubulin

Release of C-terminal tyrosine from tubulin and microtubules at steady state

Microtubule protein preparations purified by cycles of assembly-disassembly contain the enzyme tubulinyltyrosine carboxypeptidase (TTCPase). Using these preparations, containing tubulinyl[14C]tyrosine, we studied the release of [14C]tyrosine from assembled and non-assembled tubulin under steady-state conditions. It was found that both states of aggregation were detyrosinated at similar rates by the action of the endogenous TTCPase. However, practically no release of [14C]tyrosine from the non-assembled tubulin pool was found when microtubules were previously eliminated from the incubation mixture. These results indicated that non-assembled tubulin requires to interact with microtubules to be detyrosinated. This interaction seems to occur through the incorporation of dimers into microtubules, since when the capability of tubulin to incorporate into microtubules was diminished by binding of colchicine a concomitant decrease in the rate of release of tyrosine was observed. When detyrosination was accelerated by increasing the concentration of TTCPase relative to the microtubule protein concentration, microtubules were found to be detyrosinated faster than was non-assembled tubulin. Using exogenous TTCPase in an incubation system in which the formation of microtubules was not allowed, tubulinyl[14C]tyrosine and tubulinyl[14C]tyrosine-colchicine complex were shown to have similar capabilities to act as substrates for this enzyme. Free colchicine was shown not to affect the activity of TTCPase.

Distinct populations of microtubules: tyrosinated and nontyrosinated alpha tubulin are distributed differently in vivo

A unique post-translational modification of tubulin has previously been described in which a tyrosine residue is reversibly added to the C terminus of the alpha-tubulin subunit. We have prepared peptide antibodies that specifically react (shown by competitive immunoassay and Western blots) with the tyrosinated (Tyr) and nontyrosinated (Glu) forms of alpha-tubulin. Immunofluorescence with these antibodies demonstrated that the distributions of Tyr and Glu tubulin in fixed cells were markedly different. Tyr tubulin was found throughout the interphase network of microtubules and in the metaphase spindle, whereas Glu tubulin was present in a limited subset of interphase microtubules and was absent from the astral fibers of the metaphase spindle. Double immunofluorescence showed that Glu and Tyr microtubules comprised distinct subsets of the total cellular microtubules. These results suggest that tyrosination is involved in the establishment of separate populations of microtubules that may functionally distinct.

Tubulin tyrosinolated in vivo can be different from that tyrosinolated in vitro

Tubulin can be tyrosinolated, in the presence of ATP, by tubulin-tyrosine ligase, and tyrosine can be released by the same enzyme in the presence of ADP plus inorganic phosphate. There is however a 'non-substrate' component of tubulin which can not be tyrosinolated or detyrosinolated by this enzyme. Tubulin tyrosinolated in vivo was found to be the non-substrate species in HeLa cells, and the substrate species in cultured neuronal cells. In this respect HeLa tubulin resembled membrane-associated tubulin from brain, and neuronal cell tubulin resembled brain cytosolic tubulin.

Inhibition of brain tubulinyl-tyrosine carboxypeptidase by endogenous proteins

When a 25-50% ammonium-sulphate-insoluble fraction from a bovine brain preparation was chromatographed on a cellulose phosphate column, several protein fractions which inhibit the activity of tubulinyl-tyrosine carboxypeptidase were obtained. One of these fractions exhibited activity of fructose-bisphosphate aldolase (EC 4.1.2.13) and the enzyme accounted for more than 95% of the protein of this fraction as judged by sodium dodecyl sulphate-polyacrylamide gel electrophoresis. The inhibitory activities of the two protein fractions which had the highest activity per mg of protein were practically abolished by pretreatment with pronase; preincubation with trypsin, on the other hand, caused only a partial inactivation of the inhibitors. The inhibitory activities were little affected by heating at 90 degrees C for 5 min. Preincubation with purified tubulinyl-tyrosine carboxypeptidase caused a great decrease of the inhibitory activities of these two fractions, leaving open the possibility that these inhibitors act as substrates of the carboxypeptidase.

A rat monoclonal antibody reacting specifically with the tyrosylated form of alpha-tubulin. I. Biochemical characterization, effects on microtubule polymerization in vitro, and microtubule polymerization and organization in vivo

The antigenic site recognized by a rat monoclonal antibody (clone YL 1/2) reacting with alpha-tubulin (Kilmartin, J.V., B. Wright, and C. Milstein, 1982, J. Cell Biol., 93:576-582) has been determined and partially characterized. YL 1/2 reacts specifically with the tyrosylated form of brain alpha-tubulin from different mammalian species. YL 1/2 reacts with the synthetic peptide Gly-(Glu)3-Gly-(Glu)2-Tyr, corresponding to the carboxyterminal amino acid sequence of tyrosylated alpha-tubulin, but does not react with Gly-(Glu)3-Gly-(Glu)2, the constituent peptide of detyrosylated alpha-tubulin. Electron microscopy as well as direct and indirect immunofluorescence microscopy shows that YL 1/2 binds to the surface of microtubules polymerized in vitro and in vivo. Further in vitro studies show that the antibody has no effect on the rate and extent of microtubule polymerization, the stability of microtubules, and the incorporation of the microtubule-associated proteins (MAP2) and tau into microtubules. In vivo studies using Swiss 3T3 fibroblasts injected with YL 1/2 show that; when injected at low concentration (2 mg IgG/ml in the injection solution), the antibody binds to microtubules without changing their distribution in the cytoplasm. Injection of larger concentration of YL 1/2 (6 mg IgG/ml) induces the formation of microtubule bundles, and still higher concentrations cause the aggregation of microtubule bundles around the nucleus (greater than 12 mg IgG/ml).

Multiple forms of tubulin in Polytomella and Chlamydomonas: evidence for a precursor of flagellar alpha-tubulin

The quadriflagellate alga polytomella agilis contains several alpha-tubulins with distinct isoelectric points (McKeithan, T.W., and J.L. Rosenbaum, 1981, J. Cell Biol., 91:352-360). While alpha-3 is the major component in flagella, alpha-1 predominates in cytoskeletal microtubules. For determination of whether the differences in alpha- tubulins are due to distinct genes or to posttranslational modification of a common alpha-tubulin precursor, poly A+ RNA was isolated from deflagellated and control (nonregenerating) cells and translated in vitro. Approximately twice as much alpha-1 was synthesized using RNA from deflagellated as compared to control cells; however, there was no detectable synthesis in vitro of alpha-3 in either. These results suggest that alpha -3 tubulin is formed in vivo by posttranslational modification of a form co- migrating with, and possibly identical to, cytoskeletal alpha-tubulin. In the related alga chlamydomonas reinhardii deflagellation greatly stimulates synthesis of tubulin and tubulin mRNA. As in polytomella, the principal alpha-tubulin synthesized both in vivo and in vitro following deflagellation in chlamydomonas is more basic than the major alpha-tubulin and appears to correspond to alpha-1 tubulin in polytomella. The conversion of alpha-1 to alpha-3 receives additional support from in vivo labeling and pulse-chase experiments. In addition, in both polytomella and chlamydomonas some conversion of alpha-1 to alpha-3 appears to occur even when protein synthesis is inhibited.

Modulation of some parameters of assembly of microtubules in vitro by tyrosinolation of tubulin

Using tyrosinolated and detyrosinolated tubulins, we have compared several parameters of microtubule assembly in vitro. Rates and extents of polymerization were the same under all conditions, but microtubules assembled from detyrosinolated tubulin in the presence of crude microtubule-associated proteins (MAPs) or subsaturating MAP-2 contained a smaller proportion of the MAPs. Preliminary results indicate that this may be a function of the phosphorylation state of MAP-2. Tyrosinolated tubulin assembled into relatively shorter microtubules in the presence of saturating MAP-2. When assembly was induced with substoichiometric concentrations of taxol, in place of MAPs, the rate and extent of assembly were about twice as great with tyrosinolated tubulin.

Enzymatic detyrosination of tubulin tyrosinated in rat brain slices and extracts

Tubulin was tyrosinated in slices and in extracts of brain of rats of 3, 25, and 120 days of age by successive incorporation of [14C]tyrosine and [3H]tyrosine, respectively. The release of the incorporated amino acid was measured by using tubulinyl-tyrosine carboxypeptidase, carboxypeptidase A, and tubulin-tyrosine ligase. With the carboxypeptidases no differences in either the rates or the extents of the release of tyrosine between these two differently labeled tubulins were found. Differences were found when the detyrosination was catalyzed by the ligase and these were attributed to a higher inactivation of tubulin labeled in slices than of that labeled in extracts.

Total tubulin and its aminoacylated and non-aminoacylated forms during the development of rat brain

The amount of total tubulin in the soluble fraction of rat brain was measured by a method based on the purification of tubulin previously labeled by incorporation of [14C]tyrosine in the C terminus of its alpha-chain. The tubulin content decreased from 2.01 to 1.30 nmol/mg protein when the animals passed from 4 to 30 days of age and then remained practically constant. The amounts of aminoacylated and non-aminoacylated tubulin present in the soluble brain extracts were determined from the incorporation of [14C]tyrosine into the free acceptor sites of tubulin preparations, that were preincubated without carboxypeptidase A or with this enzyme to eliminate tyrosine and phenylalanine from the C terminus of the alpha-chain of tubulin. The values obtained were corrected for the inactivation of tubulin to accept [14C]tyrosine that occurred during the isolation and incubation of the soluble fractions. The ratio non-aminoacylated/aminoacylated tubulin increased from 1.62 +/- 0.03 in the 4-day-old rats to 2.11 +/- 0.17 in the 120-day-old rats. The aminoacylatable tubulin, that is the sum of aminoacylated plus non-aminoacylated tubulin, decreased from 1.71 to 0.75 nmol/mg protein from 4-day-old to 30-day-old rats respectively and then remained practically constant. The amount of aminoacylatable tubulin is lower than that of total soluble tubulin. Therefore there is a fraction of tubulin that is unable to accept tyrosine. This non-aminoacylatable tubulin fraction increases with the age of the animal so that in the 120-day-old rats this tubulin species accounts for 48% of the total soluble tubulin.

Tubulinyl-tyrosine carboxypeptidase from chicken brain: properties and partial purification

Tyrosine can be released from tubulinyl-tyrosine by the action of a brain carboxypeptidase. The molecular weight of this enzyme found by gel filtration through a column of Sephadex G-200 was 90,000. The enzyme was very unstable in a purified preparation in which the activity per milligram of protein was increased 250-fold with respect to the starting material. The precise magnitude of the purification cannot be stated because of the unknown amount of endogenous tubulinyl-tyrosine in the material to be assayed. A comparative study was done between tubulinyl-tyrosine carboxypeptidase (TTCP) activity and pancreatic carboxypeptidase A (CPA, EC 3.4.12.2) activity using tubulinyl-[14C]tyrosine as substrate. The most remarkable differences found are: MgCl2 (2 mM), phenyl acetate (10 mM), or EDTA (5 mM) increased the TTCP activity whereas the CPA activity was strongly inhibited by these compounds. Iodoacetate (2 mM) and ZnCl2 (0.1 mM) inhibited the TTCP activity more than the CPA activity. Contrarily, mercaptoethanol (50 mM) and dimethyl sulfoxide (5%) showed a stronger inhibitory effect on CPA than on TTCP. Of several N-carbobenzoxy dipeptides (Z-dipeptides) tested the greatest inhibitory effects on TTCP activity were obtained with Z-Glu-Tyr and Z-Glu-Phe, although strong inhibitory effects on CPA were also obtained with other Z-dipeptides.