EP4739299A1

EP4739299A1 - Models of cone photoreceptor degeneration

Info

Publication number: EP4739299A1
Application number: EP24742477.3A
Authority: EP
Inventors: Serge Picaud; José Alain Sahel; Valérie FRADOT; Manon LECLERCQ; Alexandre DENTEL
Original assignee: Centre National de la Recherche Scientifique CNRS; Institut National de la Sante et de la Recherche Medicale INSERM; Sorbonne Universite
Current assignee: Centre National de la Recherche Scientifique CNRS; Institut National de la Sante et de la Recherche Medicale INSERM; Sorbonne Universite
Priority date: 2023-07-07
Filing date: 2024-07-05
Publication date: 2026-05-13
Also published as: WO2025012134A1; CN121752263A; WO2025012666A1

Abstract

The present invention provides a method for preparing a model of cone degeneration, comprising a step of activating a specific cell death in at least one cone photoreceptor. Also provided are methods for isolating compounds capable of protecting conae photoreceptors from degeneration.

Description

MODELS OF CONE PHOTORECEPTOR DEGENERATION.

INTRODUCTION

The retina is a light-sensitive neuronal tissue located at the posterior part of the eyeball. Retinas of humans have two distinct regions. The retinal periphery has low spatial acuity and is responsible for night-vision and different aspects of motion vision. The fovea (or macula) is at the retinal centre and drives high spatial acuity vision that is essential for reading and face recognition. Primates are the only mammals with a fovea. Retina is arranged in three layers of cells, namely the outer nuclear layer (ONL), the inner nuclear layer (INL), and the retinal ganglion cell (RGC) layer. The ONL harbours photoreceptors, which are unique neurons dedicated to converting light into electrochemical signals and as such essential for vision. Two different types of photoreceptors are distinguished: Rod photoreceptors respond to dim light and enable vision at night, whereas cone photoreceptors respond to daylight and mediate high-resolution and colour vision

Retinal neurodegeneration associated with the dysfunction or death of photoreceptors is a major cause of incurable vision loss. Photoreceptor death is the common cause in many retinal disorders, such as age-related macular degeneration (AMD), and retinitis pigmentosa (RP). In most of these diseases (e.g., AMD), the pathology appears to be linked primarily to a loss of cones. It is thus particularly important to be able to prevent cone photoreceptors degeneration.

To this day, there is no effective therapy for retinal degenerative diseases which thus remain poorly treatable, presenting an urgent need for developing novel treatment strategies.

In particular, diseases affecting photoreceptors, and especially cone photoreceptors, are essentially untreatable. None of the available pharmacological treatments aim at preventing the degeneration of the photoreceptors, and in particular of the cone photoreceptors. The only available therapies are directed to the later stages of the diseases, when the majority or all retinal cells have been lost after cone photoreceptors degeneration has occurred.

Preventing cone cell loss appears as a main target in therapeutic strategies.

This situation thus creates a strong need for models that can accurately reproduce photoreceptor degeneration, in particular cone photoreceptor degeneration, and allow the testing of new treatments for retinal neuroprotection.

However, no satisfactory such model appears to be available to date. Notably no in vitro or in vivo model is specific for cone degeneration, thereby hampering the isolation and testing of potential drugs for treating retinal diseases.

There is thus a need for an in vitro and/or in vivo model of cone photoreceptor degeneration.

SUMMARY OF THE INVENTON

In a first aspect, the present disclosure relates to a method for preparing a model of cone degeneration. The method disclose herein comprises a step of activating a specific cell death, notably ferroptosis, in at least one cone photoreceptor. In particular, the activation of cell death, notably ferroptosis, is specific to the cone photoreceptor, i.e., cell death is activated in the cone photoreceptor, but not in other retinal cell types, such as rod photoreceptors and retinal pigment epithelium cells, at identical concentrations of cell death (in particular ferroptosis) activators.

In an instance, activation of cell death, e.g., ferroptosis, in the method disclosed herein results in inhibition of the Xc- system, GPX-4 inhibition, activation of lipid peroxidation, and/or increase of intracellular iron levels in the at least one cone photoreceptor. Notably, the inhibition of the Xc- system and/or of GPX-4 is preferably specific, i.e., the Xc- system and/or GPX-4 are inhibited in cone photoreceptors whilst they are substantially unaffected in other retinal cell types, such as rod photoreceptors and retinal pigment epithelium cells, at identical concentrations of cell death (in particular ferroptosis) activators. Likewise, lipid peroxidation is preferably specifically activated and/or increased in cone photoreceptors. Finally, intracellular iron levels are preferably specifically increased in cone photoreceptors.

In an instance, activation of cell death, e.g., ferroptosis, in particular specific activation of cell death in cone photoreceptors (i.e. cell death is not substantially activated in other retinal cell types such as rod photoreceptors and retinal pigment epithelium cells, in particular at identical concentrations of cell death activator) results in one or more of: • reduced viability of the at least one cone photoreceptor, in particular specific reduced viability of the at least one cone photoreceptor (i.e. the viability of other retinal cell types such as rod photoreceptors and retinal pigment epithelium cells, is not substantially reduced, in particular at identical concentrations of cell death activator),

• increased cell membrane permeability to dyes in the at least one cone photoreceptor, in particular specific increased cell membrane permeability to dyes in the at least one cone photoreceptor (i.e. cell membrane permeability to dyes is not substantially increased in other retinal cell types such as rod photoreceptors);

• reduced intracellular glutathione level in the at least one cone photoreceptor, in particular specific reduced intracellular glutathione level in the at least one cone photoreceptor;

• increased intracellular NADP/NADPH ratio in the at least one cone photoreceptor, in particular specific increased intracellular NADP/NADPH ratio in the at least one cone photoreceptor);

• increased intracellular formation of lipid peroxides in the at least one cone photoreceptor, notably 5-oxo-6, 8, 11 ,14-eicosatetraenoic acid (5oxoETE), in particular specific increased intracellular formation of lipid peroxides, notably 5 -oxo-6, 8,11 ,14-eicosa tetraenoic acid (5oxoETE) in the at least one cone photoreceptor;

• increased gene expression of AIFM2 (Apoptosis-inducing factor 2), SLC3A2 (Solute Carrier Family 3 Member 2), HSBP1 (Heat Shock Protein Family B (Small) Member 1 ), and Keapl (Kelch-like ECH -associated protein 1 ), in particular specifically in cone photoreceptors;

• increased protein expression of ferritin and/or nuclear receptor coactivator 4 (NCOA4), in particular specifically in cone photoreceptors.

In an instance, activation of cell death, e.g., ferroptosis, comprises contacting the at least one cone photoreceptor with at least one cell death inducer. Preferably, the cell death inducer is a ferroptosis inducer. More preferably, the at least one cell death, e.g., ferroptosis inducer, is a compound selected in the group consisting of glutamate, RSL3, erastin, an erastin derivative (preferably MEH, PE, or AE), imidazole ketone erastin (IKE), sulfasalazine, sorafenib, sorafenib analogs (SRS13-45 and SRS13- 60), altretamine, artesunate, artemisinin, dihydroatemisinin, artemether, an artemisinin derivative, ML-162 (DPI7), ML-210 (DPI10), RSL5, 4,4'- diisothiocyanostilbene-2,2'-disulfonic acid, temozolomide, MMRi62, t-tertiary-butyl hyperoxide, FIN56, buthionine sulfoximine, NDP4928, statins, brequinar, withaferin A (WA), auranofm, almitrine, lanperisone, HG106, DPI2, DPI12, DPI13, DPI17, DPI18, DPI19, 31MEW44, and ML160.

In an instance, the method disclosed herein comprises a prior step of obtaining a retinal tissue. In particular, the retinal tissue may be obtained from an eye or by production from stem cells. In a preferred instance, the retinal tissue comprises photoreceptors, notably cone photoreceptors, rod photoreceptors, and/or their progenitors. In a further preferred instance, the retinal tissue is a retinal tissue from a vertebrate including both human or non-human animal tissue, preferably, a nonhuman mammal, more preferably a non-human primate.

In a specific instance, the method disclosed herein comprises a further prior step of dissociating the retinal tissue enzymatically and/or mechanically and/or chemically, thereby obtaining a population of isolated retinal cells.

In a specific instance, the method disclosed herein comprises a further prior step of incubating the population of isolated retinal cells with peanut agglutinin (PNA) and recovering the PNA-bound cone photoreceptor.

In another instance, the retinal tissue is a retinal explant or a retinal organoid maintained in culture. For example, the retinal explant or retinal organoid is cultivated in CCh-independent medium.

In a preferred instance, the method disclosed herein comprises the further step of contacting the retinal tissue with a polycarbonate membrane. Preferably, the retinal tissue is contacted with the membrane so that photoreceptors are facing away from the membrane. According to a more preferred instance, contacting the at least one cone photoreceptor with at least one cell death inducer (e.g., a ferroptosis inducer) results in cone degeneration, in particular specific cone degeneration, or microglial migration to the outer nuclear layer of the retinal tissue. Even more preferably, the contacting the at least one cone photoreceptor with at least one cell death inducer (e.g., a ferroptosis inducer) results in specific cone degeneration, i.e., degeneration of cone photoreceptors is induced or increased, whilst the other retinal cell types, such as rod photoreceptors or retinal pigment epithelium cells are substantially unaffected at identical concentrations of cell death inducer (e.g. ferroptosis inducer).

In another instance, the retinal tissue is comprised in a complete eye.

In a preferred instance, contacting the at least one cone photoreceptor with at least one cell death inducer (e.g., a ferroptosis inducer) comprises administering the at least one cell death inducer (e.g., a ferroptosis inducer) in the subretinal space, the suprachoroidal space, the anterior chamber, the vitreous humour, the subconjunctival space, or on the corneal surface. In another preferred instance, administration of the at least one cell death inducer (e.g., a ferroptosis inducer) can be through systemic administration (inhalator, intravenous, intramuscular or intraperitoneal) or under solution, gel or implant.

According to another preferred instance, contacting the at least one cone photoreceptor with at least one cell death inducer (e.g., a ferroptosis inducer) results in the formation of subretinal deposits, preferably in the fovea and/or in the macula.

In another preferred instance, contacting the at least one cone photoreceptor with at least one cell death inducer (e.g., a ferroptosis inducer) results in a decrease, in particular in a specific decrease, of cone photoreceptor activity. More preferably, contacting the at least one cone photoreceptor with at least one cell death inducer (e.g., a ferroptosis inducer) results in a decrease, in particular in a specific decrease, of cone photoreceptor activity measured by ERG or mfERG.

In another aspect, the present disclosure relates to a model of cone degeneration obtainable by the method described herein.

In yet another aspect, the present disclosure provides a method of screening for compounds capable of showing protective and/or anti -degenerative properties on cone photoreceptors, the method comprising:

• providing the model of cone degeneration described above,

• contacting the model of cone degeneration with a candidate compound,

• assessing whether the candidate compound induces a protective and/or anti -degenerative effect, in particular a specific protective and/or anti- degenerative effect, on cone photoreceptors. Preferably, the candidate compound induces a protective and/or anti- degenerative effect, in particular a specific protective and/or anti -degenerative effect, on cone photoreceptors, which comprises at least one of:

• maintaining or increasing, in particular specifically maintaining or increasing, viability of the at least one cone photoreceptor,

• preventing, in particular specifically preventing, the increase in cell membrane permeability to dyes in the at least one cone photoreceptor;

• maintaining or increasing, in particular specifically maintaining or increasing, intracellular glutathione level in the at least one cone photoreceptor;

• maintaining or decreasing, in particular specifically maintaining or decreasing, the intracellular NADP/NADPH ratio in the at least one cone photoreceptor;

• maintaining or decreasing, in particular specifically maintaining or decreasing, the intracellular formation of lipid peroxides in the at least one cone photoreceptor, notably 5-oxo-6,8,11 ,14-eicosatetraenoic acid (5oxoETE);

• maintaining or increasing the gene expression of AIFM2 (Apoptosis-inducing factor 2), SLC3A2 (Solute Carrier Family 3 Member 2), HSBP1 (Heat Shock Protein Family B (Small) Member 1 ), and Keapl (Kelch-like ECH-associated protein 1 ), in particular specifically in cone photoreceptors;

• maintaining or decreasing the protein expression of ferritin and/or nuclear receptor coactivator 4 (NCOA4), in particular specifically in cone photoreceptors;

• preventing the proliferation, activation or migration of microglial cells;

• preventing the loss of cone outer segments;

• preventing, in particular specifically preventing, the decrease in cone activity, preferably cone activity as determined by ERG or mfERG; and

• preventing the thinning of the retinal thickness as indicated on histological section or in vivo imaging as with OCT.

LEGENDS OF THE FIGURES

FIGURE 1. Evidence of ferroptosis occurring in pure cone photoreceptors.

Mean values and standard errors of the mean are shown. A-B. Brightfield image of cone photoreceptors after 3 days of incubation with control (Ctrl = 0.2% DMSO, A) and treated solutions (3pM imidazole ketone erastin, IKE, B). C- D. Cone photoreceptor staining with calcein labelling viable cells in control (C) and treated (D) conditions. E. Glutamate concentration-dependent cell viability of pure cone photoreceptors with or without adding 1 mM L-Cystine (n=7, p<0.0001 , IC50 = 4.15pM). F. Levels of glutathione in pure cone photoreceptors in control, 3pM IKE and 500pM glutamate mediums over 3 days (D1 , D2, D3 n=18, p<0.0001 ). G. NADP/NADPH percentages in pure cone photoreceptors in control and 3pM IKE conditions at 3 days (n=4, p=0.0159). H. Concentration-dependent plot of pure cone photoreceptors cell viability after treatment with RSL3, from 0.01 nM to 25nM (n=3, IC50 =8.67nM). I. Western Blot revealing 5 -lipoxygenase (5LOX) and 5-lipoxygenase-activating protein (FLAP) bands in pure cone photoreceptors, absent in rod photoreceptors. J. Lipidomic analysis showing the amount of 15- and 5-hydroxyeicosatetraenoic acid (15-HETE and 5-HETE, respectively) and 5-Oxo-6,8,11 ,14-eicosatetraenoic acid (5oxoETE) detected in pure cone photoreceptors in control and 3pM IKE conditions (n= 6, p=0.0079 (“), p=0.0159 (*)). K. Pure cone photoreceptors viability in control and 3pM IKE conditions when adding various treatments (with antiferroptotic or iron chelating effects): L-Cystine, Zileuton (ZEN), Ferrostatin-1 (FST1 ) and deferiprone (DF).

FIGURE 2. Alteration of mRNA expression levels for proteins involved in ferroptosis in control versus treated (3 M IKE) pure porcine cone photoreceptors (n=6). Solute Carrier Family 11 Member 2 (SLC11 A2) is involved in the releasing of Fe²⁺ from the endosome into the cell cytoplasm; its decline suggests an iron metabolism dysregulation (p=0.0260). Solute Carrier Family 3 Member 2 (SLC3A2) corresponds to the heavy chain of the System Xc- (p=0.0022). Expression of transferrin receptor (TFRC) modestly increased under treatment condition (p=0.0286). Ferroptosis suppressor protein 1 (AIFM2) is involved to reduce ferroptosis mechanism (p=0.0022). Heat shock protein family B small member 1 (HSPB1 ) carries a role in the iron uptake and may promote ferroptosis²² (p=0.0022).

FIGURE 3. Immunofluorescence localisation of Flap on retinal sections from 3 different species (pig, rat and non-human primate) with respect to the DAPI nuclear staining. In all species, Flap is localised in cone photoreceptors.

FIGURE 4. Immunofluorescence of Ferritin heavy chain (B, G) and NCOA4 (C, H) in control (A-E) and treated (F-J) isolated porcine cone photoreceptors. The IKE treatment increased the fluorescence intensities (G, H) as compared to control cone photoreceptors. IKE condition (G, H) showing more immunolabelling than controls (B, C). E. Zoom of the merged labelling showing control cells with a more intense ferritin heavy chain immunolabelling than NCOA4. J. Zoom on treated cells showing a more intense NCOA4 immunolabelling than heavy chain.

FIGURE 5. Cone photoreceptors degeneration and microglial migration in ex vivo retinal explants. A - H. Histology of retinal sections showing an increase in permeability to ethidium in the first nuclear retinal row representing cone photoreceptors in the treated explant (20pM IKE, F) versus the control tissue (0.2% DMSO, B). Note also the disorganisation of Flap-immunolabeled cone photoreceptors in the treated retina (G) as compared to the control condition (C) as well as the presence of large amoeboid microglial cells in the treated retina (H) up to the outer nuclear layer (ONL) not present in the control retina (D). The nuclear layers are underlined in the control (A) and treated retina (E) by DAPI. I-M. Cone photoreceptor destruction indicated by the loss of their Flap-immunopositive outer segment on a treated flat mounted retinal explant (20pM IKE) and their protection by antiferroptotic agents FST-1 (L) and DF (M) as compared to the control condition (J). The quantification of cone photoreceptor outer segments demonstrates a significant decrease in treated explant and a rescue by the aniferroptotic agents (I, n=3). N-O. Microglial cell distributions in the retinal explant thickness showing their migration toward the outer retina in treated retina (IKE, RSL3) and a prevention of this effect by DF or FST-1 (n=3) . Scale bars = 20pm.

ONL: outer nuclear layer, INL: inner nuclear layer, GCL: ganglion cell layer, Iba1 : ionised calcium-binding adapter molecule 1 , Ctrl: control condition, IKE: imidazole ketone erastin, OS: outer segment, FST-1 : ferrostatin-1 , DF: deferiprone

FIGURE 6. Maintenance of different retinal cell types on sections from treated porcine retinal explants (20pM IKE) as compared to the control condition (Ctrl). Rod photoreceptors were immunolabelled against rhodopsin, bipolar cells against PKC alpha and Muller cells against vimentin. Note that retinal structure underlined by the DAPI stained nuclei, and the morphologies of the labelled cells appear preserved in the treated condition, without evident cell loss. FIGURE 7. Effects of RSL3 on retinal explants. RSL3 permeability to ethidium revealed dying cells in control (0.2% DMSO, B) and treated (200nM RSL3, F) retinae. C, G. Morphological changes of Flap-immunolabelled cone photoreceptors in a treated retinal explant (G) as compared to a control explant (C) with the migration of iba1 - immunopositive microglial cells toward the outer nuclear layer (ONL) in treated explant (H) with respect to the control condition (D). I-L. Loss of Flap-immunopositive cone outer segments (OS) in retinal explants treated by RSL3 (J) and their preservation by FST-1 (K) or DF (L) as related to the control condition (I). M. Quantification of the RSL3 toxicity on cone photoreceptor outer segments and their preservation by DF in retinal explants (n=3, p=0.0286). N. Distribution of microglial cells showing their migration towards the outer nuclear layer (ONL) in the presence of RSL3 as compared to the control condition (see Fig. 5N) with the effects of DF or FST-1 (n=3).

Scale bars = 20pm. ONL: outer nuclear layer, INL: inner nuclear layer, GCL: ganglion cell layer, Iba1 : ionised calcium-binding adapter molecule 1 , Ctrl: control condition, IKE: imidazole ketone erastin, OS: outer segment, FST-1 : ferrostatin-1 , DF: deferi prone.

FIGURE 8. Cone photoreceptor degeneration and microglial migration in rats following in vivo subretinal delivery of ferroptosis inducers. A. Schematic representation of the experimental protocol and timeline. Rats were delivered 20 M IKE or 200nM RSL3 (treated) or 0.2% DMSO (control, Ctrl) in the subretinal space. Retinas were retrieved at day 6 (D6) to assess microglial migration (K-O, n=4) or at day 30 (D30) to assess cone photoreceptor degeneration (H-J, n=4). B-G. Sample of in vivo retinal imaging of rats: fundus photography (B, C) and optical coherence tomography (OCT) (D-G) showing the induced retinal detachment (DO, B,D) beside the optic nerve (ON), with fully reattached retina at D6 (C, E); the injection site (arrowheads) and the OCT slice (dashed lines) are shown on fundus photography; (F, G) OCT of the injected area at D6 showing hyperreflective punctuations (arrows) spotted in the inner nuclear layer (INL) and at the outer plexiform layer (OPL)/outer nuclear layer (ONL) junction, more abundantly in the treated condition (G). H, I. Retinal sections (same rats as in B-G) within the injected area, collected at D30 showing a decrease in cone-arrestin immunolabelled cone photoreceptors (CAR) in treated animals (I) as compared to control (H) with the corresponding DAPI nuclear stainings. The eccentricity relative to the optic nerve is the same in both cases. J. Quantification of cone densities (in elements/mm) measured on retinal sections in treated and control rats at D30 (n=4). Outer segments (OS) and perikarya of cones were quantified separately. K. Quantification of microglial density on whole-mounted retinae of treated and control rats at D6. Three retinal plans were evaluated separately: the level of the outer segments (OS), the ONL plan and the pedicles (Ped) plan, which corresponds to the OPL/ONL junction. Each field corresponds to a 500x500pm window, repeated over the entire injected area. L-O. Distributions of cone-arrestin positive cells and Iba1 -positive microglial cells in the whole-mounted retinae of treated and control rats collected at D6, showing two different plans on the z axis (illustrated in H-l): OS (L, M) and Ped (N, O). The samples were picked in the detached area but away from the injection site to illustrate the change in cone OS morphology and the invading microglial cells for treated animals. The eccentricity relative to the optic nerve is the same in both cases.

Scale bars: 50pm. INL: inner nuclear layer, OPL: outer plexiform layer, ONL: outer nuclear layer, EZ: ellipsoid zone, RPE: retinal pigment epithelium, DAPI: 4', 6- diamidino-2-phenylindole, CAR: cone arrestin, Iba1 : ionised calcium binding adaptor molecule 1 , Ctrl: control.

FIGURE 9. Macular changes in non-human primates (NHP) following submacular delivery of 20pM IKE. A. Schematic representation of the experiment protocol. The subretinal delivery of 20pM IKE in the right eye and DMSO diluted in PBS at 1 /500 in the left eye allowed to detach the superior hemifovea in both eyes. B-C. Fundus photography at baseline (B), 1 (C) and 3 months (D) following the subretinal injection of 20pM IKE. Demarcation line at 1 month (arrowheads in C) and macular pigmentation surrounding subretinal deposits occurring at 3 months (highlighted in D) are shown. E-F. OCT vertical b-scans through the foveal spot of NHP three months after a subretinal delivery of 1 /500 DMSO (ctrl, E) and 20pM IKE (F). Note the eroded ellipsoid zone (EZ, black asterisk), punctuate hyperreflectivities (arrows) and subretinal deposits (arrowhead). G. OCT horizontal b-scan through the upper parafoveal area within the subretinal delivery of 20pM IKE. The EZ is eroded (black asterisks) within the detached area borderlines (dashed lines). Note also bunches of hyperreflective foci mainly located in the ONL and below the inner limiting membrane (arrows). The uninjected area (outside of the dashed lines) remains perfectly normal. H-J. OCT horizontal b-scans through a macular subretinal deposit (arrowheads) that appeared at M2 (I) and grew until M3 (J). K-L. En face adaptive optics imaging of the foveal photoreceptor layer at 2 degrees of eccentricity, within the detached area, 3 months following the subretinal delivery. Cones are less reflective and clusters with vanishing cones appeared following 20pM IKE delivery (K), whereas the cone mosaic of the control side remains regular (L). M-N. Illustration of the ONL thickness changes in the macular area of NHP 3 months after the subretinal delivery. The grey scale shows ONL thickness changes (in pm). O. Cone density changes (in cones/deg²) measured at 2 degrees within and without the injected area using adaptive optics, 3 months after the subretinal delivery (n=2). P. Changes in N1 - and P1 -wave amplitudes as a percentage of baseline measured on multifocal electroreti nogram (n=2). Only hexagons within the injected area were analysed.

FIGURE 10. Ferroptosis pathways elucidated in mammalian cone photoreceptors. Ferroptosis ultimately drives ROS formation that leads to cone photoreceptor degeneration. Three subcellular pathways must be considered in ferroptosis in cones. Firstly, the System Xc-, that can be inhibited by high extracellular concentration of glutamate of by ferroptosis inducers (e.g. , IKE), stimulates the entry of L-cystine to produce glutathione. Inhibition of the glutathione redox cycle results in glutathione peroxidase 4 (GPX4) inactivation and accumulation of NADPH. Additionally, the reduced activity of GPX4 enzyme leads to the production of peroxidised lipids such as 5-oxo-ETE via the lipoxygenase pathway. Finally, a high ferritinophagy activity of NCOA4 leads to a dysregulated iron metabolism, resulting in an excessive amount of iron release. Radical species are formed by Fenton’s reaction, and allow for reactive oxygen species production via interaction with peroxidised lipids.

GPX4: Glutathione peroxide 4, 12-LOX: 12 lipoxygenase, 15-LOX: lipoxygenase, 5-LOX: 5 lipoxygenase, FLAP: 5-lipoxygenase-activating protein, 12-HPETE: 12 hydroxyperoxyeicosatetraenoic, 15-HPETE: 15 hydroxyperoxyeicosatetraenoic, 5- HPETE: 5 hydroxyperoxyeicosatetraenoic, 5-oxo-ETE: 5-0xo-eicosatetraenoic, 12- HETE: 12-hydroxyeicosatetraenoic, 15-HETE: 15-hydroxyeicosatetraenoic, 5-HETE: 5- hydroxyeicosatetraenoic, STEAP 3: STEAP family member 3, DMT1 : Dimetal transporter 1 , NCOA4: nuclear receptor coactivator 4, Keapl : Kelch-like ECH- associated protein 1 , NFE2L2: Nuclear factor (erythroid-derived 2)-like 2. ROS: reactive oxygen species.

FIGURE 11. Cell viability after pro-ferroptotic treatment on different cell type in the retina. Cone photoreceptors (Cone PRs) are represented by a light grey line and a round shape, ARPE-19s by a dark grey line and a triangular shape, and RPE primary cells by a black line and a square shape. A. Cells were treated with IKE (Imidazole Ketone Erastin) from 0.20 pM to 25 pM. All cone PRs died at 25 pM while 40% of ARPE-19 (n=3, p=0.0001 ) and 80% of RPE primary cells (n=3, p=0.0019) remained alive. B. Cells were treated with RSL3 (Ras Selective Ligand 3) from 0.01 pM to 10 pM. All PR cone cells died at all concentrations, whereas at 5 pM 50% of ARPE-19 (n=3, p=0.0008) and primary RPE cells remained alive. C. Cells were treated with glutamate from 0.001 mM to 1 mM showing no effect on ARPE-19 and primary RPE cells. In contrast, cone PRs all died at 1 mM compared to ARPE-19 cells (n=3, p=0.0050) and primary RPE cells.

FIGURE 12. Ferroptosis inhibitors reverse cone degeneration in rd1 mice between P15 and P45 ex vivo and in vivo. A-D. PNA-immunolabeled retinal explants of rd1 mice exposed to a control solution (A) and to several ferroptosis inhibitors in culture for 30 days (from P15 to P45) (B-D). E. Automated cone counting on retinal explants of rd1 mice exposed in culture with various ferroptosis inhibitors (Ctrl, HTHQ, 1 ,3-Dicaffeoylquinic acid, SRS16-86, Astilbin, caffeic acid, Tinoridin Hydrochloride n=3; GW5074 n=6; Nobergenin n=2; Mangiferin n=5). White bars represent antioxidants or radical scavengers, gray bars represent molecules involved in the Xc system pathway, checkerboard bars represent molecules acting on the lipoxygenase pathway, and vertical bars represent molecules involved in the iron pathway. F-l. PNA- immunolabeled whole-mounted retinae of rd1 mice daily injected intraperitoneally with the vehicle (Ctrl, F and H) or a 2 mg/kg GW 5074 solution for 30 days (from P15 to P45) (G and I). The asterisks indicate the position of H and I. J. Automated cone counting on whole-mounted retinae after daily intraperitoneal injections (n=5) . Scale bars = 50 m in A-D, H, I; 300 m in F, G. PNA: peanut agglutinin, Ctrl: control, HTHQ: 1-0-hexyl-2,3,5-trimethylhydroquinone, 1-3-DCFA: 1 ,3- Dicaff eoylquinic acid, RA: Rosmarinic acid, SMNL: Sonlicromanol, CA: Caffeic acid, Tinoridin HCL: Tinoridin hydrochloride, OH puerarin: Hydroxy purrarain, 5-ACQ: Chlorogenic acid.

FIGURE 13. Ferroptosis inhibitors do not reverse rod degeneration in rd1 mice between P15 and P45 ex vivo and in vivo. A. Automated cone counting on retinal explants of rd1 mice exposed in culture with various ferroptosis inhibitors that rescued cones (Ctrl, HTHQ, 1 ,3-Dicaffeoylquinic acid, SRS16-86, Astilbin, caffeic acid, Tinoridin Hydrochloride n=3; GW5074 n=6; Nobergenin n=2; Mangiferin n=5). B. Automated cone counting on whole-mounted retinae after daily intraperitoneal injections (n=5).

FIGURE 14. Rods immunolabeled with rhodopsin (Rho) and DAPI in the nonhuman primate retina injected with the control solution and with 20 pM IKE.

DESCRIPTION

Definitions

The term “about” or “approximately” refers to the normal range of error for a given value or range known to the person of skills in the art. It usually means within 20%, such as within 10%, or within 5% (or 1% or less) of a given value or range. Reference to “about” a value or parameter herein includes (and describes) embodiments that are directed to that value or parameter per se.

As used herein, the terms “activate”, “stimulate, or “induce”, and any grammatical derivative thereof, refer to a comparative increase in a specified response of a designated material (e.g., expression, enzymatic activity) in the presence of a specified reagent. The reagent is described herein as an “activator”.

The term “animal” is used herein to include all animals. In some instances, the non-human animal is a vertebrate, notably a mammal. Examples of animals are human, non-human primates, mice, rats, cows, pigs, horses, chickens, ducks, geese, cats, dogs, etc. The term “animal” also includes an individual animal in all stages of development, including embryonic and fetal stages.

“Cell death”, as used herein, refers to any mechanism and/or pathway whereby a cell undergoes a series of events which ultimately would lead to the death of the cell. Cell death may be induced in individual cells as a consequence of numerous internal and external stimuli including, but not limited to, genetic predisposition, toxic chemicals or processes, heat, cold, rapid environmental changes, radiation, viruses, prions, bacteria, disruption of nutrient balance, or exposure to by-products and signalling from other cells undergoing cell death. “Cell death” encompasses both specific cell death and necrosis. “Specific cell death” or “regulated cell death” or “active cell death”, as used, herein refers to a group of processes which are characterised by specific mechanism and/or pathways whose activation leads to the death of the cell. Specific cell death as used herein includes, but is not limited to, apoptosis or programmed cell death, necroptosis, parthanatos, entotic cell death, netotic cell death, parthanatos, lysosome-dependent cell death, autophagy-dependent cell death, alkaliptosis, oxeiptosis, NETosis, pyronecrosis, pyroptosis, and ferroptosis/oxitosis. Many specific cell death pathways are known in the art; see e.g., Vanden Berghe et al. Nat Rev Mol Cell Biol. 15(2) : 135-47, 2014; Green. Cell. 177(5): 1094-1107, 2019; Tang et al. Cell Res 29: 347-364, 2019; Lomphithak & Fadeel. Toxicol Sci. kfad008, 2023.

“Cell death inducer” or “cell death activator” refers to an agent which induces, promotes or activates cell death.

As used herein, “contacting” means bringing a compound and optionally one or more additional compounds, e.g., therapeutic agents, into close proximity to the sample such as cells in need of such modulation. This may be accomplished using conventional techniques of drug delivery to the subject or in the in vitro situation by, e.g., providing the compound and optionally other compounds, e.g., therapeutic agents, to a culture media in which the cells are located.

The term “decreased” or “reduced”, as used herein, refers to the activity of a protein at least 1 -fold (e.g. , 1 , 2, 3, 4, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 1000, 10,000- fold or more) lower than its reference value. “Decreased” or “reduced”, as it refers to the activity of a protein of a subject, signifies also at least 5% lower (e.g. , 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%), 95%), 99%), or 100%) than the activity of the protein in the reference sample or with respect to the reference value for said protein. The term “decreased” or “reduced”, as used herein, also refers to the level of a biomarker of a subject at least 1 -fold (e.g. , 1 , 2, 3, 4, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 1000, 10,000- fold or more) lower than its reference value. “Decreased” or “reduced”, as it refers to the level of a biomarker of a subject, signifies also at least 5% lower (e.g., 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%), 95%), 99%), or 100%) than the level in the reference sample or with respect to the reference value for said marker. As used herein, “electroretinogram (ERG)” refers to a diagnostic test which measures electrical activity of the retina in response to a light stimulus. ERG can for example record a massed potential from the whole retina; see e.g., full-field ERG. Alternatively, multifocal ERG (mfERG) assesses ERG activity in small areas of retina, whilst pattern ERG (pERG) assesses macular retinal ganglion cell (RGC) activity. ERG can be an objective measure of retinal function that can be recorded under physiological conditions. For example, ERG can be used to provide diagnostic information, monitor the progression of retinal diseases and disorders, or a combination thereof. ERG can be used to determine the effectiveness of any of the various therapeutic compounds disclosed herein. Examples of ERG values from healthy subjects and patients can be found in Lorenz et al., Invest. Ophthalmol. Vis. Sci. 49:5235-5242, 2008.

As used herein the term “ex vivo” refers to a process in which cells, which are removed from a living organism, are cultured outside the organism (e.g., in a cell culture plate, flask, bag or test tube).

As used herein, “ferroptosis” refers to a form of cell death understood in the art as involving generation of reactive oxygen species mediated by iron, and characterised by, in part, lipid peroxidation. In ferroptosis, lethality occurs as a result of peroxidation of polyunsaturated fatty acids (PUFAs)that self-propagate unless halted by the lipid peroxidase, glutathione peroxidase 4 (GPX4).

“Ferroptosis inducer” or “ferroptosis activator” refers to an agent which induces, promotes or activates ferroptosis. As used herein, the term “Ferroptosis inducer” or “ferroptosis activator” refers to any compound or means that when brought into contact with a cell, e.g., a cone photoreceptor, induces ferroptosis in that cell, and/or a compound when administrated to a subject result in the induction of, or in the increase of, ferroptotic cell death, e.g., of cone photoreceptors.

As used herein, “GPX4” refers to glutathione peroxidase 4, a glutathione metabolism enzyme.

The term “GPX4 inhibitor” as used herein refers to any agent that inhibits the activity of the enzyme glutathione peroxidase 4 (GPX4). A GPX4 inhibitor can be either a direct or indirect inhibitor. GPX4 is a phospholipid hydroperoxidase which catalyses the reduction of hydrogen peroxide and organic peroxides, thereby protecting cells against membrane lipid peroxidation, or oxidative stress. An indirect inhibitor blocks the formation of or depletes the concentration of glutathione. A non-limiting example is buthionine sulfoximine (BSO). A direct inhibitor of GPX4 acts to prevent binding of either or both glutathione or a lipid-hydroperoxidase in the GPX4 active site. GPX4 has a selenocysteine in the active site that is oxidised to a selenenic acid by the peroxide to afford a lipid-alcohol. The glutathione acts to reduce the selenenic acid (-SeOH) back to the selenol (-SeH). When this catalytic cycle is disrupted, cell death occurs through the specific cell death process known as ferroptosis. Non-limiting examples of direct GPX inhibitors include RSL3 and ML162.

The term “increased”, as used herein, refers to the activity of a protein, at least 1 -fold (e.g., 1 , 2, 3, 4, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 1000, 10,000- fold or more) greater than its reference value. “Increased”, as it refers to the activity of a protein of a subject, signifies also at least 5% greater (e.g. , 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 99%, or 100%) than the activity of the protein in the reference sample or with respect to the reference value for said protein. The term “increased”, as used herein, also refers to the level of a biomarker of a subject at least 1 -fold (e.g. , 1 , 2, 3, 4, 5, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 1000, 10,000-fold or more) greater than its reference value. “Increased”, as it refers to the level of a biomarker of a subject, signifies also at least 5% greater (e.g. , 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 99%, or 100%) than the level in the reference sample or with respect to the reference value for said marker.

As used herein, the terms “inhibit” or “inhibition” refer to a comparative decrease in a specified response of a designated material (e.g., expression, enzymatic activity) in the presence of a specified reagent. The reagent is described herein as an “inhibitor”.

As used herein the term “isolated” refers to at least partially separated from the natural environment e.g., the human body.

As used herein, “measure” or “determine” refers to any qualitative or quantitative determinations. As used herein, the term “lipid peroxidation” means the consequence of free radical damage to lipids. Lipid peroxidation forms a number of oxidation products, including lipid peroxides and aldehydes such as malondialdehyde (MDA) and 4- hydroxynonenal (4-HNE).

As used herein, the terms “lipid peroxides” or “peroxided lipids” refer to the oxidation products of phospholipids and polyunsaturated fatty acids (PUFAs).

As used herein, a “lipoxygenase” is an enzyme that oxidises polyunsaturated fatty acids or alkenes with two oxygen atoms to eicosanoids. It catalyses the first reaction in the so-called lipoxygenase pathway, which is involved in the response to external trauma and stress. Examples of lipoxygenases in humans include 15-LOX, 15- LOX-2, 12-LOX, 12R-LOX, eLOX-3, and 5-LOX. In some instances, lipoxygenase activity requires the presence of an activator. For example, the helper protein 5-LOX activating protein (FLAP) is necessary for the activation of the 5-LOX enzyme. As used herein, the term “lipoxygenase” also includes those coactivators, notably FLAP.

A “lipoxygenase inhibitor”, as used herein, refers to an agent that inhibits the activity of a lipoxygenase. Exemplary lipoxygenase inhibitors include PD146176 and ML351.

As used herein, “optical coherence tomography (OCT)” refers to a non- invasive imaging test that uses light waves to capture cross-sectional images of a retina. OCT can be used to distinguish layers of the retina, map and measure thickness, inform treatment decisions, provide diagnostic information, monitor disease progression, or a combination thereof. Notably, OCT can be used to quantify cone degeneration. In embodiments, subjects who have not been diagnosed with a retinal degenerative disease have a peak cone density in the fovea of about 200,000 cells/mm² as measured by OCT, and/or a foveal outer nuclear layer thickness of about 100 m as measured by OCT. Preferably, patients diagnosed with retinal degenerative disease have a peak cone density in the fovea of less than about 200,000 cells/mm² as measured by OCT. Patients diagnosed with a retinal degenerative disease can have a foveal outer nuclear layer thickness of less than about 100 pm as measured by OCT. Additional examples of OCT values from healthy subjects and patients diagnosed with a retinal degenerative disease can be found in Lorenz et al., 2008. By “organoids”, it is herein referred to tiny, self-organised three-dimensional tissue cultures that are derived from stem cells. An “organoid” as used herein usually replicates much of the complexity of an organ, or to express selected aspects of it like producing only certain types of cells. Organoids can range in size from less than the width of a hair to five millimetres. There are potentially as many types of organoids as there are different tissues and organs in the body. To date, researchers have been able to produce organoids that resemble the brain, the retina, kidney, lung, intestine, stomach, and liver, and many more are on the way.

A “polyunsaturated fatty acid (PUFA)”, as used herein, refers to a fatty acid containing at least two -CH=CH- groups, such as linoleic acid, linolenic acid, and arachidonic acid (AA).

As used herein, the term “specifically” means that the compound, at a given concentration, affects one cell type (namely cone photoreceptors), while not substantially affecting other retinal cell types such as rod photoreceptors or retinal pigment epithelial cells, at the same given/identical concentration. In particular, when referring to an in vivo use, the term “specifically” means that the compound affects one cell type (namely cone photoreceptors) before substantially affecting or without substantially affecting other retinal cell types such as rod photoreceptors or retinal pigment epithelial cells.

As used herein, the term “substantially” refers to the qualitative condition of exhibiting total or near-total extent or degree of a characteristic or property of interest. One of ordinary skill in the biological arts will understand that biological and chemical phenomena rarely, if ever, go to completion and/or proceed to completeness or achieve or avoid an absolute result. For example, “substantially” refers to refers to at least about 60%, at least about 70%, at least about 75%, at least about 80%, at least about 85%, at least about 90%, at least about 95%, at least about 98%, at least about 99%, or about 100%. The term “substantially” is therefore used herein to capture the potential lack of completeness inherent in many biological and chemical phenomena.

The term “system Xc-” refers to an amino acid antiporter that mediates the exchange of extracellular-L-cystine and intracellular L-glutamate across the cellular plasma membrane. The system Xc - exchanges intracellular glutamate for extracellular cystine, thereby supporting intracellular glutathione (GSH) synthesis as well as non-vesicular glutamate release thus leading to glutathione production and oxidative protection. System Xc- is a heterodimeric amino acid transporter which consists of xCT, a light-chain subunit that confers cystine transport function and is encoded by the SLC7A11 gene, and the CD98 heavy-chain subunit, which localises system Xc- to the plasma membrane. As used herein, inhibition of system Xc- refers to the prevention of glutamate exit and/or cystine entry by ligands. As used herein, activation of the system Xc- refers to the stimulation of glutamate exit and/or cystine entry by ligands. System Xc- is a heterodimeric amino acid transporter which consists of xCT, a light-chain subunit that confers cystine transport function and is encoded by the SLC7A11 gene, and the SLC3A2 heavy-chain subunit, which localises system Xc- to the plasma membrane.

Preparation of a model of cone photoreceptor degeneration

The present disclosure provides a method for preparing a model of cone photoreceptor degeneration.

The present inventors have surprisingly found that that degeneration of cone photoreceptors is caused by activation of specific cell death pathways. Notably, they have shown that inducers of ferroptosis such as glutamate, erastin, imidazole ketone erastin (IKE), and RSL3 cause a specific drop in cone photoreceptors viability. This loss of viability of cone photoreceptors is accompanied by a decrease in glutathione (GSH) levels, an increase in lipid peroxides, and an increase of iron metabolism, notably an increase of intracellular iron levels. In sharp contrast, no such effect was observed in retinal pigment epithelium cells. Not only is this specific effect observed in pure cone photoreceptors in culture, it is also apparent in an ex vivo model of retina, wherein in addition microglia migrated as they do in vivo. The significance to in vivo pathological processes was further emphasised by the finding that administration of inducers of ferroptosis into the eyes of non-human primates leads not only to cone loss, but also to formation of subretinal deposits and decrease in electroreti nogram amplitude, features highly similar to those exhibited by AMD patients. Surprisingly, even though ferroptosis mediators such the Xc- system and the GPX-4 enzyme are present in both cone and rod photoreceptors (Hu et al. Eur J Neurosci. 28(8):1491 -502, 2008), administration of such inducers of ferroptosis into the eyes does not affect rod photoreceptors viability, further emphasising the specificity of the effects of ferroptosis on cone photoreceptors.

Thus, contacting cone photoreceptors with cell death inducers (e.g., ferroptosis inducers) triggers cone degeneration. This is particularly helpful because it provides an easy and reproducible method for preparing a model of cone photoreceptor degeneration. Moreover, the method is not limited to the preparation of an in vitro model, but can also be applied to the preparation of ex vivo and/or in vitro models, thus allowing a wider range of applications.

In a first aspect the disclosure provides a method for preparing a model of cone photoreceptor degeneration. The method comprises a step of activating a specific cell death in at least one cone photoreceptor, more particularly a step of specifically activating a specific cell death in at least one cone photoreceptor, i.e. no specific cell death is substantially activated in other retinal cell types such as rod photoreceptors and retinal pigment epithelial cells, at identical concentrations of cell death (in particular ferroptosis) activators.

Accordingly, the present disclosure provides a method for preparing a model of cone degeneration, comprising a step of activating a specific cell death in at least one cone photoreceptor, more particularly a step of specifically activating a specific cell death in at least one cone photoreceptor, i.e. no specific cell death is substantially activated in other retinal cell types such as rod photoreceptors and retinal pigment epithelial cells, at identical concentrations of cell death (in particular ferroptosis) activators.

The specific cell death according to the present disclosure can be any type of regulated cell death known in the art; see e.g., Vanden Berghe et al. Nat Rev Mol Cell Biol. 15(2):135-47, 2014; Green. Cell. 177(5):1094-1107, 2019; Tang et al. Cell Res 29: 347-364, 2019; Lomphithak & Fadeel. Toxicol Sci. kfad008, 2023.

Preferably, the specific cell death activated in the method described herein is ferroptosis. Ferroptosis is known to be regulated by many signalling pathways, including, but not limited to, the glutamate/cystine anti porter system Xc-, glutathione peroxidase-4 (GPX-4), the lipid metabolism pathway, and the iron metabolic pathway. Indeed, it has been shown that each of inhibition of the system Xc-, inhibition of GPX- 4, lipid peroxidation, and modulation of iron metabolism is sufficient to induce ferroptosis. Indeed, the present inventors have shown any of these mechanisms can induce cell death and lead to cone degeneration.

Accordingly, activation of cell death according to the method described herein results in inhibition of the Xc- system, GPX-4 inhibition, stimulation of lipid peroxidation, and/or increase in intracellular iron levels, in particular results in specific inhibition of the Xc- system, specific inhibition of GPX-4, specific stimulation of lipid peroxidation and/or specific increase in intracellular iron levels in cone photoreceptors, i.e. for example the Xc- system and GPX-4 are not substantially inhibited in other retinal cell types such as rod photoreceptors and retinal pigment epithelial cells at identical concentrations of cell death activators.

In a particular instance of the method described herein, activation of cell death, e.g., ferroptosis, results in one or more of the following features:

• reduced viability of the at least one cone photoreceptor, in particular specific reduced viability of the at least one photoreceptor (i.e. the viability of other retinal cell types such as rod photoreceptors and retinal pigment epithelium cells, is not substantially reduced, in particular at identical concentrations of cell death activator),

• increased permeability of the cell membrane to dyes in the at least one cone photoreceptor, in particular specific increased cell membrane permeability to dyes in the at least one cone photoreceptor (i.e. cell membrane permeability to dyes is not substantially increased in other retinal cell types such as rod photoreceptors),

• increased intracellular NADP/NADPH ratio in the at least one cone photoreceptor, in particular specific increased intracellular NADP/NADPH ratio in the at least one cone photoreceptor;

• increased intracellular formation of lipid peroxides in the at least one cone photoreceptor, notably 5-oxo-6, 8, 11 ,14-eicosatetraenoic acid (5oxoETE), in particular specific increased intracellular formation of lipid peroxides, notably 5-oxo-6,8,11 ,14-eicosatetraenoic acid (5oxoETE) in the at least one cone photoreceptor; • increased gene expression of AIFM2 (Apoptosis-inducing factor 2), SLC3A2 (Solute Carrier Family 3 Member 2), HSBP1 (Heat Shock Protein Family B (Small) Member 1 ), and Keapl (Kelch-like ECH-associated protein 1 ), in particular specifically in cone photoreceptors;

These features can be assayed by any technique or method known in the art.

In particular, viability of the cone photoreceptor cells can be measured by any means known to the skilled person. For example, the viability of these cells can be assayed in vitro with cell membrane permeability to dyes e.g., calcein AM, MTT, alamarBlue, ethidium, CellTiterGlo. Alternatively, cone viability can be assayed in an ex vivo model of retina, by determining the number of cones after labelling the retinal tissue with a reagent specific of cone outer segments or cones (e.g., an anti-FLAP antibody, an anti-opsin antibody, the peanut agglutinin lectin an anti-cone arrestin antibody). In yet another instance, cone viability can be assayed in vivo by optical coherence tomography (OCT) or adaptative optics retinal imaging. For example, OCT is widely employed clinically to characterise potential retinal degeneration in patients, because it permits to define the location and nature of the changes in the retina and adjacent structures and objectively evaluates the thickness of the retina and surrounding structures. The prevention of decrease of cone degeneration can thus be measured via OCT. All these techniques are well known to the skilled person. The cone cell survival can also be demonstrated on histological samples by measuring the thickness of the retina, measuring the number of nuclei in the ONL (e.g. after DAPI staining), or determining the number of cones after immunolabelling the retinal tissue with a reagent specific of cone outer segments or cones (e.g., an anti-FLAP antibody, anti-opsin antibody, the peanut agglutinin lectin or an anti-cone arrestin antibody). It is understood that the skilled person will be capable of adapting each of these methods as required by the circumstances to effectively measure the cone photoreceptor viability.

Many commercial assays for measuring glutathione levels are available to the skilled person, including the one used in the examples. Likewise, there are many available commercial assays for measuring NADP/NADPH ratio, including the one used in the examples. Quantitative analysis of lipid peroxides in biological samples can be performed in a number of ways known to the skilled person, including liquid chromatography-mass spectrometry (LC-MS/MS) and ultra performance liquid chromatography (UPLQ-MS/MS (Chhonker et al. Bioanalysis. 10(24):2027-2046, 2018).

When expression levels are measured at the polynucleotide level, it may be notably performed using well known technologies such as high throughput sequencing, quantitative PCR or nucleic acid microarray technologies (including cDNA and oligonucleotide microarrays). These technologies are now used routinely by those skilled in the art and thus do not need to be detailed here. Alternatively, tissue microarrays coupled to fluorescent in situ hybridisation may be used.

When expression levels are measured at the protein level, it may be notably performed using specific antibodies, in particular using well known technologies such as western blot, ELISA or ELISPOT, antibodies microarrays, or tissue microarrays coupled to immunohistochemistry.

Preferably, activation of cell death, e.g., ferroptosis, results in one of the above features. More preferably, activation of cell death, e.g., ferroptosis, results in 2, 3, 4, 5, 6, or 7 of the features. These features can be assayed by any technique or method known to the skilled person. The skilled person will notably refer to the examples for particular instances of such assays.

In a specific instance, the method disclosed herein comprises a step of contacting at least one cone photoreceptor with a cell death inducer (e.g., a ferroptosis inducer), thereby activating, in particular specifically activating, cell death (e.g., ferroptosis) in said cone photoreceptor. Cell death inducers as described herein can act through any of the pathways leading to cell death. For example, cell death inducers as used herein encompass inhibitors of the system Xc-, inhibitors of GPX-4, inducers of lipid peroxidation, and modulators of iron metabolism which lead to increased intracellular iron levels.

The cell death inducer, e.g., ferroptosis inducer, of the present disclosure, may for example act through inhibition of system Xc-, the transmembrane cystineglutamate antiporter, which imports cystine into cells. Cystine, the cysteine disulfide, is required for the biosynthesis of glutathione (GSH), which is a cofactor and cosubstrate for GPX4. Depletion of GSH leads to loss of GPX4 activity, resulting in accumulation of lethal lipid peroxides and ferroptotic cell death. Glutamate, erastin and its more potent analogues imidazole ketone erastin (IKE) and piperazine erastin (PE), as well as the clinically used drugs sulfasalazine and sorafenib, belong to this class of cell death inducers.

In some other instances, the cell death inducer, e.g., ferroptosis inducer, acts through direct inhibition of GPX4. (1 S, 3R)-RSL3 (henceforth RSL3) covalently interacts with GPX4 and inhibits its enzymatic activity, resulting in ferroptotic cell death.

In other instances, the cell death inducer, e.g., ferroptosis inducer, promote lipid peroxidation. For example, ferroptosis inducer 56 (FIN56) and caspaseindependent lethal 56 (CIL56) deplete GPX4 protein and mevalonate-derived coenzyme Q10, which is an endogenous lipophilic antioxidant that suppresses lipid peroxidation.

In some other instances, the cell death inducer, e.g., ferroptosis inducer, acts by modulating the iron metabolism, e.g., by promoting accumulation of intracellular iron, notably oxidised iron. For example, ferroptosis inducer endoperoxide (FIN02), acts by oxidising iron, driving lipid peroxidation, and indirectly inactivating GPX4 enzymatic function in cells.

Many compounds are described to have an activity as inducer or enhancer of specific cell death, e.g. , inducers of ferroptosis, and such compounds, known to the skilled person, shall be encompassed by the present disclosure.

Agonists or inducers of cell death, e.g., ferroptosis inducers, may be for example small molecules. Such inducers are well known to the skilled person and are disclosed in e.g., Hassannia et al. Cancer Cell. 35(6):830-849, 2019; Du & Guo. Cell Death Discov. 8(1 ):501 , 2022; Nie et al. Cancer Cell Int. 22(1 ):12, 2022; Zhang et al. Front Nutr. 9:844757, 2022; Cai et al. Front Oncol. 13:1119369, 2023; Yu et al. Nanoscale Adv. 5(5):1271 -1290, 2023.

Preferably, the cell death inducer is a small molecule selected in the group consisting of glutamate, RSL3, erastin, an erastin derivative (preferably MEH, PE, or AE), imidazole ketone erastin (IKE), sulfasalazine, sorafenib, sorafenib analogs (SRS13- 45 and SRS13-60), altretamine, artesunate, artemisinin, dihydroatemisinin, artemether, an artemisinin derivative (such as artesunate, artemisinin, and dihydroartemisinin (DHA)), ML-162 (DPI7), ML-210 (DPI10), RSL5, 4,4'- diisothiocyanostilbene-2,2'-disulfonic acid, temozolomide, MMRi62, t-tertiary-butyl hyperoxide, FIN56, CIL56, B-elemene, buthionine sulfoximine, NDP4928, statins, brequinar, withaferin A (WA), auranofm, almitrine, lanperisone, albinoside A, CDDO- Im, BAY 11 -7085, Withaferin A, NSC-207895, Bigelovin, Bufotalin, ruscogenin, Pifithrin- p, Bardoxolone methyl, Omeveloxolone, Erianin, Brussatol, KRA-533, Mulberrofuran G, RGB-286638, Xanthatin, Dodecyl gallate, Piperlongumine, Octyl gallate, phenylethyl B-isothiocyanate (PEITC), Tanshinone II A, Cryptotanshinone, Ungeremine, Erianin, VAS 3947, Nordihydroguaiaretic acid, GPX4-IN-3, HG106, DPI2, DPI12, DPI13, DPI17, DPI18, DPI19, 31MEW44, ML160, and ML162.

In other instances, the cell death inducer (e.g. , ferroptosis inducer) disclosed herein is a nanomaterial. Such nanomaterials capable of inducing specific cell death include notably iron-based nanomaterials and non-iron based nanomaterials. Preferably, the nanomaterials which can be used in the method of the disclosure include, e.g., ferumoxytol, Cornell dots (ultrasmall poly[ethylene glycol] -coated silica nanoparticles), C’dots, SPION, GA-Fe(ll), UPDA-PEG, WS2, MoS2, CPMNS, LDL-DHA nanoparticles, ChA CQDs, and Zinc oxide nanoparticles.

In addition, a cell death inducer (e.g. , a ferroptosis inducer) may be a compound selected from an antibody, or other antigen binding compound, a cytokine, a T-cell receptor, a nucleic acid. For example, the cell death inducer (e.g., a ferroptosis inducer) is an anti-TfR1 antibody, such as e.g., 3F3 anti-ferroptotic membrane antibody (3F3-FMA), anti-TfR1 3B8 2A1 antibody, and anti-TfR1 FI68.4 antibody. Alternatively, the cell death inducer (e.g. , a ferroptosis inducer) is interferon-y.

In some instances, the inducer of cell death is a nucleic acid expressing a cellular molecule, e.g., a protein, involved in regulating ferroptosis. For example, the nucleic acid may lead to an increased expression of a protein which activates ferroptosis. Such proteins include e.g. , p53, BAP1 , PKCBII, NCOA4. Vectors for controlling the expression of a transgene are well known in the art. Such vectors can be introduced into the cone photoreceptor by any means known to the skilled person.

In some instances, the agonist or inducer of specific cell death (e.g., ferroptosis) can be an inhibitory nucleic acid targeting expression of a cellular molecule involved in regulating ferroptosis. Nucleic acids, such as shRNA and anti- sense nucleic acids can target cellular components, including cysteine-glutamate antiporter (system X- c), glutathione peroxidase 4 (GPX4), NOX, and ALOX15. shRNAs and inhibitory RNAs for such cellular targets include such shRNAs as e.g., miR-672-3p and miR-129-5p, and are described in further details in US 2015/0079035, US 2015/0175558, US 2010/0081654, WO 2015/051149, WO 2015/084749, WO 2013/152039, and WO 2015/109009. Other cellular targets and corresponding inhibitory nucleic acids are described in Gao et al. Cell /?es.26:1021 -1032, 2016.

It is understood that the at least one cone photoreceptor used in the method disclosed herein can be provided in various ways. Notably, the cone photoreceptor used in the present method can be part of a retinal tissue.

In this instance, the disclosure comprises a prior step of providing a retinal tissue.

The retinal tissue may be complete, i.e., comprising all the retinal cell types, or it may be partial, i.e., comprising only some of the retinal cell types. In some other instances, the cone photoreceptor is an isolated cone photoreceptor (i.e., an in vitro model). Such isolated cone photoreceptors are available as populations of substantially purified cones, which are obtained from the retinal tissue. The retinal tissue may be an organoid or a retinal explant (i.e., an ex vivo model). Alternatively, the retinal tissue may be comprised within the whole eye, preferably the whole eye of a living subject (i.e., an in vivo model), more preferably the whole eye of a non-human living subject, such as e.g., a non-human living primate. Each of these models present its own advantages and uses, as will be detailed further below.

Advantageously, the retinal tissue can be an adult tissue. Preferably, the retinal tissue can be obtained from any vertebrate. Preferably, the retinal tissue is a mammalian retinal tissue (primate or non-primate retinal tissue), e.g., a human retinal tissue, or an animal but non-human retinal tissue such as a pig, rat, or mouse retinal tissue. In some instances, the retinal tissue is of human origin. In some other instances, the retinal tissue is a retinal tissue from a non-human primate.

In vitro models of cone photoreceptor degeneration

In a specific aspect, the model of cone degeneration obtained by the method of the disclosure is an in vitro model. In vitro models of cone degeneration are particularly useful, because they allow easy screening of therapeutic compounds, e.g. , compounds which maintain or restore, in particular specifically maintain or restore, cone photoreceptor viability after specific cell death (e.g., ferroptosis) has been induced in those cells.

Advantageously, the method of preparation of in vitro model for cone degeneration comprises the isolation of a substantially pure population of adult cone photoreceptor cells.

Accordingly, the method disclosed herein comprises a prior step of providing a retinal tissue. Advantageously, the retinal tissue comprises photoreceptors. In particular, the retinal tissue comprises cone photoreceptors. The retinal tissue may also comprise rod photoreceptors. The retinal tissue may also comprise progenitors of cone and/or rod photoreceptors. In addition, the retinal tissue may further comprise other cell types including e.g., bipolar cells, ganglion cells, horizontal cells, and amacrine cells, and/or adjacent supporting tissue, such as e.g., the retinal pigment epithelium.

The retinal tissue can notably be obtained from stem cells. Methods for generating cone photoreceptors and/or rod photoreceptors from stem cells have been described and can easily be used by the skilled person to obtain the retinal tissue (WO 2018/154295; WO 2020/178222; Khalili et al. Stem Cell Res. 33:215-227, 2018).

Preferably, the retinal tissue is used to obtain a substantially pure population of adult cone photoreceptor cells. Processes for isolating a substantially pure population of adult cone photoreceptor cells have been previously described; see e.g., WO 2005/103232A2; Balse et al. Invest Ophthalmol Vis Sci. 46(1 ):367-74, 2005.

In a first step, the retinal tissue is dissociated from the extracellular matrix. Dissociation of extracellular matrix of the retinal tissue or retinal tissue fragments can be achieved by any means available to the skilled person, as long as it allows for retinal cell dissociation without leading to cone cell lysis. Preferably, dissociation is performed enzymatically (e.g., with papain and/or trypsin) and/or mechanically and/or chemically. This step yields a population of isolated retinal cells. After dissociation, cone photoreceptors may be identified by their morphology, specific markers, or characteristic features, such as e.g., the specific binding of peanut germ agglutinin (PNA) to cone photoreceptors or by specific antibodies.

Advantageously, the method comprises a further step of isolating cone photoreceptors from the population of isolated retinal cells, thereby yielding a substantially pure population of cone photoreceptors. Preferably, the isolation step comprises incubating the population of isolated retinal cells with an antibody or a lectin, preferably PNA, and recovering the PNA-bound cone photoreceptor.

Advantageously, PNA is bound to a solid support, e.g., via anti-PNA antibody. Anti -PNA antibodies are commercially available. Any solid support appropriate to cell binding can be used, e.g., a cell culture plate, a glass coverslip, magnetic beads coupled with antibody anti-PNA.

The cone photoreceptors can be isolated from the dissociated retinal tissue with PNA by any means suitable to the binding of a lectin to a cell, such as e.g., by panning, MACS (magnetic activated cell sorting), or FACS (fluorescent activated cell sorting technique). Preferably, the cone photoreceptors are isolated from the dissociated retinal tissue by panning.

In another instance, the method of the disclosure comprises a further step of growing the substantially pure cone photoreceptors in a conditioned method. Muller glial cells synthesise and extracellularly release trophic factors which promote the survival and development of cone cells. Preferably, the present method hence advantageously comprises a further step of in vitro growing the substantially pure cone photoreceptors in a conditioned culture medium obtainable by collection of the culture medium after in vitro culture of Muller glial cells.

Ex vivo models of cone photoreceptor degeneration

In vitro models are suitable for isolating anti -degenerative compounds and studying their effects on isolated cells. However, they do not allow for the investigation the effects of these compounds on a whole tissue made of complex interactions between different cell types arranged in specific structures and organisational units. The information obtained with a substantially pure population of a specific cell type, e.g., a substantially pure population of cone photoreceptors, remain limited to that specific cell type, e.g., cone photoreceptors. Notably, it is difficult to assess in these conditions the effects of compounds preventing cone degeneration on other cell types, such as e.g., rod photoreceptor or RPE cells. On the other hand, the ex vivo model used in the examples has allowed the inventors to show that the induction of specific cell death in cone photoreceptors is followed by a microglial migration towards the outer nuclear layer (ONL). Microglia migration is well known to the skilled person and can be determined by any means known in the art (Rashid et al. Front Immunol. 10:1975, 2019).

In a specific aspect, the model of cone degeneration obtained by the method of the disclosure is an ex vivo model. According to this aspect, the retinal tissue is, for example, an organoid or a retinal explant maintained in culture.

In a first instance, the retinal tissue is an organoid. As the skilled person will know, organoids are three-dimensional in vitro culturing models that originate from self-organising stem cells. Organoids can mimic the in vivo structural and functional specificities of body organs. Concerning the retina, stem cells, in particular induced pluripotent stem cells (iPSCs), can differentiate into RPE cells and cells of the neural retina, including rods and cones. The process of harvesting somatic cells and reprogramming them into virus-free human iPSCs (hiPSC) and from there to retinal organoids has been standardised in various laboratories (WO 2020/178222; Sharma et al. Int J Mol Sci. 21 (22):8484, 2020; Cobb et al., Transl Vis Sei Technol. 10(10):9, 2021 ).

Alternatively, the retinal tissue is a retinal explant. The retinal explant of the disclosure comprises live cone photoreceptor cells. Preferably, the retinal explant comprises, in addition to live cone photoreceptors, other live cells such as e.g., rods, retinal pigment epithelial cells, ganglion cells, bipolar cells, amacrine cells, horizontal cells, astrocytes, and/or Muller cells.

Preferably, the retinal explant of the disclosure comprises a layer of rods and cones.

In particular instances, the retinal explant comprises, in addition to the layer of rods and cones, a nerve fibre layer, a ganglion cell layer, an inner plexiform layer, an inner nuclear layer, an outer plexiform layer, an outer nuclear layer, and/or an external limiting membrane. In some instances, the retinal explant comprises all of: the nerve fibre layer, the ganglion cell layer, the inner plexiform layer, the inner nuclear layer, the outer plexiform layer, the outer nuclear layer, the external limiting membrane, and the layer of rods and cones. In certain instances, the non-human primate retinal explant comprises, in addition to the layer of rods and cones, an inner a nerve fibre layer, a ganglion cell layer, an inner plexiform layer, an inner nuclear layer, an outer plexiform layer, an outer nuclear layer, an external limiting membrane, and/or the retinal pigment epithelium. In some instances, the retinal explant comprises all of: the nerve fibre layer, the ganglion cell layer, the inner plexiform layer, the inner nuclear layer, the outer plexiform layer, the outer nuclear layer, the external limiting membrane, the layer of rods and cones, and/or the retinal pigment epithelium.

Any source of retina can be used for the present disclosure. In some instances, the retina is a human retina. In other instances, the retina is of non-human origin, e.g., non-human primate, bovine, pig or rodent origin. Human retina can be easily obtained from cornea banks where said retina are normally discarded after the dissection of the cornea. Adult human retina has a large surface (about 1 100 mm²) and can therefore be easily separated to a number of experimentally subregions.

The organoid and/or the retinal explant are maintained in culture. Several culture media suitable for growing organoids and/or retina explants have been previously described (see e.g., Fradot et al. Hum Gene Ther. 22(5):587-93, 2011 ; Belhadj et al. J Vis Exp. 2020 Nov 25; (165); Marie et al. Cell Death Dis. 2020 Aug 29; 11 (8) :711 , 2020). The skilled person will refer to the examples for a specific instance of a culture medium which can be used for maintaining the organoid or the retinal explant alive.

Advantageously, the present method comprises the further step of contacting the retinal tissue, i.e., the organoid or the retinal explant, with a polycarbonate membrane. The retinal tissue may be contacted with the membrane in any orientation, i.e., with the photoreceptors facing towards the membrane or with the photoreceptors facing away from the membrane. In some instances, it is advantageous that the photoreceptors are facing away from the membrane.

In vivo models of cone photoreceptor degeneration In another aspect, the retinal tissue is comprised in a whole eye. Preferably, the eye is the eye of a living animal, more preferably a non-human animal.

Such an in vivo model is particularly useful for studying the physiological relevance of drugs aiming at preventing of cone photoreceptor degeneration. This is especially true for cases where the intact eye has to be considered once translational therapy-oriented studies are being conducted. This allows notably studying not only the effects of the therapy on the cone photoreceptors themselves, but also the effects on the other cell types of the retina, and further the even more complex interactions with other organs. Such studies are required before proceeding to the clinic. The present in vivo model of cone photoreceptor degeneration is thus particularly useful for the characterisation of potential drugs against retinal degeneration pathologies.

Indeed, the inventors have shown that the present in vivo model faithfully reproduces the features of retinal degeneration pathologies such as e.g., AMD. In particular, when an inducer of specific cell death, e.g., a ferroptosis inducer, is injected into the eye of a non-human primate, cone loss was observed, notably a massive loss of cone outer segments. Further, the loss of cone receptors in this model was accompanied by an increased number of microglial cells in outer retinal layers. Retinal thickness was diminished, as shown by histological sections or by OCT (Optical Coherence Tomography). Subretinal deposits were detected, in the fovea and/or in the macula. Moreover, these histological observations were correlated with physiological results showing that the cone activity in the retina was decreased, as assayed by electroreti nogram (ERG) or multifocal electroreti nogram (mfERG); see e;g. Asanad S, Karanjia R. Multifocal Electroreti nogram. [Updated 2022 Oct 9]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2023 Jan.

Accordingly, activation of specific cell death, e.g., ferroptosis, in the method of preparing an in vivo model of cone photoreceptor degeneration as disclosed herein, results in one or more of:

• proliferation, activation and/or migration of microglial cells to the ONL;

• loss of cone outer segments;

• formation of subretinal deposits, preferably in the fovea and/or in the macula; • decreased cone activity in the retina, preferably as indicated by ERG or mfERG; and

• thinning of the retinal thickness as indicated on histological section or in vivo imaging as with OCT.

The method disclosed herein comprises contacting an inducer of specific cell death, e.g., a ferroptosis inducer, with at least one cone photoreceptor. In the case of the in vivo model described herein, such contacting can be performed by administering the cell death inducer (e.g., a ferroptosis inducer) to the animal. Administration is local to the ocular or adnexal tissues. Preferably, the composition is administered intravitreally, subretinally, or topically. Topical ophthalmic formulations include eye drops. Preferably, the methods do not include systemic administration. Local ocular administration has several advantages as the eye is an immune-privileged environment and compounds administered to the eye function locally and have little or no systemic dissemination.

Optionally, the method further comprises the administration of a pharmaceutically acceptable carrier.

The phrase “pharmaceutically acceptable” refers to compositions, polymers and other materials and/or dosage forms which are, within the scope of sound medical judgment, suitable for use in contact with the tissues of human beings and animals without excessive toxicity, irritation, allergic response, or other problem or complication, commensurate with a reasonable benefit/risk ratio.

A “pharmaceutically acceptable carrier” refers to, for example, pharmaceutically acceptable materials, compositions or vehicles, such as a liquid or solid filler, diluent, excipient, solvent or encapsulating material, involved in carrying or transporting any supplement or composition, or component thereof, from one organ, or portion of the body, to another organ, or portion of the body. Each carrier must be “acceptable” in the sense of being compatible with the other ingredients of the supplement and not injurious to the patient. Optionally, a pharmaceutically acceptable carrier is non- pyrogenic. Some examples of materials which may serve as pharmaceutically acceptable carriers include: (1 ) sugars, such as lactose, glucose and sucrose; (2) starches, such as corn starch and potato starch; (3) cellulose, and its derivatives, such as sodium carboxymethyl cellulose, ethyl cellulose and cellulose acetate; (4) powdered tragacanth; (5) malt; (6) gelatine; (7) talc; (8) excipients, such as cocoa butter and suppository waxes; (9) oils, such as peanut oil, cottonseed oil, sunflower oil, sesame oil, olive oil, corn oil and soybean oil; (10) glycols, such as propylene glycol; (11 ) polyols, such as glycerine, sorbitol, mannitol and polyethylene glycol; (12) esters, such as ethyl oleate and ethyl laurate; (13) agar; (14) buffering agents, such as magnesium hydroxide and aluminium hydroxide; (15) alginic acid; (16) pyro gen -free water; (17) isotonic saline; (18) Ringer's solution; (19) ethyl alcohol; (20) phosphate buffer solutions; and (21 ) other non-toxic compatible substances employed in pharmaceutical formulations.

Models of cone photoreceptor degeneration and uses thereof

In another aspect, the present disclosure relates to a model of cone photoreceptor degeneration which is obtainable by the methods disclosed herein. The model can be any of the models disclosed herein, i.e., an in vitro model, an ex vivo model, or an in vivo model.

The model disclosed herein is particularly useful for screening and/or testing compounds with potential therapeutic activity. Indeed, a preliminary screening of a compound library with the in vitro model retrieved several compounds capable of restoring cell viability, demonstrating the usefulness of the model of cone degeneration described herein.

In another aspect, the present disclosure thus relates to a method of screening for compounds capable of showing protective and/or anti -degenerative properties on cone photoreceptors. Preferably, the method comprises: a) providing the model of cone degeneration described herein, b) contacting the model of cone degeneration with a candidate compound, c) assessing whether the candidate compound induces a protective and/or anti -degenerative effect on cone photoreceptors.

The assessment of the protective and/or anti -degenerative effect can be done by any means known to the skilled person. Preferably, a protective and/or anti- degenerative effect on cone photoreceptors is induced by the candidate compound, when contacting the model of cone degeneration with a candidate compound results in at least one of: • maintaining or increasing, in particular specifically maintaining or increasing, viability of the at least one cone photoreceptor,

• maintaining or increasing intracellular glutathione level, in particular specifically maintaining or increasing intracellular glutathione level, in the at least one cone photoreceptor;

• maintaining or decreasing the intracellular NADP/NADPH ratio, in particular specifically maintaining or decreasing the intracellular NADP/NADPH ratio, in the at least one cone photoreceptor;

• maintaining or decreasing the intracellular formation of lipid peroxides, in particular specifically maintaining or decreasing the intracellular formation of lipid peroxides, in the at least one cone photoreceptor, notably 5-oxo- 6,8,11 ,14-eicosatetraenoic acid (5oxoETE);

• preventing an increase in the cell membrane permeability to dyes, in particular specifically of cone photoreceptors.

The effect can also translate into any combination of the above properties. Notably, an effect against the degeneration of cone photoreceptors can result in at least 1 , at least 2, at least 3, at least 4, at least 5, or at least 6 of the above properties. Preferably, an effect against the degeneration of cone photoreceptors results in all 7 of the properties.

As described herein, the various cone degeneration models (in vitro, ex vivo, in vivo) described herein offer various levels of analysis.

For example, the screening method described herein can be used to identify one or more compounds having protective and/or anti -degenerative properties using the in vitro model, i.e., a model of cone degeneration wherein specific cell death is induced in a substantially pure population of cone cells. The compounds thus identified can then be further screened according to the method described herein in the ex vivo model in order to identify those compounds which still display an effect against the degeneration of the cone photoreceptors in the context of a whole organ. This effect translates into an increased migration of the microglial cells in the ONL or in any one of:

• maintaining or increasing viability, in particular specifically maintaining or increasing viability, of the at least one cone photoreceptor,

• Preventing an increase in cell membrane permeability to dyes.

The effect can also translate into any combination of the above properties. Notably, an effect against the degeneration of cone photoreceptors can result in at least 1 , at least 2, at least 3, at least 4, at least 5, at least 6, or at least 7 of the above properties. Preferably, an effect against the degeneration of cone photoreceptors results in all 8 of the properties.

Finally, the in vivo model can be used in the screening method of the disclosure to identify the compounds which present a therapeutic effect. Preferably, a protective and/or anti -degenerative effect on cone photoreceptors is induced in vivo by the candidate compound, when contacting the model of cone degeneration with a candidate compound results in at least one of:

• Preventing an increase in cell membrane permeability to dyes, in particular specifically of cone photoreceptors;

• preventing the proliferation, activation or migration of microglial cells;

• preventing the loss of cone outer segments;

• preventing the decrease in cone activity in the retina, preferably as indicated by ERG or mfERG; and

The effect can also translate into any combination of the above properties. Notably, an effect against the degeneration of cone photoreceptors can result in at least 1 , at least 2, at least 3, at least 4, at least 5, at least 6, at least 7, at least 8, at least 9, or at least 10 of the above properties. Preferably, an effect against the degeneration of cone photoreceptors results in all 11 properties.

The invention in all its aspects is illustrated further in the following examples. The examples do not, however, limit the scope of the invention, which is defined by the appended claims.

EXAMPLES

EXAMPLE 1

METHODS

Data availability

Animal model: rats

Eight-weeks-old Long Evans wild-type rats were obtained from Janvier Labs (ISO 9001 certification). Controls and treated were administered a subretinal injection at day 0 (DO) of either 0.2% DMSO (D2650, Millipore Sigma, Burlington, MA) in PBS (14190- 094 Life Technologies Europe B.V., Bleiswijk, The Netherlands) or 0.2% of a 20pM IKE in DMSO solution diluted in PBS, respectively.

Surgical and dissection procedures in rats

Rats were anesthetized by an intraperitoneal (IP) injection of 40mg/kg ketamine (Ketamidor 100mg/ml, Axience SAS, Pantin, France) and 0.14mg/kg medetomidine (Domitor 0.85mg/ml, Vetoquinol S.A., Paris, France). Local anesthesia was achieved with oxybuprocaine eyedrops (1.6mg/0,4ml, Thea, Clermont-Ferrand, France). Mydriasis in the left eye was obtained with 0.5% tropicamide eyedrops (Mydriaticum, Thea, Clermont-Ferrand, France). Tear gel (Lubrithal, Dechron, Shrewsbury, UK) was used as a lens-eye interface during the whole procedure. Using an ophthalmic microscope (Lumera 700, Carl Zeiss, Oberkochen, Germany), a conjunctivoscleral channel was created with a 30 Gauge (G) needle (BD Microlance 3, Becton, Dickinson S.A., Fraga, Spain). A flat contact lens (Cover slips, mini, 8mm, World Precision Instruments, Sarasota, FL) allowed for visualisation of the retinal plane. Nasal subretinal delivery of 9pL of control or IKE solution was performed using a 30G cannula with a nonbeveled metal tip on a 10pL syringe (Hamilton, Reno, NV) mounted on a microinjector (Micro 4, World Precision Instruments, Sarasota, FL). After surgery, chloramphenicol-retinol eye ointment (Ophtalon 10mg/g, TVM, Lempdes, France), IP injection of 1 mL of 5% glucose monohydrate (Osalia, Paris, France) and subcutaneous injection of 0.9mg/kg atipamezole (Antidorm 4.27mg/ml, Axience SAS, Pantin, France) were administered. In vivo anatomical investigations, i.e., optical coherence tomography (OCT, Bioptigen, Durham, NC) and funuds photography (Micron IV, Phoenix-Micron, Inc., Bend, OR), were performed at DO to ensure the subretinal delivery and the day of the sacrifice to confirm retinal reattachment. Sacrifice, using intracardiac injection of 1 ml/kg pentobarbital (Exagon, Axience SAS, Patin, France), was performed at day 6 (D6) for one group of animals (n=4) to assess microglial migration, and at D30 (D30) for one group (n=8) to assess cone degeneration. Left eyes were collected and immediately immersed in 4% paraformaldehyde (PFA, J61899.AP ThermoFisher Scientific, Waltham, MA) for 2 hours, then stored in PBS at +4°C. Globes were dissected to obtain whole retinae, immersed in a 24-well plate (#351147 Corning Inc., Corning, NY) for immunohistochemistry. In addition, labelled eyecup slices (with CAR and Iba1 ) were performed to assess another quantitative analysis. To do so, eyes were collected from rats and the cornea was removed. Then, eyes were immerged in 4% PFA for 1 h before successive sucrose baths from 10% to 30%. Lastly, lens was gently removed and the eye cup placed into a freezing medium in liquid nitrogen. They were stocked at -20° C until to be cut into slices of 12pm with a cryostat (CM3050 S, Leica, Wetzlar, Germany).

Mammalian eyes

Porcine eyes were obtained from a local slaughterhouse in agreement with the local regulatory department and the veterinarians from the French Ministry of Agriculture (agreement FR75105131 ).

Pure cone photoreceptors

Pig eyes were dissected to obtain retinae that were cut into pieces and then digested by papain at 4U/mL (LSO 3124, Worthington,) and L-Cysteine (5.5 mM, Sigma- Aldrich) for 20 minutes at 37° C. The enzyme activity was stopped by Neurobasal-A (NBA) medium (10888022, ThermoFisher Scientific, Waltham, MA, USA) which contained 5% Fetal Bovine Serum (FBS) and DNase I at 15pg/ml (D4263, Millipore Sigma, Burlington, MA). A series of short centrifugations (30sec, 5 to 6 time) was performed in order to collect supernatant excluding the first one which contains mainly rod photoreceptors. Pellets were resuspended with a P1000 to dissociate the remaining cells between each centrifugation. Afterwards, the cell suspension was centrifuged for 10 minutes at 800 rpm and pellets were resuspended in NBA medium containing 1% of L-glutamine (G3126, Millipore Sigma, Burlington, MA) and 1% B27 supplement (17504044, ThermoFisher Scientific, Waltham, MA, USA).

Cone PR were purified by lectin-panning selection as previously described (Balse et al. (2005) IOVS 46:3672), with slight modifications to obtain millions of cones in a suspension. The cells were then seeded in 384 well plates (781091 , Greiner, Ulis, France) to get 3000cells/well using a robot (Viaflo Assist, Integra-Biosciences, Saint- Ouen, France) and incubated at 37° C and 5% of CO? for 3 days.

Retinal explants

Porcine and non-human primate retinae were cultivated as previously described (Greferath et al. (2016) Ophthalmology 123:1320). After isolation of retinae in CO?- independent medium, pieces from the region behind the optic nerve were obtained using biopsy punches of 2mm, were placed on polycarbonate membrane (140652, Thermo, Waltham, Massachusetts) and photoreceptors were turned upwards. The explants were kept in culture in a CO? incubator at 37° C for 3 days.

Treatments

To induce ferroptosis, the following components were used: Glutamate (49621 , Sigma Aldrich, Missouri, USA) between 1 pM to 500pM in vitro, Imidazole Ketone Erastine (IKE) (HY-114481 , MedChemExpress, Monmouth, USA) either at 3pM on purified cones photoreceptors or at 20pM for ex vivo and in vivo experiments, and RSL3 (HY- 114481 , MedChemExpress, Monmouth, USA) either at 20nM, 200nM and 10pM respectively for cells, ex vivo and in vivo experiments. To inhibit ferroptosis, Ferrostatin-1 (FST-1 , HY-100579, MedChemExpress, Monmouth, USA) was used at 50nM on purified cones or at 400nM on retinal explants. Also, Deferiprone (HY-B0568, MedChemExpress, Monmouth, USA) was efficient at 50pM on pure cones and Zileuton (HY-14164, MedChemExpress, Monmouth, USA) at 20pM on purified cones.

Cell viability measurement After 3 days of incubation, cell viability was measured with Calcein (C1430, Thermo, Waltham, Massachusetts) at 1 /4000 which is incubated for 1 hour at 37° C. The number of viable cells was calculated with an automated fluorescence microscope, ArrayScan (Cellomics ArrayScan VTI HCS reader, Thermo, Waltham, MA).

Screening on primary cells

Purified cone photoreceptors were seeded into 384-well plates with clear flat bottom (781091 , Greiner, Ulis, France) by a robot having a 384-well head (BRAVO, Assist, Integra-Biosciences, Saint-Ouen, France) in order to have 4*10³ cells/well. In the first two columns, cells were treated with only DMSO (D2650, Millipore Sigma, Burlington, MA) and in the rest of the plate cells were treated with, either RSL3 (HY- 114481 , MedChemExpress, Monmouth, USA) at 20nM final or IKE (HY-114481 , MedChemExpress, Monmouth, USA) at 3pM final. 17 hours after the induction of cell degeneration, a ferroptosis library compounds (HY-L051 , MedChemExpress, Monmouth, USA) was added to cells by a robot having a 384-well head (BRAVO, Assist, Integra- Biosciences, Saint-Ouen, France) in order to obtain a final concentration at 10pM of compounds. Each treatment was diluted into neurobasal medium (10888022, ThermoFisher Scientific, Waltham, MA, USA) and 1 /100 of L-glutamine (G3126, Millipore Sigma, Burlington, MA). The cells were then incubated for another 2 days at 37°C and 5% CO2. Then on the 3rd day after seeding, cell viability was measured.

Results were analysed with TIBCO Spotfire® software (California, USA). The robust Z’ factor was calculated for each plate, and those above 0.45 were selected. Then, for each plate, the percentage of live cells was calculated from the positive controls. The average number of cells counted is calculated and reported at 100%. This ratio of positive controls was reported to all the molecules in the plate. Next, all the plates from the same treatment (IKE or RSL3) were displayed in order to select the compounds having a cell survival higher than 50% and we extract these compounds on Excel.

Li pi domic analysis

Purified cone photoreceptors either treated or not with 3pM IKE were washed 3 times with PBS, scrapped in 100pL and snap-frozen with liquid nitrogen immediately after collection and stored at -80° C until extraction. Extraction and analysis of eicosanoids was performed by the lipidomic facility core: MetaToul-Lipidomique (I2MC, Inserm, Toulouse, France), MetaboHUB-ANR-1 1— INBS-OO1O.

Western blot

Cells were collected, separated from medium by centrifugation and whole cellular proteins were extracted in lysis buffer (10mM HEPES [pH 7], 100mM NaCl, 2mM EDTA, 0.5% NP-40 and protease inhibitor cocktails (Millipore Sigma, Burlington, MA). After a 13000rpm centrifugation, the supernatant was kept for further WB analysis. Samples were then loaded in 4% to 15% Mini-PROTEAN TGX Tris-glycine buffer SDS- PAGE and transferred onto a 0.2pm Trans-Blot Turbo nitrocellulose membrane (BioRad, California, USA)). Membranes were blocked for 1 hour at room temperature in 1 x Tris-Buffer Saline (TBS, 10mM Tris-HCl [pH 8], 150 mM NaCl) supplemented with 5% (w/v) dried skim milk powder. Primary antibody incubation was carried out overnight at 4°C. The presence of 5LOX and FLAP in total lysates at the expected molecular weight was confirmed by Western blotting with anti-5LOX (1 :200; Abeam AB169755) and anti-FLAP (1 :500; Abeam AB85227). The secondary HRP -coupled antibody used was goat anti-rabbit (1 :20,000; 111 -035-003, Jackson ImmunoResearch Laboratories, Pennsylvania) for detection of both 5LOX and FLAP. In between and after antibodies incubations, membranes were extensively washed in TBS containing 2.5% Tween-20 (TBS-T). Western blots were visualised using the enhanced chemiluminescence method (ECL Prime detection reagent, Amersham, UK).

NADPH assay

The NADP/NADPH assay kit (ab176724, Abeam, Cambridge, UK) was used to quantify the level of NADP and NADPH respectively according to the manufacturer’s instructions.

Glutathione (GSH) assay

We quantified the level of GSH according to the manufacturer’s instruction for GSH assay kit (V6912, Promega, Madison, Wl).

Cryostat sections

After 3 days of incubation, retinal explants were fixed with 4% paraformaldehyde (15714, Electron Microscopy Sciences, Hatfield, Pennsylvania) for 1 h at room temperature and washed 3 times with PBS. Then, retinal explants were conserved at +4°C or transversally sectioned. The last were cryopreserved in successive sucrose baths of 10%, 20% and 30%. Then, samples were frozen in a tissue freezing medium (72592, Electron Microscopy Sciences, Hatfield, Pennsylvania) by immersion in liquid nitrogen until the tissue freezing medium appeared stiff. Samples were conserved at -20° C until cutting in transversal slices of 10pm with a cryostat (CM3050 S, Leica, Wetzlar, Germany). Retinal cross sections were conserved at -20° C.

Immunolabelling

Samples (fixed retinal explants and retinal cross-sections) were firstly permeabilised with 0.5% Triton 100X (T8787, Sigma-Aldrich, Missouri, USA) diluted in PBS 1 x. Then, unspecific areas were blocked with a saturated buffer composed of 10% of Normal Donkey Serum (S30-100ML, Millipore Sigma, Burlington, MA) diluted in PBS 1 x. The primary and secondary antibodies used in this study are listed in table 1 .

Table 1. List of antibodies.

The PNA lectin diluted at 1 /50 (L21409, Thermo, Waltham, Massachusetts) and rabbit FLAP antibody with a working concentration at 5pg/mL (ab85227, Abeam, Cambridge, UK) are targeting antigenic sites that are at the extracellular side of the cell membrane. For those labelling, no permeabilisation step was done.

Finally, retinal explants were washed 3 times in PBS and flat-mounted with Permafluor (TA-030-FM, Thermo, Waltham, Massachusetts) in Cellvis plates (P12-1.5H- N, IBL, Gerasdorf, Austria) for fluorescence microscopy. Retinal transversal slices and whole-mounted retinae were also mounted with Permafluor on coverslips or in a 6-well plate (P06-1.5H-N, Cellvis, Mountain View, CA), respectively.

Confocal microscopy

Retinae were imaged using a laser-scanning confocal microscope (Fluoview V- 1000, Olympus, Tokyo, Japan). In rats’ retinae, cone counting (outer segments and perikaryon per mm) was performed manually on retinal slices.

CQ1 imaging

The CQ1 confocal microscope (Yokogawa Electric Corporation, Tokyo, Japan) was used to image whole-mounted retinae of rats and flat mounted porcine explants. For injected rats’ retinae, fields of the inferonasal and infero temporal petals were manually reviewed in 4 z-axis planes defined by the cone labelling: the outer segments (OS) plane, the cone nucleus (CN) plane, the inner fibre (IF) plane and the cone pedicle (CP) plane. Emphasis was given to microglial migration. Representative samples were extracted from the CQ1 software. Z stack images were analysed using Cell pathfinder software (Yokogawa, Tokyo, Japan) which enables 2D, 3D or slice quantification. Algorithms were generated (one for the porcine explants and one for the wholemounted retinae of rats) to quantify cones perikarya or outer segments and microglial cells according to the z axis. Extracted quantification data were analysed using TIBCO Spotfire® (California, USA).

Non-human primates (NHP)

Two NHP were used in this study: a 5-year-old male and a 15-year-old female. Animals were born in captivity and originated from approved suppliers with AAALAC certification (SARL Bioprim, Bazieges, France; Cynologics-Silabe, Niederhausbergen, France). Ethical approvals were obtained from the local ethical committee CETEA n°44 of MIRCen and then the French Ministry of Education and Research. A full description of housing and perioperative care is available in Dentel et al. (2023) Ophthalmol. Sci. 3:100316.

IKE toxicity in NHP was assessed by a subretinal delivery of a 20pM IKE solution in one eye and a control solution (0.2% DMSO diluted in PBS) in the fellow eye. A comprehensive description of anesthesia, surgery and data acquisition is available in Dentel et al. Briefly, the subretinal injection was performed in order to detach the superior hemifovea in each eye. Anatomical (slit-lamp exam, optical coherence tomography -OCT-, adaptive optics -AO-) and functional (full-field and multifocal electroretinogram, ffERG and mfERG, respectively) investigations were performed at baseline (within a week prior to surgery), at day 3 (to confirm that retinal reattachment was completed), at one month, and then monthly until month 4. The embedded software of the rtx1 (AOdetect, Imagine Eyes, Orsay France) was used to quantify adaptive optics cone mosaic parameters. Also, the additionally used power spectrum spacing method to quantify cone mosaic in NHP is detailed in Dentel et al.

Analyses and statistics

Analyses, graphs and charts were achieved using GraphPad Prism 8.4.0 Software (GraphPad Software Inc., San Diego, CA). Treatments and controls were statistically compared with the Mann-Whitney t-test and were considered to be significant when p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) or p < 0.0001 (****). In charts, data were normalized to controls when possible and represented as mean +/- standard error of the mean.

The Mann-Whitney t-test was used to compare controls with treatments in the in vitro and ex vivo analyses. In rats, differences in cone density and microglial density were compared using a multiple t-test and statistical significance determined using the Holm-Sidak method. In NHPs, differences in cone density, regularity, spacing values and ERG amplitude waves were compared using a two-tailed t test. For mfERG in NHP, each hexagon was compared with the contralateral matching hexagon, considering the symmetry between the two eyes.

RESULTS

Glutamate-elicited ferroptosis in purified cone photoreceptors

Photoreceptors release their neurotransmitter, glutamate, at their synaptic terminals but they only express glutamate transporters (not classic ionotropic receptors). When applying glutamate on purified porcine cone photoreceptors (Fig. 1 A- D), their viability decreased by 20% already with 10pM; and up to 82% at the highest tested concentration of 500pM (Fig. 1E). This range of glutamate is encountered physiologically as indicated by the glutamate affinity of receptors and transporters (Gielen (2010) medecine/sciences 26:65). To define if this toxicity can be related to the expression of the system Xc anti porter, which has been reported in cone photoreceptor by immunocytochemistry (Dun et al. (2006) Cell Tissue Res 324:189; Bridges et al. (2004) Ophthalmol. Vis. Sci. 45:2906-2914), we applied glutamate in the presence of 1mM L-cystine. At every glutamate concentration, L-cystine prevented the toxicity of glutamate (n=7, p<0.0001 , IC50 = 4.15pM) (Fig. 1E). Because the system Xc transporter has a very distinct pharmacology from other glutamate transporters or receptors, we examined how the different agonist and antagonist of this transporter affected cone survival. Erastin and imidazole ketone erastin (IKE), two functional inhibitors of the System Xc, produced toxicity on isolated cone photoreceptors, similarly than with glutamate but in a lower dose range in accordance with their greater affinity (fig. 1A-D, F). L-Cystine also rescued cones exposed to IKE (n=6, p=0.0022) (Fig. 1K).

To assess the molecular mechanisms that underlie this glutamate-elicited toxicity in cone photoreceptors, we measured GSH levels in cones exposed to 500pM glutamate and imidazole ketone erastin (IKE 3pM). A 50% decrease in GSH at day 1 and later was observed under both conditions when compared to a control condition (Fig. 1 F, n=18, p<0.0001 ). Cone photoreceptor cell death is therefore due to a dysregulation of the cellular levels of GSH, induced by a System Xc dysfunction. Knowing that GSH is involved in a redox cycle where its oxidised form is reduced with a hydrogen atom carried by NADPH (Dixon et al. (2019) Annu. Rev. Cancer Biol. 3:35), we measured the NADPH/NADP ratio in cones treated with 3pM IKE. This condition increased the amount of NADPH from 25% in controls to 85% in treated cultures (n=4, p=0.0159) (Fig. 1G). The intracellular NADPH/NADP ratio is indirectly related to the activity of glutathione peroxidase 4 (GPX4) during the oxidation reduction of glutathione. When GPX4 becomes inactive, this cycle also becomes inactive and can induce this accumulation of NADPH suggesting that the redox cycle is downregulated when the System Xc is inhibited (Azuma et al. (2022) J. Biol. Chem. 298:101824). To investigate if GSH is crucial for cones survival, a specific inhibitor of GPX4, Ras-selective lethal 3 (RSL3), was used. This inhibitor showed a toxic effect on cone photoreceptor at very low concentrations (n=3, IC50 =8.67nM) (Fig. 1H), highlighting the key role of GPX4 in cone survival. RNA extraction from treated and control pure cone photoreceptors was then performed, and we showed that expressions of proteins involved in regulating the glutathione redox cycle and downstream mechanisms were altered (Fig. 2).

GPX4 is also known to prevent peroxided lipids formation that react with free radical species (Xie et al. (2016) Cell Death Differ. 23:369; Dixon et al. 2019). Those peroxided lipids are formed by lipoxygenases such as 5 -lipoxygenase (5LOX), which is activated by 5-lipoxygenase-activating protein (FLAP). To examine presence of these mechanisms in cone photoreceptors, western blots were performed, and we discovered the presence of both 5LOX and FLAP in cones but not in rod photoreceptors (Fig. 11). Specific immunostaining on retinal tissues from different species revealed the specific FLAP localisation in cone photoreceptors (Fig. 3). To assess the role of this peroxided lipid pathway in glutamate-elicited cell death, cone cell extracts were analysed by lipidomic after 2 days incubation with 3pM IKE. Analyses revealed enhanced amounts of peroxided lipids such as 5-Oxo-6,8,11 ,14-eicosatetraenoic acid (5-oxo-ETE) (n=6, p= 0.0079 (“), p= 0.0159 (*)) (Fig. 1 J), a metabolic product of 5LOX. To demonstrate the contribution of this lipid peroxidation in the observed glutamate-elicited toxicity, we tested the effects of Zileuton (ZEN), an enzymatic antagonist of 5LOX, in the presence of IKE. ZEN rescues 80% of cone photoreceptors (Fig. 1K, n=3, p=0.0238), suggesting that lipid peroxidation plays a key role in glutamate-elicited cone degeneration.

Lipid peroxidation induces iron metabolism changes during ferroptotic processes. Ferri tinophagy processes can be highlighted by immunolabelling ferritin and nuclear receptor coactivator 4 (NCOA4), a ferritinophagy inducer. Cones treated with IKE showed an increase in immunolabelling intensity of both ferritin heavy chain and NCOA4. Also, NCOA4 seemed to be slightly more intense than ferritin heavy chain, suggesting an increase of the activity of iron metabolism in treated cones (Fig. 4). Ferritinophagy processes can be inhibited by Ferrostatin-1 (FST1 , a specific ferroptosis inhibitor) and deferiprone (DF, an iron chelator preventing Fenton’s reaction by limiting intracellular accumulation of free iron). Both rescued cones from IKE-induced toxicity, by 68.5% for FST1 (n=8, p=0.0047) and by 77.2% for DF (n=5, p=0.0079) (Fig. 1K). In addition, other antiferroptotic molecules were tested after treatment with RSL3 at 20nM and IKE at 3pM, targeting different components involved in ferroptosis activation (Table 2). Thus, it seems that pure cones photoreceptors exposed to proferroptotic agents (e.g., IKE) can be rescued by antiferroptotic treatments, enhancing the role of ferroptosis in cone survival.

TABLE 2. Anti ferroptotic molecules after treatment by either 20nM RSL3 or 3 M IKE

A preliminary screening was performed on cone photoreceptors treated either by 20nM RSL3 or 3pM IKE. A day after inducing ferroptosis in cells, library compounds were added at 10pM. Best hits were recorded in the table below, according to each treatment. Some compounds did not increase cell survival over 50% depending on treatment and were not reported in the table (/). Cells treated with only 20nM RSL3 or 3pM IKE showed only 30% and 25% of cell survival respectively, 3 days after treatments (n=3) .

Molecules involved in system Xc- pathway

Ferroptosis of cone cells in an ex vivo retinal model

To show that cone photoreceptors can degenerate by ferroptosis in an integrated retinal model without retinal pigment epithelium, we applied IKE on freshly prepared porcine retinal explants. Specificity and selectivity of ferroptosis in retina was assessed by examining damaged cells (cell membrane permeability to ethidium dye) after incubation for 3 days. Whereas ethidium labelled few scattered cells in the control condition, the first row of photoreceptor nuclei was labelled by ethidium in the IKE-treated explants (Fig.5B,F). This outstanding layer corresponds to cone photoreceptors’ nuclei in the porcine retina (Pattnaik et al. (2000) J. Neurosci. 29:6789), as indicated by the Flap immunolabeling (Fig.5 C,G). Figure 5G illustrates the morphological alterations of cone photoreceptors with 20pM IKE. Other types of retinal cells were also labelled, showing no major qualitative changes (Fig. 6). In situ cellular toxicity was then assessed in cone photoreceptors by counting (automatically with the CellPath software of the CQ1©) at the level of outer segments (OS) immunolabelled by Flap (Fig. 5 J-M). 20pM IKE induced a 50% decrease of the amount of cone OS (n=3, p=0.0286) (Fig. 5 I). 1 pM FST-1 and 100pM DF rescued 84% and 78.4% of OS, respectively (n=3, p=0.1000 and p=0.0286, respectively) (Fig. 5 I). Similarly, when cone degeneration was triggered by 200nM RSL3 in retinal explants (Fig.7), 100pM DF rescued 78.5% of OS (n=3, p=0.0286) (Fig. 7M). These data are suggestive for ferroptosis being an effective and selective pathway of cone photoreceptor degeneration in an integrated retinal tissue.

Additionally, microglial migration from the inner retina (normal localisation) toward the outer retina (normally free of microglia) seemed to be triggered in retinal tissues exposed to ferroptosis inducers. Activation of microglial cells is suggested by their increased volume and modified shape, with cellular extensions embracing cones (Fig. 5, H). Automated quantitative analyses revealed significant intraretinal microglial displacement toward the outer nuclear layer (ONL) triggered by both 20pM IKE and 200nM RSL3 (n=3) (Fig. 5N). The migration was partially prevented by ferroptosis inhibitors (FST-1 and DF) (n=3) (Fig. 50; Fig. 7N). This microglial cell morphological change and their migration toward the photoreceptor layers provide another evidence of cone degeneration by ferroptosis.

Cone degeneration and microglial migration in rats

To assess whether proferroptotic agents enable for initiation of degenerative and inflammatory processes in living retinae, ferroptosis inducers were injected in the subretinal space of Long Evans rats (20pM IKE, see Fig. 8A). At day 6, optical coherence tomography (OCT) showed numerous punctuate hyperreflectivities suggestive of microglial cells in rats injected with ferroptosis inducers (arrows in Fig 8G). The same retinae were then examined with cone and microglial immunostaining. Cone photoreceptors density was evaluated at day 30 on retinal slices through the injected area. OS and perikarya were substantially reduced in animals treated by ferroptosis inducers through subretinal delivery (Fig 8H-I). The quantification on retinal sections showed a reduction of perikarya approximately by one third, and of OS by almost a half with ferroptosis inducers (Fig. 8J). Microglial migration within the injected area was quantified at D6 on whole mounted retinae, showing an increase of microglial cells at the level of the OS and at the cone synaptic terminals (Fig. 8K). Illustration of microglial cells spotted with high density in the photoreceptor layers day 6 is provided in Fig. 8 L-0. These data suggest that triggering ferroptosis in living rat retinae leads to cone degeneration and microglial migration toward the photoreceptor layers.

Macular degeneration in non-human primates A ferroptosis inducer was similarly applied in vivo on retinae of living nonhuman primates. A subretinal delivery of 20pM IKE was performed in one eye and the control solution (1 /500 DMSO) in the fellow eye with a subsequent follow-up of 4 months (see Fig 9A). A previous study showed that injection of the vehicle solution in the subretinal space had no functional or anatomical ocular or systemic adverse effect in NHP (6 months of follow-up to date in 2 NHPs), but resulted in a transient change of photoreceptors identified with adaptive optics ophthalmoscopy, which underwent a full recovery after 4 months (Dentel et al. 2023).

In eyes injected with 20pM IKE, changes in retinal structure and function were observed in vivo. Firstly, macular pigmentation surrounding foveal subretinal deposits occurred within 3 months (Fig. 9B-D). OCT also showed eroded ellipsoid zone (i.e., zone of inner and outer segments of photoreceptors) and punctuate hyperreflectivities spotted in the ONL only within the detached area with 20pM IKE (Fig 9E-G). Several subretinal deposits appeared at the foveal level at 1 month (Fig 9F) and at the macular level at 2 months (Fig 9H-J). Adaptive optics imaging (a technology that allows for in vivo cone photoreceptor visualisation) of several subretinal deposits appeared identical to those seen in patients affected by age-related macular degeneration (AMD) and showed substantial vanishing of cones within the detached area with 20pM IKE (Fig. 9K-L). Embedded OCT software allowing for ONL segmentation through the injected area showed significant ONL thickness changes within the injected area with 20pM IKE (Fig. 9M, N).

Adaptive optics was also used to quantify cone density in the injected area. The effect of a standard subretinal delivery of DMSO 0.2% on cone reflectivity in adaptive optics imaging has already been reported (Dentel et al. 2023). Cone loss was observed in eyes injected with 20pM IKE: cone density decreased of 661.00 cones/deg² versus 37.50 cones/mm² at 4 months in a 2-degree superior eccentricity area (n=2, p<0.001 ). No changes were observed in a 2-degree inferior eccentricity (i.e., outside of the injected area).

Functional alterations of cone photoreceptors were assessed by multifocal electroreti nogram (mfERG): changes from baseline in the N1 -wave amplitudes (an indicator of photoreceptor’s hyperpolarized response to light) and P1 -wave amplitudes (indicating inner retinal layers’ functionality) were analysed. N1 -wave amplitudes were reduced significantly in eyes injected with 20pM IKE (-52.19%) versus eyes injected with control solution (-2.95%) (n=2, p=0.0396). A non-significant decreasing trend of P1 -wave amplitudes was also noted (-33.03% versus -3.83%, n=2, p=0.0874). The subretinal delivery of a ferroptosis inducer at the macular level in non-human primates resulted in alterations highly reminiscent of clinical features that can be seen in macular degeneration.

EXAMPLE 2: Ferroptosis does not affect primary retinal pigment epithelium cells

MATERIALS AND METHODS

Purification of porcine cone photoreceptors

Pig eyes were dissected to obtain retinae, which were cut into pieces and digested with 4 U/ml papain (LSO 3124, Worthington,) and L-cysteine (5.5 mM, Sigma-Aldrich) for 20 minutes at 37° C. The enzymatic reaction was stopped by adding Neurobasal-A (NBA) medium (10888022, Thermo Fisher Scientific, Waltham, MA, USA) containing 5% fetal bovine serum (FBS) and 15 pg/ml DNase I (D4263, Millipore Sigma, Burlington, MA). Several brief centrifugations (30 s at 110 x g, 5 to 6 times) were performed and the supernatants were collected, excluding that from the first centrifugation, which contained mostly rod photoreceptors. Pellets were resuspended by using a P1000 pipette to dissociate the remaining cells between centrifugations. The cell suspension was then centrifuged for 10 minutes at 110 x g and pellets were resuspended in NBA medium containing 1% L-glutamine (G3126, Millipore Sigma, Burlington, MA) and 1% B27 supplement (17504044, Thermo Fisher Scientific, Waltham, MA, USA).

Cone photoreceptors were purified by lectin panning, with a slightly modified version of a previously described method (Balse et al, IOVS 2005) to obtain millions of cones in suspension. The cells were then used to seed in a transparent 384-well plates (781091 , Greiner, Ulis, France) or in a white 384-well plates (781080, Greiner, Ulis, France) at a density of either 3,000 or 4,000 cells/well with a robot (Viaflo Assist, Integra-Biosciences, Saint-Ouen, France) and incubated for three days at 37° C under an atmosphere containing 5% CO?.

Purification of porcine retinal pigmented epithelial (RPE) cells

Porcine eyes were dissected and the retinas were removed to obtain an eyecup with RPE cells at the fundus. Trypsin 0.25% (25200-056, Thermo Fisher Scientific, Waltham, MA, USA), previously heated, was added to the eyecup and incubated for one hour at 37° C. The cells were then removed from the eyecup by flushing with a pipette and were transferred to a tube containing DMEM medium (41966-029, Thermo Fisher Scientific, Waltham, MA, USA) and 20% FBS (A31605-01 , Thermo Fisher Scientific, Waltham, Massachusetts) (=D20). After centrifugation at 110 x g for 5 min, the pellet was resuspended in D20 and the cells were seeded in 60 mm Petri dishes and incubated at 37 °C and 5% CO2. The following day, the medium was changed.

The cells were then used to seed in a white 384-well plates (781080, Greiner, Ulis, France) at 4,000 cells/well with a robot (Viaflo Assist, Integra-Biosciences, Saint-Ouen, France) and incubated for three days at 37° C under an atmosphere containing 5% CO?.

ARPE- 19 cell line

The cell line was obtained from ATCC, distributed by LGC. The cells were used between P15 and P30 and seeded in D20 medium.

Treatments

The following compounds were used to induce ferroptosis: glutamate (49621 , Millipore Sigma, Burlington, MA) at a concentration of 1 pM to 1 mM in vitro, IKE (HY-114481 , MedChemExpress, Monmouth, USA) at a concentration of 200 nM to 25 pM, including 3 pM for purified cone photoreceptors and 20 pM for ex vivo and in vivo experiments and RSL3 (HY-100218A, MedChemExpress, Monmouth, USA) at concentrations of 10 nM to 10 pM, including 20 nM, 200 nM and 10 pM for cells, ex vivo and in vivo experiments, respectively. For the inhibition of ferroptosis, FST-1 (HY-100579, MedChemExpress, Monmouth, USA) was used at a concentration of 50 nM on purified cones or at 400 nM on retinal explants. DF (HY-B0568, MedChemExpress, Monmouth, USA) was effective at a concentration of 50 pM on pure cones and ZEN (HY-14164, MedChemExpress, Monmouth, USA) was effective at 20 pM on purified cones (see Figures 1 K, 5I and 7M).

CellTiter Gio (CTG) A cell viability assay was performed using CellTiter Gio (G7573, Promega, Madison, Wl) on RPE cells, ARPE-19 cells and cone photoreceptors. Viability was measured according to the manufacturer's instructions.

RESULTS

IKE and RSL3 do not alter primary RPE cells in vitro

The inventors further investigated the sensitivity threshold of retinal pigment epithelium cells (RPE) following administration of a pro-ferroptotic agent. It is well known in the literature that RPE cells can undergo ferroptosis-mediated death (Lee et al. (2022) Oxid. Med. Cell. Longev. 2022:1792894). However, the question is to determine the most sensitive cell type between RPE cells and cone PRs. The inventors therefore used the ARPE-19 cell line, which has been used in all research to characterize the RPE cells, and performed a range of concentrations of IKE, RSL3 and glutamate. When the cells were treated with IKE, a weaker effect was observed on ARPE-19 showing 34.3% living cells at 25 pM compared to 0% in cones (n=3, p = 0.0001 ) (Fig. 11A). Similarly, treatment of ARPE-19 with RSL3 resulted in a higher mortality rate, however, at concentrations considerably higher than those used to kill the cones. At 10 pM RSL3, only 5.3% of ARPE-19 cells were still alive, but at 5 pM, only 50% of cells remained alive (n=3, p = 0.0008) (Fig. 11 B), whereas the EC50 of RSL3 on cones is 1 .58 nM, demonstrating a much greater sensitivity to GPX4 inhibition in cones. Finally, treatment with glutamate had no effect on ARPE-19, unlike the cone PRs where at 1 mM all died (n=3, p= 0.0050) (Fig. 11C).

However, the inventors were also interested in comparing the effect of pro-ferroptotic molecules between the ARPE-19 cell line and primary porcine RPE cells. There was a major difference with IKE treatment, which did not affect primary cells at all, even at 25 pM (n=3, p = 0.0019) (Fig. 11A). But survival trends were similar between RSL3 and glutamate treatments (Fig. 11 B, C).

Conclusions

These results show that cone photoreceptors are more sensitive to pro-ferroptotic agents than RPE cells. Thus, a lower level of oxidative stress can trigger the death of cone photoreceptors without triggering the death of the RPE cells. Furthermore, it should be noted that the articles talking about ferroptosis in RPE cells use a cell line that divides easily and can therefore become immortal (Kozlowski et al. (2015) Curr.

Eye Res. 40:501 ). However, ferroptosis is very well described in RPE cells.

EXAMPLE 3: Ferroptosis rescued cones from degeneration in RD1 mice ex vivo and in vivo

MATERIALS AND METHODS

Animal model: mice

Mice experiments and procedures were approved by the Local Animal Ethics Committee Charles Darwin CEEACD #5 and performed in the approved facilities associated to the Institut de la Vision (Paris, France) in accordance with European Directive 2010/63/UE. All experimental work were performed according to the institutional policies on biosecurity and safety procedures. Mice were enclosed in a controlled environment and maintained under a reverse half-day dark/light cycle, with ad libitum access to food and water except during surgery.

The C3H/HeNRj line (rd1 obtained by crossing a Bagg Albino female and a DBA male) was purchased from Janvier Laboratories (Le Genest Saint-Isle, France, ISO 9001 certification), for production and breeding.

Retinal explants of rd1 mice

Eyes were collected at P15, cleaned and incubated for 20 min at 37 °C and 5% CO? in a solution of 1 /10 L-Cysteine (Millipore Sigma, Burlington, MA) at 3.5 mg/10 mL in a CO? independent medium containing glucose (6.5 g/L) in which papain (LSO 3124, Worthington) was diluted at 1 /50. Eyes were then immersed for 5 min in neurobasal medium (10888022, Thermo Fisher Scientific, Waltham, MA, USA) with 1% of L- glutamine (G3126, Millipore Sigma, Burlington, MA) = NBAg, and 10% SVF (A31605-01 , Thermo Fisher Scientific, Waltham, Massachusetts) on ice. Dissection was performed at +4°C to obtain the whole retina with the RPE. Whole-mounted retinae were flattened onto polycarbonate membranes (140652, Thermo Fisher Scientific, Waltham, Massachusetts). Tested components were diluted to 10 pM in NBAg and 1 /50 B27 supplement (17504044, Thermo Fisher Scientific, Waltham, MA, USA) and changed twice a week for 30 days (until P45 is reached). Explants were then fixed with 4% PFA (15714, Electron Microscopy Sciences, Hatfield, PA) for 1 h. Finally, PNA 488 immunolabeling (L21409, Thermo Fisher Scientific, Waltham, AAA) was performed and explants were mounted in Cellvis plates (P12-1.5H-N, IBL, Gerasdorf, Austria) with mounting medium for imaging.

Intraperitoneal injections in rd1 mice

Rd1 mice were daily injected intraperitoneally from P15 to P45 with 2 mg/kg GW 5074 (HY-10542, MedChemExpress, Monmouth, USA) diluted in 40% PEG300 (HY-Y0873, MedChemExpress, Monmouth, USA) and 5% Tween80 (HY-Y1891 , MedChemExpress, Monmouth, USA) in PBS (14190144, Thermo Fisher Scientific, Waltham, MA, USA) or the vehicle only (control group). Eyes were recovered at P45 and fixed for 1 h with 4% PFA (15714, Electron Microscopy Sciences, Hatfield, PA). Retinae were retrieved for PNA 488 immunolabeling (L21409, Thermo Fisher Scientific, Waltham, MA). The CQ1© confocal microscope (Yokogawa Electric Corporation, Tokyo, Japan) was used to image rd1 mice retinae at X10.

Imaging and statistics

The Cell PathFinder analysis software was used to quantify the number of PNA-labeled objects in the hand-drawn center of retinal explants for ex vivo experiments, and in whole-mounted retinae for in vivo experiments. The number of detected objects (= cones) was quantified over the entire area and then reported per mm².

A Mann-Whitney U two-sided test was used to assess differences of cone density in retinal explants and in injected mice.

RESULTS

Various ferroptosis inhibitors were screened on isolated porcine cones after initial exposure to 20 nM RSL3 and 3 pM IKE to seek for protective components to prevent cone loss in retinal degeneration.

The inventors then tested the neuroprotective effect of several of these components on a well-studied model of retinal degeneration, the rd1 mouse (Leveillard et al. (2004) Nat. Genet. 36:755). In vivo in rd1 mice, cone degeneration is early, starting at P15 with subsequent OS loss at P15-P20; only half of the cell bodies remain at P45. Retinae were retrieved from rd1 mice at P15 and the retinal explants were exposed to ferroptosis inhibitors in culture until P45. Several treatments have provided cone protection (Fig. 12 A-E). However, while this model leads to a rapid degeneration, the inventors may have shown molecules that can rescue cone photoreceptor degeneration in a slower degenerating disease such as AMD.

Thanks to automatic cone counting, the inventors have identified five antioxidant molecules with a significant neuroprotective effect (1 ,3-Dicaffeoylquinic acid (n=3, p=O.OO55), HTHQ (1 -0-hexyl-2,3,5-trimethylhydroquinone, n=3, p=0.0176), rosmarinic acid (n=4, p=0.0205), nobergenin (n=2, p=0.0256), tinoridine hydrochloride (n=3, p=0.0055)). On the other hand, the inventors have identified five molecules which inhibits ferroptosis process, with a significant neuroprotective effect: two of them acts directly on 5-LOX (SRS16-86 (n=3, p=0.0001 ) and caffeic acid (n=3, p=0.0055)), two of them play a role in iron metabolism (Mangiferin (n=5, p=0.0005) and astilbin (n=3, p=O.OO55). The last molecule involved in specific ferroptosis inhibition (through the system Xc pathway) demonstrated a 138.92% neuroprotective effect on pure cone photoreceptors and also proved effective on retinal explants, which led to in vivo injections with this molecule GW 5074 (n=6, p=0.0001). It was injected intraperitoneally and daily at 2mg/kg from P15 to P45. Immunolabeling on retinae retrieved at P45 showed a rescue of +81.48% of cones in rd1 mice injected with GW 5074 in comparison with the controls daily injected with the vehicle only (1597 vs 880 cones per mm², n=5 p=0.0079) (Fig. 12J).

The findings demonstrate that a ferroptosis inhibitor can reverse cone degeneration in an animal model of retinal degeneration ex vivo and in vivo.

EXAMPLE 4: Ferroptosis inhibitors do not rescue rods in RD1 mice

MATERIALS AND METHODS

Similarly to cones, rods were assessed on retinal explants of rd1 mice following exposure to ferroptosis inhibitors in culture from P15 to P45, as described in Example 3 above. The inventors also conducted the same assessment following intraperitoneal injections of GW 5074 that was described in Example 3 above.

Retinae were labeled with Rho, were imaged with the CQ1 and Cell Pathfinder was used to count rods. Whereas cones were plotted in “fold change”, results dealing with rods were plotted in density (rods per mm²) because their amounts were not enough to consider the changes from baseline in each retina (changes sometimes exceeded the total amount of rods). RESULTS

There were no significant differences between controls and ferroptosis inhibitors, either in retinal explants or in vivo (Fig. 13).

EXAMPLE 5: Ferroptosis inducers do not alter rods in non-human primates MATERIALS AND METHODS

Similarly to cone labeling in a NHP injected with 20 pM IKE in the subretinal space, rods were labeled with a rhodopsin antibody.

RESULTS

No significant changes were observed at the level of the outer segments of rods in the retina injected with 20 pM IKE with respect to the control (Fig. 14).

REFERENCES

1. Reichert, C. 0. et al. Ferroptosis Mechanisms Involved in Neurodegenerative Diseases. Int. J. Mol. Sci. 21 , 8765 (2020).

2. Xie, Y. et al. Ferroptosis: process and function. Cell Death Differ. 23, 369-379 (2016).

3. Tang, D., Chen, X., Kang, R. & Kroemer, G. Ferroptosis: molecular mechanisms and health implications. Cell Res. 31 , 107-125 (2021 ).

4. Btasiak, J., Sktodowska, A., Ulihska, M. & Szaflik, J. P. Iron and age-related macular degeneration. Klin. Oczna 111 , 174-177 (2009).

5. Zhao, T., Guo, X. & Sun, Y. Iron Accumulation and Lipid Peroxidation in the Aging Retina: Implication of Ferroptosis in Age-Related Macular Degeneration. Aging Dis. 12, 529-551 (2021 ).

6. Gielen, M. Fonctionnement des recepteurs-canaux du glutamate - Des proteines responsables de la transmission synaptique excitatrice. medecine/sciences 26, 65-72 (2010).

7. Dun, Y. et al. Expression of the cystine-glutamate exchanger (xc-) in retinal ganglion cells and regulation by nitric oxide and oxidative stress. Cell Tissue Res. 324, 189-202 (2006).

8. Bridges, C. C. et al. Induction of cystine-glutamate transporter xc- by human immunodeficiency virus type 1 transactivator protein tat in retinal pigment epithelium. Invest. Ophthalmol. Vis. Sci. 45, 2906-2914 (2004).

9. Dixon, S. J. & Stockwell, B. R. The Hallmarks of Ferroptosis. Anna. Rev. Cancer Biol. 3, 35-54 (2019).

10. Azuma, K. et al. Mitochondrial glutathione peroxidase 4 is indispensable for photoreceptor development and survival in mice. J. Biol. Chem. 298, 101824 (2022).

11. Pattnaik, B., Jellali, A., Sahel, J., Dreyfus, H. & Picaud, S. GABA c Receptors Are Localised with Microtubule-Associated Protein 1 B in Mammalian Cone Photoreceptors. J. Neurosci. 20, 6789-6796 (2000).

12. Dentel, A. et al. Adaptive Optics Flood Illumination Ophthalmoscopy in Nonhuman Primates. Ophthalmol. Sci. 3, 100316 (2023).

13. Leveillard, T. Cancer metabolism of cone photoreceptors. Oncotarget 6, 32285- 32286 (2015).

14. Dixon, S. J. et al. Ferroptosis: An Iron-Dependent Form of Non-Apoptotic Cell Death. Cell 149, 1060-1072 (2012). 15. Woodburn, S. C., Bollinger, J. L. & Wohleb, E. S. The semantics of microglia activation: neuroinflammation, homeostasis, and stress. J. Neuroinflammation 18, 258 (2021 ).

16. Di Pierdomenico, J., Garcia-Ayuso, D., Agudo-Barriuso, M., Vidal-Sanz, M. & Villegas-Perez, M. P. Role of microglial cells in photoreceptor degeneration. Neural Resen. Res. 14, 1186-1190 (2019).

17. Ren, X. & Leveillard, T. Modulating antioxidant systems as a therapeutic approach to retinal degeneration. Redox Biol. 57, 102510 (2022).

18. Totsuka, K. et al. Oxidative stress induces ferroptotic cell death in retinal pigment epithelial cells. Exp. Eye Res. 181 , 316-324 (2019).

19. Joseph, K. et al. Oxidative Stress Sensitizes Retinal Pigmented Epithelial (RPE) Cells to Complement-mediated Injury in a Natural Antibody-, Lectin Pathway-, and Phospholipid Epitope-dependent Manner*. J. Biol. Chem. 288, 12753-12765 (2013).

20. Augustin, S. et al. Melanophages give rise to hyperreflective foci in AMD, a disease-progression marker. J. Neuroinflammation 20, 28 (2023).

21. Greferath, U., Guymer, R. H., Vessey, K. A., Brassington, K. & Fletcher, E. L. Correlation of Histologic Features with In Vivo Imaging of Reticular Pseudodrusen. Ophthalmology 123, 1320-1331 (2016).

22. Fradot, M. et al. Gene therapy in ophthalmology: validation on cultured retinal cells and explants from postmortem human eyes. Hum. Gene Ther. 22, 587-593 (2011 ).

23. Liu, Y. et al. Heat Shock Proteins and Ferroptosis. Front. Cell Dev. Biol. 10, (2022).

24. Yang et al. Novel Programmed Cell Death as Therapeutic Targets in Age-related Macular Degeneration? Int J Mol Sci. 21 (19):7279 (2020).

25. Lin et al. Roles and Mechanisms of Regulated Necrosis in Corneal Diseases: Progress and Perspectives. J Ophthalmol. 2022:2695212 (2022).

26. Zhang et al. Molecular Mechanisms of Iron Mediated Programmed Cell Death and Its Roles in Eye Diseases. Front Nutr. 9:844757 (2022).

27. Rashid et al. Microglia in Retinal Degeneration. Front Immunol. 10:1975 (2019).

28. Monge et al. Subretinal drusenoid deposits: An update. Taiwan J Ophthalmol. 12(2) : 138-146 (2022).

29. Asanad S, Karanjia R. Multifocal Electroreti nogram. [Updated 2022 Oct 9]. In: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing; 2023 Jan. 30. Liu et al. The Deubiquitylase OTUB1 Mediates Ferroptosis via Stabilization of SLC7A11. Cancer Res. 79 (8): 1913-1924 (2019).

31. Wu et al. Cancer stem cell regulated phenotypic plasticity protects metastasized cancer cells from ferroptosis. Nat Commun. 13(1 ):1371 (2022).

32. Fingl & Woodbury, “General Principles” in Goodman and Gilman’s The Pharmaceutical Basis of Therapeutics, Chapter 1 , latest edition, Pagamonon Press.

33. Balse et al. Purification of Mammalian Cone Photoreceptors by Lectin Panning and the Enhancement of Their Survival in Glia-Conditioned Medium. IOVS 46(1 ):367- 374 (2005).

34. Lee, J. -J. et al. 5 -Lipoxygenase Inhibition Protects Retinal Pigment Epithelium from Sodium Iodate-Induced Ferroptosis and Prevents Retinal Degeneration. Oxid. Med. Cell. Longev. 2022, 1792894 (2022).

35. Kozlowski, M. R. The ARPE-19 Cell Line: Mortality Status and Utility in Macular Degeneration Research. Curr. Eye Res. 40, 501-509 (2015).

36. Leveillard T, Mohand-Said S, Lorentz 0, et al. Identification and characterization of rod-derived cone viability factor. Nat Genet. 36(7):755-759 (2004).

37. Lorenz et al. A Comprehensive Clinical and Biochemical Functional Study of a Novel RPE65 Hypomorphic Mutation. Invest. Ophthalmol. Vis. Sci. 49:5235-5242 (2008).

38. Vanden Berghe et al. Regulated necrosis: the expanding network of non- apoptotic cell death pathways. Nat Rev Mol Cell Biol. 15(2): 135-47 (2014).

39. Green. The Coming Decade of Cell Death Research: Five Riddles. Cell. 177(5) : 1094-1107 (2019).

40. Tang et al. The molecular machinery of regulated cell death. Cell Res 29: 347- 364 (2019).

41. Lomphithak & Fadeel. Die hard: cell death mechanisms and their implications in nanotoxicology. Toxicol Sci. kfad008 (2023).

42. Hassannia et al. Targeting Ferroptosis to Iron Out Cancer. Cancer Cell. 35(6):830-849 (2019).

43. Du & Guo. Recent progress in ferroptosis: inducers and inhibitors. Cell Death Discov. 8(1 ):501 (2022).

44. Nie et al. Induction and application of ferroptosis in cancer therapy. Cancer Cell Int. 22(1 ):12 (2022). 45. Cai et al. Ferroptosis and tumor immunotherapy : A promising combination therapy for tumors. Front Oncol. 13:1119369 (2023).

46. Yu et al. Recent trends in emerging strategies for ferroptosis-based cancer therapy. Nanoscale Adv. 5(5):1271 -1290 (2023).

47. Gao et al. Ferroptosis is an autophagic cell death process. Cell Res.26:1021 - 1032 (2016).

48. Khalili et al. Induction of rod versus cone photoreceptor-specific progenitors from retinal precursor cells. Stem Cell Res. 33:215-227 (2018).

49. Sharma et al. The Rise of Retinal Organoids for Vision Research. Int J Mol Sci. 21 (22):8484 (2020).

50. Cobb et al. Transitioning into GMP-Compliance : Alternative Methods for Producing Retinal Organoids for Transplantation. Transl Vis Sci Technol. 10(10):9 (2021 ).

51 . Belhadj et al. Long-Term, Serum-Free Cultivation of Organotypic Mouse Retina Explants with Intact Retinal Pigment Epithelium. J Vis Exp. 165:e61868 (2020)

52. Marie et al. Phototoxic damage to cone photoreceptors can be independent of the visual pigment: the porphyrin hypothesis. Cell Death Dis. 11 (8):711 (2020).

53. Hu et al. Characterization of the cystine/glutamate transporter in the outer plexiform layer of the vertebrate retina. Eur J Neurosci. 2008 28(8):1491 -502

Claims

1. A method for preparing a model of cone degeneration, comprising a step of activating a specific cell death in at least one cone photoreceptor.

2. The method of claim 1 , wherein activation of cell death results in inhibition/activation of the Xc- system, GPX-4 inhibition, lipid peroxidation, and/or modulation of iron metabolism.

3. The method of claim 1 or 2, wherein activation of cell death results in one or more of:

• reduced viability of the at least one cone photoreceptor,

• Increased permeability of the cell membrane to dyes

• reduced intracellular glutathione level in the at least one cone photoreceptor;

• increased intracellular NADP/NADPH ratio in the at least one cone photoreceptor;

• increased intracellular formation of lipid peroxides in the at least one cone photoreceptor, notably 5-oxo-6,8,11 ,14-eicosatetraenoic acid (5oxoETE);

• increased gene expression of AIFM2 (Apoptosis-inducing factor 2), SLC3A2 (Solute Carrier Family 3 Member 2), HSBP1 (Heat Shock Protein Family B (Small) Member 1 ), and Keapl (Kelch-like ECH-associated protein 1 );

• increased protein expression of ferritin and/or nuclear receptor coactivator 4 (NCOA4).

4. The method any one of claims 1 to 3, wherein activation of cell death comprises contacting the at least one cone photoreceptor with at least one cell death inducer.

5. The method of claim 4, wherein the at least one cell death inducer is a compound selected in the group consisting of glutamate, RSL3, erastin, an erastin derivative (preferably MEH, PE, or AE), imidazole ketone erastin (IKE), sulfasalazine, sorafenib, sorafenib analogs (SRS13-45 and SRS13-60), altretamine, artesunate, artemisinin, dihydroatemisinin, artemether, an artemisinin derivative , ML-162 (DPI7), ML-210 (DPI10), RSL5, 4,4'-diisothiocyanostilbene-2,2'-disulfonic acid, temozolomide, MMRi62, t-tertiary-butyl hyperoxide, FIN56, buthionine sulfoximine, NDP4928, statins, brequinar, withaferin A (WA), auranofm, almitrine, lanperisone, HG106, DPI2, DPI12, DPI13, DPI17, DPI18, DPI19, 31MEW44, and ML160.

6. The method of any one of claims 1 to 5, comprising a prior step of obtaining a retinal tissue from an eye or by production from stem cells and comprising photoreceptors, notably cone photoreceptors rod photoreceptors or their progenitors.

7. The method of claim 6, wherein the retinal tissue is a retinal tissue from a vertebrate including both human or non-human animal tissue, preferably, a nonhuman mammal, more preferably a non-human primate.

8. The method of claim 7, comprising a further prior step of dissociating the retinal tissue enzymatically and/or mechanically and/or chemically, thereby obtaining a population of isolated retinal cells.

9. The method of claim 8, comprising a further prior step of incubating the population of isolated retinal cells with peanut agglutinin (PNA) and recovering the PNA-bound cone photoreceptor.

10. The method of claim 6, wherein the retinal tissue is a retinal explant or a retinal organoid maintained in culture.

11 . The method of claim 10, comprising the further step of contacting the retinal tissue with a polycarbonate membrane, so that photoreceptors are facing away from the membrane.

12. The method of any one of claims 6 to 11 , wherein contacting the at least one cone photoreceptor with at least one cell death inducer results in cone degeneration or microglial migration to the outer nuclear layer of the retinal tissue.

13. The method of claim 6, wherein the retinal tissue is comprised in a complete eye.

14. The method of claim 13, wherein contacting the at least one cone photoreceptor with at least one cell death inducer comprises administering the at least one cell death inducer subretinal space, the suprachoroidal space, the anterior chamber, the vitreous humor, the subconjunctival space, on the corneal surface, through systemic administration (inhalator, intravenous, intramuscular or intraperitoneal) or under solution, gel or implant.

15. The method of anyone of claims 13 or 14, wherein contacting the at least one cone photoreceptor with at least one cell death inducer results in one or more of:

• proliferation, activation and/or migration of microglial cells;

• loss of cone outer segments;

• formation of subretinal deposits, preferably in the fovea and/or in the macula;

• decreased cone activity in the retina, preferably cone activity as determined by ERG or mfERG; and

16. A model of cone dystrophy or macular degeneration obtainable by the method of any one of claims 1 to 15.

17. A method of screening for compounds capable of showing protective and/or anti- degenerative properties on cone photoreceptors, the method comprising: providing the model of cone dystrophy or macular degeneration of claim 16, contacting the model of cone dystrophy or macular degeneration with a candidate compound, assessing whether the candidate compound induces a protective and/or anti- degenerative effect on cone photoreceptors.

18. The method of screening of claim 17, wherein the candidate compound induces a protective and/or anti -degenerative effect on cone photoreceptors comprises at least one of:

• maintaining or increasing viability of the at least one cone photoreceptor,

• maintaining or increasing intracellular glutathione level in the at least one cone photoreceptor;

• maintaining or decreasing the intracellular NADP/NADPH ratio in the at least one cone photoreceptor;

• maintaining or decreasing the intracellular formation of lipid peroxides in the at least one cone photoreceptor, notably 5-oxo-6,8,11 ,14-eicosatetraenoic acid (5oxoETE); • maintaining or increasing the gene expression of AIFM2 (Apoptosis-inducing factor 2), SLC3A2 (Solute Carrier Family 3 Member 2), HSBP1 (Heat Shock Protein Family B (Small) Member 1 ), and Keapl (Kelch-like ECH-associated protein 1 ); • maintaining or decreasing the protein expression of ferritin and/or nuclear receptor coactivator 4 (NC0A4);

• preventing the increase in cell membrane permeability to dyes,

• preventing the proliferation, activation or migration of microglial cells;

• preventing the loss of cone outer segments; • preventing the decrease in cone activity, preferably cone activity as determined by ERG or mfERG; and