Consensus phylogenetic tree of eukaryotes. The branches where Hog domain containing proteins are found are indicated with red dots. With permission from Sandra Baldauf, (see, also [71]).

Distribution of Hint superclass genes in the three domains of life. Hint genes can be divided into several different classes: inteins; Bil-A (bacterial intein-like genes type A); Bil-B; a new class referred to here as Bil-C [10,13]; Vint (VWA domain and Hint domain proteins) [10]; and Hog.

Structural features of Hh proteins. (a) Signal peptide sequence for protein export (SS, yellow), the amino-terminal signaling domain (HhN, green), and the autocatalytic carboxy-terminal domain (HhC, black) are indicated. Both HhN and HhC domains are also found in proteins other than the Hh family, and are therefore globally referred to as 'Hedge' and 'Hog', respectively. The Hog domain itself can be separated into two regions; the first two-thirds shares similarity with self-splicing inteins, and this module has been named Hint, whereas the carboxy-terminal third binds cholesterol in Hh proteins and has been named the sterol-recognition region (SRR) [14]. In Hog proteins other than Hh, that is, Hh-related proteins, this region is referred to as ARR (adduct recognition region) [21], as the nature of the adduct is not known. (b) Intramolecular autoprocessing of Hh. Acids and bases assisting in catalysis are not shown (figure adapted from [14,70]).

One possible scenario for the evolution of hh and hh-related genes in metazoa. Different phylogenetic branches are outlined, and gene families known at present are shown. Dotted lines indicate uncertain evolutionary connections. Hedgling genes are currently known only from sponges and Cnidaria [8,10,19]. The Hh family could have originated in two possible ways. (a) The Hedge domain evolved concomitantly with the Hog domain from a protist Hog protein before the emergence of the Metazoa. A duplication of the Hedge domain and merger with an extracellular protein gave rise to the Hedgling gene. (b) No hh gene existed at the emergence of sponges. The Hedge domain of a Hedgling gene duplicated and merged with a Hog gene to give rise to hh in early Eumetazoa. Cnidaria, Lophotrochozoa and nematodes contain both Hh as well as other Hog family genes. The phylogenetic analysis cannot unequivocally resolve whether these other families originated from a single ancestor in Eumetazoa - as shown here with dotted lines - or whether, at least in some phyla, duplication and divergence from a hh gene gave rise to new families in particular phyla.

Neighbor-joining phylogenetic tree of eukaryote Hog domain protein sequences. The Hh, Groundhog (Grd), Warthog (Wrt), Quahog (Qua), and new Lophohog families are indicated. Sequence names are color-coded according to phyletic divisions, except for sponges. Chromadorea and Enoplea are two major nematode divisions. Protist is loosely used to encompass all non-metazoans. The Hint domains of Vint proteins were used as outgroup and bootstrap values ≥ 40% are shown. Most of the sequences and the analysis methods are described in [10]. Additional sequences were added to this analysis from sponges [19], and BLAST searches were carried out at JGI [72] of the genomes of L. gigantea and Capitella I ECS-2004. From Capitella I ECS-2004 one hh and one hh-related gene were retrieved, and from L. gigantea one hh and six other Hog genes were retrieved. Capitella Cap_213608 and L. gigantea Lg_236513, which encodes an export signal peptide, form a clade, although not with high bootstrap significance. Interestingly, this clade clusters with the Cnidarian hh-related genes - although bootstrap values are insignificant. Five L. gigantea Hog genes (Lg_173620, Lg_173619, Lg_237232, Lg_FC606200, Lg_229767) form a distinct clade, but these genes are very divergent from the Hog domains of the other metazoan genes. These genes encode only a few residues upstream of the Hog domain (7-15), and lack an export signal peptide. This unusual structure is confirmed by multiple expressed sequence tags (ESTs) for each gene. Do these genes represent a highly divergent form of Hog-only proteins in this gastropod, or do they stem from another organism, perhaps some ciliated protozoan parasite found in L. gigantea [73]? More analysis will be necessary to resolve this.

A simplified Hh signaling pathway, constructed from combined Drosophila and mammalian data. Hh is targeted to the endoplasmic reticulum by its signal peptide, is palmitoylated at its amino terminus by Rasp/Skinny Hedgehog (Ski), and autoprocessed. Lipidated HhN (M-HhN) is released by Dispatched (Disp) and forms multimers or associates with lipoproteins (LP) in the extracellular environment [32]. A number of molecules can interact with M-HhN and propagate or modulate its trafficking: the Dally-like protein (Dlp), which is modified by the heparan sulfate (HS) polymerases Tout-velu (Ttv), Sister of tout-velu (Sotv), and Brother of tout-velu (Botv), all members of the EXT family; the Hedgehog-interacting protein (Hip); and the Growth-arrest-specific 1 (Gas1) protein. Hip and Gas1 are not present in Drosophila. Megalin (Meg) is most probably involved in the recycling of M-HhN. Ihog is thought to function as co-receptor for M-HhN. M-HhN acts as an antagonistic ligand that represses the function of the receptor Patched (Ptc), a 12-transmembrane protein related to Disp. Binding of M-HhN to Ptc results in internalization. Smoothened (Smo) is a seven-pass membrane receptor, which is key for the transmission of the signal to the nucleus in the Hh pathway. Smo is inhibited by Ptc when not bound by M-HhN. When the inhibitory function of Ptc is released by M-HhN, Smo can translocate to the plasma membrane or - in mammals - to the primary cilium, and active Smo is phosphorylated (red P). Ptc may secrete pro-vitamin D3 or related compounds (D3) to inhibit Smo. Conversely, oxysterols (Oxy) can indirectly activate Smo [52,55]. The Hh pathway downstream of Smo displays some important differences between Drosophila and mammals. In Drosophila, when Smo is active, the signal passes through a complex comprising the kinesin-like molecule Costal 2 (Cos2), Fused (Fu), Suppressor of fused (Su(fu)) and Cubitus interruptus (Ci), leading to the release of Ci, which can then enter the nucleus to activate transcription. When Smo is inhibited, the Cos2/Fu/Su(fu)/Ci complex remains associated with microtubules, Ci is phosphorylated and is cleaved by Cos2. The Ci fragment now acts as a transcriptional repressor. In mammals, the targeting of Smo to primary cilia is essential for signal transduction. No obvious equivalents of Cos2 and Fu exist in mammals. Instead, Su(fu) has a more prominent role in inhibiting the pathway. Gli1, Gli2, and Gli3 are the mammalian homologs of Ci; Gli1 and Gli2 activate transcription when Smo is active, whereas Gli3 is processed and becomes a repressor when Smo is inhibited. A number of components in the pathway, in particular downstream of Smo, are not shown in this figure.