Evolution of hedgehog and hedgehog-related genes, their origin from Hog proteins in ancestral eukaryotes and discovery of a novel Hint motif

Abstract

Background: The Hedgehog (Hh) signaling pathway plays important roles in human and animal development as well as in carcinogenesis. Hh molecules have been found in both protostomes and deuterostomes, but curiously the nematode Caenorhabditis elegans lacks a bona-fide Hh. Instead a series of Hh-related proteins are found, which share the Hint/Hog domain with Hh, but have distinct N-termini.

Results: We performed extensive genome searches such as the cnidarian Nematostella vectensis and several nematodes to gain further insights into Hh evolution. We found six genes in N. vectensis with a relationship to Hh: two Hh genes, one gene with a Hh N-terminal domain fused to a Willebrand factor type A domain (VWA), and three genes containing Hint/Hog domains with distinct novel N-termini. In the nematode Brugia malayi we find the same types of hh-related genes as in C. elegans. In the more distantly related Enoplea nematodes Xiphinema and Trichinella spiralis we find a bona-fide Hh. In addition, T. spiralis also has a quahog gene like C. elegans, and there are several additional hh-related genes, some of which have secreted N-terminal domains of only 15 to 25 residues. Examination of other Hh pathway components revealed that T. spiralis - like C. elegans - lacks some of these components. Extending our search to all eukaryotes, we recovered genes containing a Hog domain similar to Hh from many different groups of protists. In addition, we identified a novel Hint gene family present in many eukaryote groups that encodes a VWA domain fused to a distinct Hint domain we call Vint. Further members of a poorly characterized Hint family were also retrieved from bacteria.

Conclusion: In Cnidaria and nematodes the evolution of hh genes occurred in parallel to the evolution of other genes that contain a Hog domain but have different N-termini. The fact that Hog genes comprising a secreted N-terminus and a Hog domain are found in many protists indicates that this gene family must have arisen in very early eukaryotic evolution, and gave rise eventually to hh and hh-related genes in animals. The results indicate a hitherto unsuspected ability of Hog domain encoding genes to evolve new N-termini. In one instance in Cnidaria, the Hh N-terminal signaling domain is associated with a VWA domain and lacks a Hog domain, suggesting a modular mode of evolution also for the N-terminal domain. The Hog domain proteins, the inteins and VWA-Vint proteins are three families of Hint domain proteins that evolved in parallel in eukaryotes.

Figure 1

Protein sequence logo of Hh Hog domains. Central section of the protein sequence logo that was generated from aligned Hog domains of diverse Hh proteins using LogoBar. For the full image see Additional file 2. The color scheme is similar to the one used in the multiple sequence alignments (N,Q,S,T: green; C: yellow; P: pink; G: orange; K,R: red; A,I,L,M,V: blue; F,W,Y: cyan blue; H, purple, D,E: magenta; gaps: white). The extend of the Hint domain and the SRR region are indicated above the logo with a red line. Red boxes underneath the logo indicate the different motifs A, B, F, J, K, L.

Figure 2

Multiple sequence alignment of Hog domains, part 1. Multiple sequence alignment in this and other figures was carried out using first MUSCLE and then imported into Clustal_X. Manual adjustments to the alignment were carried out using SeaView. Color coding was modified from default Clustal_X color coding by marking all cysteine residues in yellow, small hydrophobic residues in light blue and large hydrophobic residues in cyan blue. The Hint domain, as well as the C-terminal SRR or ARR regions are indicated above the alignment. Motifs A, B, F, J, K, and L are indicated with red rectangles underneath the alignment. Species abbreviations are shown in Table 3. Note that not all sequences in this alignment are complete.

Figure 3

Multiple sequence alignment of Hog domains, part 2. Continuation of the multiple sequence alignment of Figure 2.

Figure 4

Multiple sequence alignment of Hog domains, part 3. Continuation of the multiple sequence alignment of Figure 3.

Figure 5

Phylogenetic tree of Hog domains. Phylogenetic trees were built from aligned Hog domains (Figure 2 – 4). The Neighbor joining tree was created using the default settings of Clustal_X. Bootstrap values of 1000 trials are indicated in the figure. In this and subsequent phylogenetic tree figures Enoplea sequences are highlighted in light green, Cnidarian sequences in yellow, Choanoflagellate sequences in light red and fungal sequences in blue. The Hh sequences are marked with Hh and the nematode Hh-related sequences are marked with NemaHog. The root was placed between the red algae/plant sequences and the remaining sequences. Some incomplete sequences were omitted in this tree. Additional phylogenetic analyses were also carried out, for example by omitting the protist sequences and using the fungal sequence GmGIN1 as outgroup (Additional files 5, 6, 7). Overall, the results were very similar.

Figure 6

Maximum likelihood phylogenetic tree of Hog domains. A Maximum likelihood phylogenetic tree was constructed using the same data as in Figure 5. Phyml default values were used, and bootstrap values for 100 trials are shown.

Figure 7

Phylogenetic tree of Wart domains. A multiple sequence alignment of Wart domains (see Additional file 4) was used to generate at Neighbor joining tree with the default settings of Clustal_X. B. malayi sequences are highlighted in light blue. This tree is unrooted. Results of 1000 bootstrap trials are shown.

Figure 8

Phylogenetic tree of Ground and Ground-like domains. A multiple sequence alignment of Ground and Ground-like domains (see Additional file 8) was used to generate a Neighbor joining tree with the default settings of Clustal_X. For grd-1, grd-2 and grd-11 the four Ground domains were extracted manually prior to alignment; the R1 to R4 postscripts indicate the repeat number. B. malayi sequences are highlighted in light blue. This tree is unrooted. Results of 1000 bootstrap trials are shown.

Figure 9

Phylogenetic tree of Hedge domains. A multiple sequence alignment of Hedge and Hedgehog proteins (see Additional file 10) was used to generate at Neighbor joining tree with the default settings of Clustal_X. This tree is unrooted. Results of 1000 bootstrap trials are shown.

Figure 10

Multiple sequence alignment of Enoplea Hog proteins with a new upstream motif. Multiple sequence alignment of Enoplea Ts Xhog1, Ts Xhog2, and XC Xhog5 reveals new conserved regions upstream of the Hog domain.

Figure 11

Multiple sequence alignment of two cnidarian Hog proteins with a new upstream motif. Pairwise sequence alignment of cnidarian Nv 120428 and Acm DY579185.

Figure 12

Pairwise sequence alignment of Alveolata aCm and aCp Hog.

Figure 13

Pairwise sequence alignment of red algae rPy and rPh Hog2.

Figure 14

Multiple sequence alignment of moss pPp Hog, pSl Hog and pSm Hog.

Figure 15

Multiple sequence alignment of jakobid jJl Hog1, Hog2, and Hog3.

Figure 16

Multiple sequence alignment of haptophyte hPh Hog1, Hog2, and Hog3.

Figure 17

Multiple sequence alignment of VWA domain – Hint-like domain proteins, part 1. Proteins containing a VWA merged to a Hint-like domain were discovered in Tetrahymena, several fungal species, as well as several other eukaryote branches, including choanoflagellates. The VWA domain and the Hint-like domain (Vint) with motifs A and B of the Hint domain are marked in the alignment. A new domain between the VWA and Vint domain is marked with Vwaint. Four proteins also have an Ubox upstream of the VWA domain. An alignment of selected Vint domains to Hh Hog domains is presented in additional file 15. A. thaliana At5g60710 is not a Vint protein, but one of the best matching VWA domain containing proteins. While it lacks the Vint domain, it does have some weak similarity to the Vwaint domain, and upstream of the VWA domain is a Ring finger, which shares similarity with the Ubox motif. It would be worthwhile to investigate this similarity with a detailed evolutionary analysis in the future.

Figure 18

Multiple sequence alignment of VWA domain – Hint-like domain proteins, part 2. Continuation of the multiple sequence alignment of Figure 17.

Figure 19

Multiple sequence alignment of VWA domain – Hint-like domain proteins, part 3. Continuation of the multiple sequence alignment of Figure 18.

Figure 20

Summary of the evolution of hh and hh-related genes. For detailed discussion of the evolution of the Hog proteins see text. The right side shows the different types of ORFs found in different organisms. The sizes are not to scale. The "Hedge" domain is marked in green, the Qua domain in orange, and the Hog domain in black, with yellow bars representing the conserved cysteine residues. T stands for poly-threonine repeats. Red 'X's mark branches where a gene loss occurred.