ORIGINAL ARTICLE 2 Abnormal Mesenchymal Differentiation in the Superior Semicircular Canal of Brn4/Pou3f4 Knockout Mice

Steven E. Sobol, MD, MSc; Xiuyin Teng, MSc; E. Bryan Crenshaw III, PhD

Objective: To examine the developmental time course of the SSCC in Brn4 knockout mice was initially detect- of the mutant phenotype and cellular mechanisms that able at P14. Interestingly, the mutant SSCC is indistin- result in malformations of the superior semicircular ca- guishable from control mice at earlier neonatal time points. nal (SSCC) in Brn4 knockout mice. Mutations in the Brn4/ In mutant neonates, there is persistence of immature wo- Pou3f4 result in characteristic inner ear abnormali- ven bone with high cellularity surrounding the perilym- ties in mutant mouse pedigrees, and the findings in these phatic space of the SSCC. These findings are not pres- mice are similar to those in human X-linked deafness type ent in control animal specimens, which demonstrate III. appropriate lamellar bony architecture.

Design: Mutant and control mice were killed at vari- Conclusions: In Brn4 knockout mice, constriction of the ous neonatal time points to assess the development of SSCC with narrowing of the bony labyrinth develops in the SSCC. Measurements of SSCC diameter were made the postnatal period at approximately P14. The persis- on paint-perfused specimens at postnatal day (P) 0, P7, tence of immature bone in affected mice indicates that P10, and P14. Histologic evaluation of the SSCC was signaling abnormalities disrupt normal mesenchymal dif- made on hematoxylin-eosin–stained sections at P10. ferentiation in the SSCC.

Results: A dysmorphic constriction of the superior arc Arch Otolaryngol Head Neck Surg. 2005;131:41-45

ORMAL INNER EAR DEVEL- have revealed multiple defects in affected opment results from a individuals, including stapes footplate fixa- complex series of mor- tion, dilatation of the internal auditory phogenetic changes that canal (IAC), reduced diameter of the semi- occur in response to in- circular canals, and varying degrees of coch- teractions between epithelial and mesen- lear dysplasia.5,7 High-resolution com- N 1 Author Affiliations: Division chymal embryonic tissues. The molecu- puted tomography has demonstrated that of Pediatric Otolaryngology and lar signaling events underlying normal there is a reduction of bone thickness at the Mammalian Neurogenetics inner ear development have yet to be fully cochlear modiolus.8,9 Group, The Center for elucidated. The recent development of X-linked deafness type 3 results from Childhood Communication, genetic approaches provides the means to mutations in the POU3F4 gene,10,11 a mem- The Children’s Hospital of Philadelphia (Drs Sobol and study the underlying pathogenesis of he- ber of the POU domain family of tran- Crenshaw and Ms Teng); and reditary auditory dysfunction in animal scription factors. POU domain play Department of models of human congenital malforma- critical roles in the development of the cen- Otorhinolaryngology–Head and tions. tral nervous system and the inner ear as Neck Surgery, University of X-linked deafness type 3 is the most well as other organ systems.12-22 Linkage Pennsylvania School of common type of X-linked hearing loss in analysis has localized the POU3F4 gene to Medicine (Drs Sobol and humans, accounting for 0.8% of all cases the Xq13-21.1 region.23,24 Crenshaw), Philadelphia. of hereditary hearing loss2,3 and character- The Brn4 gene (also called Pou3f4)is Dr Sobol is currently with ized by early-onset mixed hearing loss with the murine ortholog of the human POU3F4 the Department of a high incidence of perilymphatic gusher gene. This gene is expressed early in the Otolaryngology–Head and Neck 4,5 Surgery, Emory University during stapedectomy. Many patients with development of the nervous system and is School of Medicine, X-linked deafness type 3 display abnor- the only gene in the POU domain family Atlanta, Ga. mal vestibular responses on electronystag- expressed in the mesenchyme of the de- Financial Disclosure: None. mographic studies.5,6 Radiologic studies veloping inner ear.25-29 The gene has been

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 131, JAN 2005 WWW.ARCHOTO.COM 41

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 localized by immunocytochemical analysis to the mes- (P) 0 (9 controls, 8 mutants), P7 (9 controls, 12 mutants), P10 enchymal cells of the developing otic capsule and plays (5 controls, 20 mutants), and P14 (14 controls, 10 mutants). a role in the regulation of normal mesenchymal remod- These whole-mount preparations were obtained by bisecting eling, epithelial-mesenchymal interactions, and mesen- the head in the sagittal plane, dissecting out the temporal bone, chymal-mesenchymal interactions, which are vital for nor- and placing the specimens in Bodian fixative (75% ethanol, 15% 30,31 water, 5% formalin, and 5% glacial acetic acid) for 24 to 48 hours. mal labyrinthine development. Preparations were then dehydrated through a graded series of Mutations in Brn4 result in a number of characteristic ethanols and cleared in methyl salicylate. Inner ears were per- behavioral abnormalities in mutant mouse pedigrees, in- fused through the oval window with a suspension of 1% gloss cluding hearing loss, vertical head bobbing, changes in gait, paint in methyl salicylate using a 33 gauge standard Hamilton and reduced whisker mobility.30-32 A number of studies have needle. These paint fills visualized the bony labyrinth within evaluated auditory function as well as middle and inner ear the temporal bones of these mice. structure in Brn4 knockout mice. Evaluation of the audi- tory brainstem response in mutant mice has demon- HISTOLOGIC ANALYSIS strated severe hearing loss and evidence of cochlear dys- function.31 Abnormal fibrocyte structure within the spiral Anesthetized mutant and control animals (6 of each) were fixed ligament of Brn4 knockout mice has been hypothesized to by cardiac perfusion using 4% paraformaldehyde in phosphate- result in an abnormal endolymphatic potential, which may buffered saline on P10. Preparations were obtained by bisect- explain these auditory brainstem response results. Knock- ing the head in the sagittal plane, dissecting out the temporal out alleles of the Brn4 gene have demonstrated multiple bone, and placing the specimens in 4% paraformaldehyde for malformations of tissues derived from the otic mesen- overnight immersion fixation. Specimens were washed 3 times using phosphate-buffered chyme, including hypoplasia of the temporal bone, en- saline and decalcified in 0.1M EDTA for 2 to 3 days. After de- larged internal auditory meatus, dysplastic fibrocytes, and 30,31 calcification, temporal bone specimens were serially dehy- malformed stapes. Other morphologic abnormalities drated in alcohol, embedded in paraffin, and sectioned (8 µm) identified in knockout mice have included cochlear hypo- in the coronal plane (cross section through the SSCC) in prepa- plasia, a reduction in the number of cochlear turns, and ration for staining with hematoxylin-eosin. endolymphatic hydrops. Vestibular abnormalities of Brn4 knockout mice give DATA ANALYSIS rise to the vertical head-bobbing phenotype detected in the mutant mice. The vertical nature of the aberrant head Measurements of SSCC diameter were made on paint- motion prompted evaluation of the superior semicircu- perfused specimens and reported in micrometers as lar canal (SSCC, which lies in the vertical plane) by paint mean±standard error of the mean (SEM). On sections stained perfusion in studies by Phippard et al.30,32 Findings in- with hematoxylin-eosin, the narrowest point of the SSCC peri- cluded a constriction of the bony labyrinth surrounding lymphatic space was identified for each sample. To account for the canal with a reduction in the thickness of the bone the natural variability that may exist between specimens, a fixed in this area in adult mutant mice. This was not observed point anterior to the common crus that could be consistently in the posterior semicircular canal and was only rarely located was identified in all samples. Multiple cross-sectional area measurements were made at these points on the hema- observed in the horizontal canal. toxylin-eosin–stained sections by tracing and analysis using the The first objective of the present study was to delin- National Institutes of Health Image J program (version 1.32; eate the timing of the development of the SSCC constric- available at http://rsb.info.nih.gov/ij/. The mean cross- tion by analyzing the canal diameter of Brn4 knockout sectional area of the narrowest and fixed points was reported and control mice at different postnatal time points us- in square micrometers ± SEM. ing paint perfusion analyses. The second goal was to in- Statistical analysis was performed using a 2-tailed t test with vestigate the histologic findings within the mesenchy- PϽ.01 used for statistical significance. All animal protocols used mal compartment of mutant and control animal SSCCs during this research were approved by the institutional ani- at a time point preceding the constriction. We postu- mal care and use committee at The Children’s Hospital of Phila- lated that the vestibular abnormalities observed in Brn4 delphia in accordance with regulations established by the Na- tional Institutes of Health. knockout mice were the result of aberrant mesenchy- mal differentiation in the developing SSCC. RESULTS METHODS SSCC DIAMETER Mouse pedigrees containing a null mutation of the Brn4 gene have been described.30 CD1 and wild-type littermate mice served Although there was a trend of decreasing SSCC diam- as controls. Normal and Brn4 knockout mice were killed at sev- eter in Brn4 knockout mice with increasing develop- eral defined postnatal time points to assess the development mental age, no statistically significant difference was of the SSCC. found between mutant mice and controls at P0, P7, and P10. The mean±SEM diameters of the SSCC in P0, P7, PAINT PERFUSION ANALYSIS and P10 mutant mice were 194.8±4.2 µm, 149.5±10.3 µm, and 126.1±8.3 µm, respectively. The mean diam- In preparation for paint perfusion analyses, Brn4 knockout and eters of the SSCC in P0, P7, and P10 control mice were control mice were anesthetized and killed by cervical disloca- 197.4±4.4 µm, 168.2±4.2 µm, and 143.0±8.1 µm, re- tion. Temporal bone preparations were made on postnatal day spectively.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 131, JAN 2005 WWW.ARCHOTO.COM 42

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 A discrete constriction of the superior arc of the SSCC 250 was present in P14 Brn4 knockout mice but not con- Control Mutant trols. The mean±SEM diameter of the narrowest por- tion of the SSCC was significantly smaller in P14 mu- tant mice (78.3±10.1 µm) than in controls (171.4±4.1 200 µm) (PϽ.001) (Figure 1 and Figure 2). ∗

SSCC PERILYMPHATIC CROSS-SECTIONAL AREA ANALYSIS 150

The mean±SEM cross-sectional area of the perilym- phatic space at the narrowest point of the SSCC was sig- 100 nificantly smaller in mutant mice (16.3 ϫ 103±1.3 ϫ 103 µm SSCC Diameter, µm2) than in controls (21.8 ϫ 103 ±0.8 ϫ 103 µm2) (PϽ.01). The mean cross-sectional area of the perilym- phatic space at the fixed point of the SSCC was not sig- 50 nificantly different in mutant mice (56.2ϫ 103±6.5ϫ103 µm2) and controls (53.3ϫ103±12.9ϫ103 µm2)(Figure 3).

0 MORPHOLOGIC ANALYSIS P0 P7 P10 P14

Evaluation of hematoxylin-eosin–stained sections Figure 1. Paint perfusion analyses demonstrate a trend of decreasing demonstrated a deficiency of mature lamellar bone superior semicircular canal (SSCC) diameter from postnatal day (P) formation surrounding the perilymphatic space in the 0 through P14 in Brn4 knockout mice. Results at P14 demonstrate a statistically significant difference in the diameter of the SSCC in mutant mice SSCC of mutant mice. There was persistence of imma- compared with controls. Error bars indicate standard error of the mean; ture woven bone with high cellularity throughout the asterisk indicates PϽ.001.

A B

C D

Figure 2. Medial view of paint-perfused superior (SSCCs) of postnatal day (P) 0 mice (A and B) and P14 mice (C and D). Control mice are shown in panels A and C; the Brn4 knockout mice in panels B and D. The arrow in panel B indicates the normal superior arc of the SSCC in a mutant mouse at P0. The arrow in panel D indicates the position of the SSCC constriction at P14 in a mutant mouse. Note the absence of this narrowing in the mutant mouse at P0 and in the controls. The scale bar represents 500 µm.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 131, JAN 2005 WWW.ARCHOTO.COM 43

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 SSCC. These findings were not present in control ani- COMMENT mal specimens, which demonstrated appropriate lamellar bony architecture surrounding the perilym- The findings in this study demonstrate that the Brn4/ phatic space (Figure 4). Pou3f4 gene inactivation results in aberrant differentia- tion of the bone surrounding the SSCC and a charac- teristic constriction in the superior arc of this canal. Our first goal was to determine the time course of the SSCC constriction. We observed a trend of decreasing Control Mutant SSCC diameter from P0 through P14, which was sig- 70 nificantly different from the control group at P14. Although there was no statistical difference in the

60 canal diameter at P10, cross-sectional area measure-

3 ments of the perilymphatic space on histologic sec- 10

× tions were significantly smaller in Brn4 knockout ani- 2 50 mals than in controls. These findings suggest that the observed narrowing of the SSCC may be due to defi- 40 ciencies in mesenchymal remodeling that occur post- natally. Once the developmental time course of the SSCC 30 ∗ constriction was delineated, we evaluated the morpho- logic changes in the mutant SSCC at a time point imme- SSCC Cross-sectional Area, µm 20 diately preceding the constriction. In control animal specimens, there was appropriate lamellar bony archi-

10 tecture surrounding the perilymphatic space. Histologic evaluation of Brn4 knockout mice demonstrated a defi- ciency of lamellar bone formation. Instead there was 0 Narrow Point Fixed Point persistence of immature woven bone with high cellular- ity found throughout the SSCC. Figure 3. Cross-sectional area measurement of the superior semicircular Our observations are consistent with other temporal canal (SSCC) perilymphatic space. Note that the narrowest point measure is bone defects seen in Brn4 knockout mice that are asso- significantly smaller in mutant mice than in controls, whereas there is no 30 difference between the 2 groups at a fixed point anterior to the common ciated with osteogenic abnormalities. Phippard et al dem- crus. Error bars indicate standard error of the mean; asterisk, PϽ.01). onstrated that adult mutant mice have thinning of the

A B

Figure 4. Cross-sectional views of the superior semicircular canal in control (A) and Brn4 knockout (B) mice at the narrow point (hematoxylin-eosin, original magnification ϫ20). Note the hypercellularity and persistence of immature woven bone surrounding the perilymphatic space in mutant mice (white arrow). These findings were not present in control animal specimens, which demonstrated lamellar bony architecture surrounding the perilymphatic space (black arrow).

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 131, JAN 2005 WWW.ARCHOTO.COM 44

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 stapes footplate, an enlarged internal auditory canal, and 8. Phelps PD, Reardon W, Pembrey M, et al. X-linked deafness, stapes gushers and a reduction in the thickness of bone surrounding the a distinctive defect of the inner ear. Neuroradiology. 1991;33:326-330. 9. Tang A, Parnes LS. X-linked progressive mixed hearing loss: computed tomog- SSCC. Fibrocytes within the spiral ligament of mutant raphy findings. Ann Otol Rhinol Laryngol. 1994;103:655-657. mice were found to be dysplastic with a reduced num- 10. de Kok YJ, Merkx GF, van der Maarel SM, et al. A duplication/paracentric inver- ber of cytoplasmic extensions and mitochondria.30,31 This sion associated with familial X-linked deafness (DFN3) suggests the presence finding was consistently present throughout all turns of of a regulatory element more than 400 kb upstream of the POU3F4 gene. Hum the and suggests that Brn4 is essential for nor- Mol Genet. 1995;4:2145-2150. 11. de Kok YJ, van der Maarel SM, Bitner-Glindzicz M, et al. Association between mal fibrocyte differentiation. Because Brn4 is expressed X-linked mixed deafness and mutations in the POU domain gene POU3F4. Science. exclusively within the mesenchymal tissues of the de- 1995;267:685-688. veloping inner ear, we hypothesize that its gene product 12. Corcoran LM, Karvelas M, Nossal GJ, Ye ZS, Jacks T, Baltimore D. Oct-2, al- plays a critical role in mesenchymal cell differentiation. though not required for early B-cell development, is critical for later B-cell matu- Mutation of the Brn4 gene may alter osteoprogenitor cell ration and for postnatal survival. Genes Dev. 1993;7:570-582. 13. Erkman L, McEvilly RJ, Luo L, et al. Role of transcription factors Brn-3.1 and Brn- differentiation, which may explain the abnormal bone 3.2 in auditory and visual system development. Nature. 1996;381:603-606. structure seen in mutant mice. 14. Feldhaus AL, Klug CA, Arvin KL, Singh H. Targeted disruption of the Oct-2 locus In conclusion, we have demonstrated that a constric- in a B cell provides genetic evidence for two distinct cell type-specific pathways tion of the SSCC in Brn4 knockout mice results from nar- of octamer element-mediated gene activation. EMBO J. 1993;12:2763-2772. rowing of the perilymphatic space. The persistence of im- 15. Anderson MG, Perkins GL, Chittick P, Shrigley RJ, Johnson WA. Drifter, a [sic] Drosophila POU-domain , is required for correct differentia- mature bone in the SSCC in mutant mice suggests that tion and migration of tracheal cells and midline glia. Genes Dev. 1995;9:123- there are signaling abnormalities preventing normal dif- 137. ferentiation of the canal. These findings may help us fur- 16. Li S, Crenshaw EB, Rawson EJ, Simmons DM, Swanson LW, Rosenfeld MG. ther understand the pathologic mechanisms underlying Dwarf locus mutants lacking three pituitary cell types result from mutations in X-linked deafness type 3 in humans. the POU-domain gene pit-1. Nature. 1990;347:528-533. 17. Nakai S, Kawano H, Yudate T, et al. The POU domain transcription factor Brn-2 is required for the determination of specific neuronal lineages in the hypothala- Submitted for Publication: June 9, 2004; final revision mus of the mouse. Genes Dev. 1995;9:3109-3121. received August 24, 2004; accepted September 16, 2004. 18. Pfaffle RW, DiMattia GE, Parks JS, et al. Mutation of the POU-specific domain of Correspondence: E. Bryan Crenshaw III, PhD, 712 Pit-1 and hypopituitarism without pituitary hypoplasia. Science. 1992;257: Abramson Research Center, 34th and Civic Center Bou- 1118-1121. 19. Radovick S, Nations M, Du Y, Berg LA, Weintraub BD, Wondisford FE. A muta- levard, Philadelphia, PA 19104-4399 (crenshaw@email tion in the POU-homeodomain of Pit-1 responsible for combined pituitary hor- .chop.edu). mone deficiency. Science. 1992;257:1115-1118. Funding/Support: This work was supported by award R01 20. Schonemann MD, Ryan AK, McEvilly RJ, et al. Development and survival of the DC-03917 from the National Institute on Deafness and endocrine and posterior pituitary gland requires the neuronal POU Other Communication Disorders (Dr Crenshaw). Dr So- domain factor Brn-2. Genes Dev. 1995;9:3122-3135. 21. Weiss S, Gottfried I, Mayrose I, et al. The DFNA15 deafness mutation affects POU4F3 bol’s participation in this project was supported by a Pe- stability, localization, and transcriptional activity. Mol Cell Biol. 2003; diatric Fellowship in Otolaryngology at The Children’s 23:7957-7964. Hospital of Philadelphia. 22. Bermingham JR Jr, Scherer SS, O’Connell S, et al. Tst-1/Oct-6/SCIP regulates a Previous Presentation: This study was presented at the unique step in peripheral myelination and is required for normal respiration. Genes annual meeting of the American Society of Pediatric Oto- Dev. 1996;10:1751-1762. 23. Brunner HG, van Bennekom CA, Lambermon EMM, et al. The gene for X-linked laryngology; May 2-3, 2004; Phoenix, Ariz. progressive mixed deafness with perilymphatic gusher during stapes surgery Acknowledgment: We appreciate advice from Kyung Ahn, (DFN3) is linked to PGK. Hum Genet. 1988;80:337-340. BS; John Germiller, MD, PhD; Brad Rittase, BS; Lara Wine- 24. Wallis C, Ballo R, Wallis G, et al. X-linked deafness with stapes fixation in a Mau- Lee, BS; Nancy Solowski, MD; and James G. Davis, PhD. ritanian kindred: linkage to Xq probe pDP34. Genomics. 1988;3:299-301. 25. Alvarez-Bolado G, Rosenfeld MG, Swanson LW. Model of forebrain regionaliza- tion based on spatiotemporal patterns of POU-III , REFERENCES birthdates, and morphological features. J Comp Neurol. 1995;355:237-295. 26. Le Moine C. Young WS. RHS2, a POU domain-containing gene, and its expres- 1. Fritzsch B, Barald KF, Lomax MI. Early embryology of the vertebrate ear. In: Ru- sion in developing and adult rat. Proc Natl Acad SciUSA. 1992;89:3285-3289. bel EW, Papper AN, Fay RR, eds. Development of the Auditory System. New York, 27. Mathis JM, Simmons DM, He X, Swanson LW, Rosenfeld MG. Brain 4: a novel NY: Springer-Verlag; 1998. mammalian POU domain transcription factor exhibiting restricted brain specific 2. Olson NR, Lehman RH. Cerebrospinal fluid otorrhea and congenitally fixed stapes. expression. EMBO J. 1992;11:2551-2561. Laryngoscope. 1968;78:352-360. 28. Heydemann A, Nguyen LC, Crenshaw EB III. Regulatory regions from the Brn4 3. Friedman RA, Bykhovskaya Y, Tu G, et al. Molecular analysis of the POU3F4 gene promoter direct LACZ expression to the developing forebrain and . in patients with clinical and radiographic evidence of X-linked mixed deafness Brain Res Dev Brain Res. 2001;128:83-90. with perilymphatic gusher. Ann Otol Rhinol Laryngol. 1997;106:320-325. 29. Phippard D, Heydemann A, Lechner M, et al. Changes in the subcellular local- 4. Nance WE, Setleff R, McLeod A, et al. X-linked mixed deafness with congenital ization of the Brn4 gene product precede mesenchymal remodeling of the otic fixation of the stapedial footplate and perilymphatic gusher. Birth Defects Orig capsule. Hear Res. 1998;120:77-85. Artic Ser. 1971;07:64-69. 30. Phippard D, Lu L, Lee D, et al. Targeted mutagenesis of the POU-domain gene 5. Cremers CW, Hombergen GCJH, Scaff JJ, et al. X-linked progressive mixed deaf- Brn4/Pou3f4 causes developmental defects in the inner ear. J Neurosci. 1999; ness with perilymphatic gusher during stapes surgery. Arch Otolaryngol. 1985; 19:5980-5989. 111:249-254. 31. Minowa O, Ikeda K, Sugitani Y, et al. Altered cochlear fibrocytes in a mouse model 6. Cremers CW. Audiologic features of the X-linked progressive mixed deafness of DFN3 . Science. 1999;285:1408-1411. syndrome with perilymphatic gusher during stapes surgery. AmJOtol. 1985; 32. Phippard D, Boyd Y, Reed V, et al. The sex-linked fidget mutation abolishes Brn4/ 6:243-246. Pou3f4 gene expression in the embryonic inner ear. Hum Mol Genet. 2000; 7. Glasscock ME. The stapes gusher. Arch Otolaryngol. 1973;98:82-91. 9:79-85.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 131, JAN 2005 WWW.ARCHOTO.COM 45

©2005 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021