The cytochrome P450 genesis locus: the origin and evolution of animal cytochrome P450s

Abstract

The neighbourhoods of cytochrome P450 (CYP) genes in deuterostome genomes, as well as those of the cnidarians Nematostella vectensis and Acropora digitifera and the placozoan Trichoplax adhaerens were examined to find clues concerning the evolution of CYP genes in animals. CYP genes created by the 2R whole genome duplications in chordates have been identified. Both microsynteny and macrosynteny were used to identify genes that coexisted near CYP genes in the animal ancestor. We show that all 11 CYP clans began in a common gene environment. The evidence implies the existence of a single locus, which we term the 'cytochrome P450 genesis locus', where one progenitor CYP gene duplicated to create a tandem set of genes that were precursors of the 11 animal CYP clans: CYP Clans 2, 3, 4, 7, 19, 20, 26, 46, 51, 74 and mitochondrial. These early CYP genes existed side by side before the origin of cnidarians, possibly with a few additional genes interspersed. The Hox gene cluster, WNT genes, an NK gene cluster and at least one ARF gene were close neighbours to this original CYP locus. According to this evolutionary scenario, the CYP74 clan originated from animals and not from land plants nor from a common ancestor of plants and animals. The CYP7 and CYP19 families that are chordate-specific belong to CYP clans that seem to have originated in the CYP genesis locus as well, even though this requires many gene losses to explain their current distribution. The approach to uncovering the CYP genesis locus overcomes confounding effects because of gene conversion, sequence divergence, gene birth and death, and opens the way to understanding the biodiversity of CYP genes, families and subfamilies, which in animals has been obscured by more than 600 Myr of evolution.

Figure 1.

Metazoan phylogeny showing the 2R and 3R WGD events and total number of CYP coding genes in each genome. This phylogeny is based on the best current knowledge of metazoan evolution [34]. Gene counts were determined based on protein predictions from the respective genome projects (see §2). Phylogenetic divisions are based on currently accepted consensus; where no consensus exists a polytomy is displayed. 525 Ma is the minimum age for vertebrates based on the vertebrate fossils in the Chengjiang biota. Other estimated divergence times are from Edgecombe et al. [36].

Figure 2.

Maximum-likelihood phylogeny of CYP protein sequences. Branches represent individual sequences, and are coloured according to species or lineage (e.g. Insecta, Vertebrata). The tree is rooted with CYP51, and all CYP51 genes in various genomes cluster at or near the root. Large highlighted blocks indicate the major clans—Clan 2 (pink), Clan 3 (blue), Clan 4 (green) and Mito Clan (yellow). While the clan ordering is robust to different alignments, the relative position of specific subclans and individual families is sensitive to alignment and does not display high bootstrap values (not shown).

Figure 3.

Synteny mapping with emphasis on ZDHHC12, CRAT and DOLPP1 and their neighbours. These three genes maintain syntenic relationships from Acropora to human. CYP Clan 74, mito Clan and Clan 2 P450s are linked to these genes and/or their neighbours. Connecting lines are colour-coded by species and they indicate a neighbour relationship, though not all genes are shown.

Figure 4.

Distribution of CYP clans in animals and fungi. Presence of a clan is indicated by a closed circle. Absence is indicated by an open circle. The asterisk denotes a probable lateral gene transfer from animals to the filamentous fungi. Some clan losses are evident, particularly in insects and Crustacea (Ecdysozoa), which are known to lack CYP51 and thus require dietary cholesterol. Note also that CYP Clan 19 does not appear until cephalochordates.

Figure 5.

All 11 CYP clans can be linked directly or indirectly to Hox gene clusters. P450 genes are boxed. Hox genes are aligned down the centre with chr and Mb locations given. Non-Hox regions are labelled as non-Hox. Orthologues and ohnologues are aligned as much as possible to highlight sytenic relationships as seen with EXOC6 in lines M and N. Not all genes in the region are shown because of space limitations.

Figure 6.

Genes in the vicinity of human CYP2W1, CYP2D6 and chicken CYP2W1. The human CYP2W1 region is telomeric. Connecting lines between (b) and (c) mark ohnologues. Not all genes from the regions are included.

Figure 7.

Human paralogons covering the CYP2D6 and CYP2W1 regions, including two CYP3 pseudogenes (arrows mark P450s). Probable ohnologues are connected by lines. The regions shown are boxed in red in the ideograms.

Figure 8.

Direct linkage between different CYP clans including Hox clusters (A–D). Each bar indicates a neighbour relationship between the clan members listed on the top of the figure, in the species listed on the left.

Figure 9.

Synteny mapping of CYP16 family members showing linkages to CYP26 family members, the CYP Clan 46 and the CYP Clan 2 via IDE and DICER1 and to the CYP7 and mito Clans via HEXB. P450 genes are in shaded boxes. CYP16B1 has a direct linkage to CYP20 in lancelet.

Figure 10.

Synteny mapping of the CYP46 clan to CYP7 and CYP26 clans via AK7, to ZDHHC12 via MRPL1 (see figure 2 for ZDHHC12 linkages), to CYP2 and mito clans via STOML1. CYP46 is connected by direct linkages to CYP Clans 2 and 4.

Figure 11.

The P450 genesis locus in animals.