Preliminary phylogenetic analysis of the BC domain

We constructed a first preliminary tree of the biotin carboxylase (BC) domain sequences. It showed different clades corresponding to the diverse substrate specificities of each biotindependent carboxylase (data not shown). The eukaryotic ACC biotin carboxylases grouped within one clade clearly separated from the rest of the tree by a long branch, which probably reflected an increased divergence induced by the fusion of BC in a single ACC polypeptide in eukaryotes. Since these very divergent sequences could introduce noise in the phylogenetic reconstruction, we decided to withdraw the polypeptidic ACC eukaryotic sequences to recalculate a new tree. The resulting topology (see Figure A below) showed several clades that we assigned to the functions ACC, PYC, MCC, UCA and one group of PCCs using wellcharacterized enzymes as reference. This tree was used to classify the BC sequences into these functional groups according to their phylogenetic position. An important number of sequences including the acyl-CoA carboxylases, the GCC and some sequences known to carry out a PCC function but that could also carboxylate other substrates, were not specifically clustered together into unique monophyletic-monofunctional groups. They instead formed a largely unresolved paraphyletic assemblage of deep branches (black branches of the tree), from which the other well-defined functional groups branched off.