InteractiveFly: GeneBrief

interference Hedgehog: Biological Overview | References

The iHog gene (interference hedgehog), identified by RNA interference in Drosophila cultured cells, encodes a type 1 membrane protein shown in this study to bind and to mediate response to the active Hedgehog (Hh) protein signal. ihog mutations produce defects characteristic of Hh signaling loss in embryos and imaginal discs, and epistasis analysis places ihog action at or upstream of the negatively acting receptor component, Patched (Ptc). The first of two extracellular fibronectin type III (FNIII) domains of the Ihog protein mediates a specific interaction with Hh protein in vitro, but the second FNIII domain is additionally required for in vivo signaling activity and for Ihog-enhanced binding of Hh protein to cells coexpressing Ptc. Other members of the Ihog family, including Drosophila Boi and mammalian CDO and BOC, also interact with Hh ligands via a specific FNIII domain, thus identifying an evolutionarily conserved family of membrane proteins that function in Hh signal response (Yao, 2006; full text of article).

Activity of the Hedgehog (Hh) signaling pathway is required for normal regulation of cell proliferation and differentiation during a diverse array of patterning events, ranging from embryonic segmentation in insects to neural tube differentiation in vertebrates. The Hh pathway also plays a homeostatic role in postembryonic tissues through its regulation of stem cell and precursor cell proliferation, and aberrant activity of the Hh pathway is associated with the initiation and growth of a variety of deadly human cancers. Despite the importance of this signaling pathway in development and disease, however, significant gaps remain in the understanding of Hh signal transduction (Yao, 2006).

The mature Hh ligand is derived from the Hh protein precursor by autoprocessing and lipid modification that generates an amino-terminal signaling peptide (HhN) dually modified by palmitoyl and cholesteryl adducts. Pathway activity is triggered by stoichiometric binding of this dually lipidated Hh ligand to Ptc, a 12-transmembrane transporter-like protein that in the absence of Hh acts catalytically to suppress activity of the seven-transmembrane protein Smoothened (Smo). The release of Smo inhibition by HhN binding to Ptc permits activation of an intracellular signal cascade that in turn activates latent cytoplasmic transcription factors, the zinc finger protein Ci (Cubitus interruptus) in Drosophila or the homologous Gli proteins in vertebrates, and these proteins subsequently stimulate transcriptional activation of pathway target genes (Yao, 2006).

The identification of Ptc as a Hh receptor is based on genetic epistasis studies demonstrating that Ptc functions downstream of Hh and upstream of Smo and other pathway components. Biochemical studies have further demonstrated that ShhN, the amino-terminal signaling domain of the mammalian Hh family member Sonic hedgehog, binds with high affinity to mammalian cells expressing murine Patched (mPtch). The signaling potencies of mutationally altered ShhN proteins in this cultured cell-based assay correlate with their apparent mPtch binding affinities. In addition to its cell-autonomous role in pathway activation, Ptc/HhN binding also results in sequestration and degradation of the HhN protein, thus preventing movement to distal cells and restricting the tissue range of Hh action in a cell nonautonomous fashion (Yao, 2006).

Despite the importance of Ptc/Hh binding in regulating both Hh tissue distribution and cell autonomous response to the Hh signal, it is curious that Ptc protein is mainly localized in vesicular structures within the cytoplasm, with barely detectable levels in the plasma membrane. The Hh receptor has not been biochemically defined and isolated, and the possibility remains that other pathway components may be involved in Hh binding to target cells. An RNAi-based genome-scale screen has recently identified several candidates for such action, including Dally-like (Dlp), a member of the glypican family of heparan sulfate proteoglycans (HSPGs) (Yao, 2006).

A second membrane-associated Hh pathway component, previously known as CG9211 and here referred to as ihog (interference hedgehog), was identified in this RNAi screen. RNAi targeting of Ihog in cultured cells reveals its requirement for Hh response in cultured cells, and genetic analysis shows that loss of ihog function in embryos and imaginal discs leads to patterning defects associated with loss or reduction of Hh pathway activation. The ihog gene encodes a type I transmembrane protein with four immunoglobulin-like (Ig) domains and two fibronectin type III (FNIII) domains in its extracellular region. A biochemical interaction between HhN and the Ihog extracellular domain maps to the first FNIII domain (FN1), but signaling function in vivo requires the additional presence of the second FNIII domain (FN2). The binding of HhN to cultured cells is greatly enhanced by coexpression of Ptc and Ihog, and both FN1 and FN2 of Ihog are required to mediate this synergistic effect. A Drosophila homolog and two mammalian homologs can also interact directly with Hh ligands through a specific FNIII domain, suggesting that proteins in this family constitute conserved components of the Hh signal reception machinery (Yao, 2006).

The embryonic and imaginal phenotypes of ihog mutations are somewhat weaker than those of mutations affecting other pathway components such as hh and smo. Of possible relevance, the Drosophila genome contains a related gene, boi, (brother of ihog). The Ihog and Boi proteins, as well as the related mammalian proteins BOC and CDO (Kang, 2002; Aglyamova, 2007; Okada, 2006; Tenzen; 2006), contain amino-terminal clusters of four or five immunoglobulin (Ig) domains followed by two or three fibronectin type III (FNIII) domains, with a predicted transmembrane domain and cytoplasmic carboxy-terminal cytotail that displays no apparent homology to other family members or other proteins. The two FNIII domains of the Ihog and Boi proteins are particularly well conserved, with 54% amino acid identity as compared to 42% identity in the Ig domains or 45% identity between the two proteins as a whole, and these domains are most closely related to the second and third FNIII domains of BOC and CDO. Transfection of a Boi expression construct can rescue RNAi-mediated loss of Ihog in cl-8 cells; and Boi, like Ihog, appears to function at or upstream of the level of Ptc. Boi expression is also capable of rescuing RNAi-mediated loss of dlp function in cl-8 cells (Yao, 2006).

Ihog protein is detected as a single species of about 100KD in Western blots of lysates from Drosophila embryos and from various Drosophila cultured cell lines. Ihog is ubiquitously expressed throughout embryogenesis and in the wing imaginal disc. Confirming this distribution, ihog RNA was found by RT-PCR to be expressed in embryos, the wing imaginal disc, and in cl-8 and S2-R+ cultured cells. Two isoforms of boi cDNA were characterized, one full-length and the other lacking coding sequences for the signal sequence and the Ig domains. The full-length boi transcript, like the ihog transcript, is expressed in embryos, in the wing imaginal disc, and in S2 cells, but in cl-8 cells only the defective RNA is detected (Yao, 2006).

Given these nearly identical functional properties of Boi and Ihog proteins in cl-8 cultured cells, it seems likely that Boi expression may at least partially compensate for loss of ihog in embryos and imaginal discs and thus account for the intermediate phenotypes of ihog mutations. The relatively strong effect of ihog RNAi in cl-8 cells, where its pathway role was discovered, may be due to the absence of boi expression. Both ihog and boi are expressed in S2 cells, where hyperphosphorylation of Smo provides a rapid and direct readout of Hh pathway activation. Combined RNAi against both gene products indeed produced additive effects in this assay. A similar RNAi-based analysis in embryos was precluded by stability of the Ihog protein, and no boi mutation or nearby transposable element for imprecise excision is currently available (Yao, 2006).

Ihog function at or upstream of the level of Ptc suggests a possible role for Ihog as a central or accessory component of the Hh receptor. This possibility was further investigated by examining the subcellular localization of Ihog and by asking whether it can interact with the Hh protein. Consistent with its predicted structure, the Ihog protein appears to be localized mainly to the surface of embryonic cells and can be detected on the cell surface in both permeabilized and nonpermeabilized S2-R+ cultured cells (Yao, 2006).

The Ihog protein likely is modified by glycosylation, as it was enriched by Concanavalin A (Con A) Sepharose chromatography. Ihog also can be cross-linked to membrane-impermeable biotin, indicating that it is a cell surface glycoprotein. The levels of Ihog protein in cl-8 cells were not affected by Hh stimulation, consistent with its lack of spatial modulation in embryonic and imaginal disc expression (Yao, 2006).

Hh binding by Ihog raises the question of how Ihog may interact with Ptc, a previously characterized Hh receptor. Binding was measured of HhN enzymatically tagged by fusion to Renilla luciferase (HhN-Ren) to intact cells coexpressing Ihog and Ptc; tagging of HhN at this site is compatible with signaling activity, although the resulting HhN-Ren protein, like the HhN in conditioned medium used for signaling assays, is not cholesterol modified. COS1 monkey cells were selected for these experiments to avoid potential interference from endogenous Drosophila proteins that may function in binding. It was found that transfection for expression of Ptc only slightly increased HhN-Ren binding above endogenous levels and that transfection for expression of Ihog increased binding less than 2-fold. Cotransfection for expression of Ptc and Ihog in contrast produced a more than ten-fold increase in binding of HhN-Ren, far greater than the additive effects of Ptc and Ihog alone. Similar effects were noted for expression of Boi, alone and together with Ptc (Yao, 2006).

S2-R+ cells offer the possibility of reducing background binding by RNAi-mediated knockdown of specific endogenous proteins through bathing of cells in dsRNA. Significant binding of HhN-Ren to untransfected S2-R+ cells was noted, and RNAi targeting of ihog or ptc reduced this basal binding 2- to 5-fold, indicating that endogenous Ptc and Ihog indeed contribute to basal HhN-Ren binding activity. It was also found that coexpression of Ihog and Ptc synergistically increased HhN-Ren binding to levels many fold higher than those produced by expression of either protein alone. To more accurately compare synergistic binding to that of Ptc or Ihog individually, expression of one component with RNAi was reduced while transfecting for expression of the other. It was found that HhN-Ren binding with coexpression of Ihog and Ptc was 59-fold higher than that produced by combined Ihog expression and RNAi-mediated knockdown of ptc and 30-fold higher than that seen with combined expression of Ptc and RNAi-mediated knockdown of ihog (Yao, 2006).

The two closest vertebrate homologs of ihog are Cdo and Boc (Kang, 2002; Aglyamova, 2007; Okada, 2006; Tenzen; 2006). Cdo-/- mice display mild holoprosencephaly (HPE) (Cole, 2003), more severe forms of which are associated with embryonic loss of Hh signaling. Using an established reporter assay involving a Gli-luciferase reporter construct in NIH 3T3 cells, it was found that shRNA constructs targeting either Cdo or Boc, but not a control shRNA construct, inhibited cell response to ShhN, thus suggesting that HPE in Cdo-/- mice may be due to defective Hh signal response (Yao, 2006).

Upon testing of various human Fc fusions for a direct interaction with ShhN in vitro, it was found that only the third FNIII domain of CDO and BOC can precipitate ShhN from conditioned media. Thus, although CDO and BOC differ from their Drosophila counterparts in number of FNIII domains and, in the case of CDO, in the number of Ig domains, these mammalian proteins appear to retain the ability to interact with a Hh ligand via a specific FNIII domain (Yao, 2006).

Previous studies of ShhN binding utilized a series of altered proteins with structure-based alanine substitutions for four groups of conserved surface residues (A, B, C, and D). Using the Fc fusion to the FN3 domain of CDO, it was found that surface A and B mutants are as efficiently coprecipitated as wild-type ShhN but that the surface C mutant was not coprecipitated and the surface D mutant only poorly precipitated. Similar results were noted for FN3 of BOC. The ShhN determinants required for in vitro interaction with the FN3 domains of CDO and BOC thus are very similar to those required for binding of ShhN to cells expressing Ptch, consistent with the possibility of a cooperative role for CDO/BOC proteins with Ptch in binding and reception of the Hh signal in mammals (Yao, 2006).

Wing imaginal disc clones lacking Ihog function display a cell-autonomous loss or reduction in the expression of Hh pathway targets. Epistasis analysis in cultured cells revealed that Hh pathway activation by RNAi of cos2 or by expression of a constitutively activated form of Smo is not blocked by RNAi of ihog and that RNAi of ptc reverses the loss of Hh response caused by ihog RNAi. Furthermore, the ptc embryonic cuticle phenotype prevails in ihog ptc double mutant embryos, placing Ihog function at or upstream of the level of Ptc. Ihog protein is predominantly localized at the cell surface and binds specifically to Hh in vitro, as demonstrated biochemically and by crystallographic analysis of an Ihog:Hh complex (McLellan, 2006). Finally, Hh binding to cultured cells is synergistically augmented by coexpression of Ihog with Ptc. On the basis of its cell-autonomous role in mediating Hh signal response, its localization to the cell surface, and its binding to the Hh signal directly and in synergy with Ptc, it is concluded that Ihog functions as a component of the Hh signal reception machinery. This role appears to be conserved in mammals, since mutation of the mouse homolog Cdo produces holoprosencephaly (Cole, 2003), and CDO and its close relative BOC both bind to the Shh signal in vitro and contribute to Shh response in cultured cells (Yao, 2006).

The embryonic and imaginal disc phenotypes of ihog mutations are intermediate in severity, likely due to overlapping expression of boi, a closely related member of the Ihog family whose Hh pathway function in cultured cells can substitute for that of ihog. Boi interacts with HhN in vitro and synergistically with Ptc in vivo. Residual pathway function in embryos and imaginal disc cells lacking ihog may well be due to expression of boi in these tissues, and more severe phenotypes thus would be predicted to result from the combined absence of Boi and Ihog function, once boi mutations become available. Consistent with this prediction, RNAi-mediated targeting in S2-R+ cells of both ihog and boi produced a stronger effect on Smo accumulation and phosphorylation than targeting of either gene alone (Yao, 2006).

Although the binding interactions between Ihog and HhN are mediated by FN1 in vitro (McLellan, 2006), in vivo signaling function depends upon the additional presence of FN2. This requirement for FN2 likely relates to the finding that combined expression of Ihog and Ptc in cells produces 30- to 60-fold higher levels of HhN binding than either protein alone; this synergistic binding of Hh, like signaling, requires the presence of FN1 and FN2. Much, though not all, Ptc protein is localized in intracellular vesicles, and neither Ihog nor Ptc proteins detectably change localization when coexpressed. Despite this predominant intracellular localization for Ptc, HhN-Ren binding by Ptc/Ihog nevertheless appears to occur on the surface, since binding studies were carried out at 4°C, preventing endocytosis, and bound HhN-Ren is sensitive to treatment with trypsin. The basis of synergistic binding seems likely to be an increased affinity for Hh, as Scatchard analysis of HhN-Ren binding to cells expressing Ihog and Ptc yields an estimate of HhN affinity as much as two orders of magnitude stronger than that for Ihog alone (McLellan, 2006). This increase in affinity could be based on simultaneous interaction of Hh with Ptc and Ihog, since the Ptch/ShhN interface defined by mutational analysis is adjacent to and partially overlapping with the Ihog/Hh interface identified crystallographically (McLellan, 2006) (Yao, 2006).

The Ihog family of proteins has previously been studied primarily in mammals, where CDO and BOC were identified as members of a distinctive subgroup of the Ig/FNIII family. One of the distinguishing features of the Ihog/CDO subgroup is a higher degree of conservation within their membrane-proximal FNIII domains as compared to the Ig domains, which contrasts with higher conservation for the Ig domains in the Robo receptors and other subgroups of the larger Ig/FNIII family. Interestingly, these FNIII domains are critically important for Hh signaling function in functional dissection of the Drosophila proteins (Yao, 2006).

The CDO/BOC proteins were initially linked to myogenesis based on their overexpression in C2C12 and 10T1/2 cells, which promoted increased levels of myogenic transcription factors and myotube differentiation. The myogenesis-promoting effects of CDO and BOC were attributed to a promyogenic interaction of these proteins with cadherins (Kang, 2003) and with neogenin, a netrin receptor (Kang, 2004). Consistent with a role in myogenesis, loss of Cdo function in mouse embryos caused a reduction or delay in expression of promyogenic transcription factors Myf-5, MyoD, and myogenin and a delay in muscle development (Cole, 2004). It is interesting to note, however, that expression of Myf-5 and MyoD is also reduced in Shh-/- embryos, particularly in the epaxial domain of the newly forming somite and later in the epaxial dermomyotome. An effect on embryonic Shh signal response thus might account, at least in part, for the effect of Cdo loss on embryonic myogenesis (Yao, 2006).

Loss of Cdo function also produced a mild form of HPE in mice (Cole, 2003). The role of Hh signaling in HPE is clearly established from genetic analysis in mice and humans (Muenke, 2001), and the HPE phenotype of Cdo-/- mice thus may well be accounted for by a partial reduction of Hh pathway activation. Morpholino oligonucleotide-based disruption of boc expression in zebrafish embryos also has implicated BOC function in axonal growth guidance for ventrally projecting forebrain neurons (Connor, 2005). This defect could be due to an effect on Hh pathway activity, given the role of Hh signal response in ventrally directed axonal guidance of commissural neurons in the developing spinal cord (Yao, 2006).

The primary importance of individual Ihog FN1 and FN2 domains in Hh ligand binding and response suggests that other parts of the Ihog protein, which also are evolutionarily conserved, may play functional roles in other signaling pathways. Further genetic and biochemical analysis of variant forms of Ihog family proteins will be required to identify such roles and to learn how such pathways may be integrated with Hh signaling through use of a common receptor (Yao, 2006).

Search PubMed for articles about Drosophila iHog

Aglyamova, G. V. and Agarwala, S. (2007). Gene expression analysis of the hedgehog signaling cascade in the chick midbrain and spinal cord. Dev. Dyn. 236(5): 1363-73. Medline abstract: 17436280

Cole, F. and Krauss, R. S. (2003). Microform holoprosencephaly in mice that lack the Ig superfamily member Cdon. Curr. Biol. 13: 411-415. Medline abstract: 12620190

Cole, F., Zhang, W., Geyra, A., Kang, J. S. and Krauss, R. S. (2004). Positive regulation of myogenic bHLH factors and skeletal muscle development by the cell surface receptor CDO. Dev. Cell 7: 843-854. Medline abstract: 15572127

Connor, R. M., et al. (2005). BOC, brother of CDO, is a dorsoventral axon-guidance molecule in the embryonic vertebrate brain. J. Comp. Neurol. 485: 32-42. Medline abstract: 15776441

Kang, J. S., et al. (2002). BOC, an Ig superfamily member, associates with CDO to positively regulate myogenic differentiation. EMBO J. 21: 114-124. Medline abstract: 11782431

Kang, J. S., et al. (2003). Promyogenic members of the Ig and cadherin families associate to positively regulate differentiation. Proc. Natl. Acad. Sci. 100: 3989-3994. Medline abstract: 12634428

Kang, J. S., et al. (2004). Netrins and neogenin promote myotube formation. J. Cell Biol. 167: 493-504. Medline abstract: 15520228

McLellan, J. S., et al. (2006). Structure of a heparin-dependent complex of Hedgehog and Ihog. Proc. Natl. Acad. Sci. 103(46): 17208-13. Medline abstract: 17077139

Muenke, M. and Beachy, P. A. (2001). Holoprosencephaly. In: C. Scriver, A. Beaudet, W. Sly and D. Valle, Editors, The Metabolic and Molecular Bases of Inherited Disease, McGraw-Hill, New York pp. 6203-6230.

Okada, A., et al. (2006). Boc is a receptor for sonic hedgehog in the guidance of commissural axons. Nature 444(7117): 369-73. Medline abstract: 17086203

Tenzen, T., et al. (2006). The cell surface membrane proteins Cdo and Boc are components and targets of the Hedgehog signaling pathway and feedback network in mice. Dev. Cell 10(5): 647-56. Medline abstract: 16647304

Yao, S., Lum. L. and Beachy, P. (2006). The ihog cell-surface proteins bind Hedgehog and mediate pathway activation. Cell 125: 343-357. Medline abstract: 16630821

date revised: 20 October 2007

Home page: The Interactive Fly © 2006 Thomas Brody, Ph.D.