SEARCH

SEARCH BY CITATION

Keywords:

  • cochlea;
  • gene therapy;
  • semicircular;
  • canal;
  • vector delivery

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. LITERATURE CITED

This article presents a unique approach for the delivery of gene therapy vectors into the cochlea of the laboratory rat. Mice and guinea pigs are established in vivo models for cochlear gene therapy each of which has distinct advantages and disadvantages. The rat has some of the molecular advantages of a mouse model combined with size advantages for surgical approaches. Vector delivery via cochleostomy or injection through the round window causes concomitant sensorineural hearing loss and is therefore not suitable for studies where the change in hearing is being followed. Compared to the mouse, the rat does not demonstrate easily recognizable landmarks that allow for use of the semicircular canal as an approach to the inner ear. We analyzed sagittal and coronal temporal bone sections of Long Evans rats and identified the bony entrance of the facial nerve as a crucial landmark for canalostomy. Auditory brainstem response and distortion product otoacustic emission measurements revealed minimal differences in the hearing threshold after adenovirus vector application when large volumes of vector were infused to the inner ear. Canalostomy and infusion of adenoviral vectors also resulted in temporary balance disturbance in the rat. Immunohistochemical assessment after delivery of a green fluorescent protein expressing vector showed significant GFP expression in the cochlea. Anat Rec, 2012. © 2012 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. LITERATURE CITED

In vivo gene transfer to the mammalian inner ear has been achieved in different rodent models (Maeda et al., 2009). Direct approaches to the perilymphatic system via cochleostomy or round window membrane injection have compromised hearing unless delivery volume is minimized (Praetorius et al., 2003). The posterior semicircular canal (PSCC) in mice instead is relatively easy to approach surgically and inoculation of adenovirus through canalostomy efficiently introduces transgenes to cochlea and vestibular system without any hearing loss (Jero et al., 2001; Kawamoto et al., 2001). Compared to the mouse model, the rat does not demonstrate easily recognizable landmarks that allow for the use of the semicircular canal as an approach to the inner ear. The vestibular organs and the cochlea are embedded in a thick bone that makes approaches to the vestibular labyrinth of the rat challenging.

Developing the rat as a model for inner ear gene therapy is vital because of the spectrum of behavioral and electrophysiology studies that can be performed on the rat. The study of tinnitus in particular has focused on rat models (Turner et al., 2006; Ralli et al., 2010). (Zhang and Kaltenbach, 1998). (Imig and Durham, 2005).

Gene therapy has the potential for altering the function of the inner ear, allowing analysis of central auditory function in this model. A prerequisite for this is achieving delivery of vector to the inner ear without altering hearing. The surgical anatomy of the rat middle ear is quite challenging. The internal carotid artery (stapedial artery) courses directly through the stapes crura as it passes from the neck into the brain. Exposure of stapes and oval window often necessitate trauma with destruction of the remaining ossicular chain, tympanic membrane, or cochlea and thereby compromising auditory function (Judkins and Li, 1997). Previously, a utricular application of pHSVlacZ by opening the bulla has been described and proved efficient. β-galactosidase reporter gene was in nearly all neurons of vestibule and cochlea (Praetorius et al., 2002).

Our surgical approach avoids opening of the middle ear and identifies the exit of the facial nerve from the temporal bone as landmark for the canalostomy. Preoperative and postoperative auditory brainstem response (ABR) audiometry and distortion product otoacustic emissions (DPOAEs) indicated no hearing loss after delivery of vector to the inner ear.

Adenoviral and adenovirus-associated vectors (AAV)-mediated gene transfer are upcoming therapies for protection and regeneration of auditory and vestibular hair cell function (Huang et al., 2009; Husseman and Raphael, 2009; Luebke et al., 2009). We propose, that the rat is an appropriate animal model for examining central auditory pathway effects of cochlear gene therapy.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. LITERATURE CITED

Animals

Three-month-old male Long Evans rats (n = 14) were purchased from the Jackson Laboratory (Bar Harbor, ME). All experimental protocols were approved by the Institutional Animal Care and Use Committee at the University of Kansas Medical Center. Adult Long Evans rats (n = 4) were anesthetized with an i.m. injection of a mixture of ketamine (50 mg/kg), atropine (0.05 mg/kg), and xylazine (10 mg/kg) in sodium chloride 0.09%.

Serial Sagittal Temporal Bone Sections

Animals were anesthetized with a mixture of phenobarbital (5.8 g/kg) and phenytoin sodium (750 mg/kg) (Beuthanasia-D-Special) and sacrificed via intracardiac perfusion with 4% paraformaldehyde in phosphate buffered saline (PBS). The temporal bones were removed and postfixed in 4% paraformaldehyde in PBS at 4°C overnight. After rinsing in PBS three times for 30 min, the temporal bones were decalcified in 10% ethylenediaminetetraacetic acid (EDTA) for 5 days. The specimens were rinsed in PBS, rehydrated, and cleared. Finally, the temporal bones were embedded in gelatin 20% in double distilled water (DDH2O). The blocks were frozen at −20°C and cut in 100 μm frozen serial sections in sagittal orientation. The sections were submerged as free-floating in PBS in a 12-well plate and then dehydrated in two cycles of 70 and 95% ethyl alcohol (ETOH) for 10 min. Each section was then stained with eosin for 90 sec. The specimens were rehydrated subsequently in 95% ETOH, 70% ETOH, and DDH2O for 15 min each. After clearing, the free-floating sections were photographed under a Nikon SMZ 1500 Stereo Microscope.

Delivery of Vector to the PSCC

To minimize trauma to the cochlea, the vector was delivered to the PSCC. A dorsal postauricular incision was made to expose the cartilaginous external auditory canal (EAC) and the dorsally located facial nerve. Muscular tissue was separated and minor bleeding stopped with the bipolar forceps. The facial nerve was followed to his exit in the temporal bone. Unlike in the mouse, there are no clear external markings for the PSCC in the rat. Analysis of sagittal temporal bone sections suggested that the exit point of the facial nerve served as an approximate marker for the later most extension of the posterior semicircular canal. Bone was carefully removed posterior and perpendicular to the facial nerve with a 1-mm diamond burr. The leakage of perilymph usually indicated a successful canalostomy. Subsequently, 5 μL of vector was injected using a 30# Hamilton syringe with 0.1 μL graduations and a 36 gauge needle. Injections consisted of 5.0 × 108 particle units (pu) of Ad28hcMV.EGFP. Pieces of muscular tissue and blood were used to seal the canalostomy. The animals were allowed to recover for 48 hr and successfully injected rats showed a brief period of circling behavior postsurgery.

Vector Production

Ad28 serotype vectors were used for all experiments. The adenovector backbone used for all experiments was deleted of adenovirus regions E1A, E1B, E3, and E4. The production system for these adenovectors provides robust replication of the adenovector and purified stocks at 5 × 1011 to 2 × 1012 total pu/mL, with a total particle to active particle ratio ranging from 3 to 10 pu/fluorescent focus unit. Total particles were determined by spectrophotometric assay that has been standardized and qualified to reliably and robustly quantify the total particles within a single lot of adenovector. Adenovector lots were purified, aliquoted, and stored at −80°C. Individual aliquots were used for each experiment to prevent loss of activity associated with freeze-thaw cycles. Expression was driven by a hcmv promoter, and the expression cassette contained an optimized artificial splice site at the 5′ end of the open reading frame and a SV40 polyadenylation site and transcriptional stop site at the 3′ end of the open reading frame (GenVec, Gaithersburg, MD).

Immunohistochemistry

Gelatin sections (10 μm) cut parallel to the modiolus were mounted on silane-coated slides (Sigma-Aldrich, St. Louis, MO) and dried overnight at 40°C. The slides were then rehydrated in PBS for 30 min at same temperature to dissolve excessive gelatin. After rehydrating the samples in PBS, sections underwent antigen retrieval in a microwave using 1:10 Dako® Target Retrieval Solution (DAKO Corporation, Carpinteria, CA) for 30 sec. The sections incubated for 30 min at room temperature in blocking solution, containing 0.3% Triton X 100 and 10% FBS in PBS. After rinsing another time in PBS, the sections were treated with myosin VIIa rabbit polyclonal antibody (Proteus® BioSciences, Ramona, CA) for inner hair cell staining and anti-GFP (Invitrogen®, Carlsbad, CA) for staining of vector transfected supporting cells. Both primary antibodies were diluted 1:50. The tissue was incubated overnight at 4°C in a humid chamber. After three rinses in PBS, immunohistochemical detection was carried out with an anti-rabbit IgG (1:50; Alexa Fluor 555 nm; Invitrogen®) and anti-chicken IgG (1:50; Alexa Fluor 488 nm; Invitrogen®). The secondary antibody incubated for 3 hr at room temperature in a humid chamber. After a final step of rinsing, the sections were mounted with Prolong Gold antifade reagent containing DAPI and then coverslipped. Microscopic pictures were taken under a Nikon Confocal Microscope C1 Si (Nikon®, Kanagawa, Japan).

ABR and DPOAE Measurement

ABRs were measured on the left ear of each animal using the Smart EP Program (Intelligent Hearing Systems Corp, Miami, Florida). Animals were anesthetized as described above and kept warm on a heating pad (37°C). Needle electrodes were placed on the vertex (+), behind the left ear (−), and behind the opposite ear (ground). Tone bursts were presented at 2, 4, 8, 16, 32 kHz, with a duration of 500 μs using a high-frequency transducer. Recording was carried out using a total gain equal to 100 K and using 100-Hz and 15-kHz settings for the high-and low-pass filters. A minimum of 128 sweeps were presented at 90 dB sound pressure level (SPL). The SPL was decreased in 10-dB steps. Near the threshold level, 5-dB SPL steps using up to 1024 presentations were carried out at each frequency. Threshold was defined as the SPL at which at least one of the waves could be identified in two or more repetitions of the recording. Hearing measurements were generally performed before surgery and 48 hr post vector delivery. To evaluate the functional damage on outer hair cells (OHC), DPOAEs were recorded on the left ear using the IHS Program described above. The distortion products were measured for pure tones from 2 to 32 kHz using a high-frequency transducer. The Etymotic 10B + Probe was inserted to the EAC. L1 level was set to 65 dB and L2 level to 55 dB. Frequencies were acquired with an F2–F1 ratio of 1.22 using 16 sweeps. DPOAEs were presented as response dB-SPL to different frequency levels.

Statistics

One-way ANOVA and two-way repeated-measures ANOVA with a Scheffe post hoc test were used to calculate P values. P values of < 0.05 were considered significant. Statistical analysis was performed with SPSS 17.0 Version for Windows XP.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. LITERATURE CITED

Anatomy of the Rat Temporal Bone

Unlike mice, rats do not have a prominent PSCC at on the surface of the temporal bone. To determine the optimal approach to the rat vestibular system, we performed serial sections of the temporal bone in sagittal plane. As can be seen in Fig. 1B–D, the PSCC lies near to the surface of the temporal bone and can be traced back to into the vestibule (Fig. 1G). The most consistent landmark defining the position of the semicircular canal in the rat is the facial nerve (N.VII; in Fig. 1A marked with red arrow), which courses outside the temporal bone in close proximity to the EAC. The eustachian tube (ET) lies superior to the PSCC and consists of a bony and a cartilaginous part (Fig. 1C–F).

thumbnail image

Figure 1. Serial sagittal sections of a right rat temporal bone photographed at 40x magnification. The temporal bone was embedded in 20% gelatin, stained with H&E, and cut in 100 μm sections in sagittal plane to mimic surgical orientation. The facial nerve (N.VII) exits dorsal of the EAC, the temporal bone (A). Through the sequentially orientated sagittal sections, the PSCC (PSC) can be traced lateral and superior to the facial nerve (between green arrows in BD) and followed on its way into the vestibule (G). The ET is located superior to the labyrinth. Scale bar = 2.5 mm.

Download figure to PowerPoint

Surgical Approach to the PSCC

Based on the anatomical position of the facial nerve, we developed a surgical approach to the semicircular canal (Fig. 2). The hair behind the left ear was shaved and the skin cleaned with the use of local antiseptic. Thereafter, a semilunar incision, 15 mm in length, was made in close proximity to the pinna. The wound was opened with the use of a small retractor and any injury of the ear canal avoided. Fascia was dissected away to get a direct visualization of the EAC and the temporalis muscle. The facial nerve (N.VII) could be traced in the upper angle formed by the EAC and the temporalis (Fig. 1A). The temporalis muscle was transected dorsal to the facial nerve and hemostatsis was obtained with bipolar coagulation. The inferior portion of the temporal bone was then exposed. Using a 1-mm diamond burr, bone was drilled posterior and perpendicular to N.VII, until the posterior canal was open (Fig. 2C,D). Generally, the correct position of the drilled hole was confirmed by a leakage of perilymph. Thereafter, 5 μL (5 × 108) of Ad28hCMV EGFP was injected using a 36# Hamilton syringe (Fig. 1E,F). The canalostomy was sealed with a piece of muscle and then the wound sutured.

thumbnail image

Figure 2. Surgical approach to the PSCC on the left ear. After exposure via a postauricular incision, the facial nerve (N.VII) is traced within a triangle formed by the EAC and the temporalis muscle (A). Partial transection and removal of the temporalis muscle reveals the temporal bone (B) and the exit point of the facial nerve (white arrow) (B). Using a 1-mm diamond burr, bone is carefully removed posterior and perpendicular to N. VII until the canal is open and perilymph leaks (C, D). About 5 μL of Ad28hCMV.EGFP is injected into the perilymph with a 30# Hamilton syringe (E). After injection (F), the canalostomy is sealed with a piece of muscle and blood and the animal is allowed to recover for 48 hr. Animals that are successfully injected show a brief period of circling behavior in the postoperative period. Scale bar = 1 mm A, B, C, E, and F; 0.2 mm D.

Download figure to PowerPoint

Mortality, Morbidity, and the Effect of Surgery upon the Rat

Ten animals in total were treated with Ad28hCMV.EGFP. Postoperative assessment of hearing function happened 48 hr after vector delivery and animals were then euthanized for histological processing. After surgery most rats showed a slight tilted head with inclination to the operated ear, probably due to the partial removal of the temporalis muscle. All animals were able to eat and drink independently. Successful canalostomy and vector application was also confirmed by postoperative vestibulopathy with a short period of circling behavior and nystagmus.

Comparison of Preoperative and Postoperative Hearing Function

Distortion product otoacoustic emissions (DPOAE) and absolute hearing threshold in ABR audiometry were measured before and 2 days after vector delivery. Canalostomy did not change function of OHC or absolute thresholds (Fig. 4A,B). Using two-way repeated-measures ANOVA and a Scheffe post hoc test, there was no statistical differences in preoperative and postoperative auditory function for ABR thresholds (Fig. 1A). Nonetheless, the statistical analysis revealed a significant difference in ABR thresholds and DPOAE responses across frequencies. ABR thresholds (Fig. 3A) are higher at 2 kHz than at 4 kHz (**P = 0.003), 16 kHz (*P = 0.017), and 32 kHz (**P = 0.008). DPOAE responses (Fig. 3B) at the F2 frequency 2.2 kHz are lower than at 6.2 kHz (**P = 0.001), 8.8 kHz (**P < 0.001), 12.5 kHz (**P < 0.001), and 17.6 kHz (P = 0.001). Otoacoustic emissions at the F2 frequency 35.3 kHz are also significantly lower than at 8.8 kHz (*P = 0.012) and 12.5 kHz (*P = 0.015). Those significant differences across frequencies might also be a good indicator for preciseness and reliability in our measurements of hearing function.

thumbnail image

Figure 3. Comparison of ABR thresholds (A) and DPOAE responses (B) before and after vector delivery. Using two-way repeated-measures ANOVA, there was no statistical difference in between groups before and after surgical application of a Ad28hCMV EGFP vector. Canalostomy and application of vector did not change hearing thresholds and DPOAE responses. Nonetheless, ANOVA statistical analysis and the Scheffle post hoc test revealed a significant difference in ABR hearing thresholds and DPOAE responses across frequencies in both treatment groups. ABR thresholds (A) are higher at 2 kHz than at 4 kHz (**P = 0.003), 16 kHz (*P = 0.017), and 32 kHz (**P = 0.008). DPOAE responses (B) at the F2 frequency 2.2 kHz are significantly lower than at 6.2 kHz (**P = 0.001), 8.8 kHz (**P < 0.001), 12.5 kHz (**P < 0.001), and 17.7 kHz (**P < 0.001). The response levels at the F2 frequency 35.3 kHz are also significantly lower than at 8.8 kHz (*P = 0.012) and 12.5 kHz. (*P = 0.015).

Download figure to PowerPoint

thumbnail image

Figure 4. Distribution of EGFP in the inner ear 48 hr after delivery of Ad28eGFP via canalostomy. Coexpression of EGFP (green) and Myosin VII (red) were visualized under a confocal microscope (A). Supporting cells surrounding inner and OHC in the organ of corti showed abundantly EGFP expression 2 days after vector application (A, B). A slightly weaker expression pattern could be seen in utricle (not shown), saccule (C), and ampulla (D). Phase contrast microscopy accentuated cell borders between vestibular hair cells and supporting cells in the saccule (E) demonstrating GFP expression limited to the vestibular neuroepithelium.

Download figure to PowerPoint

Characterization of GFP Transgene Expression Post Canalostomy

Immunohistochemical staining allowed analysis of GFP expression post Ad28hCMV vector inoculation (Fig. 4A–E). Also, the relationship of transfected cells to vestibular and cochlear hair cells could be determined. However, the most abundant expression of green fluorescent protein (GFP) was detected in the stria vascularis (not shown). Cytoplasm of supporting cells in the organ of Corti appeared intensively green fluorescence.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. LITERATURE CITED

In this study, we identified the exit of the facial nerve from the temporal bone as a crucial landmark for the canalostomy of the PSCC in the rat. Injection of adenovector serotype 28 via canalostomy successfully transfected vestibular and cochlear supporting cells with enhanced GFP (EGFP). The surgical approach via canalostomy and subsequent vector administration were able to preserve hearing function.

Transgene expression using adenoviral or AAV has been demonstrated in inner ears of guinea pigs and mice. Cochleostomy of the basal turn in guinea pigs and treatment with different recombinant AAV serotypes expressed hCMV-promoter driven EGFP transgene in inner and OHC, Hensen's cells, spiral limbus, and ligament. Adenoviral-mediated transduction of LacZ in the same species expressed β-galactosidase in connective tissue within the spiral ligament and fibrocytes lining the perilymphatic fluid spaces. No significant threshold shifts between artificial perilymph and AAV vectors were observed 7 days postoperatively (Konishi et al., 2008). The guinea pig temporal bone houses a relatively large labyrinth, which is completely covered by a bony lamina. The superior and lateral semicircular canals are easily accessible from the dorsal bulla, (Wysocki, 2005) but canalostomy has not been described yet in the species. The mouse model has the advantage of a well-studied genome. The small size of the inner ear poses a particular challenge for surgical procedures, but canalostomy in mice appears effortless due to the exposed position of the PSCC under the postauricular neck muscles. Studies have shown that vector administration via a canalostomy in the PSCC yields better hearing preservation than administration via a cochleostomy (Praetorius et al., 2003). However, transfection in the cochlea is less after the canalostomy approach compared to cochleostomy. Adenoviral vector Ad.CMV-lacZ application expressed transgene in mesothelial cells lining the perilymphatic space of the PSCC, utricle, saccule, and cells close to Reissner's membrane (Kawamoto et al., 2001). Previously, a vestibular approach for gene transfer via utriculostomy has been described in rats but necessitated opening of the middle ear. Treatment with a HSV-1- derived amplicon vector expressed lacZ in cochlear spiral ganglion cells, the organ of corti, and vestibular neurons (Praetorius et al., 2002). Our study showed a different transgene distribution pattern for inoculation with the Ad28hCMV EGFP vector. EGFP was expressed in vestibular and cochlear supporting cells, Deiter's cells, stria vascularis, and spiral ganglion cells.

Rats represent an intermediate model that allows electrophysiology recordings on the central auditory pathway and behavioral testing in scientifically established procedure.

Vestibular surgical approaches with the intention to preserve hearing necessitated up to date the opening of the bulla and therefore have a greater risk of injury for middle-ear structures such as the ossicular chain or the internal carotid (stapedial) artery.

In an endolymph delivery simulation model, dissected from P1–P4 Sprague Dawley rats, atoh 1 gene transfer induced ectopic vestibular hair cell-like cells (Huang et al., 2009). Thus, future perspectives of our developed surgical approach in the rat model may lie in cochlear gene therapy and the analysis of its impact on remodeling at the level of the central auditory pathway.

LITERATURE CITED

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. LITERATURE CITED