Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Gene Therapy (2001) 8, 789–794 2001 Nature Publishing Group All rights reserved 0969-7128/01 $15.00 www.nature.com/gt RESEARCH ARTICLE A modified adenovirus can transfect cochlear hair cells in vivo without compromising cochlear function AE Luebke1,2, JD Steiger1, BL Hodges3 and A Amalfitano3 1 Department of Otolaryngology and 2Neuroscience Program, University of Miami School of Medicine, Miami, FL; and 3Department of Pediatrics, Division of Medical Genetics and Department of Genetics, Duke University Medical Center, Durham, NC, USA The loss of cochlear hair cells, or the loss of their capacity to transduce acoustic signals, is believed to be the underlying mechanism in many forms of hearing loss. To develop viral vectors that allow for the introduction of genes directly into the cochleae of adult animals, replication-deficient (E1−, E3−) and replication-defective (E1−, E3−, pol−) adenovirus vectors were used to transduce the bacterial -galactosidase gene into the hair cells of the guinea pig cochlea in vivo. Distortion product otoacoustic emissions, which monitor the functional status of outer hair cells, were measured throughout the viral infection periods to identify hair cell ototoxicity. The results demonstrated that the use of the (E1−, E3−) adenovirus vectors containing CMV-driven LacZ, compromised cochlear function when gradually introduced into scala tympani via an osmotic pump. However, when (E1−, E3−, pol−) adenoviral vectors containing CMV-driven LacZ were used to transduce cochlear hair cells, there was no loss of cochlear function over the frequency regions tested, and -galactosidase (gal) was detected in over 80% of all hair cells. Development of a viral vector that infects cochlear hair cells without virusinduced ototoxic effects is crucial for gene replacement strategies to treat certain forms of inherited deafness and for otoprotective strategies to prevent hair cell losses to treat progressive hearing disorders. Moreover, in vivo (E1−, E3−, pol−) adenovirus mediated gene-transfer techniques applied to adult guinea pig cochleae may be useful in testing several hypotheses concerning what roles specific genes play in normal cochlear function. Gene Therapy (2001) 8, 789–794. Keywords: cytomegalovirus promoter; LacZ reporter; replication-defective; cochlea; hair cell; distortion-product otoacoustic emissions Introduction The introduction of genes into cells has become an effective method for the expression of proteins, both for experimental manipulation and for therapeutic purposes. One model for manipulating cochlear functions in vivo relies upon transgenic mouse technologies, that allows for the over-expression or ‘knock-out’ of gene constructs potentially critical to cochlea hair cell function. Several problems with traditional transgenic mouse techniques include the expense and time involved in isolating the transgenic lines of mice. In addition, due to the small size of the mouse cochlea, isolated hair cell experiments are exceedingly difficult. Finally, the mouse strain most commonly used for the initial recombination events in embryonic stem cells, the 129SvEv strain, is extremely resistant to noise-damage,1 and thus is not an ideal model system in which to elucidate all aspects of cochlear function. Transgenic studies in mice using bacterial artificial chromosomes (BACs) containing cochlear-specific promoters have directed transgene expression to cochlear hair cells.2 However, these transgenes are necessarily expressed throughout development. Thus, it is not clear if the phenotype exhibited is due specifically to the somatic Correspondence: AE Luebke, University of Miami Ear Institute (M805), PO Box 016960, Miami, FL 33101-6960, USA Received 22 November 2000; accepted 14 February 2001 transgene expression, or if it is a consequence of previously altered cochlear development. Gene delivery to the nervous system, on the other hand, offers a potential alternative to traditional transgenic approaches, and has been extensively studied, primarily using viral vectors. Of the numerous viral vectors tested, only adenoviral vector-mediated gene transfer to cochlear hair cells has been effective in vitro, as herpes simplex virus (HSV) and adeno-associated virus (AAV) failed to infect cultured hair cells.3–6 HSV, vaccina virus, lentivirus, and AAV vectors have also failed to infect cochlear hair cells when tested in vivo.7–9 Interestingly, replication-deficient adenoviral vectors (ie E1−, E3−), using the Rous sarcoma virus (RSV) promoter to drive LacZ expression, have failed to transduce cochlear hair cells in vivo when approximately 107 plaque-forming units of virus were injected into scala tympani.10 In addition, the studies using adenoviral constructs to infect cochlea cells in vivo, either did not assess cochlea function, or used a functional test (ie auditory brainstem response or ABR) that evaluated principally the ascending auditory system at the auditory nerve and brainstem level, thus, only grossly examined hair cell function. In contrast, replication-deficient (E1−, E3−) adenovirus vectors, when used at higher titers (⬎108 p.f.u. per cochlea) with the CMV promoter driving green fluorescent protein expression (GFP), have been demonstrated to effectively transduce cochlear hair cells. Unfortunately, the latter is achieved at a high price, with a complete loss Modified adenovirus can transfect cochlear hair cells AE Luebke et al 790 of outer hair cell function presumably due to the toxic nature of this class of vector.11 The present study was designed to determine if a recently described class of modified adenovirus (E1−, E3−, pol−) vector could be used to efficiently deliver genes to the cochlea in vivo, without sacrificing cochlear function. This class of vector is advantageous in that it has a demonstrated improved persistence and decreased toxicity in murine models of hepatic gene transfer.12,13 Since cochlear outer hair cells are more sensitive than inner hair cells to most of the common factors that impair hearing (eg excessive sounds, drug ototoxicity, bacterial or viral diseases, genetic defects, aging, etc), DPOAE testing, which primarily assesses outer hair cell function, was used to evaluate the effects of (E1−, E3−, pol−) transduction on peripheral cochlear function. It was anticipated that the outcome of these experiments would set the course for future interventions using gene therapy to prevent permanent deafness. Results In this current study, the primary aim was to determine whether the use of an improved adenovirus vector would be capable of efficiently transducing cochlear hair cells in vivo, while preserving cochlear function. One of the principal findings of an earlier study was that an (E1−, E3−) adenovirus-based vector, when slowly infused into the scala tympani of the cochlea, was found to be capable of efficiently transducing cochlear hair cells in vivo, but also resulted in severely compromised cochlear function.11 To ensure that there were no deleterious effects on cochlear function due to the implantation of the perfusion pump, or the perfusion of substances into the cochlea, three animals were chronically infused with artificial perilymph for 8 days. These control experiments demonstrated that there was no change in cochlear function as assessed using DPOAE measures in the form of the DPgram, and shown here for a representative control guinea pig subject in Figure 1b. Here, the pre- (closed circles) and post-perfusion (open circles) status of outer hair cell function, as measured with the DP-gram, are compared. From these data it can be seen that following perfusion cochlear function was unchanged, with the 8-day postperfusion DP-gram being similar to the baseline DPgram values. Figure 2a demonstrates the effect that perfusion of the replication-deficient (E1−, E3−) adenovirus encoding -gal had on DPOAE levels for three guinea pigs. When this virus was perfused into the left ear using titers of 5 × 108 -gal-forming units (b.f.u.)/cochlea, the DPOAEs were reduced, in that by 8 days after virus infection (open symbols), there were no detectable emissions remaining, confirming our previous results with the GFP reporter adenovirus vector.11 In these ears there were no signs of local inflammatory responses or acute otitis media infections. In contrast, the non-perfused contralateral ears showed no reductions in DPOAE levels as shown in Figure 2b, suggesting that the vector-induced loss of cochlear function was a locally mediated response directed only to the perfused ear. In contrast, when a modified (E1−, E3−, pol−) adenovirus vector encoding LacZ was infused into the left scala tympani, using titers of 5 × 108 b.f.u./cochlea, of another three guinea pigs as shown in Figure 2c, there Gene Therapy Figure 1 Fluids can be systematically infused into the cochlea without compromising cochlear function. (a) Schematic showing the placement of a microcatheter into the basal turn of the left cochlea that connects to an externalized microperfusion pump. (b) DPOAE measurements in the form of a DP-gram elicited by equilevel primary tones (ie L1 = L2 = 65 dB SPL) are shown for an implanted ear of a control experiment guinea pig perfused with artificial perilymph. These data indicate that there was essentially no significant changes in DPOAE levels between the 30-min post-pump (or pre-perfusion) surgical implantation (filled circles) and 8 day postinfusion (open circles) measurements obtained after infusion. was no significant loss in DPOAEs and hence no change in cochlear function for up to 8 days (open symbols) following virus infection. As before, the related contralateral ears also showed no changes in cochlear function throughout the virus infection period, as seen by the open-symbol functions of Figure 2d. To ascertain that the modified (E1−, E3−, pol−) adenovirus actually transduced cochlear hair cells, immunohistochemistry using an antibody directed to -gal was performed on both cochleae from animals that received the modified adenovirus. These results showed that the transgene (LacZ) was detected in inner hair cells and outer hair cells as shown in Figure 3e and f from a crosssection of the cochlear whole-mount shown in Figure 3c. As the functional integrity of outer hair cells was assessed using DPOAEs, the focal plane for the whole-mount photomicrographs of Figure 3a–d are shown at this level. Outer hair cells within all frequency regions of the cochlear were clearly transduced by the modified adenovirus. Figure 3a and b display the most apical turn (turn 4) of the guinea pig cochlea, and Figure 3a provides evidence that approximately 50% of these apical low-frequency outer hair cells were transduced by the modified adenovirus, as can be visualized by the specific labeling of the Modified adenovirus can transfect cochlear hair cells AE Luebke et al 791 Figure 2 Adenovirus (E1−, E3−) infects the infused cochlea, and compromises cochlear function. (a) Equilevel DPOAE measurements (L1 = L2 = 70 dB SPL) are shown for the left (perfused) and right (non-perfused) ears of three animals that received (E1−, E3−) adenovirus at 5 × 108 b.f.u./cochlea. (a) Compared with 30 min post-pump implantation (closed symbols), by 8 days post-infusion (open symbols) there was a significant loss in DPOAEs in the infused cochlea. (b) In contrast, the non-infused ears, did not exhibit any significant losses in DPOAE levels suggesting that the reduced emissions in the perfused ears were due to a local virally mediated response. (c and d) DPOAE measurements (L1 = L2 = 70 dB SPL) are shown for the left (perfused) and right (non-perfused) ears of three animals that received (E1−, E3−, pol−) adenovirus at 5 × 108 b.f.u./cochlea. (c) A modified adenovirus (E1−, E3−, pol−) infected the infused cochlea without compromising cochlear function. Compared with 30 min post-pump implantation (open symbols), there were no significant DPOAE changes by 8 days post-infusion (closed symbols). This finding is in sharp contrast to those with the (E1−, E3−) adenovirus in which the DPOAEs were reduced and remained at noise floor levels by 8 days postinfusion (panel a). (d) As noted above, there was no significant change in the DPOAEs measured in the right (non-perfused) ear. -gal protein. Figure 3b shows the equivalent frequency region in the contralateral cochlea demonstrating no detectable -gal present in these outer hair cells 8 days after viral infection. The mid- to high-frequency regions of the cochleae showed even greater outer hair cell transduction using the modified adenovirus as shown by the micrographs of Figure 3c and d. Approximately 97% of all outer hair cells in these cochlear regions were transduced by the modi- a b c d e f Figure 3 Photomicrographs of whole mounts of -galactosidase immunostained (E1−, E3−, pol−) portions of cochlear turns at 8 days post-perfusion for gp 68 of Figure 2. (a) Adenovirus infection with LacZ expression driven by the CMV promoter stained outer hair cells (OHCs) in the lowfrequency regions (apical portions) of the cochlea. (b) No LacZ expression was detected in hair cells of the corresponding right (non-perfused) contralateral ear over the comparable low-frequency regions of the cochlea. (c) Adenovirus infection with LacZ expression driven by the CMV promoter outer hair cells (OHCs) in the high-frequency regions (basal portions) of the cochlea. (d) Again, no LacZ expression was detected in hair cells of the related right (non-perfused) contralateral ear, over the comparable highfrequency regions (basal portions) of the cochlea. (e and f) Plastic-embedded cross-section of the cochlear whole-mount shown in c. All scale bars represent 20 m unless noted otherwise. fied adenovirus. This increase in expression in the basal cochlear turns is not surprising as the virus was slowly infused into the scala tympani adjacent to the most basal turn of the cochlea. Brownian motion and the flow of perilymph from a basal to apical course through the cochlea14 could have contributed to the virus vector diffusion. Similar to the results shown in Figure 3b with an apical turn from the contralateral cochlea, there was no transgene expression detected in hair cells of the basal turn of the corresponding contralateral cochlea as shown in Figure 3d. The percentage of hair cells transduced divided by the total hair cells present are shown in Table 1 for both basal (high frequencies) and apical (low frequencies) cochlear regions. Discussion These experiments demonstrated that adenovirus-based vectors can efficiently transduce cochlear hair cells in vivo, and that ototoxicity associated with the use of early generation (E1−,E3−) adenovirus vectors can potentially be avoided with the use of (E1−, E3−, pol−) adenovirusGene Therapy Modified adenovirus can transfect cochlear hair cells AE Luebke et al 792 Table 1 Tabulation of results showing percent (%) transduction, defined as ((No. hair cells transduced/No. total hair cells) × 100) in both the high (basal) and low frequency (apical) regions of the cochlea, for both adenovirus vectors tested in this study Adenovirus type (5 × 108 p.f.u./cochlea) E1−, E3−, pol− E1−, E3− Cochlear function +++ −−− % Transduction basal cochlea (high frequencies) % Transduction apical cochlea (low frequencies) IHCs (%) OHCs (%) IHCs (%) OHCs (%) 99 13 97 — 90 2 51 0 Both adenovirus vectors were tested at equivalent titers (5 × 108 b.f.u./cochlea) with LacZ as the reporter gene, and the cytomegalovirus (CMV) promoter driving transgene expression. In all cases, artificial perilymph was used as the carrier solution. based vectors. The modified adenovirus used here was previously shown to reduce adenoviral toxicity when other quantitative measures of virus toxicity (ie liver enzymes, etc) were measured in mice that had received (E1−, E3−) and (E1−, E3−, pol−) adenovirus vectors injected directly into the bloodstream.12 The present study also demonstrated that cochlear function is a sensitive measure of virus toxicity when virus is directly infused into scala tympani. Exogenous DNA can be efficiently delivered and expressed for up to 8 days in adult cochlear hair cells using (E1−, E3−, pol−) adenovirus vectors encoding genes important for cochlear function. Longer-term studies are currently underway to ascertain how long this class of vector persists in guinea pig cochleae. Since many genes have developmental roles, adenovirus-mediated gene delivery to adult animals will facilitate experiments on gene function independent of developmental consequences. Finally, the modified adenovirus perfusion technique allows expression of a gene specifically in the cochlea, without expression in other parts of the brain, a problem that can limit the usefulness of current transgenic technologies. Previous in vitro experiments using replicationdeficient adenovirus vectors encoding either the CMV promoter driving the LacZ transgene, or the CMV promoter driving the green fluorescent protein transgene, also exhibited robust transgene expression when adenovirus vectors were used to transduce cultured hair cells.3,5 In contrast, previous in vivo studies using adenovirus-mediated gene transfer to hair cells encoding the RSV promoter driving expression of the LacZ gene failed to demonstrate efficient transduction of cochlear hair cells,10,15 which is in contrast to this report. There are a number of explanations that may account for the divergent results in this study to the previous in vivo adenovirus gene-transfer experiments. Aside from the differences in promoters used, previous adenovirus transduction attempts utilized titers of adenovirus (107/cochlea) that were at least 10-fold lower than the amounts utilized the current study. Furthermore, unpublished studies in our laboratory have confirmed that attempts of virus infection at lower titers fail to transduce cochlear hair cells in vivo. Moreover, when cochlear function was monitored using auditory brainstem responses with low titer adenovirus infection, there was also no loss of these evoked potentials, which is an indirect indication that the hair cells were not transduced.16 This outcome was in contrast to the loss of cochlear function observed by measuring DPOAEs in the current study using higher Gene Therapy titers of (E1−, E3−) adenovirus vectors that were slowly perfused into the cochlea. Finally, differences in vector delivery also may have contributed to the different outcomes between the present results and those of previous reports. For example, earlier studies used a 3-min infusion of adenovirus into scala tympani, in contrast to the slow infusion (1 l/h) of adenovirus into the scala tympani that was used here. Using this slow-infusion approach, the virus remained in the catheter for approximately 50 h at 37°C, yet there was no significant loss of titer (⬍10%) from the initial titer to the titer after 50 h at 37°C. It has been demonstrated that adenovirus transduction of cells can be greatly enhanced by increasing contact time of the virus with the cell,17 which could have also contributed to the transduction of the sensory hair cells in the current study. Previous in vivo gene-transfer studies did report transgene expression in spiral ganglion cells and cells of the cochlear aqueduct in the contralateral ear, an observation not noted in the present study.18 Possibly the slowinfusion method used here did not approach the pressure gradients associated with bolus infusions which reduces the driving force for virus to reach the cochlear aqueduct, and could account for the observed differences. A benefit of the perfused ear showing transgene expression with undetectable expression found in the contralateral ear is that this approach allows for intra-animal comparisons for controls. In addition, no transgene expression was detected in the brains of these animals which further confirms that the technique described here targets virus only to the perfused cochlea. In summary, somatic gene transfer and expression of various genes into the guinea pig cochlear hair cells over extended time periods can now be envisioned. The welldescribed characteristics of this animal model for studies of normal and abnormal hearing19,21 will facilitate experimental designs aimed at evaluating both the causes of hearing loss and the potential to treat human deafness. Materials and methods Subjects The study was performed on 10 adult pigmented guinea pigs (Ncr/2 strain) weighing 300–350 g and purchased from the Charles River Laboratories (Boston, MA, USA). The experimental design consisted of a pre/post-infusion comparison, with DPOAEs measured before and following perfusion of the adenoviral vectors into the cochlea using commercially available osmotic pump assemblies Modified adenovirus can transfect cochlear hair cells AE Luebke et al that were chronically implanted as described below. Additionally, the contralateral cochlea was examined both functionally, with DPOAEs, and immunohistochemically, for -galactosidase expression (-gal), to ensure that the perfusion was specific to the pump-implanted cochlea. Cochlear function assessment using DPOAEs Before DPOAE testing, the animals were sedated by an intramuscular (i.m.) injection of 40 mg/kg of ketamine hydrochloride and 1 mg/kg of acepromazine. DPOAEs at 2f1-f2 were elicited and measured, conventionally, using equilevel (L1 = L2) primary tones produced by ER-2 speakers and an ER-10B+ microphone assembly (Etymotic Research, Elkgrove, IL, USA). Stimulus generation and response acquisition were computer-controlled using an on-board digital signal processor, along with customized software controlling stimulus presentation and response analysis.22 DPOAEs were obtained in the form of level/frequency functions, or DP-grams (see example in Figure 1b) for geometric-mean (GM) frequencies (ie (f1 × f2)0.5 ) in 0.1-octave steps, from 1.4–17.8 kHz (f2 = 1.8–19 kHz). Primary tone-levels ranged from 45–75 dB SPL, in systematic 5-dB steps, with f2/f1 = 1.2. Adenovirus construction and titering A kanamycin-resistant shuttle plasmid vector was constructed containing the CMV promoter plus the LacZ gene within the E1 region. The shuttle plasmid was then digested with PmeI, and electroporated into the BJ5183 recombinogenic strain of E. coli with either the pAdEasy (E1−, E3−) plasmid, obtained from the Vogelstein Laboratory,23 or the (E1−, E3−, pol−) pAdEDpol plasmid,24 both of which encode ampicillin resistance genes. The pAdEDpol plasmid was constructed as follows: the NheI subfragment, encompassing the adenovirus polymerase gene of the pAdEasy plasmid, was replaced with an identical subfragment that had been previously modified to include a 608 bp deletion within the adenovirus polymerase gene. The resultant plasmid was referred to as pAdEDpol. After recombination, kanamycin-resistant clones were screened by BstXI and PmeI digestion to confirm successful generation of the respective full-length recombinant adenovirus vector genomes. (E1−, E3−) DNA was isolated and was digested with PacI and transfected into 293 cells. Similarly, the (E1−, E3−, pol−) DNA was digested with PacI and transfected into an (E1+, E3+, pol+) expressing cell line (C-7) that transcomplements the growth of the polymerase-deleted adenovirus vectors.25 Both adenovirus vectors were then amplified in their respective cell lines, and confirmed to have the correct construction by restriction enzyme mapping of the vector genomes. Two CsCl purifications were performed to generate high titer vector preparations, followed by a -gal assay using limiting dilutions, to determine the titers of both viruses in -gal forming units (b.f.u.). Osmotic-pump surgery The implant surgery, performed under aseptic conditions, secured the osmotic pump (model 2001 (1 l/h), Alza Co, Palo Alto, CA, USA) in the middle ear. Specifically, the animals were anesthetized using an intramuscular injection of 40 mg/kg ketamine hydrochloride and 5 mg/kg of xylazine. A skin incision was made initially at the midline to expose the dorsal surface of the skull. Using a dental burr, a 1.5-mm hole was drilled through the calvarium at a vertex site 1 cm posterior to the bregma suture, and a self-tapping stainless-steel screw was introduced to anchor the pump’s cannula to the skull. The cannula was fashioned according to the specifications detailed elsewhere,26 and included a drop of silicon at its distal terminus in the cochlear base. Another incision extending along the dorsal skin to a post-auricular locus, allowed the exposure of the bullar portion of the temporal bone. A small defect was created in the bulla using the tip of a scalpel blade and widened sufficiently to visualize the round window, as shown in Figure 1a. The cannula of the pump was filled with the infusion substance and clamped at the pump end. Using a fine sharpened metal probe, a small hole was created through the cochlear bone at its base, and the tip of the cannula was inserted until the silicon drop was seated securely against the bone, thus, extending the cannula about 0.5 mm into scala tympani. A drop of cyanoacrylate cement was applied at the bulla defect to anchor the cannula at this site. Following confirmation of cannula placement, the remainder of the defect was covered with carboxylate cement. To accommodate the body of the osmotic pump, a subcutaneous pocket was then made between the scapulae. Before seating the pump, its flow moderator was inserted into the cannula. To secure the cannula, its middle portion was looped around the vertex screw, cemented with methyl methacrylate to the skull and the skin incision closed. 793 Cochlear perfusions In all experiments, artificial perilymph (145 mm NaCl, 2.7 mm KCl, 2 mm MgSO4, 1.2 mm CaCl2, 5 mm HEPES buffer) was used as the carrier solution. Three guinea pigs received control cochlear perfusions with artificial perilymph only. The respective adenovirus vectors were perfused into six guinea pig cochleae using 5 × 108 b.f.u. of the virus per cochlea. All aspects of this experiment were reviewed and approved by the University of Miami’s Institutional Animal Care and Use Committee. -Galactosidase immunohistochemistry After perfusion, the presence of the transgene was detected using immunohistochemistry for the -gal protein. Immunohistochemistry was used rather than the enzyme histochemistry, since endogenous -gal activity present in the guinea pig cochlea (A Luebke, unpublished observation) prevented detection of -gal activity derived from the vector genome. In addition, enzyme histochemistry to detect LacZ gene expression has been shown to underestimate the transfection efficiency.25 Toward this end, the animals were terminated with an overdose of pentobarbital, and perfused with 4% paraformaldehyde. Both cochleae were then harvested and decalcified in 0.1 m EDTA. Following decalcification, the cochlear halfturns were microdissected and processed to detect the presence of the -gal using an anti--gal antibody (Promega, Madison, WI, USA). Cochlear half-turns were incubated overnight in the -gal antibody (1:1000) in 0.01 m PBS at room temperature. After washing with PBS, the tissues were incubated for 1 h with a donkey-anti-rabbit secondary antibody (1:800) conjugated to biotin (Jackson Immunoresearch Laboratories, West Grove, PA, USA), and then washed and incubated in ABC reagent for 1 h Gene Therapy Modified adenovirus can transfect cochlear hair cells AE Luebke et al 794 (Vector Laboratories, Burlingame, CA, USA), according to the manufacturer’s specifications. Finally, the tissues were incubated in diaminobenzidine, cleared in glycerol, and mounted on depression slides for light-microscopy viewing. To obtain cross-sections of these whole mounts, the sections were embedded in plastic (araldite) and cut into thick sections following the methodology described by Bohne.28 11 12 13 Acknowledgements This work was supported by grants from the Public Health Service DC03086 (AEL), DK52925 (AA), Muscular Dystrophy Association, USA (AA), and funds from the University of Miami’s Stanley Glaser Research Foundation and Chandler Chair (AEL). We would like to thank Dr Ken Muller for his assistance with plastic sectioning. 14 15 16 17 References 1 Yoshida N et al. Acoustic injury in mice: 129/SvEv is exceptionally resistant to noise-induced hearing loss. Hear Res 2000; 141: 97–106. 2 Zuo J, Treadaway J, Buckner TW, Fritzsch B. Visualization of alpha 9 acetylcholine receptor expression in hair cells of transgenic mice containing a modified bacterial artificial chromosome. Proc Natl Acad Sci USA 1999; 96: 14100–14105. 3 Dazert S, Battaglia A, Ryan AF. Transfection of neonatal rat cochlear cells in vitro with an adenovirus vector. Int J Dev Neurosci 1997; 15: 595–600. 4 Staecker H, Gabaizadeh R, Federoff H, Van De Water TR. Brainderived neurotrophic factor gene therapy prevents spiral ganglion degeneration after hair cell loss. Otolaryngol Head Neck Surg 1998; 119: 7–13. 5 Holt JR et al. Functional expression of exogenous proteins in mammalian sensory hair cells infected with adenoviral vectors. J Neurophysiol 1999; 81: 1881–1888. 6 Van de Water TR, Staecker H, Halterman MW, Federoff HJ. Gene therapy in the inner ear. Mechanisms and clinical implications. Ann NY Acad Sci 1999; 884: 345–360. 7 Derby ML, Sena-Esteves M, Breakefield XO, Corey DP. Gene transfer into the mammalian inner ear using HSV-1 and vaccinia virus vectors. Hear Res 1999; 134: 1–8. 8 Hann JJ et al. Transgene expression in the guinea pig cochlea mediated by a lentivirus-derived gene transfer vector. Hum Gene Ther 1999; 10: 1867–1873. 9 Luebke AE, Peel AL, Muller CD, Foster PK. DPOAE function and transgene expression in guinea pig cochlea using AAVdirected gene-transfer methods. Assoc Res Otolaryngol Abstr 1999; 22: 81. 10 Raphael Y, Frisancho JC, Roessler BJ. Adenoviral-mediated gene Gene Therapy 18 19 20 21 22 23 24 25 26 27 28 transfer into guinea pig cochlear cells in vivo. Acta Otolaryngol 1996; 116: 125–131. Luebke AE, Steiger JD, Hodges BL, Amalfitano A. Cochlear function during (E1-) and (E1, E2b-) adenovirus transduction of guinea pig hair cells. Mol Ther 2000; 1: 580–581. Hu H, Serra D, Amalfitano A. Persistence of an (E1−, polymerase−) adenovirus vector despite transduction of a neoantigen into immune-competent mice. Hum Gene Ther 1999; 10: 355–364. Hodges BL et al. Multiply deleted (E1, polymerase−, and pTP−) adenovirus vector persists despite deletion of the preterminal protein. J Gene Med 2000; 2: 250–259. Salt AN, Inamura N, Thalmann R, Vora AR. Evaluation of procedures to reduce fluid flow in the fistulized guinea-pig cochlea. Acta Otolaryngol 1991; 111: 899–907. Yagi M et al. Hair cell protection from aminoglycoside ototoxicity by adenovirus-mediated overexpression of glial cell linederived neurotrophic factor. Hum Gene Ther 1999; 10: 813–823. Stover T, Yagi M, Raphael Y. Cochlear gene transfer: round window versus cochleostomy inoculation. Hear Res 1999; 136: 124– 130. Mittereder N, March KL, Trapnell BC. Evaluation of the concentration and bioactivity of adenovirus vectors for gene therapy. J Virol 1996; 70: 7498–7509. Stover T, Yagi M, Raphael Y. Transduction of the contralateral ear after adenovirus-mediated cochlear gene transfer. Gene Therapy 2000; 7: 377–383. Dallos P. Membrane potential and response changes in mammalian cochlear hair cells during intracellular recording. J Neurosci 1985; 5: 1609–1615. Preyer S et al. Frequency response of mature guinea-pig outer hair cells to stereociliary displacement. Hear Res 1994; 77: 116– 124. Cheatham MA, Dallos P. The dynamic range of inner hair cell and organ of Corti responses. J Acoust Soc Am 2000; 107: 1508–1520. Martin GK et al. Locus of generation for the 2f1-f2 vs 2f2-f1 distortion-product otoacoustic emissions in normal-hearing humans revealed by suppression tuning, onset latencies, and amplitudes correlations. J Acoust Soc Am 1998; 103: 1957–1971. He TC et al. A simplified system for generating recombinant adenoviruses. Proc Natl Acad Sci USA 1998; 95: 2509–2514. Amalfitano A et al. Production and characterization of improved adenovirus vectors with the E1, E2b, and E3 genes deleted. J Virol 1998; 72: 926–933. Amalfitano A, Chamberlain JS. Isolation and characterization of packaging cell lines that coexpress the adenovirus E1, DNA polymerase, and preterminal proteins: implications for gene therapy. Gene Therapy 1997; 4: 258–263. Prieskorn DM, Miller JM. Technical report: chronic and acute intracochlear infusion in rodents. Hear Res 2000; 140: 212–215. Couffinhal T et al. Histochemical staining following LacZ gene transfer underestimates transfection efficiency. Hum Gene Ther 1997; 8: 929–934. Bohne BA. Location of small cochlear lesions by phase contrast microscopy prior to thin sectioning. Laryngoscope 1972; 82: 1–16.