Ihog and Boi are essential for Hedgehog signaling in Drosophila

Background The Hedgehog (Hh) signaling pathway is important for the development of a variety of tissues in both vertebrates and invertebrates. For example, in developing nervous systems Hh signaling is required for the normal differentiation of neural progenitors into mature neurons. The molecular signaling mechanism underlying the function of Hh is not fully understood. In Drosophila, Ihog (Interference hedgehog) and Boi (Brother of Ihog) are related transmembrane proteins of the immunoglobulin superfamily (IgSF) with orthologs in vertebrates. Members of this IgSF subfamily have been shown to bind Hh and promote pathway activation but their exact role in the Hh signaling pathway has remained elusive. To better understand this role in vivo, we generated loss-of-function mutations of the ihog and boi genes, and investigated their effects in developing eye and wing imaginal discs. Results While mutation of either ihog or boi alone had no discernible effect on imaginal tissues, cells in the developing eye disc that were mutant for both ihog and boi failed to activate the Hh pathway, causing severe disruption of photoreceptor differentiation in the retina. In the anterior compartment of the developing wing disc, where different concentrations of the Hh morphogen elicit distinct cellular responses, cells mutant for both ihog and boi failed to activate responses at either high or low thresholds of Hh signaling. They also lost their affinity for neighboring cells and aberrantly sorted out from the anterior compartment of the wing disc into posterior territory. We found that ihog and boi are required for the accumulation of the essential Hh signaling mediator Smoothened (Smo) in Hh-responsive cells, providing evidence that Ihog and Boi act upstream of Smo in the Hh signaling pathway. Conclusions The consequences of boi;ihog mutations for eye development, neural differentiation and wing patterning phenocopy those of smo mutations and uncover an essential role for Ihog and Boi in the Hh signaling pathway.


Background
The Hedgehog (Hh) signaling pathway is essential for proper embryonic development, and aberrant Hh pathway activity is the cause of several human congenital defects and cancers [1][2][3]. In developing nervous systems, Hh is an important factor governing neural fate specification, neural precursor proliferation, and axon guidance [3][4][5]. For example, Hh signaling specifies neural cell fate identity in the developing neural tube of vertebrates [4], and is essential for the normal differentiation of photoreceptors in the Drosophila compound eye [6]. Despite its importance in these and many other developmental events in diverse species, the molecular signaling mechanism underlying the function of Hh is still not fully elucidated.
Hh is a secreted protein that elicits concentrationdependent effects [3,4]. Genetic and biochemical experiments have led to a model where Patched (Ptc), a 12-pass transmembrane protein, is involved in sensing the extracellular Hh concentration [2]. In the absence of Hh, Ptc maintains the 7-pass transmembrane protein Smoothened (Smo) in a repressed state. Under these conditions, the Cubitus interruptus (Ci) transcription factor is proteolytically cleaved and acts as a transcriptional repressor. Conversely, in the presence of Hh, Ptc-mediated repression of Smo is relieved and this leads to the stabilization of full-length Ci (Ci155), which activates transcription of Hh target genes. Ihog (Interference hedgehog; CG9211) and Boi (Brother of Ihog; CG13756) are two related immunoglobulin superfamily (IgSF) members composed of four immunoglobulin-like (Ig) domains, two fibronectin type III (FN3) repeats, a transmembrane domain, and a cytoplasmic tail [7][8][9]. The first FN3 domain of Ihog is required and sufficient for direct binding to Hh [9,10]. In transcription reporter assays in cultured cells, modulation of Ihog and Boi levels affected the strength of responses to Hh [8,9]. Although Ptc plays a critical role in sensing the Hh morphogenic gradient, the identification of Ihog and Boi as proteins that bind to Hh and promote pathway activation raised questions about their exact role in the Hh signaling pathway. To better understand this role in vivo, we generated loss-of-function mutations of the ihog and boi genes in Drosophila. We have investigated the effects of these mutations in developing eye and wing imaginal discs, and found that Ihog and Boi are functionally redundant and are required for Hh signaling in these tissues. This work uncovers an essential role for Ihog and Boi in the Hh signaling pathway.
We characterized a UAS-ihog:myc transgenic line with basal levels of expression in the absence of a Gal4 driver ( Figure 2D). Eclosion and adult viability were fully rescued in the presence of UAS-ihog::myc (Table 1), demonstrating that death at pupal stages was caused by loss of Ihog function and not CG10158. Together, the data indicate that zygotic Ihog is important for adult viability only in the absence of maternal Ihog.
The viability and lack of overt Hh-like phenotypes in ihog DC1 mutants prompted us to also investigate Boi, since it behaves similarly to Ihog in Hh transcription reporter assays in vitro [9]. We used chemical mutagenesis to generate a boi mutation (boi C1 ) in a genetic screen ( Figure 1B; Additional file 1; see Materials and methods). boi C1 causes a premature stop codon at amino acid Trp626 ( Figure 1B). This mutation is predicted to truncate the Boi protein within the second FN3 domain, and fails to encode the transmembrane domain and cytoplasmic tail ( Figure 1C). Therefore, the boi C1 mutation is predicted to severely disrupt the properties of Boi. The boi gene is situated on the X chromosome and therefore males are boi C1/hemizygotes. Similar to ihog, boi C1/C1 female flies and boi C1/male flies were viable and fertile, but unlike ihog there was no maternal effect.
Thus, zygotic mutants for either ihog DC1 or boi C1 were viable and exhibited no overt phenotype reminiscent of mutations of components of the Hh signaling pathway. However, larvae that were double mutants for ihog DC1 and boi C1 died 24 to 48 hours after hatching, and never reached third instar (L3), suggesting that ihog and boi might act redundantly. In support of this idea, viability of double mutants was rescued in the presence of UASihog::myc, suggesting that Ihog could supplant Boi function in these flies ( Figure 2C).

Effects of ihog and boi mutations in the developing visual system
In the developing Drosophila retina, neuronal differentiation within the eye imaginal disc proceeds stepwise in the wake of a constriction in the disc epithelium called the morphogenetic furrow (MF) [11]. Hh signaling is required for the initiation of the MF at the posterior margin of the disc, and for its wave-like progression across the disc that marks the boundary between undifferentiated anterior cells and differentiating posterior cells [6]. Previous studies have demonstrated that clones of cells in the developing eye disc that lack the essential Hh mediator Smo retard MF progression and exhibit a cell-autonomous delay of differentiation into photoreceptors (R-cells) [12,13].
To begin to investigate whether ihog and boi could play a role in the developing visual system, we studied their expression in wild-type L3 larvae using in situ hybridization ( Figure 1D, E). Transcripts for both genes were observed throughout the developing eye disc, but were enriched in anterior regions characterized by undifferentiated cells compared to the posterior regions where there is ongoing differentiation of retinal cell types. Transcripts for neither ihog nor boi were detected in the optic lobe and other regions of the brain at this stage. We confirmed the complete loss of ihog mRNA in ihog DC1 mutants ( Figure 1F) and thereby demonstrated the specificity of the ihog probe. We could not do likewise with the boi probe, since it was predicted to hybridize with transcripts from the boi C1 nonsense mutation. However, the boi probe is unlikely to cross-hybridize with ihog transcripts as the ihog probe does not detect boi transcripts in ihog DC1/DC1 mutants. Together, these data revealed overlapping expression of ihog and boi in the developing eye disc, supporting the idea that they might act redundantly in this context.
To test this directly, we needed to generate mosaic animals because, as noted above, double mutant larvae died prior to L3. In boi C1/hemizygotes, we used FLP-mediated mitotic recombination to render the majority of cells in the developing eye discs also mutant for ihog DC1/DC1 [14]. Eye development was severely compromised in these mosaic animals ( Figure 2B), leading to structural defects reminiscent of those found in smo and hh mosaics [15][16][17]. Control mosaics in which ihog DC1/DC1 eye discs were generated in boi C1/+ heterozygotes had normal eyes ( Figure  2A). Additional controls with normal eyes were boi C1/-; ihog DC1/+ males (not shown) and rescued double mutants (genotype: boi C1/-; ihog DC1/DC1 ; UAS-ihog::myc/+; Figure  2C). These results indicated that Ihog and Boi are together required for the proper development of the Drosophila eye.

Ihog and Boi are required for Hh signaling and neuronal differentiation in the eye imaginal disc
To determine whether the disruption of eye development that we observed was due to perturbation of the Hh signaling pathway, we first examined the distribution of Ci155, a Gli-related zinc-finger protein that regulates transcription of Hh-responsive genes. Hh signaling inhibits proteolytic processing of Ci155 to the truncated repressor form Ci75 [3,18]. Using an antibody that detects Ci155 but not Ci75, one can monitor Hh signaling in cells that accumulate high levels of Ci155 [19]. In eye discs of L3 larvae, Ci155 normally accumulates to high levels just anterior to the advancing MF, marking Hh signaling that is required for MF progression [12,20,21]. In control experiments, we generated ihog DC1/DC1 clones in boi C1/+ heterozygotes and found no effect on Ci155 accumulation just anterior to the MF (eight of eight clones; Figure 3A-B''). In contrast, ihog DC1/DC1 clones in boi C1/mutants had no detectable Ci155 expression at the MF (11 of 11 clones; Figure 3C-D''), indicating a loss of Hh signal transduction in double mutant cells.
Like smo mutant clones [12,15], we found that boi;ihog double mutant clones failed to express Elav, a neuronspecific marker for R-cells in the eye disc (green fluorescent protein (GFP)-negative in Figure 3E-E''). In The eye is small, poorly shaped, and many ommatidia are missing. Some double mutant ommatidia appear to differentiate in these mosaics but have a roughened appearance, which is consistent with smo mutations. (C) Addition of Ihog rescues viability and eye morphology in a double mutant (genotype: boi C1 /Y; ihog DC1/DC1 ; UAS-ihog::myc/+). (D) RT-PCR to demonstrate loss of ihog transcripts in ihog DC1/DC1 mutants, and to characterize a transgenic line with low-level, constitutive expression of UAS-ihog::myc in the absence of a Gal4 driver.
controls, cells posterior to the MF that were mutant for only ihog (GFP-negative in Figure 3A, B) or only boi (GFP-positive in Figure 3E) expressed Elav normally, indicating that simultaneous loss of both family members is required to affect R-cell differentiation.
Elevated Ci155 levels drop sharply in cells posterior to the MF in a Cullin-3-dependent proteolytic process associated with the onset of R-cell differentiation [12,15]. As in the case of smo mutant clones [12,15], the absence of R-cells in clones lacking ihog and boi is accompanied by ectopic accumulation of Ci155 in clones located posterior to the MF ( Figure 3E-E''). Taken together, our data suggest that Ihog and Boi are functionally redundant and, like Smo, are essential for the differentiation of R-cells in response to Hh.

Ihog and Boi are required cell-autonomously for both high-and low-threshold responses to Hh pathway activation
To determine whether Ihog and Boi also function in Hh signaling elsewhere, and to explore whether they mediate high-threshold and low-threshold responses to a Hh morphogen gradient in vivo, we examined their role in wing development. Subdivision of the developing imaginal wing disc into anterior and posterior compartments involves the posterior-specific expression of the selector gene Engrailed [22], which programs posterior cells to produce and secrete Hh, and simultaneously prevents their response to Hh by blocking expression of Ci [23,24]. Therefore, only anterior cells limited to a broad stripe along the anterior/posterior compartment boundary can respond to Hh, and they do so by upregulating the expression of Hh target genes in a manner that depends on their proximity to the boundary and therefore the concentration of Hh to which they are exposed [25]. Cells immediately adjacent to the compartment boundary respond to high levels of Hh by expressing Ptc, a direct transcriptional target of Hh signaling [26]. Cells positioned more anteriorly respond to lower levels of Hh marked by high Ci155 accumulation, though all anterior cells have low baseline levels of Ci155. In control experiments, levels of Ptc and Ci155 were unaffected in ihog DC1/DC1 mutants (not shown), in boi C1/hemizygous cells (GFP-positive in Figure 4C, D) and in ihog DC1/DC1 clones in boi C1/+ heterozygotes (12 of 12 clones; GFP-negative in Figure 4A, B).
In contrast, cells lacking both ihog and boi in anterior clones situated near the compartment boundary did not express Ptc and did not accumulate high levels of Ci155 (12 of 12 clones; Figure 4C-D'''), similar to smo clones [27]. Ci155 normally accumulates to high levels even at low concentrations of Hh [28,29], and so the failure to elicit this low-threshold response indicates that removal of ihog and boi results in complete loss of Hh pathway activation. Importantly, while cells lacking both ihog and boi lost the ability to respond to Hh, cells immediately adjacent to these clones expressed Ptc and accumulated Ci155 normally (arrowheads in Figure 4D'''), indicating that Ihog and Boi are required cell-autonomously in Hh responding cells.

Ihog and Boi sequester Hh activity
In response to Hh, upregulation of Ptc levels in anterior cells near the compartment boundary reduces the anterior range of Hh activity [27]. This finding has been fundamental to the prevailing model that Ptc is a Hh receptor that can directly bind and sequester Hh protein. An alternative hypothesis is that Ptc interacts indirectly with Hh, perhaps by tethering another protein that directly binds Hh [27]. Since Ihog and Boi bind directly to Hh in vitro [9], and since Ptc and Ihog have been shown to synergize to increase binding of Hh to cultured cells [9], we wondered whether Ihog and Boi could likewise limit the range of Hh activity within the anterior compartment of the wing disc. Therefore, we studied large anterior clones that spanned the usual Ptc expression domain, and examined Ptc and Ci levels in cells immediately anterior to these clones. In cells that were immediately anterior to control ihog DC1/DC1 clones in boi C1/+ heterozygotes, there was no evidence for Hh pathway activation. In contrast, high levels of both Ptc and Ci155 were observed in cells immediately anterior to clones lacking Ihog and Boi ( Figure 4C-D'''),  indicating that the loss of both Ihog and Boi fails to restrict the movement of Hh through the clone.

Ihog and Boi act upstream of Smo
Activation of the Hh signaling pathway leads to increased accumulation of Smo protein at the surface of cells near the compartment boundary of the wing disc [30]. This accumulation of Smo at the cell surface is required for its signaling activity, and therefore marks Smo activation by Hh [30][31][32][33]. Thus, to test whether Ihog and Boi could  Ihog and Boi are required for compartment-specific cell affinity Cells of the wing imaginal disc normally do not cross the boundary between the anterior and posterior compartments, and implementation of this boundary is thought to involve mechanisms that control compartment-specific cell affinity and adhesion [34]. Clonal analysis has demonstrated that Hh signaling is required to maintain cell affinities in the anterior compartment. For example, if anterior cells that are adjacent to the boundary are mutant for smo or ci [35][36][37], they sort out from other anterior cells and segregate into posterior territory. To determine if Ihog and Boi are involved in maintaining compartment-specific cell affinities, we examined the segregation behavior of clones near the anterior-posterior compartment boundary. To do this, clones of anterior origin were unambiguously identified if they expressed baseline levels of Ci, and if an adjacent wildtype sister clone, the twin-spot, was situated in the anterior compartment since, by definition, the mutant clone and twin-spot must arise from the same compartment.
In controls (ihog DC1/DC1 clones in boi C1/+ heterozygotes), anterior-derived clones exclusively occupied anterior territory and defined a straight border with posterior cells in the normal position of the boundary (14 of 14 clones; Figure 4A-B'''). In contrast, double mutant clones (ihog DC1/DC1 clones in boi C1/hemizygotes) often straddled the boundary and formed tight borders with both anterior and posterior cells (8 of 16 clones; Figure  6). This modified segregation behavior indicates that anterior cells lacking both Ihog and Boi sort out from the anterior compartment into posterior territory. These clones do not readily integrate into posterior territory because they do not possess affinity for posterior cells conferred by expression of Engrailed [35]. Consistent with the lack of a known role for Hh signaling in the posterior compartment, double mutant posterior clones respected the compartment boundary and did not sort into anterior territory (11 of 11 clones; Figure 6A'). Together, these results indicate that, like Smo or Ci, Ihog and Boi are required to maintain cell affinities in the anterior compartment.

Ihog and Boi are essential components of the Hh pathway
Using genetic approaches in Drosophila, we show that ihog and boi act redundantly and are required for viability. Mutation of ihog and boi in the developing eye disc prevents Hh signaling and causes severe disruption of photoreceptor differentiation. Similarly, mutation of ihog and boi in the developing wing disc completely abrogates Hh signaling in a cell-autonomous manner. All these phenotypic effects are identical to mutations of the essential Hh signaling mediator smo and thus indicate that the transmembrane proteins Ihog and Boi are absolutely required for the Hh signaling pathway. We surmise that the essential role for ihog and boi in Hh signaling was not discovered previously because they are redundant with one another, and therefore refractory to conventional forward genetic screens in Drosophila.
A previous study of the lethal mutation ihog KG reported defects that mimicked those caused by mutations of the Hh signaling pathway [9]. That ihog KG disrupts gene(s) in addition to ihog is indicated by the fact that zygotic mutation of ihog KG alone is lethal, while the null allele ihog DC1 is not. ihog KG was reported to disrupt Ptc expression in the wing disc, and the patterning of wing veins [9], but we did not observe these effects with ihog DC1 null mutations ( Figure 4A'', B''). In addition, germline clones of ihog KG were found to disrupt patterning of the embryonic cuticle [9]. Therefore, it was unexpected that our maternal/zygotic ihog DC1 null mutants survived at least to pupal stages, and that there were no overt Hh-like phenotypes in the occasional flies that managed to escape the lethal phase and reach adulthood. Indeed, such flies looked remarkably normal. Perhaps the phenotypes observed in ihog KG are rare, or due to the disrupted function of the neighboring gene CG10158 or another gene in addition to ihog.
In Drosophila, genetic experiments have suggested that Ptc is the Hh receptor [38] and subsequent work has shown Ptc co-localization with Hh in S2 cells [39]. This has led to the model that Ptc binds Hh, though experiments have yet to demonstrate direct physical contact. In vertebrates, expression of the Ptc ortholog Ptc1 in cells promoted Sonic hedgehog (Shh) binding, suggesting that Ptc1 is a receptor for Shh [40,41]. However, as noted by the authors in those studies, Shh could use additional receptors and Ptc could affect their ability to bind to Shh. Addressing the important question of whether Ptc/Ptc1 bind directly to Hh/Shh is likely to require purified proteins in cell-free systems, a challenge for transmembrane proteins with complex topologies like Ptc. Interestingly, Ihog and Boi have been shown to bind directly to Hh [9,10]. Here we found that clones mutant for both ihog and boi phenocopied the effects of smo mutations for Hh signaling. Thus, our data demonstrate that Ihog and Boi are essential components of the Hh pathway. Coupled with our finding that Ihog and Boi act upstream of Smo and sequester Hh activity, this raises the possibility that Ihog and Boi are Hh-binding components of the Hh receptor complex. While preparing this manuscript, another research group came to a similar conclusion [42]. Using biochemical experiments, they further suggest that Ihog/Boi, together with Ptc and Hh, form a complex and that the presence of Ihog/ Boi in this complex is essential to allow Ptc to bind to Hh. However, since purified proteins were not used in these experiments, it remains to be determined whether the interaction between Hh and Ptc is direct. Alternatively, Ptc could affect the ability of an additional protein (perhaps Ihog or Boi) to recruit Hh. Thus, the question as to whether Ptc binds directly to Hh, in the absence or the presence of Ihog/Boi, remains to be elucidated. Additional experiments will be required to understand the exact contributions of Ihog, Boi and Ptc to the reception and transduction of the Hh signal. Nonetheless, our experiments and those of Zheng and colleagues identify Ihog and Boi as essential components of a Hh receptor complex.
In addition to Ihog and Boi, other membrane-tethered Hh-binding proteins, such as the glypican family members, have been identified [8,[43][44][45]. Although it is possible that other developing tissues may use different Hhbinding molecules as sensors for Hh, our data indicate that Ihog and Boi are critical to elicit Hh signaling and that, at least in developing imaginal tissues, no other molecule can compensate for loss of Ihog and Boi.
Ihog and Boi: transmembrane IgSF proteins with potential to mediate cell affinity In addition to the role of Ihog and Boi in wing cell fate specification (Zheng et al. [42] and this study), we show for the first time that Ihog and Boi are required for maintenance of the antero-posterior boundary of the wing disc. Formation of this boundary is thought to involve cell-surface recognition molecules responsible for affinity among anterior cells. Since ci mutant clones lose affinity for the anterior compartment [37], it is likely that such adhesion molecules are transcriptionally regulated by the Hh pathway. Despite broad screening efforts to identify these molecules [46], they remain unidentified. While the loss of cell affinity in boi;ihog double mutant clones is likely a consequence of loss of Hh signaling, it is an intriguing possibility that Ihog and/or Boi could also be Ci-regulated recognition molecules promoting the differential compartment-specific cell affinity. Consistent with this hypothesis, the vertebrate orthologs of Ihog and Boi, called Cdon (Cell adhesion molecule-related/down-regulated by oncogenes) and Boc (Brother of Cdon), can interact with each other and form higher-order cell-surface complexes with the wellcharacterized adhesion molecules N-and M-cadherin [7,47]. These complexes mediate cell-cell interactions between muscle precursor cells and are enriched at sites of contact between myoblasts. Similar to Cdon and Boc, Ihog and Boi could mediate cell-cell interactions, and might play a direct role in compartment-specific cell affinity.

Ihog and Boi and their vertebrate orthologs in nervous system development
Although Cdon and Boc have also been shown to play a role in Hh signaling in vertebrates [9,[48][49][50], their absolute requirement for Hh signaling has not been demonstrated, and mouse embryos mutant for both cdon and boc do not display a smo-like phenotype (our unpublished data). Additionally, Hh binds to Ihog/Boi in a manner that is different from Shh binding to Cdon and Boc [10,51], suggesting that a direct extrapolation of our work in flies to Cdon and Boc could be subject to added complexities.
Nevertheless, there are enough interesting parallels between Ihog and Boi and their vertebrate orthologs in the developing nervous system to warrant further investigation. We have shown here that Ihog and Boi are essential for the differentiation of R-cells in the developing fly visual system. Interestingly, overexpression studies in chick have shown that Boc and Cdon promote Shhinduced differentiation in the neural tube and that Cdon is required for Shh-dependent cell fate specification in the neural tube [48,50]. Furthermore, Boc has been shown to be essential for axon guidance through the activation of a non-canonical, transcription-independent Shh signaling pathway mediated by the activity of Src-family kinases [49,52,53]. It will be interesting to determine whether Ihog and Boi also function in axon guidance and whether they may contribute to non-canonical Hh signaling in Drosophila.

Conclusions
We conclude that the transmembrane proteins Ihog and Boi are essential components of the Hh signaling pathway. Ihog and Boi act upstream of Smo and sequester Hh activity, raising the possibility that Ihog and Boi are Hhbinding components of the Hh receptor complex. These findings advance our understanding of the molecular signaling mechanism underlying the function of Hh.

Materials and methods
Fly stocks ihog DC1 was generated by Flippase (FLP)-mediated recombination between Flippase recognition target (FRT) sites in pBac{RB}CG10158 e02576 and pBac{RB} iHog e02142 [54,55]. After out-crossing to w 1118 , the resulting chimeric pBac element was excised, and the predicted deletion of 3,479 bp was confirmed by DNA sequencing and RT-PCR ( Figure 2D). A deficiency stock that uncovers the ihog locus (Df(2L)Exel7029) was also used, and is designated ihog Df herein.
boi C1 was identified in a screen for mutations that suppress ectopic wing veins caused by ap GAL4 -driven expression of P{EP}EP1447, a UAS-based P-element situated upstream of the boi gene ( Figure S1A). Briefly, starved EP1447 male flies were exposed to 25 mM ethylmethanesulfonate (EMS) overnight, then crossed to ap GAL4 virgin females. boi C1 was identified among approximately 15,000 adult progeny screened for suppression of ectopic wing veins. boi C1 proved to be a viable mutation, and so PCR and DNA sequencing were used to compare the boi coding region in boi C1 homozygotes with the parental line EP1447. An induced G-to-A transition was found in boi C1 , resulting in a nonsense mutation at Trp626.
To create UAS-ihog::myc, the entire coding sequence of ihog (minus stop codon) was PCR-amplified from an ihog cDNA (GH03927), and cloned into pCR2.1-TOPO (Invitrogen, Carlsbad, CA, USA). This fragment was excised and cloned into p5MT2Stp (BamHI/SalI) to add five copies of the myc epitope to the carboxyl terminus. UAS-ihog::myc was generated by KpnI/NotI excision and cloning into pUASt. Transgenic flies were generated in a w-background by standard microinjection procedures (BestGene Inc., San Diego, CA, USA).