Three Strategies to Induce Neurotrophic Keratitis and Nerve Regeneration in Murine Cornea

Summary

Here, we propose three different methods of damaging the sensory fibres innervating the cornea. These methods facilitate the study of axon regeneration in mice. These three methods, which are adaptable to other animal models, are ideal for the study of corneal innervation physiology and regeneration.

Abstract

The cornea is a transparent tissue that covers the eye and is crucial for clear vision. It is the most innervated tissue in the body. This innervation provides sensation and trophic function to the eye and contributes to preserving corneal integrity. The pathological disruption of this innervation is termed neurotrophic keratitis. This can be triggered by injury to the eye, surgery, or disease. In this study, we propose three different protocols for inflicting damage on the innervation in ways that recapitulate the three types of cases generally encountered in the clinic.

The first method consists in making an abrasion of the epithelium with an ophthalmic burr. This involves the removal of the epithelial layer, the free nerve endings, and the subbasal plexus in a manner similar to the photorefractive keratectomy surgery performed in the clinic. The second method only targets the innervation by sectioning it at the periphery with a biopsy punch, maintaining the integrity of the epithelium. This method is similar to the first steps of lamellar keratoplasty and leads to a degeneration of the innervation followed by regrowth of the axons in the central cornea. The last method damages the innervation of a transgenic mouse model using a multiphoton microscope, which specifically localizes the site of cauterization of the fluorescent nerve fibers. This method inflicts the same damage as photokeratitis, an overexposure to UV light.

This study describes different options for investigating the physiopathology of corneal innervation, particularly the degeneration and regeneration of the axons. Promoting regeneration is crucial for avoiding such complications as epithelium defects or even perforation of the cornea. The proposed models can help test new pharmacological molecules or gene therapy that enhance nerve regeneration and limit disease progression.

Introduction

The cornea, which is the transparent surface of the eye, is composed of three distinct layers: the epithelium, the stroma, and the endothelium. This organ has the highest density of innervation in the body and is composed mainly of sensory fibers (types Aδ and C) originating from the ophthalmic branch of the trigeminal ganglion. Sensory fibers penetrate the periphery of the cornea in the mid-stroma in the form of big bundles that branch out to cover the surface. They then bifurcate to pierce the Bowmann's membrane and form the subbasal plexus, which is easily recognizable by the formation of a vortex in the center of the cornea. Those fibers terminate as free nerve endings at the external surface of the epithelium. They are able to transduce thermal, mechanical, and chemical stimuli and to release trophic factors that are essential for epithelium homeostasis^1,2. Neurotrophic keratitis (NK) is a degenerative disease affecting the corneal sensory innervation. This rare disease stems from a decrease or a loss of corneal sensitivity that results in lower tear production and poor healing properties of the cornea³. NK progresses through three well-described stages, from stage 1 where patients suffer epithelial defects, to stage 3 where stromal melting and/or corneal perforation occur⁴.

Clinically, the origins of this disease can be diverse. Patients can lose corneal innervation after physical injury to the eye, surgery, or through chronic diseases, such as diabetes^5,6. To date, the NK pathogenesis process remains poorly understood, and therapeutical options for this sight-threatening condition are very limited. Therefore, a better understanding of the characteristics of epithelial defects is needed to better understand the mechanisms behind the regeneration of those fibers and potentially promote them. Here, we propose several models of corneal injury that induce NK in mice.

The first model is the abrasion of the epithelial layer of the cornea with an ocular burr. This model has mainly been studied in the context of the regeneration of the epithelium in different animals, such as rodents and fish^7,8,9, and to test molecules promoting corneal healing^10,11. Physiologically, it takes 2-3 days for the epithelial cells to close the wound. The physiological pattern of the innervation, however, takes more than four weeks to recover from the abrasion^12,13. During the surgery, the ocular burr removes the epithelial layer of the cornea that contains the subbasal plexus and the fibers' free nerve endings. This procedure can be clinically compared to patients with photorefractive keratectomy (PRK) to correct eye refractive defects. The procedure consists of removing the epithelium of the cornea and then reshaping the stroma with a laser¹⁴. Patients can experience several side effects following such surgery, such as a decrease of corneal nerve density for 2 years and a reduction in sensitivity for a duration of 3 months to one year post-surgery¹⁵. Given that the surgery induces a fragility of the corneal microenvironment, this model could help investigate these side effects and develop therapeutical approaches that would promote faster reinnervation, thus reducing the side effects in question.

The second model consists of sectioning the axons at the periphery of the cornea with a biopsy punch, inducing a Wallerian degeneration of the central innervation ¹⁶. Clinically, this method could be compared to anterior lamellar keratoplasty, in which the surgeon realizes a partial trephination of the cornea to remove a part of the anterior thickness of the cornea and replace it with a donor transplant ¹⁷. Following lamellar keratoplasty, patients may suffer from a number of symptoms including dry eye, loss of corneal innervation and graft rejection¹⁸. This axotomy model performed on corneal nerves could provide insight into the mechanisms of fiber degeneration, which occurs after a graft, followed by the axons' regeneration.

The third method damages the corneal nerves with a laser. By using a multiphoton microscope on the cornea of anesthetized animals, degeneration of the nerves localized in the optical field is induced as a result of reactive oxygen species (ROS) formation, which leads to DNA damage and cellular cavitation¹⁹. This method recapitulates the corneal photodamage induced by overexposure to natural UV (sunburn), which also triggers ROS formation, leading to DNA damage²⁰. Patients who suffer from corneal sunburn experience great pain, as the deterioration of epithelial cells deprives the corneal fibers' extremities of all.

The three methods described here are designed to enable the investigation of the NK pathogenesis process and axon regeneration. They are easily reproducible and precise. Moreover, they allow quick recovery and easy monitoring of the animals.

Protocol

All experiments were approved by the National Animal Experiment Board.

1. Preparations

1. Prepare an anesthetic solution of ketamine-xylazine for anesthesia. Inject ketamine at 80 mg/kg and xylazine at 10 mg/kg by diluting 200 µL of ketamine (100 mg/mL) and 125 µL of xylazine (20 mg/mL) in 2,175 mL of sterile 0.9% NaCl.
2. Prepare 0.02 mg/mL buprenorphine solution as an analgesic solution by adding 100 µL of 0.3 mg/mL buprenorphine to 1,400 mL of sterile 0.9% NaCl.
3. Prepare the fluorescent staining solution.
   1. Use a fine scale to weigh 10 mg of fluorescein salt and dilute it in 10 mL of phosphate-buffered saline (PBS) to obtain a 0.1% fluorescein solution.
   2. Protect the solution from the light and shake it for 5 min. Use a 10 mL syringe and a syringe filter of 0.2 µm to filter the solution.
      NOTE: The fluorescent solution has to be protected from the light and can be stored at +4 °C for 5 days.
4. Preparation of the mouse
   1. Weight the mouse and induce anesthesia by performing an intraperitoneal injection of 10 µL/g of mouse weight (MW) of the anesthetic solution. When the mouse stops moving, place it on a heated plate (37 °C) and verify that it is completely anesthetized by pinching its toes.
   2. Apply a drop of artificial tear on the eye undergoing the surgery and a drop of ocular gel on the contralateral eye.
   3. Once the mouse is unresponsive to pinch, inject the analgesia (0.02 mg/mL buprenorphine) at 5 µL/g of MW to provide 0.1 mg/kg buprenorphine subcutaneously in the neck.

2. Abrasion of the cornea

1. Follow step 1.3 to prepare the fluorescent staining solution.
2. Follow step 1.4 to prepare the mouse.
3. Before doing the abrasion, dip the ocular burr in 70% EtOH and then in PBS to clean it.
4. Place the mouse on a heated plate (37 °C) on the side to easily access the eye undergoing the surgery.
   1. Use a cotton swab to absorb the artificial tear and brush the eyelashes away without touching the eye.
   2. Turn on the ocular burr, open the eyelid of the mouse with two fingers, and simultaneously block the vibrissae to avoid them getting stuck into the burr.
5. Localize the pupil or the center of the eye and apply the ocular burr on the surface of the eye, and do circular movements. By looking closely, the surface of the removed epithelium can be observed. Otherwise, do approximately 20 circular movements on the cornea.
6. Put the mouse back on its belly and check if the ocular gel is still present in the contralateral eye.
7. Apply a drop of fluorescent staining solution on the abraded eye and wait for 20 s. Absorb the drop with a tissue without touching the eye and rinse it with a drop of artificial tear. Absorb the drop of the artificial tear with a tissue and illuminate the eye with the blue cobalt lamp.
   NOTE: The abrasion has to have a circular shape. It requires some training to obtain it.
8. Apply a drop of ocular gel on the abraded eye and let the mouse wake up on the heated pad. When the mouse is moving on its own, put it back in its cage.
9. Check the well-being of the animal during the following 2 days.
10. After the use of the ocular burr, use a soft tissue to remove epithelial cells stuck into the head of the burr and dip it in 70% EtOH and then PBS.
    ​NOTE: Once the epithelial cells are removed with the soft tissue, make sure that no pieces of tissue are stuck in the head of the burr.

3. Axotomy of the corneal nerves

1. Follow step 1.4 to prepare the mouse.
2. Place the mouse under a binocular loupe on the side to access the eye undergoing the surgery. Put the bottom of a Petri dish under the head of the mouse for the eye to be horizontal.
   NOTE: The following steps will be performed by looking through the binocular loupe.
3. Remove the artificial tear with a cotton swab and remove the eyelashes without touching the eye.
4. Place smooth curved pliers under the eye of the animal in order to pop it out from the orbit. Make sure to close the pliers enough for the eye not to be able to go back in the orbit once pressure is applied, but not excessively to prevent optic nerve damage.
5. Apply the biopsy punch of 2.5 mm vertically on the eye without pressure to ensure total contact of the punch with the surface of the cornea.
   NOTE: Once the punch is applied against the cornea, the hand of the experimenter might interfere with the field of view.
6. Then, start applying pressure and twist the punch several times.
   NOTE: Depending on the pressure, the number of twists may vary, but generally it is around 5. It requires training to feel the right pressure and movement. If too much pressure is applied, the eye ruptures. In this case, liquid can be observed exiting the lesion and the eye get soften. The animal has to be sacrificed.
7. Remove the biopsy punch. A circle must be visible on the cornea where the biopsy punch was applied.
   NOTE: The next steps do not require the binocular loupe.
8. Check if the contralateral eye still has ocular gel and apply some on the axotomized eye.
9. Let the mouse wake up on a heated pad (37 °C) and put it back in its cage when it is moving on its own.
10. Check the well-being of the animal during the following 2 days.

4. Cornea whole-mount processing

1. Sample collection and fixation.
   1. Weigh the mouse and induce anesthesia by performing an intraperitoneal injection of 10 µL/g of mouse weight (MW) of the anesthetic solution.
   2. When the mouse stops moving and does not have reflexes by toe pinching, perform a cervical dislocation.
   3. Pop the eye out of the orbit using two fingers and enucleate the eye by cutting the optic nerve using curved scissors.
   4. Place the eye in a 2 mL plastic tube containing PBS.
   5. Fix the eye by replacing the PBS with 4% paraformaldehyde and placing it on a shaker (30 rpm) at room temperature (RT) for 20 min.
   6. Rinse 3 times for 10 min with PBS at RT on a shaker (30 rpm).
2. Dissection of the cornea
   1. Place a drop of PBS on a piece of parafilm inside a Petri dish.
   2. Place the eye in this drop of PBS.
   3. Place the Petri dish under the binocular loupe.
   4. Dissect the cornea using microdissection scissors and forceps. Cut above the ciliary body to keep the limbus.
   5. Remove the lens, the ciliary body, and the iris with fine forceps.
   6. Place the cornea back into a 2 mL plastic tube.
3. Immunofluorescence protocol
   1. Incubate the cornea in 2 mL of 2.5% fish skin gelatine, 5% goat serum, and 0.5% detergent in PBS for 1 h at RT on a shaker (30 rpm).
   2. Incubate the cornea in 150 µL of a solution of 2.5% fish skin gelatine, 5% goat serum, and 0.1% detergent in PBS containing the primary antibody diluted at 1/1000 (anti βIII tubulin antibody) overnight at 4 °C on a shaker (30 rpm).
   3. Rinse 3 times for 1 h with 0.1% detergent in PBS at RT on a shaker (30 rpm).
   4. Incubate the cornea in the same solution as for step 4.3.2, containing the secondary antibody diluted at 1/500 overnight at 4 °C on a shaker (30 rpm).
   5. Rinse 3 times for 1 h with PBS at RT on a shaker (30 rpm).
   6. Incubate the cornea for 10 min with 150 µL of DNA intercalator diluted at 1/5000 at RT on a shaker (30 rpm).
   7. Rinse 2 times for 5 min with PBS.
4. Mounting of the cornea
   1. Place the cornea on a slide, epithelium facing the slide, and use a scalpel to create four sections without reaching the center.
      NOTE: Do the sections opposite to each other to form a flower shape.
   2. Place the cornea endothelium facing the slide and remove the PBS around the cornea using a tissue.
      NOTE: Do not remove the PBS under the cornea; otherwise, air bubbles may appear. If so, add PBS to the cornea and remove the air bubble. Then restart from step 4.3.2.
   3. Add model paste on the four corners of a rectangular coverslip.
      NOTE: The model paste elevates the coverslip as the cornea is a thick tissue.
   4. Drop 50 µL of a mounting medium on the cornea and lay carefully the coverslip above to avoid bubble formation.
   5. Seal the coverslip with nail polish.
5. Imaging
   1. Place the slide under an epifluorescent microscope.
   2. Define the size of the mosaic and the depth of the image and start the acquisition.
   3. Apply a deblurring and deconvolution program on the acquired image.
      ​NOTE: Step 4.6.3 is an option selected on the acquisition software (Large Volume Computational Clearing [LVCC])

5. Localized laser ablation of corneal nerves

1. Turn on the upright multiphoton microscope and the laser combined with an optical parametric oscillator. Activate the heated stage of the microscope (37 °C) and set the lasers at 850 nm and 1100 nm, corresponding to the required wavelengths for the mouse model.
2. Use a power meter to set the power of the lasers simultaneously at around 20 mW.
3. Follow step 1.4 to prepare the mouse and apply ocular gel on both eyes.
   NOTE: The animal used for this protocol has to be from a transgenic mouse line (MAGIC-Marker transgenic mouse²¹) that expresses an endogenous fluorophore in the corneal nerve fibers.
4. Install the animal in a head holder. Turn the animal 90° for the eye of interest to face the ceiling.
5. Strap the vibrissae of the animal to clear the eye area. Strap the animal's body to lessen the head movement induced by breathing.
6. Position a circular coverslip on the eye above the ocular gel. The coverslip has to be horizontal. Do not hesitate to move the head of the animal for the coverslip to be positioned correctly.
7. Add a drop of PBS above the coverslip. Place the animal on the microscope stage carefully and set the objective turret at the right distance.
8. Use the aqueous 20x objective and epifluorescence to localize the eye and the innervation through the eyepiece.
9. Activate the lasers and define the depth that needs to be illuminated.
10. Acquire 1024 x 1024 (pixels) image at an acquisition speed of 70%-85% of the maximum scan speed.
11. Once the image is acquired, take the animal off the microscope and remove the coverslip and the straps. Remove the animal from the head holder, add more ocular gel if needed, and let it wake up in a cage on a heated plate.
    NOTE: Destruction of the innervation can be observed 1 week after the imaging. This protocol can be repeated once a week for several weeks to increase the damage caused by the lasers.
12. When the animal is moving on its own, put the cage back in the stable and check its well-being for the following 2 days.

Results

This study proposes several protocols for inflicting damage on corneal innervation in mice. While similar protocols have been used to investigate the physiopathology of the healing of the epithelium, we chose to adapt and develop new methods of investigating corneal innervation regeneration. To observe the innervation, we used two techniques. First, we employed an immunofluorescence technique to stain the nerve fibers using a pan-neuronal antibody (BIII tubulin) and the nuclei with an intercalator. Second, we took advant...

Discussion

Neurotrophic keratitis is considered a rare disease, affecting 5 in 10,000 individuals. However, people suffering from NK due to a physical injury such as chemical burns, or syndromes such as diabetes or multiple sclerosis are not included in those statistics³. Furthermore, this condition remains significantly underdiagnosed²², and the prevalence of the disease is underestimated. There is a strong need for new treatments and therapy that would promote axon regeneration as a...

Materials

Name	Company	Catalog Number	Comments
0.2 µm seringe filter	CLEARLINE	51733
0.5 mm rust ring remover	Alger Equipment Company	BU-5S
2 mL plastic tubes	Eppendrof	30120094
Algerbrush burr, Complete instrument	Alger Equipment Company	BR2-5
Anti-beta III Tubulin antibody	Abcam	ab18207
Antigenfix	Diapath	P0016
Artificial tear	Larmes artificielles Martinet	N/A
Buprecare	Animalcare	N/A
Cotton swab	Any provider	N/A
Dissecting tools	Fine Science Tools	N/A
Fluorescein	Merck	103887
Gelatin from cold water fish skin	Sigma	G7765
Goat serum	Merck	S26
Head Holder	Narishige	SGM 4
Heated plate	BIOSEB LAB instruments	BIO-HE002
Hoechst 33342	Thermo Fisher Scientific	H3570
Imalgene 1000	BOEHRINGER INGELHEIM ANIMAL HEALTH France	N/A	French marketing authorization numbre: FR/V/0167433 4/1992
LAS X software	Leica	N/A	Large volume computational clearing (LVCC) process
Laser Chameleon Ultra II	Coherent	N/A
Laser power meter	Coherent	N/A
Leica Thunder Imager Tissue microscope	Leica	N/A
Multi-photon Zeiss LSM 7MP upright microscope	Zeiss	N/A
Ocry-gel	TVM lab	N/A
Parametric oscillator	Coherent	N/A
Penlights with blue cobalt filtercap	Bernell	ALPEN
Petri dish	Thermo Scientific	150318	Axotomy protocol
Petridish	Thermo Scientific	150288	Cornea whole-mount processing
Rompun 2%	Elanco	N/A	French marketing authorization numbre: FR/V/8146715 2/1980
Sterile biopsy punch 2.5 mm	LCH medical	LCH-PUK-25
Triton X-100	VWR	0694
Vectashield	EuroBioSciences	H-1000	Mounting medium