The Novel Mouse Mutation Oblivion Inactivates the PMCA2 Pump and Causes Progressive Hearing Loss

Progressive hearing loss is common in the human population, but we have few clues to the molecular basis. Mouse mutants with progressive hearing loss offer valuable insights, and ENU (N-ethyl-N-nitrosourea) mutagenesis is a useful way of generating models. We have characterised a new ENU-induced mouse mutant, Oblivion (allele symbol Obl), showing semi-dominant inheritance of hearing impairment. Obl/+ mutants showed increasing hearing impairment from post-natal day (P)20 to P90, and loss of auditory function was followed by a corresponding base to apex progression of hair cell degeneration. Obl/Obl mutants were small, showed severe vestibular dysfunction by 2 weeks of age, and were completely deaf from birth; sensory hair cells were completely degenerate in the basal turn of the cochlea, although hair cells appeared normal in the apex. We mapped the mutation to Chromosome 6. Mutation analysis of Atp2b2 showed a missense mutation (2630C→T) in exon 15, causing a serine to phenylalanine substitution (S877F) in transmembrane domain 6 of the PMCA2 pump, the resident Ca2+ pump of hair cell stereocilia. Transmembrane domain mutations in these pumps generally are believed to be incompatible with normal targeting of the protein to the plasma membrane. However, analyses of hair cells in cultured utricular maculae of Obl/Obl mice and of the mutant Obl pump in model cells showed that the protein was correctly targeted to the plasma membrane. Biochemical and biophysical characterisation showed that the pump had lost a significant portion of its non-stimulated Ca2+ exporting ability. These findings can explain the progressive loss of auditory function, and indicate the limits in our ability to predict mechanism from sequence alone.


Introduction
PMCA2 is one of four isoforms of the plasma membrane Ca 2+ pumps of mammalian cells [1,2]. The expression of PMCA2 and PMCA3 is largely restricted to brain and muscle, whereas PMCA1 and 4 are ubiquitously expressed. PMCA2 and PMCA3 are more active in exporting Ca 2+ than the ubiquitous isoforms [3], probably due to their higher affinity for the activator calmodulin. PMCA2, however, is peculiar in its very high activity even in the absence of calmodulin [4,5]. In the ear, PMCA2 is expressed at high levels in outer hair cell stereocilia and apical membranes and at moderate levels in inner hair cell stereocilia and in the spiral ganglion [6][7][8][9][10]. It actively extrudes Ca 2+ that has entered the hair cell during mechanoelectrical transduction [11]. This maintains the low intracellular levels of Ca 2+ and may create a relatively higher concentration of Ca 2+ in the endolymph surrounding the stereocilia, contributing to the maintenance of the electrochemical gradient needed for transduction to occur [11]. Bulk concentration of Ca 2+ in mammalian cochlear endolymph is estimated at ,20 mM [12]. PMCA2 is also important in maintaining sufficient extracellular Ca 2+ in the vestibular system for the formation of the otoconia, the calcium carbonate crystals needed for sensing gravity and acceleration [13].
Transcripts for PMCA2 undergo alternative splicing at two sites. Site A is closer to the N-terminus and site C closer to the Cterminus. In the PMCA2 variant expressed in stereocilia [14,15], the splicing introduces three exons at site A, generating variant w, and two alternative exons at site C, generating variant a. The Csite insert leads to a truncated pump that contains only about half of the original calmodulin binding domain [4,5]. The doubly inserted w/a variant, seen in stereocilia, has an unusually limited ability to increase activity rapidly when challenged with a Ca 2+ pulse, but has about the same high non-stimulated activity as the full-length z/b variant [16].
The Oblivion (Obl) mutant was identified as a new mouse mutant with progressive hearing loss from a large scale ENU mutagenesis screen [17]. The aim of this screen is to provide new models for deafness, especially progressive deafness which is common in the human population, and to identify the genes and underlying pathology in these new mutants. Here we report that the progressive hearing loss in Obl is due to a missense mutation in the gene Atp2b2, encoding PMCA2. We describe the hearing impairment and hair cell pathology in the mutants, the dysfunction of Ca 2+ export by the mutated PMCA2 pump cloned and overexpressed in model cells and in cultures of utricles from the mutant mice.

Progressive Hearing Loss in Oblivion Mutants
Obl/+ heterozygotes have a normal Preyer reflex at one month old, but by two months only 58% offspring from Obl/+6+/+ matings showed a Preyer reflex (Table S1), suggesting progressive hearing loss in Obl/+ mice. No vestibular defect, indicated by head-tossing or circling behaviour, was seen in these heterozygotes, although no detailed analysis of vestibular function was performed.
To measure auditory thresholds, auditory brainstem responses (ABR), a reflection of cochlear and brainstem neural activity, were recorded in P20, P59-62 and P89-91 mice on their original C3HeB/FeJ genetic background ( Figure 1). ABRs of wild-type mice showed an improvement in thresholds below 12 kHz from P20 to P59-62, perhaps indicative of maturation of the auditory system. From P20 to P89-91, wild type mice showed mild and progressive elevations of thresholds above 12 kHz.
Obl/+ mice demonstrated a severe and age-related progressive hearing loss. Obl/+ mice showed significantly raised thresholds at all frequencies, compared to age-matched wild-type controls (ttest, p,0.05), of up to 60-70 dB or more. In contrast to the Preyer reflex tests, even at P20 Obl/+ showed large threshold elevations. At P59-62, the heterozygotes showed further threshold elevations which were most severe at higher frequencies, above 18 kHz. By P89-91, high frequency losses were compounded by severe losses across the entire range measured. This indicated a progressive hearing loss in Obl/+ mice. The improvement of low frequency (3-6 kHz) ABR thresholds between P20 and P59-62 may indicate maturation of the developing auditory system between these ages.
Obl/Obl homozygous mutants show a very severe hearing and vestibular phenotype and are significantly smaller (10.5 g, SD 2.02) than age and sex matched Obl/+ littermates (17.3 g, SD 2.11; t-test, p,0.05, 31-34 days old). They (a) fail to develop a Preyer reflex; (b) fall from side to side whilst walking; (c) are unable to right themselves; and (d) curl towards their belly when lifted by their tail and do not display a reaching response when lowered towards a surface. Homozygotes also show hind limb stiffness and appear ataxic, which are not general features of vestibular dysfunction.

Structure of the Inner Ear
The gross morphology of the middle ear ossicles and inner ear appeared normal in Oblivion mutants, both heterozygotes and homozygotes. Scanning electron microscopy in Obl/+ mutants at 3-4 months of age showed degeneration of hair cells, with the basal turn more severely affected than the apex, and outer hair cells (OHCs) more affected than inner hair cells (IHCs), a pattern that is commonly reported in damaged cochleas (Figures 2 and 3). . Auditory brainstem response thresholds in Oblivion heterozygous and wildtype animals. Obl/+ mutants show significantly raised auditory brainstem response (ABR) thresholds compared to wildtype mice at P20 (circles), P59-62 (squares), and P89-91 (triangles). Mean ABR thresholds with standard error bars are given for +/+ mice (filled symbols) and Obl/+ mice (open symbols). The dotted line without symbols indicates the maximum dB SPL output for the sound system at each frequency. The position along the length of the cochlear duct that best responds to each frequency (represented by % of total distance from the base) is indicated by the frequency place map at the top of the graph. Mean click ABR thresholds are plotted at an arbitrary point on the frequency axis; the position does not relate to the frequency content of the click stimulus. doi:10.1371/journal.pgen.1000238.g001

Author Summary
Progressive hearing loss is very common in the human population, but we know little about the causes. Environmental and genetic factors each may contribute. Knowledge of the genetic variants involved in hearing loss and understanding of the molecular and cellular mechanism of their action will aid the development of better treatments. One of the few genes known to be involved, in both mouse and humans, is Atp2b2, which encodes a calcium pump. We have discovered a new mutation in this gene leading to hearing loss in the mouse mutant oblivion. The mutation leads to a serine to phenylalanine substitution in a transmembrane domain. Mutations affecting such transmembrane domains are usually expected to interfere with the normal process of inserting the protein in the membrane and transporting it to its final destination on the plasma membrane of the cell. Surprisingly, the pump is produced and is targeted to the plasma membrane, in both cultured cells expressing the mutant gene and sensory hair cells from the oblivion mutant inner ear. However, we show it has impaired calcium pumping ability, which can account for the progressive hearing loss as well as the progressive degeneration of the sensory hair cells that we observe in the mutants.
Obl/Obl homozygotes were more severely affected than heterozygotes. However, there were many remaining hair cells with relatively normal appearance in the mutants, including a Wshaped arrangement of stereocilia, especially in the apical turn. Stereocilia fusion was seen in some, an early indicator of hair cell degeneration. At P20, no significant hair cell loss was detected in Obl/+ mutants compared to their littermate controls ( Figure 3A and 3B), despite the fact that we saw significantly raised ABR thresholds in another cohort of P20 heterozygotes ( Figure 1). Hair cell counts from the basal and middle turns at P75 showed no significant OHC degeneration in the middle turn and no significant IHC loss throughout the cochlea in Obl/+ heterozygotes ( Figure 3C and 3D). By P121, there was significant OHC and IHC loss in basal and middle turns in Obl/+ ( Figure 3E and 3F). This suggests that the hair cell loss seen in these mutants is a secondary consequence of the hair cell not functioning correctly, rather than being the primary cause of raised thresholds in Obl/+ mutants.
In Obl/Obl mutants at P30 there was highly variable hair cell degeneration, both within and between animals. In some regions there was scattered hair cell loss with a pattern similar to that seen in heterozygotes ( Figure 2G and 2H), while in some regions towards the base there was complete degeneration of the organ of Corti with a complete absence of specialised cells, including supporting cells such as pillar cells ( Figure 2I).

Mapping and Identification of the Oblivion Mutation
Obl/+ mutants on a C3HeB/FeJ background were outcrossed to C57BL/6J and mutant F1 progeny were backcrossed to the original C3HeB/FeJ strain. Backcross litters were assessed for absence of a Preyer reflex and a genome-wide scan was performed on their DNA using 60 polymorphic microsatellite markers. We identified a region of linkage on chromosome 6 ( Figure 4A) between markers D6Mit104 and D6Mit218, corresponding to a physical distance of 16 Mb. This region contained a good candidate gene: Atp2b2. Genomic DNA was used to sequence the 19 coding exons of the gene, including the splice sites. We identified a C/T heterozygous peak in Obl/+ mutants, suggesting a CRT transition (2630CRT) in exon 15 of Atp2b2, predicted to change a serine to a phenylalanine (S877F; Figure 4) in the mutant allele. This change was also confirmed by a restriction test assay that was used to genotype the colony. This assay was used to screen 19 inbred strains for the Obl mutation and none were found to have it, suggesting that it is not a common polymorphism. We found non-complementation of Obl with the deafwaddler mutant allele, Atp2b2 dfw , confirming that the missense mutation we found was the pathogenic mutation (see Text S1).

Functional Analysis of the Mutated PMCA2 Pump Overexpressed in Model Cells
To investigate the effects of the serine to phenylalanine change on the functionality of the pump, mammalian expression plasmids for the S877F and the wild type variant of the PMCA2 pump were prepared and expressed in CHO cells. Appropriate controls (Western blotting and quantitative immunocytochemistry) established that the two pump variants were expressed at about the same levels, and were correctly delivered to the plasma membrane ( Figure 5). CHO cells were transfected with the Ca 2+ sensitive photoprotein aequorin (cytAEQ, [3]) and stimulated with ATP, an agonist of purinergic P2Y receptors that produces InsP 3 generating a cytosolic Ca 2+ transient. Under the experimental conditions, the height of the Ca 2+ peak, and the kinetics of the return of the Ca 2+ transient to baseline were controlled primarily by the PMCA pump: the much larger amounts of the overexpressed PMCA2 pump overshadowed the endoplasmic reticulum Ca 2+ pump (SERCA) (see [16]) and the contribution of plasma membrane Ca 2+ influx channels opened by emptying of intracellular stores to the shaping of the Ca 2+ trace was disregarded, as their effect would be the same in the wild type and Obl measurements. The overexpressed Obl pump did not further depress the limited ability of the wild type w/a pump to control the height of the Ca 2+ peak ( Figure 6). The mutation, however, severely affected the resting activity of the pump that drove the return of the Ca 2+ trace to baseline after the peak. The half time of the declining phase was 64.1563.02 sec (n = 6) in control, 6.5560.72 sec (n = 9) in the wild type w/a variant and 45.5065.97 sec (n = 4), p,0.001, in Obl.

Ca 2+ Dynamics in the Hair Cell Stereocilia of Wild Type, Obl/+, and Obl/Obl Mice
To characterize Ca 2+ dynamics in the stereocilia of hair cells, cultures of immature utricular maculae were obtained from wild type and mutant mice (see Materials and Methods). Ca 2+dependent changes in fluorescence evoked by the photorelease of intracellular caged Ca 2+ (4 ns single UV pulse) were monitored with a temporal resolution of 6 ms using confocal laser scanning microscopy. Immunofluorescence labelling with isoform specific antibodies showed that PMCA2 was correctly located in the stereocilia of homozygous mutant organotypic cultures of utricular maculae ( Figure 7A) and organ of Corti ( Figure 8). Figure 7B shows a macular hair cell loaded with Fluo-4. Fluorescence changes in the stereocilia were monitored repeatedly by a line-scan positioned along the hair bundle and extending into the cell soma (dashed line). Time-dependent post UV pulse changes in fluorescence at different parts of the line scan are illustrated in Figure 7C. The time course of the stimulus-evoked changes in fluorescence (DF), normalized to basal (pre-stimulus) fluorescence (F 0 ), are compared for wild type (wt, blue trace) heterozygous Oblivion mice (Obl/+, green trace) and homozygous Oblivion mice (Obl/Obl, black trace) in Figure 7D. The time courses matched well at the peak, although the [Ca 2+ ] i transient decayed more slowly for the Obl/Obl mice. To highlight the differences, the traces are re-plotted in Figure 7E on an expanded time scale (solid lines), together with their respective confidence intervals (dashdotted lines). A single exponential fit to the first 10 s of these transients yielded significantly longer decay time constants for the Obl/+ and Obl/Obl cultures: t wt = 2.860.4 s for wt, t Obl/+ = 4.261 s for Obl/+ mice (p = 0.02, if compared to wt using the ANOVA test) and t Obl/Obl = 6.961.3 s for Obl/Obl mice (p,0.01). No significant divergence in the early phase (first 2 s) of the decay was found for Obl/+ mice relative to wt, consistent with the lack of evident phenotypic vestibular defects in these mice. At later times, the decay tended to diverge possibly due to the contribution of complex processes such as Ca 2+ -induced Ca 2+ release. These contributions were not investigated further.

Discussion
Progressive, age-related hearing loss affects 60% of humans over the age of 70 [18]. The condition is a multifactorial disorder to which genetic variation, disease and environmental influences such as acoustic trauma are all contributing factors, making identification of the genes involved in humans difficult. The mouse is an ideal resource to study the genetics of progressive hearing loss due to the possibility of controlling both genetic background and the environment.
The present study has identified a new ENU-induced allele of Atp2b2, Atp2b2 Obl in the Obl mouse mutant. This allele contains a C to T missense mutation in exon 15 of Atp2b2, causing a nonconservative amino acid substitution of serine by phenylalanine in transmembrane domain 6 of the PMCA2 pump [1,19]. The serine at position 877 is highly conserved in both human and mouse and also between other members of the PMCA family ( Figures 4C and  S2). The S877F mutation is of special interest on at least two accounts. One is the finding that substitutions in the transmembrane residues of the PMCA pump frequently impair its correct plasma membrane targeting [20]. However, the mutant Obl pump was correctly delivered to the plasma membrane in both the model cells and in native stereocilia. In another Atp2b2 mutant allele, Wriggle mouse Sagami (wri), a missense mutation in transmembrane domain 4 completely abolished the expression of PMCA2 protein in stereocilia of cochlear hair cells [21]. The second reason for interest stems from studies of the SERCA pump of endoplasmic reticulum, considered a model for all P-type pumps. Conserved residues within transmembrane domain 6 of SERCA have been shown to be components of the channel through which Ca 2+ is translocated [22] and are also present in the PMCA pump. It is pertinent to quote at this point an earlier study on some point mutations in transmembrane domain 6 of another isoform of the pump, PMCA4 [23]. In that study, one of the mutants (S877A) appears to have higher activity than the wild type pump. The reason for the different effect of the S877A mutation of PMCA4 and the S877F mutation of PMCA2 that we report here is not clear. However, the PMCA4 study is difficult to compare with the present one, since it analyzed the pump in crude microsomes of overexpressing COS cells, and offered no information on the localization of the expressed pump in the plasma membrane.
The Obl mutation is the seventh mutation in mouse Atp2b2 to be reported. In the deafwaddler (dfw) mouse a missense mutation in Atp2b2 leads to a partial loss of function [8] and PMCA2 retains 30% of its Ca 2+ pumping activity [24]. A spontaneous T692K mutation in Atp2b2 led to clear ataxic behaviour, with normal mRNA levels, in a second unnamed mutant [25]. A further four mutations of mouse Atp2b2 have been described: deafwaddler 2J (dfw 2J ), deafwaddler 3J (dfw 3J ), wriggle mouse sagami (wri) and a targeted null mutation. Analysis of mRNA transcripts and protein levels suggests that these latter four mutations are all null alleles [13,21,26]. Homozygotes for these alleles show severe ataxia by 10 days of age as well as profound deafness. The profound deafness and ataxia detected in Obl homozygotes is similar to the phenotype of these other known Atp2b2 null mutants [13,26,27]. Noncomplementation between Obl and dfw confirmed that these two mouse mutants are likely to be allelic. The mutations identified in the Atp2b2 gene are shown in Figure S1.
Both the Obl and the dfw mutations significantly reduce the nonactivated Ca 2+ exporting ability of the PMCA2 protein. Observations on organotypic cultures showed that the defect of the pump observed in model cells also reduced its ability to remove the Ca 2+ produced by UV photolysis in stereocilia. Thus, the Obl mutant has lost a significant portion of the non stimulated, longerterm Ca 2+ exporting ability in respect to the w/a PMCA2 pump.
In the present work we show that hearing loss in Obl/+ mice is detected at P20 and progresses in its severity with increasing age. Threshold shifts of up to 60-70 dB were found at frequencies corresponding to the basal and middle turns of the cochlea, where the majority of outer hair cell loss was detected. Analysis of Obl/+ mice at P20 showed no OHC or IHC loss despite the fact that they had a significant hearing impairment, confirming the previous suggestion that although hair cells are present they are not functioning normally [28].
IHC loss was detected in the base and middle of the cochlea in Obl/+ mutants at 4 months of age. The early degeneration of the OHCs seen in heterozygous Obl (and dfw) mice, leading to complete degeneration of the organ of Corti in the base of the cochlea, is similar to that seen in some human patients with agerelated hearing loss [29,30]. It is not clear why hair cell dysfunction leads to hair cell degeneration in these mouse mutants, but prolonged abnormal calcium homeostasis may contribute to hair cell death. In dfw mice at approximately P60, distortion product otoacoustic emission studies show that PMCA2 is important for the correct functioning of outer hair cells, especially at high frequencies [31]. Mice expressing the Atp2b2 dfw2J allele, [9] demonstrated a lowered endocochlear potential and reduced endolymphatic calcium concentration, and thus have a reduced capacity for mechanoelectrical transduction. Taken together, these features may explain why ABR thresholds in Obl/+ mice are elevated above controls, but are still recordable. An interesting observation from the ABRs recorded in P59-62 and P89-91 Obl/+ mice is that click thresholds are more sensitive than the best tone threshold, by a factor of 15-22 dB. Clicks produce a more synchronised activation of a wider region of the basilar membrane compared to tone pips, and this may produce a summation of activity, reflected in lower click-evoked ABR thresholds.
When Obl was placed on a mixed C3HeJ/FeB and C57BL/6J background, the progressive hearing loss in heterozygotes seemed more severe (data not shown). Onset and severity of progressive hearing loss seen in heterozygotes of other Atp2b2 mutant alleles have been shown to vary considerably depending on the genetic background on which the mutation arose [13,21,26,28,32]. This is due to the presence of modifier alleles, one of which has been identified as the G753A variant of Cdh23 at the ahl locus (also known as modifier of deafwaddler, mdfw) [32,33]. Interactions between heterozygous Atp2b2 mutations and Cdh23 mutations have been shown to worsen the progressive hearing loss seen in some human patients. Heterozygous mutations (ATP2B2 V586M ) increase the severity of the progressive hearing loss seen in human patients with mutations in CDH23 [34]. The ATP2B2 V586M mutation reduced the level of PMCA2 protein produced by 50%, although on its own it was not sufficient to cause hearing impairment in humans [34]. More recently, hearing loss has been reported in a human patient with an ATP2B2 G293S mutation and a CDH23 T1999S mutation. The parents of the patient carried either the ATP2B2 or the CDH23 mutation, but had no hearing impairment [16].
The ATP2B2 gene has recently been implicated in the deafness in 3p-syndrome. The syndrome is characterised by developmental delay, growth retardation and craniofacial abnormalities (see [35]), which is sometimes, but not always, associated with a severe sensorineural hearing loss. In 3p-syndrome patients with a hearing loss, a deletion in the 3p25.3 locus was mapped to a region containing 18 genes including ATP2B2. It is likely that haploinsufficiency of ATP2B2 is responsible for the deafness associated with this syndrome [36].
As progressive hearing loss is so common in the human population and we know so little about its molecular basis, identifying other mutations in and modifiers of the Atp2b2 locus in mouse inbred strains may be of importance in identifying new loci involved in progressive and age-related hearing loss in humans.

Mice
The founder mouse carrying the Oblivion mutation (gene symbol Obl, original identifier DEA14) arose from the ENU mutagenesis program at Munich [17]. Mutations were generated by injecting 3 doses of 80-90 mg/kg bodyweight of N-ethyl-Nnitrosourea (ENU) into C3HeB/FeJ males. F1 progeny were screened for a range of phenotypes, including deafness and balance disorders. The founder Obl/+ mutant was identified by the absence of a Preyer reflex at 3 months. The mutant colony was maintained on the C3HeB/FeJ background, and all ABR and

Inner Ear Clearing and Ossicle Dissection
Half heads were fixed in Bodian's fixative and cleared with glycerol using a standard protocol. The inner ear was examined for signs of malformation. Middle ear ossicles were dissected out and studied. Six Obl/+ mutants and six littermate controls between 3-4 months age were analysed.

Scanning Electron Microscopy
Inner ears were fixed in 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer, the organ of Corti was exposed and samples were processed using the osmium tetroxide-thiocarbohydrazide (OTOTO) method [37]. After critical point drying and sputter coating with gold, samples were examined using a Phillips XL30 scanning electron microscope at 10 kV or a Hitachi S4800 FE Scanning Electron Microscope at 5 kV. Initial characterisation was performed on 3 Obl/+ and 3 littermate controls at 3-4 months of age. Hair cell degeneration was assessed in the basal turn (20-30% of the total distance from the base of the cochlear duct) and middle turn (55-65% from the base) at 20 (P20), 72-75 and 121 days old. Between 3 and 7 Obl/+ and +/+ mice were examined at each age. Hair cells with intact stereocilia bundles were counted over a stretch of at least 200-300 mm of the cochlear duct. Hair bundles that were damaged or showed fusion were still counted as being present. A two tailed T-Test was performed on hair cell counts for each hair cell row in each region at 95% confidence intervals, with the number of cases and standard deviations included in the analysis. A two tailed T-test was carried out on the weights of Obl/Obl mutants compared to littermate (Obl/+) controls in the same way.

Preyer Reflex and Auditory Brainstem Response (ABR)
The Preyer reflex (pinna flick) was detected using a custom-built click box to deliver a calibrated 20 kHz sound burst at 90 dB SPL. Up to 88 mice from Obl/+6+/+ matings were tested weekly from 3 to 8 weeks of age, although the numbers of mice at each timepoint varied. For ABR recordings, a separate cohort of mice were anaesthetised (urethane 2 mg/g) and subcutaneous needle electrodes inserted on the vertex (active), and over the left (reference) and right (ground) bullae. A calibrated sound system was used to deliver free-field click (0.01 ms duration) and tone pip (various frequencies from 3-42 kHz of 5 ms duration, 1 ms rise/fall time) stimuli at a range of intensity levels in 3 dB (or multiple) steps. Averaged responses to 512 stimuli, presented at 21.1 s 21 , were analysed and thresholds established as the lowest sound intensity giving a visually-detectable ABR response. ABR recordings were obtained from a total of 61 mice, 28 at P20 (+/+, n = 9; Obl/+, n = 19), 14 at P59-62 (+/+, n = 5; Obl/+, n = 9) and 19 at P89-91 (+/+, n = 5; Obl/+, n = 14).

Mapping the Obl Mutation
Obl/+ mutants on a C3HeB/FeJ background were outcrossed to C57BL/6J wild type females. Obl/+ F1 progeny were then backcrossed to +/+ animals from the original C3HeB/FeJ strain. Offspring from these backcross matings were examined at 2 months of age or older (as this is the age at which Obl/+ mice on the original genetic background show profound hearing loss) using the Preyer reflex. Tail and pinna tissue were collected for DNA preparation. A total of 255 backcross mice were analysed (129 Obl/+, 126 +/+). A genome-wide scan was conducted with 60 microsatellite markers approximately 25 cM apart, that had been shown to be polymorphic between C3HeB/FeJ and C57BL/6J inbred strains (Table S2). Additional microsatellite markers used for fine mapping of the Obl mutation were: D6Mit36; D6Mit104; D6Mit150; D6Mit115; D6Mit218; D6Mit254. PCR was performed using standard techniques.

Mutation Screening and Confirmation
Sequence analysis was performed on genomic DNA in Obl/+ mutants and littermate controls using primers designed to amplify the coding exons and splice sites of the Atp2b2 gene. The primer sequences are listed in Table S3. PCR was performed using standard techniques and products were cleaned using magnetic bead separation (Ampure) and sequenced using BigDye Terminator Cycle sequencing kit (Applied Biosystems). Sequence traces were analysed using Gap4 software [38]. To confirm the mutation identified in exon 15 and for genotyping of the colony, a PCRbased genotype test was designed. The 2630CRT missense mutation did not change a restriction enzyme recognition site, so primers were designed flanking the mutation site that would incorporate a StyI recognition site in the wildtype (C) allele, but not in the mutant (T) allele. Primers OblRTF (59-CTT CTT CTC CCT GCC ACT GTC GTA G) and OblRTR (59-CCA CCG AGA CAC CGG TCC CGG TTC) were used for PCR. The 111 bp PCR product was digested with StyI (New England Biolabs) which cuts the wildtype allele giving an 89 bp fragment while the mutant allele remains uncut. This genotyping tool was used to establish whether the sequence change in Obl DNA was a polymorphism in a total of 17 inbred strains: Cloning and Mutagenesis of pmRFP-PMCA2 w/a mRFP was amplified from pCDNA3.1/zeo-mRFP (kindly donated by Dr. M. Zaccolo, Padua, Italy) using the following primers, forward: 59-GCGCTAGCATGGCCTCCTCCGAG-GACGTCA-39 and reverse: 59-GCAGATCTGAGGCGCC-GGTGGAGTGGCGG-39, bearing restriction sites for NheI and BglII, respectively (in bold). The PCR product was then digested with NheI and BglII and inserted in pEGFP-c1 (Clontech, Palo Alto, CA) digested with NheI and BglII to create pmRFP-c1. PMCA2 w/a in pMM2 (kindly provided by Dr. Strehler, Rochester, MN) was excised by independent digestion with SalI-EcoRI and EcoRI-KpnI and inserted into XhoI-KpnI sites of pmRFP-c1 in a three-part ligation reaction resulting in pmRFP-PMCA2wa. The construct was controlled by sequencing. Site-directed mutagenesis was carried out to obtain the mutant cloned in the appropriate vector. pmRFP-PMCA2w/a was used as target and experiments were performed according to the manufacturer's standard protocol (Stratagene, Cedar Creek, TX) The following primers were used: Obl 59 CATCATGGACACATTTGCTTTCCTGGCCCTGGCAACA-GAGC 39 (forward) and 59 GCTCTGTTGCCAGGGCCAG-GAAAGCAAATGTGTCCATGATG 39 (reverse)

Immunolocalization of the Expressed Pumps and Membrane Fluorescence Computation
CHO cells were grown in Ham's F12 medium, supplemented with 10% fetal calf serum (FCS). Before transfection, they were seeded onto 13 mm glass coverslips and allowed to grow to 50% confluence. Transfection with 3 mg of plasmid DNA (or 1.5 :1.5 mg in the case of co-transfection) was carried out with a Caphosphate procedure [39]. Immunocytochemistry quantified the expressed pump proteins in the plasma membrane of transfected cells. CHO cells expressing the PMCA2 variants, were stained with polyclonal isoform-specific PMCA antibody 2N (Affinity Bioreagent, Inc., Golden, CO) or a monoclonal antibody recognizing all pump isoforms (5F10, Affinity Bioreagent, Inc., Golden, CO), at a 1:100 dilution in PBS. Staining was carried out with Alexa 488 labelled anti-rabbit or anti-mouse secondary antibodies (Molecular Probes, Invitrogen Corp., Carlsbad, CA) at a 1:50 dilution in PBS. Cells were imaged on a spinning disk confocal microscope (Ultraview; Perkin-Elmer) by using a X60 oilimmersion objective at a N.A. (PlanAPo; Nikon, Tokyo, Japan). Regions of interest were selected by applying an edge-finding (Sobel) digital filter, thus limiting the analysis to plasma membrane areas. The total fluorescence intensity in membrane-delimiting regions of interest was quantified with software developed in our laboratory. For each construct fluorescence was averaged over a total of 50 cells in 3 different slides.

Ca 2+ Measurements with Recombinant Aequorin
Transfected cytAEQ were reconstituted by incubating CHO cells for 1-3 h with 5 mM coelenterazine in Dulbecco's modified Eagle's medium (D-MEM) supplemented with 1% FCS, at 37uC in a 5% CO 2 atmosphere. Additions to the KRB medium (1 mM CaCl 2 , 100 mM ATP) were made as specified in the figure legends. The experiments and luminescence calibration into [Ca 2+ ] values were carried out according to [40]. The experiments were terminated by lysing the cells with 100 mM digitonin in a hypotonic Ca 2+ -rich solution (10 mM CaCl 2 in H 2 O) to discharge the remaining aequorin pool. Briefly, a 13-mm round coverslip with the transfected cells was placed in a perfused thermostated chamber in close proximity to a low-noise photomultiplier, with a built-in amplifier discriminator. The output of the discriminator was captured by a Thorn-EMI photon-counting board and stored in an IBM-compatible computer for further analyses. Luminescence was calibrated off-line into [Ca 2+ ] values by using a computer algorithm based on the Ca 2+ response curve of wt aequorin. Data are reported as mean6SD. Statistical differences were evaluated by Student's 2-tailed t-test for unpaired samples. A p value,0.01 was considered statistically significant.

Preparation of Organotypic Utricle Cultures
To access utricular maculae of wild type or mutant mice between postnatal day 3 (P3) and P4, the otic capsule was opened medially and the endolymphatic compartment of the macula cut open. The otolithic membrane was removed after 15 min incubation in dissection saline to which 0.1 g/l bacterial subtilisin (type XXIV; Sigma-Aldrich, St. Louis, MO) had been added. Dissection saline was composed of Hank's Balanced Salt Solution (HBSS; part number H6648, Sigma-Aldrich) with 10 mM HEPES, 10.000 U/l penicillin and 25 mg/l fungizone. HBSS contained (in g/l): 0.4 KCl, 0.06 KH 2 PO 4 (anhydrous), 0.35 NaHCO 3 , 8.0 NaCl, 0.048 Na 2 HPO 4 (anhydrous), 1 D-glucose. The epithelium was fixed by Cell-Tak (BD Biosciences, Bedford, MA), mixed with 90% NaHCO 3 , to the lateral side of a glass capillary (1.5 mm diameter, 5 mm length), which had been previously glued to a microscope slide by a small drop of Sylgard Silicon Elastomer (Dow Corning, Wiesbaden, Germany). Cultures were preserved for one day at 37uC in a complete medium of 95% D-MEM/Ham's F-12 (1:1) (concentration 1X, liquid form, containing L-glutamine but no HEPES; Gibco, Invitrogen Corp., Carlsbad, CA) and 5% fetal bovine serum.

Immunolocalization of PMCA2 in the Stereocilia of Hair Cells in Utricle Cultures
Organotypic cultures dissected from P3 wild type and mutant mice pups were maintained over night at 37uC in D-MEM/Ham's F-12 (1:1) medium with 5% fetal bovine serum. Tissue preparations were fixed in paraformaldehyde 4% for 20 min at room temperature, rinsed in washing solution (PBS containing BSA 2%) and permeabilised with washing solution containing Triton 0.1% for 1 h at room temperature. Incubation with primary PMCA antibody 2N was carried out overnight at 4uC using a 1:100 dilution in washing solution. FITC-conjugated rabbit anti-IgG antibody (Invitrogen) was used as secondary antibody for pump detection (1:200 dilution in washing solution, 2 h at room temperature). The preparation was mounted on a coverslip and imaged up side down on an inverted spinning disk confocal microscope (Ultraview; Perkin-Elmner) using a 606 oil-immersion objective at a 1.4 N.A. (PlanApo; Nikon, Tokyo, Japan).

Whole Cochlea Immunohistochemistry
Cochleae dissected from P5 mice were fixed in 4% paraformaldehyde for 20 min at room temperature, rinsed in PBS containing 2% BSA (rinse solution) and permeabilized for 1 hour at room temperature with 0.1% Triton, dissolved in rinse solution. Tissues were stained with polyclonal isoform-specific PMCA antibody 2N (Affinity Bioreagent, Inc., Golden, CO) by incubation overnight at 4uC with specific polyclonal antibodies (2.5 mg/ml) (Invitrogen) diluted in rinse solution. A FITC conjugated rabbit anti-IgG antibody (5 mg/ml, Invitrogen) was used as secondary antibody, following incubation in rinse solution for 2 hours at room temperature. F-Actin was stained by incubation for 1 hour at room temperature with rhodamine phalloidin (7 mM, Invitrogen, R415) diluted in rinse solution. Stained samples were sandwiched between two coverslips and imaged with a 606 oilimmersion objective (NA 1.4, Plan Apo, Nikon, Tokyo, Japan) attached to an inverted microscope (Eclipse 200, Nikon) equipped with a Nipkow disk confocal scanning head (Ultraview, Perkin Elmner, USA). Confocal fluorescence images were captured with a scientific grade cooled CCD camera (Orca, Hamamatsu Photonics, Hamamatzu City, Shizuoka, Japan).

Confocal Imaging of Organotypic Utricle Cultures
Cultures were loaded with 10 mM cell permeant Fluo-4 AM (Invitrogen) for 50 min at 37uC in D-MEM supplemented with 10 mM cell permeant NP-EGTA AM (Invitrogen), 25 mM sulfinpyrazone and Pluronic F-127 (0.1% w/v). For de-esterification, cultures were transferred to an experimental chamber mounted on the stage of a confocal imaging setup (Biorad Radiance 2100) incorporating an upright microscope (Eclipse E600FN, Nikon, Tokyo, Japan) and superfused for 20 min with a medium composed of HBSS supplemented with 4.4 g/l glucose and 2 mM anhydrous CaCl 2 (pH 7.4, Osm 330). Experiments were performed with a 1006 water-immersion objective (N.A. 1.00, LUMPlanFl, Olympus, Tokyo, Japan) using the same perfusion medium. Fluo-4 fluorescence was excited by the 488 nm line of an argon laser coupled by fiber optics to the confocal microscope. Fluorescence emission was selected around 528 nm using a narrow-band (50 nm) interference filter. Fluorescence images of utricle hair cells in the organotypic cultures were acquired with a resolution of 5126512 pixel by scanning at 512 lines per second under control of the Biorad Laser Sharp software. To be retained for subsequent recording, imaged cells had hair bundles extending for their entire length in a single confocal plane (bundle planarity condition), a possibility afforded by having the culture attached to a curved surface.

Confocal Line-Scan Recording and UV Photolysis of Caged Ca 2+ in Utricular Hair Cells
Dynamic fluorescence data were acquired in the 'linescan' mode to produce a scan series of fluorescence intensity values, F, measured in the photomultiplier tube PMT units from a value of 0 to a saturating value of 255. Laser intensity and PMT gain were adjusted to accommodate the dynamic range of changes in F. Typical background fluorescence values, measured from regions devoid of obvious cellular structures, were ,1 PMT units, while pre-stimulus (basal) levels, F 0 , averaged over the entire length of the hair bundle, were ,8 PMT units. In all experiments pinhole aperture was adjusted to the same value, yielding confocal section with 3 mm thickness. To assay the Ca 2+ extrusion activity of the PMCAs, an area of ,3000 mm 2 , comprising a few hair cells in the cultured utricular macula, was exposed to UV radiation generated by an aircooled 337 nm pulsed nitrogen laser (Model VSL-337ND-S, Spectra Physics, Mountain View, CA, USA) connected to the microscope through a 600 mm optical fiber. UV light was directed onto the sample by reflection off a 400 DCLP dichromatic beam splitter (Chroma) positioned at 45u just above the microscope objective lens. A single laser pulse (4 ns) delivering a maximum of 326 mJ of energy (at the laser output) was used to photorelease Ca 2 from the caged state (NP-EGTA bound to Ca 2+ ) in the brief time interval between the 2500 th and the 2501 st scan line. In a typical record lasting about 120 s, 20 000 consecutive lines were acquired.

Off-Line Analysis of Fluorescence Transients
All data were analyzed offline on a personal computer using the Matlab 7.0 (The MathWorks, Inc., Natick, MA) software environment. Data are presented as DF/F 0 where DF = F2F 0 . In these expressions, raw pixel values are spatial averages along the hair bundle. Maximal percent fluorescence changes, (F max 2F 0 )/F 0 , were about 280%. To estimate the slow time constant, t, of recovery to baseline, transients peaking at DF max , were fitted by a single exponential function during the first 10 s from the UV pulse. Data are given as mean6standard error of the mean (S.E.M.).

Figure S1
Mutations identified to date affecting PMCA2 protein in mice. Three mutations are missense mutations leading to amino acid substitutions (dfw, wri, Obl), two are small deletions that lead to frame shift mutations and premature truncation of the PMCA2 protein (dfw 2J , dfw 3J ), and one is a targeted null allele (Atp2b2 tm1Ges ) [8,13,21,26]. Adapted from [21].