Although the precision of centriole duplication has aroused interest for over a century, its molecular mechanism remains almost entirely unknown. What proteins are required for a pre-existing centriole to induce assembly of a new centriole? One strong candidate is centrin, an EF hand-containing calcium binding protein first identified biochemically in green algae. Analysis of centrin mutants in the haploid unicellular alga Chlamydomonas (also known as'green yeast' because of its powerful genetics) suggests that centrin is required for proper centriole segregation, and that defects in centrin lead to defects in centriole duplication. However, these studies leave several important questions about centrin function unanswered. First, centriole duplication still occurs in the Chlamydomonas centrin mutant, albeit to a reduced extent. Is this because the mutation, which was not a null, retains some residual centrin function, or because centrin is not strictly required for centriole duplication? Second, although the centrioles of Chlamydomonas are virtually identical to those of animal cells, is it possible that centrin might play a different role in animals?To resolve these questions, Jeffrey Salisbury [1xCentrin-2 is required for centriole duplication in mammalian cells. Salisbury, J.L. et al. Curr. Biol. 2002; 12: 1287–1292Abstract | Full Text | Full Text PDF | PubMed | Scopus (202)See all References[1] (who, by the way, was the first to discover centrin in algae) and colleagues have used RNA interference (RNAi) to reduce the levels of centrin-2 (the centriole-specific centrin isoform) in human tissue culture cells. RNAi of centrin-2 in HeLa cells leads to progressive loss of centrioles, consistent with a complete block in centriole replication. This result confirms that centrin is involved in centriole duplication in animal cells, which was suggested in previous work with algae. Moreover, these results show that a significant reduction in centrin levels can, in fact, cause a complete block in centriole duplication. This implies that the requirement for centrin is absolute.Interestingly, although centriole duplication is blocked, cell division continues and bipolar spindles are still observed, even when centrioles are absent. However, these acentriolar spindles are abnormal, with broad, non-focused poles. In addition, multinucleate cells are observed. These results are consistent with a function for centrioles in both spindle assembly and cytokinesis.These results indicate that centrin plays a conserved role in centriole duplication in eukaryotes ranging from algae to animals. The yeast homolog of centrin, CDC31, is a component of the half-bridge structure involved in spindle-pole body duplication. Thus, these results suggest that centriole duplication might involve a structure similar to the half-bridge in yeast.