By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
Ror1 is one of the two vertebrate Ror-family members of orphan receptor-tyrosine kinases (Yoda et al.,2003; Green et al.,2008; Minami et al.,2009). Like its closely related family member Ror2, Ror1 contains an extracellular cysteine-rich domain (CRD) similar to the CRD domain of the Wnt-receptor frizzled (Saldanha et al.,1998). It has been demonstrated that Ror2 can bind to Wnt-ligands as well as to the BMP/TGFβ-family member, Gdf5 (Oishi et al.,2003; Sammar et al.,2004; Liu et al.,2008). Ror2 activity can mediate non-canonical Wnt-signaling through interaction with Wnt5a, and modulates the canonical Wnt-signaling pathway through differential interaction with the Frizzled receptor (Hikasa et al.,2002; Mikels and Nusse,2006; He et al.,2008; Li et al.,2008). Ror1 has been recently demonstrated to be capable of binding the Wnt5a ligand as well. However, whether this binding results in functional activity is currently not known (Fukuda et al.,2008). Both Ror genes are widely expressed during embryonic development, in limbs, perichondrium of the developing long bones, teeth, heart, lung, liver, gut, urogenital tract, and hippocampal neurons (Oishi et al.,1999; Al-Shawi et al.,2001; Matsuda et al.,2001; Paganoni and Ferreira,2003; Rodriguez-Niedenfuhr et al.,2004; Schwabe et al.,2004). Mutations in ROR2 in humans cause autosomal dominant brachydactyly type B (BDB) and are associated with an autosomal recessive form of Robinow syndrome (RS) (Afzal et al.,2000; Oldridge et al.,2000; Schwabe et al.,2000; van Bokhoven et al.,2000; Hamamy et al.,2006). Loss of Ror2 in mice results in phenotypic changes resembling RS (DeChiara et al.,2000; Takeuchi et al.,2000; Schwabe et al.,2004). The point mutation Ror2-W749X, linked to human BDB, behaves as a recessive mutation in mouse causing brachydactyly and models recessive RS (Raz et al.,2008). In addition, ROR family member have been implicated in tumor formation in humans. ROR1 has been found to be overexpressed in patients with acute and chronic lymphoblastic leukemia (Shabani et al.,2007,2008; Daneshmanesh et al.,2008) and was found to act as a survival kinase in HeLa cervical carcinoma cells (MacKeigan et al.,2005). ROR2 has been found to be overexpressed in squamous cell carcinoma, renal cell carcinoma, and metastatic melanoma, and to regulate osteosarcoma cell invasiveness (Enomoto et al.,2009; Kobayashi et al.,2009; Morioka et al.,2009; Wright et al.,2009; O'Connell et al.,2010). Previously it was reported that loss of Ror1 in mice results in perinatal lethality due to respiratory defects, but that these mice lack any abnormalities in skeletogenesis (Nomi et al.,2001). Here, we have re-examined the Ror1 mutants and found that the mice have subtle skeletal defects at birth; they showed fusions of the sternebrae, a cleft in the basisphenoid bone, and abnormal development of the cervical vertebral element C2. Homozygous mice survived in our facility and displayed abnormal synchondrosis in the cranial base, postnatal growth retardation, and age-related skeletal changes. Furthermore, we observed additional phenotypic defects in Ror1−/− mutants, such as female infertility probably due to an imperforated hymen, kidney defects, and occasionally enlarged seminal vesicles in Ror1−/− males.
Skeletal Defects in Ror1−/− Embryos
Ror1 mutant embryos displayed very distinct skeletal defects, which were apparent prior to birth. The sternum of Ror1−/− embryos was shortened and showed fusion of caudal sternebrae (S) involving to a variable degree S1–S4 (Fig. 1A and Table 1). In addition, the xyphoid process was abnormally bifurcated showing two ossification centers in all mutants examined at birth or early postnatal stages (Fig. 1A). Defects in sternal development were already detectable at E15.5 where the two sternal bands were still separated from each other in the caudal region of Ror1−/− embryos, while they were already almost completely fused along the entire length in the wild-type littermate controls (Fig. 1B, E). The sternal-costal junctions, demarcations between the end of the ribs and the sternal tissue, can be distinguished by their reduced reactivity with the proteoglycan staining reagent alcian blue and low levels of Collagen 2α1 (Col2α1) transcript, appeared right at the sternal bands in the wild-type (see arrowheads in Fig. 1C), while they were located more laterally in the Ror1 mutants (Fig. 1F). In situ hybridization for the prehypertrophic marker, Ihh, revealed that the onset of chondrocyte differentiation occurs normally in the Ror1 mutant sterni (Fig. 1D, G). However, in the caudal part the Ihh expression domains were not absolutely restricted to the intercostal segments (arrowheads in Fig. 1G). At E17.5, the abnormal shape and location of the sternal-costal junctions were still present in the mutant sterni. The sternal bands had either begun to fuse abnormally, visible by the presence of alcian blue–positive and Col2α1 (not shown) -expressing cells in the midline, or were still separated by a small band of none-chondrogenic cells (data not shown). However, as in the wild-type, chondrocytes within the sternal bands had maturated into hypertrophic chondrocytes that expressed Collagen 10α1 (Col10α1) and Osteopontin (Opn) and produced mineralized matrix in all mutants examined (Fig. 1I, J, M, N). In contrast to the controls where this process was restricted to the intercostals regions, hypertrophic cells differentiated also at the ends of ribs in the mutants leading to the formation of continuous hypertrophic domains in the region of sternebrae 3–4 (Fig. 1M, N). Concomitantly, Ihh expression was altered and expressed all around the hypertrophic regions (Fig. 1K, R). The abnormal expression of maturation markers, such as Ihh and Col10α1, suggests that the absence of functional Ror1 affects the normal organization of cells into a growth plate-like structure in the intercostal regions of the sternum, instead those are now located at the costal-sternal junctions. This defect was even more apparent at P0 where Col10α1 and Ihh were expressed almost all around the fused sternebrae elements, including the regions at the ends of the ribs (data not shown).
Table 1. Summary of the Observed Sternebrae Fusion Defects in Mutant and Control (Wt/Het) Animals
Occasionally a slight misalignment of sternum halves was observed (n=3; e.g. WT in Fig. 1C).
In addition to the defects in the sternum, we detected skeletal abnormalities in the cervical vertebral element C2 and the base of the skull in newborns (P0). The cervical element C2 appeared wider in the mutants (n=9) and was split in some mutants (n=4/9; Fig. 2A). In the skulls of mutant newborns, clefts were present in the posterior region of the basisphenoid bone and in the anterior borders of the exoccipitale bones in 100% of the mutants analyzed (n=9; Fig. 2B). No alterations were found in the suture regions in the dorsum of the skull (data not shown). At P5, the clefts had developed into holes, which were still visible at later stages and in adults (Fig. 2C, D, and data not shown). Furthermore, we noticed in the cranium of P5 and P19 heads a bony bridge within the spheno-occipitale synchondrosis connecting the basioccipitale with the basisphenoid bone (arrows in Fig. 2C, D). This resulted in premature synostosis of the spheno-occipitale synchondrosis in the cranial base and an overall shortening of the skull with a more dome-shaped appearance in X-rays at the adult stage (Fig. 2E). At P5, it was also apparent that the vomeral bones were hypoplastic (Fig. 2C). In the P17 and P19 mutant heads, we noticed a hole in the exoccipitale bones, which probably was the result of the cleft already visible in the newborn skulls (n=5; Fig. 2D). The fusion of the basisphenoid and basioccipitale bone was also clearly visible by microCTs of aged mutant specimens (Fig. 2F). None of these skeletal defects was visible in heterozygous Ror1 animals.
As previously described, Ror1 is expressed at E10.5 and E11.5 in the developing limbs, somites, and head (see Supp. Fig. S1A,B, which is available online). At E12.5, we noticed a prominent expression of Ror1 in the mesoderm of the body wall and at E13.5 in the sternal bars (see Supp. Fig. S1C–E). Ror1 appears also to be expressed in perichondrium of the skeletal elements in the developing limb (see Supp. Fig. S1F, G). Furthermore, we noticed Ror1 expression in the roof of the oral cavity and in the area of the developing pituitary gland (see Supp. Fig. S1H, I). Thus, Ror1 is expressed in or near the skeletal elements/structures affected by the loss of Ror1 activity.
Growth retardation of the Ror1 mutant pups became apparent at P2 (see Fig. 4), associated with morphological and molecular changes in the growth plates of long bones (Fig. 3B), while there were no changes detectable at E17.5 or in newborns (Fig. 3A and data not shown). At P2, the prehypertrophic, Ihh-expressing, and the hypertrophic, Col10α1-expressing zones of the mutant growth plates were shorter compared to wild-type or heterozygous littermate controls (Fig. 3B). In the wild-type growth plates at P2, only the last 3–4 rows of hypertrophic chondrocytes had mineralized matrix, while the hypertrophic chondrocytes towards the articular region were slightly smaller and had a non-mineralized matrix (Fig. 3B). Similarly, Opn was expressed in the last 3–4 rows of hypertrophic chondrocytes in the wild-type growth plate (Fig. 3B). In contrast, in the Ror1 mutant growth plate, only 3–4 rows of hypertrophic cells were present, which all expressed Opn and had a slightly hyper-mineralized matrix (Fig. 3B). Thus, loss of Ror1 seems to affect the differentiation of a distinct subpopulation of hypertrophic cells and prehypertrophic chondrocytes. Examination of P5 specimens revealed a similar phenotype in the growth plate, with a hyper-mineralized matrix and reduced zones of Ihh- and Col10α1-expressing chondrocytes in the mutant (Fig. 3C). In addition, skeletal preparations as well as sections through the long bones revealed that the formation of the secondary ossification center was delayed in Ror1 mutants (Fig. 3C and data not shown).
Ror1 Mutants Show Postnatal Growth Retardation and Have a Reduced Life Expectancy
At postnatal day 2, Ror1 mutant pups could easily be identified by their reduction in size and weight (Fig. 4A, B). Statistically significant reduction in the weight was apparent already at P1. However, the size reduction was less apparent at that age (Fig. 4A and data not shown). In contrast, no significant differences were detected between Ror1 heterozygous and wild-type animals. Within the first 6 postnatal days, mutant pups reached only approximately 50% of the weight of their heterozygous and wild-type littermates (Fig. 4A). The life expectancy of the mutants was reduced; only 60% of the mutants (32/53) survived until weaning (Fig. 4C). The reduced life expectancy might in part be due to the occasional kidney defects (see below) or physical weakness of the runted pups compared to their littermates. Ror1−/− pups, which survived until after weaning, stayed smaller and reached only 60–75% of the weight of their littermates (Fig. 4D, E). Accordingly, their long bones were ∼20% shorter compared to wild-type and heterozygous littermates (see Fig. 3D). X-rays of 3-month-old female mice suggested a decrease in bone density and in cortical thickness (Fig. 4D, see also inset). This finding prompted us to examine volumetric bone mineral density (BMD) and bone geometry by peripheral quantitative computed tomography. Femurs from Ror1 mutant male and female mice were characterized by bonedensitometry, which revealed reductions in the cross-sectional area, total BMD, and cortical thickness in the shaft and metaphyseal region (Table 2). Cortical bone osteopenia at the femoral shaft might be slightly more pronounced in females than in males (Table 2). Likewise, the reduction in weight and size was more prominent in females than in males at the age of 3 months (Fig. 4E and data not shown). Together, this suggests a possible involvement of sex hormones in the expressivity of the mutant phenotype. Mutant males aged for 13–18 months still showed severe weight reduction (32–60%). However, the difference in size was relatively small (5–10%) compared to their littermate controls (Table 3).
Table 2. Bone Densitometry Data of Femurs of 3–4-Month-Old Micea
Cross-sect area (mm)
Total BMD Trab (mg/cm3)
Cort. thickness (mm)
BMD, bone mineral density; Cort., cortical; wk, weeks; F, female; M, male; —, not applicable.
Table 3. Comparison of Weight and Size in 13–18-Month- (M) Old Ror1 Mutants and Littermates
Cystic right kidney, left seminal vesicle enlarged.
Abnormal Ossifications Develop in the Spine of Aged Ror1 Mutants
In alizarin red/alcian blue–stained skeletal preparations of aged mutant mice (15–18 months), we observed abnormal mineralization in the intervertebral discs of the lower thoracic region (n=2/2; T10–T13; arrowheads in Fig. 5A) and occasionally within un-fused sternal synchondroses (data not shown). Abnormal ossification within the axial skeleton could also be observed in microCTs of the lower thoracic spine of 15–18-month-old male mice (n=2/3; arrows in Fig. 5B). In addition, we noticed that there were abnormal ossifications between the spine and the proximal ends of the ribs (white arrowheads in Fig. 5C). The reduction in cortical bone thickness was also visible in the microCTs of the ribs and vertebral bodies (Fig. 5C and data not shown).
IGFI Serum Levels Are Reduced in Ror1 Mutants
Given that the hole in the basisphenoid bone of Ror1−/− embryos was located directly below the pituitary gland (see Fig. 6A), we suspected that the postnatal growth retardation might be due to altered hormone production by the pituitary gland. Thus, we examined the specification of different hormone-producing cells using immunohisto-chemical staining for growth hormone (GH), thyroid-stimulating hormone beta (TSHβ), luteinizing hormone alpha and beta (LHα, β), and adrenocorticotropin (ACTH) on E17.5 and P2 sagittal sections. However, no obvious differences in hormone staining were detected between wild-type and Ror1 mutant pituitaries at both stages of development (Fig. 6A and data not shown). Older pituitary glands of Ror1-deficient animals were slightly hypoplastic and had a different shape compared to wild-type and heterozygous littermate controls (Fig. 6B and data not shown). Western blots on pituitary gland extracts of P17 and P19 old littermates revealed, however, no difference in GH levels (Fig. 6C). Although GH production seemed not to be affected, its possible that the levels of secreted GH might be altered in Ror1−/− animals. Since GH synthesis and release from pituitary somatotrophs is pulsatile, the serum GH levels fluctuate accordingly. Thus, instead of measuring the serum GH levels we measured the serum levels of IGF-I, since they directly depend on GH levels. The IGF-I levels were indeed reduced in 3-month-old male and female Ror1 mutant animals (Fig. 6D).
Urogenital Defects in Ror1 Mutant Animals
Approximately 90% of the Ror1 mutant females had an abnormal vagina with an imperforated hymen and could, therefore, not reproduce (Fig. 7A). With age, these Ror1−/− females accumulated liquid and cellular material in their uteri resulting in a swelling of the uteri and abdomen (Fig. 7B, C). Almost all Ror1 mutant males tested could reproduce normally, but 3/7 dissected Ror1−/− males showed abnormal seminal vesicles that were either cystic or solid and unilaterally enlarged (Fig. 7D, D′). Approximately 30% (5/17) of the Ror1−/− mice had abnormal kidneys, where double kidneys and double ureters were present unilaterally (Fig. 7E). All of the mutants with kidney defects had been found dead within the first week, with the exception of one male that was sacrificed at the age of 13 months and had a cystic right kidney (see Table 1). In addition to the urogenital defects, we noticed that the mutant animals had almost no subcutaneous fat and that the visceral fat pads associated with the urogenital system were also reduced in size (Fig. 7D, F).
Ror1 mutants have previously been reported to die shortly after birth due to defects in lung maturation (Nomi et al.,2001). Currently, we have no explanation for the increased postnatal survival of Ror1 mutants that were bred in Vienna, particularly since Ror1 mice were maintained in a C57Bl/6 background. Nevertheless, we cannot rule out subtle differences between the C57Bl/6 strains in the two geographically distinct locations. The fact that we still observed a high lethality rate for the Ror1 mutants within the first few days suggests a fitness weakness of Ror1 mutants. In contrast to Ror2, loss-of Ror1 activity resulted only in minor skeletal defects, affecting primarily the sternum, the basisphenoid bone, and the axial element C2 during embryonic development. Ror1 and Ror2 have been implicated as co-receptors in Wnt-signaling and for Ror2 it was shown that it can modulate Gdf5/BMP receptor signaling (Sammar et al.,2004; Green et al.,2008). As such, it is interesting to note that a similar phenotype in the basisphenoid bone has been observed in embryos double mutant for Tcf4/Lef1 during development at E15.5 and in mutants for Bmp7 or conditional mutants of the Tgf-beta receptor 2, Tgfbr2ΔCol2 (Jena et al.,1997; Baffi et al.,2004; Brugmann et al.,2007). Furthermore, it is interesting to note that double mutants for the heparan sulfate 6-O-endosulfatase genes Sulf1 and Sulf2, which alter sulfate modification of heparan molecules in the extracellular matrix, showed a similar phenotype in the basisphenoid bone (Holst et al.,2007; Ratzka et al.,2008). In these mutants, signaling of several growth factors (including Wnts, hedgehogs, and fibroblast growth factors) is probably affected. Mutants for Gli3Δ699/Δ699, which encodes only a N-terminal truncated version of Gli3 solely acting as a repressor, also have a hole in the basisphenoid bone (Bose et al.,2002). Fusions of sternebrae or ectopic mineralization within the fibrocartilage separating the sternebrae have been reported also in Sulf1/2 double mutants and in mutants for the TGFβ/BMP family members, Bmp5 and Gdf5 (Storm and Kingsley,1996; Holst et al.,2007; Ratzka et al.,2008). Thus, given the phenotypic similarities, it might be possible that in analogy to Ror2, Ror1 might not only interact with Wnt-ligands altering Wnt-signaling, but might also modulate TGFβ/BMP signaling.
Besides the patterning defects in certain skeletal elements, we noticed postnatal alterations in the growth plates of the long bones of Ror1 mutants. Here the zones of prehypertrophic and hypertrophic chondrocytes were reduced from postnatal day 2 onwards. Based on our analyses, it seemed that primarily the zone of immature hypertrophic chondrocytes, in which the cells do not have a mineralized matrix, nor do they express Opn, is severely reduced or maybe even lost. This specific defect in chondrocyte maturation might contribute to the fact that the appearance of the secondary ossification centers was delayed as well. Currently, we have no explanation for the molecular nature of the specific defect. The previous study by Nomi and colleagues suggests that Ror1 has redundant functions with Ror2 during embryonic skeletal development (Nomi et al.,2001). However, the study gave no insights into the underlying mechanism explaining the further shortening of the skeletal elements particularly in the stylopode region. Ror2 seems to be regulating chondrocyte maturation primarily at the level of prehypertrophic chondrocytes (Schwabe et al.,2004). Our analysis suggests that in contrast to Ror2, Ror1 might be required at a slightly different step of chondrocyte maturation acting primarily in immature hypertrophic chondrocytes. This could explain the additional decrease in overall length and of the mineralized zone of the stylopode elements, humerus and femur, upon loss of Ror1 in a Ror2 mutant background. This would also explain why there was no significant effect on zeugopode elements such as the ulna and radius upon additional removal of Ror1 (Nomi et al.,2001). Here the differentiation of prehypertrophic and hypertrophic chondrocytes is already severely delayed in Ror2 single mutants, while it still occurs in the stylopode at E15.5 (Schwabe et al.,2004).
In addition to the specific reduction in immature hypertrophic chondrocytes, the postnatal growth deficiency might in part be influenced by the reduction in IGF-I serum levels but it is probably also influenced by sex hormones given that the differences in size are much more pronounced in females than in males. Whether the reduced IGF-I serum levels are related to altered growth hormone production due to the morphological alteration in the basisphenoid bone and the associated pituitary gland deformation or if they might be due to local Ror1 activity in IGF-I-producing organs is currently unclear. While Ror1 is expressed fairly broadly in the organism, it is not expressed in the liver, which is the major site of IGF-I production (Al-Shawi et al.,2001; Baskar et al.,2008). Concomitantly, liver-specific deletion of Igf-1 does not affect growth of the appendicular skeleton (Yakar et al.,1999; Sjogren et al.,2002). Thus, local IGF-I production in the growth plate of the long bones is probably affected in Ror1 mutants, given the reduction in the hypertrophic zone that is the predominant region in which Igf-1 is transcribed postnatally (Reinecke et al.,2000; Smink et al.,2002). IGF-I has been shown to stimulate hypertrophic chondrocyte differentiation (Mushtaq et al.,2004). Thus, a reduction in IGF-I-producing hypertrophic chondrocytes might exaggerate the phenotype. Our phenotypic analysis suggests that Ror1 activity is required for postnatal growth and hypertrophic chondrocyte differentiation and that the phenotype is possibly mediated in part via the GH/IGF-I system. Since Ror1 is fairly broadly expressed and not highly expressed in the chondrocytes, it is currently unclear to what extent the growth plate phenotype is due to a Ror1 activity requirement particularly in chondrocytes. In addition, it is well known that cortical bone mass and size is severely reduced in IGF-I-ablated mice (Moerth et al.,2007). Therefore, it remains to be determined whether the cortical bone osteopenia observed in Ror1 mutants is caused by lack of Ror1 and concomitantly reduced Wnt signaling in periosteal osteoblasts, or by an indirect effect through the GH/IGF-1 axis.
Furthermore, our phenotypic analysis revealed a requirement for Ror1 in the urogenital system. The relatively low penetrance of the kidney defects makes it difficult to study the underlying mechanisms. In addition to Ror1, Ror2 is also expressed in the urogenital tract (Al-Shawi et al.,2001). Interestingly, we noticed in Ror2 mutants a very similar kidney phenotype with double ureters, unilateral or bilateral double kidneys, cystic kidneys, or even sporadic agenesis of one kidney in 30–40% of Ror2 mutants (C. Hartmann, unpublished observation). Thus, it is possible that Ror1 and Ror2 act redundantly in kidney development, although no defect in kidney development has been reported in the previous double mutant analysis (Nomi et al.,2001). Ror1 and Ror2 could possibly modulate Wnt or TGF-β/BMP-signaling during kidney development, as Wnt-signaling as well as TGF-β/BMP-signaling have been functionally implicated in different aspects of kidney development (Martinez and Bertram,2003; Pulkkinen et al.,2008).
The generation of the Ror1 mutant allele has been previously described (Nomi et al.,2001). The Ror1 strain was maintained in a C57Bl/6 background. Mutants were generated by intercrosses of heterozygous mice, or through breeding of homozygous males with heterozygous females to increase the number of mutant offspring. Genotyping of mice was performed by PCR using the following primer pairs: for detection of the 1.1-kb wild-type allele the WT_for-primer 5′ GGCAACAAATGGCAAGAAAGTG GTGTC and WT_rev-primer 5′ GAAA TGGAATCCTTAGACTCCGTTATC were used. For detection of the 1.1-kb mutant allele, the WT_for-primer was used in combination with the neo-primer 5′ ATCGCCTTCTATCGCCTT CTTGACGAG.
Skeletal Preparations, MicroCT, and Bone Densitometry
For the alcian blue/alizarin red staining of newborns, postnatal, and adult mouse skeletons, the mice were sacrificed, skinned, eviscerated, fixed in 95% ethanol, and stained according to McLeod (1980). Harvested organs were visually inspected for abnormalities. Three-dimensional medium-resolution images were obtained from the skulls, spine, ribs, and digits of Ror1−/− and control mice (Ror1+/−; Ror1+/+) using microcomputed tomography (eXplore locus SP, GE Healthcare, London, Ontario, Canada). Scans were taken at 28-μm isotropic resolution and 720 projections were acquired over an angular range of 360°. Images were reconstructed and thresholded to distinguish bone voxels with MicroView software version 5.2.2 (GE Healthcare, Buckinghamshire, UK). One threshold was chosen for all specimens or as previously described (Amarilio et al.,2007). Volumetric bone mineral density (BMD) of the femurs was measured by peripheral quantitative computed tomography (pQCT) using a XCT Research M+ pQCT machine (Stratec Medizintechnik, Pforzheim, Germany) as described (Schneider et al.,2009). One slice (0.2 mm thick) in the mid-diaphysis of the femur and 3 slices in the distal femoral metaphysis located 1.5, 2, and 2.5 mm proximal to the articular surface of the knee joint were measured. BMD values of the distal femoral metaphysis were calculated as the mean over 3 slices. A voxel size of 0.070 mm and a threshold of 600 mg/cm3 were used for calculation of BMD.
Histology and In Situ Hybridizations
For histology and section in situ hybridizations, material from embryos and pups was dissected, washed in PBS, and fixed overnight in 4% PFA/PBS, dehydrated to 70% ethanol, and processed using a standard program of the Tissue-Tek VIP5 Vacuum Infiltration Processor (Sakura, Torrance, CA). Processed tissue was embedded in paraffin and sectioned at 5 μm. Hematoxylin/eosin staining was performed using standard protocols. Alcian blue/van Kossa staining on sections was performed as follows: tissue was rehydrated, washed twice in deionized water, exposed for 60 min in 2% silver-nitrate solution to a 60-W lamp, washed three times in deionized water, incubated for 2 min in 1% acetic acid, stained with alcian blue solution (pH 2.2) for 15 min, washed in 1% acetic acid, dehydrated into 75% ethanol, counterstained for 30 sec with eosin, destained with 100% ethanol and xylene, and mounted using DPX (Fluka, St. Louis, MO). Whole mount and section in situ hybridizations were done as previously described (Murtaugh et al.,1999; Riddle et al.,1993).
Immunohistochemistry, Western Blot, and Elisa for IGF1 Serum Levels
Immunohistochemistry for the following hormone precursors, GH (1:10,000), TSHβ (1:2,000), LHα (1:2,000), LHβ (1:2,000), and ACTH (1:3,000) was performed on paraffin sections of the pituitaries from E17 embryos and P2 mice. Antibodies were obtained from the National Hormone and Pituitary Program and immunohistochemistry was performed using the Ventana. Briefly, sections were dewaxed and rehydrated; antigen-retrieval was done using citrate buffer pH 6.0. Western blots to determine GH protein levels were performed on 50-μg protein extracts from dissected pituitaries from P17 & P19 mice (α-GH at a dilution of 1:2,000). IGFI serum levels were determined in duplicates from the plasma of Ror1−/− and control mice at 2 and 3 months of age using an IGF-I Elisa assay following the manufacturer's instructions (Quantikine Kit, R&D Systems, Minneapolis, MN).
We thank Christiane Schüler, Martin Glösmann, and Claudia Bergow for help with the pQCT analysis, and Vukoslav Kommenovic for technical assistance. The IMP is supported by Boehringer Ingelheim. N.L. was supported by the Austrian Science Found (FWF grant P19281-B16).