2.1 Ethical statement
The animal experiments described below complied with the ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines. Care and use of the animals followed the NIH Guide for Care and Use of Laboratory Animals guidelines34. All procedures were approved by The Institutional Animal Care and Use Committee (IACUC) at Northwestern University (IS00008596 and IS00008707) and by the Navy Bureau of Medicine and Surgery (BUMED)
2.2 Animals
Ten-week-old CBA/CaJ male mice (stock number 000654) were shipped from Jackson Laboratory (Bar Harbor, ME) to the Center for Comparative Medicine at Northwestern University, Chicago. The mice were acclimated for two weeks in the facility before surgery. Three animals were housed together in an autoclaved, individually ventilated cage of 36 cm (L) x 18 cm (W) x 12 cm (H), having wood shavings and cotton pads for the nest and hiding place building. The light cycle was 14 hours, followed by 10 hours of dark. The lights were “on” at 6 am and “off” at 8 pm. Water and Teklad (Envigo, Indianapolis, IN) LM-485 Mouse/Rat sterilizable diet pellets were given ad libitum. Post-surgery, mice were individually housed to prevent chewing and scratching at the sutures.
2.3 Osmotic pump implantation
2.3.1 The catheter for the osmotic pump
The catheter inserted into the cochlea was custom fabricated to reduce the tubing diameter from 1 mm at the port of the osmotic pump to 130 µm, where it enters the cochlea. Supplemental Figure 1 shows the fabrication steps (Supplemental Table 1 lists the materials and sources). A 2 cm long segment of polyimide tubing (Micolumen, Oldsmar, FL), 130 µm outer and 100 µm inner diameter, was inserted into the Tygon® ND100-65 tubing (Saint Gobain Performance Plastic, Akron, OH), 840 µm inner diameter. The connection was sealed with SILASTIC® MDX4-4210 medical-grade elastomer mixed with its curing agent (Dow Corning, Midland, MI). The fabricated catheters were then placed in an oven for overnight polymerization to form the silicone rubber at 50°C (Supplemental Figure 1). Before sterilization by ethylene oxide gas exposure, the quality of each catheter was individually tested (Supplemental Figure 1). The catheter length was adjusted for proper pump placement when the pump and its catheter were used in the surgery.
2.3.2 Materials required and preparation of the surgery
Before the surgery, all instruments (Supplemental Figure 2 and Supplemental Table 2) were autoclaved. Catheters that connect the osmotic pump and the cochlea were sterilized using ethylene oxide gas (Supplemental Table 3). We also gathered the materials required for anesthesia (Supplemental Table 4) and collected the disposable and miscellaneous items (Supplemental Table 5) for the surgery.
2.3.3 The surgical team
A two-person team was optimal for the procedure; one person was trained in performing surgeries under aseptic conditions. The second, a support person, was responsible for inducing the anesthesia, mounting the mouse in the head holder, shaving, and cleaning the surgical field, keeping medical records, monitoring, and adjusting the anesthesia, and supervising the animal's recovery after surgery.
2.3.4 Animal anesthesia
For pain management, we injected a subcutaneous dose of 0.05 mg/kg immediate-release Buprenorphine analgesic35,36 at least an hour before the induction of the anesthesia, followed thirty minutes later by an intraperitoneal injection of 1 mL of 0.9% NaCl United States Pharmacopeia (USP) solution for hydration. The anesthesia was induced with 3% of the inhalation anesthetic isoflurane in 0.3 L/min oxygen by placing the mouse into a commercially available induction box (23 cm (length) x 10 cm (width) x 10 cm (height)). The mouse was then transferred from the induction box to a water-based heating pad covered with a paper underpad and placed into a custom-made head holder. Isoflurane, 1-3% in 0.3 L/min oxygen, delivered via a nose cone, was used to maintain the level of anesthesia (Figure 1(a) and Supplemental Figure 3).
2.3.5 Preparation of the surgical field
With a commercially available Wahl clipper (Supplemental Table 2), the animal’s fur was removed well beyond the expected incision line (Figure 1(a)). The area was cleaned three times with betadine solution and 75% alcohol in alternate order, wiping the cleaning fluid away from the center of the surgical field. Drapes under the animal and drapes covering the nose cone and exposed animal parts created a sterile field. Towel Clamps (Supplemental Figure 2 and Supplemental Table 2: Tool #1) held the drapes in place.
2.3.6 Access to the bulla
Figure 1 shows in detail the surgical access to the bulla. First, in the supine position, the mouse is mounted to the nosecone of the custom-made head holder (Figure 1(a)). The surgery begins with an approximately 2 cm skin incision using a small pair of sharp scissors (Supplemental Table 2: Tools #2 and #3) starting close to the right shoulder, lateral from the midline, towards the mandible along the dashed white arrow (Figure 1(a)). Using blunt dissection, the tissue over the submandibular glands (SMGs) was removed (Figure 1(b)), and the anterior belly of the digastric muscle (AD) became visible. The white dashed line in Figure 1(b) highlights the boundary between the left and right SMG, along which they were separated (Figure 1(c), Supplemental Table 2: Tool #2). Retracting the right SMG exposed the entire digastric muscle with its anterior (AD) and posterior bellies (PD). The PD was the landmark to approach the bulla. The white arrow in Figure 1(c) points to the location of the bulla below the PD. Scissors were not used for the surgery's following steps to minimize the risk of cutting small blood vessels. Instead, the tissue was dissected carefully with sharp forceps (Supplemental Table 2: Tools #4 and #5) to expose the digastric muscle's origin. Figure 1(d) shows the bulla after elevating the PD; the bulla and outer ear canal are visible (dashed box, Figure 1(d) and Supplemental Table 2: Tools #4 and #5). It is important not to manipulate the facial nerve (thick arrow). After the dissection, the muscles were retracted with a custom-made tool (Supplemental Figure 2(c), Supplemental Table 2: Tool #6). Careful muscle dissection with pointed forceps exposed the bulla (Figure 2(a)-(b) and Supplemental Table 2: Tools #4 and #5).
2.3.7 Opening the bulla and creating the cochleostomy
The surgical bullostomy and cochleostomy are made using a sharpened pick, motorized drill with a burr, or a CO2 laser. Any tools will work for the bullostomy (Figure 2(c)), but the cochleostomy is more challenging (Figure 2(d)-(e)-(f)). The burr and the pick often impede the cochlea's direct view, frequently resulting in a much larger cochleostomy than required to insert the osmotic pump’s catheter. In contrast to the burr, the CO2 laser allows adequate visual control over the cochleostomy site while creating the opening (Figure 2(f)). However, using the laser bears the risk of nicking the stapedial artery. To use the pick for making or enlarging the cochleostomy (Supplemental Figure 2 and Supplemental Table 2: Tool #7), the tip of the pick should be less than 100 µm in diameter.
2.3.8 ALZET® pump placement
Supplemental Figure 4 describes the intraoperative catheter and osmotic pump assembly, and Figure 3 the surgical placement. For the pump assembly, the catheter, the osmotic pump, and its access port were placed (Supplemental Figures 4(a)-(b) on the sterile drape. The ALZET® pump was loaded with the drug using the filling needle. The catheter was shortened (Supplemental Figure 4(b)) and connected to the access port (Supplemental Figure 4(c)). Before joining the catheter to the pump, it was flushed with the drug, ensuring no air bubbles had been introduced (Supplemental Figures 4(d)-(e)). The fluid drop at the catheter's tip (Supplemental Figure 4(f)) confirms the pump's successful assembly. The assembled ALZET® pump was inserted into a pocket under the skin over the thorax (Figure 3(a)), which was made by blunt dissection with a pair of pointed scissors. Two pairs of forceps (Supplemental Figure 2 and Supplemental Table 2; Tools #4 and #5) were used to manually insert the pump towards the shoulder blades and push it into its final position. Since the osmotic pump was placed on the animal’s back, its catheter was positioned below the sternocleidomastoid muscle (Figures 3(b) and 3(c)). After tunneling the catheter, its tip was gently bent (Figures 3(b)-(c)) and inserted with a pair of fine forceps (Supplemental Figure 2 and Table 2; Tool #5) through the cochleostomy (Figure 3(c)). The cochleostomy was slightly larger than the catheter's outer diameter. Additional tissue was required to seal the opening.
The cochleostomy's exact location and the catheter orientation determined whether the catheter’s orifice was toward scala tympani or scala vestibuli. Cochleostomy sites closer to the stapes and the catheter's orientation toward the cochlear base placed the tubing tip into scala vestibuli. In contrast, cochleostomy locations closer to the stapedial artery and the bulla's medial wall with catheter orientations toward the bulla's medial wall placed the catheter's tip into the scala tympani. In some experiments, we confirmed the catheter tips' locations through micro-computed tomography with synchrotron radiation (see the section below: Image acquisition and tomographic reconstruction, Figure 4(a)). After placing the catheter, we stabilized it in its position by filling the middle ear cavity with dental acrylic using a 1 mL syringe and 22G needle (Figure 3(e)). During the 5 minutes of acrylic polymerization, we covered the bulla with the SMG and skin to limit desiccation (Figure 3(f)).
2.3.9 Skin closure and suture
The skin incision was closed with two layers using interrupted sutures (Supplemental Figure 2 and Supplemental Table 2: Tool #8 or #11 and #9). For the first layer, we used 6-0 (VICRYL) absorbable suture material and 6-0 (ETHICON) non-resorbable suture for the skin. During this last step, the percentage of isoflurane was decreased to a minimum of 1-1.5% under continuous monitoring of the mouse’s anesthesia level. The closed incisions were cleaned with two alternating swaps of Betadine solution and alcohol (75%). The anesthetic delivery was discontinued, and the animal was removed from the head holder and placed in the supine position in a recovery box on top of a heating pad. The animal was under continuous supervision until sternal and freely ambulant. The recovery time was, on average, 6 minutes (min) and 52 seconds (s) (2 – 19 min, n = 45) after stopping the flow of isoflurane (Figure 5). The pump was visible on the mouse's back but did not affect the animal’s movements and activity (Figure 4B, Supplemental Video 1).
2.3.10 Post-surgical pain management
For post-surgical pain management, we injected Buprenex subcutaneously (SC) every 12 hours over 48 hours.
2.3.11 Post-surgical animal handling
Animal handling after the surgery was gentle to avoid dislocation and damage to the newly implanted osmotic pump. Mice were single-housed in fresh and clean cages after the surgery to prevent the risk of injury by cage mates. Some of the former cotton bedding/nest was transferred into a new cage to reduce mouse stress from being left alone in a new environment. A food paste made of water-saturated food pellets was placed in the cage close to the mouse nest for easy access. It was essential to change the food every day. Visual mouse pain monitoring included but was not limited to, looking for the shape of eyelids, porphyrin patches around the eyes, back positioning, the status of nest building, the animal’s activity level, and gait during frequent mouse observations 37-41. The animals’ weight was monitored over two weeks.
2.4 Image acquisition and tomographic reconstruction
All cochleae were imaged at the 2-BM-B beamline of the Advanced Photon Source (APS) using monochromatic radiation with photon energies of 22 kilo-electron volts (keV). The detector-sample distance was 600 mm for phase contrast. A 5x objective lens was used in the detector system, resulting in an approximately 3x3 mm2 field of view. Over a range of 180 degrees, we captured projections of each sample at increments of 0.12 degrees. The exposure time for a single image was 0.2-0.3 s. At the beginning and end of each image series, flat field images (no object in the beam path) were recorded; after the series, a dark field image (the radiation beam was blocked) was captured.
The projections were used to reconstruct the cochlea. Custom-written phase retrieval software for non-interferometric phase imaging with partially coherent X-rays was used42-44. Reconstructions were on a 2048 x 2048 grid with custom-written software45. The reconstructions resulted in 1.45 µm isotropic voxels. The spatial resolution was determined from the system's response to a sharp discontinuity in the image, such as a bony edge. The parameter measured was the distance required for the gray values to fall from 90% to 10%. The resulting distance was 4.4 µm.
2.5 Events observed during and after surgery
Under optimal anesthesia, the mouse’s expected heart rate is between 300 to 450 beats per minute (bpm), with an oxygen saturation above 96% and a breathing rate of around 55-65 breaths/min46. Four out of 45 mice (8.9%) had an episode with fewer than 50 bpm breathing rates. Decreasing the isoflurane concentration to 0.5% for 1 to 3 minutes restored normal breathing and respiratory rates. While little to no bleeding during the surgeries is typical, in 6 out of 45 mice (13.3%), for example, bleeding from the jugular vein occurred after the retractor nicked the vessel. Gentle pressure with a sterile cotton-tipped applicator stopped the bleeding. Bleeding also occurred while making the cochleostomy (3 mice, 6.7%). Since the mouse’s head was not tightly fixed in the head holder, it moved slightly from the animal’s breathing. The stapedial artery moved similarly, and in rare instances, into the beam path of the CO2 laser or drilling while creating the cochleostomy. Bleeding by opening the stapedial artery could be severe, and it was difficult to stop. It is important to record bleeding events and episodes of hypoxia as they may affect mouse hearing.
The vestibular system is close to the cochlear base and can be affected by the catheter’s placement. Symptoms indicating vestibular stimulation or damage (VDs) included a tilted head, shaking the head, a twirling body, and the mouse's inability to extend arms toward the cage to grab it while held by the tail. Vestibular symptoms showed in 19 out of 45 mice up to 72 hours post-surgery and did not recover over time. All mice showing vestibular symptoms had the catheter implanted toward the base of the cochlea.
Overall surgical and recovery times were not affected by the above-described incidents (Figure 5). After surgery, the mice's body weight decreased by 3.56 ± 3.12% and 6.28 ± 9.90% at 24 hours and one week, respectively. The body weight was back or above pre-surgery two weeks after the surgery. On average, mice presenting signs of vestibular damage lost more weight (11.95 ± 12.73% one-week post-surgery). Two mice from the VD group were euthanized for losing more than 25% of body weight during the first week after the surgery (Supplemental Figure 5). No other animal required euthanasia in this study.