CCR2(+) monocytes infiltrate atrophic lesions in age-related macular disease and mediate photoreceptor degeneration in experimental subretinal inflammation in Cx3cr1 deficient mice

Abstract

Atrophic age-related macular degeneration (AMD) is associated with the subretinal accumulation of mononuclear phagocytes (MPs). Their role in promoting or inhibiting retinal degeneration is unknown. We here show that atrophic AMD is associated with increased intraocular CCL2 levels and subretinal CCR2(+) inflammatory monocyte infiltration in patients. Using age- and light-induced subretinal inflammation and photoreceptor degeneration in Cx3cr1 knockout mice, we show that subretinal Cx3cr1 deficient MPs overexpress CCL2 and that both the genetic deletion of CCL2 or CCR2 and the pharmacological inhibition of CCR2 prevent inflammatory monocyte recruitment, MP accumulation and photoreceptor degeneration in vivo. Our study shows that contrary to CCR2 and CCL2, CX3CR1 is constitutively expressed in the retina where it represses the expression of CCL2 and the recruitment of neurotoxic inflammatory CCR2(+) monocytes. CCL2/CCR2 inhibition might represent a powerful tool for controlling inflammation and neurodegeneration in AMD.

Keywords: age-related macular disease; chemokines; monocyte; neurodegeneration; neuroinflammation.

Figure 1. Intraocular CCL2 levels and CCR2⁺ inflammatory monocytes are increased in atrophic AMD

1. A. CCL2 ELISA of aqueous humours of geographic atrophy (GA) patients and control subjects (n = 18 GA patients n = 22 control patients, student t-test p < 0.0001; Mann–Whitney test p < 0.0001).

2. B–D. CCL2 immunohistochemistry (red staining) on macular sections of (B) control donor tissues, (C and D) within the GA lesion.

3. E–H. CCR2 immunohistochemistry (red staining) on macular sections of (E) control donor tissues, inset: major retinal vessel containing erythrocytes and leucocytes, (F) within the GA lesion, inset: adjacent to GA lesion (G) laminar deposit (H) soft drusen.

4. I–L. (I) CCR2 (green staining), (J) CD18 (red staining), (K) merge double labelling of a GA lesion. (L) Quantification of intraretinal and subretinal CCR2/CD18 positive cells expressed as CCR2⁺ cells/500 mm of the atrophic lesion (n = 10 GA donor maculae from 7 patients: age, mean (SD): 84 (8.8) and n = 5 control maculae from 5 patients: age, mean (SD): 83 (8.8), *students t-test p = 0.001).

5. M, N. IBA-1 immunohistochemistry (red staining) on macular sections of (M) control donor tissues, (N) within the GA lesion.

6. O. (O) Quantification of intraretinal and subretinal IBA-1 positive cells expressed as IBA-1⁺ cells/500 mm of the atrophic lesion (subjects same as above, *students t-test p = 0.005). B–L and N and O: representative images from 5 healthy donors, 7 donors (10 eyes) with GA and 3 donors with age related maculopathy (4 eyes), controls omitting the primary antibody showed no staining. All values are represented as mean ± SEM. CTL: control; GA: geographic atrophy; ONL: outer nuclear layer; INL: inner nuclear layer; Ch: choroid; iR: inner retina; sR: subretinal; Scale bar B–J = 50 μm.

Figure 2. CCL2 expression in age and light-challenged Cx3cr1^−/− mice

1. A. Quantitative RT-PCR of Ccl2 mRNA normalized with β-actin mRNA of 2- and 18-month-old C57BL/6 and Cx3cr1^−/− mouse retina (n = 4 per group, *two-way Anova, Bonferroni p < 0.001).

2. B. Quantitative RT-PCR of Ccl2 mRNA normalized with β-actin mRNA of non-injured (NI) and at day 4 and day 14 of the light-challenge model of 2- to 3-month-old C57BL/6 and Cx3cr1^−/− mice (n = 5–6 per group, *two-way Anova, Bonferroni p < 0,05).

3. C. CCL2 ELISA protein quantification of retinal protein extracts from non-injured (NI) and at day 14 (d14) of the light-challenge model of 2- to 3-month-old C57BL/6 and Cx3cr1^−/− mice (expressed as pg/mg total retinal protein; n = 4 per group, *two-way Anova, Bonferroni p < 0.001).

4. D–G. Immunohistochemistry CCL2 (red; arrows), and IBA-1 (green; arrows colocalization yellow) of the subretinal side (D and E) and vitreal aspect (F and G) of a retinal flatmount from a Cx3cr1^−/− mouse at day 14 (d14) of the light-challenge model (representative of 3 independent experiments, immunostaining omitting the primary antibody or performed on Ccl2^−/− mice served as negative controls).

5. H. Quantitative RT-PCR of Ccl2 mRNA normalized with β-actin mRNA of 2- to 3-month-old C57BL/6, Ccl2^−/−, Ccr2^−/−, Cx3cr1^GFP/GFP and Cx3cr1^GFP/GFPCcr2^RFP/RFP mice at day 4 of the light-challenge model (d4) (n = 5–6 per group, *two-way Anova, Bonferroni p < 0.01).

6. I. Quantitative RT-PCR of Ccl2 mRNA normalized with S26 of whole-eye-lysats (set as 1) and FACS-sorted GFP^highLy6C^neg and GFP^lowLy6C^high cells pooled from 8 eyes of PBS perfused Cx3cr1^GFP/GFP mice after 4 days of light challenge.

7. J. (J) Quantitative RT-PCR of Ccl2 mRNA normalized with S26 mRNA of Cx3cr1^+/+ and Cx3cr1^−/− monocyte derived Mφs cultivated for 18 h with or without photoreceptor outer segments (POSs, n = 4 per group, two-way Anova Bonferroni *p < 0.001). All values are represented as mean ± SEM. Scale bar D–G = 50 μm; CTL: control; +POS: +photoreceptor outer segments.

Figure 3. CCL2 mediates subretinal MP accumulation in age and light-challenged Cx3cr1^−/− mice

1. A–D. 12 month-old IBA-1 stained RPE-flatmounts of C57BL/6 (A), Ccl2^−/− (B), Cx3cr1^−/− (C) and Cx3cr1^−/−Ccl2^−/− (D).

2. E. Quantification of subretinal MPs/mm² at 3, 9 and 12 months (n = 6 per group, *two-way Anova Bonferroni p < 0,001).

3. F. Quantification of subretinal MPs/mm² non-injured and at day 4 and 14 of the light-challenge model of 2- to 3-month-old C57BL/6J, Ccl2^−/−, Cx3cr1^−/− and Cx3cr1^−/−Ccl2^−/− mice (n = 6–9 per group, *two-way Anova Bonferroni p < 0.001).

4. G. Quantification of IBA-1 positive subretinal MPs/mm² in non-injured and at day 14 of the light-challenge model of 2- to 3-month-old C57BL/6J, Ccl2^−/−, CCR2^−/−, Cx3cr1^GFP/GFP, Cx3cr1^GFP/GFPCcl2^−/− and Cx3cr1^GFP/GFPCcr2^RFP/RFP mice (n = 6–9 per group, *two-way Anova Bonferroni p < 0.001).

5. H–K. Representative fundoscopic photographs of 12-month-old C57BL/6 (H), Ccl2^−/− (I), Cx3cr1^−/− (J) and Cx3cr1^−/−Ccl2^−/− (K) (n > 8 per group). All values are represented as mean ± SEM. Scale A–D = 100 μm.

Figure 4. CCL2 deficiency protects Cx3cr1^−/− mice from age and light-induced photoreceptor degeneration

1. A–D. Micrographs, taken 1000 μm from the optic nerve of 12-month-old C57BL/6J (A), Ccl2^−/− (B), Cx3cr1^−/− (C, arrows: photoreceptor nuclei; star: nucleus of a subretinal macrophage) and Cx3cr1^−/−Ccl2^−/− (D).

2. E–G. (E) Photoreceptor nucleus rows at increasing distances (−3000 μm: inferior pole, +3000 μm: superior pole) from the optic nerve (0 μm) in 12-month-old C57BL/6J, Ccl2^−/−, Cx3cr1^−/− and Cx3cr1^−/−Ccl2^−/− mice (n = 4–7, the area under the curve in Cx3cr1^−/− mice tested significantly different from the other strains *one-way Anova Bonferroni p < 0.05). Hoechst (blue staining), TUNEL (red staining), rhodopsin (green staining) labelling of Cx3cr1^−/− mice at d14 of the light-challenge model in sections (F inset showing rhodopsin double labelling) and confocal microscopy of flatmounts with z-stack projections (G).

3. H–K. TUNEL stained retinal flatmounts of d14 light-challenged C57BL/6 (H), Ccl2^−/− (I), Cx3cr1^−/− (J) and Cx3cr1^−/−Ccl2^−/−(K) mice.

4. L. Quantification of TUNEL positive cells per retina (n = 4–6 per group, *one-way Anova Bonferroni p < 0.01). All values are represented as mean ± SEM. ONL: outer nuclear layer. Scale bars: A–D and G = 50 μm; F and H–K = 100 μm.

Figure 5. Subretinal MP accumulation and photoreceptor degeneration in Cx3cr1^−/− mice is dependent on CCL2 mediated monocyte recruitment

1. A, B. RT-PCR of relative (A) Cx3cr1 and (B) Ccr2 mRNA expression normalized with S26 mRNA in C57BL/6 blood monocytes (Mo), retina (R) and retinal microglial cells (MC) (n = 3 independent cell preparations, all groups significantly different from each other for Cx3cr1 by one-way Anova Bonferoni p < 0.05; Mo significantly different from MC and R for Ccr2 by two-way Anova Bonferoni p < 0.001).

2. C, D. Confocal microscopy of (C) RFP fluorescence and (D) merged RFP and GFP fluorescence of the outer plexiform layer of non-injured Cx3cr1^+/GFPCcr2^+/RFPmice.

3. E. Quantitative RT-PCR of Ccr2 mRNA normalized with S26 mRNA of Cx3cr1^+/+ and Cx3cr1^−/− freshly prepared monocytes and monocyte derived Mφs cultivated for 18 h in direct contact with photoreceptor outer segments (POS) (n = 4 per group, two-way Anova Bonferroni p < 0.001 significant difference in monocytes compared to Mφs; no difference between genotypes).

4. F, G. (F) RFP fluorescence and (G) merged RFP and GFP fluorescence of Hoechst stained chroidal/RPE flatmount after 4 days of light-challenge of Cx3cr1^GFP/GFPCcr2^+/RFP mice.

5. H. Quantification of RFP positive subretinal cells after 4 days of light-challenge in Cx3cr1^+/GFPCcr2^+/RFP, Cx3cr1^GFP/GFPCcr2^+/RFP and Cx3cr1^GFP/GFPCcr2^RFP/RFP mice (n = 4 per group, *one-way Anova Bonferroni p < 0.01).

6. I. Quantification of the percentage of CD115⁺ (inset green staining) and CD115⁺EdU⁺ cells (inset red/green staining) of all Hoechst positive leukocytes of blood smears from day 4 light-challenged Cx3cr1^−/− mice that all received three daily intraperitoneal EdU injections and one daily intravenous injection of either control or clodronate liposome (n = 4 per group, two-way Anova Bonferroni p < 0.001 significant difference in the number of CD115⁺ and CD115⁺EdU⁺ cells in clodronate liposome treated animals compared to controls).

7. J, K. IBA-1 (green) EdU (red) double labelled chroidal/RPE flatmount of intraperitoneally EdU-injected light-challenged Cx3cr1^−/− mice receiving control liposome injections (J) or clodronate liposome injections (K).

8. L. Quantification of subretinal IBA-1⁺ and IBA-1⁺EdU⁺ cells of control liposome and clodronate liposomes treated 4 days light-challenged Cx3cr1^−/− mice (n = 4 mice per group, two-way Anova Bonferroni p < 0.001 significant difference in the numbers of subretinal *IBA-1⁺ MPs and †IBA-1⁺EdU⁺ MPs between clodronate liposome and control liposome treated mice).

9. M, N. TUNEL stained retinal-flatmounts of control liposome (M) and clodronate liposome (N) treated light-challenged Cx3cr1^−/− at d14.

10. O, P. O) Quantification of IBA-1 positive subretinal MPs/mm² (n = 4–7 per group, *Mann–Whitney p = 0.03) and (P) TUNEL positive photoreceptor nuclei in control liposome and clodronate liposome treated light-challenged Cx3cr1^−/− at d14 (n = 4–7 * Mann–Whitney p < 0.0001). All values are represented as mean ± SEM. R: Retina; MC: microglial cells; Mo: monocytes; POS: photoreceptor outer segments; lipo: empty control liposomes; lipo-clo: clodronate liposomes. Scale bars = 50 μm.

Figure 6. Photoreceptor toxicity of wildtype and Cx3cr1^−/− monocytes and microglial cells on retinal explants

1. A–C. Confocal microscopy of flatmounts with z-stack projections (Hoechst in blue) of TUNEL (red) stained retinal-flatmounts of cultured for 18 h without MPs (A), in contact with C57BL/6 monocytes (B) and Cx3cr1^−/− monocytes (C).

2. D. Quantification of TUNEL⁺nuclei/mm² in the ONL of the different groups (n = 8 per group, one-way Anova Bonferroni *significant difference between explants without MPs and C57BL/6 monocytes p < 0.01; † significant difference between explants with C57BL/6 monocytes and Cx3cr1^−/− monocytes p < 0.01; $ significant difference between Cx3cr1^−/− monocytes and Cx3cr1^−/− MC p < 0.001). All values are represented as mean ± SEM. Mo: monocytes; MC: microglial cells. Scale bar = 50 μm.

Figure 7. Blocking CCR2/CCL2 axis protects Cx3cr1^−/− mice from MP accumulation and photoreceptor degeneration

1. A, B. IBA-1 stained RPE-flatmounts of PBS (A) and CCR2 inhibitor RS 102895 (B) treated light-challenged Cx3cr1^−/− mice at d14.

2. C–E. C) Quantification of IBA-1 positive subretinal MPs/mm² (n = 6–10 per group, *Mann–Whitney p = 0.01). TUNEL stained retinal-flatmounts of PBS (D) and CCR2 inhibitor (E) treated light-challenged Cx3cr1^−/− mice at d14.

3. F. Quantification of TUNEL positive photoreceptor nuclei per retina (n = 4 per group, *Mann–Whitney p = 0.028). All values are represented as mean ± SEM. Scale bar = 100 μm.