Comparative genomics of Lbx loci reveals conservation of identical Lbx ohnologs in bony vertebrates

Abstract

Background: Lbx/ladybird genes originated as part of the metazoan cluster of Nk homeobox genes. In all animals investigated so far, both the protostome genes and the vertebrate Lbx1 genes were found to play crucial roles in neural and muscle development. Recently however, additional Lbx genes with divergent expression patterns were discovered in amniotes. Early in the evolution of vertebrates, two rounds of whole genome duplication are thought to have occurred, during which 4 Lbx genes were generated. Which of these genes were maintained in extant vertebrates, and how these genes and their functions evolved, is not known.

Results: Here we searched vertebrate genomes for Lbx genes and discovered novel members of this gene family. We also identified signature genes linked to particular Lbx loci and traced the remnants of 4 Lbx paralogons (two of which retain Lbx genes) in amniotes. In teleosts, that have undergone an additional genome duplication, 8 Lbx paralogons (three of which retain Lbx genes) were found. Phylogenetic analyses of Lbx and Lbx-associated genes show that in extant, bony vertebrates only Lbx1- and Lbx2-type genes are maintained. Of these, some Lbx2 sequences evolved faster and were probably subject to neofunctionalisation, while Lbx1 genes may have retained more features of the ancestral Lbx gene. Genes at Lbx1 and former Lbx4 loci are more closely related, as are genes at Lbx2 and former Lbx3 loci. This suggests that during the second vertebrate genome duplication, Lbx1/4 and Lbx2/3 paralogons were generated from the duplicated Lbx loci created during the first duplication event.

Conclusion: Our study establishes for the first time the evolutionary history of Lbx genes in bony vertebrates, including the order of gene duplication events, gene loss and phylogenetic relationships. Moreover, we identified genetic hallmarks for each of the Lbx paralogons that can be used to trace Lbx genes as other vertebrate genomes become available. Significantly, we show that bony vertebrates only retained copies of Lbx1 and Lbx2 genes, with some Lbx2 genes being highly divergent. Thus, we have established a base on which the evolution of Lbx gene function in vertebrate development can be evaluated.

Figure 1

Molecular phylogenetic analysis of Lbx sequences. The tree shows a maximum likelihood analysis of Lbx protein sequences. Dark grey boxes indicate orthologous groups while light grey boxes indicate gene linkage. Bootstrap values below 70 have been removed. Only two Tetraodon Lbx proteins are included as the third sequence (found on scaffold 8483) is incomplete. Note that the Lbx sequences separate into two clear groups, Lbx1 and Lbx2, which are supported by bootstrap values of 79 and 83 (boxed), respectively. The chicken protein previously denoted as Lbx3 [18], the zebrafish protein previously denoted as Lbx1 [14] and one of the fugu, Tetraodon, stickleback and medaka sequences group with mammalian Lbx2 sequences. The two novel zebrafish Lbx sequences and the remaining teleost sequences group with mammalian, chicken and frog Lbx1. The tree shows significantly longer branch lengths for the Lbx2 proteins indicating that these sequences may be evolving at a quicker rate than the Lbx1 proteins. For common names of species see additional file 2.

Figure 2

Human paralogons of the LBX/TLX cluster loci. Schematic representation of the four human LBX/TLX paralogons plus paralogous genes occurring at the NKX3.2 locus, deduced from analysis of the human genome using NCBI map viewer Homo sapiens Build 36.2 (September 2006). Genes are represented by boxes with gene names appearing above and gene subfamilies indicated by numbers inside. Letters on the far left indicate paralogon designation, while chromosomal locations are indicated on the far right. Numbers between genes are approximate intergenic distances in Kb. Background shading indicates paralogous genes. Blue and red dots indicate sites of inversions. Note that LOXL1, AUP1 and PCGF1 genes appear only at the LBX2 locus and KAZALD1, POLL, DPCD, FBXW4 and MGEA5 appear only at the LBX1 locus. BTRC, FGF, NPM and KCNIP genes are linked with LBX and TLX genes at more than one locus, suggesting that these genes were acquired by the LBX/TLX cluster during or before the two rounds of vertebrate genome duplication. Also note that a KCNIP4 and a SLIT2 gene are associated with NKX3.2 on chromosome 4. A KCNIP3 gene is located on the long arm of chromosome2, however at a distance to LBX2, suggesting that it is not an original component of the LBX2 paralogon.

Figure 3

Genomic organisation of human LBX/TLX cluster paralogons and putative orthologous counterparts. Genomic organisation of human (Hs) Lbx/Tlx paralogons and the putative orthologous counterparts identified in Mus musculus (Mm), Xenopus tropicalis (Xt), Gallus gallus (Gg), Danio rerio (Dr), Tetraodon nigroviridis (Tn) and Takifugu rubripes (Fr). For simplicity, only three teleosts are included in this figure: additional data for Oryzias latipes (medaka) and Gasterosteus aculeatus (stickleback) can be found in additional file 1. The main genes characteristic of each paralogon are included in the figure; additional genes and linked genes found further away can be found in additional file 1. Schematic representations of the orthologous genomic regions are depicted in panels 1–4, which correspond to human paralogons 1–4 in figure 2. Gene orthology is indicated by colour code and was inferred from molecular phylogenetics (see additional files). Genes appearing in more than one orthologous region are underlined. Numbers at the ends of each line indicate chromosome (chr) or scaffold (s) numbers. A parallel red line marks gaps in the sequence, while triple black lines indicate large intergenic distances. Blue dots mark sites of inversions, red arrows the end of a contig. Boxes off-line represent genes for which genomic sequences are incomplete; where no chromosome or scaffold number is given, no linkage data is available. In some cases, "missing" genes can be found on different scaffolds or chromosomes. This data is shown in additional file 1. In addition, we do not show the dispersed 2^nd teleost Lbx2 paralogon that no longer contains an Lbx or Tlx gene, but this information is included in additional file 1. The genes our phylogenetic analysis identified as Lbx1 are found in a similar genomic environment, linked to orthologs of the genes present at the human LBX1/TLX1 locus. In addition, non-mammalian Lbx1 genes are linked to the Slc2a15 gene not found at any other Lbx locus. Lbx2 genes are linked to Loxl3, Aup1 and Pcgf1 genes (genomic information incomplete for chicken and frog), while Tlx3 genes are linked to the orthologs of human KCNIP1, NMP1, FGF18 and BTRC2. In teleosts, Fgf24 and Npm4 are part of the Lbx2 locus, while in all species examined, Fgf17 and Npm2 are no longer linked to Lbx/Tlx genes.

Figure 4

A model of the evolution of the vertebrate Lbx/Tlx loci. Lbx loci in extant vertebrates arose from a region containing an Lbx, Tlx, Fgf8/17/24/18 and an Npm gene. After one round of whole genome duplication (1R WGD) Lbx1/4 and Lbx2/3 precursors were produced linked to Fgf8/17 – Npm2/3 and Fgf24/18 – Npm4/1 precursors, respectively. During the second round of whole genome duplication (2R WGD), four loci where produced. By the time of the divergence of the lobed-fined and ray-finned fish, the Lbx4, Tlx4 and Lbx3 genes (shown in grey) had been lost.