Three Drosophila beta-tubulin sequences: a developmentally regulated isoform (beta 3), the testis-specific isoform (beta 2), and an assembly-defective mutation of the testis-specific isoform (B2t8) reveal both an ancient divergence in metazoan isotypes and structural constraints for beta-tubulin function

Expression of ribosomal protein L22e family members in Drosophila melanogaster: rpL22-like is differentially expressed and alternatively spliced

Several ribosomal protein families contain paralogues whose roles may be equivalent or specialized to include extra-ribosomal functions. RpL22e family members rpL22 and rpL22-like are differentially expressed in Drosophila melanogaster: rpL22-like mRNA is gonad specific whereas rpL22 is expressed ubiquitously, suggesting distinctive paralogue functions. To determine if RpL22-like has a divergent role in gonads, rpL22-like expression was analysed by qRT-PCR and western blots, respectively, showing enrichment of rpL22-like mRNA and a 34 kDa (predicted) protein in testis, but not in ovary. Immunohistochemistry of the reproductive tract corroborated testis-specific expression. RpL22-like detection in 80S/polysome fractions from males establishes a role for this tissue-specific paralogue as a ribosomal component. Unpredictably, expression profiles revealed a low abundant, alternative mRNA variant (designated 'rpL22-like short') that would encode a novel protein lacking the C-terminal ribosomal protein signature but retaining part of the N-terminal domain. This variant results from splicing of a retained intron (defined by non-canonical splice sites) within rpL22-like mRNA. Polysome association and detection of a low abundant 13.5 kDa (predicted) protein in testis extracts suggests variant mRNA translation. Collectively, our data show that alternative splicing of rpL22-like generates structurally distinct protein products: ribosomal component RpL22-like and a novel protein with a role distinct from RpL22-like.

Cooperativity between the beta-tubulin carboxy tail and the body of the molecule is required for microtubule function

Using Drosophila spermatogenesis as a model, we show that function of the beta-tubulin C-terminal tail (CTT) is not independent of the body of the molecule. For optimal microtubule function, the beta-tubulin CTT and body must match. beta2 is the only beta-tubulin used in meiosis and spermatid differentiation. beta1-tubulin is used in basal bodies, but beta1 cannot replace beta2. However, when beta1 is co-expressed with beta2, both beta-tubulins are equally incorporated into all microtubules, and males exhibit near wild type fertility. In contrast, co-expression of beta2beta1C and beta1beta2C, two reciprocal chimeric molecules with bodies and tails swapped, results in defects in meiosis, cytoskeletal microtubules, and axonemes; males produce few functional sperm and few or no progeny. In these experiments, all the same beta-tubulin parts are present, but unlike the co-assembled native beta-tubulins, the "trans" configuration of the co-assembled chimeras is poorly functional. Our data thus reveal essential intra-molecular interactions between the CTT and other parts of the beta-tubulin molecule, even though the CTT is a flexible surface feature of tubulin heterodimers and microtubules. In addition, we show that Drosophila sperm tail length depends on the total tubulin pool available for axoneme assembly and spermatid elongation. D. melanogaster and other Drosophila species have extraordinarily long sperm tails, the length of which is remarkably constant in wild type flies. We show that in males of experimental genotypes that express wild type tubulins but have half the amount of the normal tubulin pool size, sperm tails are substantially shorter than wild type.

Deficiency screen identifies a novel role for beta 2 tubulin in salivary gland and myoblast migration in the Drosophila embryo

The Drosophila embryonic salivary gland is an epithelial organ formed by the coordinated invagination and migration of primordial cells. To identify genes that regulate gland migration we performed a deficiency screen of the third chromosome. Here, we report on the analysis of the beta 2 tubulin isoform (beta2t) that maps at 85D15. We show that, in beta2t mutant embryos, salivary glands did not complete their posterior migration and that migration of fusion competent myoblasts and longitudinal visceral muscle founder cells between the gland and circular visceral mesoderm was delayed. We also demonstrate that gland migration defects correlate with reduced betaPS and alphaPS2 integrin expression in the surrounding mesoderm and that beta2t genetically interacts with genes encoding integrin alphaPS1 and alphaPS2 subunits. Our studies reveal for the first time that beta2t is expressed in embryogenesis and that beta2t plays an important role in salivary gland and myoblast migration, possibly through proper regulation of integrin adhesion proteins.

Axoneme beta-tubulin sequence determines attachment of outer dynein arms

Axonemes of motile eukaryotic cilia and flagella have a conserved structure of nine doublet microtubules surrounding a central pair of microtubules. Outer and inner dynein arms on the doublets mediate axoneme motility [1]. Outer dynein arms (ODAs) attach to the doublets at specific interfaces [2-5]. However, the molecular contacts of ODA-associated proteins with tubulins of the doublet microtubules are not known. We report here that attachment of ODAs requires glycine 56 in the beta-tubulin internal variable region (IVR). We show that in Drosophila spermatogenesis, a single amino acid change at this position results in sperm axonemes markedly deficient in ODAs. Moreover, we found that axonemal beta-tubulins throughout the phylogeny have invariant glycine 56 and a strongly conserved IVR, whereas nonaxonemal beta-tubulins vary widely in IVR sequences. Our data reveal a deeply conserved physical requirement for assembly of the macromolecular architecture of the motile axoneme. Amino acid 56 projects into the microtubule lumen [6]. Imaging studies of axonemes indicate that several proteins may interact with the doublet-microtubule lumen [3, 4, 7, 8]. This region of beta-tubulin may determine the conformation necessary for correct attachment of ODAs, or there may be sequence-specific interaction between beta-tubulin and a protein involved in ODA attachment or stabilization.

Structurally similar Drosophila alpha-tubulins are functionally distinct in vivo

We used transgenic analysis in Drosophila to compare the ability of two structurally similar alpha-tubulin isoforms to support microtubule assembly in vivo. Our data revealed that even closely related alpha-tubulin isoforms have different functional capacities. Thus, in multicellular organisms, even small changes in tubulin structure may have important consequences for regulation of the microtubule cytoskeleton. In spermatogenesis, all microtubule functions in the postmitotic male germ cells are carried out by a single tubulin heterodimer composed of the major Drosophila alpha-84B tubulin isoform and the testis-specific beta 2-tubulin isoform. We tested the ability of the developmentally regulated alpha 85E-tubulin isoform to replace alpha 84B in spermatogenesis. Even though it is 98% similar in sequence, alpha 85E is not functionally equivalent to alpha 84B. alpha 85E can support some functional microtubules in the male germ cells, but alpha 85E causes dominant male sterility if it makes up more than one-half of the total alpha-tubulin pool in the spermatids. alpha 85E does not disrupt meiotic spindle or cytoplasmic microtubules but causes defects in morphogenesis of the two classes of singlet microtubules in the sperm tail axoneme, the central pair and the accessory microtubules. Axonemal defects caused by alpha 85E are precisely reciprocal to dominant defects in doublet microtubules we observed in a previous study of ectopic germ-line expression of the developmentally regulated beta 3-tubulin isoform. These data demonstrate that the doublet and singlet axoneme microtubules have different requirements for alpha- and beta-tubulin structure. In their normal sites of expression, alpha 85E and beta 3 are coexpressed during differentiation of several somatic cell types, suggesting that alpha 85E and beta 3 might form a specialized heterodimer. Our tests of different alpha-beta pairs in spermatogenesis did not support this model. We conclude that if alpha 85E and beta 3 have specialized properties required for their normal functions, they act independently to modulate the properties of microtubules into which they are incorporated.

Expression of beta 1 tubulin (beta Tub56D) in Drosophila testis stem cells is regulated by a short upstream sequence while intron elements guide expression in somatic cells

Stem cell differentiation to mature spermatozoa is a morphogenetic process that is highly dependent on microtubular arrays. In the early, mitotically active stages of spermatogenesis, only the beta 1 tubulin isotype is expressed. Analysis of transgenic flies containing beta 1-lacZ gene fusions revealed that this expression is regulated by sequences located between positions -45 and -191 upstream of the transcription initiation site. Furthermore, beta 1 tubulin is a major component of cyst cells. Expression in these cells is driven by enhancer elements located in the beta 1 tubulin gene intron. These enhancer elements also guide expression in combination with the hsp70 basal promoter. In addition, redundant enhancer elements in the intron drive expression in the testis wall. Our data show that within a single tissue, the male gonad, expression of the beta 1 tubulin gene is under cell-type-specific control mediated by independent cis-acting elements. Therefore in the germ line, control of beta 1 tubulin expression is strictly governed by promoter-proximal elements, while for the somatic parts of the testis, enhancer elements confer less stringent expression control.

Either alpha-tubulin isogene product is sufficient for microtubule function during all stages of growth and differentiation in Aspergillus nidulans

The filamentous fungus Aspergillus nidulans has two genes encoding alpha-tubulin, tubA and tubB, which are differentially required at distinct stages during the life cycle. The tubA gene is required during vegetative growth for mitosis and nuclear migration (B. R. Oakley, C. E. Oakley, and J. E. Rinehart, Mol. Gen. Genet. 208:135-144, 1987; P. Doshi, C. A. Bossie, J. H. Doonan, G. S. May, and N. R. Morris, Mol. Gen. Genet. 225:129-141, 1991). The tubB gene is not required for any detectable aspect of vegetative growth or asexual reproduction but is essential during sexual development prior to the first meiotic division (K. E. Kirk and N. R. Morris, Genes Dev. 5:2014-2023, 1991). In this study, we determined whether the role of each alpha-tubulin gene is to provide a specific isotype necessary for a particular microtubule function or whether either alpha-tubulin isotype, if present in sufficient quantities, can participate effectively in all types of microtubule. Strains carrying a deletion allele of tubB (tubB delta) produce no ascospores from a cross. When one copy of a plasmid containing the region upstream of the tubB gene fused to the tubA coding region was integrated into a tubB delta strain, ascosporogenesis proceeded beyond the tubB delta block and resulted in the formation of sexual spores. However, irregular numbers of spores formed in some asci during development, and the ascospores had greatly diminished viability and aberrant morphologies. These defects were nearly corrected when two additional copies of the tubA coding region were integrated into the tubB delta strain. These results indicate that the tubA alpha-tubulin isotype can form functional microtubules during sexual development in the absence of tubB protein. In a reciprocal set of experiments, we examined whether upregulation of tubB can complement the tubA4 mutation, which causes supersensitivity to benomyl during vegetative growth. When tubA4 strains integrated a plasmid containing an alcohol-inducible promoter joined to the tubB coding region and subsequently overexpressed the tubB isotype, the benomyl supersensitivity normally caused by the tubA4 allele was relieved. These results indicate that when enough tubB alpha-tubulin is supplied, strains lacking functional tubA isotype can still form microtubules which effectively carry out mitosis and nuclear migration.

Either alpha-tubulin isogene product is sufficient for microtubule function during all stages of growth and differentiation in Aspergillus nidulans

The filamentous fungus Aspergillus nidulans has two genes encoding alpha-tubulin, tubA and tubB, which are differentially required at distinct stages during the life cycle. The tubA gene is required during vegetative growth for mitosis and nuclear migration (B. R. Oakley, C. E. Oakley, and J. E. Rinehart, Mol. Gen. Genet. 208:135-144, 1987; P. Doshi, C. A. Bossie, J. H. Doonan, G. S. May, and N. R. Morris, Mol. Gen. Genet. 225:129-141, 1991). The tubB gene is not required for any detectable aspect of vegetative growth or asexual reproduction but is essential during sexual development prior to the first meiotic division (K. E. Kirk and N. R. Morris, Genes Dev. 5:2014-2023, 1991). In this study, we determined whether the role of each alpha-tubulin gene is to provide a specific isotype necessary for a particular microtubule function or whether either alpha-tubulin isotype, if present in sufficient quantities, can participate effectively in all types of microtubule. Strains carrying a deletion allele of tubB (tubB delta) produce no ascospores from a cross. When one copy of a plasmid containing the region upstream of the tubB gene fused to the tubA coding region was integrated into a tubB delta strain, ascosporogenesis proceeded beyond the tubB delta block and resulted in the formation of sexual spores. However, irregular numbers of spores formed in some asci during development, and the ascospores had greatly diminished viability and aberrant morphologies. These defects were nearly corrected when two additional copies of the tubA coding region were integrated into the tubB delta strain. These results indicate that the tubA alpha-tubulin isotype can form functional microtubules during sexual development in the absence of tubB protein. In a reciprocal set of experiments, we examined whether upregulation of tubB can complement the tubA4 mutation, which causes supersensitivity to benomyl during vegetative growth. When tubA4 strains integrated a plasmid containing an alcohol-inducible promoter joined to the tubB coding region and subsequently overexpressed the tubB isotype, the benomyl supersensitivity normally caused by the tubA4 allele was relieved. These results indicate that when enough tubB alpha-tubulin is supplied, strains lacking functional tubA isotype can still form microtubules which effectively carry out mitosis and nuclear migration.

Microtubule behavior during guidance of pioneer neuron growth cones in situ

The growth of an axon toward its target results from the reorganization of the cytoskeleton in response to environmental guidance cues. Recently developed imaging technology makes it possible to address the effect of such cues on the neural cytoskeleton directly. Although high resolution studies can be carried out on neurons in vitro, these circumstances do not recreate the complexity of the natural environment. We report here on the arrangement and dynamics of microtubules in live neurons pathfinding in response to natural guidance cues in situ using the embryonic grasshopper limb fillet preparation. A rich microtubule network was present within the body of the growth cone and normally extended into the distal growth cone margin. Complex microtubule loops often formed transiently within the growth cone. Branches both with and without microtubules were regularly observed. Microtubules did not extend into filopodia. During growth cone steering events in response to identified guidance cues, microtubule behaviour could be monitored. In turns towards guidepost cells, microtubules selectively invaded branches derived from filopodia that had contacted the guidepost cell. At limb segment boundaries, microtubules displayed a variety of behaviors, including selective branch invasion, and also invasion of multiple branches followed by selective retention in branches oriented in the correct direction. Microtubule invasion of multiple branches also was seen in growth cones migrating on intrasegmental epithelium. Both selective invasion and selective retention generate asymmetrical microtubule arrangements within the growth cone, and may play a key role in growth cone steering events.

A codon change in beta-tubulin which drastically affects microtubule structure in Drosophila melanogaster fails to produce a significant phenotype in Saccharomyces cerevisiae

The relative uniformity of microtubule ultrastructure in almost all eukaryotic cells is thought to be a consequence of the conserved elements of tubulin sequence. In support of this idea, a mutation in a beta-tubulin gene of Drosophila melanogaster, occurring at a highly conserved position, produces U-shaped microtubules, suggesting a defect in either nucleation or packing during assembly (M. T. Fuller, J. H. Caulton, J. A. Hutchens, T. C. Kaufman, and E. C. Raff, J. Cell Biol. 104:385-394, 1987, and J. E. Rudolph, M. Kimble, H. D. Hoyle, M. A. Subler, and E. C. Raff, Mol. Cell. Biol. 7:2231-2242, 1987). Surprisingly, we find that introducing the same mutation into the sole beta-tubulin gene of Saccharomyces cerevisiae has virtually no consequences for microtubule structure or function in that organism.

A codon change in beta-tubulin which drastically affects microtubule structure in Drosophila melanogaster fails to produce a significant phenotype in Saccharomyces cerevisiae

The relative uniformity of microtubule ultrastructure in almost all eukaryotic cells is thought to be a consequence of the conserved elements of tubulin sequence. In support of this idea, a mutation in a beta-tubulin gene of Drosophila melanogaster, occurring at a highly conserved position, produces U-shaped microtubules, suggesting a defect in either nucleation or packing during assembly (M. T. Fuller, J. H. Caulton, J. A. Hutchens, T. C. Kaufman, and E. C. Raff, J. Cell Biol. 104:385-394, 1987, and J. E. Rudolph, M. Kimble, H. D. Hoyle, M. A. Subler, and E. C. Raff, Mol. Cell. Biol. 7:2231-2242, 1987). Surprisingly, we find that introducing the same mutation into the sole beta-tubulin gene of Saccharomyces cerevisiae has virtually no consequences for microtubule structure or function in that organism.

Two Drosophila beta tubulin isoforms are not functionally equivalent

We have tested the functional capacity of different beta tubulin isoforms in vivo by expressing beta 3-tubulin either in place of or in addition to beta 2-tubulin in the male germ line of Drosophila melanogaster. The testes-specific isoform, beta 2, is conserved relative to major metazoan beta tubulins, while the developmentally regulated isoform, beta 3, is considerably divergent in sequence. beta 3-tubulin is normally expressed in discrete subsets of cells at specific times during development, but is not expressed in the male germ line. beta 2-Tubulin is normally expressed only in the postmitotic germ cells of the testis, and is required for all microtubule-based functions in these cells. The normal functions of beta 2-tubulin include assembly of meiotic spindles, axonemes, and at least two classes of cytoplasmic microtubules, including those associated with the differentiating mitochondrial derivatives. A hybrid gene was constructed in which 5' sequences from the beta 2 gene were joined to protein coding and 3' sequences of the beta 3 gene. Drosophila transformed with the hybrid gene express beta 3-tubulin in the postmitotic male germ cells. When expressed in the absence of the normal testis isoform, beta 3-tubulin supports assembly of one class of functional cytoplasmic microtubules. In such males the microtubules associated with the membranes of the mitochondrial derivatives are assembled and normal mitochondrial derivative elongation occurs, but axoneme assembly and other microtubule-mediated processes, including meiosis and nuclear shaping, do not occur. These data show that beta 3 tubulin can support only a subset of the multiple functions normally performed by beta 2, and also suggest that the microtubules associated with the mitochondrial derivatives mediate their elongation. When beta 3 is coexpressed in the male germ line with beta 2, at any level, spindles and all classes of cytoplasmic microtubules are assembled and function normally. However, when beta 3-tubulin exceeds 20% of the total testis beta tubulin pool, it acts in a dominant way to disrupt normal axoneme assembly. In the axonemes assembled in such males, the doublet tubules acquire some of the morphological characteristics of the singlet microtubules of the central pair and accessory tubules. These data therefore unambiguously demonstrate that the Drosophila beta tubulin isoforms beta 2 and beta 3 are not equivalent in intrinsic functional capacity, and furthermore show that assembly of the doublet tubules of the axoneme imposes different constraints on beta tubulin function than does assembly of singlet microtubules.

Genetic and molecular analysis of a Caenorhabditis elegans beta-tubulin that conveys benzimidazole sensitivity

Benzimidazole anti-microtubule drugs, such as benomyl, induce paralysis and slow the growth of the nematode Caenorhabditis elegans. We have identified 28 mutations in C. elegans that confer resistance to benzimidazoles. All resistant mutations map to a single locus, ben-1. Virtually all these mutations are genetically dominant. Molecular cloning and DNA sequence analysis established that ben-1 encodes a beta-tubulin. Some resistant mutants are completely deleted for the ben-1 gene. Since the deletion strains appear to be fully resistant to the drugs, the ben-1 product appears to be the only benzimidazole-sensitive beta-tubulin in C. elegans. Furthermore, since animals lacking ben-1 are viable and coordinated, the ben-1 beta-tubulin appears to be nonessential for growth and movement. The ben-1 function is likely to be redundant in the nematode genome.

The major component of the paraflagellar rod of Trypanosoma brucei is a helical protein that is encoded by two identical, tandemly linked genes

The flagellum of the parasitic hemoflagellate Trypanosoma brucei contains two major structures: (a) the microtubule axoneme, and (b) a highly ordered, filamentous array, the paraflagellar rod (PFR). This is a complex, three-dimensional structure, of yet unknown function, that extends along most of the axoneme and is closely linked to it. Its major structural component is a single protein of 600 amino acids. This PFR protein can assume two different conformations, resulting in two distinct bands of apparent molecular masses of 73 and 69 kD in SDS-gel electrophoresis. Secondary structure predictions indicate a very high helix content. Despite its biochemical similarity to the intermediate filament proteins (solubility properties, amino acid composition, and high degree of helicity), the PFR protein does not belong in this class of cytoskeletal proteins. The PFR protein is coded for by two tandemly linked genes of identical nucleotide sequence. Both genes are transcribed into stable mRNAs of very similar length that carry the mini-exon sequence at their 5' termini.

Mutations that encode partially functional beta 2 tubulin subunits have different effects on structurally different microtubule arrays

The testis-specific beta 2 tubulin of Drosophila is required for assembly and function of at least three architecturally different microtubule arrays (Kemphues et al., 1982). Two recessive male-sterile mutations in the B2t locus that encode partially functional, stable, variant forms of beta 2 tubulin cause defects in only certain microtubule-based processes during spermatogenesis. These mutations could thus identify aspects of beta tubulin primary structure critical for function only in specific microtubule arrays. In males carrying the B2t6 mutation, meiotic chromosome segregation and nuclear shaping are normal and flagellar axonemes are formed, but there is a subtle defect in axoneme structure; the outer doublet microtubules fill in with a central core normally seen only in the central pair and accessory microtubules. In homozygous B2t7 males, chromosome movement is usually normal during meiosis but cytokinesis often fails, cytoplasmic microtubules are assembled and nuclear shaping appears to be normal, but the flagellar axoneme lacks structural integrity. In contrast, the B2t8 allele affects a general property of tubulin, the ability to form normal side-to-side association of protofilaments (Fuller et al., 1987), and causes defects in meiosis, axoneme assembly and nuclear shaping. Certain combinations of these beta 2 tubulin mutations show interallelic complementation; in B2t6/B2t8 males functional sperm are produced and both variant subunits are incorporated into mature sperm, in the absence of wild-type beta 2 tubulin. Comparison of the phenotypes of the three partially functional beta 2 tubulin alleles reveals some aspects of tubulin primary structure more important for function in specific subsets of microtubule arrays, and other aspects required for the construction of microtubules in general.

A gene encoding the major beta tubulin of the mitotic spindle in Physarum polycephalum plasmodia

The multinucleate plasmodium of Physarum polycephalum is unusual among eucaryotic cells in that it uses tubulins only in mitotic-spindle microtubules; cytoskeletal, flagellar, and centriolar microtubules are absent in this cell type. We have identified a beta-tubulin cDNA clone, beta 105, which is shown to correspond to the transcript of the betC beta-tubulin locus and to encode beta 2 tubulin, the beta tubulin expressed specifically in the plasmodium and used exclusively in the mitotic spindle. Physarum amoebae utilize tubulins in the cytoskeleton, centrioles, and flagella, in addition to the mitotic spindle. Sequence analysis shows that beta 2 tubulin is only 83% identical to the two beta tubulins expressed in amoebae. This compares with 70 to 83% identity between Physarum beta 2 tubulin and the beta tubulins of yeasts, fungi, alga, trypanosome, fruit fly, chicken, and mouse. On the other hand, Physarum beta 2 tubulin is no more similar to, for example, Aspergillus beta tubulins than it is to those of Drosophila melanogaster or mammals. Several eucaryotes express at least one widely diverged beta tubulin as well as one or more beta tubulins that conform more closely to a consensus beta-tubulin sequence. We suggest that beta-tubulins diverge more when their expression pattern is restricted, especially when this restriction results in their use in fewer functions. This divergence among beta tubulins could have resulted through neutral drift. For example, exclusive use of Physarum beta 2 tubulin in the spindle may have allowed more amino acid substitutions than would be functionally tolerable in the beta tubulins that are utilized in multiple microtubular organelles. Alternatively, restricted use of beta tubulins may allow positive selection to operate more freely to refine beta-tubulin function.

A gene encoding the major beta tubulin of the mitotic spindle in Physarum polycephalum plasmodia

The multinucleate plasmodium of Physarum polycephalum is unusual among eucaryotic cells in that it uses tubulins only in mitotic-spindle microtubules; cytoskeletal, flagellar, and centriolar microtubules are absent in this cell type. We have identified a beta-tubulin cDNA clone, beta 105, which is shown to correspond to the transcript of the betC beta-tubulin locus and to encode beta 2 tubulin, the beta tubulin expressed specifically in the plasmodium and used exclusively in the mitotic spindle. Physarum amoebae utilize tubulins in the cytoskeleton, centrioles, and flagella, in addition to the mitotic spindle. Sequence analysis shows that beta 2 tubulin is only 83% identical to the two beta tubulins expressed in amoebae. This compares with 70 to 83% identity between Physarum beta 2 tubulin and the beta tubulins of yeasts, fungi, alga, trypanosome, fruit fly, chicken, and mouse. On the other hand, Physarum beta 2 tubulin is no more similar to, for example, Aspergillus beta tubulins than it is to those of Drosophila melanogaster or mammals. Several eucaryotes express at least one widely diverged beta tubulin as well as one or more beta tubulins that conform more closely to a consensus beta-tubulin sequence. We suggest that beta-tubulins diverge more when their expression pattern is restricted, especially when this restriction results in their use in fewer functions. This divergence among beta tubulins could have resulted through neutral drift. For example, exclusive use of Physarum beta 2 tubulin in the spindle may have allowed more amino acid substitutions than would be functionally tolerable in the beta tubulins that are utilized in multiple microtubular organelles. Alternatively, restricted use of beta tubulins may allow positive selection to operate more freely to refine beta-tubulin function.