An evaluation of elongation factor 1 alpha as a phylogenetic marker for eukaryotes.

Elongation factor 1 alpha (EF-1 alpha) is a highly conserved ubiquitous protein involved in translation that has been suggested to have desirable properties for phylogenetic inference. To examine the utility of EF-1 alpha as a phylogenetic marker for eukaryotes, we studied three properties of EF-1 alpha trees: congruency with other phyogenetic markers, the impact of species sampling, and the degree of substitutional saturation occurring between taxa. Our analyses indicate that the EF-1 alpha tree is congruent with some other molecular phylogenies in identifying both the deepest branches and some recent relationships in the eukaryotic line of descent. However, the topology of the intermediate portion of the EF-1 alpha tree, occupied by most of the protist lineages, differs for different phylogenetic methods, and bootstrap values for branches are low. Most problematic in this region is the failure of all phylogenetic methods to resolve the monophyly of two higher-order protistan taxa, the Ciliophora and the Alveolata. JACKMONO analyses indicated that the impact of species sampling on bootstrap support for most internal nodes of the eukaryotic EF-1 alpha tree is extreme. Furthermore, a comparison of observed versus inferred numbers of substitutions indicates that multiple overlapping substitutions have occurred, especially on the branch separating the Eukaryota from the Archaebacteria, suggesting that the rooting of the eukaryotic tree on the diplomonad lineage should be treated with caution. Overall, these results suggest that the phylogenies obtained from EF-1 alpha are congruent with other molecular phylogenies in recovering the monophyly of groups such as the Metazoa, Fungi, Magnoliophyta, and Euglenozoa. However, the interrelationships between these and other protist lineages are not well resolved. This lack of resolution may result from the combined effects of poor taxonomic sampling, relatively few informative positions, large numbers of overlapping substitutions that obscure phylogenetic signal, and lineage-specific rate increases in the EF-1 alpha data set. It is also consistent with the nearly simultaneous diversification of major eukaryotic lineages implied by the "big-bang" hypothesis of eukaryote evolution.

Autoantibodies to a group of centrosomal proteins in human autoimmune sera reactive with the centrosome.

OBJECTIVE: Human autoantibodies reacting with protein components of the microtubule organizing center of the cell, the centrosome, are rare and have not been extensively studied. We therefore investigated the number, type, and frequency of autoantibodies reactive with centrosomal proteins in a cohort of human sera. METHODS: To establish the type of autoantibodies found in autoimmune sera reactive with the centrosome, we used a prototype human serum, which was chosen for its intense reactivity with the centrosome throughout the cell cycle, to screen a HeLa complementary DNA (cDNA) (expression) library. Positive cDNA clones were sequenced and classified as encoding either known centrosomal autoantigens, known centrosomal proteins but unknown as human autoantigens, or previously unknown centrosomal antigens. To investigate whether these centrosomal autoantibody classes were characteristic of centrosomal-reactive sera, sera from 21 subjects with centrosomal reactivity by indirect immunofluorescence were characterized by Western blotting for reactivity to recombinant protein from each of the classes of centrosomal antigens. Clinical features were studied by retrospective chart review. RESULTS: In each of the sera, autoantibodies that recognize a group of centrosomal proteins were identified. This group included known centrosomal autoantigens (pericentrin and pericentriolar material 1 [PCM-1]), the human homolog of a known mouse centrosomal protein, ninein, which was previously unknown as a human autoantigen, and a novel centrosomal protein (Cep250). Autoantibodies to PCM-1 were the least common (8 of 21 subjects; 38%) while those to ninein, Cep250, and pericentrin occurred at roughly equal frequencies (17 subjects [81%], 17 subjects [81%], and 19 subjects [90%], respectively). There was no apparent correlation between serum autoantibody reactivity and the clinical diagnosis. CONCLUSION: Each of the autoimmune sera contained autoantibodies that reacted with a group of centrosomal proteins. We found that the centrosomal component ninein, first identified in mice, has a human homolog that is an autoantigen. Also, anticentrosomal sera contained antibodies to previously undetected centrosomal components. One of these novel antigens was identified and was designated Cep250. Thus, a characteristic of sera reactive with the centrosome is that they contain antibodies to a group of centrosomal proteins.