Our search for an animal model that spontaneously develops ELH, hearing loss and degeneration postnatally led us to investigate the PhexHyp-Duk mice in the BALB/cUrd and later in the B6 background. Collectively, data from both studies show that (a) PhexHyp-Duk mice typically develop adult onset, asymmetric, progressive hearing loss closely followed by the onset of ELH (as observed by distention of the Reissner’s membrane), (b) functional degeneration precedes structural degeneration, according to ABR and histological data, (c) the major degenerative correlate of hearing loss and ELH in the mutants is the primary loss of spiral ganglion cells, (d) neuronal loss and hair cell loss are downstream events and not the cause of initial hearing loss, (e) PhexHyp-Duk mice develop ELH without evidence of ED obstruction, and occlusion of ED is not a prerequisite for the development of ELH in patients, (f) the expression of the cochlear duct pathology could vary depending on the genetic background and (g) hearing loss could be mildly exacerbated due to susceptibility to otitis media.
gene is one of the genes implicated in the pathogenesis of human X-linked hypophosphatemic (XLH) rickets. Fishman et al. concluded that hearing impairment is not a feature of XLH in childhood; however, 3 of 10 adult patients with XLH show sensorineural hearing loss, suggesting that hearing loss in adults is linked to XLH (Fishman et al., 2004
). Though no inner ear abnormalities other than hearing loss were reported in these patients, future studies involving vestibular testing are needed, as are genetic tests for mutation of the human Phex
Phex mouse mutants lack the ability to mineralize newly formed bone matrix. The non-mineralized bone is spongy and its composition is altered. It is presumed that the thickened otic capsule in PhexHyp-Duk mice represents a defective bone matrix, and whether this is responsible for some of the observed findings or whether Phex has a more direct role in cochlear fluid homeostasis are unclear. Temporal and spatial localization of the Phex gene expression within the cochlear duct may shed some light on this issue.
Lack of observable cochlear duct pathology and significant hearing loss in B6-PhexH-D mouse clearly demonstrates that the bone abnormality alone cannot trigger ELH and associated hearing loss. Rather, other unknown mechanism(s) may be involved in the cochlea, which is suppressed or modified in B6 background. Indeed, mild hearing loss (as measured by elevation in threshold by 10–15 dB SPL) was observed in one ear in some of B6-PhexH-D/Y mice; the contra-lateral ear showed normal thresholds, though thickening of the otic capsule was patent in both ears. Should defective middle ear bones contribute to hearing loss, then both ears in B6-PhexHyp-Duk/Y mice would have registered some degree of hearing loss. Otitis media was observed in 4/5 mutant ears that showed elevated threshold, indicating that defective bone conductance is not a significant issue in the Phex mouse, but otitis media could account for the asymmetric hearing loss observed in Phex mutants.
The onset of hearing loss for most PhexHyp-Duk/Y mice tested occurs between P21-30. Physiological and morphological examination of PhexHyp-Duk/Y mice from birth to late adulthood showed that, in most cases, deterioration of auditory function begins as early as 3–4 weeks after birth with relatively better preservation of hearing in one ear compared to the other. Among all the mice included in this study, none was totally deaf at a young age, which demonstrates that PhexHyp-Duk mice are not born deaf but develop hearing loss postnatally over a period of time. Longitudinal ABR recordings from 8 PhexHyp-Duk/Y mice along with controls clearly show a progressive pattern of hearing deterioration with some element of fluctuation. Results in suggest that mutants tend to acquire low frequency hearing loss earlier than high frequency hearing loss, but more mutants need to be examined to confirm this trend.
Histological analysis clearly shows that cochlear pathology is the dominant factor associated with hearing loss, and it is likely that otitis media minimally contributes to asymmetric hearing loss in PhexHyp-Duk
/Y mice. One interesting observation is the relative preservation of hair cells in the apical turn of the cochlea at a time when most of the spiral ganglion cells had degenerated (, ). This finding has been described in human temporal bone specimens of patients with MD (Nadol, 1990
; Nadol et al., 1987
). This feature is unlike what is observed in other mouse models of sensorineural deafness, where loss of spiral ganglion cells is secondary to hair cell loss (Alagramam et al., 2000
). Further, in PhexHyp-Duk
/Y mice, degeneration of neuronal cells in the cochlea shows an apex to base progression. However, the fact that the first signs of neuronal degeneration are observed some time after P40 clearly shows that functional degeneration is well underway before structural degeneration.
It should be noted that aging mice show a tendency to lose apical spiral ganglion cells before inner hair cells. In these cases, primary neuronal loss is linked to changes in supporting cells and Reissner’s membrane. Though hydrops has not been reported in aging mouse models, it has been proposed that the changes in supporting cells and Reissner’s membrane may be due to disorders of inner ear fluid homeostasis (Keithley et al., 2004
; Ohlemiller and Gagnon, 2004
; Willott and Erway, 1998
). Due to the presence of hydrops in the young PhexHyp-Duk
/Y mutants (P21-90), one might expect a defect in cochlear fluid balance associated with strial pathology. Interestingly, no obvious strial pathology was observed in the PhexHyp-Duk
/Y mutant (), suggesting that the proximate cause of initial hearing loss does not involve the stria. However, appearance of normal stria does not prove that the endocochlear potential (EP) is normal; measurement of EP at various time points would be necessary to assess the status of stria in the ear.
/Y mouse, however, is not the only hydropic mouse noted in recent years. Several others have been described. The Slc26A4 (former PDS gene) knockout mouse develops ELH as early as embryonic day 15, and the mutant mouse is profoundly deaf (Everett et al., 2001
). This mutation is associated with embryonic cochlear malformations not seen in most ELH conditions in patients, including MD, making this model less attractive for use in ELH modeling. Another hydropic mouse that develops ELH in utero is the Foxi1 null mouse. Foxi1 acts as an upstream gene that regulates pendrin expression, hence the null mouse exhibits phenotypic features that are similar to the Scl26A4 knockout mouse (Hulander et al., 2003
). Finally, a recently described Brn-4-deficient mouse also develops ELH in utero (Xia et al., 2002
). The Brn-4 gene belongs to transcription factors of the Pou3f4 family that regulate the transcription of connexin 26, connexin 31, Na/K ATPase and Na–K–Cl cotransporter (Phippard et al., 1999
). This prenatal development of ELH deviates from postnatal development of ELH observed in patients, making this model less attractive.
The focus of this report is not whether Phex is a candidate gene for MD or other conditions leading to ELH. Rather, the point is that PhexHyp-Duk mice are valuable models to use in the study of the pathophysiology linked to ELH since the ear phenotype observed in this mutant is reminiscent of the auditory dysfunction characteristic of patients with ELH condition. The potential advantages of the Phex model in the study of ELH are: (a) compared to the guinea pig model, surgery is not required, making it amenable to use by more researchers, (b) the phenotype is reproducibly seen in mutant males in the BALB/curd background (there is a degree of uncertainty with the surgical model and a small percentage of animals die after surgery), (c) identification of genetic factors associated with susceptibility to ELH condition would be easier with a mouse model than with other animal models and (d) since the development of hearing loss and severe signs of neuronal degeneration occur post weaning, this model offers a new vehicle to test putative otoprotective agents.
There are some limitations to the Phex
mouse model. First, episodic vertigo is an important clinical feature in patients with ELH condition. It may be difficult to document episodic vertigo in this model. Whether the observed vestibular dysfunction is episodic or not will require more systematic, long-term studies including quantitative assessment of the vestibular evoked potential (Alagramam et al., 2005
). Second, the extent of Reissner’s membrane distention is presently used as a measure of the severity of ELH. Although this is a traditional approach, it may not accurately reflect the onset and/or the severity of the disorder in the endolymph fluid homeostatis. Measurement of the EP along with the physical marker (membrane distention) may be a better way to assess severity of ELH in the Phex
mutants. Third, mice have a 2-year life span; patients identified with ELH condition are typically 30-years or older. It is hard to equate late onset hearing loss linked to ELH condition in patients to an animal model that lives for about 2-years. It should be noted that these limitations are shared by other animal models as well. Fourth, susceptibility of the Phex
mutant to otitis media appears to increase the threshold by 10–15 dB in the affected ear. This, however, is not a significant problem for three reasons. Data from the B6 mice clearly separates the level of hearing loss associated with OM compared to the more severe hearing loss observed with PhexHyp-Duk
mice in BALB/cUrd background. The hearing loss caused by cochlear pathology is much more severe than 10–15 dB loss observed in B6-PhexHyp-Duk
mice. ABR data from mutant ears without OM can be used to correlate hearing loss with ELH and degeneration in this model.
In summary, we have established the profile of the inner ear phenotype in the PhexHyp-Duk mutant at various time points from birth to adulthood (), and we compare the phenotype in two genetic backgrounds. The spontaneous development of ELH, hearing loss and sensory cell loss are attractive features of the PhexHyp-Duk mouse model as they seem to mirror many of the pathologic features associated with ELH-related disorders both clinically and experimentally. The PhexHyp-Duk mouse is a potentially useful model to explore the relationship among ELH, hearing loss and neuronal degeneration in the cochlea.
Fig. 16 Approximate timeline of events in the inner ear of PhexHyp-Duk/Y. Dotted line indicates onset of phenotype, mild hearing loss (HL), signs of endolymphatic hydrops (ELH), spiral ganglion cell (SGC) degeneration, or hair cell degeneration, with some variability (more ...)