New tubulins in protozoal parasites

More than 20 years ago, biochemical analysis of the eukaryotic cell cytoskeleton revealed the major component proteins. The heterodimeric (α/β) protein tubulin was defined as the building block of microtubules, assembled in a polar manner into specifically arranged protofilaments in the microtubule wall [1]. The next two members of the tubulin protein superfamily were both discovered by genetic means — γ tubulin in Aspergillus [2] and δ tubulin in Chlamydomonas [3]. The γ tubulin is essential for microtubule function and is located in centroso mes and other microtubule-organising centres [4]. The δ tubulin is encoded by the UNI3 gene in Chlamydomonas and a uni3-1 mutation resulted in flagellar basal bodies that possess doublet rather than triplet microtubules [3]. These four members of the tubulin superfamily can be characterised by their distinct intracellular locations and expression patterns, which are reflected in unique sequence characteristics. The large number of tubulin sequences available in current databases, coupled with the considerable divergence of those sequences, complicates the task of reliable identification and characterisation of tubulin family members. During the Saccharomyces cerevisiae genome project, sequencing revealed the presence of a tubulin gene that was only around 30% identical to the yeast α and β tubulins. This Tub4 protein was conjectured to be a novel tubulin rather than an α, β or γ tubulin [5]. However, subsequent analysis of the completed S. cerevisiae genome and molecular and biochemical studies have led to an accepted view that Tub4 is the budding yeast γ tubulin [4]. Consequently, it has been suggested that caution is required in using certain types of sequence analysis methods to classify novel tubulin sequences [6]. Within the genome of the protozoan parasite Trypanosoma brucei, the genes for α and β tubulin exist as a cluster of repeated α/β pairs [7]. Recently, we identified the single γ tubulin gene in T. brucei [8]. We then conducted a search by PCR and other means for the presence of the T. brucei homologue of δ tubulin. To our surprise, after we cloned the T. brucei δ tubulin homologue, we also identified two new divergent tubulin-like sequences. The general features can be readily visualised in the automatically generated alignment illustrated in Figure 1. Both of these new sequences are also present within the T. brucei genome project databases at the Sanger Centre and TIGR as partial or complete sequences (see Figure 1). Although it was possible to conclude …