Bid

Cell (1996)

0 Elsevier,

83

87,83-93

Paris

Original

article

The tubulin gene family of Paramecium: Characterization and expression of the aPT1 and aPT2 genes which code for a-tubulins with unusual C-terminal amino acids, GLY and ALA Pascale Dupuis-Williams

a*, Catherine Klotz a, Honor6 Mazarguil b, Janine Beisson a

a Centre de Ghe’tique Moltkulaire, Centre National de la Recherche Scientifique, Associe’ h I’llniversite’ Pierre et Marie Curie, 91198 Gif-sur-Yvette; b Institut de Pharmacologic et de Biologie Structurale, Centre National de la Recherche Scientijique, 31400 Toulouse, France (Received 15 July 1996; accepted 26 September 1996)

- The ciliated protozoan Paramecium harbours a particularly large diversity of microtubule networks, ranging from the elaborate and stable ciliary axonemes and basal bodies to very dynamic cytoplasmic, cortical or intranuclear arrays. Their organization and individual cycle of assembly/disassembly are well known and extensive immunocytochemical studies of the post-translational modifications in the various microtubule systems have been reported. However, in order to better understand the biogenesis of these multiple and diverse microtubule arrays, it seemed necessary to characterize the tubulin gene family. We show that P tetraurelia possesses four a- and three Pgenes and we report the cloning and sequencing of two intronless a-genes, @Tl and clpT2, which code for very similar polypeptides, differing only by their unusual C-terminal amino acids, respectively GLY and ALA. Partial sequencing of the two other a-genes suggests an absence of any further isotype diversity. In an attempt to study the expression of @Tl and cPT2, polyclonal antibodies were raised against the twelve C-terminal amino acids corresponding to the deduced polypeptide sequences. The reactivity of these anti-sequence antibodies was studied on blots of soluble tubulin and in situ and compared with that of other well characterized anti-a-tubulin antibodies. The molecular data show that in Paramecium, like in other ciliates, microtubule diversity does not arise from tubulin isotype diversity. The immunocytological data indicate that the native C-terminal sequences are predominantly detected in transient or nascent microtubule arrays and lead us to propose: 1) that the C-terminal TYR, absent in Paramecium and in most cilate species, has no intrinsic functional role; and 2) that post-translational modifications do not seem directly instrumental in the geometry and functions of microtubule arrays.

Summary

tubulin

multigene

family

/ cytoskeleton

/ tubulin

diversity

/ post-translational

Introduction Microtubules are engaged in a variety of functions as diverse as cell division, cell shapedetermination, intracellular transport and locomotion, related first to their intrinsic dynamic properties [37], and to the great diversity in their biochemical properties, organization, stability, and associated proteins (MAPS). The major componentsof microtubules, the a- and Ptubulins, are encoded by multigenic families whose complexity has increased in the course of evolution of eukaryotes [45, 481. The diversity resulting from differences in primary sequenceis further increased by post-translational modifications: acetylation, detyrosilation/tyrosylation of a-tubulins, phosphorylation of Ptubulins [30], glutamylation [21] and polyglycylation [61] of both sub-units. Over the last 20 years, researchon a large variety of organismshas aimed at sorting out the role of this molecular microheterogeneity in the functional and structural differentiation of microtubule arrays. The functional significance of tubulin sequencediversity, first postulated by Fulton and Simpson [25], has been

* Present &dress: Ldboratoire lake, Ecole Vauquelin.

SupCrieure 75005 Paris,

de neurobiologie de Chimie et Physique France

de la diversit Industrielles,

cellu10, rue

modifications

challenged by several lines of evidence. Not only do tubulins of various origins coassemble into microtubules in vitro, but gene disruption [33, 661 or transfection experiments [12], and microinjection experiments using exogenous tubulins [29, 741 all seemto show that an apparently specialized isoform can be replaced by any exogenousisoform to form a functional microtubule. Furthermore, in vivo, microtubules can be copolymers of the various available tubulin isoforms [46]. In several lower eukaryotes, all the microtubule arrays share a single a- or Pgene product [3, 52, 771. Then, to account for microtubule diversity, many studieswere devoted to the role of post-translational modifications, but have globally been elusive: in no case did the modifications appear instrumental in determining directly the properties of microtubule arrays and at any rate acetylation of a-tubulin was proved to be totally dispensable [26, 381. Conversely, genetic analysis of tubulin mutants in several specieshas more recently provided evidence that some microtubule systems specifically require, for assemblyor function, a given a- or Ptubulin gene product; such is the case in Drosophila (the @ gene is required for spermatogenesis[34, 501 and one a-tubulin for the ovocyte meisosisand cleavage mitosis [50]), in Caenorhabditis (the met-7 gene for touch sensing[65]), and in Aspergillus (the tubB gene for ascoporogenesis[36]). These studiesalso suggestedthat the functional specificity of microtubule net-

works

depends

on the ratio of the different

iaotyptta con>-

posing the tubulin pool. Furthermore, isotype specificit), was shown to influence the self-assembly properties 01 tubulins [53. 721 and the biological properties of microtu hules suchas their resistanceto cold [ 17. 5 I]. All in all it appearsthat different organismsmight adopt different strategiesand that the organization and functions of microtubule arrays can be controlled at diffe.rent levels. isotype diversity, ratio of particular isotypes, post-trznsla~, tional modifications, interactions with MAPS, and presunably properties of the MTOCs. This is why.a full understanding of the basis of microtubule diversity requires a comprehensive description of all these parameters. which have seldom been studied together in a single cell type. Hence, unicellular organismslike Purarner.Gn may be of special interest. Indeed, the ciliate P tetnrureliu displays a particularly high diversity of microtubule arrays. Cytological and immunological data distinguish 16 microtubule networks. which differ by their geometry, dynamic properties and functions [ 14, 20, 231 and extensive analysis of their post-translational modifications have been carried out 12. 5, ‘,I. As a necessary complement, we have undertaken a molecular study of the tubulin family in P tetruuretia and we show that it is composed of four a- and three Ptubulin genes, significantly more than in the other ciliates that have been studied: Tetrahymena [3], Euplntes [43] and Stylon~chia [16. 321. We have cloned and sequencedtwo of the a-genes, which encode proteins differing only in their C terminal amino acids, respectively ALA and GLY. Partial sequencedata on the two other a-genes seemto preclude any further isotype diversity, We have raised antibodies against the twelve C-terminal amino acids encoded by the two genes.Their preferential reaction on the most dynamic microtubule arrays and in the early stagesof assembly ot more stable arrays indicates that in structurally and functionally diverse microtubules arrays, the C-terminal sequenceof the a-tubulins, whether native oi weakly modified, remains accessible. These observations, along with the molecular data, concur to the conclusion that, in Pntw mccium. neither t,ubulin diversity nor post-translational modifications suffice to account for the diversity of organization and functions of the microtubule networks. Materials and methods Cells

und odture

conditions

Wild type cells of stock d4-2 of P tt’tncureiia

(701 were grown at

ham) by rarttlum-priming (Bocfrringer.