Post-translational processing and renal expression of mouse Indian hedgehog.

The full-length mouse Indian hedgehog (Ihh) cDNA was cloned from an embryonic 17.5-day kidney library and was used to study the post-translational processing of the peptide and temporal and spatial expression of the transcript. Sequence analysis predicted two putative translation initiation sites. Ihh translation was initiated at both initiation sites when expressed in an in vitro transcription/translation system. Expression of an Ihh mutant demonstrated that the internal translation initiation site was sufficient to produce the mature forms of Ihh. Ihh post-translational processing proceeded in a fashion similar to Sonic and Drosophila hedgehog; the unprocessed form underwent signal peptide cleavage as well as internal proteolytic processing to form a 19-kDa amino-terminal peptide and a 26-kDa carboxyl-terminal peptide. This processing required His313 present in a conserved serine protease motif. Ihh transcript was detected by in situ RNA hybridization as early as 10 days postcoitum (dpc) in developing gut, as early as 14.5 dpc in the cartilage primordium, and in the developing urogenital sinus. In semiquantitative reverse transcription-polymerase chain reaction experiments, Indian hedgehog transcript was first detected in the mouse metanephros at 14.5 dpc; transcript abundance increased with gestational age, becoming maximal in adulthood. In adult kidney, Ihh transcript was detected only in the proximal convoluted tubule and proximal straight tubule.

The full-length mouse Indian hedgehog (Ihh) cDNA was cloned from an embryonic 17.5-day kidney library and was used to study the post-translational processing of the peptide and temporal and spatial expression of the transcript. Sequence analysis predicted two putative translation initiation sites. Ihh translation was initiated at both initiation sites when expressed in an in vitro transcription/translation system. Expression of an Ihh mutant demonstrated that the internal translation initiation site was sufficient to produce the mature forms of Ihh. Ihh post-translational processing proceeded in a fashion similar to Sonic and Drosophila hedgehog; the unprocessed form underwent signal peptide cleavage as well as internal proteolytic processing to form a 19-kDa amino-terminal peptide and a 26-kDa carboxyl-terminal peptide. This processing required His 313 present in a conserved serine protease motif. Ihh transcript was detected by in situ RNA hybridization as early as 10 days postcoitum (dpc) in developing gut, as early as 14.5 dpc in the cartilage primordium, and in the developing urogenital sinus. In semiquantitative reverse transcription-polymerase chain reaction experiments, Indian hedgehog transcript was first detected in the mouse metanephros at 14.5 dpc; transcript abundance increased with gestational age, becoming maximal in adulthood. In adult kidney, Ihh transcript was detected only in the proximal convoluted tubule and proximal straight tubule.
Mouse Indian hedgehog (Ihh) 1 is a member of a multigene family that includes Hedgehog (Drosophila) and its vertebrate homologue, Sonic hedgehog (1,2). The Hedgehog proteins are secreted extracellular signals that communicate with neighboring cells to regulate the production of patterning molecules such as Wingless and Decapentaplegic (3,4). Hedgehog (Hh) regulates the anterior-posterior patterning of the imaginal disc structures in Drosophila, while Sonic hedgehog (Shh) carries out a similar function in the vertebrate limb in concert with the fibroblast growth factor-4 (5)(6)(7)(8)(9)(10)(11). In addition, Shh affects the dorsoventral patterning of the mouse neural tube and somites leading to the induction of floor plate cells, motor neurons, and sclerotome (12)(13)(14)(15). Both Hh and Shh undergo autoproteolytic cleavage to generate a functional amino-terminal peptide, with inducing activity, and a carboxyl-terminal peptide that can tether the precursor protein to the cell membrane (14 -20).
Two additional vertebrate hedgehog genes, Desert hedgehog (Dhh) and Ihh have also been identified (2). For Ihh, partial cDNA sequences are available for human and mouse with expression reported in the embryonic lung of human, the developing gut and cartilage of chick, and the adult kidney in both mouse and human (2,(21)(22)(23). In the developing cartilage, Ihh is produced by proliferating chondrocytes of the prehypertrophic growth zone and signals to the surrounding perichondrium to induce the release of parathyroid hormone-related protein (PTHrP) (24,25). Once bound to its receptor in the undifferentiated chondrocytes, PTHrP blocks cells from entering the hypertrophic pathway. Thus, this Ihh/PTHrP feedback loop can regulate chondrocyte differentiation to balance the growth and ossification of long bones. While the complete chick Ihh cDNA sequence has been reported recently, the post-translational processing of the peptide has not been described.
In this report the entire coding region of the mouse Ihh cDNA was identified, its protein products analyzed, and the embryonic expression pattern determined. The mouse Ihh cDNA contains two putative translation initiation sites, unique to vertebrate hedgehog family members. Ihh undergoes proteolytic processing like Hh and Shh. Ihh transcript expression in the developing gut and in the growth zone of cartilage of developing long bones was confirmed by in situ RNA hybridization. In addition, Ihh transcript was detected as early as 14.5 dpc in the developing mouse kidney and its abundance increased with gestational age. Furthermore, the Ihh transcript localized to the proximal convoluted and proximal straight tubule in the adult kidney.

MATERIALS AND METHODS
Animals-CD-1 mice (Charles River Laboratories) were bred to obtain mouse embryos. Detection of a vaginal plug was used to define the first day of gestation. At the appropriate time point, pregnant females were sacrificed and kidneys or other organs were dissected from isolated embryos and rapidly frozen in liquid nitrogen.
In a separate experiment, 10 g of total RNA extracted from 16.5 dpc mouse embryonic tissues was hybridized with the same random-labeled 32 P-Ihh cDNA probe as described above. Ethidium bromide staining of the gel was used to assess the equivalence of RNA loaded in each lane.
Molecular Cloning-Using synthetic oligonucleotide primers complementary to the published partial cDNA sequence of mouse Indian hedgehog (GenBank™ accession number X76291), the PCR was used to amplify an Ihh cDNA fragment from a 17.5 dpc mouse kidney cDNA library. The amplification product was DNA sequenced to assure Taq DNA polymerase fidelity to the previously published partial Ihh cDNA sequence. The PCR-amplified 1.1-kb Ihh cDNA fragment was radiolabeled and used to screen approximately 2.5 ϫ 10 5 recombinants from an oligo(dT)-primed ZAPII mouse embryonic 17.5-day kidney cDNA library. Filters were hybridized and washed as described previously (26). Six positive clones were identified, purified to homogeneity, in vivo excised according to the protocol of the manufacturer (Stratagene), and restriction-mapped. The two longest clones were sequenced along both strands over their entire length. DNA sequencing was performed using the dideoxy nucleotide termination method. Sequenase T7 DNA polymerase (U. S. Biochemical Corp.), its reagent kit, and synthetic oligonucleotides were employed according to the directions of the manufacturer. Gel compressions were resolved with 7-deaza-dGTP and/or formamide. DNA sequence was assembled and analyzed using Sequence Analysis Software Package of the Genetics Computer Group (version 8.1).
Eukaryotic Expression Constructs-The full-length Ihh cDNA (2103 bp) was excised from pBluescript (pBS-Ihh) and subcloned into a cytomegalovirus promoter-based eukaryotic expression vector pcDNA3 (Invitrogen, San Diego, CA) and designated pcDNA3-Ihh. Next, an expression Ihh cDNA plasmid was prepared that deleted the domain NH 2terminal to Met 39 and added a COOH-terminal Myc-epitope tag (EQKLISEEDL) (designated Ihh-Met 39 -Myc). The PCR was used to amplify a cDNA fragment that encoded a 5Ј HindIII restriction site, Ihh nucleotides 289 -1530 including the putative Ihh Kozak's consensus preceding Met 39 , the Ihh cDNA open reading frame, the Myc-epitope, a stop codon, and a 3Ј XbaI site. Synthetic oligonucleotides used included 5Ј-[ACACAAGCTTTACCCGGCCATGTCTCCC]-3Ј for the sense primer and 5Ј-[GGTCTAGATCACAAGTCCTCCTCCGAGATCAATTTCTGCT-CGCTTCCTGCCCCAGACATGCCCAGT]-3Ј for the antisense primer. The amplified cDNA product was prepared and ligated directly into the eukaryotic TA cloning vector pCR 3.1-Uni (Invitrogen) according to the manufacturer's instructions. The subcloned DNA amplification product was bidirectionally sequenced.
Site-directed mutagenesis was performed using the unique site elimination method (U.S.E. mutagenesis kit, Pharmacia Biotech Inc.). A synthetic oligonucleotide, 5Ј-[CTCACGCCTGCCGCCCTGCTCTTCAT-T]-3Ј, that substitutes a GC for CA at nucleotides 1120 and 1121 of the Ihh cDNA was used with the ScaI/MluI U.S.E. selection/toggle primer: 5Ј-[CTGTGACTGGTGACGCGTCAACCAAGTC]-3Ј (Pharmacia) to mutate the pcDNA3-Ihh plasmid as directed by the manufacturer. The isolated plasmid, designated pIHH-H313A, was bidirectionally sequenced over the region of the mutation to confirm the desired mutation.
In Vitro Expression of Ihh cDNA-Using an in vitro transcriptiontranslation reticulocyte lysate assay (TNT-Promega Corp., Madison, WI), 1.0 g of pcDNA3 alone or the indicated Ihh expression constructs were transcribed in vitro with T7 RNA polymerase and translated in a reaction mix containing 12.5 l of reticulocyte lysate and 20 Ci of [ 35 S]methionine (Amersham Corp., SJ1015) according to the manufacturer's recommendations. Where indicated, 2.5 l of canine pancreatic microsomes (Boehringer Mannheim) were added to the reaction mix. The reactions were incubated at 30°C for 90 min. Five microliters of each reaction mix was reduced and resolved on a 12.5% SDS-polyacrylamide gel electrophoresis (PAGE). The gel was fluorinated in Entensify (DuPont NEN) according to the manufacturer's instructions before autoradiography.
Bacterial Fusion Protein Expression and Generation of Antisera-The bacterial expression vector pGEX-KT and the host Escherichia coli TG1 (gifts of Dr. J. Dixon, University of Michigan) were used to produce an NH 2 -terminal Ihh glutathione S-transferase fusion protein, GST-Ihh-N. A PCR-amplified nucleotide fragment encoding amino acids 81-140 of Ihh and engineered with BamHI (5Ј) and EcoRI (3Ј) restriction sites, so as to maintain the appropriate reading frames, was subcloned into the pGEX vector. Synthetic oligonucleotides used included 5Ј-[ATTGGATCCAACCTCGTGCCTCTTGCCTAC]-3Ј as the sense primer and 5Ј-[ATTGAATTCTCAAGGCGGTCGGCACCCGTGTTC]-3Ј as the antisense primer. The construct was sequenced along both strands to assure Taq polymerase fidelity and maintenance of the appropriate reading frame. A 32-kDa fusion protein was expressed and purified as described previously (26). Two rabbits were immunized at 4-week intervals with 100 g of fusion protein by intramuscular injection.
Expression and Analysis of Eukaryotic Constructs-COS-7 cells (ATCC) were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, penicillin, and streptomycin. Two ϫ 10 5 cells plated on a 35-mm tissue culture Petri dish (Falcon) were grown overnight, then transiently transfected with 2 g of the appropriate eukaryotic expression construct mixed with 6 l of Lipo-fectAMINE (Life Technologies, Inc.) according to the manufacturer's instructions. After 48 h, the conditioned medium was removed and centrifuged for 15 min at 16,000 ϫ g in a microcentrifuge at 4°C. The cell layer was washed twice with ice-cold phosphate-buffered saline, then treated with lysis buffer plus protease inhibitors. Five-hundred microliters of lysis buffer (RIPA: 50 mM Tris, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.1% SDS, 0.5% deoxycholate, plus protease inhibitors) was added and allowed to incubate for 3 min before shearing with a 22.5-gauge needle four times and centrifuging for 15 min at 16,000 ϫ g in a microcentrifuge at 4°C. Protease inhibitors consisted of 50 g/ml phenylmethylsulfonyl fluoride, 1 g/ml leupeptin, 1 g/ml pepstatin, and 1 g/ml aprotinin.
For immunoblotting, 40 l of cell lysate or 10 l of conditioned medium was reduced and separated by 12.5% SDS-PAGE. Proteins were transferred to nitrocellulose and immunoblotted as described previously (26). The primary antibodies were diluted 1:500 in TBS-T (Trisbuffered saline ϩ 0.1% Tween 20) for rabbit polyclonal antiserum. The Myc monoclonal antibody (9E10) was diluted to a final concentration of 6 g/ml in TBS-T. Secondary antibody consisted of horseradish peroxidase-conjugated goat anti-rabbit IgG or goat anti-mouse IgG (Bio-Rad). Blots were developed using the ECL chemiluminescent reagent (Amersham).
Reverse Transcription-PCR-Adult rat kidneys were harvested, collagenase-treated, and the nephrons were microdissected as described previously (27). These nephrons were further sectioned by morphology into separate segments consisting of the distal convoluted tubule, outer medullary collecting duct, proximal straight tubule, inner medullary collecting duct, proximal convoluted tubule, thick ascending limb, macula densa containing segment, and glomerulus. Total RNA from these segments was isolated in a commercially prepared phenol-4 M guanidine isothiocyanate reagent (TRI-Reagent; Molecular Research Center, Cincinnati, OH) and was reverse-transcribed using an oligo(dT) primer. Samples were incubated at 65°C for 5 min, then ramped to 37°C over 5 min. After 5 min at 37°C, 100 units of Maloney murine leukemia virus reverse transcriptase (Life Technologies, Inc.) was added to each reaction, except the reverse transcriptase negative control, and samples were incubated for an additional 55 min at 37°C. The cDNA generated by reverse transcription was used as a template for the PCR as described (27). Sense (IHH-B) (5Ј-[GTCTCTTGCTAGAAGAGA]-3Ј) and antisense (IHHT7Ј) (5Ј-[GCCTGCAGGGAAGGTCAT]-3Ј) synthetic oligonucleotides corresponding to the 3Ј-untranslated region of mouse Ihh were used in the PCR. Samples were denatured at 94°C for 3.5 min then cycled 30 times for 1 min at 94°C, 58°C for 1 min, and 72°C for 2 min. Final extension was carried out at 72°C for 8 min. To control for variation in tissue mass and reverse transcription reaction efficiency, cDNA concentration was adjusted so that equal aliquots yielded constant amounts of PCR amplification product for ␤-actin as described previously (28).
To assess expression of the cloned products, PCR was performed using serial 10-fold dilutions of cDNA from each time point. As a positive control, a known copy number of plasmid DNA containing specific cloned fragments were serially diluted and used in PCR reactions. The cDNA reaction mix from reverse transcriptase minus reactions was also included in each PCR run; these samples were consistently negative. To assess product abundance, 12 l of each PCR reaction was separated by electrophoresis in a 5% acrylamide gel.
Metanephric kidney from timed pregnant CD-1 mice were obtained by dissection of embryos at 11.5, 14.5, and 17.5 dpc. Additionally, mouse kidney tissue was harvested from newborn and adult mice. Tissues were snap-frozen in liquid nitrogen in TRI-Reagent as described above. Isolated total RNA was reverse-transcribed and subjected to the PCR as described above. Oligonucleotides were selected to minimize sequence similarity to other hedgehog family members. In control experiments, unlabeled amplified product was sequenced to ensure sequence identity with Indian hedgehog.
In Situ RNA Hybridization-A BamHI digest of pBS-Ihh was performed to eliminate the 5Ј-untranslated region and the coding region of Ihh. Religation of this construct resulted in a 540-bp fragment from the 3Ј-untranslated region of Ihh within pBluescript and designated pBS-Ihh-3ЈUT. Single-stranded RNA probes used in hybridizations were transcribed in vitro from appropriately linearized pBS-Ihh-3ЈUT template using either T3 (sense) or T7 (antisense) RNA polymerase and labeled with [ 33 P]uridine 5Ј-triphosphate. After DNase digestion, probes were precipitated with 10% trichloroacetic acid and collected on nitrocellulose filters (Millipore). Probes were eluted from the filters in 20 mM EDTA, pH 8.0, 0.1% SDS at 65°C, ethanol-precipitated, and partially degraded with 0.2 N NaOH on ice for 30 -60 min. Following neutralization with 1 M acetic acid, probes were reprecipitated with ethanol and resuspended in 50% formamide, 10 mM dithiothreitol. This 0.9-kb Ihh cDNA was prepared to correspond to the 3Ј-untranslated region of Ihh to minimize hybridization to other hedgehog family members. In situ RNA hybridization was carried out as described previously (29). Briefly, frozen, paraformaldehyde-fixed 8-m sections were digested in buffer containing 0.125 mg/ml proteinase K for 10 min at room temperature. Hybridization was done overnight at 50°C in buffer containing 50% formamide, 0.3 M NaCl, 10 mM Tris, 10 mM NaPO 4 , pH 6.8, 5 mM EDTA, 1 ϫ Denhardt's solution, 10% dextran sulfate, 10 mM dithiothreitol, and 1 mg/ml tRNA. Posthybridization washes were performed with a final stringency of 50% formamide, 2 ϫ SSC at 37°C, followed by RNase digestion. Hybridized sections were dipped into Kodak NTB-1 photographic emulsion and photographed as described previously (29). The Indian hedgehog primary peptide sequence is shown aligned with representative hedgehog family members in Fig.  1. The largest predicted Ihh open reading frame encodes a polypeptide sequence of 449 amino acids with a predicted relative molecular mass of 49 kDa. The Indian hedgehog cDNA contains an in frame stop codon followed by two AUGs near its 5Ј end that are favorable for translation initiation as defined by Kozak (30). These AUGs begin at nucleotide 184 (Met 1 ) and nucleotide 298 (Met 39 ), respectively. If both putative translation initiation sites are utilized, the Ihh cDNA would encode a 49-kDa and a 45-kDa protein product. Immediately following Met 39 is a 27-amino acid sequence which closely fits von Heijne's consensus for a signal peptide (31). Comparison with this consensus predicts that the signal peptide cleavage likely follows the (Ϫ3, Ϫ1) rule and occurs preceding Cys 66 . Cleavage of this signal peptide would yield a mature protein of 384 amino acids with a predicted relative molecular mass of 42 kDa. Primary sequence alignment reveals that mouse Ihh shows 89% identity and 96% similarity to mouse Shh over the NH 2terminal region, from the end of the signal peptide to the proteolytic processing site following Cys 241 .

Isolation and Molecular Cloning of Full-length Mouse
Analysis of in Vitro Ihh Expression and Processing- Fig. 2A outlines the potential Ihh precursor proteins and their proteolytic cleavage products. To test whether both Met 1 and Met 39 could function as translation initiation sites, the 2103-bp fulllength Ihh cDNA (pcDNA3-Ihh) was translated in a rabbit reticulocyte lysate. Two predominant protein products of 49 and 45 kDa were expressed (Fig. 2B, lane 3). Pancreatic microsomes were added to the reactions to study post-translational modifications such as signal peptide cleavage and core glycosylation. Expression of pcDNA3-Ihh in the presence of micro-somes yielded protein products of 49, 45, and 42 kDa (Fig. 2B,  lane 4) and additional smaller peptides (discussed below). To test whether the 45-kDa protein product initiated at Met 39 , the Ihh cDNA was deleted from its 5Ј end to nucleotide 289, thereby eliminating Met 1 but preserving the Kozak's consensus sequence surrounding Met 39 . In addition, this construct (pIhh-Met 39 -Myc) encoded a Myc-epitope tag placed immediately 5Ј to the stop codon. When programmed into the reticulocyte lysate system, pIhh-Met 39 -Myc yielded a 46-kDa protein (Fig. 2B,  lane 7). In the presence of microsomes, a 43-kDa protein was also observed (Fig. 2B, lane 8). These proteins had slightly reduced electrophoretic mobility than predicted since the 10amino acid Myc-epitope tag added 1 kDa to their relative molecular masses. When compared with the protein products observed with expression of the full-length Ihh cDNA, it was noted that the 49-kDa band was not observed. Taken together, these results suggested that both Met 1 and Met 39 function as translation initiation sites when expressed in vitro.
Sequence analysis showed that Ihh contains a conserved serine protease motif as well as a conserved proteolytic cleavage motif identified in other Hedgehog species, Gly-Cys-Phe (16, 18) (Fig. 1). In addition to the unprocessed Ihh proteins, pcDNA3-Ihh in vitro translation yielded protein products of 26, 23, and 22 kDa in the absence of microsomes (Fig. 2B, lane 3). This was consistent with the Ihh-N**, Ihh-C*, and Ihh-N* products predicted ( Fig. 2A). When microsomes were added, some persistence of these products was seen. Additionally, a 25-kDa band and a less intense 19-kDa band were observed (Fig. 2B, lane 4). Upon addition of microsomes, core glycosylation of Ihh-C* should result in an increase in the molecular mass from 23 to 25 kDa; the appearance of the protein product designated Ihh-C (Fig. 2B, lane 4) was consistent with this prediction. Moreover, following addition of microsomes, signal peptide cleavage from Ihh-N** or Ihh-N* resulted in a protein of 19 kDa, predicted to be Ihh-N (Fig. 2B, lane 4). Expression of Ihh-Met 39 -Myc in vitro resulted in protein bands at 22, 24, and 46 kDa. These bands were consistent with Ihh-N*, Ihh-C*, and Ihh*, the size of the latter two products were increased slightly by the Myc-epitope tag. Also in the absence of microsomes, no protein was seen at 26 kDa which was consistent with the disappearance of Ihh-N**. In the presence of microsomes, the smaller products included the aforementioned 22-and 24-kDa proteins in addition to a band at 26 kDa and a less intense band at 19 kDa. These latter two protein bands were believed to represent Ihh-C and Ihh-N after post-translational modifications, respectively. Core glycosylation of Ihh-C* at Asn 320 could explain a 2-kDa increase in molecular mass seen in Ihh-C. Likewise, signal peptide cleavage of Ihh-N* would explain the presence of a 19-kDa protein, Ihh-N.
Substitution of alanine for the conserved histadine residue (His 313 ) within the serine protease motif of full-length Ihh (pIhh-H313A) yielded protein products of 49 and 45 kDa (Fig.  2B, lane 5). The 42-kDa protein lacking its signal peptide was also observed with the addition of microsomes (Fig. 2B, lane 6). However, the more rapidly migrating peptide bands were not observed either in the presence or absence of microsomes. Therefore, the 26-, 23-, and 22-kDa proteins required His 313 within the serine protease motif for their generation. Like Shh and Hh, this suggests that Ihh undergoes internal proteolytic processing dependent upon the presence of an autoproteolytic serine protease domain (16,17).
Generation of Ihh-N Polyclonal Antiserum-To better identify the Ihh protein fragments and characterize Ihh processing in mammalian cells, a rabbit polyclonal antiserum was raised to a bacterially expressed NH 2 -terminal Ihh fusion protein.
Ihh-N polyclonal antiserum recognized glutathione S-transfer-ase (GST only) and GST-Ihh-N-terminal peptide fusion protein (amino acids 83-140, GST-Ihh-N) by immunoblotting (Fig. 3). Preadsorption of the Ihh-N antiserum with a molar excess of GST essentially eliminated the detection of GST but failed to block the detection of GST-Ihh-N, demonstrating that this Ihh-N antiserum specifically recognized the Ihh-N peptide.
Expression of Ihh in Mammalian Cells-COS-7 cells were transiently transfected with the full-length Ihh construct (pcDNA3-Ihh), pIhh-Met 39 -Myc, pIhh-H313A, and vector controls, to determine if Ihh was processed in mammalian cells in a manner similar to that observed in the in vitro transcription/ translation system. Following transfection with pcDNA3-Ihh, proteins of 42 and 19 kDa were detected in the cell layer by immunoblotting with the Ihh-N antiserum. The 42-kDa band corresponded to the intact Ihh lacking the signal peptide, similar to the in vitro translated product following addition of microsomes (Fig. 2B, lanes 4, 6, and 8). The 19-kDa protein corresponded to the Ihh-N after proteolytic cleavage, as observed in vitro (Fig. 2B, lanes 4 and 8). The 19-kDa NH 2terminal peptide was found associated only with the cell layer and was not detected in the conditioned medium. Consistent with observations made in the in vitro transcription/translation system, expression of pIhh-H313A in COS-7 cells produced only a 42-kDa protein as detected by immunoblotting (Fig. 4B,  lane 4). No 19-kDa protein was detected, consistent with failure of proteolysis of Ihh into its NH 2 -and COOH-terminal peptides. COS-7 cells transiently transfected with pIhh-Met 39 -Myc expressed Ihh products with electophoretic mobility similar to those expressed by pcDNA3-Ihh transfected cells. When an aliquot of the cell layer of pIhh-Met 39 -Myc-transfected cells was resolved by SDS-PAGE and immunoblotted with Ihh-N antiserum, 43-and 19-kDa peptides were detected (Fig. 4B, lane 3). The 43-kDa protein had a slightly reduced electrophoretic mobility relative to that derived from pcDNA3-Ihh transfected cells due to the presence of its Myc epitope tag. The Ihh-N protein (19 kDa) had identical mobility to that derived from pcDNA3-Ihh.
To attempt to detect the expression of the COOH-terminal Ihh peptide, COS-7 cells were transiently transfected with an expression vector encoding COOH-terminal epitope-tagged Ihh (pIhh-Met 39 -Myc). An aliquot of conditioned medium or lysed cell layer was separated by SDS-PAGE, transferred to nylon, and immunoblotted with a Myc-epitope monoclonal antibody 9E10. Only the 43-kDa unprocessed form of Ihh was detected in the cell layer. No signal corresponding the Ihh-C was detected in the conditioned medium (data not shown). In similar experiments, pIhh-Met 39 -Myc-transfected COS-7 cells were metabolically labeled and cell layer and conditioned medium immunoprecipitated with 9E10. Again, a product corresponding in electrophoretic mobility to Ihh-C was not detected in either fraction (data not shown).
Ihh Is Expressed in Developing Stomach, Gut, Kidney, and FIG. 2. Proteolytic processing of Ihh. A, schematic diagram of proposed Ihh processing as predicted from amino acid sequence. The unprocessed 449-amino acid Ihh protein has a predicted relative molecular mass of 49 kDa (Ihh**). Translation initiation beginning at an internal AUG, A 298 , should yield a 411-amino acid Ihh protein of 45 kDa (Ihh*). Signal peptide cleavage from either Ihh** or Ihh* results in a 42-kDa product (Ihh). Predicted internal cleavage at the conserved proteolytic cleavage site (Gly 240 -Cys 241 ) of Ihh** should yield an Ihh-C* fragment (209 amino acids) of 23 kDa and an Ihh-N** peptide (240 amino acids) with a predicted molecular mass of 26 kDa. Cleavage of Ihh* at this same proteolytic cleavage site should produce the same Ihh-C* fragment with a smaller 22-kDa NH 2 -terminal fragment, Ihh-N*. The Ihh-N fragment represents Ihh-N** or Ihh-N* devoid of its signal peptide and should be 19 kDa in size. The predicted Ihh-C fragment represents an Ihh-C* fragment that has undergone glycosylation at Asn 320 and increased its apparent molecular mass to 26 kDa. Ab-N indicates the 58-amino acid epitope to which Ihh-N polyclonal antiserum is directed. B, Ihh processing in an in vitro transcription/ translation reticulocyte lysate system. A transcription and translation reticulocyte lysate system containing [ 35 S]methionine was programmed with pcDNA3, pcDNA3-Ihh, pIhh-Met 39 -Myc, or pIhh-H313A. As indicated, lysates were programmed with or without Ihh transcript in the presence (lanes 2, 4, 6, and 8) or absence (lanes 1, 3, 5, and 7) of pancreatic microsomes. Products were resolved on a 12.5% SDS-PAGE under reducing conditions. Relative molecular mass standards are shown at left of each gel. Liver in the 16.5 dpc Mouse Embryo-To better localize the Ihh transcript during embryonic mouse development, in situ RNA hybridization was performed using 33 P-labeled riboprobes transcribed from an Ihh cDNA template. Serial section of 10, 12, 14.5, and 16.5 dpc mouse embryos were hybridized with Ihh sense or antisense probes. The control 33 P-labeled Ihh sense riboprobe produced no detectable signal. However, the antisense Ihh riboprobe could detect expression in the gut at all embryonic time points investigated (Fig. 5, light and dark fields, top left). Higher power examination of the light field demonstrated localization in the luminal portion of the gut (Fig. 5, top right). Ihh transcript was also detected in the cartilage primordium of the cervical and tail vertebrae at 14.5 dpc (Fig. 5, top left). Closer inspection of the cartilaginous structures of the tail and cervical vertebrae suggested that Ihh is expressed in chondrocytes of these developing bones (Fig. 5,  bottom left and right). Less prominent Ihh transcript expression existed in the region of the developing urogenital sinus at 14.5 dpc (Fig. 5, top left). It should be noted that no expression of Ihh was detected in the developing mouse kidney at any embryonic time point, at least at the level of sensitivity of in situ hybridization.
Northern blot of total RNA extracted from multiple 16.5 dpc embryonic mouse organs was probed with a 1.1-kb Ihh cDNA probe. Ihh transcript expression was detected in the developing stomach and gut with lower levels of expression in the kidney and liver (Fig. 6).
Ihh Transcript in Kidney Increases with Developmental Maturation-Although Ihh transcript was undetectable in the embryonic kidney by in situ hybridization, data presented above clearly indicated its renal expression at various time points of renal development. Northern analysis demonstrated Ihh tran-script expression in the metanephric kidney at 16.5 dpc, its cDNA was cloned from a 17.5 dpc kidney library, and its transcript was highly expressed on Northern blot in the adult kidney. To better investigate the renal expression of Ihh, semiquantitative reverse transcription-polymerase chain reaction (RT-PCR) was utilized to establish an embryonic time course of Ihh transcript expression in the metanephros. Total RNA was extracted from dissected mouse metanephric kidney at 11.5, 14.5, and 17.5 dpc in addition to newborn and adult kidney. Indian hedgehog transcript was first detected by amplification from the mouse metanephros at 14.5 dpc. Indian hedgehog transcript abundance increased with gestational age and was maximal in adulthood (Fig. 7).
Ihh Transcript Localizes to the Proximal Tubular Structures of the Adult Kidney-The RT-PCR was used to localize the Ihh transcript to a specific nephron segment in microdissected adult kidney. Nephrons from collagenase treated adult rat kidneys were microdissected by morphology into separate segments. Total RNA was isolated from these segments and subjected to RT-PCR. As seen in Fig. 8, Ihh transcript was detected only in the proximal convoluted tubule and proximal straight tubule and not in other portions of the adult nephron.

DISCUSSION
This report describes the cloning of the full-length cDNA for mouse Indian hedgehog, a mammalian hedgehog family member. Sequence analysis revealed that mouse Ihh contained two putative translation initiation sites, which were shown to be utilized in vitro, a feature not shared by chick Ihh or any other vertebrate hedgehog family member. Expression of mouse Ihh in vitro and in mammalian cells demonstrated proteolytic processing similar to Hh and Shh (16,17). Analysis of Ihh transcript expression demonstrated Indian hedgehog expression in both the embryonic and adult kidney. Surprisingly, Ihh transcript expression was demonstrated as early as 14.5 dpc and increased with developmental maturation of the kidney. In the adult kidney, Ihh transcript was present only in the proximal convoluted and proximal straight tubule and was not detected elsewhere in the nephron.
The mouse Ihh cDNA contains two putative translation initiation sites within a favorable context (30). Although both sites are utilized in vitro, the mature Ihh NH 2 -terminal peptide can be produced from precursor proteins initiating at either site. While the two translation initiation sites are not conserved among other vertebrate hedgehog genes, the Drosophila hedgehog gene does contain two translation initiation sites, which precede a hydrophobic signal peptide and are utilized in vitro. This "internal" signal peptide led to early speculation that Hh represented a Type II transmembrane protein (1). However, subsequent experiments have shown that Hh is a secreted protein (18). Based on functional data recently reported on chick Ihh, it appears that mouse Ihh is similarly a secreted protein.
A serine protease motif and an internal cleavage site are necessary for proteolytic processing in Hh and Shh (16 -18). Mouse Ihh undergoes similar processing events upon in vitro translation and upon expression in eukaryotic cells. The conserved His residue present in the serine protease motif is necessary for proteolytic cleavage of all hedgehog proteins examined to date. Using specific antiserum, the cleaved Ihh amino-terminal peptide could easily be detected in the cell layer of transfected mammalian cells. We were unsuccessful in identifying the Ihh-C species in either the cell layer or the cell medium using a Myc-epitope-tagged Ihh precursor and the monoclonal antibody 9E10. The stability of the COOH-terminal cleavage product is not known, although Lee  that Hedgehog may undergo an additional cleavage step at its COOH-terminal end when expressed in vivo but not in vitro. The increased molecular weight of the Myc-tagged Ihh translated in vitro would be consistent with absence of further proteolytic processing of the COOH terminus in this system. Terminal cleavage of Ihh-C in vivo could explain our inability to detect this fragment with the anti-Myc antibody in transfected mammalian cells.
In this study in situ RNA hybridization detected Ihh transcript in the developing gut epithelium, cartilage, and urogenital sinus in good agreement with previous reports (22). Ihh is expressed in the columnar epithelial cells lining the length of the intestine, including the rectum. In the duodenum at 14.5 dpc, Ihh is expressed in the more differentiated epithelial cells of the villi, whereas Sonic hedgehog is expressed in the more FIG. 5. In situ hybridization demonstrating Ihh expression in mouse developing cartilage, GI tract, and genitourinary tract. In situ RNA hybridization was performed using cryosectioned 10.5, 12, 14.5, and 16.5 dpc mouse embryos. Sections were hybridized with a 33 P-labeled sense or antisense riboprobe corresponding to the 3Ј-untranslated cDNA of Ihh. Comparison of the low power light field and dark field images (top left) revealed Ihh expression in the GI tract, the cartilage primordium of the cervical and tail vertebrae, and in the region of the developing urogenital sinus. Top right, a higher power view demonstrates localization in the luminal portion of the gut. Bottom left, closer inspection of the cartilaginous structures of the tail and cervical vertebrae (bottom right) suggests that Ihh is expressed in chondroblasts of these developing bones.
FIG. 6. Survey of Ihh mRNA expression in embryonic 16.5 dpc mouse tissues. 10 g/lane of total RNA isolated from 16.5 dpc mouse tissues was separated, transferred to a nylon membrane, and probed with a radiolabeled Ihh cDNA fragment. The blot was washed under stringent conditions and exposed to film at Ϫ80°C with intensifying screens overnight. Integrity of RNA and loading equivalency was assessed by ethidium bromide staining shown below. The position of 28 and 18 S ribosomal RNA is indicated on the right. undifferentiated cells remaining in the crypts (22). Abundant Ihh mRNA has also been detected in cartilage as early as 11.5 dpc. This expression is highest in chondrocytes in the growth zone regions of the cartilage with a lower level of expression in the hypertrophic zone (22). In developing chick bone, Ihh is produced by cells making a transition to hypertrophic or differentiated chondrocytes and appears to signal to neighboring perichondrial fibroblasts (24). Via a negative feedback loop that involves the induction of PTHrP in the perichondrium, Ihh secretion results in the inhibition of premature differentiation by chondroblasts in the bone growth plate (24,25).
In the kidney, Northern blotting confirmed Ihh transcript expression at 16.5 dpc and semiquantitative RT-PCR could detect Ihh as early as 14.5 dpc. Yet, Ihh transcript could not be localized in the developing kidney by in situ hybridization. RT-PCR allowed the localization of Ihh transcript to a specific terminally differentiated tubular epithelia in the adult kidney. Given that other hedgehog species function in early inductive events, it was surprising to find that the abundance of metanephric Ihh transcript increased with gestational age. This temporal pattern of Ihh expression suggested that Ihh is expressed in more differentiated metanephric cell phenotypes. This expression pattern is similar to Ihh expression in the duodenum, where Ihh transcript is found in more differentiated epithelium of the villi (22). That Ihh expression in differentiated tubular epithelia can signal adjacent, proliferating cells and maintain their undifferentiated state is a testable hypothesis that will require further investigation.