Hsp70 sequences indicate that choanoflagellates are closely related to animals.

Over 130 years ago, James-Clark noted a remarkable structural similarity between the feeding cells of sponges (choanocytes) and a group of free-living protists, the choanoflagellates. Both cell types possess a single flagellum surrounded by a collar of fine tentacles. The similarity led to the hypothesis that sponges, and, by implication, other animals, evolved from choanoflagellate-like ancestors. Phylogenetic analysis of ribosomal DNA neither supports nor refutes this hypothesis. Here, we report the sequence of an hsp70 gene and pseudogene from the freshwater choanoflagellate Monosiga ovata. These represent the first nuclear-encoded protein-coding sequences reported for any choanoflagellate. We find that Monosiga and most bilaterian hsp70 genes have high GC contents that may distort phylogenetic tree construction; therefore, protein sequences were used for phylogenetic reconstruction. Our analyses indicate that Monosiga is more closely related to animals than to fungi. We infer that animals and at least some choanoflagellates are part of a clade that excludes the fungi. This is consistent with the origin of animals from a choanoflagellate-like ancestor.

Genomic organization of the Hoxa4-Hoxa10 region from Morone saxatilis: implications for Hox gene evolution among vertebrates.

The physical mapping of Hox gene clusters from a limited number of vertebrates has shown an overall conservation in gene organization in which major evolutionary changes appear to be primarily restricted to the deletion of one or more genes, with the exception of the amplification of additional clusters as postulated from zebrafish. We have sequenced a 31 kb region of the HoxA cluster from the teleost Morone saxatilis (striped bass), both to provide a detailed physical map of this region and to better understand the nature of Hox cluster evolution among vertebrate taxa. We identified five linked Hox genes: Hoxa4, Hoxa5, Hoxa7, Hoxa9, and Hoxa10, which are organized similarly to those of other vertebrates. Furthermore, we have documented the absence of the Hoxa6 and Hoxa8 genes within the 31 kb contig. Comparison of our results to those published for other vertebrates suggests that the absence of Hoxa6 is a common characteristic of teleosts, whereas the absence of Hoxa8 is common to vertebrates in general, with the possible exception of zebrafish. Further comparisons between the HoxA genes from Morone with those from the pufferfish, Fugu rubripes, revealed the likely presence of a previously unreported Hoxa7 gene, or gene fragment, in the Fugu genome, which suggests that the Hoxa7 gene, unlike Hoxa6 or Hoxa8, is present in teleosts. In addition to these differences in vertebrate Hox cluster structure, we also observed a marked reduction in the length of the Hoxa4--a10 region between vertebrate lineages representative of teleosts and mammals. Comparative analysis of HoxA cluster organization among teleosts and mammals suggests that cluster length reduction and lineage-specific gene loss events are hallmarks of Hox cluster evolution.