Grafting or pruning in the animal tree: lateral gene transfer and gene loss?

Julie C. Dunning Hotopp

Lateral gene transfer (LGT), also known as horizontal gene transfer, into multicellular eukaryotes with differentiated tissues, particularly gonads, continues to be met with skepticism by many prominent evolutionary and genomic biologists. A detailed examination of 26 animal genomes identified putative LGTs in invertebrate and vertebrate genomes, concluding that there are fewer predicted LGTs in vertebrates/chordates than invertebrates, but there is still evidence of LGT into chordates, including humans. More recently, a reanalysis of a subset of these putative LGTs into vertebrates concluded that there is not horizontal gene transfer in the human genome. One of the genes in dispute is an N-acyl-aromatic-L-amino acid amidohydrolase (ENSG00000132744), which encodes ACY3. This gene was initially identified as a putative bacteria-chordate LGT but was later debunked as it has a significant BLAST match to a more recently deposited genome of Saccoglossus kowalevskii, a flatworm, Metazoan, and hemichordate. Using BLAST searches, HMM searches, and phylogenetics to assess the evidence for LGT, gene loss, and rate variation in ACY3/ASPA homologues, the most parsimonious explanation for the distribution of ACY3/ASPA genes in eukaryotes involves both gene loss and bacteria-animal LGT, albeit LGT that occurred hundreds of millions of years ago prior to the divergence of gnathostomes. ACY3/ASPA is most likely a bacteria-animal LGT. LGTs at these time scales in the ancestors of humans are not unexpected given the many known, well-characterized, and adaptive LGTs from bacteria to insects and nematodes.