In Vivo Multimodal Imaging and Analysis of Mouse Laser-Induced Choroidal Neovascularization Model

Neovascular processes are a characteristic feature of several prevalent ocular pathologies. Most prominently, exudative age-related macular degeneration or Wet AMD for short, proliferative diabetic retinopathy, and retinopathy of prematurity. Together, these disorders account for most instances of legal blindness worldwide and are associated with additional ocular complications, such as vitreous hemorrhage and neovascular glaucoma.

However, despite their prevalence, therapeutic options for ocular neovascular disorders are limited. The current standard of care for neovascularization associated with AMD and proliferative diabetic retinopathy is the intravitreal injection of humanized antibodies directed against a critical mediator of angiogenesis and neovascularization, vascular endothelial growth factor or VEGF. However, despite the efficacy in halting disease progression and improving functional vision, intravitreal injections are costly and far from being without risk.

Complications can include infections, endophthalmitis, retinal detachment, and ocular hemorrhage. Therefore, there's an urgent need for the development of novel treatment options that are efficacious, safe, and less costly. In order to advance drug development for neovascular disorders, small animal models are of critical importance.

Such models need to be reproducible, have established and validated readouts and endpoints, and ideally use a clinically-relevant reference compound to serve as positive control. In this protocol, we will present the mouse choroidal neovascularization, or CNV model, which is one of the most commonly employed models for both the study of pathophysiological mechanisms contributing to CNV and the development of novel anti-neovascular agents. In the CNV model, Bruch's membrane is ruptured using an argon laser.

Thereby initiating neovascular processes originating from the choroid. Employing longitudinal in vivo imaging by spectral-domain optical coherence tomography, or SD-OCT for short, and fluorescein angiography, provides a means to follow the proliferation and regression of CNV. And thereby, to assess the efficacy and time course of novel pharmaceutical interventions.

Recent advances in image processing furthermore allow for the automated segmentation for the measurement of retinal thickness, providing a methodology free of investigator bias to evaluate the presence of edema. Herein, we will discuss the usefulness of the new InVivoVue Diver software by Leica Microsystems for the automated segmentation of retinal layers in the mouse CNV model. Lastly, we will discuss how the histological analysis of retinal whole mounts can complement longitudinal in vivo imaging in this model.

A single drop of tropicamide is applied onto each eye for pupillary dilation. Subsequently, the animal is placed gently into the rodent alignment stage. The mouse is adjusted to expose the eye and secured using the nose holder and a piece of laboratory tape placed gently over the back.

Next, a drop of lubricant is applied onto the eye to moisturize, and any excess liquid is carefully removed using filter paper swabs. Lastly, the stage is aligned in front of the OCT device for baseline OCT imaging. in vivo spectral-domain optical coherence tomography imaging is performed at baseline, prior to laser application, in order to verify the absence of any retinal abnormalities.

Gently remove the mouse from the holder. Place one drop of Viscotears Gel on a cover slip to applanate the cornea. Orient the mouse with the optic nerve head in the center, and focus the laser beam on the retinal pigment epithelium.

Make three laser shots by avoiding retinal blood vessels at four, eight, and 12 o'clock positions around the optic nerve. Inspect the firmness of the eye after all laser shots for the absence of any retinal bleeding. As before, align the mouse in the holder and perform fluorescein angiography and OCT imaging to confirm damage of Bruch's membrane.

For fluorescein angiography, first focus at laser burn areas using the infrared reflectance mode, which allows to visualize the sites laser was administered. Carefully inject 0.1 milliliters of 5%fluorescein sodium salt, for approximately 20 grams of mouse body weight without changing the mouse eye position. Start taking fluorescein angiography images at the level of the choroid and at the level of the retina, approximately every 30 seconds.

Align the eye for OCT imaging as before and start imaging. Averaging the SD-OCT signal helps to better visualize the detailed retinal morphology, as shown in this sequence. Once the SD-OCT imaging has been performed, carefully remove the mouse from the holder.

Apply lubricant for both eyes. At this point, you may choose to reverse anesthesia by subcutaneous injection of the alpha-2 antagonist for medetomidine, atipamezole, or wait for animal recovery from anesthesia. Repeat the in vivo SD-OCT and FA imaging in anesthetized animals on follow-up days five, 10, and 14.

Use the automated segmentation feature by the InVivoVue Diver software for retinal thickness measurements. For the healthy retina, the total retinal thickness is considered as the thickness of all layers, from the nerve fiber layer to the RPE. In the healthy retina, automated segmentation can accurately detect individual retinal layers, as can be seen in this example.

It is noteworthy to mention, that in places where CNV is present, manual measurement of retinal thickness should be performed for each CNV lesion site separately. From the nerve fiber layer to an imaginary line passing through the RPE layer. This sequence shows the workflow for manual thickness measurements at sites where CNV lesions are present.

The nerve fiber layer is located at the top, the choroid at the bottom of the image. Using the mouse or touchpad, the borders of the nerve fiber layer and the choroid are manually determined, and the software automatically calculates the total retinal thickness based on these coordinates. This figure shows serial images taken every 20 seconds at the choroidal level.

Image one was acquired in infrared reflectance mode. Whereas the others, after intraperitoneal fluorescein injection. White arrows in image one point to lasered sites, which show fluorescein leakage at later time points, highlighted by the white arrows in image 18.

This image shows FA serial imaging acquired at the retinal level. The white arrow in image 18 points to a lasered site, which shows fluorescein leakage appearing faster than the other two, outlined in circles in image 18. This summary figure shows a time course of SD-OCT images.

At baseline, immediately after lasering, and the follow-up time points on days five, 10, and 14. In summary, the protocol for the longitudinal in vivo imaging of CNV lesions using SD-OCT and FA imaging allows for the fast, multimodal, and reliable classification of CNV pathology and the presence of retinal edema. Thank you for your interest.