Abstract
Cellular, molecular, and physiological studies have demonstrated an important signaling role for ATP and related nucleotides acting via P2 receptors in the cochlea of the inner ear. Signal modulation is facilitated by ectonucleotidases, a heterologous family of surface-located enzymes involved in extracellular nucleotide hydrolysis. Our previous studies have implicated CD39/NTPDase1 and CD39L1/NTPDase2, members of the ectonucleoside triphosphate diphosphohydrolase (E-NTPDase) family, as major ATP-hydrolyzing enzymes in the tissues lining the cochlear endolymphatic and perilymphatic compartments. NTPDase1 hydrolyzes both nucleoside triphosphates and diphosphates. In contrast, NTPDase2 is a preferential nucleoside triphosphatase. This study characterizes expression of these E-NTPDases in the mouse cochlea by immunohistochemistry. NTPDase1 can be immunolocalized to the cochlear vasculature and neural tissues (primary auditory neurons in the spiral ganglion). In contrast, NTPDase2 immunolabeling was principally localized to synaptic regions of the sensory inner and outer hair cells, stereocilia and cuticular plates of the outer hair cells, supporting cells of the organ of Corti (Deiters’ cells and inner border cells), efferent nerve fibers located in the intraganglionic spiral bundle, and in the outer sulcus and root region of the spiral ligament. This differential expression of NTPDase1 and 2 in the cochlea suggests spatial regulation of P2 receptor signaling, potentially involving different nucleotide species and hydrolysis kinetics.
T
Ectonucleotidases play a major role in modulation of P2 receptor-mediated effects by hydrolyzing extracellular nucleotides. A large number of nucleotide-hydrolyzing enzymes have been cloned from mammalian tissues, including members of the ectonucleoside triphosphate diphosphohydrolase (E-NTPDase) family that hydrolyze nucleoside 5′-tri- and -diphosphates. Three members of this family (NTPDase1–3) are expressed in the mammalian brain (Zimmermann and Braun 1999). Their membrane topography includes a large extracellular domain containing five apyrase conserved regions implicated in ATP hydrolysis and two transmembrane domains at the N- and C-terminals (Zimmermann 2001). E-NTPDase activity is coupled with ecto-5′-nucleotidase activity to produce adenosine that exerts its effect through P1 receptors.
ATP-hydrolyzing activity in the cochlea is largely mediated by E-NTPDases (Vlajkovic et al. 1996,1998a,b), NTPDase1 and NTPDase2 in particular (Vlajkovic et al. 1999). NTPDase1 has comparable catalytic activity for ATP and ADP, whereas NTPDase2 is a preferential ATPase (Zimmermann 2001). These enzymes are likely the key elements for termination of P2 receptor signaling in the cochlea. NTPDases may prevent receptor desensitization by the hydrolysis of extracellular nucleotides (Enjyoji et al. 1999), and NTPDase1, jointly with ecto-5′-nucleotidase, facilitates the salvage of adenosine via nucleoside transporters (Plesner 1995; Sen et al. 2001). The aim of this study was to provide immunocytochemical localization of NTPDase1 and NTPDase2 expression in mouse cochlear tissues, extending the evidence for tissue-specific modulation of P2 receptor signaling.
Materials and Methods
Animals
Experiments described in this study were approved by the University of Auckland Animal Ethics Committee. Adult Swiss white mice (2–3 months old) were deeply anesthetized with Nembutal (Virbac Laboratories; Auckland, NZ), and perfused transcardially with cold 4% paraformaldehyde (PFA; Scharlau Chemie, Barcelona, Spain) in 0.1 M phosphate-buffered solution (PB: Na2HPO4 77.5 mM; NaH2PO4 22.5 mM, pH 7.4). The cochleae were removed and post-fixed in 4% PFA overnight at 4C.
Immunohistochemistry
NTPDase1 and NTPDase2 antibodies were raised in rabbits by direct inoculation of the encoding cDNA in pcDNA3 (Enjyoji et al. 1999). The specificity of these antibodies was previously verified using cytochemistry and Western blots on transfected cells and their protein extracts, respectively (Braun et al. 2000; Heine et al. 2001; Sévigny et al. 2002). An additional control for specificity was immunohistochemistry of tissues from NTPDase1-null mice (Enjyoji et al. 1999; Braun et al. 2000). These studies showed no crossreactivity between NTPDase1 and 2 antibodies.
Mouse cochleae fixed in 4% PFA were decalcified in 5% EDTA/PB solution for 7 days, then cryoprotected overnight in 30% sucrose/PB solution. The sections were then rinsed in 0.1 M PB, snap-frozen in isopentane at −80C, and cryosectioned at 20 μm. Sections were placed in 48-well plates (Nalge Nunc; Naperville, IL), rinsed with sterile 0.1 M PBS (Na2HPO4 77.5 mM; NaH2PO4 22.5 mM; NaCl 154 mM, pH 7.4), and permeabilized with 1% Triton X-100 for 1 hr. Nonspecific binding sites were blocked with 2% non-fat milk powder and 1.5% normal goat serum (Vector Laboratories; Burlingame, CA). Primary antibodies to NTPDase1 or NTPDase2 were applied at a titer of 1:5000 in 0.1 M PBS containing 1.5% normal goat serum (Vector) and 0.1% Triton X-100 and incubated overnight at 4C. Vectastain Elite ABC kit (Vector) containing a secondary biotinylated goat anti-rabbit antibody and avidin–biotin–peroxidase complex was used for detection of NTPDase immunoreactivity. Incubations with the secondary antibody and avidin–biotin complex were done for 40 min at room temperature (RT). The reaction was visualized using diaminobenzidine (DAB; Vector) as a chromogen. Negative control sections were incubated with the preimmune rabbit serum (titer 1:5000) overnight at 4C. Brain tissue was used as a positive control as described by Braun et al. (2000) (data not shown). The cochleae were photographed on positive transparency color film (Fujichrome Provia 400F) using a microscope with Nomarski differential interference contrast optics (Zeiss Axioskop; Thornwood, NY) and the images were scanned on a ScanMaker 35t plus (Microtek Lab; Redondo Beach, CA).
E-NTPDase immunoperoxidase histochemistry. (
E-NTPDase confocal immunofluorescence. (
Confocal Immunofluorescence
Laser scanning confocal microscopy of whole mounts of cochlear tissues was used to provide high resolution of immunostaining. After fixation in 4% PFA, cochlear tissues were microdissected from the modiolus and rinsed in 0.1 M PBS. The blocking and permeabilization steps were the same as described above for immunohistochemistry. Primary NTPDase1 and NTPDase2 antibodies (titer 1:5000) were applied overnight at 4C. The tissues were then incubated with Alexa 488 goat anti-rabbit secondary antibody (Molecular Probes, Eugene, OR; dilution 1:200) for 2 hr at RT. The tissues were rinsed several times in PBS, mounted in fluorescence medium (DAKO; Carpinteria, CA), and screened for NTPDase1- or NTPDase2-specific immunofluorescence using a confocal microscope (TCSAD; Leica Leisertechnik, Heidelberg, Germany).
Results
NTPDase1 Immunolocalization
The most prominent NTPDase1 immunoreactivity was observed in the cochlear vasculature (Figure 1A). Immunopositive blood vessels were detected in the spiral ligament, stria vascularis, spiral limbus, osseous spiral lamina, and Rosenthal's canal (Figures 1A, 1C, and 1D). NTPDase1-specific immunostaining appeared to be associated with endothelial cells/smooth muscle cells of the radiating arterioles and endothelial cells of the capillaries (Figure 2B). No immunoreactivity was observed in the sensory and supporting cells of the organ of Corti (Figures 1B and 2A). NTPDase1-specific immunoreactivity was also detected in the perikarya of the spiral ganglion neurons (SGNs) (Figure 1D). In the spiral ligament, a narrow zone of fibrocytes bordering the otic capsule consistently showed immunolabeling (Figure 1C), whereas Reissner's membrane, which separates scala media from scala vestibuli (not shown), was unstained. Nonspecific staining was observed in the tectorial membrane and osseous spiral lamina in control sections stained with the preimmune serum (Figure 1A inset). Staining in the acellular tectorial membrane, attributable to binding to collagen fibers, is frequently observed in immunocytochemical studies in the cochlea (Altschuler et al. 1985).
NTPDase2 Immunolocalization
NTPDase2 immunoreactivity was differentially distributed in the tissues bordering scala media. It was observed in the organ of Corti (Figure 1E) but was notably absent from Reissner's membrane (not shown) and marginal cells of the stria vascularis (Figure 1G). Strong immunostaining was observed in both inner and outer hair cell regions. Immunoreactivity in the reticular lamina corresponds to the cuticular plates and stereocilia of the outer hair cells, but not their lateral walls (Figures 1F, 2C, and 2D). The pronounced punctate staining beneath the outer hair cells is most likely attributable to labeling of efferent nerve endings, based on their size and number. Supporting Deiters’ cells that form cups underneath the outer hair cells showed moderate immunostaining. Apart from the inner hair cells, strong immunolabeling in the inner hair cell region may also include supporting inner border cells. Intense NTPDase2 immunoreactivity was detected in the inner spiral plexus and outer spiral bundle regions, which provide afferent innervation to the inner and outer hair cells, respectively (Figure 1F). Nerve fibers projecting through the osseous spiral lamina to these regions were stained immunopositive with NTPDase2 antibody (Figure 1E). As for the lateral wall tissues, strong immunostaining was found in the spiral prominence, outer sulcus cells, and their root processes, which are intimately associated with the blood vessels of the spiral ligament (Figures 1G and 2E). Spiral ganglion cell bodies were not stained (compare with NTPDase1 labeling), whereas the efferent nerve fibers of the intraganglionic spiral bundle showed intense immunoreactivity (Figure 1H). Apart from the tectorial membrane, control sections stained with the preimmune serum (Figures 1F, 1G, and 1H insets; Figure 2F) lacked immunolabeling.
Discussion
This study demonstrates an extensive NTPDase1 and NTPDase2 distribution in mouse cochlear tissues, consistent with the postulated role for E-NTPDases in setting the level and kinetics of neurohumoral nucleotide signaling in the inner ear (Vlajkovic et al. 1998a, b, 1999). The expression of E-NTPDases in the cochlear vasculature, hair cell stereocilia, spiral ganglion neurons, and their dendritic processes that form afferent synapses with hair cells indicates a likely involvement of E-NTPDases in the regulation of blood flow, sound transduction, and auditory neurotransmission, all of which are affected by P2 receptor signaling (Housley 1998,2000,2001; Housley and Thorne 2000).
The differential expression of NTPDase1 and 2 in vascular, sensory, neural, and secretory tissues is consistent with the wide range of proposed actions of extracellular nucleostides in the cochlea (Housley 2000,2001), and implies that nucleotide di- and triphosphates may play different physiological roles at these sites.
The most prominent NTPDase1 immunostaining was observed in blood vessels of the stria vascularis and spiral ligament. Abundant NTPDase1 immunoreactivity in the cochlear vasculature was comparable to NTPDase1 signal levels in blood vessels of the mouse brain (Braun et al. 2000). Extracellular ATP appears to be a potent modulator of cochlear blood flow, acting via P2X and P2Y receptors (Ren et al. 1997; Muñoz et al. 1999). By regulating the concentrations of extracellular nucleotides, NTPDase1 may effectively control P2 receptor influence on cochlear blood flow and also ADP-mediated platelet activation (Kaczmarek et al. 1996; Enjyoji et al. 1999).
Both E-NTPDases were also present in neural tissues of the cochlea. NTPDase1 was strongly expressed by the spiral ganglion cell bodies, whereas NTPDase2 immunostaining was more prominent at the dendritic processes of the spiral ganglion neurons providing afferent innervation to the inner and outer hair cells. Co-localization of E-NTPDases and the P2X2 receptor at the SGN cell bodies and their afferent synapses with the inner and outer hair cells (Salih et al. 1998; Housley et al. 1999; Järlebark et al. 2000) imply a functional role of ATP in hair cell synaptic transmission. In addition, NTPDase2-specific immunostaining was observed in the efferent fibers of the intraganglionic spiral bundle, composed of axons arising in the superior olivary complex.
The present study demonstrates strong NTPDase2 expression at the outer hair cell stereocilia extending to the cuticular plates. In contrast, NTPDase1 was absent from the organ of Corti. This co-localization of NTPDase2 and P2X2 receptor (Housley et al. 1998a, b, 1999; Järlebark et al. 2000) at the hair cell stereocilia underlines the possible direct involvement of P2 receptor signaling in sound transduction.
The NTPDase2 immunolabeling of Deiters’ cells shown here further demonstrates P2 receptor signal termination mechanism, because P2X and P2Y receptor expression has previously been reported for these cells (Chen and Bobbin 1998; Housley et al. 1999; Matsunobu and Schacht 2000). Activation of P2 receptors leads to prolonged elevation of intracellular Ca2+ and concomitant change in the stiffness of the cells (Chen and Bobbin 1998). Deiters’ cells provide protective dampening during acoustic overstimulation (Fridberger et al. 1998), and that effect may be mediated by extracellular ATP (Bobbin 2001). By regulating the concentrations of extracellular nucleotides, NTPDase2 can provide fine tuning of the Deiters’ cell response to ATP under stressor conditions.
A number of studies have indicated that extracellular ATP levels in the cochlear endolymphatic space regulate hearing sensitivity through activation of P2X receptors that reduce the electrochemical driving force for sound transduction via an electrical shunt across the cochlear partition (Housley and Thorne 2000). This model is fully supported by NTPDase2 expression in the spiral prominence, outer sulcus cells, and their root processes in the spiral ligament, which is a major pathway for K+ cycling between the scala media and scala vestibuli (Marcus and Chiba 1999; Wangemann 2002). These sites also exhibit high P2X2 receptor expression (Järlebark et al. 2000). NTPDase activity could also modulate the endocochlear potential by hydrolyzing extracellular ATP required for P2X2 receptor-mediated stimulation of cation absorption by outer sulcus cells (Lee et al. 2001).
In conclusion, this study provides a description of NTPDase1 and NTPDase2 distribution in mouse cochlea. The results indicate tissue-specific actions of NTPDase1 and 2, which are well placed to provide spatial regulation of P2 receptor signaling in the cochlear endolymphatic compartment.
Footnotes
Acknowledgements
Supported by the Auckland Medical Research Foundation, Deafness Research Foundation of New Zealand, and Health Research Council of New Zealand. JS was supported by the American Liver Foundation and the Canadian Institutes of Health Research and SCR by NIH, USA.
We thank Dr Norbert Braun and Professor Herbert Zimmermann (AK Neurochemie, Biozentrum der J.W. Goethe-Universität, Frankfurt am Main, Germany) for their kind gift of the NTPDase2 plasmid used for generation of the NTPDase2 antibody.