Address correspondence and reprint requests to Dr Jochen Schacht, Kresge Hearing Research Institute, 1301 East Ann Street, Ann Arbor Michigan, USA. E-mail: firstname.lastname@example.org
Aminoglycoside antibiotics strongly bind to phosphoinositides and affect their membrane distribution and metabolism. Kanamycin treatment also disrupts Rac/Rho signaling pathways to the actin cytoskeleton in the mouse inner ear in vivo. Here, we investigate the influence of kanamycin on phosphoinositide signaling in sensory cells (hair cells) of the mouse cochlea. Immunoreactivity to phosphatidylinositol-3,4,5-trisphosphate (PIP3) decreased in the organ of Corti, especially in outer hair cells, after 3–7 days of drug treatment, whereas imunoreactivity to phosphatidylinositol-4,5-bisphosphate (PIP2) increased. Immunoreactivity to PIP2 was present at the apical poles of outer hair cells, but appeared in their nuclei only after drug treatment. Furthermore, nuclear PIP2 formed a complex with histone H3 and attenuated its acetylation in outer hair cells. In agreement with reduced PIP3 signaling, phosphorylated Akt decreased in both the cytoplasm and nuclei of outer hair cells after kanamycin treatment. This study suggests that kanamycin disturbs the balance between PIP2 and PIP3, modifies gene transcription via histone acetylation and diminishes the PI3K/Akt survival pathway. These actions may contribute to the death of outer hair cells, which is a consequence of chronic kanamycin treatment.
Phosphoinositides in their different incarnations regulate a wide variety of cellular processes at the plasma membrane, in the cytoplasm and in the nucleus (Irvine 2003; Parker 2004; De Matteis et al. 2005). Aminoglycoside antibiotics such as neomycin, gentamicin or kanamycin strongly bind to phosphoinositides (Schacht 1979), alter their metabolism (Schacht 1976) and disrupt membrane structures containing polyphosphoinositides (Lodhi et al. 1979). These properties have made the drugs useful probes for studying phosphoinositide metabolism and function (Janmey and Stossel 1989; Arbuzova et al. 2000; Holz et al. 2000). Phosphoinositides may also be cellular targets of aminoglycosides in their adverse effects on tissues, notably in the death of both sensory cells in the inner ear (ototoxicity) and proximal tubule cells in the kidney (nephrotoxicity).
Although early studies of ototoxicity in vivo had demonstrated changes in inner ear phosphoinositide metabolism (Orsulakova et al. 1976), only recently have more details of potential physiological mechanisms emerged. The phosphatidylinositol-3-kinase/Akt pathway and its downstream effector nuclear factor κB (NF-κB) play a central role in cell growth and survival in many tissues (Song et al. 2005; Woodgett 2005) including the inner ear (Nagy et al. 2005). NF-κB is suppressed by aminoglycoside-induced insult in vivo, and its activation can rescue the sensory hair cells of the cochlea (Jiang et al. 2005). Furthermore, aminoglycosides disturb the structural integrity of the actin cytoskeleton in the inner ear via an action on small GTPases (Jiang et al. 2006b), another potential link to phosphoinositides because the actin regulatory proteins and the assembly of actin fibers involves these lipids (Cooper and Schafer 2000; Takenawa and Miki 2001).
Phosphoinositides also constitute a nuclear signaling network regulated independently from other cell compartments (Irvine 2003; Martelli et al. 2005; Gonzales and Anderson 2006). The precise function of these lipids in the nucleus has yet to be determined but may include complex structural and regulatory roles such as influencing pre-mRNA splicing and chromatin structure (Osborne et al. 2001; Boronenkov et al. 1998), as well as regulating gene transcription by histone binding (Yu et al. 1998). The effect of aminoglycosides on nuclear phosphoinositide signaling in vivo has, to the best of our knowledge, not been studied yet.
In this study, we investigate cytoplasmic and nuclear phosphoinositide signaling pathways in the organ of Corti in vivo in order to gain a more detailed understanding of aminoglycoside–phosphoinositide interactions. The animal model used is the adult mouse (CBA/J strain), which receives chronic injections of kanamycin that primarily destroy the outer hair cells (Wu et al. 2001). The selected dosing regimen yields a slow progressive ototoxic action, which allows the determination of early cellular responses and their relationship to cell death and survival pathways (Jiang et al. 2005, 2006a).
Materials and methods
Kanamycin sulfate was purchased from USB Corporation (Cleveland, OH, USA; Cat. #17924; Lot #110755), ketamine (Ketaset®) from Fort Dodge Animal Health (Fort Dodge, IA, USA), xylazine (TranquiVed®) from Vedco Inc. (St Joseph, MO, USA), and for western blotting detection reagents, from GE Health Care (Piscataway, NJ, USA). BenchMarkTM Protein ladders were obtained from InvitrogenTM Life Technologies (Carlsbad, CA, USA). Anti-histone H3, anti-acetyl-histone H3 (Lys9), anti-phospho-histone H3 (Ser10), anti-histone H2A, anti-acetyl-histone H2A (Lys5), anti-Akt1/2 and anti-phospho-Akt1/2 (Thr308 and Ser473) polyclonal antibodies were obtained from Cell Signaling Technology Inc. (Beverly, MA, USA); anti-PIP2 and anti-PIP3 monoclonal IgM were from Echelon Research Laboratories Inc. (Salt Lake City, UT, USA), and anti-PI4Kβ polyclonal antibody was from Upstate Biotechnology (Lake Placid, NY, USA). Agarose-conjugated protein A/G was purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA, USA), agarose-conjugated anti-mouse IgM was from Sigma-Aldrich Inc. (St Louis, MO, USA), antibodies to phosphoserine were from Zymed Laboratories Inc. (South San Francisco, CA, USA) and anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody was from Chemicon International (Temecula, CA, USA). Horseradish peroxidase-conjugated secondary antibodies for western blotting were purchased from Jackson Immunoresearch Laboratories (West Grove, PA, USA), fluorescent secondary antibodies (Alexa 488 and Alexa 546) and rhodamine phalloidin, propidium iodide, as well as Hoechest 33342 were from Molecular Probes Inc. (Eugene, OR, USA). CompleteTM mini EDTA-free protease inhibitor cocktail tablets were from Roche Diagnostic GmbH (Mannheim, Germany).
Animals and drug administration
Male CBA/J mice were delivered at an age of 4 weeks from Harlan Sprague–Dawley Co. (Indianapolis, IN, USA) and divided into three groups (saline-treated control and kanamycin treatment for 3 and 7 days). The animals had free access to water and a regular mouse diet (Purina 5025; Purina, St Louis, MO, USA), and were acclimated for 1 week before the start of experimental procedures. Experimental protocols were approved by the University of Michigan Committee on the Use and Care of Animals. Animal care was under the supervision of the University of Michigan's Unit for Laboratory Animal Medicine.
Drug dosing followed our earlier protocol for studying aminoglycoside-induced hearing loss (Wu et al. 2001). Experimental mice received a dose of 700 mg of kanamycin base/kg body weight twice daily by subcutaneous injection. Cochleae were collected 3 h after the last injection on the third and seventh day. At this time, the drug did not significantly affect auditory function and morphology. Only continuous treatment with the same dose of kanamycin for 11 days or more will result in outer hair cell death (Jiang et al. 2006a).
Extraction of total protein
The cochleae were removed rapidly and dissected in ice-cold 10 mm phosphate-buffered saline (PBS). The dissections included the sensory and supporting cells of the organ of Corti, the supporting structures of the lateral wall (stria vascularis and spiral ligament) and the cochlear portion of the spiral ganglion. Tissue from one mouse cochlea was homogenized in ice-cold radioimmunoprecipitation (RIPA) buffer using a micro Tissue Grind Pestle (Kontes Glass Company, Vineland, NJ, USA) for 30 s. The homogenates were kept on ice for 15 min and then centrifuged at 15 000 g at 4°C for 10 min.
Extraction of nuclear protein from cochlear homogenates
Cochleae were removed rapidly and dissected in 10 mm ice-cold PBS. Tissues from three mice were pooled and homogenized in cytoplasmic lysis buffer [10 mm sodium HEPES, pH 7.9, additionally containing 10 mm KCl, 1 mm EDTA, 1 mm EGTA, 5 mm dithiothreitol (DTT), 10 mm each of the phosphatase inhibitors NaF and sodium β-glycerophosphate, 1 μg/mL (p-amidinophenyl) methanesulfonyluoride, and 1/10 tablet/mL of Complete Mini EDTA-free protease inhibitor cocktail] by using a micro Tissue Grind Pestle for 10 s. The homogenates were kept on ice for 15 min and then centrifuged at 750 g at 4°C for 10 min. The crude nuclear pellet was washed twice with a cytosolic lysis buffer and centrifuged each time at 15 000 g at 4°C for 5 min. The washed nuclear pellets were resuspended in a nuclear lysis buffer (50 mm Tris-HCl, pH 7.5, containing 10% glycerol, 400 mm KCl, 1 mm EDTA, 1 mm EGTA, 5 mm DTT, and the aforementioned phosphatase and protease inhibitors) and kept on ice for 30 min. The suspensions were centrifuged at 15 000 g at 4°C for 10 min, and the supernatant was collected as the nuclear protein extract. The nuclear extract was stored at − 80°C until analyzed. Protein concentrations were measured by the Bio-Rad Protein Assay (Bio-Rad, Hercules, CA, USA).
Mice were decapitated and the temporal bones were quickly removed. Cochleae were immediately perfused with and fixed overnight in 4% paraformaldehyde at 4°C. Cryostat sections of 5 µm were incubated in 0.5% Triton X-100 for 15 min at room temperature (22–24°C) and then washed three times with PBS. The sections were blocked with 10% goat serum for 30 min at room temperature, followed by the application of the primary antibody at 4°C for 72 h. Concentrations of anti-PIP2, anti-PIP3, anti-histone H3, anti-acetyl-histone H3 (Lys9), anti-phospho-histone H3 (Ser10), anti-phospho-Akt1/2 (Thr308 and Ser473) and anti-PI4Kβ were 1 : 50, and anti-Akt1/2 was 1 : 100. The sections and surface preparations were washed three times with PBS, and the secondary antibody (either Alexa 488 or Alexa 546 conjugated) at a concentration of 1 : 500 was applied at 4°C overnight in darkness. Finally the preparations were incubated with either Hoechst 33342 or propidium iodide (2 µg/mL in PBS) at room temperature for 40 min for fluorescent visualization of the nucleus. After being washed with PBS, the slides were mounted and photographed with laser confocal microscope (Zeiss LSM 510; Carl Zeiss Microimaging Inc., Thornwood, NY).
Western blot analysis
Either total protein (50 μg each) or nuclear protein (30 μg each) was separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE). A 12% polyacrylamide gel was used for the separation of histone, a 10% gel for Akt1/2, and a 7.5% gel for PI4Kβ. After electrophoresis, the proteins were transferred onto nitrocellulose membranes (Pierce, Rockford, IL, USA) and blocked with 5% non-fat dry milk in PBS with 0.1% Tween 20 (PBS-T). The membranes were then incubated overnight with anti-Akt1/2 (1 : 500), anti-phospho-Akt1/2 (Thr308 and Ser473) 1 : 250, anti-histone H3 (1 : 500), anti-acetyl-histone H3 (Lys9) 1 : 500, anti-phospho-histone H3 (Ser10) 1 : 200, anti-histone H2A and anti-acetyl-histone H2A (Lys9) 1 : 250 and anti-PI4-kinase-β (1 : 100) rabbit polyclonal antibodies. After three washes with PBS-T buffer, the membranes were incubated with the secondary antibody (goat anti-rabbit IgG) at a concentration of 1 : 10 000 for 1 h. Following extensive washing of the membrane, the immunoreactive bands were visualized by enhanced chemiluminescence (ECL; GE Healthcare, Piscataway, NJ, USA). The membranes were then stripped and restained with anti-GAPDH at a concentration of 1 : 20 000 to confirm the consistency of protein loads.
Cochlear nuclear extracts (300 μg) were incubated with 2 μg of anti-PIP2 antibody in an Eppendorf tube (a minimum of 300 μL per tube). The tube was rotated overnight at 4°C and then 2 μg of agarose-conjugated anti-IgM were added and the tube again rotated overnight at 4°C. Cochlear total proteins (300 μg) were pre-cleared by incubation with 20 μL of protein A/G-agarose beads for 1 h at 4°C. Four μg of anti-PI4Kβ antibody were added to the supernatant, rotated for 6 hr at 4°C, and then 20 μL of protein A/G-sepharose beads were added for incubation overnight at 4°C. Beads were collected by centrifugation for 1 min at 1200 g at 4°C and the supernatant was removed. The pellets were washed three times with lysis buffer, and after the final wash the pellets were re-suspended in 20 µL of 2X electrophoresis sample buffer (including 0.5 m Tris-HCl, 4% SDS, 20% glycerol, 0.02% bromophanol blue and 10%β-mercaptoethanol). Subsequently, anti-histone H2A (1 : 250), anti-histone H3 (1 : 500) and anti-phospho-serine (1 : 100) were detected by using western blotting.
Data were statistically evaluated by Student's t-test and by analyses of variance with a Student's Newman–Keuls test for significance (p < 0.05) using primer of biostatistics software (McGraw-Hill Software, New York, NY, USA).
PIP3 decreases in outer hair cells after kanamycin treatment
In untreated and saline-treated CBA/J mice, PIP3 was primarily located in the membranes of most cells of the organ of Corti, including outer and inner hair cells, pillar cells and supporting cells. The distribution of staining also suggested its presence in the cytoplasm of supporting cells (Fig. 1a). During kanamycin treatment for 3 days PIP3 gradually decreased in the membranes of outer hair cells, but there was no obvious change in the inner hair cells, pillar cells and supporting cells (Fig. 1b). After kanamycin treatment for 7 days the cochlear structure was still intact (Fig. 1c), but PIP3 had disappeared completely from the outer hair cells and the tops of pillar cells (Fig. 1d).
PIP2 increases in the membranes and nuclei of outer hair cells after kanamycin treatment
The immunostaining for PIP2 was characterized by a weak punctate presence in the apical portion (cuticular plates) of outer hair cells in saline-treated control mice (Fig. 2). In contrast to the disappearance of PIP3, PIP2 increased in the apex of outer hair cells after kanamycin treatment for 3 days forming dense dotted arrays. The staining patterns maintained a similar appearance after 7 days of treatment.
At the nuclear level, PIP2 was essentially absent (i.e. below the detection threshold) from the nuclei of outer hair cells in control animals (Fig. 3a). As PIP2 increased in the cells, immunostaining first moved to the edge of the outer hair cell nuclei and then gradually further into the nuclei (Figs 3b and c).
Nuclear PIP2 forms a complex with and inhibits the acetylation of histone H3
Immunoprecipitation of nuclear extracts from cochlear tissues (including but not limited to the organ of Corti) with an anti-PIP2 antibody followed by blotting against histone H3 showed that a complex between PIP2 and histone H3 was formed, increasing in intensity with kanamycin treatment for 7 days (Fig. 4a). However, PIP2 did not form a complex with either histone H2A or β-actin (data not shown).
Under the same treatment conditions kanamycin decreased the acetylation of histone H3 in the tissue extracts, whereas the total histone did not change (Figs 4b and c). For a localization of acetylated histone H3 we then stained a surface preparation of the organ of Corti for immunoreactivity to acetyl-histone H3 (Fig. 5). Staining was heavy in the nuclei of outer hair cells of control animals and decreased during kanamycin treatment for 3 days, further diminishing at 7 days. In contrast, levels of phospho-histone H3 did not change after kanamycin treatment for 3 and 7 days (data not shown).
Kanamycin treatment inhibits phosphorylation of Akt in outer hair cells
The phosphorylation of Akt reflects the activity of the phosphatidylinositol-3-kinase pathway. As PIP3 decreased after kanamycin treatment, we examined the total Akt1/2 and phosphorylated Akt1/2 (Thr308 and Ser473) in inner ear tissue extracts (Fig. 6). Kanamycin did not affect the level of total Akt1/2. However, phosphorylated Akt1/2 was reduced by kanamycin treatment for 3 days and reduced further after 7 days of treatment.
In sections of the cochleae of saline-treated control animals, phosphorylated Akt1/2 was localized in both the cytoplasm and nuclei of outer and inner hair cells, and supporting cells (Fig. 7a). After kanamycin treatment for 3 days, phosphorylated Akt1/2 decreased gradually in both the cytoplasm and nuclei of outer hair cells (Fig. 7b), and decreased even further after treatment for 7 days. Although phosphorylated Akt1/2 essentially disappeared from the outer hair cells, it remained present in supporting cells after 7 days treatment of kanamycin (Fig. 7c).
The present study clearly indicates phosphoinositides as in vivo targets of aminoglycosides in the cochlea. Of particular interest is the drug-induced shift between the levels of PIP2 and PIP3, and the emergence of PIP2 in the nuclei of outer hair cells. Both of these events appear to have distinct consequences for the reaction of the cell to a drug challenge.
PIP3 is a key regulator of several pathways, including Akt as a prominent downstream target. This pathway may serve the maintenance of homeostasis in the survival of cells either during development or under stress, and it may serve such a function in the mammalian cochlea (Jiang et al. 2005; Nagy et al. 2005). As the level of PIP3 is decreased by kanamycin treatment in the cochlea, the phosphorylation of Akt correspondingly is attenuated to the point that it escapes detection in immunocytochemical staining of outer hair cells. This predominant attenuation of the pathway in outer hair cells is significant in the context that these cells are the primary targets of the ototoxic actions of aminoglycosides (Forge and Schacht 2000), and will die during continued drug treatment (Wu et al. 2001; Jiang et al. 2006a). Surviving supporting cells, in contrast, have increased levels of phospho-Akt in their nuclei. Although the nuclear presence of the PI3K/Akt pathways has been documented, little is known about its functional significance (Neri et al. 2002). Some targets of Akt, notably the forkhead family of transcription factors, reside in the nucleus where they promote the transcription of apoptotic genes (Nicholson and Anderson 2002). Upon phosphorylation by active phospho-Akt, these factors may be exported and sequestered in the cytoplasm. It is intriguing to speculate that nuclear phospho-Akt therefore contributes to the survival of supporting cells during a drug challenge.
Although the nuclear role of phosphoinositides is less firmly established than their plasma membrane/cytoplasmic pathway, the lipids may contribute to the regulation of gene expression by modulating DNA replication, gene transcription and apoptosis (Irvine 2003; Martelli et al. 2005; Gonzales and Anderson 2006). Histones are potential targets of nuclear phosphoinositides as PIP2, but not PIP3, binds to histone H1 and histone H3, regulating gene transcription (Yu et al. 1998). Histone binding increases as nuclear PIP2 increases with kanamycin treatment, correlating with an attenuation of histone acetylation. Histone modification is an important switch in the control of gene expression (Verdone et al. 2005). Furthermore, phosphoinositides can regulate histone phosphorylation and this phosphorylation can result in cell apoptosis (Enomoto et al. 2003; Prigent and Dimitrov 2003). The unchanged levels of phospho-histone H3 after kanamycin treatment for 3 and 7 days is in agreement with our previous finding that cell death in the inner ear does not occur until after either 10 or 11 days (Jiang et al. 2006a). Histone acetylation also regulates gene expression but, as far as we can determine, phosphoinositides have not yet been linked to histone acetylation. It is interesting, however, that oxidative stress can inhibit histone acetylation (Berthiaume et al. 2006) because oxidative stress is one of the mechanisms by which aminoglycosides exert their ototoxic actions (Priuska and Schacht 1995; Lesniak et al. 2005).
Our results also add to our knowledge of phosphoinositides in the inner and outer hair cells in particular. Drug effects on these lipids in the cochlea have long been recognized (Orsulakova et al. 1976), as has their presence in outer hair cells (Williams et al. 1987; Montcouquiol and Corwin 2001; Hirono et al. 2004). Their intracellular localization, however, has been disputed, possibly because of the different detection methods employed. PIP2 is often detected by transfecting cells with plasmids that direct expression of green fluorescent protein (GFP)-tagged pleckstrin homology (PH) domains (Varnai and Balla 1998), a procedure which may have some limitations (Balla et al. 2000). In contrast, anti-PIP2 antibodies can efficiently and selectively recognize PIP2 (Fukami et al. 1988; Thomas et al. 1999), including PIP2 present in the hair bundles of sensory cells (Hirono et al. 2004). Our localization of PIP2 in the membranes and the apical plates of outer hair cells is in agreement with the literature (Hirono et al. 2004) but, in addition, we show the dynamic behavior of phosphoinositides and their nuclear presence.
We cannot say with certainty which reactions are responsible for the shift in PIP2 and PIP3 in the hair cells and the nuclear appearance of PIP2 during drug treatment. In vitro, aminoglycosides interfere with polyphosphoinositide metabolism by binding to the polar head groups of these lipids (Schacht 1976; Lodhi et al. 1979), but their actions in vivo will depend, in part, on tissue penetration and distribution of the drugs vs. the compartmentalization of phosphoinositides. The most parsimonious explanation, nevertheless, is an inhibition of the kinase reaction(s) from PIP2 to PIP3 leading to an accumulation of PIP2 and a corresponding decrease in PIP3 in the membranes. For the nuclear emergence of PIP2, a similar mechanism may be assumed as aminoglycosides may enter the nucleus (Sha and Schacht, unpublished). Furthermore, as all enzymatic machinery and potential precursors are present in the nucleus, a translocation of increased PIP2 from other cell compartments seems unlikely. Interestingly, phosphatidylinositol-4-kinase, which synthesizes PIP2 from phosphatidylinositol-5-phosphate, was not found immuno-cytochemically in the nuclei of outer hair cells in both normal and kanamycin-treated mice (data not shown).
In summary, this study suggests that kanamycin alters phosphoinositide signaling by affecting PIP2 and PIP3. Increased nuclear PIP2 increases complex formation between PIP2 and histone H3, attenuating the acetylation of histone H3, which may induce an inhibition of gene transcription. A consequence of low levels of PIP3 may be the inhibition of the activities of the PIP3/Akt pathway, contributing to outer hair cell death.
This research was supported by Research Grant RO1 DC-03685 and Core Center Grant DC-05188 from the National Institute on Deafness and Other Communication Disorders, National Institutes of Health.