Subretinal mononuclear phagocytes induce cone segment loss via IL-1β

Abstract

Photo-transduction in cone segments (CS) is crucial for high acuity daytime vision. For ill-defined reasons, CS degenerate in retinitis pigmentosa (RP) and in the transitional zone (TZ) of atrophic zones (AZ), which characterize geographic atrophy (GA). Our experiments confirm the loss of cone segments (CS) in the TZ of patients with GA and show their association with subretinal CD14(+)mononuclear phagocyte (MP) infiltration that is also reported in RP. Using human and mouse MPs in vitro and inflammation-prone Cx3cr1(GFP/GFP) mice in vivo, we demonstrate that MP-derived IL-1β leads to severe CS degeneration. Our results strongly suggest that subretinal MP accumulation participates in the observed pathological photoreceptor changes in these diseases. Inhibiting subretinal MP accumulation or Il-1β might protect the CS and help preserve high acuity daytime vision in conditions characterized by subretinal inflammation, such as AMD and RP.

Keywords: age-related macular degeneration; human; human biology; interleukin; macrophages; medicine; mouse; neuroscience; outer segments; retinitis pigmentosa; subretinal space.

Figure 1.. Rhodopsin, cone arrestin, and L/M cone opsin staining on central sections from control and geographic atrophy patients.

(A–D) Representative micrographs of immunohistochemical detection of rhodopsin (red), cone arrestin (green, A–D and left insets of B, C and D), or L/M cone opsin (green, right insets of B, C and D) and Hoechst nuclear stain (blue, RPE autofluorescence visible in orange) on central sections of control donors (A) and donors with geographic atrophy (GA, panel B–D) at a distance greater than 1000 μm from the atrophic zone (AZ) (B), at the boundary of the AZ and transitional zone (TZ, panel C; white arrow indicates the margin of the autofluorescent RPE), and within the AZ (D). (E and F) Quantification of the number of arrestin⁺cone somata (E, one way ANOVA, Dunnett’s post test *p=0,0024) and rhodopsin⁺rod somata (F, one way ANOVA, Dunnett’s post test *p<0.0001) in control- and GA-donors (10 GA samples from 9 donors and in 6 control samples from 6 donors). ONL: outer nuclear layer; c-arrestin: cone-arrestin; RPE: retinal pigment epithelium; CTL: control; AZ: atrophic zone; TZ: transitional zone; GA: geographic atrophy. Scale bar = 50 μm. DOI: http://dx.doi.org/10.7554/eLife.16490.002

Figure 2.. CD14, peanut agglutinin and cone arrestin staining on central flatmount preparations from control and geographic atrophy patients.

Representative micrographs of immunohistochemical detection (confocal Z stack projections) of CD14 (white), cone arrestin (red), peanut agglutinin (PNA,green), and Hoechst nuclear stain (blue, RPE autofluorescence visible in orange) of RPE/choroid- (A and E) and retinal- flatmounts (B–D and F–H) of a control donor (A–D) and a donor with geographic atrophy (GA, panel E–H). (E) 3D reconstruction at higher magnification of a CD14⁺cell on RPE flatmounts in an orthogonal (E, upper inset) and perpendicular (E, lower inset). (F) Representative confocal micrographs of CD14 (F, upper inset) -IBA-1 (F, lower inset) double labeling. (D and H) Oblique and perpendicular (insets) 3D reconstruction views of the outer aspect of the retinal flatmounts. The margin of the atrophic zone (AZ), recognized by the loss of RPE (E) and by the irregular cone distribution and a thinned outer nuclear layer (G) is indicated by the yellow dotted line. (G) PNA/arrestin pattern distant (>1000 μm) from the AZ of a patient with GA (G inset). Experiments on flatmounts from 4 different control and patients with GA gave similar results. AZ : atrophic zone; TZ : transitional zone. Scale bars A and E = 100 μm; B–C and F–G = 50 μm; inset D and H = 20 μm. DOI: http://dx.doi.org/10.7554/eLife.16490.003

Figure 3.. Evaluation of rods and cones in monocyte/retinal explant co-cultures.

(A and B) Orthogonal projection of confocal Z stack images of the outer nuclear layer of TUNEL- (red), Hoechst nuclear dye- (blue) stained mouse retinal explants after 18 hr of culture (A) or co-culture with human monocytes (hMo, panel B). (C) TUNEL(red)/rhodopsin(green) co-staining of a section of a mouse retinal explant co-cultured with hMo (the inset represents a longer green exposure of the outer nuclear layer). (D) Quantification of TUNEL⁺nuclei in mouse retinal explants cultured with or without hMo (n = 4/group, Mann Whitney *p=0,028). (E and F) Orthogonal projection of confocal Z stack images of the photoreceptor segments of mouse retinal explants after 18 hr of mono-culture (E) or co-culture with hMo (F) after cone arrestin (red)/peanut agglutinin (PNA, green) staining. (G) Quantification of cone numbers in retinal explants cultured with or without hMo (n = 10/group). (H and I) Oblique-, and perpendicular (insets) -3D reconstruction views of confocal Z stack images of the photoreceptor segments of mouse retinal explants after 18 hr of mono-culture (H) or co-culture with hMo (I) after peanut agglutinin (PNA, green) and cone-arrestin (red, insets) staining. (J) Quantification of cone segment volume in mouse retinal explants cultured without or with the indicated numbers of hMos (n = 6/group, Mann Whitney *p<0,0004). (K) Quantification of cone segment volume in mouse retinal explants cultured without or with 100 000 Cx3cr1-deficient bone marrow-derived Mos (n = 4–6/group, Mann Whitney *p<0,0061). (L and M) Orthogonal-, and perpendicular (insets) -3D reconstruction views of confocal Z stack images of the photoreceptor segments of macaque retinal explants after 18 hr of mono-culture (L) or co-culture with hMo (M) after peanut agglutinin (PNA, red) and cone-arrestin (green) staining. (N) Quantification of cone segment volume in macaque retinal explants cultured without or with 100 000 hMos (n = 8/group, Mann Whitney *p<0,0002). hMo/Mo : human monocyte; arr: cone-arrestin; CS : cone segments; PNA : peanut agglutinin. Scale bar : A–F = 50 μm; H and I = 10 μm. DOI: http://dx.doi.org/10.7554/eLife.16490.004

Figure 4.. The influence of IL-1β retinal explants co-cultures.

(A) Quantitative RT-PCR of RT-PCR of IL-1β and IL-18 normalized with S26 mRNA in fresh human monocytes (hMo), hMo cultured alone or with a retinal explant for 18 hr (both on polycarbonate filters, n = 5, ANOVA, Dunnett’s post test*p=0,0079). (B and C) Oblique and perpendicular (insets) 3D reconstruction views of 18 hr peanut agglutinin (PNA)-stained retinal explant (B) and IL-1β (50 ng/ml) exposed explant (C). (D) Quantification of cone segment volume in retinal explants cultured with or without IL-1β (n = 6/group, Mann Whitney *p=0,0087). (E and F) Oblique and perpendicular (insets) 3D reconstruction views of 18 hr peanut agglutinin (PNA)-stained retinal explant co-cultured with human monocytes without (E) or with IL-1 receptor antagonist (F, 10 mg/ml). (G) Quantification of cone segment volume in Mo/retinal co-cultures with or without an IL-1 receptor antagonist (n = 6/group, Kruskal-Wallis, Dunn’s post test *p=0,0022 versus 'without hMo'; †p=0,0182 versus “with hMo without IL1-Ra). RetEx: retinal explant; CTL: control; CS: cone segment; hMo: human Monocytes. Scale bar B, C, E, F = 10 μm. DOI: http://dx.doi.org/10.7554/eLife.16490.005

Figure 5.. Cone segments in light-induced subretinal inflammation of C57BL6/J and Cx3cr1^GFP/GFP-mice and the effect of pharmacological IL-1β inhibition.

(A–C) Orthogonal projections of confocal Z stack images of the photoreceptor segments of peanut agglutinin (PNA, red) stained Cx3cr1^GFP/GFP-mice (expressing GFP under the Cx3cr1 promoter) that were kept under normal light conditions (A), or exposed to an inflammation-inducing light-challenge (LC) and treated with PBS (B) or IL-1β receptor antagonist (C, IL-1Ra). (D and E) Quantification of cone segment volume (D) and subretinal mononuclear phagocytes (E) in normal light raised and LC-mice (n = 5–8/group; Kruskal-Wallis, Dunn’s post test *p=0,0109 versus Cx3cr1^GFP/GFP-mice in normal light; †p=0,0028 versus light-challenged Cx3cr1^GFP/GFP-mice). PNA: peanut agglutinin; GFP: green fluorescent protein; IL-Ra: IL-1β receptor antagonist LC: illuminated/light-challenged. Scale bar 30 μm. DOI: http://dx.doi.org/10.7554/eLife.16490.006

Author response image 1.. Additional examples of CD14-stained RPE flatmounts

DOI: http://dx.doi.org/10.7554/eLife.16490.007