Microsporidia are obligate intracellular parasites that were thought to be
an ancient eukaryotic lineage based on molecular phylogenies using ribosoma
l RNA and translation elongation factors. However, this ancient origin of m
icrosporidia has been contested recently, as several other molecular phylog
enies suggest that microsporidia are closely related to fungi. Most of the
protein trees that place microsporidia with fungi are not well sampled, how
ever, and it is impossible to resolve whether microsporidia evolved from a
fungus or from a protistan relative of fungi. We have sequenced beta-tubuli
ns from 3 microsporidia, 4 chytrid fungi, and 12 zygomycete fungi, expandin
g the representation of beta-tubulin to include all four fungal divisions a
nd a wide diversity of microsporidia. In phylogenetic trees including these
new sequences, the overall topology of the fungal beta-tubulins generally
matched the expected relationships among the four fungal divisions, althoug
h the zygomycetes were polyphyletic in some analyses. The microsporidia con
sistently fell within this fungal diversification, and not as a sister grou
p to fungi. Overall, beta-tubulin phylogeny suggests that microsporidia evo
lved from a fungus sometime after the divergence of chytrids. We also found
that chytrid alpha- and beta-tubulins are much less divergent than are tub
ulins from other fungi or microsporidia. In trees in which the only fungal
representatives were the chytrids, microsporidia still branched with fungi
(i.e., with chytrids), suggesting that the affiliation between microsporidi
an and fungal tubulins is not an artifact of long-branch attraction.