The genomic environment around the Aromatase gene: evolutionary insights

Abstract

Background: The cytochrome P450 aromatase (CYP19), catalyses the aromatisation of androgens to estrogens, a key mechanism in vertebrate reproductive physiology. A current evolutionary hypothesis suggests that CYP19 gene arose at the origin of vertebrates, given that it has not been found outside this clade. The human CYP19 gene is located in one of the proposed MHC-paralogon regions (HSA15q). At present it is unclear whether this genomic location is ancestral (which would suggest an invertebrate origin for CYP19) or derived (genomic location with no evolutionary meaning). The distinction between these possibilities should help to clarify the timing of the CYP19 emergence and which taxa should be investigated.

Results: Here we determine the "genomic environment" around CYP19 in three vertebrate species Homo sapiens, Tetraodon nigroviridis and Xenopus tropicalis. Paralogy studies and phylogenetic analysis of six gene families suggests that the CYP19 gene region was structured through "en bloc" genomic duplication (as part of the MHC-paralogon formation). Four gene families have specifically duplicated in the vertebrate lineage. Moreover, the mapping location of the different paralogues is consistent with a model of "en bloc" duplication. Furthermore, we also determine that this region has retained the same gene content since the divergence of Actinopterygii and Tetrapods. A single inversion in gene order has taken place, probably in the mammalian lineage. Finally, we describe the first invertebrate CYP19 sequence, from Branchiostoma floridae.

Conclusion: Contrary to previous suggestions, our data indicates an invertebrate origin for the aromatase gene, given the striking conservation pattern in both gene order and gene content, and the presence of aromatase in amphioxus. We propose that CYP19 duplicated in the vertebrate lineage to yield four paralogues, followed by the subsequent loss of all but one gene in vertebrate evolution. Finally, we suggest that agnathans and lophotrocozoan protostomes should be investigated for the presence of aromatase.

Figure 1

The MHC-paralogon. Two sets of paralogy regions are intact (A and B) while the remaining 2 are broken (C and D) (adapted from [20]). The map location of the MHC, CYP19 and the surrounding genes (and paralogues) is shown. In parentheses the genomic distance in megabases to the p telomere.

Figure 2

Physical maps of the genomic environment around CYP19 in Homo sapiens, Tetraodon nigroviridis and Xenopus tropicalis.

Figure 3

Neighbor-joining phylogenetic trees from alignment of the putative protein sequences of AP4E1 (A), SSC-2S (B), COL (C), DMXL (D), PPBP (E) and TMOD (F). Figures at nodes are scores from 1000 bootstrap resamplings of the data. Arrow denotes duplication timing. Ag – Anopheles gambiae; Ce – Caenorhabditis elegans; Dm – Drosophila melanogaster; Ci – Ciona intestinalis; Hs – Homo sapeins; Mm – Mus musculus; Gg – Gallus gallus; Xt – Xenopus tropicalis; Fr – Fugu rubripes; Tn – Tetraodon nigroviridis.

Figure 4

Evolutionary model for the origin of the aromatase gene family. Symbols from the two conserved DNA blocks from figure 2. Dotted line boxes denote conservation of synteny but not gene order. Black bars denote "en bloc" (or genome) duplications. Star indicates tandem duplication of TMOD2/3. Horizontal curve arrow indicates gene inversion of the DNA block containing TMOD3, TMOD2, PPBP and SCGIII.

Figure 5

Nucleotide and predicted amino acid sequence of the AmphiCYP19 (partial sequence) (A); alignment of CYP19 sequences; dashes denote insertions; black line, I-helix region; red line, Ozol's peptide region; green line, Aromatic region and blue line, Heme binding region. Bf, B. floridae, Ds, D. sabina, Am, Alligator mississippiensis, Gg, G. Gallus, Hs, H. sapiens, Mm, M. musculus, Xt, X. tropicalis, Tn, T. nigroviridis, Dr, D. rerio, Fr, Fugu rubripes (B); Neighbor-joining phylogenetic tree from the alignment of the putative protein sequences of CYP19 genes, figures at nodes are scores from 1000 bootstrap resamplings of the data ; an insertion of the TnCYP19b predicted protein sequence was kept out of the alignment (C).