Hearing Research

Hearing Research

Hearing Research 228 (2007) 11–21

www.elsevier.com/locate/heares

Research paper

Plasma membrane Ca2+-ATPase isoforms in frog crista ampullaris: Identification of PMCA1 and PMCA2 specific splice variants Mariarosa Polimeni a

a,*

, Ivo Prigioni b, Giancarlo Russo b, Daniela Calzi b, Luciana Gioglio

a

Dipartimento di Medicina Sperimentale – Sezione di Anatomia Umana Normale, Universita` di Pavia, Via Forlanini 8, I-27100 Pavia, Italy b Dipartimento di Scienze Fisiologiche Farmacologiche Cellulari e Molecolari, Universita` di Pavia, Via Forlanini 6, I-27100 Pavia, Italy Received 24 July 2006; received in revised form 14 December 2006; accepted 14 December 2006 Available online 24 January 2007

Abstract Ca2+ ions play a pivotal role in inner ear hair cells as they are involved from the mechano-electrical transduction to the transmitter release. Most of the Ca2+ that enters into hair cells via mechano-transduction and voltage-gated channels is extruded by the plasma membrane Ca2+-ATPases (PMCAs) that operate in both apical and basal cellular compartments. Here, we determined the identity and distribution of PMCA isoforms in frog crista ampullaris: we showed that PMCA1, PMCA2 and PMCA3 are expressed, while PMCA4 appears to be negligible. We also identify PMCA1bx, PMCA2av and PMCA2bv as the major splice variants produced from PMCA1 and PMCA2 genes. PMCA2av appears to be the major Ca2+-pump operating at the apical pole of the cell, even if PMCA1b is also expressed in the stereocilia. PMCA1bx is, instead, the principal PMCA of hair cell basolateral compartment, where it is expressed together with PMCA2 (probably PMCA2bv) and PMCA3. Frog crista ampullaris hair cells lack a Na/Ca exchanger, therefore PMCAs are the only mechanism of Ca2+ extrusion. The coexpression of specific isozymes in the different cellular compartments responds to the need of a fine regulation of both basal and dynamic Ca2+ levels at the apical and basal pole of the cell.  2006 Elsevier B.V. All rights reserved. Keywords: Frog; Semicircular canal; Crista ampullaris; Hair cells; Calcium pump; Isozymes

1. Introduction Calcium ions play a pivotal role in the transduction process and excitability of hair cells, the inner ear sensory recepAbbreviations: PMCA, plasma membrane Ca2+-ATPase; SERCA, sarcoplasmic-endoplasmic reticulum Ca2+-ATPase; PBS, phosphate-buffered saline; Ig, immunoglobuline; BSA, bovine serum albumine; DAB, diaminobenzidine; RT-PCR, reverse transcriptase-polymerase chain reaction; M-MLV-RTase, moloney-murine leukemia virus-reverse transcriptase; cDNA, complementary deoxyribonucleic acid; TBE, tris-borate/ EDTA electrophoresis buffer; X-gal, 5-bromo-4-chloro–3-indolyl-b-Dgalactoside; IPTG, isopropylthio-b-D-galactoside; IP3R, inositol triphosphate receptors * Corresponding author. Tel.: +39 382 987659; fax: +39 382 422117. E-mail addresses: [email protected] (M. Polimeni), [email protected] (I. Prigioni), [email protected] (G. Russo), [email protected] (D. Calzi), [email protected] (L. Gioglio). 0378-5955/$ - see front matter  2006 Elsevier B.V. All rights reserved. doi:10.1016/j.heares.2006.12.016

tors. Mechanical stimulation of the hair cells, i.e. ciliary bundle deflection, leads to the opening of transduction channels (see review, Hudspeth et al., 2000), which are cationselective channels, that also allow calcium ion influx (Lumpkin et al., 1997; Ricci and Fettiplace, 1998). Calcium entering into the stereocilia is involved in fast and slow adaptation processes (Eatock et al., 1987; Crawford et al., 1991) and also triggers active hair bundle motion (Ricci et al., 2000). Another main route of Ca2+ entry in hair cells is represented by voltage-gated Ca2+ channels of the basolateral membranes (Fuchs et al., 1990; Prigioni et al., 1992; Platzer et al., 2000). Ca2+ entering in this way is involved in the activation of Ca2+-activated potassium channels (Roberts et al., 1990; Fettiplace and Fuchs, 1999) and in triggering neurotransmitter release from presynaptic active zones (Parsons et al., 1994; Spassova et al., 2001).

12

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

Due to the central role of calcium ions in regulating hair cell activity, the control of the intracellular calcium concentration in these cells is critical. Together with mobile Ca2+ buffers (Ricci et al., 1998; Hackney et al., 2005), three mechanisms are known to lower cytoplasmic calcium concentration: the plasma membrane Ca2+-ATPases (PMCAs) (Gioglio et al., 1998; Yamoah et al., 1998; Boyer et al., 2001; Dumont et al., 2001), the sarcoplasmicendoplasmic reticulum calcium pumps (SERCAs) (Tucker and Fettiplace, 1995) as well as Na/Ca exchangers (Ikeda et al., 1992; Boyer et al., 1999). PMCAs seem to be the most relevant mechanism of Ca2+ extrusion, at least in the hair bundles which lack Ca2+ intracellular stores and are exposed to the endolymph whose ionic composition is incompatible with the activity of a Na/Ca exchanger (Yamoah et al., 1998). PMCAs are encoded by four independent genes (PMCA1-4) and each primary transcript can be subjected to differential splicing (Keeton et al., 1993). These rearrangements occur at two main sites (A and C) and affect regulatory elements of the protein (acid phospholipid sensitivity, phosphorilation sites and calmodulin binding sites) but not the functional domains (ATP-binding and Ca2+transport domains) (Keeton et al., 1993; Carafoli, 1994). The alternative splice variants at site A are designated by the letters at the end of the alphabet (v, w, y. . .), those at site C by letters at the beginning of the alphabet (a, b, c. . .). This variety of isoforms and splice variants often shows differences in tissue or cellular distribution and some authors correlate these expression patterns to their physiological properties (Stauffer et al., 1995; Garcia and Strehler, 1999). The mammalian cochlea expresses specific splice variants derived from all of the four PMCA genes (1b, 2b, 3a, 3c, 4b) (Crouch and Schulte, 1996; Furuta et al., 1998). In amphibians, the only cloned PMCAs are PMCA1 and PMCA2 and specifically their splice variants 1bx, 2av, 2bv and 2cv (Dumont et al., 2001). Among these numerous PMCA splice variants, PMCA1b and PMCA2a were found to be the most abundant in hair cells of mammals and amphibians; immunolabelling experiments demonstrated that PMCA1b represents the major isozyme of both hair and supporting cell basolateral membranes, while PMCA2a appears to be the only Ca2+-pump present in hair bundles (Dumont et al., 2001). We used this knowledge as our basis to identify the expression and the distribution of PMCA isoforms and their splice variants in crista ampullaris of frog semicircular canals. We chose this model for the following reasons: (1) using a cytochemical method we found a marked Ca2+ATPase activity in hair cell bundles and apical membranes (Gioglio et al., 1998); (2) PMCAs in frog canal hair cells would play a primary role in Ca2+ extrusion since these cells lack a Na/Ca exchanger (Martini et al., 2002); (3) the lack of the Na/Ca exchanger could be compensated by the expression of different PMCA isozymes in specific membrane domains.

The main purpose of this study was to identify and localize PMCA isoforms in hair cells of the frog crista ampullaris. Our immunolabelling experiments revealed that at least three PMCA isoforms (PMCA1, PMCA2 and PMCA3), even though at different levels, are coexpressed in hair cell specific cell membrane domains. In addition, RT-PCR experiments showed that PMCA1bx, PMCA2av and PMCA2bv splice variants are the major isozymes present in the ampulla. 2. Materials and methods The experiments were performed on vertical posterior semicircular canals of adult frogs (Rana esculenta) purchased from local suppliers. Frogs weighing 20–30 g were anaesthetized by immersion in 0.1% tricaine metanesulphonate (MS222) solution (Sandoz, Basel, Switzerland). After decapitation, the otic capsule was opened to expose the semicircular canal, and the ampulla was dissected in Ringer solution as previously described (Prigioni et al., 1990; Masetto et al., 1994). 2.1. Immunohistochemistry The PMCA proteins were detected by indirect immunoperoxidase or indirect immunofluorescence on frog crista ampullaris sections. Ampullae were fixed for 1 h at 4 C in 3% paraformaldehyde and 0.5% glutaraldehyde dissolved in Millonig’s buffer (pH 7.25). Specimens were dehydrated with gradients of alcohol, cleared in xylene and embedded in paraffin. Serial longitudinal sections (5 lm thick) were cut from each ampulla using a Leitz microtome; the sections were then deparaffinized, rehydrated, and rinsed in PBS. Endogenous peroxidase reactivity was blocked by incubating the sections for 20 min in 3% hydrogen peroxide (Sigma, St. Louis, MO) diluted in absolute methanol and then for 20 min in sodium borohydrate 0.5% (Sigma) in PBS. After rinsing in PBS for 30 min, non-specific binding of Ig was prevented by incubation of sections in 50 mM NH4Cl for 40 min, followed by incubation for 30 min in PBS containing 10% normal goat serum (Vector Labs, Burlingame, CA), 2% BSA and 0.3% Triton X-100 (Sigma). Sections were incubated overnight at 4 C with primary antibodies (Affinity BioReagents, Golden, CO) diluted in PBS. We used the PMCA mouse monoclonal antibody 5F10 (1:200 dilution) that detects an epitope present in all of the four PMCAs isoforms and is highly conserved among different species. To detect the different PMCA isozymes we used antibodies raised against the N-terminus of rat or human PMCAs: rabbit policlonal antibodies against PMCA1 (1:200–1:600 dilution), PMCA2 (1:200 dilution) and PMCA3 (1:50 dilution), and a mouse monoclonal antibody against PMCA4 (1:25–1:50 dilution). From the analysis of frog published sequences we determined that the epitopes used to generate the anti-PMCA1 and PMCA2 antibodies were conserved at 83% and 100%, respectively,

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

in the frog homologous proteins. This analysis was not possible for PMCA3 and PMCA4 because these isoforms have not yet been cloned in the frog. Moreover, the epitope used to generate the anti-PMCA3 antibody was quite well conserved in mammals and showed a 75% identity with the protein in crayfishes where three out of four substitutions maintained aminoacid charge. Similarly the epitope used to generate the anti-PMCA4 antibody was well conserved through the species (88% identity between mammals and birds). For immunoperoxidase labelling, tissue sections were washed with PBS and incubated with the biotinylated goat anti-rabbit or horse anti-mouse secondary antibody (Elite kit, Vector Labs) for 1 h at room temperature, rinsed with PBS and incubated with the avidin–biotin-peroxidase solution (Elite kit, Vector Labs) for 1 h at room temperature. After the PBS washes, the immunohistochemical reaction was visualized by incubating sections with a diaminobenzidine solution (DAB kit, Vector Labs) for 2–4 min at room temperature and then rinsing in distilled water. Sections were counterstained with ematoxylin, dehydrated with alcohol gradients followed by xylene, mounted and observed with a Leitz ortoplan light microscope. For immunofluorescence labelling, after the primary antibody treatment, tissue sections were washed in PBS and then incubated for 1 h at room temperature with the secondary antibodies: Cy5-coniugated goat anti-rabbit (1:1200 dilution; Jackson ImmunoResearch Laboratories Inc., West Grove, PA) or Alexa 594-coniugated goat anti-mouse (1:1500 dilution, Molecular Probes Inc., Eugene, OR). Finally, sections were incubated for 5 min at room temperature with Hoecst 33258 (Sigma) to label the nuclei and mounted in PBS/glycerol. Labelled sections were observed using a TCS SP2 LEICA confocal microscopy system equipped with a LEICA DM IRBE inverted microscope. Three laser lines were used for the excitation of the different fluorochromes and specifically: a He/Ne laser (em 633 nm) for Cy5, an Ar/Vis laser (em 543 nm) for Alexa 594 and an Ar/UV laser (em 351/364 nm) for Hoecst. The bandwidths to collect fluorescence emissions were 650/720 nm for Cy5, 590/650 nm for Alexa 594 and 400/490 nm for Hoecst. The objective lenses used were LEICA PLANAPO 20X dry (NA = 0.70) and a LEICA HCX PLANAPO 60X oil (NA = 1.32). Optical sections of about 200 nm were collected at intervals of 930 nm. Images were processed using the LCS software and Adobe Photoshop (version 7.0, San Jose, CA). Non-specific staining was tested by omitting the primary antibodies. Sections from rat cerebral and cerebellar cortex were used as positive control (data not shown). 2.2. Immunoelectron microscopy After dissection the ampullae were included into small blocks of agar that were attached to a Teflon plate. Blocks were cut along the longitudinal axis of the crista using a vibroslicer (Campden-Instrument, UK) to obtain slices

13

about 100 lm thick. Sections were then fixed and treated for PMCAs immunoperoxidase labelling as for immunohistochemistry. After the DAB-incubation, sections were postfixed 1 h at 4 C in 3% paraformaldehyde and 0.5% glutaraldehyde, and 1 h at 4 C in OsO4 dissolved in Millonig’s buffer (pH 7.25). Specimens were dehydrated in graded ethanol series, processed through propylene oxide and embedded in epoxy resin. Ultrathin sections stained with uranyl acetate and lead citrate were examined with a Zeiss 109 transmission electron microscope. As for immunohistochemistry the primary antibody was omitted for negative controls. 2.3. RT-PCR screening Ampullae from five frogs were dissected and immediately frozen in liquid nitrogen; the pool was then homogenized in a guanidinium thyocyanate solution. The same procedure was used to collect sacculi samples. Frog brain expresses most of the known PMCA splice variants and for this reason it was chosen as positive control for all the experiments: the tissue was dissected in cold Ringer solution and immediately homogenized in a guanidinium thyocyanate solution. Total RNAs were prepared from the ampulla, sacculus and brain homogenates by the guanidinium thyocyanateacid phenol extraction method as previously described (Polimeni et al., 1996) and their integrity was verified by gel electrophoresis through denaturing agarose gel. For each sample, 2 lg of total RNA was primed with random hexamers and reverse transcribed with M-MLV-RTase for 1 h at 42 C in a 20 ll reaction. The cDNAs were subjected to PCR with nested primers flanking splice region A or C in the PMCA genes to check for variability at these sites. No evidence of alternative splicing occurring at PMCA1 splice region A have ever been reported (Strehler and Zacharias, 2001), so we performed PCR using nested primers for splice region C only; for PMCA2 we used nested primers for both splice regions A and C. PMCA1 and 2 primers were chosen on the basis of published frog sequence (Dumont et al., 2001): their name, sequence and annealing temperature are indicated in Table 1. No sequence data were available for frog PMCA3 and PMCA4. The amplification protocol comprised a 90 s denaturation step at 94 C, followed by 35 amplification cycles of 50 s at 94 C, 40 s at 50–62 C (depending on primers), 45 s at 72 C, and a final 10 min extension step at 72 C. PCR with external primers was performed on 5% of the cDNA; amplification with the internal primers was performed on 3% of the first PCR reaction product. No RTase was added in the negative control samples. PCR products were separated by electrophoresis through 1.5% agarose gel in TBE. The size of each PCR product was estimated by comparison with standard molecular weight markers (lambda HindIII, 50 bp and 25 bp DNA ladders from Invitrogen, CA).

14

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

Table 1 Sequence and annealing temperature (Ta) of nested primers used for amplification of PMCA1 region C and PMCA2 regions A and C PMCA isoform

Splice region

Primers

PMCA1

C

External primers Internal primers

PMCA2

A

External primers Internal primers

C

External primers Internal primers

Ta (C) F1Cf1 F1Cr1 F1Cf2 F1Cr2 F2Af1 F2Ar1 F2Af2 F2Ar2 F2Cf1 F2Cf1rev2 F2Cf2 F2Cf2rev2

gcagagagggagttacgccg catttgaagtccaaggagc ggtcagatcttatggtttag gcactcttctgcctgctgc ctcatgtgatggaaggctc tgcctatctgcactgccag gaagaatgcttgtaacagc ggtcagcttgccctgtagg cgcctgacacagaaggagg caggctgtgaatggggctc gggagttaagaagagggc gttttgtttcagggcagg

55 50 55 50 62 60

In order to clearly identify the different splice variants expressed in the ampulla, PCR products were gel purified and subcloned in the pGEM-T vector using the pGEM-T Easy system (Promega, Madison, WI) and the obtained plasmids were transformed into XL1blue competent cells. Colonies were plated on selective medium supplemented with X-gal/IPTG and recombinant clones were selected by the b-galactosidase a-complementation assay. Plasmids were purified from liquid bacterial coltures, digested with restriction enzymes to verify the size and orientation of cloned inserts and sequence verified. Animal experiments have been approved by the Animal Care and Use Committee of the University of Pavia. 3. Results The crista ampullaris of the frog posterior canal is shaped like a dumbbell with two enlarged peripheral regions curving along the ampullar wall separated from a central isthmus by a thin intermediate region (Hillman, 1976). The sensory epithelium (Fig. 1A) consists of hair cells surrounded by supporting cells, reaching the surface of the crista, and basal cells located close to the basement membrane (Gioglio et al., 1995). Sensory cells have a convex apical surface with a kinocilium and a long bundle of stereocilia, while supporting cells have a flattened apical surface with numerous and short microvilli (Gioglio et al., 1995). To investigate the presence of PMCAs in sensory epithelium of the crista we used 5F10 monoclonal antibody that binds all known PMCA isoforms (Caride et al., 1996). An evident labelling was observed in all three crista regions either by using immunoperoxidase staining or immunofluorescence confocal microscopy (Fig. 1A and B). The 5F10 antibody strongly labelled hair bundles and basolateral membranes of hair cells, while a weaker reactivity was seen on cell apical membranes (Fig. 1A and B). No significant differences were seen in the labelling of hair cells from the peripheral, intermediate and central region (Fig. 1A). The antibody also labelled apical and basolateral membranes of supporting cells, even if the staining was less evident

Fig. 1. 5F10 immunolabelling of PMCA isoforms in frog crista epithelium. (A) Immunoperoxidase staining is shown on a half crista where the central (CR), intermediate (IR) and peripheral (PR) regions are indicated. The antibody strongly labels hair bundles (arrows) and basolateral membranes of hair cells, while a lower staining is detectable on their apical membranes. A discrete labelling is present on apical (arrow heads) and basolateral membranes of supporting cells. No labelling is detectable in basal cells (stars). (B) Confocal micrograph showing a higher magnification of the central region: hair cells basolateral membranes and hair bundles exhibit a clear immunolabelling. (C, D) Negative control obtained by omitting the primary antibody in the immunoperoxidase (C) and in the immunofluorescence (D) experiments. Bars = 10 lm.

when compared with that of hair cells. Finally, immunoreactivity was absent in basal cells. In control preparations, where the primary antibody was omitted, no labelling was detected in the crista epithelium (Fig. 1C and D). To examine the specific distribution of the different isozymes we used a set of antibodies raised against the four PMCA isoforms. The anti-PMCA1 antibody intensely labelled the basolateral membranes of both hair and supporting cells (Fig. 2A and B). Most hair bundles showed a discrete PMCA1 labelling mainly localized at the proximal region of the stereocilia. This was clearly apparent at higher magnification (Fig. 2A insert, B). In addition

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

15

Fig. 2. PMCA1 immunolocalization. (A) The immunoperoxidase reaction shows a clear staining of basolateral membranes of hair and supporting cells. Apical membranes were stained in most supporting cells while only an occasional labelling was detectable on hair cells apical membranes. The hair bundles labelling is mainly localized in the proximal region of the stereocilia: the insert shows a higher magnification (2·) of the hair bundle, highlighted in the figure. (B) Staining of the proximal region of hair bundles and of the hair cell basolateral membranes by immunofluorescence. Bars = 20 lm. (C) Immunoelectron microscopy of a hair cell apical region showing the labelling of the single stereocilia of the bundle as well as the basolateral membrane staining. In this cell the occasional labelling of the apical membrane is also detectable. Bar = 2 lm.

immunoelectron microscopy showed that PMCA1 labelling involves all stereocilia of the hair bundle (Fig. 2C). Apical membranes were labelled in most supporting cells (Fig. 2A), while they were only occasionally labelled in hair cells (Fig. 2A and C). The anti-PMCA2 antibody strongly labelled the whole hair bundle in all sensory cells of the crista. An evident labelling was also detectable on apical and basolateral membranes of these cells (Fig. 3A and B). PMCA2 staining on hair cell basolateral membranes showed a spot distribution consisting of large and medium-sized particles which were particularly evident by immunofluorescence (Fig. 3B). The anti-PMCA2 antibody also labelled the apical and basolateral membranes of supporting cells (Fig. 3A and B). The localization of PMCA2 at hair cell level using immunoelectron microscopy is shown in Fig. 3C. Consistent with immunoperoxidase and immunofluorescence experiments, apical and basolateral membranes as well as stereocilia were labelled with the antibody against the PMCA2. The antibody against the PMCA3 isoform stained hair cells basolateral membranes (Fig. 4A). The labelling consisted of large sized spots that were more clearly detectable by immunofluorescence (Fig. 4B, B 0 ) and were particularly

evident in the central region of the crista. The immunoperoxidase showed an occasional labelling of hair bundles and a weak labelling of supporting cell apical membranes (Fig. 4A). The immunoelectron microscopy confirmed PMCA3 expression in the basolateral membranes of hair cells and excluded the labelling of the neighbouring supporting cell membranes (Fig. 4C). A very faint labelling was detectable with the antiPMCA4 antibody at the apical membranes of both hair and supporting cells (Fig. 5A and B). This staining was obtained only with very high concentrations of the antibody either by immunoperoxidase or immunofluorescence microscopy. In addition, almost no staining was detectable by immunoelectron microscopy (Fig. 5C). Taken together these data indicate that PMCA1 and PMCA2 are the major PMCA isoforms expressed in the frog crista ampullaris. Since PMCA1 and PMCA2 are the only isozymes cloned in amphibians (Dumont et al., 2001), we performed an RT-PCR screening in order to investigate which of the known splice variants are expressed in the crista as well as in the sacculus. Our analysis showed that specific splice variants from PMCA1 and PMCA2 genes are expressed in the ampulla. All the currently known PMCA1 isoforms correspond to

16

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

Fig. 3. PMCA2 immunolocalization. (A) The immunoperoxidase staining intensely labels the hair bundles and apical membranes of hair cells; an evident labelling is detectable also at the basolateral membranes of hair cells and in supporting cells. (B) The immunofluorescence clearly shows the apical labelling of hair cells and the spot distribution of PMCA2 at their basolateral membranes. Bars = 20 lm. (C) Immunoelectron microscopy detection of PMCA2 in hair (arrows) and supporting (arrow head) cells of the peripheral region. Bar = 2 lm.

the x variant, suggesting that this messenger RNA is not subjected to alternative splicing at site A (Strehler and Zacharias, 2001). In addition, PMCA1x has been shown to be the only splice variant expressed in bullfrog sacculus (Dumont et al., 2001). Using nested primers for PMCA1 splice region C we obtained a clear 439 bp band corresponding to the PMCA1b splice variant (Fig. 6, lane C). For the PMCA2 isoform we used two sets of nested primers. The first set of primers amplified a single 410 bp band from splice region A, corresponding to the PMCA2v frog specific splice variant (Fig. 7A, lane C). The second set of primers amplified two bands of 463 bp and 236 bp from region C that correspond to the splice variants PMCA2a and PMCA2b, respectively (Fig. 7B, lane C). Altogether, these data indicate that a single PMCA1 splice variant, corresponding to PMCA1bx, is expressed in the ampulla while alternative splicing of the PMCA2 mRNA give rise to two different isozymes, PMCA2av and PMCA2bv. Our RT-PCR analysis on sacculus samples showed that the 1b variant was expressed from the PMCA1 gene (Fig. 6, lane S), while the PMCA2 splicing results in the 2av variant (Fig. 7A and B, lane S). It is worthy of interest to note that PMCA2bv splice variant is abundant in frog crista ampullaris, but is lacking in frog sacculus, pointing out that ampullar and saccular PMCA expression patterns are not completely overlapping.

4. Discussion Our data demonstrate that at least three PMCA isoforms (PMCA1, 2 and 3) are expressed, at different levels, in frog crista ampullaris and that only two cell types of the sensory epithelium, hair cells and supporting cells, show immunolabelling. The presence of PMCAs in these cells was revealed by the immunoreactivity observed using the PMCA antibody 5F10, which reacts with all four isoforms in mammals (Caride et al., 1996), as well as in a wide range of species (de Talamoni et al., 1993; Benaim et al., 1995). We also found that among the different PMCAs, PMCA1 and PMCA2 are the prominent isoforms. Recent immunocytochemical experiments showed that PMCA2 is the principal Ca2+-pump of hair bundles and PMCA1 is the major isoform of hair cell basolateral membranes in both frog sacculus and rat auditory and vestibular hair cells (Dumont et al., 2001). These and other data (Strehler and Zacharias, 2001; Chicka and Strehler, 2003) indicating specific functions and localizations for distinct PMCA splice variants, have generated the current view regarding the segregation of different PMCA isoforms in discrete membrane domains of hair cells. Here, we show that in frog crista ampullaris, even though PMCA2 is the principal Ca2+-pump of the hair bundle, a second PMCA, i.e. PMCA1, is expressed in the

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

17

Fig. 4. PMCA3 immunolocalization. (A) The immunoperoxidase reaction shows PMCA3 spot distribution on hair cell basolateral membranes (arrows), an occasional labelling of hair bundles and a weak labelling of supporting cell apical membranes (arrow heads). (B, B 0 ) Confocal and bright-field Nomarski images showing a clear spot distribution of PMCA3 on the basolateral membranes of a hair cell from the central region. (C) Immunoelectron micrograph showing the spot distribution of PMCA3 on hair cell basolateral membrane (arrows) and the absence of labelling on the neighbouring supporting cell plasma membrane. h = hair cell, s = supporting cell. Bar = 10 lm for A, B and B 0 , Bar = 1 lm for C.

stereocilia. However, the localization of PMCA1 appears to be restricted to the proximal region of the stereocilia. In addition PMCA1 and PMCA2 are coexpressed in the hair cell basolateral membrane domain. Since, in our experiments, we used antibodies raised against the N-terminus of rat or human PMCAs, we performed a BLAST analysis to verify the conservation of these epitopes in frog PMCA1 and PMCA2 proteins and we concluded that these antibodies were also specific for frog PMCAs (see Section 2). An occasional faint PMCA1 labelling of the hair bundles has also been seen by other authors in frog sacculus hair cells and in rat outer hair cells (Dumont et al., 2001). Moreover, immunoprecipitation experiments performed using antibodies for frog specific PMCA1 splice variants, suggest the expression of a second PMCA1 isozyme in frog macula residua (Dumont et al., 2001). We found that hair cells of the frog crista ampullaris also express the PMCA3 isoform which is known to be present in excitable cells such as in the nervous system and skeletal muscle (Strehler and Zacharias, 2001). In hair cells, PMCA3 appears to be restricted to the basolateral domain, similarly to what was already seen in rat utricular and canal hair cells (Dumont et al., 2001). This isozyme has a clear spot distribution that strictly resembles that of the

intracellular IP3R (Rossi et al., 2006) in the basal region of the hair cell where IP3-sensitive stores have been identified. These stores are possibly involved in the modulation of glutamate release at the synaptic pole of the hair cell. Moreover, in frog semicircular canal, a functional role of Ca2+ stores in afferent transmission has been demonstrated (Hendricson and Guth, 2002; Lelli et al., 2003). These data suggest that the PMCA3 isozyme could be involved in regulating cytosolic Ca2+ levels at the synaptic pole of the cell, and consequently in neurotransmitter release modulation. A very faint labelling was detectable with the antiPMCA4 antibody even when we used antibody concentrations four times higher than those used for PMCA1 and PMCA2. We concluded that PMCA4 expression in frog crista ampullaris is negligible. Although no PMCA4 immunolabelling was detected in frog saccular hair cells (Dumont et al., 2001), in situ hybridization experiments revealed very low levels of PMCA4 in rat cochlea with the exception of a transient peak of expression at P12 in inner hair cells (Furuta et al., 1998). A moderate PMCA4 expression is present in the hair bundles of the outer and inner hair cells in deafwaddler mutant mice lacking PMCA2 (Wood et al., 2004). These data suggest that PMCA4 may have a developmental function and may be

18

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

Fig. 5. PMCA4 immunolocalization. Both immunoperoxidase (A) and immunofluorescence (B) reveal only a very faint labelling of the apical region of both hair and supporting cells. Bars = 20 lm. (C) Immunoelectron microscopy showing the absence of labelling on a hair cell and on the neighbouring supporting cells. Bar = 2 lm.

up-regulated to compensate the loss of other PMCAs. It is therefore not surprising that we did not detect this isozyme in adult frog crista ampullaris. Our data also show that the apical and basolateral membranes of supporting cells express PMCA1 and PMCA2 isoforms. Although the function of Ca2+-pumps in supporting cells is not known, it is worth mentioning that these cells are polarized cells with high metabolic and secretory activity; furthermore, supporting cells are probably involved in the control of the ionic environment around the hair cells (Sugihara and Furukawa, 1996). Therefore, it is not surprising that supporting cells are provided with fine systems for Ca2+ homeostasis control. Our molecular screening revealed that in the frog crista, PMCA1bx is the major PMCA1 splice variant, while PMCA2av and PMCA2bv are the major splice variants for PMCA2. We suggest that in the frog crista, PMCA2av could be the principal pump of the hair cell apical compartment in agreement with the current view (Chicka and Strehler, 2003). PMCA2 has a very low Km for Ca2+ and it is able to bring the ion concentration at lower levels than other isozymes (Elweess et al., 1997; Strehler and Zacharias, 2001). This feature appears to be tailored for hair bundles, which lack intracellular Ca2+ stores (Yamoah et al., 1998). PMCA2a density on stereocilia membranes of the frog sacculus appears very high (2000/lm2) and the pump

appears able to extrude most of Ca2+ that enters the stereocilium during transduction (Yamoah et al., 1998; Lumpkin and Hudspeth, 1998; Dumont et al., 2001). This intense pumping could prevent the diffusion of Ca2+ ions from the stereocilia into the cell soma and play a pivotal role in the mechano-transduction process (Dumont et al., 2001) and in maintaining the normal endolymph Ca2+ concentration (Wood et al., 2004). Our study shows a coexpression of PMCA1 and PMCA2 in the stereocilia, and suggests that PMCA1bx is the only PMCA1 splice variant expressed in the crista. All published data seems to indicate a basolateral localization for the PMCA1bx isoform: in fact the b form can interact with the PDZ proteins (DeMarco and Strehler, 2001) and the x variant is supposed to target the protein in this domain (Chicka and Strehler, 2003). The unusual localization of PMCA1 we detected in the proximal region of the hair bundles could be explained in different ways: in the apical region PMCA1bx isozyme could interact with specific PDZ proteins expressed in the stereocilia, like harmonin or whirlin recently identified in mammals (Verpy et al., 2000; Mburu et al., 2003); another possibility is that a second PMCA1 splice variant, that could not be evidenced by our molecular screening, is expressed in the hair cell apical compartment, as already suggested by Dumont et al. (2001) in frog sacculus.

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

Fig. 6. RT-PCR analysis of frog PMCA1 splice variants at site C. The diagram illustrates frog PMCA1 isozyme structure with the ten putative transmembrane domains (numbered boxes) and its known splice variants at site A (x) and C (b). Both sacculus (S) and semicircular canal (C) samples express PMCA1b isozyme. B = brain, positive control; - = negative control; mwm = molecular weight markers.

Our study also indicates that in frog crista ampullaris PMCA2 could operate in the hair cell basolateral domain. We suggest that PMCA2bv could be the PMCA2 splice

19

variant that operates at the hair cell basolateral membranes together with the PMCA1bx. This hypothesis is supported by the observation that both the b splice variants of these isozymes have high affinity for synapse-associate PDZ proteins (Garcia and Strehler, 1999; DeMarco and Strehler, 2001). This interaction, due to the presence of a minimal consensus sequence at the C-terminal of the protein, is apparently limited to the PMCA b splice forms (Strehler and Zacharias, 2001). Moreover, PMCA2b PDZ-binding partners are likely binding to PMCA 1b and 3b as well (DeMarco and Strehler, 2001). The need to have different isozymes in the basolateral membranes may be due to the fact that the hair cells of frog crista ampullaris lack a Na/Ca exchanger (Martini et al., 2002). It is worth mentioning that PMCA2bv appears peculiar to the ampulla since it is absent in the otolith organs in which the Na/Ca exchangers are expressed (Boyer et al., 1999). It is known that most of Ca2+ entering into hair cells follows the route of voltage-gated Ca2+ channels, and thus the two isozymes could play different roles in controlling variations of intracellular Ca2+ levels. Differences in the function of these splice variants have been described: PMCA2 splice variants shows the highest affinity for calmodulin, while PMCA1 splice variants has the lowest Kd for ATP (Strehler and Zacharias, 2001). Since PMCA2 has the highest Ca2+ pumping activity (Elweess et al., 1997) it is likely that it is involved in dynamic Ca2+ regulation, while PMCA1 could maintain basal Ca2+ levels. In conclusion, our data indicate that in frog crista ampullaris different PMCA isoforms are coexpressed in hair cell

Fig. 7. RT-PCR analisys for frog PMCA2 splice variants. The diagram illustrates frog PMCA2 isozyme structure with the ten putative transmembrane domains (numbered boxes) and its known splice variants at site A (v, w and z) and C (a, b and c). Both semicircular canal (C) and sacculus (S) samples express the PMCA2v variant from splice region A (panel A). At site C, the alternative splicing generates the PMCA2a and PMCA2b variants in the semicircular canal, and the PMCA2a variant only in the sacculus (panel B). B = brain, positive control; - = negative control; mwm = molecular weight markers.

20

M. Polimeni et al. / Hearing Research 228 (2007) 11–21

apical and basolateral domains in which they could play different roles in Ca2+ homeostasis. Acknowledgements This work was supported by the Italian ‘‘Ministero dell’Istruzione, dell’Universita` e della Ricerca, MIUR’’: FIRB (Project RBAU01WWXZ, 2001). We thank Mr. Maurizio Alloni and Mr. Domenico Burrone for their expert technical assistance. References Benaim, G., Moreno, S.N., Hutchinson, G., Cervino, V., Hermoso, T., Romero, P.J., Ruiz, F., de Souza, W., Docampo, R., 1995. Characterization of the plasma-membrane calcium pump from Trypanosoma cruzi. Biochem. J. 306, 299–303. Boyer, C., Art, J.J., Dechense, C.J., Lehouelleur, J., Vuatrin, J., Sans, S., 2001. Contribution of plasmalemma to Ca2+ homeostasis in hair cells. J. Neurosci. 15, 2640–2650. Boyer, C., Sans, A., Vautrin, J., Chabbert, C., Lehouelleur, J., 1999. K+dependence of Na+–Ca2+ exchange in type I vestibular sensory cells of guinea pig. Eur. J. Neurosci. 11, 1955–1959. Carafoli, E., 1994. Biogenesis: plasma membrane calcium ATPase: 15 years of work on the purified enzyme. Faseb J. 8, 993–1002. Caride, A.J., Filoteo, A.G., Enyedi, A., Verma, A.K., Penniston, J.T., 1996. Detection of isoform 4 of the plasma membrane calcium pump in human tissues by using isoform-specific monoclonal antibodies. Biochem. J. 316, 353–359. Chicka, M.C., Strehler, E.E., 2003. Alternative splicing of the first intracellular loop of plasma membrane Ca2+-ATPase isoform 2 alters its membrane targeting. J. Biol. Chem. 278, 18464–18470. Crouch, J.J., Schulte, B.A., 1996. Identification and cloning of site C splice variants of plasma membrane Ca-ATPase in the gerbil cochlea. Hear. Res. 101, 55–61. Crawford, A.C., Evans, M.G., Fettiplace, R., 1991. The actions of calcium on the mechano-electrical transducer current of turtle hair cells. J. Physiol. (Lond.) 434, 369–398. DeMarco, S.J., Strehler, E.E., 2001. Plasma membrame Ca2+-ATPase isoforms 2b and 4b interact promiscuously and selectively with members of the membrane associated guanilate kinase family of PDZ (PSD95/Dlg/ZO-1) domain-containing proteins. J. Biol. Chem. 276, 21590–21600. de Talamoni, N., Smith, C.A., Wasserman, R.H., Beltramino, C., Fullmer, C.S., Penniston, J.T., 1993. Immunocytochemical localization of the plasma membrane calcium pump, calbindin-D28k, and parvalbumin in Purkinje cells of avian and mammalian cerebellum. Proc. Natl. Acad. Sci. USA 90, 11949–11953. Dumont, R.A., Lins, U., Filoteo, A.G., Penniston, J.T., Kachar, B., Gillespie, P.G., 2001. Plasma membrane Ca2+-ATPase isoform 2a is the PMCA of hair bundles. J. Neurosci. 15, 5066–5078. Eatock, R.A., Corey, D.P., Hudspeth, A.J., 1987. Adaptation of mechano-eletrical transduction in hair cells of the bullfrog sacculus. J. Neurosci. 7, 2821–2836. Elweess, N.L., Filoteo, A.G., Penniston, J.T., 1997. Plasma membrane Ca2+ pump isoforms 2a and 2b are unusually responsive to calmodulin and Ca2+. J. Biol. Chem. 272, 17981–17986. Fettiplace, R., Fuchs, P.A., 1999. Mechanisms of hair cell tuning. Annu. Rev. Physiol. 61, 809–834. Fuchs, P.A., Evans, M.G., Murrow, B.W., 1990. Calcium currents in hair cells isolated from the cochlea of the chick. J. Physiol. (Lond.) 429, 553–568. Furuta, H., Luo, L., Hepler, K., Ryan, A.F., 1998. Evidence for differential regulation of calcium by outer versus inner hair cells: plasma membrane Ca-ATPase gene expression. Hear. Res. 123, 10–26.

Garcia, M.L., Strehler, E.E., 1999. Plasma membrane calcium ATPase as critical regulators of calcium homeostasis during neuronal cell function. Front. Biosci. 4, 869–882. Gioglio, L., Congiu, T., Quacci, D., Prigioni, I., 1995. Morphological features of different regions in frog crista ampullaris (Rana esculenta). Arch. Histol. Cytol. 58, 1–16. Gioglio, L., Russo, G., Marcotti, W., Prigioni, I., 1998. Localization of Ca-ATPase in frog crista ampullaris. NeuroReport 9, 1309–1312. Hackney, C.M., Mahendrasingam, S., Penn, A., Fettiplace, R., 2005. The concentrations of calcium buffering proteins in mammalian cochlear hair cells. J. Neurosci. 24, 7867–7875. Hendricson, A.W., Guth, P.S., 2002. Transmitter release from Rana pipiens vestibular hair cells via mGluRs: a role for intracellular Ca2+ release. Hear. Res. 172, 99–109. Hillman, D.E., 1976. Morphology of peripheral and central vestibular systems. In: Linas, R., Precht, W. (Eds.), Frog Neurobiology. Springer-Verlag, Berlin, GE, pp. 452–480. Hudspeth, A.J., Choe, Y., Mehta, A.D., Martin, P., 2000. Putting ion channels to work: mechanoelectrical transduction, adaptation, and amplification by hair cells. Proc. Natl. Acad. Sci. USA 97, 11765– 11772. Ikeda, K., Saito, Y., Nishiyama, A., Takasaka, T., 1992. Na2+–Ca2+ exchange in the isolated cochlear outer hair cells of the guinea-pig studied by fluorescence image microscopy. Pflugers Arch. 420, 493– 499. Keeton, T.P., Burk, S.E., Shull, G.E., 1993. Alternative splicing of exons encoding the calmodulin-binding domains and C termini of plasma membrane Ca2+-ATPase isoforms 1, 2, 3, and 4. J. Biol. Chem. 268, 2740–2748. Lelli, A., Perin, P., Martini, M., Ciubotaru, C.D., Prigioni, I., Valli, P., Rossi, M.L., Mammano, F., 2003. Presynaptic calcium stores modulate afferent release in vestibular hair cells. J. Neurosci. 23, 6894–6903. Lumpkin, E.A., Marquis, R.E., Hudspeth, A.J., 1997. The selectivity of the hair cell’s mechanoelectrical transduction channel promotes Ca2+ flux at low Ca2+ concentrations. Proc. Natl. Acad. Sci. USA 94, 10997–11002. Lumpkin, E.A., Hudspeth, A.J., 1998. Regulation of free Ca2+ concentration in hair-cell stereocilia. J. Neurosci. 18, 6300–6318. Martini, M., Rossi, M.L., Farinelli, F., Moriondo, A., Mammano, F., Rispoli, G., 2002. No evidence for calcium electrogenic exchanger in frog semicircular canal hair cells. Eur. J. Neurosci. 16, 1647–1653. Masetto, S., Russo, G., Prigioni, I., 1994. Differential expression of potassium currents by hair cells in thin slices of frog crista ampullaris. J. Neurophysiol. 72, 443–455. Mburu, P., Mustapha, M., Varela, A., Weil, D., El-Amraoui, A., Holme, R.H., Rump, A., Hardisty, R.E., Blanchard, S., Coimbra, R.S., Perfettini, I., Parkinson, N., Mallon, A.M., Glenister, P., Rogers, M.J., Paige, A.J., Moir, L., Clay, J., Rosenthal, A., Liu, X.Z., Blanco, G., Steel, K.P., Petit, C., Brown, S.D., 2003. Defects in whirlin, a PDZ domain molecule involved in stereocilia elongation, cause deafness in the whirler mouse and families with DFNB31. Nat. Genet. 34, 421–428. Parsons, T.D., Lenzi, D., Almers, W., Roberts, W.M., 1994. Calcium triggered exocytosis and endocytosis in an isolated presynaptic cell: capacitance measurements in saccular hair cells. Neuron 13, 875–883. Platzer, J., Engel, J., Scrott-Fischer, A., Stephan, K., Bova, S., Chen, H., Zheng, H., Striessing, J., 2000. Congenital deafness and sinoatrial node dysfunction in mice lacking class D L-type Ca2+ channels. Cell 102, 89–97. Polimeni, M., Giorgi, S., DeGregorio, L., Dragani, T.A., Molinaro, M., Cossu, G., Bouche, M., 1996. Differentiation dependent expression in muscle cells of ZT3, a novel zinc finger factor differentially expressed in embryonic and adult tissues. Mech. Dev. 54, 107–117. Prigioni, I., Masetto, S., Russo, G., Taglietti, V., 1992. Calcium currents in solitary hair cells isolated from frog crista ampullaris. J. Vestib. Res. 2, 31–39. Prigioni, I., Russo, G., Valli, P., Masetto, S., 1990. Pre- and post-synaptic excitatory action of glutamate agonists of frog vestibular receptors. Hear. Res. 46, 253–260.

M. Polimeni et al. / Hearing Research 228 (2007) 11–21 Ricci, A.J., Fettiplace, R., 1998. Calcium permeation of the turale hair cell mechanotransducer channel and its relation to the composition of endolymph. J. Physiol. (Lond.) 506, 159–173. Ricci, A.J., Crawford, A.C., Fettiplace, R., 2000. Active hair bundle motion linked to fast transducer adaptation in auditory hair cells. J. Neurosci. 20, 7131–7142. Ricci, A.J., Wu, Y.C., Fettiplace, R., 1998. The endogenous calcium buffer and the time course of transducer adaptation in auditory hair cells. J. Neurosci. 18, 8261–8277. Roberts, W.M., Jacobs, R.A., Hudspeth, A.J., 1990. Colocalization of ion channels involved in frequency selectivity and synaptic transmission at presynaptic active zones of hair cells. J. Neurosci. 10, 3664–3684. Rossi, M.L., Prigioni, I., Gioglio, L., Rubbini, G., Russo, G., Martini, M., Farinelli, F., Rispoli, G., Fesce, R., 2006. IP3 receptor in the hair cells of frog semicircular canal and its possible functional role. Eur. J. Neurosci. 23, 1775–1783. Spassova, M., Eisen, M.D., Saunders, J.C., Parsons, T.D., 2001. Chick cochlear hair cells exocytosis mediated by dihydropyridine-sensitive calcium channels. J. Physiol. (Lond.) 535, 689–696. Stauffer, T.P., Guerini, D., Carafoli, E., 1995. Tissue distribution of the four gene products of the plasma membrane Ca2+ pump. A study using specific antibodies. J. Biol. Chem. 270, 12184–12190.

21

Strehler, E.E., Zacharias, D.A., 2001. Role of alternative splicing in generating isoform diversity among plasma membrane calcium pumps. Physiol. Rev. 81, 21–50. Sugihara, I., Furukawa, T., 1996. Inwardly rectifying currents in hair cells and supporting cells in the goldfish sacculus. J. Physiol. (Lond.) 495, 665–679. Tucker, T., Fettiplace, R., 1995. Confocal imaging of calcium microdomains and calcium extrusion in turtle hair cells. Neuron 15, 1323– 1335. Verpy, E., Leibovici, M., Zwaenepoel, I., Liu, X.Z., Gal, A., Salem, N., Mansour, A., Blanchard, S., Kobayashi, I., Keats, B.J., Slim, R., Petit, C., 2000. A defect in harmonin, a PDZ domain-containing protein expressed in the inner ear sensory hair cells, underlies Usher syndrome type 1C. Nat. Genet. 26, 51–55. Wood, J.D., Muchinsky, S.J., Filoteo, A.G., Penniston, J.T., Tempel, B.L., 2004. Low endolymph calcium concentrations in deafwaddler mice suggest that PMCA2 contributes to endolymph calcium maintenance. JARO 5, 99–110. Yamoah, E.N., Lumpkin, E.A., Dumont, R.A., Smith, P.J., Hudspeth, A.J., Gillespie, P.G., 1998. Plasma membrane Ca2+ATPase extrudes Ca2+ from hair cell stereocilia. J. Neurosci. 18, 610–624.