Studies of Otic Capsule Morphology and Gene Expression in the

Original Paper Audiology Neurotology

Audiol Neurotol 2007;12:334–343 DOI: 10.1159/000104789

Received: May 3, 2006 Accepted after revision: March 28, 2007 Published online: June 26, 2007

Studies of Otic Capsule Morphology and Gene Expression in the Mov13 Mouse – An Animal Model of Type I Osteogenesis Imperfecta Konstantina M. Stankovic a, b Arthur G. Kristiansen a Argyro Bizaki a, b Matthew Lister a, b Joe C. Adams a, b Michael J. McKenna a, b a

Department of Otolaryngology, Massachusetts Eye and Ear Infirmary, and b Department of Otology and Laryngology, Harvard Medical School, Boston, Mass., USA

Abstract Type I osteogenesis imperfecta (OI) is a disorder of skeletal bones characterized by bone fragility and blue sclera, which can result from mutations in genes encoding for type I collagen – the COL1A1 and COL1A2 genes. Fifty percent of patients with type I OI develop hearing loss and associated histopathological changes in the otic capsule that are indistinguishable from otosclerosis, a major cause of acquired hearing loss. In an attempt to elucidate molecular and cellular mechanisms of hearing loss in type I OI, we have studied the Mov13 mouse, which has served as an animal model of type I OI by virtue of exhibiting variable transcriptional block of the COL1A1 gene. We studied the morphometry of the Mov13 otic capsule and compared expression levels of 60 genes in the otic capsule with those in the tibia and parietal bone of the Mov13 and wild-type mice. The degree of transcriptional block of the COL1A1 gene and its downstream effects differed significantly between the bones examined. We found that expression levels of bone morphogenetic protein 3 and nuclear factor -B1 best distinguished Mov13 otic capsule from wild-type otic capsule, and that osteoprotegerin, caspase recruitment domain containing protein 1,

© 2007 S. Karger AG, Basel Fax +41 61 306 12 34 E-Mail [email protected] www.karger.com

Accessible online at: www.karger.com/aud

and partitioning defective protein 3 best distinguished Mov13 otic capsule from Mov13 tibia and parietal bone. Although the Mov13 mouse did not demonstrate evidence of active abnormal otic capsule remodeling as seen in type I OI and otosclerosis, studying gene expression in the Mov13 mouse has provided evidence that osteocytes of the otic capsule differ from osteocytes in other bones. Copyright © 2007 S. Karger AG, Basel

Introduction

Osteogenesis imperfecta (OI) is a dominantly inherited generalized connective tissue disorder known to result from mutations in type I collagen genes, COL1A1 and COL1A2, resulting in bone fragility with a propensity for fractures from minimal trauma. There are 6 clinical types of OI, with type I OI being the most common and mildest form. Approximately 50% of patients with type I OI develop hearing loss that may be clinically and histopathologically indistinguishable from otosclerosis [Nager, 1988], a major cause of acquired hearing loss in Caucasians. At a histologic level, both type I OI and otosclerosis are characterized by abnormal bone remodeling affecting the otic capsule [Pedersen et al., 1985; McKenna et al., 1998], which under normal conditions remodels miniMichael J. McKenna, MD Massachusetts Eye and Ear Infirmary 243 Charles Street Boston, MA 02114-3096 (USA) Tel. +1 617 573 3651, Fax +1 617 573 3939, E-Mail [email protected]

Downloaded by: Massachusetts Eye & Ear 65.112.5.1 - 6/4/2015 1:39:17 AM

Key Words Mov13 Otic capsule Type I osteogenesis imperfecta COL1A1 Osteoprotegerin

Gene Expression in the Mov13 Otic Capsule

some form of otosclerosis because this mouse was previously recognized to have progressive and profound hearing loss [Bonadio et al., 1990] of the mixed conductive and sensorineural type [Altschuler et al., 1991]. In addition, preliminary histological analysis of the Mov13 otic capsule suggested the presence of new bone formation [Altschuler et al., 1991]. We studied the detailed morphometry of the Mov13 otic capsule, the ratio of COL1A1 to COL1A2 mRNA and expression of 58 other genes involved in bone metabolism. The gene expression profile of the otic capsule was compared with that of parietal bone and tibia to determine whether fundamental differences exist between osteocytes of the otic capsule and other bones. Such differences could provide an insight into why the otic capsule normally does not undergo substantial remodeling.

Methods Histology and Morphometry All animal experiments in this study were approved by the Animal Care and Use Committee of the Massachusetts Eye and Ear Infirmary. Histological and morphometric analysis was performed on 6 ears from 3 heterozygous, 17-week old Mov13 mice (of C57BL/6 background, strain name C57BL/6-Col1a1Mov13/J) and 9 ears from five 17-week old C57BL/6J control mice. Mov13 homozygous mice were not studied because they are embryonic lethal [Schnieke et al., 1983]. Lymphoma has been described in adult Mov13 mice [Bonadio et al., 1990], and we found lymphoma involving the temporal bone in 2 adult mice. However, the 3 Mov13 mice reported in this study had no histological evidence of lymphoma involving the temporal bone. All mice were obtained from Jackson Laboratories, Maine. The animals were anesthetized by intraperitoneal injection of urethane (2 g/kg), then fixed by intracardiac perfusion with 10% formalin. Skulls containing temporal bones were dissected, postfixed for 6 h at 4 ° C, decalcified in 120 mmol/l ethylenediaminetetraacetic acid for 1 week, embedded in paraffin and serially sectioned in an axial plane at a thickness of 10 m. Select sections were stained with 0.2% azure B, pH 3, to identify cartilage [Lillie, 1965] because cartilaginous rests are unique histological features of the normal otic endochondral bone, and are thought to be developmental remnants of fetal bone [Bast and Anson, 1949]. Changes in cartilaginous rests have been implicated in the development of type I OI and otosclerosis. The area of the otic capsule and cartilaginous rests within it were measured using Olympus MicroSuite drawing software. All measurements were performed in the second turn of the cochlea because the size of this turn was most uniform among animals. All morphometric measurements were done in adult mice to avoid biasing of results due to developmental age because Mov13 mice are known to have delayed development and reach maturity at 13 weeks. A rectangle was drawn so that its medial long side connected the inferior medial and the superior medial edge of the spiral ligament (fig. 1A). The lateral side of the rectangle was a

Audiol Neurotol 2007;12:334–343

335


mally, if at all [Frisch et al., 2000]. Predominant cells within the normal otic capsule are osteocytes [Bast and Anson, 1949], which maintain bone. Osteoblasts, which build new bone, and osteoclasts, which degrade bone, are rarely found. The initiation of otic capsule remodeling in type I OI and otosclerosis is seen histologically as a localized change in the staining pattern of the otic capsule within the sites of predilection (blue mantles) [Schuknecht, 1993], which reflects an alteration in the composition of the extracellular matrix that is produced by osteoblasts and osteocytes. In addition to histopathologic and clinical similarities between type I OI and otosclerosis, there are some differences between the 2 diseases. Type I OI affects the entire skeleton, whereas otosclerosis occurs only within the temporal bones of humans [Wang et al., 1999]. Compared with otosclerosis, hearing loss in type I OI tends to have an earlier onset, more diffuse otic capsule involvement and a higher incidence of sensorineural hearing loss [Holdsworth et al., 1973; Shapiro et al., 1982; Kuurila et al., 2002]. The majority of cases of type I OI are due to mutations in the COL1A1 gene that result in null expression of the mutant allele, whereas we have shown that only a fraction of patients with otosclerosis have a similar single allele reduction in COL1A1 expression [McKenna et al., 1998; McKenna et al., 2002]. Six other unrelated loci have been implicated in otosclerosis [Stankovic and McKenna, 2006]. In addition to COL1A1 null expression, other studies have implicated functional polymorphisms within the COL1A1 gene that promote overexpression of COL1A1 and the formation of unstable COL1A1 homotrimers in the extracellular matrix [Chen et al., 2007]. The Mov13 transgenic mouse has served as an animal model for the study of COL1A1 null expression and as a model of mild OI [Bonadio et al., 1990]. The Mov13 mouse was created by insertion of a Maloney murine leukemia provirus into the first intron of the COL1A1 gene [Schnieke et al., 1983], resulting in a variable transcriptional block of the COL1A1 allele, ranging from complete block in fibroblasts [Breindl et al., 1984] to partial block in osteoblasts (5% of osteoblasts from long bones had no transcriptional block) [Kratochwil et al., 1993] and no block in odontoblasts [Kratochwil et al., 1989]. Because Mov13 homozygous mice are embryonic lethal, complete transcriptional block in Mov13 heterozygous tissue implies a 50% reduction in the ratio of the COL1A1 to COL1A2 mRNA. We undertook the current study to determine if the Mov13 mouse might constitute an animal model for otic capsule remodeling seen in type I OI and

Fig. 1. Cross-sectional histology of the wild-type (A , B) and Mov13 (C , D) otic capsule stained with azure. Measurements were done in the second turn of the cochlea (A , C) and in the apical region of greatest marrow space (B, D) within the areas outlined by straight lines. The region of greatest marrow space was thicker and contained less bone marrow in Mov13 than wild-type mice. Asterisk = Spiral ligament; black arrow in A = periosteal edge of the otic capsule facing the middle ear; black arrows in B define the limits of the measured area. Scale bar = 100 m (A) and 200 m (B). Scale bars in A and B apply to C and D, respectively.

Collection of Tissue, Extraction of Total RNA and cDNA Synthesis Samples of bone were collected from a total of 5 Mov13 heterozygous mice and 7 adult C57BL/6J mice. The animals, ranging in age from 8 to 18 weeks, were sacrificed by intraperitoneal injection of urethane (2 g/kg). Otic capsule bone surrounding the cochlea, the parietal bone and tibia were microdissected and placed in RNAlater reagent (Ambion). For each mouse, both otic capsules were removed and processed together as 1 specimen. The thickest part of the petrous bone containing the marrow space could not be easily harvested and thus was excluded from all specimens. Care was taken to scrape away mucosa of the middle ear and parts of the membranous labyrinth adherent to the otic capsule. However, some contamination with bone marrow and soft

336


tissues adherent to the otic capsule cannot be definitively excluded. Parietal bone and tibia were chosen because they exemplify developmentally distinct mechanisms of ossification: intramembranous ossification, which is characterized by direct conversion of precursor tissue into bone, and endochondral ossification, which is characterized by indirect conversions of precursor tissue into bone via a cartilage intermediate [Gilbert, 2000]. Intramembranous ossification gives rise to flat bones of the skull, including the parietal bone, whereas endochondral ossification gives rise to the otic capsule, and most of the axial and appendicular skeleton, including the tibia. Total RNA was purified using RNeasy spin columns (Qiagen) according to the manufacturers’ protocol and a modification for hypocellular, dense connective tissues [Reno et al., 1997]. Specifically, Trizol was used first, followed by chloroform, collection of the aqueous phase, addition of an equal volume of isopropanol and application of this mixture to RNeasy spin columns. The RNA was initially quantified on a Genesys 5 Spectrophotometer, followed by quantification and quality assessment using Agilent 2100 Bioanalyzer (Agilent Technologies) and RNA Pico kit. Total RNA that appeared clean and undegraded, based on the Bioanalyzer’s electropherograms, was reverse transcribed with Taqman Reverse Transcription Reagents kit (Applied Biosystems); degraded or contaminated RNA samples were discarded. Overall, RNA

Stankovic /Kristiansen /Bizaki /Lister / Adams /McKenna


tangent to the periosteal edge of the otic capsule. In a cross-section of the second turn of the otic capsule, the total area of the otic capsule, the area of cartilaginous rests and the total number of cells were determined. We also measured the widest cross-sectional area of the thickest part of the petrous bone, which contains bone marrow and is located anterior to the cochlea (fig. 1B), and the area occupied by bone marrow in that region. Statistical significance of the measurements was determined using the bootstrap resampling method [Efron and Tibshirani, 1993].

Table 1. Genes whose expression levels were studied in the otic capsule, parietal bone and tibia Gene name (accession No.)

Gene symbol

Gene name (accession No.)

Gene symbol

B cell leukemia lymphoma 2 (NM_009741.2) B cell leukemia lymphoma 10 (NM_009740.1) Bone morphogenetic protein receptor type IA (NM_009758.3) Bone morphogenetic protein receptor type 1B (NM_007560.3) Bone morphogenetic protein receptor type II (NM_007561.1) Bone morphogenetic protein 3 (NM_173404.1) Bone morphogenetic protein 4 (NM_007554) Bone morphogenetic protein 7 (NM_007557.2) Caspase recruitment domain family, member 11 (NM_175362.1) Collagen, type I, 1 (NM_007742.2) Collagen, type I, 2 (NM_007743.1) Collagen, type II, 1 (NM_031163.2) Rel (NM_009044.1) Colony-stimulating factor 1 (NM_007778.2) Connexin 26, gap junction protein 2, 26 kDa (NM_008125.2) Connexin 30, gap junction protein 6, 30 kDa (NM_008128.3) E-cadherin (epithelial), cadherin 1, type 1 (NM_009864.1) Epidermal growth factor receptor (NM_207655.1) Glial-cell-derived neurotrophic factor (NM_010275.1) GDNF family receptor 1 (NM_010279.2) GDNF family receptor 2 (NM_008115.1) Glucocorticoid receptor (NM_008173.1) Heat shock 70-kDa protein 1B (NM_008301.3) Heat shock 70-kDA protein 14 (NM_015765.1) Inhibitor of apoptosis protein 1 (NM_007465.1) Interleukin 1 (NM_010554.3) Interleukin 1 (NM_008361.2) Interleukin 6 (NM_031168.1) Interleukin 11 (NM_008350.1) Leptin (NM_008493.3) Mothers against decapentaplegic, Drosophila, homolog of 4 (NM_008540.2)

BCL2 BCL10 BMPR1A

Nuclear factor of -B, subunit 1 (NM_008689.1) Osteoprotegerin (NM_008764.2) Partitioning defective 3 homolog, protease-activated receptor 3 (NM_033620.1) Parathyroid hormone receptor (NM_011199.1) TNF receptor superfamily, NFkB activator (NM_009399.2) TNF ligand superfamily, member 11 (NM_011613.2) V-rel reticuloendotheliosis viral oncogene homolog A (NM_009045.3) V-rel reticuloendotheliosis viral oncogene homolog B (NM_009046.2) Runt-related transcription factor 2 (NM_009820.2) Suppressor of cytokine signaling 1 (NM_009896.1) Suppressor of cytokine signaling 2 (NM_007706.2) Suppressor of cytokine signaling 3 (NM_007707.2) Sclerosteosis (NM_0024449.3) Secreted protein, acidic, cysteine-rich (osteonectin) (NM_009242.1) TGF-induced apoptosis protein 2 (NM_153409.2) TGF-induced apoptosis protein 12 (NM_153407.1) TRAF-family-member-associated NFkB activator (NM_011529.1) TGF 1 (NM_011577.1) TGF receptor 1 (NM_009370.2) TGF 2 (NM_009367.1) TGF receptor 2 (NM_083851.2) Tolloid-like 1 (NM_009390.1) Toll-like receptor 4 (NM_021297.1) Toll-like receptor 5 (AF186107.1) TNF (NM_013693.1) TNF 1 receptor (NM_011609.2) TNF 1 receptor (NM_011610.3) Thyroid-stimulating hormone receptor (NM_011648.3) Vitamin D receptor (NM_009504.2)

NFkB1 OPG PAR3

BMPR2 BMP3 BMP4 BMP7 CARMA1 COL1a1 COL1a2 COL2a1 C-REL CSF1 CX26 CX30 E-CAD1 EGFR GDNF GFRA1 GFRA2 GR HSP70-2 HSP70-4 IAP2 IL1a IL1b IL6 IL 11 LEP SMAD4

from 5 otic capsules, 4 parietal bones and 6 tibias of C57BL/6J mice, and RNA from 5 otic capsules, 5 parietal bones and 3 tibias of Mov13 mice met the quality criteria. Two of the wild-type tibia samples were pooled together to increase the RNA yield; the results from the pooled and nonpooled sample were in agreement. Relative Quantitation of mRNA Using Real-Time PCR Real-time quantitative RT-PCR was used to analyze gene expression levels in a sensitive and high throughput fashion. We selected 60 genes that are known to play roles in bone remodeling, including inflammatory cytokines, and that encode structural proteins (table 1). For all genes 6-FAM linked fluorescent probes and primers were designed and optimized by Applied Biosystems. The measurements were carried out on an Applied Biosystems 7700 Sequence Detector using 96-well plates. For each well, the 25-l reaction contained: 1.25 l of the 20X probe/primer mix, 1 l of cDNA template (approximately 15 ng cDNA/l), 12.5 l Universal Master Mix (Applied Biosystems) and 10.25 l distilled


PTHR RANK RANKL RELA RELB RUNX2 SOCS1 SOCS2 SOCS3 SOST SPARC TAIP-2 TAIP-12 TANK TGFB1 TGFB1r TGFB2 TGFB2R TLL TLR4 TLR5 TNFa TNFa1R TNFb1R TSHR VDR

water. All cDNA samples were run in triplicate, and each plate was used to test expression of up to 32 genes from a single sample. Fluorescence data were collected over 45 cycles of PCR that consisted of an initial denaturation step at 95 ° C for 10 min, followed by 45 cycles of 95 ° C for 15 s and 60 ° C for 1 min. Gene expression levels were quantified relative to the 18S rRNA gene and compared between bone types using the comparative threshold cycle (CT) method, i.e. the C T method [Livak and Schmittgen, 2001]. Bioinformatics Analysis Tools of bioinformatics were applied to simultaneously analyze expression patterns of a relatively large number of genes studied in a relatively small number of samples. The bootstrap resampling method was used to determine whether variations in gene expression levels among different bone types were statistically significant at p ! 0.05 [Stankovic and Corfas, 2003]. Genes were ranked according to signal-to-noise ratio defined as the differ-


337


BMPR1B

Table 2. COL1A1 to COL1A2 mRNA ratio in Mov13 and control mice

Mouse

Otic capsule

Tibia

Parietal bone

Mov13 heterozygote C57BL/6J wild type

2.680.7 (p = 0.02, n = 5) 4.581.1 (p = 0.01, n = 5)

2.681.0 (p = 0.20, n = 3) 4.582.5 (p = 0.05, n = 5)

1.780.5 (p = 0.14, n = 4) 3.181.0 (p = 0.05, n = 5)

The values are expressed as a mean 8 standard deviation of the mean based on measurements from n animals. p = Statistical significance of the difference between COL1A1 and COL1A2 mRNA levels.

Results

Histology and Morphometry of the Otic Capsule in Mov13 Mice The otic capsule of adult Mov13 mice had no gross abnormalities and demonstrated no pathologic osteoclastic bone resorption (fig. 1C compared with fig. 1A). Detailed morphometric analysis revealed no statistically significant difference (p = 0.08) in the cross-sectional area of the otic capsule of Mov13 mice (30865.0 8 4691.3 m2) compared with C57BL/6J wild-type controls (26044.1 8 5786.2 m2), no difference (p = 0.9) in the cross-sectional area of cartilaginous rests (3245.9 8 1172.0 m2 in Mov13 mice and 3158.0 8 960.8 m2 in controls) and no difference (p = 0.4) in the total number of cells within the otic capsule (74 8 18 in Mov13 mice and 67 8 17 in controls). The cross-sectional area of the widest part of the apical region that has the greatest marrow space was 64% increased in Mov13 mice (p = 0.0006), and the area of that bone occupied by bone marrow was 32% decreased in Mov13 mice (p = 0.03) compared with wild types. Consequently, the cross-sectional ratio of bone marrow to solid bone in the apical region of greatest marrow space decreased 58% (p ! 0.00000001) in Mov13 mice. The region of greatest marrow space was the thickest part of the petrous bone and thus could not be easily harvested during dissections of the otic capsule. Therefore, the region of greatest marrow space was excluded from samples of the 338


otic capsule used for analysis of gene expression described below. Except for missing hair cells in the lower basal turn, the histology of the Mov13 cochlea appeared normal. However, the histology of the middle ear revealed effusions in 4 out of 6 ears, accompanied by mucosal edema and infiltrates of inflammatory cells. There was cholesterol granuloma in 1 middle ear. The Ratio of COL1A1 to COL1A2 mRNA in Mov13 Mice The COL1A1 to COL1A2 mRNA ratio expressed as a mean 8 standard error of the mean was 4.5 8 1.1 in the wild-type otic capsule, 4.5 8 2.2 in the tibia and 3.1 8 1.0 in parietal bone (table 2). This ratio was reduced about 50% in the otic capsule of Mov13 mice (table 2). The same trend was present in other bones but it did reach statistical significance. Nonetheless, the trend suggested that expression levels of COL1A1 and COL1A2 genes were coordinately affected in Mov13 mice because expression of the COL1A1 gene was affected to different degrees in different bones: 2.2 8 0.6 fold reduction in the otic capsule (p = 0.03), 9.0 8 4.7 fold reduction in tibia (p = 0.04) and 1.89 8 0.5 fold increase in parietal bone (p = 0.08). Expression of Multiple Genes in Mov13 Mice versus Controls The set of genes that we studied was sufficient to allow for robust distinction between Mov13 and wild-type mice in a bone-type-specific manner. The results are shown in a form of a heat map (fig. 2) where columns represent different samples, rows represent expression levels of different genes determined from real-time quantitative RTPCR, and color coding reflects row-normalized levels of gene expression with red indicating high and blue indicating low levels of expression. The wild-type and Mov13 mice are compared by separately focusing on the otic capsule (fig. 2A), tibia (fig. 2B) and parietal bone (fig. 2C). Twenty-eight genes were differentially expressed in the Stankovic /Kristiansen /Bizaki /Lister / Adams /McKenna


ence of the mean expression levels of each bone type scaled by the sum of standard deviations. GenePattern 2.0 platform [GenePattern, 2005] was used for all bioinformatics analyses. Class neighbors analysis [Golub et al., 1999] was used to determine which genes best characterize a bone type (class) or a mouse type (C57BL/6J or MOV13), and there was a priori knowledge of which sample came from what bone type or what mouse type. Genes were ranked using the signal-to-noise ratio statistic. The robustness of the genes that best characterize a bone type was tested using K-nearest neighbors cross-validation [Golub et al., 1999].

C57BL6J_1 C57BL6J_2 C57BL6J_3 C57BL6J_4 MOV13_1 MOV13_2 MOV13_3 MOV13_4 MOV13_5

C57BL6J_2 C57BL6J_3 C57BL6J_4 C57BL6J_5 C57BL6J_2 mice MOV13_1 MOV13_2 MOV13_3 C57BL6J_1 C57BL6J_2 C57BL6J_3 C57BL6J_4 C57BL6J_5 MOV13_1 MOV13_2 MOV13_3 MOV13_4 MOV13_5 NFkB1 CSF1 C-REL TANK TGFB1 BCL10 PAR3 RANK CARMA1 IL1b RELA BMP4 RELB TNFa BMP3 GRFA1 TSHR IL1A E-CAD1 CX30 SOCS2 COL1A1 HSP70-2 TNFA1R GR SOCS3 BMP7 SMAD4

C

B

A

BMP7 CARMA1 SOCS3 SOCS1 HSP70-2 CX30 TLR5 BCL10 SOCS2 PAR3 CX26 TLL RANKL TLR4 PTHR TGFB2 TAIP-12 BMPR1B TAIP-2 SOST SPARC EGFR TGFB1 BMPR2 BMP3 OPG SMAD4 TNFB1R RANK GFRA1 BCL2 COL1A2 BMPR1A TGFB2R

TLL RANK BMP7 TNFB1R CSF1 PTHR NFkB1 RELB TAIP-12 HSP70-2 BMPR1A COL1A1 CARMA1 TAIP-2 COL1A2 CX26 SOCS1 TLR4 TNFa IL1a TANK GFRA2 SPARC GR BMPR2 IL1B SMAD4 TLR5 BMP3 TNFa1R CX30 RANKL IAP2 TGFB2R

otic capsule, 34 in the parietal bone and 34 in the tibia. Statistically significant changes in gene expression were at least 1.6 fold. The figure illustrates that measurements were highly reproducible among animals and bone types because samples from the same animal type and bone type had similar color coding. When compared with wild types, Mov13 otic capsules exhibited downregulation of all genes, except bone morphogenetic protein 3 (BMP3) (fig. 2A). Similarly, all genes were downregulated in

Mov13 tibias (fig. 2B). In contrast, the majority of genes were upregulated in Mov13 parietal bones (fig. 2C). Having established that expression levels for multiple genes are different in Mov13 mice compared with wildtype controls in a bone-type-specific manner, we wanted to determine the smallest set of genes that best characterize the otic capsule. These genes, known as predictor genes, are not necessarily genes that exhibit the largest difference in expression between 2 groups but genes that



339


Fig. 2. Heat map of real-time quantitative RT-PCR data comparing the otic capsule (A), tibia (B) and parietal bone (C) in C57BL/6J and Mov13 mice. Red indicates high and blue low levels of expression.

MOV13 OTIC CAPSULE 1 MOV13 OTIC CAPSULE 2 MOV13 OTIC CAPSULE 3 MOV13 OTIC CAPSULE 4 MOV13 OTIC CAPSULE 5 MOV13 PARIETAL BONE 1 MOV13 PARIETAL BONE 2 MOV13 PARIETAL BONE 3 MOV13 PARIETAL BONE 4 MOV13 RARIETAL BONE 5 MOV13 OTIC CAPSULE 1 MOV13 OTIC CAPSULE 2 MOV13 OTIC CAPSULE 3 MOV13 OTIC CAPSULE 4 MOV13 OTIC CAPSULE 5 MOV13 TIBIA 1 MOV13 TIBIA 2 MOV13 TIBIA 3

OPG COL2A1 E-CAD1 BMPR1B SOST CX30 CX26 TGFB2R PTHR SOCS3 TLR4 SMAD4 RANKL RELA BCL2 IL1a CARMA1 COL1A2 SPARC TANK COL1A1 TGFB1 C-REL TNFa IAP2 BCL10 TNFb1R TSHR TLR5 PAR3 NFkB1 IL1b

A

have a small variance within a group in addition to a substantial difference between groups [Golub et al., 1999]. We found, using K-nearest neighbors cross-validation [Golub et al., 1999], that genes that best characterized the Mov13 otic capsule were BMP3 and nuclear factor -B1 (NFkB1). These 2 genes were used in 18 out of 20 crossvalidation steps to correctly predict 8 out of 10 samples. Genes whose expression best characterized the Mov13 tibia were TLL and TLR4; they were used 11 out of 16 times to correctly predict 6 out of 8 samples. Genes that best characterized the Mov13 parietal bone were 340


B

BMPR1B and BMP7; they were sufficient to correctly predict all 9 samples. Overall, genes that best characterize the Mov13 parietal bone and otic capsule were more robust than genes that best characterize the Mov13 tibia. In addition to comparing Mov13 and wild-type bones, bones within the Mov13 mouse were compared. The results are shown in figure 3. Thirty-two genes were differentially expressed in the otic capsule compared to parietal bone and 22 genes in the otic capsule compared to tibia.



Fig. 3. Heat map of real-time quantitative RT-PCR data comparing the otic capsule with tibia (A) and parietal bone (B) in Mov13 mice. Red indicates high and blue low levels of gene expression.

OPG SOST COL2A1 CX30 CX26 E-CAD1 TGFB2 TGFB2R BMPR1B TAIP-12 TNFa1R PAR3 TSHR C-REL IL1b SPARC COL1A2 GFRA2 TANK TNFa COL1A1 CARMA1

Discussion

The Mov13 mouse turned out not to be an animal model for active remodeling of the otic capsule associated with hearing loss in type I OI or otosclerosis. Altschuler et al. [1991] also failed to detect active remodeling in the Mov13 otic capsule, although they qualitatively described new bone formation within the otic capsule. The current study provides quantitative support for the qualitative observation of Altschuler et al. [1991], with a distinction that we note increased bone thickness in Mov13 mice limited to the apical region of greatest marrow space and accompanied by decreased area occupied by bone marrow. That apical region was not included in our gene expression assay. Although it is known that there is an interplay of osteogenesis and hematopoiesis [Kuznetsov et al., 2004], the molecular bases for the increased osteogenesis and the decreased volume of marrow space in Mov13 mice remain unknown. Our results imply that transcriptional control of the COL1A1 gene in ostecytes of the otic capsule is more similar to that in fibroblasts than to that in odontoblasts because the ratio of the COL1A1 to COL1A2 mRNA was 50% reduced in the otic capsule of Mov13 mice, and our bone samples were enriched for osteocytes, which account for 90% of all cells in the skeleton. The 50% reduction in the COL1A1 mRNA (and in the COL1A1/COL1A2 mRNA ratio) in the otic capsule is consistent with a near complete transcriptional block of the COL1A1 allele as described in fibroblasts [Breindl et al., 1984] and different from the lack of transcriptional block in odontoblasts [Kratochwil et al., 1989]. Elevations in the COL1A1 to COL1A2 mRNA ratio and subsequent elevations in collagen 1(I) to 2(I) protein ratio have been described in osteoblasts cultured from patients with osteoporosis [Mann et al., 2001]. These elevated ratios are thought to promote formation of collagen 1(I) homotrimers, and Gene Expression in the Mov13 Otic Capsule

the deposition of 1(I) homotrimers in the extracellular matrix is associated with OI in humans [Deak et al., 1985] and mice [Chipman et al., 1993]. Although we did not measure the collagen 1(I) to 2(I) protein ratio in bone samples, it is likely that the ratio is decreased in the Mov13 otic capsule based on the decreased COL1A1/COL1A2 mRNA ratio. Our histological analysis suggests that middle ear effusions are a major contributor to the conductive hearing loss in the Mov13 mice. Although our reported ratios of COL1A1 to COL1A2 mRNA in the wild-type mice are higher than previously reported [Vuust et al., 1985; Rowe et al., 1985; Dawson and Marini, 2000], the measurements are consistent across animals. This divergence may reflect a technical difference, since we used real-time quantitative PCR and prior reports used different techniques that included Northern blots [Vuust et al., 1985], dot hybridization [Rowe et al., 1985] and competitive RT-PCR analyses [Dawson and Marini, 2000]. Moreover, using stringent criteria for statistical significance, we found a significant difference between COL1A1 and COL1A2 mRNA levels only in the otic capsule, not other bones. The elevated ratio in the otic capsule may reflect a fundamental biologic difference between that bone and other bones that contributes to the unique resistance of the otic capsule to remodeling. Studying the Mov13 mouse allowed us to establish that reduced expression of the COL1A1 gene affects expression of many other genes within the otic capsule in different ways than in the tibia and parietal bone, thus suggesting the existence of bone-specific regulatory pathways downstream from COL1A1. Expression of COL1A1 did not best characterize any of the Mov13 bones, although altered expression of the COL1A1 gene was a major reason for altered expression of all other genes that we studied. This illustrates the principle that small changes in expression of 1 gene can cause large changes in expression of downstream genes, possibly through activation of transcriptional regulators. The genes identified as being highly characteristic of Mov13 bones may not be unique because we did not study the whole genome. Nonetheless, the identified genes provide insights into the molecular basis for the uniqueness of the otic capsule. In the current study, OPG emerged as the gene whose expression best characterizes the Mov13 otic capsule by virtue of its higher expression in the otic capsule than other bones. We recently found that expression of OPG also best characterizes the wild-type otic capsule (unpublished data). Our previous work showed that OPG plays an important role in the inhibition of otic capsule remodAudiol Neurotol 2007;12:334–343

341


The pattern of gene expression levels in the otic capsule was distinctly different from the other 2 bones. The genes that best characterize the Mov13 otic capsule when compared with tibia were partitioning defective protein 3 (PAR3) and osteoprotegerin (OPG); they were used 12 out of 16 times to correctly predict all 8 samples. Genes that best characterize the otic capsule when compared with parietal bone were caspase recruitment domain containing protein 1 (CARMA1) and OPG; they were used 13 out of 20 times to correctly predict 9 out of 10 samples.

eling [Zehnder et al., 2005], and that the otic capsule of the OPG knockout mice undergoes abnormal remodeling that closely resembles otosclerosis [Zehnder et al., 2006]. Because we did not observe active remodeling in the otic capsule of Mov13 mice, it is not surprising that expression levels of OPG remained high in the Mov13 otic capsule, thus contributing to the selection of OPG as a gene that is highly characteristic of the otic capsule. Another gene that characterizes both the wild-type and Mov13 otic capsule is BMP3 by virtue of its low expression in the otic capsule compared with other bones. This gene is a member of the transforming growth factor- supergene family, and it is a negative regulator of bone formation [Daluiski et al., 2001]. BMP3 was previously described in spiral ganglion cells and spiral limbus [Thomadakis et al., 1999], but its role in the otic capsule has not been studied. However, the importance of BMP pathways for the formation of the otic capsule and epithelial-mesenchymal signaling was previously established in chicken embryos [Chang et al., 1999; Gerlach et al., 2000; Chang et al., 2002]. In the current study, BMP3 was the only gene that was upregulated in the Mov13 otic capsule when compared to the wild-type otic capsule, and thus it emerged as a highly characteristic gene despite the magnitude of the upregulation being small (1.6 8 0.3 fold, p = 0.02). Although changes in mRNA levels are not necessarily translated into changes in protein levels, the increase in mRNA encoding for a bone inhibitor is consistent with the absence of active bone remodeling in Mov13 otic capsule. Other genes that best characterize the Mov13 otic capsule but not the wild-type otic capsule include NFkB1,

CARMA1 and PAR3. The roles of these genes have not been previously studied in the otic capsule. NFkB1 is a part of the NFB complex, a nuclear transcription factor that regulates expression of a large number of genes that are critical for the regulation of bone remodeling, inflammation, tumorigenesis, apoptosis, viral replication and various autoimmune diseases [Chen et al., 1999]. The measured reduction in expression of NFkB1 in both the otic capsule and tibia of the Mov13 mouse is consistent with the observation that the NFkB1/NFkB2 double knockout mouse develops osteoporosis due to a defect in osteoclast differentiation [Iotsova et al., 1997]. Coordinated reduction in CARMA1 and NFkB1 is not surprising because CARMA1 is known to induce NFkB activity [Bertin et al., 2001]. PAR3 is central to the development of junctional structures and apical-basolateral polarization in epithelial cells [Suzuki et al., 2001]. The role of PAR3 in the physiology of the otic capsule is unknown. Taken together, the results of this study suggest that there are fundamental differences between osteocytes within the otic capsule and those present in other bones. These differences may be the basis for the otic capsule being relatively refractory to remodeling.

Acknowledgments We thank Dr. Chitsuda Wacharasindhu for preliminary morphometric analysis of the Mov13 mice. We are grateful to Dr. Saumil Merchant for helpful comments on earlier versions of the manuscript. This study was supported by NIDCD grant 5RO1 DC03401-08 and funds from Mr. Lakshmi Mittal.

References

342

Bonadio J, Saunders TL, Tsai E, Goldstein SA, Morris-Wiman J, Brinkley L, Dolan DF, Altschuler RA, Hawkins JE Jr, Bateman JF, et al: Transgenic mouse model of the mild dominant form of osteogenesis imperfecta. Proc Natl Acad Sci USA 1990;87:7145–7149. Breindl M, Harbers K, Jaenisch R: Retrovirusinduced lethal mutation in collagen I gene of mice is associated with an altered chromatin structure. Cell 1984;38:9–16. Chang W, ten Dijke P, Wu DK: BMP pathways are involved in otic capsule formation and epithelial-mesenchymal signaling in the developing chicken inner ear. Dev Biol 2002; 251:380–394.


Chang W, Nunes FD, De Jesus-Escobar JM, Harland R, Wu DK: Ectopic noggin blocks sensory and nonsensory organ morphogenesis in the chicken inner ear. Dev Biol 1999; 216: 369–381. Chen F, Castranova V, Shi X, Demers LM: New insights into the role of nuclear factor-kappaB, a ubiquitous transcription factor in the initiation of diseases. Clin Chem 1999; 45:7– 17. Chen W, Meyer NC, McKenna MJ, Pfister M, McBride DJ, Fukushima K, Thys M, Van Camp G, Smith RJH: Single nucleotide polymorphisms in the COL1A1 regulatory regions are associated with otosclerosis. Clin Genet, in press.



Altschuler RA, Dolan DF, Ptok M, Gholizadeh G, Bonadio J, Hawkins JE: An evaluation of otopathology in the MOV-13 transgenic mutant mouse. Ann NY Acad Sci 1991;630:249– 252. Bast TH, Anson BJ: The Temporal Bone and the Ear. Springfield, Thomas, 1949. Bertin J, Wang L, Guo Y, Jacobson MD, Poyet JL, Srinivasula SM, Merriam S, DiStefano PS, Alnemri ES: CARD11 and CARD14 are novel caspase recruitment domain (CARD)/ membrane-associated guanylate kinase (MAGUK) family members that interact with BCL10 and activate NF-kappa-B. J Biol Chem 2001;276:11877–11882.


Kratochwil K, Ghaffari-Tabrizi N, Holzinger I, Harbers K: Restricted expression of Mov13 mutant alpha 1(I) collagen gene in osteoblasts and its consequences for bone development. Dev Dyn 1993;198:273–283. Kratochwil K, von der Mark K, Kollar EJ, Jaenisch R, Mooslehner K, Schwarz M, Haase K, Gmachl I, Harbers K: Retrovirus-induced insertional mutation in Mov13 mice affects collagen I expression in a tissue-specific manner. Cell 1989;57:807–816. Kuurila K, Kaitila I, Johansson R, Grenman R: Hearing loss in Finnish adults with osteogenesis imperfecta: a nationwide survey. Ann Otol Rhinol Laryngol 2002; 111: 939– 946. Kuznetsov SA, Riminucci M, Ziran N, Tsutsui TW, Corsi A, Calvi L, Kronenberg HM, Schipani E, Robey PG, Bianco P: The interplay of osteogenesis and hematopoiesis: expression of a constitutively active PTH/ PTHrP receptor in osteogenic cells perturbs the establishment of hematopoiesis in bone and of skeletal stem cells in the bone marrow. J Cell Biol 2004,167:1113–1122. Lillie RD: Histologic Technic and Practical Histochemistry. New York, McGraw-Hill, 1965. Livak KJ, Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCR and the 2–CT method. Methods 2001;25:402–408. Mann V, Hobson EE, Li B, Stewart TL, Grant SF, Robins SP, Aspden RM, Ralston SH: A COL1A1 Sp1 binding site polymorphism predisposes to osteoporotic fracture by affecting bone density and quality. J Clin Invest 2001; 107:899–907. McKenna MJ, Kristiansen AG, Bartley ML, Rogus JJ, Haines JL: Association of COL1A1 and otosclerosis: evidence for a shared genetic etiology with mild osteogenesis imperfecta. Am J Otol 1998;19:604–610. McKenna MJ, Kristiansen AG, Tropitzsch AS: Similar COL1A1 expression in fibroblasts from some patients with clinical otosclerosis and those with type I osteogenesis imperfecta. Ann Otol Rhinol Laryngol 2002;111:184– 189. Nager GT: Osteogenesis imperfecta of the temporal bone and its relation to otosclerosis. Ann Otol Rhinol Laryngol 1988; 97: 585– 593. Pedersen U, Melsen F, Elbrond O, Charles P: Histopathology of the stapes in osteogenesis imperfecta. J Laryngol Otol 1985; 99:451–458. Reno C, Marchuk L, Sciore P, Frank CB, Hart DA: Rapid isolation of total RNA from small samples of hypocellular, dense connective tissues. Biotechniques 1997;22:1082–1086.

Rowe DW, Shapiro JR, Poirier M, Schlesinger S: Diminished type I collagen synthesis and reduced alpha 1(I) collagen messenger RNA in cultured fibroblasts from patients with dominantly inherited (type I) osteogenesis imperfecta. J Clin Invest 1985;76:604–611. Schuknecht H: Pathology of the Ear. Philadelphia, Lea & Febiger, 1993. Schnieke A, Harbers K, Jaenisch R: Embryonic lethal mutation in mice induced by retrovirus insertion into the alpha 1(I) collagen gene. Nature 1983;304:315–320. Shapiro JR, Pikus A, Weiss G, Rowe DW: Hearing and middle ear function in osteogenesis imperfecta. JAMA 1982;247:2120–2126. Stankovic KM, Corfas G: Real-time quantitative RT-PCR for low-abundance transcripts in the inner ear: analysis of neurotrophic factor expression. Hear Res 2003;185:97–108. Stankovic KM, McKenna MJ: Current research in otosclerosis. Curr Opin Otolaryngol Head Neck Surg 2006;14:347–351. Suzuki A, Yamanaka T, Hirose T, Manabe N, Mizuno K, Shimizu M, Akimoto K, Izumi Y, Ohnishi T, Ohno S: Atypical protein kinase C is involved in the evolutionarily conserved PAR protein complex and plays a critical role in establishing epithelia-specific junctional structures. J Cell Biol 2001;152:1183–1196. Thomadakis G, Ramoshebi LN, Crooks J, Rueger DC, Ripamonti U: Immunolocalization of bone morphogenetic protein-2 and -3 and osteogenic protein-1 during murine tooth root morphogenesis and in other craniofacial structures. Eur J Oral Sci 1999;107:368– 377. Vuust J, Sobel ME, Martin GR: Regulation of type I collagen synthesis: total pro-alpha 1(I) and pro-alpha 2(I) mRNAs are maintained in a 2:1 ratio under varying rates of collagen synthesis. Eur J Biochem 1985;151:449–453. Wang PC, Merchant SN, McKenna MJ, Glynn RJ, Nadol JB Jr: Does otosclerosis occur only in the temporal bone? Am J Otol 1999; 20: 162–165. Zehnder AF, Kristiansen AG, Adams JC, Merchant SN, McKenna MJ: Osteoprotegerin in the inner ear may inhibit bone remodeling in the otic capsule. Laryngoscope 2005; 115: 172–177. Zehnder AF, Kristiansen AG, Adams JC, Merchant SN, McKenna MJ: Osteoprotegerin knockout mice demonstrate abnormal remodeling of the otic capsule and progressive hearing loss. Laryngoscope 2006; 116: 201– 206.


343


Chipman SD, Sweet HO, McBride DJ Jr, Davisson MT, Marks SC Jr, Shuldiner AR, Wenstrup RJ, Rowe DW, Shapiro JR: Defective pro-alpha 2(I) collagen synthesis in a recessive mutation in mice: a model of human osteogenesis imperfecta. Proc Natl Acad Sci USA 1993;90:1701–1705. Daluiski A, Engstrand T, Bahamonde ME, Gamer LW, Agius E, Stevenson SL, et al: Bone morphogenetic protein-3 is a negative regulator of bone density. Nat Genet 2001;27:84– 88. Dawson PA, Marini JC: Hammerhead ribozymes selectively suppress mutant type I collagen mRNA in osteogenesis imperfecta fibroblasts. Nucleic Acids Res 2000;28:4013–4020. Deak SB, van der Rest M, Prockop DJ: Altered helical structure of a homotrimer of alpha 1(I) chains synthesized by fibroblasts from a variant of osteogenesis imperfecta. Coll Relat Res 1985;5:305–313. Efron B, Tibshirani RJ: An Introduction to the Bootstrap. New York, Chapman & Hall, 1993. Frisch T, Sorensen MS, Overgaard S, Bretlau P: Predilection of otosclerotic foci related to the bone turnover in the otic capsule. Acta Otolaryngol Suppl 2000; 543:111–113. GenePattern 2.0. http://www.broad.mit.edu/genepattern/ (accessed January 30, 2006). Gerlach LM, Hutson MR, Germiller JA, NguyenLuu D, Victor JC, Barald KF: Addition of the BMP4 antagonist, noggin, disrupts avian inner ear development. Development 2000; 127:45–54. Gilbert SF: Developmental Biology, ed 6. Sunderland, Sinauer Associates Inc, 2000, part 3, chapt 14. Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek M, Mesirov JP, Coller H, Loh ML, Downing JR, Caligiuri MA, Bloomfield CD, Lander ES: Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 1999;286:531–537. Holdsworth CE, Endahl GL, Soifer N, Richardson KE, Eyring EJ: Comparative biochemical study of otosclerosis and osteogenesis imperfecta. Arch Otolaryngol 1973; 98: 336– 369. Iotsova V, Caamano J, Loy J, Yang Y, Lewin A, Bravo R: Osteopetrosis in mice lacking NFkappa-B1 and NF-kappa-B2. Nat Med 1997; 3:1285–1289.