Immunolocalization of calcium sensing and transport proteins in the murine endolymphatic sac indicates calciostatic functions within the inner ear

An exceptionally low calcium (Ca2+) concentration in the inner ear endolymph ([Ca2+]endolymph) is crucial for proper auditory and vestibular function. The endolymphatic sac (ES) is believed to critically contribute to the maintenance of this low [Ca2+]endolymph. Here, we investigated the immunohistochemical localization of proteins that are presumably involved in the sensing and transport of extracellular Ca2+ in the murine ES epithelium. Light microscopic and fluorescence immunolabeling in paraffin-embedded murine ES tissue sections (male C57BL/6 mice, 6–8 weeks old) demonstrated the presence of the calcium-sensing receptor CaSR, transient receptor potential cation channel subtypes TRPV5 and TRPV6, sarco/endoplasmic reticulum Ca2+-ATPases SERCA1 and SERCA2, Na+/Ca2+ exchanger NCX2, and plasma membrane Ca2+ ATPases PMCA1 and PMCA4 in ES epithelial cells. These proteins exhibited (i) membranous (apical or basolateral) or cytoplasmic localization patterns, (ii) a proximal-to-distal labeling gradient within the ES, and (iii) different distribution patterns among ES epithelial cell types (mitochondria-rich cells (MRCs) and ribosome-rich cells (RRCs)). Notably, in the inner ear membranous labyrinth, CaSR was exclusively localized in MRCs, suggesting a unique role of the ES epithelium in CaSR-mediated sensing and control of [Ca2+]endolymph. Structural loss of the distal ES, which is consistently observed in Meniere’s disease, may therefore critically disturb [Ca2+]endolymph and contribute to the pathogenesis of Meniere’s disease. Electronic supplementary material The online version of this article (10.1007/s00441-019-03062-2) contains supplementary material, which is available to authorized users.


Introduction
Hair cell function in the cochlea and the vestibular system is calcium (Ca 2+ )-dependent (Tanaka et al. 1980;Ohmori 1985;Kozel et al. 1998) and requires the endolymphatic Ca 2+ concentration ([Ca 2+ ] endolymph ) to be tightly controlled at a level that is unusually low (0.017-0.133 mmol/l (Salt et al. 1989)) for an extracellular fluid (e.g., [Ca 2+ ] blood plasma = 1.2 mmol/l (Diem and Lentner 1970)). To maintain this physiological [Ca 2+ ] endolymph , numerous epithelial sites within the membranous labyrinth are engaged in Ca 2+ transport across the endolymph-perilymph barrier, utilizing Ca 2+permeable channels (e.g., transient receptor potential cation channel subfamily V members [TRPV channels] (Takumida et al. 2005;Wangemann et al. 2007;Nakaya et al. 2007;Ishibashi et al. 2008)), ion exchangers (e.g., sodium-calcium exchangers [NCX] (Oshima et al. 1997;Yamauchi et al. 2010)), and Ca 2+ ATPases (e.g., plasma membrane calcium ATPase [PMCA] (Crouch and Schulte 1995;Ågrup et al. 1999)). Many of these transport proteins are directly regulated downstream targets of the calcium sensing receptor (CaSR), a G-coupled cell surface receptor that is prominently expressed in the primary tissues involved in whole-body Ca 2+ homeostasis, including the kidney and the parathyroid and thyroid glands (Blankenship et al. 2001;Topala et al. 2009;Ranieri et al. 2013). In those tissues, CaSR senses changes in Ca 2+ levels in urine and plasma and modulates the activity and expression of Ca 2+ transport proteins to increase renal Ca 2+ reabsorption (Riccardi et al. 1998) and the Ca 2+ -mediated release of calciotropic hormones (parathyroid hormone, calcitonin) (Brown and MacLeod 2001), respectively. In the inner ear, the endolymphatic sac (ES), among other epithelial tissues, is a site of CaSR gene expression (Beitz et al. 1999;Lin et al. 2018) and critically contributes to endolymphatic Ca 2+ homeostasis. This function was first indicated by in vivo animal experiments in which surgical separation of the ES resulted in a nonphysiological increase in [Ca 2+ ] endolymph and endolymphatic hydrops development in the cochlea (Ninoyu and Meyer zum Gottesberge 1986;Salt and DeMott 1994). Structural loss of the distal (extraosseous) ES, as induced in this animal model, is a histopathological hallmark and a presumed etiopathogenic factor in Meniere's disease (MD) (Eckhard et al. 2019b).
Here, we investigated the immunolocalization patterns of CaSR and Ca 2+ transport proteins in the mature murine ES epithelium to elucidate the molecular determinants of Ca 2+ homeostasis in the ES and to better understand the pathophysiological consequences of the structural loss of the ES in MD.

Animals
All animal experiments were performed according to Swiss animal welfare laws and were approved by the local veterinary authorities (Kantonales Veterinäramt Zürich; protocol no. ZH269/16). Male, 6-to 8-weekold C57BL/J6-Crl1 mice were purchased from Charles River Laboratories, Inc. (Sulzfeld, Germany). A total of 12 animals were used in this study. All animals were kept at an in-house animal facility with free access to a standard chow diet and water under a standard 12 h light/dark cycle.
Tissue preparation, fixation, decalcification, embedding, and sectioning Animals were sacrificed by CO 2 inhalation. For tissue fixation, the thoracic cavity was opened, and a cannula was inserted in the left cardiac ventricle for transcardial perfusion with 10 ml phosphate buffered saline (PBS), followed by 10 ml of a fixative, i.e., (i) 10% neutral buffered formalin (F; Lucerna-Chem AG, Luzern, Switzerland), (ii) F with 1% acetic acid (FA; Sigma-Aldrich, Steinheim, Germany), (iii) F with 0.2% glutaraldehyde (GA; Sigma-Aldrich), or (iv) FA with 0.2% glutaraldehyde (FGA). The respective fixative used in each protocol is given in Table 1. Then, the animals were decapitated, soft tissues were removed from the outer surface of the skull using a sharp scalpel, and the cranial part of the skull was opened to remove most of the cerebrum, while the cerebellum was left in situ. The tympanic bullae were opened to allow faster penetration of the fixative into the inner ear. The specimens were then immersed in the respective fixative used for transcardial perfusion for 6 h on a countertop shaker. Afterwards, the specimens were dehydrated in a graded series of ethanol solutions (50%, 70%, 95%, and 100%), embedded in paraffin and sectioned at 4 μm using an HM 355S Automatic Microtome (Thermo Fisher Scientific, Waltham, MA, USA). Sections were collected on SuperFrost Plus slides (Thermo Fisher Scientific), dried on a heating plate at 37°C overnight and stored at room temperature.

Immunohistochemical DAB labeling
Sections were deparaffinized in Histo-Clear (National Diagnostics, Atlanta, GA, USA), rehydrated in a graded series of ethanol solutions (100%, 95%, 70%, and 50%), and rinsed in tap H 2 O. When required, heat-induced antigen retrieval (HIAR) with pressurized coverslipping of the mounted tissue sections was performed (Table 1) according to a previously This antibody recognizes an epitope between amino acids 724-783 of the human erythrocyte calcium pump established protocol (Eckhard et al. 2019a). All subsequent steps were performed at room temperature. Nonspecific binding was blocked with 1% normal horse serum (NHS) for 15 min, followed by incubation with primary antibodies overnight. The sections were then incubated with biotinylated secondary antibodies for 1 h, followed by incubation with avidinbiotin-HRP complex (Vectastain ABC HRP Kit, Vector Laboratories, Burlingame, CA, USA). Visualization was performed with 3,3′-diaminobenzidine (DAB) (DAB Peroxidase [HRP] Substrate Kit, Vector Laboratories). All incubation steps were performed at room temperature, and each incubation step was followed by rinsing the sections in PBS for 5 min. All primary and secondary antibody combinations utilized in this study are listed in Table 1. Negative controls for all immunolabeling experiments were obtained by omitting the primary antibody or after preabsorption of the primary antibodies with the corresponding commercially available control peptides (for primary antibodies against TRPV5/6, PMCA1/4, NCX2, and SERCA1/2). For nuclear counterstaining, the slices were incubated for 5 s with hematoxylin diluted in distilled water (1:6) and mounted with a permanent mounting medium (VectaMount, Vector Laboratories).

Immunofluorescence double labeling
Deparaffinization, rehydration of the sections, and incubation with primary antibodies was performed as described for DABlabeling experiments. All primary antibodies used for immunofluorescence double labeling are listed in Table 2. The sections were then incubated with biotinylated antibodies raised in donkey and directed against mouse (1:400; 715-065-151, Milan Analytica AG, Rheinfelden, Switzerland) and Alexa Fluor 594-conjugated antibodies raised in goat and directed against rabbit (1:400; 111-585-144, Milan Analytica AG) for 1 h, followed by incubation with avidin-biotin-HRP complex (Vectastain ABC HRP Kit, Vector Laboratories). Biotinylated tyramine (10 min) and a subsequent second incubation with avidin-biotin-HRP complex were used for signal amplification. Visualization was performed with Alexa Fluor 488conjugated streptavidin (1:800; Milan Analytica AG) and Alexa Fluor 594-conjugated anti-goat antibodies raised in donkey (1:400; 705-585-003, Milan Analytica AG). The sections were mounted with an aqueous mounting medium containing DAPI for nuclear counterstaining (H-1200, Vectashield, Vector Laboratories).

Microscopy analysis
Images were acquired using either a Leica DMI6000 microscope (Leica, Wetzlar, Germany) or a Leica SP5 confocal laser scanning microscope (Leica) and processed with Adobe Photoshop CS5 software (Version 12.0, Adobe Systems, San Jose, CA, USA). Quantification of immunolabeled ES epithelial cells DAB-labeled (DAB + ) ES epithelial cells were quantified on light microscopic (× 40 air objective) images of hematoxylincounterstained tissue sections. Cell counting was performed in at least three non-consecutive sections from two different axial planes of the intraosseous ES (iES) and the extraosseous ES (eES). Each section was divided into ten equally long segments along the longitudinal axis of the ES. The percentage of DAB + cells among the epithelial cells with hematoxylinstained nuclei was determined in each segment. Single-and double-immunofluorescence-labeled ES epithelial cells in the eES were quantified following double immunofluorescence labeling of CaSR with either V-ATPase, TRPV5, or γENaC. The percentages of fluorescence-labeled cells among the epithelial cells with DAPI-counterstained nuclei were determined. From each double immunofluorescence labeling experiment, at least 100 labeled cells were counted, and sections from at least 3 independent specimens were analyzed.

Descriptive statistics
Statistical analyses were performed using Prism software (version 7.0, GraphPad Software). Cell counts are expressed as percentages of the number of epithelial cells in a segment of

Results
Based on the previously demonstrated similarities between the ES epithelium and the kidney distal tubular epithelium on the cellular and molecular levels, we mainly focused our immunohistochemical analysis on Ca 2+ transport-associated proteins that are known to be crucial for the renal tubular handling of calcium.
Cellular immunolocalization patterns of the CaSR and Ca 2+ transport proteins DAB immunolabeling of the CaSR, as well as of TRPV5, TRPV6, SERCA1, SERCA2, NCX2, PMCA1, and PMCA4 was visible in the iES (proximal to the operculum) and eES (distal to the operculum; Fig. 1 and Supplementary Fig. 1) portions of the ES epithelium. Staining of proteins was absent or very weak in the endolymphatic duct (ED; Fig. 1 and Supplementary Fig. 1). Negative controls (omitted primary antibodies and preabsorption of primary antibodies with control peptides (TRPV5/6, NCX2, PMCA1/4, SERCA1/2) showed no specific staining in the murine ES epithelium.
CaSR immunolabeling was present in the iES and eES portions (Fig. 1a, b) and exhibited a scattered labeling pattern within the epithelium. In CaSR-labeled cells, strong cytoplasmic and membranous labeling was found, which in many cells was polarized to the basolateral membranes (Fig. 1c-c"). Immunolabeling signals for transient receptor potential cation channel subfamily V (TRPV) members TRPV5 and TRPV6 appeared to be cytoplasmic with strong polarization towards the apical cell pole, most clearly in epithelial cells in the eES   Supplementary Fig. 1A-A"). Homogeneous cytoplasmic immunolabeling of the sarcoplasmic/endoplasmic reticulum calcium ATPase (SERCA) subtypes SERCA1 and SERCA2 was found throughout the iES and eES epithelium (Fig. 1e-e" and Supplementary Fig. 1B-B") and was consistent with their known localization in the endoplasmic reticulum (Strehler and Treiman 2004). Labeling for the plasma membrane calcium ATPase (PMCA) subtypes PMCA1 and PMCA4, as well as for the sodium-calcium exchanger NCX2, was polarized in the basolateral membranes, most clearly in the eES (Fig. 1f-f" and Supplementary Fig. 1). No immunolabeling in the ES and ED was found for PMCA2 and NCX1, nor for the Ca 2+ -binding proteins Calbindin D-28k and parvalbumin ( Supplementary Fig. 2). All immunolabeling patterns and their cellular polarization in the ES epithelium were consistent with those previously described for the distal convoluted tubule and the connecting tubule epithelium (Loffing et al. 2001;Loffing and Kaissling 2003); positive control experiments are shown in Supplementary Fig. 3).

Spatial (proximal-to-distal) immunolabeling gradients
For each protein studied, DAB-labeled cells were quantified along the longitudinal (proximal-to-distal) axis of the ES Fig. 2 Quantification of DAB + cells along the ES reveals staining gradients along the ES epithelium increasing from the iES to the eES. The y-axis shows the percentage of DAB + cells relative to all ES epithelial cells. For cell counting, both the iES and eES were each divided into ten equal segments. The ticks on the x-axis correspond to each of the ten consecutive segments of the iES (left x-axis) and eES (right x-axis). The first two segments represent the endolymphatic duct epithelium. The values are the percentages of the total number of cells in the respective ES segment and are given as mean values ± SD from six animals. a Quantification of DAB + cells for CaSR and the apical Ca 2+ channels TRPV5 and TRPV6. b Quantification of DAB + cells for the cytoplasmatic proteins SERCA1 and SERCA2. c Quantification of DAB + cells for proteins located in the basolateral membrane epithelium, reaching from the proximal transition of the iES in the ED to the distal blind end of the eES. For all of the investigated proteins, a gradual increase in the immunolabeling signal in the proximal-to-distal direction was observed. At the very distal end of the eES portion, on average, more than 50% of the epithelial cells were positively labeled for most of the investigated proteins (Fig. 2).
CaSR immunolabeling within the murine inner ear is restricted to the ES epithelium No specific CaSR immunoreactivity was found in other portions of the membranous labyrinth, including the cochlea, saccule, utricle, and semicircular canal cristae (Fig. 4).

Discussion
In the present study, the ES epithelium was identified as an exclusive site of CaSR immunolocalization within the mature murine inner ear. Several CaSR-regulated Ca 2+ transport proteins (TRPV5, PMCA1/4), as well as non-CaSR-regulated Ca 2+ channels/transporters (TRPV6, NCX2), and intracellular Ca 2+ sequestering proteins (SERCA1/2) were localized in the ES epithelium, almost all of which exhibited their highest immunolabeling intensities in the distal eES portion.
Consistent with the immunolocalization of CaSR in MRCs of the mature murine ES shown here, previous gene expression studies detected CaSR mRNA expression in MRCs of the early postnatal murine ES (Honda et al. 2017), in the adult rat ES (Beitz et al. 1999), and, moreover, in the adult rat organ of Corti, stria vascularis, Reissner's membrane, and vestibulum (Beitz et al. 1999), as well as in neuromast hair cell stereocilia of zebrafish larvae (Lin et al. 2018). The results of the present study indicate that the CaSR protein in the mature murine inner ear is exclusively expressed in the ES or that immunolabeling at other cellular sites was below the detection level. We propose that the CaSR may have a unique role in sensing and regulating [Ca 2+ ] endolymph in the ES, with significance for the overall endolymphatic Ca 2+ homeostasis in the inner ear. However, it is unknown whether the CaSR enables ES epithelial cells to sense Ca 2+ changes in the ES endolymph, the cerebrospinal fluid, or in both extracellular fluid compartments that face the ES epithelium. Moreover, the molecular signaling cascades by which the CaSR may regulate downstream effector proteins, such as TRPV5 and PMCA1/4 (Blankenship et al. 2001;VanHouten et al. 2007;Topala et al. 2009), which are also localized in the ES epithelium (Ågrup et al. 1999;Honda et al. 2017;present study), require further exploration. Although CaSR function in the ES may contribute to maintaining Ca 2+ within its narrow physiological range in the entire inner ear, its effects on [Ca 2+ ] endolymph in the vestibular and cochlear portions are probably slow and may have physiological relevance primarily on longer time scales. This assumption is supported by previous observations that (i) longitudinal fluid movements (between endolymph compartments) virtually do not occur, at least under physiological conditions (Salt et al. 1986), and that (ii) surgical separation of the ES only leads to a slow rise of [Ca 2+ ] endolymph over a time course of several weeks (Ninoyu and Meyer zum Gottesberge 1986;Salt and DeMott 1994). This may be primarily due to local ion homeostatic mechanisms within the cochlea and the vestibular portions (Lundquist 1976;reviewed in Salt 2010), as well as anatomical diffusion barriers, such as the utriculosaccular duct, which (in the adult murine inner ear) largely prevents fluid exchange between the pars inferior (cochlea, saccule) and the pars superior (utricle, semicircular canals) (Cantos et al. 2000). In addition, the CaSR may have local homeostatic functions in the ES, such as maintaining ES epithelial integrity or mediating immune responses (Cheng et al. 2014).
Among the two main epithelial cell types in the ES, i.e., MRCs and RRCs, we found that the CaSR was exclusively localized in vATPase + MRCs (Dou et al. 2004;Honda et al. 2017). This cell-type-specific distribution of CaSR resembles that in intercalated cells in the renal cortical collecting duct (CCD; (Everett 2006)). On the subcellular level, CaSR was polarized towards both membrane domains (apical and basolateral) with additional strong cytoplasmic labeling. This finding is consistent with the distribution of CaSR in the distal renal tubule epithelium (distal convoluted tubule and collecting duct (Riccardi et al. 1998)), where CaSR regulates transepithelial Ca 2+ reabsorption from the tubular fluid via TRPV5 (Topala et al. 2009) and PMCA (Blankenship et al. 2001). Notably, other shared molecular mechanisms of transepithelial ion (Na + ) transport between the ES epithelium and the distal renal tubule epithelium have been reported (Mori et al. 2017;Eckhard et al. 2019b). In contrast to CaSR, TRPV5 was differentially localized in subsets of MRCs and RRCs, thereby differing from its exclusive localization in principal cells in the renal CCD (Loffing and Kaissling 2003). However, this non-celltype-specific TRPV5 localization in the ES is consistent with the RNA sequencing analysis of the murine ES (Honda et al. 2017). These and previously noted structural differences between the epithelia of the ES and the renal CCD (Wangemann and Marcus 2017), despite their numerous shared cellular and molecular features, may be due to the different physiological requirements for maintaining extracellular Ca 2+ homeostasis in the ES endolymph and the blood plasma. In addition to CaSRregulated Ca 2+ transporters, we identified TRPV6 and NCX2 in the ES epithelium, both of which are known to enable constitutively active Ca 2+ transport in the renal tubular epithelium (Loffing and Kaissling 2003).
Loss of the proposed Ca 2+ -homeostatic function of the ES may be of significance in various pathological conditions, such as otoconial disorders and MD. The biogenesis of otoconial Ca 2+ carbonate (CaCO 3 ) crystals is believed to occur in the ES luminal microenvironment and is sensitive to changes in [Ca 2+ ] endolymph . For example, disturbed formation of otoconia was observed in Foxi1 knockout (k.o.) mice and Efnb2 k.o. mice (Hulander 2003;Raft et al. 2014). In these models, the MRC population in the ES is developmentally lacking (Hulander 2003) or mildly decreased and proximally mislocalized in the ES (Raft et al. 2014), respectively. The loss of CaSR-regulated Ca 2+ homeostasis in the ES is one possible explanation for the compromised otoconia formation in these transgenic mouse models. Another explanation is the loss of (MRC-specific) pendrinand v-ATPase-mediated anion transport that leads to disturbances of endolymphatic pH, which in turn may inhibit the pH-sensitive TRPV5/6 channels in the ES epithelium and thereby disturb [Ca 2+ ] endolymph ), present study). In MD, the structural loss of the distal ES epithelium is a consistent histopathological hallmark. Presuming that the human ES harbors Ca 2+ transport mechanisms in proximalto-distal expression gradients analogous to the murine ES, structural loss of the distal ES-a pathological hallmark in MD (Eckhard et al. 2019b)-and in particular its Ca 2+ transport functions may be crucial in the pathogenesis of endolymphatic hydrops (Ninoyu and Meyer zum Gottesberge 1986;Salt and DeMott 1994).
In conclusion, the murine distal ES epithelium presumably has a "calciostatic" function for inner ear fluid homeostasis. Human pathology of the ES likely impairs Ca 2+ homeostasis of the inner ear fluids and is therefore potentially a significant pathophysiological factor in various inner ear disorders.
Open Access This article is distributed under the terms of the Creative Comm ons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.