A Rabbit Model of Aqueous-Deficient Dry Eye Disease Induced by Concanavalin A Injection into the Lacrimal Glands: Application to Drug Efficacy Studies

All animal studies were approved by the Institutional Review Board of Stony Brook University and performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research.

1. Animals and housing

1. Acquire New Zealand White (NZW) rabbits weighing 2-3 kg.
2. House the rabbits singly in cages with strict temperature (65 ± 5 °F) and humidity (45 ± 5%) control. Lighting should have a 12 h on/off cycle.
3. Provide unlimited access to water and standard rabbit chow. Eliminate dietary enrichments as they may contain vitamin A that affects the eye.
4. Acclimate the animals for at least 2 weeks prior to baseline measures or induction of dry eye.

2. Methods of anesthesia and euthanasia

NOTE: All procedures require mild sedation except for Con A injection that requires moderate sedation.

1. For mild sedation, inject acepromazine (1 mg/kg) subcutaneously over the shoulders using a 26-gauge needle. Endpoint for mild sedation: animals maintain a relaxed head position with ear lobes no longer fully upright.
   NOTE: If the appropriate endpoint is not reached, an additional injection of acepromazine may be given. Animals should always remain awake, responsive to touching of their whiskers, and never show slowed breathing.
2. For moderate sedation, first give the animals acepromazine as above. After the endpoint is reached (see the note above), give isoflurane using a gas mask with O₂ flow set at 1 L/min and isoflurane delivery set to 5% (Figure 2).
3. Administer isoflurane until the rabbit's body tone is completely relaxed and ears are completely floppy.
   NOTE: No compensatory muscle movements should occur when the animal is turned on its side; breathing always remains spontaneous.
4. Spontaneous recovery occurs within 2-5 min: signs include spontaneous head movements and increased or normal muscular tone. After the experimental procedure is completed with moderate sedation, observe the rabbits for about 30 min or until their behavior returns to normal.
   NOTE: Ophthalmic ointment is not required during either form of sedation. 1) In mild sedation, animals are still alert and maintain a blink reflex. In moderate sedation, the inhibition of the blink reflex is so short that the ocular surface is not at risk. 2) Placement of ophthalmic ointment on the ocular surface precludes visualization of structures assessed during testing.
5. Euthanasia: Use an overdose of intravenous pentobarbital (100 mg/kg).

3. Removal of nictitating membrane

1. Perform the removal during the acclimation period (usually the first week) to allow complete and accurate evaluation of the cornea.
2. Injection to the right nictitating membrane
   NOTE: If the nictitating membrane of both eyes is to be removed, it is simplest to do this in one session. Start with one eye and proceed as described. For clarity of description, this method starts with the right eye.
   1. Place the rabbit in an appropriately sized restraining bag.
   2. Induce mild sedation as described in step 2.1.
   3. Apply 25 μL of preservative-free lidocaine to the right eye using a micropipette.
   4. Place a flexible wire lid speculum between the eyelids.
   5. Using 0.3 forceps (or equivalent), grasp the nictitating membrane at its apex and extend it over the cornea.
   6. Inject lidocaine 1% with 1:100,000 epinephrine subconjunctivally into the base of the nictitating membrane using a 26-gauge sharp needle (Figure 3A). A moderate bleb should form over the nictitating membrane.
   7. Remove the speculum.
3. Perform an identical injection to the left nictitating membrane.
4. Cutting the nictitating membrane
   1. After approximately 5 min, place the lid speculum back in the right eye. Grasp and retract the nictitating membrane at its apex using 0.3 forceps (or similar).
   2. Cut the nictitating membrane off at its base using Westcott scissors or equivalent (Figure 3B).
      NOTE: Bleeding is minimal and does not typically require cautery. Nevertheless, a high-temperature battery cautery is always kept nearby in case additional hemostasis is needed.
   3. Remove the speculum.
   4. Place topical antibiotic ointment on the eye (e.g., neomycin, polymyxin, bacitracin, and hydrocortisone).
5. Leave the Harderian gland intact. The Harderian gland is sometimes seen when the nictitating membrane is retracted.
   NOTE: If a large white mass or tissue elevation is seen in the nasal or superior subconjunctival region after the nictitating membrane is removed, the membrane was resected too close to its base allowing the Harderian gland to spontaneously prolapse. To prevent this in subsequent procedures, leave more of the nictitating membrane at the base.
6. Allow the ocular surface to heal for at least 1 week before further manipulations are made or ocular surface assays are performed.

4. Measurement of dry eye parameters and collection of tear samples

NOTE: Measure DED parameters based on study protocol needs (e.g., at baseline and specified time points thereafter). Measurements for DED should be done in the following order, with rigorous effort to faithfully replicate them each time. Test all animals at approximately the same time of day (± 1 h) to minimize circadian variations. These measurements usually require a team of two investigators.

1. Place the rabbit in a restraining bag. Induce mild sedation.
2. Tear Osmolarity²²
   1. Manually blink the eyelids 5-10 times to evenly distribute the tear layer on the ocular surface.
   2. Gently retract the lower lid.
   3. Sample tears with the TearLab Osmometer at the junction of the palpebral and bulbar conjunctiva along the lower fornix, just posterior to the base of the truncated nictitating membrane.
   4. Measure osmolarity using the TearLab Osmolarity Test following the manufacturer's instructions.
3. Tear break-up time (TBUT)
   1. Darken the room for this assay.
   2. Place a wire lid speculum between the eyelids.
   3. Apply a 50 µL drop of 0.2% fluorescein over the corneal surface using a micropipette. If even distribution of the fluorescein over the cornea is not obtained with the first drop, place a second drop.
   4. Immediately start a timer.
   5. Observe the pre-corneal tear film under a blue light. The TBUT is defined as the time taken to develop black dots, lines or obvious disruption of the fluorescein film (Figure 4). If needed, use surgical loupes that provide +1.50 magnification to better visualize early signs of break-up. Monitor for up to 1 min; if break-up as defined here occurs after 1 min, record TBUT for only 60 s.
4. Schirmer's Tear Test (STT)
   1. Apply a 25 μL drop of preservative-free lidocaine to the ocular surface.
   2. Place a Weck cell surgical spear into the inferior fornix to absorb residual lidocaine and any tear fluid. If needed, use the lower eyelid to cover the proximal end of the sponge to help keep it in place (Figure 5A).
   3. After approximately 30 s, remove the Weck sponge.
   4. Immediately insert a Schirmer's tear test strip into the space between the cornea and the palpebral conjunctiva at the mid-point of the lower lid.
   5. Immediately start a timer (Figure 5B).
   6. After 5 min, measure the length of the moistened portion of the strip; this is the STT value.
   7. Perform measures in triplicate and report the average of the 3 readings as the STT value.
5. Collection of tear samples
   1. To collect tear samples for assaying levels of various analytes in them such as lactoferrin, after the STT value is recorded at 5 min, leave the strip in place until wetting of at least 20 mm is obtained.
      NOTE: If adequate wetting does not occur after DED has been induced, advance the strip deeper into the lower fornix to help reach this endpoint in reasonable time.
   2. Cut the moistened strip and place immediately into 490 µL of chilled Tear Collection Buffer (4% BSA, 1M NaCl, 0.1% Tween-20 in PBS with proteinase inhibitor cocktail).
   3. Keep the samples on ice until they can be stored at -80 °C, where they should remain until assayed.
6. Rose Bengal Staining (RBS)
   1. Apply 50 µL of 1% preservative-free lidocaine to the cornea using a micro-pipette.
   2. After 30 s, place 25 µL of 1% rose bengal on the ocular surface and manually blink the eyelid to distribute it evenly.
   3. Immediately start a timer.
   4. At 3.5 min, place a wire lid speculum between the lids.
   5. At 4.0 min, photograph the superior conjunctival and corneal surface (Figure 6).
      NOTE: Adjust to the type of camera being used. Typical settings: digital single-lens reflex camera, aperture priority mode (aperture 13 or greater), ISO 6000, 100 mm macro lens attached with two 12.5 mm extension tubes, manual focus mode, lens at maximum magnification, and illumination from macro/ring flash set to automatic with through-the-lens mode. The ring flash's focus lamp is turned on to aid focus on the cornea.
   6. Complete all photos for both eyes within 1 min.
7. Score ocular surface staining using the NEI method²³ modified as follows. Don not grade 6 separate conjunctival zones. Score the superior conjunctiva of each eye. This is the portion of the conjunctival surface easily photographed without manipulating the globe. Manipulation could artefactually change the staining of the ocular surface.
8. Tear lactoferrin levels are a surrogate measure of tear production from the lacrimal glands. Assay tear lactoferrin in tears collected as above by using an enzyme-linked immunosorbent assay²⁴ kit following the instructions of the manufacturer.

5. Induction and treatment of dry eye

NOTE: Three portions of the orbital lacrimal gland system are injected.

1. Sedate the rabbits with acepromazine 0.2 mg/kg subcutaneously.
2. Shear off the fur in the periorbital and scalp area and completely remove any residual fur using Nair. Leave the skin entirely smooth for better visualization of the anatomical hallmarks and US-guided injection of concanavalin A (Figure 7).
3. Induce moderate sedation as described above.
4. Injection of the palpebral portion of the superior lacrimal gland (PSLG)
   NOTE: Perform injection of PSLG first.
   1. Apply to the appropriate eye 25 µL of preservative-free lidocaine 1% with a micropipette.
   2. Evert the upper eyelid and apply gentle medial pressure to the posterior orbital rim until the protuberance marking the palpebral portion of the gland is seen. The PSLG appears as a small bulbous elevation in the posterior (temporal) portion of upper lid.
      NOTE: To view the gland tissue during the learning process, apply 5% fluorescein to the area (Figure 8A). Tears can be seen streaming from the bulbous PSLG. Application of fluorescein is not needed for the administration of Con A; it is done only for illustration purposes to show the gland tissue.
   3. Using fine-toothed forceps and a 27-gauge needle on a tuberculin syringe, directly penetrate the gland using a transconjunctival approach. Advance the needle 2 mm into the tissue and inject 500 µg of Con A in a volume of 0.1 mL (Figure 8B).
      NOTE: This injection can sometimes be painful. If necessary, keep the animals under isoflurane until this injection is completed.
5. Injection of the orbital superior lacrimal gland (OSLG)
   NOTE: OSLG follows in rapid succession.
   1. Apply medial pressure to the globe causing the OSLG to protrude from the posterior incisure (see ref²⁵ for anatomy, if needed). Apply medial pressure to the globe (Figure 9, red arrow) with the protuberance of the OSLG from the posterior incisure. The protuberance serves as gross localization to find the posterior incisure.
   2. Use curved forceps with tips closed to indent the area until the bony opening in the skull is felt. This will be slit-like with an anterior/posterior direction under the protuberance.
   3. Apply modest pressure with forceps to leave an indentation in the skin, which will serve as the landmark for needle placement (Figure 10A).
   4. Insert a needle (tuberculin syringe with a 27-gauge, 5/8-inch needle) perpendicular to the skin over the indentation mark (Figure 10B) ~ ¼ inch into the incisure, then redirect the needle posteriorly and externally towards the lateral canthus aiming for the midpoint between the injection site and bony orbital rim.
      NOTE: If the incisure is not precisely targeted with the needle, the skull blocks its advancement.
   5. Once the hub of the needle is reached, slowly inject 1000 µg of Con A in a volume of 0.2 mL (Figure 10C).
6. Complete the injection of both the PSLG and OSLG within 2-3 min.
7. Remove the animal from isoflurane sedation (if not yet done). Injection of the inferior lacrimal gland (ILG) can usually be completed without further sedation.
8. Injection of the inferior lacrimal gland
   1. View the animal from the side. The prominence of the ILG can be seen along the lower anterior portion of the orbit (Figure 11A).
   2. Draw a vertical line using a surgical marking pen or suitable permanent marker on the skin where the superficial part of the ILG gland transitions from its superficial (more external) resting place on the zygomatic bone to its more medial location in the orbit. This is typically inferior to the anterior limbus (Figure 11A).
   3. Identify the end of the zygomatic bone by sweeping the vertically-held US probe across this line on the skin. The ILG transition occurs where the image of the gland changes from clearly circumscribed (hyperechoic line of the zygomatic bone is seen along the lower edge of the gland in the image) to one without a recognizable medial border (the zygomatic bone echo is no longer present, Figure 1).
   4. Observe the relative position of the hand-piece to the line drawn on the skin when the US screen shows this change. This is the "injection site" where Con A should be given.
   5. Control the depth of injection so as to place Con A into the gland at a point just medial to the zygomatic arch bone.
   6. Determine the depth of injection as follows: Set the desired depth of injection as the depth of the zygomatic bone (hyperechoic signal) plus 1 mm. Subtract this value from the known length of the needle (15 mm in this example).
   7. Insert the needle into the gland at the "injection site" ~12 mm, then slowly withdraw it until the length of the exposed needle outside the body (measured with surgical calipers) is equal to the difference calculated in 5.8.6 (Figure 12). Inject 1000 µg of Con A in 0.2 mL.
      NOTE: To ensure that the capsule of the gland is pierced and not simply pushed by the needle, the needle should be inserted ~12 mm or almost to the hub before its withdrawal begins.
   8. Repeat the US to confirm the success of the injection. The ILG should show a characteristic hypoechoic space (Figure 1).
      NOTE: The ILG injection is the one best tolerated by the animals²⁶ and is, therefore, done last.
9. Complete the entire procedure to inject all glands of both eyes within 10 min. This will require achievement of competency in the procedure.
   NOTE: A single set of injections into the 2 orbital lacrimal glands will induce acute DED lasting 1-2 weeks.
10. For DED of longer duration, inject Con A exactly as above every 7 days. Up to 6 such injections have been successfully performed.

6. Post-procedure care

1. Following injection of Con A, monitor the animals in their restraining bags for at least 10-20 min, or until the anesthetic effect has worn off.
2. Do not leave the animals unattended until they have regained sufficient consciousness to maintain sternal recumbency. Do not return them to their individual cages until they are fully recovered.
3. Post-procedure pain is usually mild and lasts less than 48 h. Assess pain with the rabbit grimace scale. If needed, give a single dose of subcutaneous ketorolac (5 mg/kg). For more severe pain, give subcutaneous buprenorphine 0.1 mg/kg every 8 h.