Evaluation of Hif1a knock-down in cones of RD^∆Hif1a and AC^∆Hif1a mice. (a) Immunofluorescence of retinal sections from RD;BPCre;ZsGreen reporter mice. Sections were cut in the dorsal/ventral (left) and temporal/nasal (right) orientation and stained for OPN1SW (red). Green fluorescence indicates cells with Cre activity. Blue: DAPI. (b–d) Higher resolution images of retinal sections of RD;BPCre;ZsGreen mice showing green fluorescence from the activated reporter alone (b), or in combination with cones expressing OPN1SW (c) or OPN1MW (d). (e) High resolution image of a retinal section of an AC;BPCre;ZsGreen mouse. (f) PCR amplification of Hif1a genomic DNA isolated from retinas of RD^∆Hif1a, AC^∆Hif1a (kd) and their respective control (ctrl) mice. The floxed, not excised (not exc) Hif1a sequence is detected at approximately 900 bp and Cre-mediated deletion results in a fragment of 270 bp (excised, exc). Higher levels of Hif1a excision are expected in AC mice based on the increased number of S-cones in these mice. (g,h) Relative expression levels of Cre (g) and Hif1a (h) mRNA in retinas of indicated mice at 8 weeks of age. mRNA levels were normalized to Actb and expressed relatively to the levels in RD^∆Hif1a mice, which levels were set to 1. Shown are means ± SD of n = 3. One-way ANOVA and Tukey’s test for multiple comparisons was used to analyze significance. ONL: outer nuclear layer, INL: inner nuclear layer, GCL: ganglion cell layer. Scale bars: 500 µm (a), and 100 µm (b–e).

Retinal gene expression in RD^∆Hif1a and AC^∆Hif1a mice during ageing. (a) HIF1-responsive genes (Vegf, Egln1, Adm) and Epas1 (Hif2a). (b) Cone-specific genes (Opn1sw, Opn1mw) and genes involved in degeneration (Casp1), gliosis (Gfap) or inflammation (Ccl2). Expression was determined in RD^∆Hif1a and AC^∆Hif1a and their respective controls at 4, 8, 12 and 26 weeks of age (as indicated). Expression levels were normalized to Actb and shown relative to 4-week-old RD^ctrl mice, the levels of which were set to 1 (dotted line). Shown are means ± SD of n = 3. Significance between knockdown and control mice was calculated at every individual time point by two-way ANOVA with Šídák’s multiple comparison test *p < 0.05; **p < 0.01; ***p < 0.001.

Retinal morphology following Hif1a inactivation in cones. (a,b) Retinal morphology analyzed at 4, 12 and 26 weeks of age in RD^ΔHif1a (a) and AC^ΔHif1a (b) mice and their respective Cre-negative controls (RD^ctrl and AC^ctrl). (c,d) Immunostaining for OPN1SW in retinal cross-sections of the central and peripheral retina of 4- and 26-week-old RD^ΔHif1a (c) and AC^ΔHif1a (d) mice and their respective controls (ctrl). Cell nuclei were counterstained with DAPI. (e,f) Blood vessels labelled with isolectin on retinal flat mounts of RD^ΔHif1a (e) and AC^ΔHif1a (f) mice and their respective controls at 4 weeks of age. Vessels are highlighted with pseudocoloring of the 3 vascular plexi (deep (D): green; intermediate (I): red; primary (P): blue). Lower panels show the intermediate plexus separately. Scale bars as indicated.

Retinal function in RD^∆Hif1a and AC^∆Hif1a mice. Averaged light-adapted (a–d) and dark-adapted (e,f) single flash ERG traces of RD^ΔHif1a and AC^ΔHif1a mice and their respective control littermates. A- and b-wave amplitudes were plotted as a function of light intensity (right panels, a–f). Photopic (a,b) and scotopic (e,f) single white-light flash intensity series at 12- (a,e) and 26- (b,f) weeks of age. Photopic single UV-light flash intensity series in 3-month-old RD^ΔHif1a (c) and AC^ΔHif1a (d) mice and their respective control littermates. Shown are means ± SD of n ≥ 3 (ctrl/kd = a (RD 5/4; AC 4/3); b (RD 4/4; AC 4/4); c (6/6); d (3/3); e (4/4); f (4/4)). Two-way ANOVA with Šídák’s multiple comparison test was used for statistical analysis. *p < 0.05; **p < 0.01.

Retinal hypoxia tolerance in the RD^∆Hif1a and AC^∆Hif1a mice. (a) RD^ΔHif1a, RD^ctrl, AC^ΔHif1a and AC^ctrl mice were kept under normoxic conditions (N) or exposed to hypoxia (H, 7% O₂, 6 h). Gene expression levels in total retinal RNA were determined immediately after hypoxic exposure. Expression levels were normalized to Actb and shown relative to the control littermates under normoxic conditions, which levels were set to 1 for each group (RD and AC). Shown are means ± standard deviation (SD) for n = 3. The differences in gene expression levels between normoxic and hypoxic conditions were tested for significance using a two-way ANOVA with Šídák’s multiple comparison test *p < 0.05; **p < 0.01; ***p < 0.001. (b) Detection of OPN1SW (red) and GFAP (green) in sections of the central or peripheral retina of RD^ΔHif1a, AC^ΔHif1a and their respective controls at 10 days after hypoxic exposure. Cell nuclei were counterstained with DAPI (blue). ONL: outer nuclear layer, INL: inner nuclear layer. All mice were 12-weeks of age. Scale bar as indicated.

Cones are not affected in RD^{ΔHif1a;Hif2a} mice. (a) Detection of Hif1a (left) and Hif2a (right) deletion fragment (exc) in genomic DNA isolated from retinas of in RD^{ΔHif1a;Hif2a} mice (kd) and their control littermates (ctrl). Not exc: not excised, floxed DNA sequence. (b) Relative expression levels of Hif1a and Hif2a mRNA in retinas of 8-week-old control (ctrl) and RD^{∆Hif1a;Hif2a} (kd) mice determined by semi-quantitative real-time PCR. (c) Retinal morphology of RD^{ΔHif1a;ΔHif2a} mice and their respective Cre-negative controls (RD^ctrl) at 12 weeks of age. RPE, retinal pigment epithelium; ONL, outer nuclear layer; INL, inner nuclear layer. (d) Relative levels of indicated mRNAs in retinas of 8, 12 and 26-week-old RD^{∆Hif1a;Hif2a} mice and their RD^ctrl littermates determined by semi-quantitative real-time PCR. (e,f) ERG recordings in 3-month-old RD^{∆Hif1a;Hif2a} mice and their control littermates. Averaged traces evoked by single flashes of white light with increasing intensities under light-adapted photopic (e) and dark-adapted scotopic (f) conditions. The a- and b-wave amplitudes (right panels) were plotted as a function of light intensity and shown as means ± SD of n = 3. (g) Cones detected with peanut agglutinin (PNA, green) and OPN1SW (red) staining in central or peripheral retinal sections of 12-week-old RD^{ΔHif1a;Hif2a} and their respective control (RD^ctrl) mice. (h,i) Light damage susceptibility of RD^{∆Hif1a;Hif2a} and RD^ctrl mice after hypoxic preconditioning. Mice were exposed to high levels of white light without (middle panels; LE) or were hypoxia-preconditioned before (right panels; HYP + LE) and analyzed 10 days following the insult by immunofluorescence (h) or semi-quantitative real-time PCR (i). Control animals were not subjected to any treatment (control). By the time of death mice were 12- to 14-week-old. Cone arrestin (CAR, red) and GFAP (green) immunoreactivity was analyzed on retinal cross-sections. Cell nuclei were counterstained with DAPI (blue). ONL: outer nuclear layer, INL: inner nuclear layer. Scale bars as indicated. mRNA levels (b,d,i) were normalized to Actb and expressed relatively to ctrl, which were set to 1. Shown are means ± SD of n = 3. Two-way ANOVA with Šídák’s multiple comparison test was used for all statistical analyses.