Introduction

Age-related macular degeneration (AMD) is a neurodegenerative retina disorder of which the early and intermediate forms are characterized by an increasing number and size of drusen and drusenoid deposits, Müller glia activation (1-3), infiltrated subretinal microglia (3,4), autofluorescence (5-7) and inward moving retinal pigment epithelium (RPE) cells (8). Advanced AMD often exhibits geographic atrophy (GA) and choroidal neovascularization (CNV). In this review, GA is referred as the Dry form and CNV as the Wet form of AMD. The current available treatment for early/intermediate AMD refers to the Age-Related Eye Disease Study (AREDS) (9,10), dietary supplements consisting of a calibrated blend of anti-oxidants. Intravitreal injections of vascular endothelial growth factor (VEGF) inhibitors are indicated for CNV (11), which however may accelerate the occurrence of GA (12,13), for which there are no FDA approved drugs.

As the aging population is globally increasing, the research and drug development for aging-related diseases, including AMD, are important, given that the AMD population is estimated to be ~196 million and to reach ~288 million by 2040 (14). Animal research has helped us to understand disease pathogenesis and pathobiology, although animal disease models have limitations being not like humans. The retina of rodents has no macula and has different subtypes of retinal neurons from humans (15). To date, researchers have investigated and developed genetic (Table 1) and non-genetic AMD murine models (Table 2). The genetic mouse model includes juvenile macular dystrophy, metabolic pathway, inflammatory and oxidative stress genes. The Table 1 displays that inflammation is essential part to drive disease progression especially into the Wet form. The phenotypes of all these genetic models get severer by light, fat diet and/or laser. Currently, we do not have a typical murine model for Dry AMD, whereas laser-induced and VEGF-A^high CNV models are typical for Wet AMD studies (44,72). Recently, a laser-induced Dry AMD model was reported (71) and is waiting for a deeper evaluation. Light damage is a classical model for retinal degeneration, and is still recommended and used for the preclinical study of Dry AMD (e.g., Guideline of Korea National Institute of Food and Drug Safety Evaluation for the Eye Health Functional Food Preclinical Studies). Depending on the source of lamps and wave lengths of lights, light induces damage of either photoreceptors or RPE, or both: usage of fluorescent lamps has induced photoreceptor damage rather than RPE, whereas usage of light-emitting diode (LED) lamps and source of blue light induce RPE degeneration (51,52). A number of pharmacologically induced animal models of Dry AMD have been reported, including: peptide amyloid beta (53,54), metals (55-57), sodium iodate (58-66), n-methyl n-nitrosourea (MNU) (68,69), 13(S)-hydroperoxy-9Z, 11E-octadecadienoic acid (HpODE) (2,76,77), and cisplatin (78-80), given via intravitreal, subretinal and/or systemic injections. In addition, immunization model using carboxyethylpyrrole (CEP) adducts has been applicable (70,81). The purpose of this review is to describe, compare and discuss the details of murine retina degeneration models induced by sodium iodate or lipid hydroperoxide HpODE and provide an example of the practical protocols commonly used to induce retina degeneration with these methods. For the detailed description of all other genetic and non-genetic AMD models, several review articles on these topics (82-84) are available. We present the following article in accordance with the Narrative Review reporting checklist (available at https://aes.amegroups.com/article/view/10.21037/aes-21-25/rc).

table1

Table 1

Genetic mice models of age-related macular degeneration

Name	Associated function	Early AMD Lipofusin, and/or deposits	Dry AMD PR and/or RPE degeneration	Wet AMD CNV and/or contribution to CNV	Reference
EGF-containing fibulin-like extracellular matrix protein1 (Efemp1), R345W/R345W	Genetic juvenile macular dystrophy (Doyne honecomb retinal atrophy)	√			Marmorstein et al., 2002 (16)
Tissue inhibitor of metalloproteinase 3 (Timp3), S156C/S156C	Genetic juvenile macular dystrophy Sorsby fundus dystrophy)	√			Weber et al., 2002 (17)
Neprilysin	Metabolism (amyloid beta)	√			Yoshia et al., 2005 (18)
5XFAD	Metabolism (amyloid beta)	√			Park et al., 2017 (19)
Apolipoprotein B100 (APO B100)	Metabolism (lipid)	√			Fujihara et al., 2009 (20)
Apo*E3-Leiden	Metabolism (lipid)	√			Kliffen et al., 2000 (21)
APOE-/-	Metabolism (lipid)	√			Dithmar et al., 2000 (22)
Complement factor H (Cfh) Y402H	Inflammation	√			Ufret-Vincenty et al., 2010 (23)
Cfh-/-	Inflammation	√			Hoh Kam et al., 2016 (24)
ATP-binding cassette transporter in rod outer segments (abcr)-/-	Genetic juvenile macular dystrophy (Stargardt disease)	√	√		Mata et al., 2000 (25); Lenis et al., 2017 (26)
MER proto-oncogene tyrosine kinase (Mertk)-/-	Genetic juvenile macular dystrophy (Stargardt disease)	√	√		Vollrath et al., 2015 (27)
Elongation of very long chain fatty acids (ELOVL4)	Genetic juvenile macular dystrophy (Stargardt disease)	√	√		Karan et al., 2005 (28)
Dicer1-/-	Metabolism (iron)	√	√		Damiani et al., 2008 (29); Kaneko et al., 2011 (30)
Ceruloplasmin (Cp)-/-/hephaestin (Heph)-/Y	Metabolism (iron)	√	√		Hahn et al., 2004 (31)
mouse cathepsin D (mcd)/mcd	Metabolism (protein)	√	√		Rakoczy et al., 2002 (32)
Aryl hydrocarbon receptor (AhR)-/-	Inflammation	√	√		Hu et al., 2013 (33); Kim et al., 2014 (3)
DJ1-/-	Inflammation	√	√		Bonilha et al., 2017 (34)
Nuclear factor erythroid 2-related factor 2 (Nrf2)-/-	Oxidative stress	√	√		Zhao et al., 2007 (35)
Superoxide dismutase (Sod)2-/-	Oxidative stress	√	√		Justilien et al., 2007 (36)
Hepcidine (Hepc)-/-	Metabolism (iron)	√	√	√	Hadziahmetovic et al., 2011 (37)
C-C motif chemokine ligand 2 (CCl2)-/-	Inflammation	√	√	√	Ambati et al., 2003 (38)
C-C motif chemokine receptor 2 (CCr2)-/-	Inflammation	√	√	√	Ambati et al., 2003 (38)
C-X3-C motif chemokine receptor 1 (Cx3Cr1)-/-	Inflammation	√	√	√	Raoul et al., 2008 (39)
CCL2-/-/Cx3Cr1-/-	Inflammation	√	√	√	Tuo et al., 2007 (40)
CD59a-/-	Inflammation	√	√	√	Herrmann et al., 2015 (41)
Transformting grwoth factor beta receptor (TGFbR)2-/-	Inflammation	√	√	√	Ma et al., 2019 (42)
Sod1 -/-	Oxidative stress	√	√	√	Imamura et al., 2006 (43)
Rhop-Vascular endothelial growth factor (VEGF); RPE65p-VEGF	Angiogenesis			√	Okamoto et al., 1997 (44); Schwesinger et al., 2001 (45)
Vldlr (very low-density lipoprotein receptor)-/-	Metabolism (lipid)			√	Hu et al., 2008 (46)
HtrA serine peptidase 1 (HTRA1) overexpression	Inflammation			√	Iejima et al., 2015 (47)

The table presents mouse genetic models of age-related macular degeneration.

table2

Table 2

Non-genetic age-related macular degeneration murine models

AMD form	Methods	Pathological Features	General species	Protocol
Dry	Light damage (48-50)	Photoreceptor degeneration; Inflammation; RPE degeneration; Oxidative stress; Subretinal microglia	BALB/C mouse, Sprague-Dawley rat	Fluorescent lamp, 200 to 15,000 lux, 1 to 24 hr, dark adaptation or not
	Blue light damage (51,52)	RPE degeneration; Oxidative stress	Wistar, Brown Norway rats	While (2,000 lux) or blue (150 lux) light-emitting diode, ~24 hr, 3 hr per day until day 21
	Amyloid beta1–40 (53,54)	Dose dependent; Inflammation; Photoreceptor degeneration; RPE degeneration	Long-Evans rat, C57BL/6 mouse	IVT, 5–15 μg
	Iron (55-57)	Depending on administration routes, effect or non-effect, RPE or photoreceptor damage	C57BL/6J mouse	i.v., IVT, subretinal
	Sodium iodate (58-67)	Dose dependent; Administration route dependent, RPE and/or photoreceptor degeneration; Photoreceptor rosettes; Müller glia activation; RPE movement	C57BL/6J, 129S6/SvEvTac, BALB/C mice, Brown-Norway, Long-Evans, Sprague-Dawley rats	i.v. (tail, femoral, retro-orbital), i.p., 20 to 100 mg/kg; subretinal injection, 5 μg/1 μL, 1 μL (rats)
	N-methyl-N-nitrosourea (68,69)	Photoreceptor degeneration	C57BL/6 mouse, Sprague-Dawley rat	i.p., 45 to 60 mg/kg
	Carboxyethylpyrrole (CEP)-adducts (70)	Retina inflammation; Photoreceptor and RPE degeneration	C57B/6J, BALB/C mice	i.p., 200 μg CEP-albumin in CFA, followed 100 μg CEP-albumin in IFA
	Laser (71)	Dose dependent; Altered outer-retina; Hypo/hyper-pigmented RPE; BrM thickening; Glia activation	C57B/6J mouse	810 nm laser, 1.3–2.5 J/spot, 400 μm, 50 sec per spot, 5–7 spots
Dry/Wet	HpODE (2)	RPE and photoreceptor degeneration; Autofluorescence; Extended degeneration; All retinal layer defect; CNV	Sprague-Dawley rat	subretinal, 30 μg/2 μL, 2 μL injection
Wet	Laser (72-74)	CNV; Inflammation; Complement activation	C57BL/6J mouse, Brown Norway rat	532 nm, 50–100 μm, 100–150 mW, 70 to 100 msec
	Polyethylene (PEG) (75)	CNV; Inflammation; Complement activation	C57BL/6J mouse	subretinal, polyethylene glycol 400, 0.5–2 mg/2 μL, 2 μL injection

BrM, Bruch’s membrane; CFA, complete Freurd’s adjuvant; CNV, choroidal neovascularization; IFA, incomplete Freurd’s adjuvant; i.v., Intravenous injection; IVT, intravitreal injection; RPE, retinal pigment epithelium.

---

Phenotype description of retina degeneration induced by sodium iodate or lipid peroxides (HpODE)

Sodium iodate-induced retina degeneration has been widely applied to different animal species, including rabbit, sheep, dog (85) and mouse via the administration routes of tail (58) or femoral vein (67), retro-orbital venous sinus (62), intraperitoneal (66), subretinal (61) and/or intravitreal (60) injections (Figure 1). Systemic injection of sodium iodate induces acutely RPE degeneration within a day, and increasing photoreceptor degeneration, followed by outer nuclear layer rosettes formation, but no significant damage in other retinal layers (86) (Figure 2). On the other hand, lipid peroxides were applied to New Zealand rabbits (77) and Sprague-Dawley (SD) rats via the administration routes of intravitreal and subretinal injections (2,76). Subretinal administration of linoleic acid peroxide (HpODE) induces acute local degeneration of RPE and photoreceptor around and at the injection site, and the degeneration is peripherally expanded through all retina layers, and finally induces CNV around 3 weeks post-injection (2,76) (Figure 2).

figure1

figure2

Sodium iodate

Depending on the administration route of sodium iodate, the features of retina degeneration are different: a systemic injection induces mainly rapid RPE degeneration, while subretinal injection causes local loss of both RPE and photoreceptors (61), whereas intravitreal injection induces photoreceptor, rather than RPE degeneration (60). The concentration of sodium iodate generally used is between 20 and 100 mg/kg; 20–30 mg/kg via retro-orbital injection is enough to trigger retina degeneration (62), whereas 40–50 mg/kg is generally acceptable for systemic intravenous (i.v.) and intraperitoneal (i.p.) injections (58,66). The proper sodium iodate amount for subretinal injection can be 5 μg in 1 μL for rats (61) and 1 to 50 μg in 100 μL for pigs (87). Thus injected rats displayed focal loss of RPE and photoreceptors, but no defects in inner retina. In pigs, 1 μg sodium iodate injection induced the same focal loss of RPE and photoreceptors and no inner retina defects, but 50 μg induced the degeneration of all retinal layers (87). Interestingly, the intravitreal injection of sodium iodate (300 or 400 μg, 50 μL injection) in rabbits induced only retinal, but no RPE degeneration (60) (Table 3). Although there are intensive studies on sodium iodate-induced retinal degeneration, relatively few compared different murine strains. For instance, it is known that C57BL/6J mice have a higher superoxide production than 129S6/SvEvTac mice (89) and that C57BL/6J mice respond severely to sodium iodate than BALB/C mice, when measured by electroretinogram (ERG) responses (66). Further, it is worth noting that sodium iodate-induced retina degeneration is partially reversible (58,64,90). Müller neurogenesis (88) and proliferating cells (64,88) are detected at early stages after damage, and areas with recovered RPE (58) are observed at later stages, around a month post-injection.

table3

Table 3

Observed phenotypic features of retina degeneration induced by sodium iodate according to the injection routes

Injection routes	Species	Injection amounts	Observed features	Reference
Retro-orbital	C57BL/6J mouse	20, 30 mg/kg	RPE and photoreceptor degeneration; AF in fundus; Hyperreflective spots in vitreous and retina of OCT	Wang et al., 2014 (62)
i.v.	C57BL/6J mouse	40 mg/kg	RPE degeneration; Reduced ERG; Macrophage infiltration, Peak time of TUNEL positive RPE cells at 14 hr after injection; Peak TUNEL positive ONL at 3 D after injection; AF in fundus; Fluorescein leak choroidal degeneration; Recovered phenotypes at 4 weeks	Moriguchi et al., 2018 (58)
	Sprague-Dawley rat	40 mg/kg	RPE degeneration; Photoreceptor degeneration; Rosettes; Reduced ERG	Yang et al., 2014 (65)
i.p.	C57BL/6J, BLAB/C mouse	50 mg/kg	Better tendency of ERG responses in BLAB/C than C57BL/6J; More distinct AF spots in OCT and Fundus in BLAB/C than C57BL/6J; No distinct differences in histology; No distinct differences in immunofluorescent staining: ZO-1, RPE65, Rhodopsin, Blue opsin between C57BL/6J and BLAB/C	Chowers et al., 2017 (66)
	Long-Evans rat	50 mg/kg	RPE degeneration; Photoreceptor degeneration; Outer and inner retina disorganization; Müller glia activation and proliferation, Müller neurogenesis; Neurogenesis	Jian et al., 2015 (88)
	Sprague-Dawley rat	50 mg/kg	RPE and photoreceptor degeneration; Rosettes; Regeneration of RPE (PCNA and RPE65 double positivity at D3 and PCNA immunoreactivity in RPE and choroid layer at D5 and D7), Regeneration of photoreceptor (PCNA immunoreactivity in ONL)	Kim et al., 2018 (64)
Subretinal	Brown-Norway rat	5 μg, 1 μL injection	Focal RPE and photoreceptor degeneration; Müller glia activation; Choricocapillaris degeneration	Bhutto et al., 2018 (61)
	Pig	1 μg, 50 μg, 100 μg, 100 μL injection	1 μg: Focal RPE and photoreceptor degeneration, 50-100μg: All retinal layers defected	Monés et al., 2016 (87)
Intravitreal	New Zealand white rabbit	300, 400 μg, 50 μL injection	Retinal degeneration; No RPE defect	Ahn et al., 2019 (60)

AF, autofluorescence; D, day; ERG, electroretinogram; i.p., intraperitoneal injection; i.v., intravenous injection; OCT, optical coherence tomography; ONL, outer nuclear layer; PCNA, proliferating cell nuclear antigen; RPE, retinal pigment epithelium.

Lipid peroxides

Lipid peroxides (linoleic, linolenic, arachidonic, docosahexaenoic acids) were first applied to rabbits via intravitreal injection. Nineteen mg of linoleic peroxides abolished a- and b-wave ERG responses by 7 days post-injection, and the same amounts of linolenic and docosahexaenoic acid peroxides abolished a- and b-wave ERG responses by 14 days post-injection, whereas arachidonic peroxide injection only slightly reduced a- and b-wave ERG responses (77) (Table 4). The subretinal injection of commercial linoleic acid peroxide, HpODE (30 μg), into the eyes of SD rats induced extended retina and RPE degeneration, retina inflammation, oxidative stress, and, finally, CNV by 3 weeks post-injection (2,76) (Figure 2). It is unclear whether the subretinal injections of linolenic, docosahexaenoic acid peroxides, and the intravitreal injection of lipid peroxides may also induce CNV.

table4

Table 4

Phenotypic features of retina degeneration induced by lipid peroxides injection

Species	Injection routes	Lipid hydroperoxides	Injection amounts	Observed features	References
New Zealand white rabbit	Intravitreal	Linoleic acid hydroperoxide	19 mg	Non-recordable level of ERG responses of a and b waves by 7 days post injection	Armstrong et al., 1982 (77)
		Linolenic acid hydroperoxide	19 mg	Non-recordable level of ERG responses of a and b waves by 20 days post-injection
		Arachidonic acid hydroperoxide	19 mg	Reduced but maintained/recovered ERG responses of a and b waves by 20 days post-injection
		Docosahexaenoic acid hydroperoxide	19 mg	Non-recordable level of ERG responses of a and b waves by 20 days post-injection
Sprague-Dawley rat	Subretinal	13(S)-hydroperoxy-9Z, 11E-octadecadienoic acid (HpODE): Linoleic acid hydroperoxide	30 μg	CNV by 3 weeks post-injection	Baba et al., 2010 (76)
Sprague-Dawley rat	Subretinal	13(S)-hydroperoxy-9Z, 11E-octadecadienoic acid (HpODE): Linoleic acid hydroperoxide	30 μg/2 μL, 2 μL injection	Extended retina degeneration, RPE degeneration, autofluorescence, oxidative stress, inflammation, CNV by 3 weeks post-injection	Kim et al., 2021 (2)

CNV, choroidal neovascularization; ERG, electroretinogram.

---

Practical protocol description of retina degeneration induced by sodium iodate and lipid peroxide HpODE

Sodium iodate (10 to 50 mg/kg) were peritoneally injected into the mice of BALB/C and C57BL/6J (8–10 weeks, 20–30 g) or SD (8–10 weeks, 200–350 g) rats for the observation of retinal degeneration, whereas HpODE was injected into the subretinal space of SD rats (6–8 weeks, 150–250 g).

Sodium iodate from Sigma (catalog number: 71702-25 g) and lipid peroxide HpODE from Cayman Chemical (catalog number: 48610-500 μg) were used. Compounds of sodium iodate and HpODE are summarized in Table 5.

table5

The stock solution of sodium iodate was prepared at a concentration of 20 mg/mL in sterile saline, and aliquots were stored at ?20 ℃. The stock of sodium iodate was replaced every 6 months, because a diminished effect was observed overtime. The stock solution was further diluted with saline and the diluted solution was injected into mice and rats at a standard dose (Table 6).

table6

Table 6

Calculation of i.p. injection amount of sodium iodate solution

SI injection amounts	Rats	Mice
50 mg/kg	5 mg/100 g; 2.5 mL (2 mg/mL)/100 g	0.5 mg/10g; 250 μL (2 mg/mL)/10 g
40 mg/kg	4 mg/100 g; 2 mL (2 mg/mL)/100 g	0.4 mg/10g; 200 μL (2 mg/mL)/10 g
20 mg/kg	2 mg/100 g; 1 mL (2 mg/mL)/100 g	0.2 mg/10g; 100 μL (2 mg/mL)/10 g
10 mg/kg	1 mg/100 g; 0.5 mL (2 mg/mL)/100 g	0.1 mg/10g; 50 μL (2 mg/mL)/10 g

Make 2 mg/mL SI solution from stock 20 mg/mL with saline. i.p., Intraperitoneal injection.

The HpODE solution (15 μg/1 μL) was prepared by dissolving HpODE in ethanol. The ethanol solution was then evaporated by nitrogen gas streaming and HpODE reconstituted with 0.02 M sodium borate buffer at the desired concentration (pH 9.0), kept on ice, and used within 24 hours. For the subretinal injection of HpODE, a rat was anesthetized with ketamine and xylazine (see below), pupil dilated with tropicamide, and cornea topically anesthetized by proparacaine hydrochloride. The HpODE solution was then injected into the subretinal space by a capillary micropipette, using a PL1-100A injector guiding the capillary into a hole previously done by a tiny needle (30 ? gauge).

The detailed materials for HpODE subretinal injection are summarized in Table 7, and the practical procedures of sodium iodate peritoneal injection, and HpODE subretinal injection were done as following.

table7

Table 7

Materials for HpODE subretinal injection

Material items	Available company and cat #	Additional information
Rats	Sprague Dawley	6–8 weeks, 150–250 g
HpODE	Cayman Chemical (cat #, 48610-500 μg)	Ethanol evaporated, and sodium borate solution added.
Boric acid	Sigma-Aldrich (cat#, B6768-500 g)	0.02 M Sodium borate buffer (pH 9)
Injector	Harvard Apparatus, PL1-100A pico-liter microinjector	N/A
Thin wall glass capillaries	World Precision Instruments, TW100-4	Tip internal diameter 25 to 30 μm, 100 mm glass capillaries; www.wpiinc.com
Flaming Micropipette Puller	Sutter Instrument Co. Model P-80 Brown	Heat 950, Pull 44, Velocity 30, Time 70; www.sutter,com
Razor	Any razor is fine	N/A
Anesthetized solution	Ketamine (KetaVed, 100 mg/mL) and Xylazine (VEDCO, 100 mg/mL)	Check the concentration of solutions available. Depending on region and country, the concentrations are different
Ocular lubricant ointment	Alcon, Duratears	Protect non-surgery eye during the anesthesia
Dilation solution	Tropicamide (1% Mydricayl, Alcon); Mydfrin (2.5% phenylephrine hydrochloride, Alcon)	Use available dilation solution depending on country and local region
Topical anesthetized solution	0.5% Proparacaine hydrochloride (Alcaine, Alcon); tetracaine	Use available topical anesthetized solution, depending on country and local region; Proparacaine is less toxic than tetracaine
Gonio solution	Hydroxypropyl methylcellulose gel (2.5% Gonak, Akorn)	When the micropipette is injected into eyeball, Gonio solution should be added to see the retina inside.
Balanced salt ophthalmic solution (BSS)	Alcon, NDC0065-0795-15	N/A
Forceps	McPherson, strait 5 mm, smooth 10.8 CM	Any blunt and safe forceps available.
Scissors	Proper scissors to remove whiskers	N/A
Needles	30 ? gauge	N/A
Antibiotic eye drops	Ofloxacin 3mg/mL	Ocuflox eye drops can be applied before ointments. Choose proper one depending on country and region
Antibiotic ointments	PREG-G-gentamycin, prednisolone acetate suspension (Allergan Inc); Neomycin and polymyxin b sulfates and bacitracin zinc ophthalmic ointment (Akorn, Inc)	Terramycin ophthalmic ointment (Pfizer) can be applied. Depending on region and country, available commercial ophthalmic antibiotics are different.
Ear punches	Kent Scientific Corporation, Nail-Clipper	N/A
Sterile Swab	Fisherbrand cat #. 23-400-116	N/A

Procedure for sodium iodate i.p. injection

Procedure of HpODE subretinal injection

Material preparation for HpODE subretinal injection

Sodium borate buffer solution (0.02M, pH9.0) for HpODE reconstitution

Dissolve 12.4 mg of boric acid in 5 mL of water, adjust to pH 9.0 with NaOH and add water to 10 mL. Sterilize by filtration through 0.2 μm filters the 0.02M sodium borate buffer solution before use.

Micropipette preparation

Pull 100 mm glass capillary micropipettes (diameter 25 to 30 μm) using P-80 Flaming micropipette puller (setting: Heat-950, Pull 44, Vel 30, Time 70) and cut off the pulled pipettes with razor, if needed.

Anesthetic solution for rats

Add 1.1 mL of xylazine (100 mg/mL) into 10 mL ketamine solution (100 mg/mL).

Inject subcutaneously (s.c.) or i.p. the mixed solution of ketamine and xylazine (100 μL per 100 g rat) into SD rats (6–8 weeks); 150 to 200 μL solution is enough for 6–8 weeks old rats.

or

Add 2.5 mL of xylazine (20 mg/mL) into 10 mL ketamine solution (50 mg/mL).

Inject (s.c. or i.p.) the mixed solution of ketamine and xylazine (200 μL per 100 g rat) into SD rats (6–8 weeks); 300 to 500 μL solution will be enough for 6–8 weeks old rats.

Lipid formulation

Evaporate the ethanol from HpODE under nitrogen gas and dilute the residual pellet in sterile 0.02 M sodium borate buffer (pH 9.0), keep on ice and use immediately, or practically within 8 hours.

HpODE subretinal injection

Before starting any procedure, set up the working surgery place and a surgical microscope (check focusing and controlling of the foot pedal), set up a picoinjector (PLI-100A) and micropipettes filled with HpODE (30 μg, 2 μL), set up an injector (pressure 40 psi, injection volume 2 μL, adjust injection time: 0.06 to 0.2 sec). The procedure of HpODE subretinal injection should be done by three operators consisting of a surgeon and two assistant staffs. The surgeon focuses on subretinal injection, while the staff helps with anesthesia (steps 1 to 3 below), and another staff takes care of the rest: tapping and centrifuging HpODE solution before use, applying GONAK solution (hypromellose 2.5% for gonioscopy) to the eye (step 11 below), injector operation, recording, etc. Keep the HpODE solution on ice during the surgical procedure.

---

Assessment of retina degeneration in murine models

In degenerative conditions of the outer retina, the dysfunction and loss of RPE and photoreceptors are the main causes of vision impairment. To assess outer retina degeneration, non-invasive ophthalmic analyses such as the fundus imaging, optical coherence tomography (OCT) and electroretinogram (ERG) can be applied, as well as other invasive assessment methods, as described below. In this section, ERG as a non-invasive measurement is described, and the invasive assessment of retinal thickness, photoreceptor loss, outer nuclear layer folds, RPE loss, subretinal microglia infiltration, and expression of inflammatory, oxidative and cell death genes are discussed. Moreover, the benefits of the sclerochoroid/RPE/retina whole mount application to observe the flat view of the disorganized subretinal and outer nuclear layers (86), and the infiltrated subretinal microglia (91,92) are discussed.

ERG

Visual dysfunction is considered to be an early physiological marker, and a sensitive indicator compared to morphological evaluation. The a- and b-waves of dark-adapted (scotopic) ERG represent the primary activities of photoreceptors and bipolar cells, respectively (Figure 3A). Light-adapted (photopic) ERG can be used to probe cone photoreceptor mediated activities in the retina. Glial cells provide support and nourishment to retinal neurons, and are essential for photoreceptor function and survival. Alterations in glia function could also indirectly modify ERG a- and b-waves (93,94). In addition, glia activity contributes to the formation of ERG c-wave response (3,95), together with RPE activity which represents two major components with opposite polarity. In contrast to a- and b-wave response, often elicited with flashlight stimuli, slow response of c-wave (usually peaking at 3–5 seconds after light onset) is commonly elicited by a long light stimulus (3,96,97). In rodent models of retina degeneration, such as sodium iodate and HpODE, ERG response of a- and b-waves is well-correlated to morphological degeneration in retina (2,66,77). Retina degeneration induced by intravitreal and subretinal HpODE is relatively slow: an extensive damage is detectable after 2–3 weeks (2), compared with systemic sodium iodate. ERG response also decreases slowly (77), compared to the sudden reduction of ERG observed within a week after sodium iodate injection (40 to 50 mg/kg) (48,58,63,98). In drug efficacy tests, ERG response in retinal degeneration is considered to be more sensitive than morphological evaluation, but both of them are, ultimately, correlated.

figure3

Retina thickness quantification

The thickness of the total retina and each retinal layer in murine models is quantified in histological sections by measuring its length in vertical paraffin sections or counting nuclei in a column of layer. The thickness is measured at a specific region (e.g., 300 to 500 μm away) or a specific length away (e.g., every 150, 250, 300 μm away) from the center of the optic nerve head (58,63,66) (Figure 3B).

Photoreceptor loss and rosettes quantification

Photoreceptor loss is reflected by the thickness of whole retina, outer nuclear layer, and/or photoreceptor outer/inner segments (66), and the same measurement method described above in retina thickness quantification is applied. The fluorescent intensity of opsin expressions—Rho, S-opsin, M-opsin—is further observed and measurable by immunofluorescence staining in sectioned and whole mount retinas (60,66,99). In sodium iodate model of retina degeneration, number and area of photoreceptor folds/rosettes are measurable in retinal sections (67) and sclerochoroid/RPE/retina whole mounts, and the representative images of the rosettes induced by sodium iodate in sclerochoroid/RPE/retina whole mounts can be seen in a paper published in this journal (86) (Figure 3C).

RPE loss quantification

RPE loss in rodents is detectable in histology sections of both pigmented and non-pigmented murine retinas. In the pigmented murine retina, RPE loss, swelling and mis-location are detected easily by lost, swelling and mis-located pigments whereas in the non-pigmented murine retina, loss, swelling and mis-location are observed with pale-tinted and big sized RPE nuclei. In addition, the immunofluorescence staining of RPE biomarkers—RPE65, ZO-1 and GLUT1- are also applicable in retina sections and whole mounts (2,58,61,64,66,99,100). Furthermore, fluorescence of phalloidin stained retina allows the counting of the number of outlined RPE cells in whole mounts (2,62) (Figure 3D).

Subretinal macrophages/microglia quantification

In outer retinal degeneration, the subretinal layer may include migrated photoreceptors, RPE cells, and microglia/macrophages. The nuclei of these cells in the subretinal layer are recognizable and countable in histology sections (3). The morphology and size of these nuclei are different from each other (3,91). The genetically fluorescence-labeled macrophages/microglia mice (e.g., CX3CR1-GFP) could also be used (101), and the immunofluorescence staining of macrophages/microglia—Iba1, CD11b—detects subretinal infiltrated microglia in retina sections and whole mounts (3,91,92), amenable for quantitative analysis.

Sclerochoroid/RPE/retina whole mount

Sclerochoroid/RPE/retina whole mount and imaging have recently been developed in mice and ferrets (92,102). The method allows to observe intact and integral subretinal and neighboring layers in the degenerative outer retina (Figure 4). The detailed protocol was previously published (91), and the method is applicable for the integral observation of the infiltrated subretinal microglia, and outer nuclear layer rosettes (86). Classic whole mount method viewing either the neural retina or the RPE side, the subretinal microglia appear separate, depending on their location, and a part of microglia neurites are on the RPE side and another part are visible at the retina side of the whole mount, so that it is hard to figure out the actual number of the subretinal microglia. Moreover, this tissue separation causes additional damage to already degenerated areas. Even in normal retinas, the separation between RPE and retinal tissue could cause artificial scars in RPE microvilli and photoreceptor segments (92). Thus, the non-separate sclerochoroid/RPE/retina whole mount and imaging method could provide reliable horizontal images, especially of the subretinal and neighboring layers. It is worth mentioning that all layers in the mouse retina are accessible by confocal microscopy in the non-separate whole mount method (91), but in the rat, the imaging of subretinal layer is not accessible (Zeiss 700 in our hands) (86), which we confide can be technically solved in the near future.

figure4

Gene regulation quantification

Gene regulation of biomarkers associated with cell death, oxidative stress, and inflammation can be obtained by RT-qPCR using specific primers (2,62,101). In sodium iodate and HpODE retina degeneration, BCL1 associated X apoptosis regulator (Bax) and Bcl2 antagonists/killer (Bak) are up-regulated (2,62). Oxidative stress genes—NAPDH oxidase 1 (Nox1), Nox2, Nox3, Nox4, Dual oxidase 1 (Doux1) or Doux2—are upregulated in sodium iodate and HpODE retinal degeneration (2; our observation). Inflammation-related genes—Tumor necrosis factor (TNF)-α, Interleukin (IL)-6, Intercellular adhesion molecule 1 (Icam1), C-C motif chemokine ligand (Ccl) 2, Ccl3, Ccl7 or Ccl8—are also upregulated after sodium iodate treatment (101) (our observation, data not shown) and HpODE (2). An additional comment on tissue preparation is that it is also acceptable to use whole posterior eyeballs, and not separate tissues coming from either the neural retina or the RPE side, because the presence of degenerated and angiogenic tissues hinder tissue separation, and infiltrated cells between the retina and the RPE could be lost during tissue separation, which means that the associated information will be missing.

---

Discussion and conclusions

The shortage of typical Dry AMD preclinical animal models is widely noticed. However, the recent reported laser-induced Dry AMD model (71) and the systemic, intravitreal and subretinal injection of sodium iodate (60,61,66), intravitreal and subretinal injection of lipid peroxides (2,76,77) might meet the requirement for the preclinical platform of Dry AMD models, which complimentary to the laser-induced and VEGF-A^high Wet AMD models (44,72). The light-damage model could also be utilized if the focus is inflammation and infiltrated microglia (48,49). Summarizing, light damage induces microglia subretinal migration, retina thinning, RPE degeneration, depending on light source, and sodium iodate induces RPE and retinal degeneration, and HpODE induces RPE and retinal degeneration plus CNV. Therefore, most of the clinical features expected in the human disease could be addressed using these existent models (Tables 1-3). Among the AMD models, sodium iodate model has been most widely studied and used for the translational study, and the different dosages and administration routes of sodium iodate provides different aspects of the phenotypes. On the other hand, the sodium iodate model is acute, and might be hard to study the intermediate/progressing form of AMD, whereas HpODE AMD model is more slowly progressed than sodium iodate model and finally includes CNV. However, HpODE was only studied in SD rats via a route of subretinal injection and albino rabbits via a route of intravitreal injection. Thus, additional studies will be valuable for translational studies for relating human AMD progressing.

Assessment and quantification methods in preclinical studies could be further standardized and harmonized. In fact, there are still gaps in the assessment methods used by animal research and human clinical trials, because invasive methods are generally allowed in animal experiments, whereas only non-invasive assessment is applicable in clinical trials. The invasive methods in animal models, which we discussed above, are correlated with the non-invasive ophthalmic assessments, including color fundus, fundus autofluorescence (FAF) and OCT. However, sensitivity and resolution of clinical ophthalmic assessment appears to be lower than invasive methods. Furthermore, the observations done by the non-invasive ophthalmic equipment sometimes cannot be fully explained because it is not possible to investigate the molecular and anatomical aspects, that requires the dissection of eye tissues. For example, FAF is associated with increasing lipofuscin (103), infiltrated microglia (3,104), photoreceptor rosettes (104-106) and migrating RPE cells (104), of which association can be fully dissected in the animals, but not in the human eyes. ERG functional data are considered as an earlier indicator of impaired vision and is more sensitive than the morphological changes and analyses, and ERG test protocols should be standardized for both clinical patient exams and experimental study with animal models. When patients’ safety is considered, the smallest light intensity possible could be used to measure ERG response before and after a period of treatment to reduce repetitive and harmful exposures to strong light intensity. However, one final issue that needs to be always kept in mind, is that murine eyes are not human and animal disease models are not exactly alike human’s, although there are many merits in animal research, which is the only mean that we have to address specific questions to the physiopathology and therapy of eye diseases.

---