Hearing Research 326 (2015) 40e48
Contents lists available at ScienceDirect
Hearing Research journal homepage: www.elsevier.com/locate/heares
Research paper
Cell-specific accumulation patterns of gentamicin in the guinea pig cochlea Ulf-Rüdiger Heinrich a, 1, Irene Schmidtmann b, 1, Sebastian Strieth a, Kai Helling a, * a b
Department of Otorhinolaryngology, Head and Neck Surgery, University Medical Center of the Johannes Gutenberg University Mainz, Germany Institute for Medical Biostatistics, Epidemiology and Informatics, University Medical Center of the Johannes Gutenberg University Mainz, Germany
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 January 2015 Received in revised form 19 March 2015 Accepted 20 March 2015 Available online 13 April 2015
nie re's disease (MD) Intratympanic gentamicin therapy has become a popular treatment modality for Me through controlled elimination of vertigo spells caused by the balance organ. However, the known ototoxic properties of aminoglycosides lead to cochlear damage. In order to gain more information about cellular preferences for aminoglycoside accumulation within the cochlea, gentamicin was immuno histochemically localized by light microscopy in male guinea pigs 1 and 7 days after intratympanic application (n ¼ 8 ears/incubation time). Differences in the gentamicin-specific cellular storage capacities were quantified by determination of the local immuno staining intensities. Gentamicin was detected in every cochlear cell type, but with spatiotemporal variability. One day after application, an intense staining reaction was found in all cell types except the spiral ganglion cells and the stria vascularis. Six days later, gentamicin staining intensities were additionally reduced in the nerve fibers and the spiral ligament. Statistic analysis revealed strong cellular associations in respect to aminoglycoside accumulation. Furthermore, associations with recorded hearing losses were identified comparing the cellular gentamicin content in the organ of Corti, in the stria vascularis, in the spiral ganglion cells and in fibrocytes of the Limbus. In the lateral wall, clear differences in cellular gentamicin accumulation were found between type I fibrocytes of the spiral ligament compared with basal and intermediate cells of the stria vascularis. This finding was unexpected as these three cell types belong to a well-developed gapjunction system which normally enables unhampered cell communication. Cellular differences in local gentamicin storage capacities, transport processes and inherent diffusion barriers are discussed. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Intratympanic gentamicin therapy is currently a successful nie re's disease approach in the treatment of vertigo spells in Me (MD) despite its known cochleotoxic properties (Blakley, 2000; Forge and Schacht, 2000; Lange et al., 2004; Helling et al., 2007). Its efficiency is based on differences in the vulnerability of hair cells in the labyrinth and cochlea (Nakashima et al., 2000). Currently, a low dose, single-shot therapy seems to be the most promising approach for MD treatment (Lange et al., 2004; Helling et al., 2007). Typically, a concentration of 10e40 mg/mL gentamicin is
* Corresponding author. Department of Otorhinolaryngology, Head and Neck Surgery, University Medical Center of the Johannes Gutenberg University Mainz, Langenbeckstrasse 1, 55131 Mainz, Germany. Tel.: þ49 6131 177361; fax: þ49 6131 176637. E-mail address:
[email protected] (K. Helling). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.heares.2015.03.010 0378-5955/© 2015 Elsevier B.V. All rights reserved.
intratympanically applied in order to eliminate the vertigo spells (Blakley, 2000). However, even in low-dose therapy, cochlear damage cannot be avoided. Biochemically, the formation of reactive oxygen and/or reactive nitrogen species (ROS and RNS) have been widely identified after the application of gentamicin (Hirose et al., 1997; Sha and Schacht, 1999; Heinrich et al., 2008; Choung et al., 2009). In addition, the induction of the inducible nitric oxide synthase (iNOS) (Takumida et al., 1999; Liu et al., 2008) and the up-regulation of the constitutively expressed endothelial nitric oxide(eNOS)-isoform (Heinrich et al., 2006) were found in the cochlea which both contributed to an nitric oxide(NO)-increase. Subsequently, the induction of numerous cell-death cascades were described, such as the activation of caspases and of Jun N-terminal kinases (JNKs) (Wang et al., 2003; Okuda et al., 2004; Jeong et al., 2010). Morphologically, gentamicin-induced cochlear damage was identified at the electron microscopic level (Heinrich et al., 2006) and was found to be strongly correlated with the extent of local
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
List of abbreviations AU BC BV CLIMP EDTA FL eNOS GBP HSP IC ID IHC IS iNOS
arbitrary unit basal cells blood vessel cytoskeleton-linking membrane protein ethylenediaminetetraacetic acid fibrocytes of the Limbus endothelial nitric oxide synthase gentamicin-binding protein heat shock protein intermediate cell interdental cell inner hair cell intrastrial fluid-filled compartment inducible nitric oxide synthase
gentamicin uptake (Imamura and Adams, 2003). Physiologically, gentamicin-dependent alterations were identified in the guinea pig model, resulting in a hearing threshold shift beginning on the second day after application concomitant with an increase in NOproduction in the lateral wall (Heinrich et al., 2008). Currently, it is assumed that aminoglycosides penetrate nearly all cell types of the cochlea (de Groot et al., 1990; Imamura and Adams, 2003) and that the most damaging step in cochlear malfunction is an exclusive and preferential drug uptake into inner ear hair cells (Xie et al., 2011). Otherwise, it was also proven that gentamicin-accumulation varied markedly between individual cell types and even within the same cell type when comparing individual animals (Imamura and Adams, 2003). Based on the findings of cochlear degeneration ranging from mild to severe damage, it can be hypothesized that this variability in toxic response is partly based on individual differences in drug accumulation capacity (Imamura and Adams, 2003). In order to obtain more information about possible cellular differences in gentamicin-accumulation patterns within the cochlea, aminoglycoside staining intensities were quantified for seven different cochlear regions by a computer program one and seven days after application. The associations of cellular storage capacities were investigated at the individual animal level. 2. Materials and methods 2.1. Subjects Fourteen healthy pigmented guinea pigs (tricolor, Charles River, Sulzfeld, Germany) weighing 200e250 g with good Preyer's reflexes and no evidence of middle ear disease were used in this study. All experiments were conducted in accordance with the German Prevention of Cruelty to Animals Act and were approved by the supervising authorities. Animals were kept on a 12:12 h light:dark cycle in the university's animal facility. 2.2. Gentamicin application A volume of 0.1 mL (4 mg/ear/animal) gentamicin (Ratiopharm, Ulm, Germany) was injected through the anterior parts of the tympanic membrane in both ears of the experimental animals (n ¼ 8) under microscopic control. After gentamicin treatment, 4 animals were killed after one day, and 4 animals after seven days. Six untreated animals served as controls.
JNK MC MD NF NO OC OHC PBS RGB RNS ROS SD SGC SL SV
41
Jun N-terminal kinase marginal cells nie re's disease Me nerve fibers nitric oxide organ of Corti outer hair cell phosphate buffered saline red-green-blue reactive nitrogen species reactive oxygen species standard error spiral ganglion cells spiral ligament stria vascularis
2.3. Cochlea preparation Animals were killed by pentobarbital e sodium (Narcoren®, Hallbergmoos, Germany; 448 mg/kg body weight) on the first or on the seventh day after gentamicin injection, respectively. Both bullae were completely removed and transferred into a solution that consisted of 0.2% picric acid, 4% para-formaldehyde and 0.1% glutardialdehyde. After decalcification with ethylenediaminetetraacetic acid (EDTA) for 3 weeks at 4 C, the cochleae were dehydrated by an increasing ethanol series followed by xylene. Specimens were embedded in paraffin. Five mm sections were prepared using a microtome (Leica RM 2165, Leica Microsystems GmbH, Wetzlar, Germany), mounted onto superfrost glass slides and deparaffinated by xylene and a decreasing alcohol series. Endogenous peroxidase was blocked by immersing the slides in 3% H2O2/methanol. After pre-incubation with 10% normal serum and 1% bovine serum albumin in phosphate buffered saline (PBS) for 20 min to avoid unspecific binding, the primary mouse polyclonal anti-gentamicin antibody (Fitzgerald Industries, MA, diluted 1:450) was overlaid overnight at 4 C. Slides were then consecutively incubated with biotinylated secondary antibody (1:250, DAKO, Hamburg, Germany) for 30 min, streptavidine peroxidase (1:200, DAKO) for 30 min, and finally Diaminobencidine/H2O2 (1.85 mM) for 1 min. All washing procedures were performed in PBS; dilutions of antibodies were prepared in PBS containing 1% bovine serum albumin at room temperature. For negative controls, the primary antibodies were omitted. In addition, control specimen which had not received gentamicin did not reveal any immuno staining reaction. All assessments were performed blinded with respect to treatments. 2.4. Quantification of immuno cytochemical staining Seven cochlear regions were analyzed in detail on the paraffin sections: the organ of Corti, stria vascularis, spiral ligament, interdental cells, fibrocytes of the Limbus-area, nerve fibers, and spiral ganglion cells. In all regions, the immuno staining intensities were determined as described recently (Heinrich et al., 2010). Briefly, images were taken from paraffin sections using a high density 1/3 type, three-chip Exwave HAD CCD red-green-blue (RGB) color video camera (Sony 3-CCD DXC-390P, Sony Corp Tokyo, Japan) connected to a ZEISS microscope (Axiovert 200, Carl Zeiss Microscopy GmbH, Jena, Germany) equipped with a halogen light source using a 40 objective. The illumination intensity of the specimens was assured by precise voltage control.
42
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
The CCD camera settings were kept constant for all analyses. All images were analyzed in Photoshop (version 7; Adobe Systems, San Jose, CA) and stored in a single file. Images of the gentamicin treated animals were stored in one single image file merging the levels of the different samples together. Areas with the same brown colors (tolerance level 8 of the computer program) were selected with the cursor of the Magic Wand tool and quantified using the Histogram command from the Image menu. Firstly, the background staining values (values without any cellular structure) were subtracted from the immuno staining intensities to the cells resulting in the “net staining intensity”. Secondly, the area of the stained tissue (given in pixel) was compared with the size of the whole analyzed section (given in pixel) and determined as a “percentage of the immuno stained area”. The average “net staining intensity” was multiplied by the “percentage of the immuno stained area” and expressed in arbitrary units (AU) for the different cell types. 2.5. Statistical approaches In this study, a linear mixed model was used for statistical analysis, taking into account that different regions within an ear might be dependent by including a random animal effect. When comparing exclusively the single cell types one and seven days after gentamicin application, analysis was stratified by cell type. In this model, gentamicin application, cochlea turn, treatment and pairwise interactions between these were included as fixed effects. When comparing the three cell types within the stria and the fibrocytes, cell type, treatment and cochlea turn were included as fixed effects. As the analyses were exploratory, no corrections for multiple comparisons were applied. The values must be regarded as descriptive. The mean values and standard errors (SD) are presented in the graphs and the P-values are indicated by one, two or three asterisks representing p < .05, p < .01 and p < .001. Associations of gentamicin staining intensities between different regions of the cochlea are described by Pearson correlation coefficients and corresponding p-values (testing whether the correlation is different from zero). The same statistical approach was applied when testing associations between hearing loss and gentamicin staining intensity different cochlea regions. Statistical analysis was performed using PROC MIXED from SAS 9.3 (SAS Institute Inc., Cary, NC.). 3. Results 3.1. Immuno histochemical localization of gentamicin For a better illustration of identified cellular differences, the general morphology of the cochlea with the main regions and cell types is presented below in a schematic drawing of a section through the second cochlear turn (Fig. 1). The corresponding paraffin sections demonstrate exemplarily local staining intensities (Fig. 1a). A clear anti-gentamicin staining reaction was seen in the spiral ligament but was found to be reduced or often absent in the nearby stria vascularis (Fig. 1b). In the Limbus-area, gentamicin was detected in the interdental cells, in the fibrocytes and in the nerve fibers of the osseous spiral lamina (Fig. 1c). In the organ of Corti, clear gentamicin-staining intensities were not only visible in the inner and outer hair cells but also in the supporting cells (Fig. 1d). 3.2. Time-dependent alterations in gentamicin staining intensities The average staining intensities were determined for seven cochlear regions one day after gentamicin injection and compared with the corresponding staining intensities on the seventh day
after application. After one day, a relatively high amount of gentamicin (represented by the dark-grey colored columns) was found in the nerve fibers, the fibrocytes of the Limbus, the interdental cells, the organ of Corti and the spiral ligament (Fig. 2). A relatively low amount of gentamicin (represented by the lightgrey colored columns) was measured in spiral ganglion cells and in the stria vascularis (Fig. 2). Seven days after application, the gentamicin staining intensities were roughly comparable with the values of the first day in the fibrocytes of the Limbus, the interdental cells and the organ of Corti. However, in contrast to the one-day-exposed group, markedly reduced gentamicin staining intensities were now seen in the spiral ligament (p ¼ .017) and in the nerve fibers (p ¼ .026). The content of gentamicin in the spiral ganglion cells and the stria vascularis was unaltered and low. The spatiotemporal differences in gentamicin accumulation are presented by schematic figures using the same grey shadings as in the corresponding bars (Fig. 2). No effects of the cochlea turn were found in any of the cell types. 3.3. Inter-animal variability in gentamicin accumulation within the cochlear The determination of the average gentamicin staining intensities at the level of the single ear revealed a wide variation (Fig. 3). One day after gentamicin application, a range between 1809 and 8515 arbitrary units (AU) was detected with a mean value of 5140 AU. Seven days after application, values between 1166 and 7223 AU were found with a mean value of 3438 AU. In respect to the great variability, there was no statistical difference between the two groups (p ¼ .115). Nevertheless, the existence of the variability has to be kept in mind when the success of therapeutic interventions is discussed. 3.4. Local differences in gentamicin accumulation within the lateral wall The most prominent differences were found correlating gentamicin accumulation with cell types (p < .0001), and with cell-type specific treatment effects (p < .0001 for cell type*treatment interaction). No differences were found regarding gentamicin accumulation and cochlea turns (p ¼ .1989), but treatment effects exhibited some difference with regard to cochlea turns (p ¼ .0009). A detailed analysis of gentamicin accumulation in the lateral wall revealed clear differences between type I fibrocytes located in the spiral ligament at the border to the stria vascularis and the three cell types of the stria vascularis: basal cells, intermediate cells and marginal cells (Fig. 4a). One day after gentamicin application, the content in the three strial cell types was considerably lower compared with the content in the fibrocytes (p < .0001). In addition, a difference in gentamicin content was found between the basal and marginal cells (p < .001). Seven days after gentamicin application, a similar pattern in accumulation was found among type I fibrocytes and the three strial cell types already identified for the one-day-treated ears (p < .001; Fig. 4b). Furthermore, there was a clear reduction in gentamicin staining intensities when comparing basal cells and intermediate cells (p ¼ 0.0036) and basal cells and marginal cells (p < .001). A difference was also found between intermediate cells and marginal cells (p ¼ .0123). Based on numerous published findings in literature, a welldeveloped gap-junction system is assumed in this cochlear region which connects the fibrocytes of the spiral ligament with the basal and intermediate cells in the stria vascularis. The various cell types which are involved in this communication system are marked in
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
43
Fig. 1. Distribution of anti-gentamicin immunoreaction in the cochlea of the guinea pig. a.) schematic drawing of the three fluid-filled compartments (Scm Scala media, Sct Scala tympani, Scv Scala vestibule) and the seven regions analyzed; b.) lateral wall with a clear gentamicin staining intensity in the spiral ligament and no staining in the stria vascularis; c.) Limbus-area with clear anti-gentamicin antibody reaction in the fibrocytes, the interdental cells and the nerve fibers; d.) intense gentamicin staining intensity in all cell types in the organ of Corti. FL fibrocytes of the Limbus, ID interdental cells, IHC inner hair cell, NF nerve fibers, OC organ of Corti, OHC outer hair cell, SL spiral ligament, SV stria vascularis, SGC spiral ganglion cells scale bar ¼ 30 mm.
Fig. 2. Mean gentamicin specific staining intensities (given in arbitrary units (AU; mean values with SD) of both experimental groups (1d and 7d) in spiral ganglion cells (SGC), nerve fibers (NF), fibrocytes of the Limbus (FL), interdental cells (ID), organ of Corti (OC), stria vascularis (SV), spiral ligament (SL). A statistic significance between the two time points was only found in nerve fibers and spiral ligament marked by asterisks (P < .05*). The cochlear regions with high amounts of gentamicin are marked by dark-grey columns, the areas with low amounts of gentamicin are visualized by light-grey columns. The corresponding drawings underneath reveal the spatiotemporal reduction of the gentamicin content. One day after application, low gentamicin levels were only found in the spiral ganglion cells and in the stria vascularis. Seven days after application, the gentamicin content was also reduced in nerve fibers and spiral ligament.
44
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
3.5. Cellular association of gentamicin accumulation
Fig. 3. Inter-animal variability in gentamicin accumulation within the cochlea one and seven days after application. The values of each single ear are presented by dots and the mean values by bars.
The statistical analysis revealed multiple associations with respect to the gentamicin immuno staining intensities by comparing the various cochlear regions one (Fig. 6a) and seven days (Fig. 6b) after aminoglycoside application. One day after intratympanic gentamicin application, strong correlations were detected when comparing the nerve fibers and the following: the organ of Corti, the interdental cells, the stria vascularis, the fibrocytes in the Limbus-area (Fig. 6a). Furthermore, there was a correlation between the interdental cells and the fibrocytes of the Limbus. Scatterplots with corresponding correlation coefficients and p-values are presented in Fig. 6a highlighting the strongest associations. Strong associations of spatial relations are visualized in Fig. 6c. For a better understanding, a drawing of the anatomy of the cochlea has been added (Fig. 6d). Similarly, the results seven days after application are presented in Fig. 6b and e. Here, strong correlations were found between the organ of Corti and the interdental cells (r ¼ 0.97, p < .001) or the fibrocytes of the Limbus (r ¼ 0.89, p ¼ .0029). A clear association (r ¼ 0.86, p ¼ .005) was also found between the interdental cells and the fibrocytes. 3.6. Associations of regional cellular gentamicin accumulation and hearing loss
dark-grey in Fig. 5a. However, based on the clear differences detected in cellular gentamicin accumulation between the fibrocytes and the strial cells, a diffusion barrier for the polycationic gentamicin must be assumed between fibrocytes and basal cells (Fig. 5b).
In order to identity those cell types which mainly contribute to hearing loss, the gentamicin concentrations in the seven cochlear regions were compared with recorded hearing losses at the level of the individual ear. The associations were determined at the first
Fig. 4. Mean gentamicin specific staining intensities (given in arbitrary units (AU; mean values with SD) in type I fibrocytes of the spiral ligament, and the three strial cell types, the basal, intermediate and marginal cells. a.) One day after gentamicin application, clear differences in accumulation were found between fibrocytes and the three strial cell types. There was also a clear difference in gentamicin accumulation between basal cells and marginal cells. b.) Seven days after gentamicin application, there were again clear differences between gentamicin staining intensities in fibrocytes and the three strial cells. Furthermore, differences were found between basal cells and intermediate cells and marginal cells and between intermediate and marginal cells.
Fig. 5. Proposed and detected diffusion barriers in the lateral wall: a.) It is widely accepted that there is a gap-junction system between fibrocytes, basal and intermediate cells which allows the free passage of molecules (communication system marked in grey); b.) Based on the present findings it must be assumed that there is a diffusion barrier for the polycationic gentamicin between fibrocytes and basal cells. BC basal cells, BV blood vessel, FI fibrocyte, IC intermediate cell, IS intrastrial fluid-filled compartment, MC marginal cells.
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
45
Fig. 6. Spatial associations of the gentamicin accumulation one and seven days after application in the seven regions analyzed. At the two time points, numerous strong associations were detected by statistical analysis (marked by dark-grey background, in each cell first the Pearson correlation coefficient, then the corresponding p-value and finally the number of observations was given). a.) One day after application, strong correlations were found between nerve fibers and organ of Corti, the stria vascularis, the interdental cells or the fibrocytes of the Limbus. In addition, a strong association was also detected between interdental cells and fibrocytes. b.) Seven days after application, strong correlations were seen between the organ of Corti and interdental cells or fibrocytes. Furthermore, there was a clear association between interdental cells and fibroytes. The corresponding drawings underneath reveal associations observed after one day (c) and after 7 days (e) and their spatial distribution in the cochlea (d). FL fibrocytes of the Limbus, ID interdental cells, NF nerve fibers, OC organ of Corti, SGC spiral ganglion cells, SL spiral ligament, SV stria vascularis.
and seventh day after application (Table 1). At the first day, an association was only identified for the spiral ganglion cells. At the seventh day, associations were found in five out of seven regions. The highest associations between the gentamicin content and hearing threshold were found in the organ of Corti and the stria
vascularis (Table 1). In addition, associations were detected in the nerve fibers, the fibrocytes of the limbus and in the interdental cells.
4. Discussion Table 1 Statistical associations between hearing loss and cellular gentamicin accumulation. Cochlear region
1d
7d
Spiral ganglion cells
r ¼ 0.96 p ¼ 0.040 n¼4 r ¼ 0.14 p ¼ 0.744 n¼8 r ¼ 0.15 p ¼ 0.727 n¼8 r ¼ 0.08 p ¼ 0.854 n¼8 r ¼ 0.003 p ¼ 0.995 n¼7 r ¼ 0.26 p ¼ 0.540 n¼8 r ¼ 0.08 p ¼ 0.852 n¼8
r ¼ 0.66 p ¼ 0.221 n¼5 r ¼ 0.85 p ¼ 0.015 n¼7 r ¼ 0.80 p ¼ 0.031 n¼7 r ¼ 0.84 p ¼ 0.019 n¼7 r ¼ 0.92 p ¼ 0.003 n¼7 r ¼ 0.91 p ¼ 0.005 n¼7 r ¼ 0.71 p ¼ 0.073 n¼7
Nerve fibers
Fibrocytes of the Limbus
Interdental cells
Organ of Corti
Stria vascularis
Spiral ligament
Each field in the second and third column contains first the Pearson correlation coefficient r, second the p-value of the test comparing r to 0 and finally the number of data pairs from which the correlation was computed.
The most remarkable findings in this study were the detection of quantitative differences in cellular gentamicin accumulation after intratympanic application, the numerous strong cellular correlations and the detection of a diffusion barrier within the gapjunction system in the lateral wall between fibrocytes and basal cells.
4.1. Discrepancies in gentamicin distribution With respect to cellular localization of gentamicin, there are divergences in the current literature. Our findings are in line with the results of Imamura and Adams (2003) and Schmid et al. (2011) who described a widespread distribution of gentamicin. In the albino guinea pig model, gentamicin was located in numerous cochlear cell types (Imamura and Adams, 2003). In another experimental approach using frozen sections of the isolated inner ear of 2e4-month old rats, gentamicin was identified by fluorescence dyes after 10 min exposure in the inner and outer hair cells, the interdental cells, the spiral Limbus, the nerve fibers, the spiral lamina, the inner sulcus cells, and the dorsal region of the spiral ligament (Schmid et al., 2011). Variations in the detection levels were related to the applied concentrations (Schmid et al., 2011). In these experiments, gentamicin was not found in some supporting
46
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
cell types in the organ of Corti, in the stria vascularis and spiral ganglion cells (Schmid et al., 2011). In contrast to our findings, a prominent signal of gentamicinfluorescence was observed by Wang and Steyger (2009) in the stria vascularis 3 h after intraperitoneal application using the mouse-model. In the chinchilla animal model, high levels of tritiated gentamicin were identified in spiral ganglion cells, in the stria vascularis, and in the spiral ligament after intratympanic application by an implanted mini-pump (Roehm et al., 2007). Summarizing the findings of the various experimental settings, it becomes evident that differences in spatial gentamicin concentration are influenced by the method of application (peritoneal, systemic, intratympanal), the duration of application (syringe, mini-pump), the animal model analyzed (guinea pig, rat, mouse, chinchilla), the preparation of the tissues (paraffin-embedded, frozen sections), and the detection technique used (immuno histochemistry, fluorescence, tritiated gentamicin). Nevertheless, the overall conclusion can be drawn that gentamicin is incorporated into many different vertebrate cell types (Poirrier et al., 2010; Heinrich and Helling, 2012). 4.2. Infiltration routes and accumulation of gentamicin After intratympanic application, gentamicin influx into the cochlea might occur not only through the round window membrane, a commonly assumed diffusion route (Goycoolea, 2001), but also via a perilymph/modiolar route postulated by Rask-Andersen et al. (2006) some years ago and by Mikulec et al. (2009). This route also includes the tunnels within the osseous spiral lamina harboring the nerve fibers (Rask-Andersen et al., 2006). In addition, it was demonstrated that gentamicin delivered to the stapes footplate resulted in higher levels of hearing loss and a reduced number of outer hair cells in the basal cochlea turn compared with animals where gentamicin was placed on the round window membrane (King et al., 2013). Thus, in intratympanic application the route of entry might differ and lead to variability in gentamicin up-take. One day after gentamicin application, the nerve fiber area in the osseous spiral lamina was involved in four correlations when comparing the intra-animal staining intensities. This finding was unlikely to be induced by chance alone. The strong associations between nerve fibers and the three nearby areas (organ of Corti, fibrocytes of the Limbus and interdental cells) justify the assumption of rapid gentamicin diffusion. Thus, it can be speculated that the tunnel with its nerve fibers plays an important role in the early gentamicin influx. Seven days after application, there were still three correlations when comparing the cellular gentamicin content of the following: the organ of Corti and the interdental cells, the organ of Corti and the fibrocytes in the Limbus-area, the fibrocytes and the interdental cells. After the first phase of a fast gentamicin distribution by the nerve fiber within the osseous spiral lamina, a second phase can be observed in which the nerve fiber area seems to be of no importance. In this phase, the distribution of gentamicin in the three areas (organ of Corti, interdental cells and fibrocytes) seems to be regulated by a different mechanism. It is already known that the three areas are functionally coupled by a gap-junction system mainly responsible for potassium recycling from the inner hair cells to the Limbus (Spicer et al., 2000). It can be speculated that this system allows not only a free ion transport (Jagger and Forge, 2006; Heinrich and Helling, 2012) but also a distribution of gentamicin. During the early phase after application, gentamicin might be temporarily attached to binding sites which are located in the cytoplasm and/or distributed in the extracellular space. In the late phase, seven days after application, the gentamicin distribution pattern might represent a cochlear saturation kinetic already
proposed some years ago (Tran Ba Huy and Deffrennes, 1988). Analyzing fluid and tissue samples from rats after a single intravenous injection or sustained infusion, the authors reported a very fast gentamicin entry rate and an extremely slow efflux (Tran Ba Huy and Deffrennes, 1988). 4.3. Possible cellular mechanisms of gentamicin uptake and binding Analyzing the cellular gentamicin content within the cochlea, research was primarily focused on the uptake into outer hair cells. Two main routes of uptake were discussed: non-selective cation channels (Myrdal et al., 2005; Marcotti et al., 2005; Karasawa et al., 2010a) and endocytosis followed by lysosome-traffic (Hashino et al., 1997). Up to now, the accumulation of gentamicin by lysosomes was the expected process for cellular storage of gentamicin (Hashino et al., 1997). In respect to the cellular saturation kinetics (Tran Ba Huy and Deffrennes, 1988), there is only limited knowledge about the nature of the intracellular gentamicin binding processes. Evidence for a high affinity of gentamicin to phosphoinositides was postulated as binding sites (Lesniak et al., 2005). Beside these “widespread unspecific binding-sites”, cellular cochlear gentamicin-accumulation might be controlled in addition by specific gentamicin-binding proteins (GBPs) such as calreticulin (Karasawa et al., 2011), the cytoskeleton-linking membrane protein (CLIMP-63) (Karasawa et al., 2010b), and heat shock protein (HSP) 70 (Miyazaki et al., 2004). As gentamicin is a polycation, it can be expected that reactions with the numerous cellular polyanionic binding sites commonly occur in every cell type. These polyanions were found to play an important role in numerous cellphysiological processes, especially proteineprotein interactions (Jones et al., 2004; Sakamat-Miller et al., 2007). Based on new findings, the cytoplasm can be regarded as a highly polyanionic environment (Sakamat-Miller et al., 2007). Therefore, non-specific electrostatic interactions between gentamicin and cellular polyanions such as actin, tubulin, and ribosomes will occur. In summary, the differences identified in the cellular gentamicin-binding capacities and the associations found between the various cell types could be based on spatial differences in cellular polyanion content. 4.4. Diffusion barrier in the lateral wall Furthermore, there was another important finding in the lateral wall. It is widely accepted that the type I fibrocytes of the spiral ligament are connected with the nearby basal cells and intermediate cells of the stria vascularis by gap junctions, whereas the median cells are not (Kikuchi et al., 2000; Forge et al., 2003). Based on this knowledge, a homogenous distribution of gentamicin must be expected in the three connected cell types. However, clear cellular differences in gentamicin accumulation were identified between the fibrocytes of the spiral ligament and the nearby strial cells in the early and late phase of distribution. This unexpected finding can only be explained by a still unknown diffusion barrier dividing the gap-junction system into two parts, one restricted to the spiral ligament and the other restricted to the stria vascularis. When comparing the gentamicin content in the four cell types (type 1 fibrocytes and three cell types of the stria vascularis), statistic tests of fixed effects revealed a cell type effect (p < .0001) but no effect of treatment (p ¼ .7003; 1d or 7d) or cochlear turn (p ¼ .1994). 4.5. Possible factors responsible for differences in gap junction diffusion properties Experimental evidence was presented for the cochlear sensory epithelium that the permeation of anionic dyes was slower than
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
that of cationic probes (Zhao, 2005). This finding is in line with the widespread distribution of gentamicin in the epithelial gapjunction system in this study. The differences identified in gentamicin content when comparing type I fibrocytes and the nearby basal cells can only be explained by a hampered gentamicin flow through the gap junctions connecting type I fibrocytes and basal cells. As gentamicin is a polycation, it must be speculated that the reduced gentamicin diffusion into the basal cells is directly connected to its charge. Recent studies had revealed that Connexin 30 and Connexin 26 form heterotypic and/or heteromeric gap-junction channels in this interface (Kikuchi et al., 2000; Forge et al., 2003). Thus, in addition to the molecular weight and size of the transported molecules, the net charge, shape and interactions with specific connexins might play an important role in molecule passage (Goldberg et al., 2004). 4.6. Associations between cellular gentamicin content and hearing loss In this study, cellular differences in gentamicin accumulation were found between one day and seven days after application. Seven days after application, numerous associations were found when comparing the staining intensities in the different cell types with the recorded hearing losses by statistical analyses. This interaction is in line with a detected gentamicin-dependent elevation of hearing threshold in hearing threshold parallel to an increase in NO-production, predominantly in the lateral wall (Heinrich et al., 2008; Heinrich and Helling, 2012). Taken these findings together, it becomes evident that gentamicin accumulation correlates in a time- and spatial-dependent manner with the increased hearing loss and with the up-regulated cochlear NOproduction. 5. Conclusion The current study revealed that the non-specific electrostatic interactions appear to be a very important factor in cellular uptake and compartmentalization (Jones et al., 2004). This should be kept in mind when pharmacological substances which differ in their polyanionic or polycationic nature are evaluated for human therapy. Conflict of interest The authors declare that they have no conflict of interest. Acknowledgments The authors gratefully acknowledge Mrs. K. Benz for her technical assistance. References re's disease. Blakley, B.W., 2000. Update on intratympanic gentamicin for Menie Laryngoscope 110 (2 Pt 1), 236e240. Choung, Y.H., Taura, A., Pak, K., Choi, S.J., Masuda, M., Ryan, A.F., 2009. Generation of highly-reactive oxygen species is closely related to hair cell damage in rat organ of Corti treated with gentamicin. Neuroscience 161 (1), 214e226. de Groot, J.C.M.J., Meeuwsen, F., Ruizentaal, W.E., Veldman, J.E., 1990. Ultrastructural localization of gentamicin in the cochlea. Hear. Res. 50 (1e2), 35e42. Forge, A., Schacht, J., 2000. Aminoglycoside antibiotics. Audiol. Neurootol. 5 (1), 3e22. Forge, A., Becker, D., Casalotti, S., Edwards, J., Marziano, N., Nevill, G., 2003. Gap junctions in the inner ear: comparison of distribution patterns in different vertebrates and assessement of connexin composition in mammals. J. Comp. Neurol. 467 (2), 207e231. Goldberg, G., Valiunas, V., Brink, P.R., 2004. Selective permeability of gap junction channels. Biochim. Biophys. Acta 1662 (1e2), 96e101.
47
Goycoolea, M.V., 2001. Clinical aspects of round window membrane permeability under normal and pathological conditions. Acta Otolaryngol. 121 (4), 437e447. Hashino, E., Shero, M., Salvi, R.J., 1997. Lysosomal targeting and accumulation of aminoglycoside antibiotics in sensory hair cells. Brain Res. 777 (1e2), 75e85. Heinrich, U.R., Helling, K., 2012. Nitric oxide e a versatile key player in cochlear function and hearing disorders. Nitric Oxide 27 (2), 106e116. Heinrich, U.R., Selivanova, O., Brieger, J., Mann, W., 2006. Endothelial nitric oxide synthase- upregulation in the cochlea of the guinea pig after intratympanic gentamicin injection. Eur. Arch. Otorhinolaryngol. 263 (1), 62e68. Heinrich, U.R., Helling, K., Sifferath, M., Brieger, J., Li, H., Schmidtmann, I., Mann, W.J., 2008. Gentamicin increases NO production and induces hearing loss in guinea pigs. Laryngoscope 118 (8), 1438e1442. Heinrich, U.R., Selivanova, O., Schmidtmann, I., Feltens, R., Brieger, J., Mann, W.J., 2010. Noise exposure alters cyclooxygenase 1 (COX-1) and 5-lipoxygenase (5LO) expression in the guinea pig cochlea. Acta Otolaryngol. 130 (3), 358e365. €nfeld, U., Clarke, A.H., 2007. Treatment of Me nie re’s disease by lowHelling, K., Scho dosage intratympanic gentamicin application e effect on otolith function. Laryngoscope 117 (12), 2244e2250. Hirose, K., Hockenbery, D.N., Rubel, E.W., 1997. Reactive oxygen species in chick hair cells after gentamicin exposure in vitro. Hear. Res. 104 (1e2), 1e14. Imamura, S.I., Adams, J.C., 2003. Distribution of gentamicin in the guinea pig inner ear after local or systemic application. J. Assoc. Res. Otolaryngol. 4 (2), 176e195. Jagger, D.J., Forge, A., 2006. Compartmentalized and signal-selective gap junctional coupling in the hearing cochlea. J. Neurosci. 26 (4), 1260e1268. Jeong, S.W., Kim, L.S., Hur, D., Bae, W.Y., Kim, J.R., Lee, J.H., 2010. Gentamicin-induced spiral ganglion cell death: apoptosis mediated by ROS and the JNK signaling pathway. Acta Otolaryngol. 130 (6), 670e678. Jones, L.S., Yazzie, B., Middaught, C.R., 2004. Polyanions and the proteome. Mol. Cell. Proteomics 3 (8), 746e769. Karasawa, T., Wang, Q., Fu, Y., Cohen, D.M., Steyger, P.S., 2010a. TRPV4 enhances the cellular uptake of aminoglycoside antibiotics. J. Cell. Sci. 121 (Pt 17), 2871e2879. Karasawa, T., Wang, Q., David, L.L., Steyger, P.S., 2010b. CLIMP-63 is a gentamicinbinding protein that is involved in drug-induced cytotoxicity. Cell. Death Dis. e102. http://dx.doi.org/10.1038/cddis.2010.80. Karasawa, T., Wang, Q., David, L.L., Steyger, P.S., 2011. Calreticulin binds to gentamicin and reduces drug-induced ototoxicity. Toxicol. Sci. 124 (2), 378e387. Kikuchi, T., Kimura, R.S., Paul, D.L., Takasaka, T., Adams, J.C., 2000. Gap junction systems in the mammalian cochlea. Brain Res. Brain Res. Rev. 32 (1), 163e166. King, E.B., Salt, A.N., Kel, G.E., Eastwood, H.T., O'Leary, S.J., 2013. Gentamicin administration on the stapes footplate causes greater hearing loss and vestibulotoxicity than round window administration in guinea pigs. Hear. Res. 304 (Oct), 159e166. Lange, G., Maurer, J., Mann, W., 2004. Long-term results after interval therapy with re’s disease. Laryngoscope 114 (1), intratympanic gentamicin for Menie 102e105. Lesniak, W., Pecoraro, V.L., Schacht, J., 2005. Ternary complexes of gentamicin with iron and lipid catalyze formation of reactive oxygen species. Chem. Res. Toxicol. 18 (2), 357e364. Liu, H.Y., Chi, F.L., Gao, W.Y., 2008. Taurine attenuates aminoglycoside ototoxicity by inhibiting inducible nitric oxide synthase expression in the cochlea. Neuroreport 19 (1), 117e120. Marcotti, W., van Netten, S.M., Kros, C.J., 2005. The aminoglycoside antibiotic dihydrostreptomycin rapidly enters mouse outer hair cells through the mechano-electrical transducer channels. J. Physiol. 567 (Pt 2), 505e521. Mikulec, A.A., Plontke, S.K., Hartsock, J.J., Salt, A.N., 2009. Entry of substances into perilymph through the bone of the otic capsule following intratympanic applications in guinea pigs: Implications for local drug delivery in humans. Otol. Neurotol. 30 (2), 131e138. Miyazaki, T., Sagawa, R., Honma, T., Noguchi, S., Harada, T., Komatsuda, A., Ohtani, H., Wakui, H., Sawada, K., Otoka, M., Watanabe, S., Jikei, M., Ogawa, N., Hamada, F., Itoh, H., 2004. 73-kDa molecular chaperone HSP73 is a direct target of antibiotic gentamicin. J. Biol. Chem. 279 (17), 17295e17300. Myrdal, S.E., Johnson, K.C., Steyger, P.S., 2005. Cytoplasmic and intra-nuclear binding of gentamicin does not require endocytosis. Hear. Res. 204 (1e2), 156e169. Nakashima, T., Teranishi, M., Hibi, T., Kobayashi, M., Umemura, M., 2000. Vestibular and cochlear toxicity of aminoglycosides e a review. Acta Otolaryngol. 120 (8), 904e911. Okuda, T., Sugahara, K., Shimogori, H., Yamashita, H., 2004. Inner ear changes with intracochlear gentamicin administration in guinea pigs. Laryngoscope 114 (4), 694e697. Poirrier, A.L., Van den Ackerveken, P., Kim, T.S., Vandenbosch, R., Nguyen, L., Lefebvre, P.P., Malgrange, B., 2010. Ototoxic drugs: difference in sensitivity between mice and guinea pigs. Toxicol. Lett. 193 (1), 41e49. Rask-Andersen, H., Schrott-Fischer, A., Pfaller, K., Glueckert, R., 2006. Perilymph/ modiolar communication routes in the human cochlea. Ear Hear. 27 (5), 457e465. Roehm, P., Hoffer, M., Balaban, C.D., 2007. Gentamicin uptake in the chinchilla inner ear. Hear. Res. 230 (1e2), 43e52. Sakamat-Miller, N., Fang, J., Seidel, C.W., Assenov, Y., Albrecht, M., Middaugh, C.R., 2007. A network-based analysis of polyanion-binding proteins utilizing human protein arrays. J. Biol. Chem. 282 (14), 10153e10163.
48
U.-R. Heinrich et al. / Hearing Research 326 (2015) 40e48
Schmid, K., Strutz, J., Gleich, O., Kwok, P., 2011. Localization of gentamicin uptake in the acutely isolated inner ear of the rat. Int. J. Physiol. Pathophysiol. Pharmacol. 3 (2), 71e87. Sha, S.H., Schacht, J., 1999. Formation of reactive oxygen species following bioactivation of gentamicin. Free Rad. Biol. Med. 26 (3e4), 341e347. Spicer, S.S., Thomopoulos, G.N., Schulte, B.A., 2000. Structural evidence for ion transport and tectorial membrane maintenance in the gerbil limbus. Hear. Res. 143 (1e2), 147e161. Takumida, M., Popa, R., Anniko, M., 1999. Free radicals in the guinea pig inner ear following gentamicin exposure. ORL J. Otorhinolaryngol. Relat. Spec. 61 (2), 63e70. Tran Ba Huy, P., Deffrennes, D., 1988. Aminoglycoside binding sites in the inner ears of guinea pigs. Antimicrob. Agents Chemother. 32 (4), 467e472.
Wang, J., van de Water, T.R., Bonny, C., de Ribaupierre, F., Puel, J.L., Zine, A., 2003. A peptide inhibitor of c-Jun N-terminal kinase protects against both aminoglycoside and acoustic trauma-induced auditory hair cells death and hearing loss. J. Neurosci. 23 (24), 8596e8607. Wang, Q., Steyger, P.S., 2009. Trafficking of systemic fluorescent gentamicin into the cochlea and hair cells. J. Assoc. Res. Otolaryngol. 10 (2), 205e219. Xie, J., Talaska, A.E., Schacht, J., 2011. New developments in aminoglycoside therapy and ototoxicity. Hear. Res. 281 (1e2), 28e37. Zhao, H.B., 2005. Connexin 26 is responsible for anionic molecule permeability in the cochlea for intercellular signaling and metabolic communications. Eur. J. Neurosci. 21 (7), 1859e1868.