The investigation of previously little known eukaryotic lineages within and close to the opisthokonts will be key to understanding the origins of Fungi, the evolution of developmental traits in Fungi and Metazoa, and ultimately the origin(s) of multicellularity [1–3]. In particular, it will help to establish which and how many developmental genes are either shared or specific to these two major eukaryotic groups. In this context, it is essential to determine the precise phylogenetic position of candidate protists that are close to Fungi, Metazoa, or opisthokonts as a whole.
The candidate organisms choanoflagellates, ichthyosporeans and Ministeria have been convincingly shown to be relatives of Metazoa (combined in a taxon termed Holozoa; ) by using molecular phylogenetics with genomic datasets (e.g., [4–8]). Yet, there are remaining questions about the exact phylogenetic positions of Capsaspora [5, 8] and Ministeria  within Holozoa. Another, less well studied group of protists are Nucleariida, a group of heterotrophic amoeboids with radiating filopodia. Nucleariids lack distinctive morphological features that might allow associating them with either animals or fungi. Their mitochondrial cristae are either discoidal-shaped or flattened [9–11]. Indeed, initial phylogenetic analyses based on single genes have been inconsistent in placing them even within opisthokonts. There has been also confusion due to the inclusion within Nucleariida of Capsaspora owczarzaki, a species that is now excluded from this group and shown to be clearly associated with Holozoa [5, 11–17].
Overall, the phylogenetic position of the 'true' nucleariids remains controversial. In a more recent phylogenetic investigation with four nuclear gene sequences (EF-1α, HSP70, actin and β-tubulin), nucleariids associate confidently with Fungi, but only when selecting two slow-evolving chytridiomycetes . When improving the taxon sampling to 18 fungal species, the bootstrap support (BS) value for fungal monophyly drops to 85%, and alternative nucleariid positions are not rejected with the approximately unbiased (AU) test [18, 19]. In this context, it seems noteworthy that Nuclearia and fungi other than chytrids are fast-evolving, and that the rate of tubulin evolution varies strongly among species of the latter dataset (correlating to some degree with the independent loss of the flagellar apparatus in non-chytrid fungi and in Nuclearia). Together, these rate differences at the gene and species levels may increase long-branch-attraction (LBA between the two fast-evolving groups) thus causing weaker support for fungal monophyly and the nucleariid-fungal sister relationship, or predicting altogether incorrect phylogenetic relationships.
These unresolved questions served as motivation for the current phylogenetic analyses that are based on broad taxon sampling, substantially more nuclear genes (available through expressed sequence tag (EST) or complete genome projects), and comparative analyses of nuclear and mitochondrial gene datasets. To this end, we sequenced several thousand ESTs each from two Nuclearia simplex strains (probably representing separate species based on the high level of sequence divergence between them), and added them to a previous dataset  along with new genome data available from Holozoa (C. owczarzaki, Amoebidium parasiticum, Sphaeroforma arctica; ) and Fungi (Allomyces macrogynus, Batrachochytrium dendrobatidis, and Mortierella verticillata). We then sequenced the mitochondrial genome of one of the two N. simplex strains. Similar to the nuclear genomes of fungi, their mitochondrial genomes also evolve at varying rates thereby introducing a considerable potential for phylogenetic artifacts. However, phylogenetic comparisons between mitochondrial and nuclear data provide valuable, cross-wise indicators of phylogenetic artifacts as the respective evolutionary rates differ between the two genomes. For instance, such comparisons have revealed inconsistencies for the positioning of Schizosaccharomyces species within Taphrinomycotina , and of Capsaspora within Holozoa [5, 7, 8].
If the nucleariids are indeed the closest known relatives of Fungi as claimed , this protist group will provide an excellent fungal outgroup that would ultimately facilitate the settling of controversial phylogenetic placement of taxa within Fungi and/or in close neighboring groups. Among the debated issues are the monophyly and appropriate classification of the traditional fungal taxa Chytridiomycota and Zygomycota. Previous analyses based on single or a few genes have been inconsistent in answering these questions, and often lack significant support [22–31]. For example, the analyses of ribosomal RNA data supports the sister relationship between Glomeromycota and Dikarya (Ascomycota plus Basidiomycota) , while analysis of genes encoding the largest and second-largest subunits of the nuclear RNA polymerase II supports the monophyly of Zygomycota in its traditional definition .
Phylogenetic positioning of the extremely fast-evolving Microsporidia (causing strong LBA artifacts in phylogenetic analyses) is another controversial issue of great interest. In some of the most recent analyses, Microsporidia have been placed either close to zygomycetes/Mucorales [32, 33], or together with Rozella allomycis . Together with environmental sequences, Rozella species form part of a large, diverse and relatively slowly evolving lineage (designated "Rozellida"). They branch as a sister clade to Fungi [24, 34], which raises the additional question whether they should be considered to be true fungi as originally proposed . Testing the above alternative hypotheses on microsporidian affinities by phylogenomic analysis will require much more data from Rozellida (a few genes are known from Rozella allomycis, but largely insufficient for inclusion in our analyses), and from a much wider range of the paraphyletic zygomycetes. Generation of genome-size data will be further critical for applying methods that reduce LBA artifacts such as removal of fast-evolving genes or sequence sites (e.g.,  and references therein).
Despite these and various other unresolved phylogenetic issues, fungal taxonomy has been substantially redefined in a recent proposal . Chytridiomycota is still treated as a phylum, but now include only Chytridiomycetes and Monoblepharidomycetes. Other traditional chytrid lineages such as Blastocladiomycota and Neocallimastigales have been elevated to phyla based on the analyses of LSU and SSU rRNA , although support with these and other molecular markers is inconclusive. In turn, the traditional phylum Zygomycota has been altogether removed from this taxonomy , because evolutionary relationships among its members are currently unresolved and suspected to be paraphyletic. Zygomycota are now reassigned into a phylum Glomeromycota plus four subphyla incertae sedis (i.e., uncertain): Mucoromycotina, Kickxellomycotina, Zoopagomycotina and Entomophthoromycotina. To revisit these somewhat contentious issues, we compared results with mitochondrial and nuclear phylogenomic datasets, and further analyzed the effect of extending fungal species sampling, with the two N. simplex strains as the outgroup.