Ocular neuroprotection by siRNA targeting caspase-2

Abstract

Retinal ganglion cell (RGC) loss after optic nerve damage is a hallmark of certain human ophthalmic diseases including ischemic optic neuropathy (ION) and glaucoma. In a rat model of optic nerve transection, in which 80% of RGCs are eliminated within 14 days, caspase-2 was found to be expressed and cleaved (activated) predominantly in RGC. Inhibition of caspase-2 expression by a chemically modified synthetic short interfering ribonucleic acid (siRNA) delivered by intravitreal administration significantly enhanced RGC survival over a period of at least 30 days. This exogenously delivered siRNA could be found in RGC and other types of retinal cells, persisted inside the retina for at least 1 month and mediated sequence-specific RNA interference without inducing an interferon response. Our results indicate that RGC apoptosis induced by optic nerve injury involves activation of caspase-2, and that synthetic siRNAs designed to inhibit expression of caspase-2 represent potential neuroprotective agents for intervention in human diseases involving RGC loss.

Figure 1

Caspase-2 is cleaved in RGC after ONC. (a) qPCR analysis of caspase-2 mRNA abundance in Thy1.1⁺ (bound to magnetic beads coupled to anti-Thy1.1 antibodies) and other retinal cells (unbound). (b) Representative microphotographs of sections of rat retina 24 h after ONC. Arrowheads show cleaved (p12) caspase-2 (C-CASP2) immunoreactivity (original magnification × 40). (c) High-power images (original magnification x 63) of top panel in (b) demonstrating cleaved caspase-2 (C-CASP2) immunoreactivity (brown) in the GCL (arrowheads). (d) Double immunohistochemical staining shows that CASP2 (green; arrowheads) is present in higher amounts at 7 days after ONC in RGC compared with intact control βIII-tubulin-positive RGC (red). The blocking peptide used as a negative control gave no CASP2⁺ immunoreactivity. (e) Percentage of CASP2/βIII-tubulin-double-positive cells among βIII-tubulin-positive cells, all in the ganglion cell layer±S.D. (i.e. RGC; calculations were made from 18 retinas per treatment and sampled from five different areas of each retina). (f) Double immunohistochemistry did not detect C-CASP2 (green) in RGC (red) of intact controls, but did localize C-CASP2 (colocalization=orange) in RGC at 7 days post-ONC scale bar (d) and (f)=25 μm; IHC=immunohistochemistry, GCL=ganglion cell layer, INL=inner nuclear layer, ONL=outer nuclear layer, IPL=inner plexiform layer, block pep=blocking peptide). (g) Percentage of βIII-tubulin-positive cells in the ganglion cell layer (RGC) also containing C-CASP2±S.D.

Figure 2

Target knockdown activity, nuclease stability and interferon response-inducing properties of siCASP2. (a) siCASP2 dose-dependent knockdown of caspase-2 mRNA in HeLa cells±S.D. (b) Analysis of siCASP2 integrity on native polyacrylamide gel following incubation in silico in rabbit vitreal fluid. ‘0' time point corresponds to siRNA aliquot dissolved in PBS (size control). (c) Induction of interferon (IFN)-responsive gene expression in rat retina/choroid after intravitreal injection of either 20 μg of siCASP2, 5–20 μg of poly-(I:C) or PBS (n=3–10), and the expression of IFN-responsive genes, IFIT1 and OAS1B 6 (not shown as no induction of gene expression was detected at this time point in any group) analysed 24 h later by quantitative RT-PCR. The results are expressed as % increase over control intact eyes (100%) (±S.D.) (IFIT1, interferon-induced protein with tetratricopeptide repeats 1; OAS1B 6, 2′-5′-oligoadenylate synthase 1B 6)

Figure 3

Localization, stability and RNAi activity of intravitreally injected siCASP2. (a) Detection of Cy3 fluorescence in isolated RGC 1 h (red line) or 18 h (green line) following single intravitreal injection of 40 μg Cy3-siRNA per rat eye. FACS analysis displaying results obtained from representative one out of two retinal tissue pools analysed per time point are shown. (b) Representative micro-autoradiographs of paraffin sections (counterstained with haematoxylin and eosin) from untreated ((i), (ii)) and siCASP2-injected rat eyes ((iii)–(iv)), which were hybridized with either siCASP2-specific ((i)–(iv)) or with a nonspecific ((v), (vi)) control ³³P-labelled probe ((i), (iii), (v)) – bright field images, ((ii), (iv), (vi)) – corresponding dark field images. In dark field images, hybridization signals are white dots over all retinal layers in sections of siCASP2-injected eyes hybridized to the specific probe ((iv)) and were especially prominent in the GCL (arrow). No signal in the GCL is detected in control sections ((ii), (vi)) (exposure 24 h, original magnification × 20). The eyes for in situ hybridization analysis were enucleated at 2 h after intravitreal injection of 20 μg siRNA. At least 10 eyes from each group were analysed from 2 to 3 different experiments. (c) Quantification of siCASP2±S.D. in retina at different times after intravitreal injection of 20 μg per eye siCASP2 (n=6). Control retinas were obtained from intact non-injected eyes. (d) Detection of siCASP2-mediated RNAi using RLM-RACE: (i) EtBr-stained agarose gel with electrophoresed RLM-RACE products from siCASP2-transfected PC12 cells (size control) and from retinas collected at 4 h after intravitreal injection of PBS (1); 20 μg siCNL (2); or 20 μg siCASP2 (3) (yellow boxes indicate gel regions excised for cloning and subsequent colony sequencing); (ii) autoradiograph of the Southern blot of the gel shown in (i) after hybridization with radiolabelled RLM-RACE junction-specific probe; and (iii) example of the results of colony sequencing showing the RLM-RACE junction corresponding to the expected siCASP2 produced cleavage site. RLM-RACE was performed on RNA extracted from a pool of two retinas from eyes injected with PBS, pool of five retinas from the eyes injected with siCNL or a pool of four retinas from the eyes injected with siCASP2

Figure 4

Protection of RGC from death by siCASP2 treatment after ONC and ONT. (a) Treatment with siCNL at 7 days post-ONC resulted in 60% RGC death compared with intact and BDNF-treated controls. Increasing the dose of siCASP2 from 5 to 20 μg enhanced RGC survival with 20–35 μg per eye, resulting in 100% RGC protection compared with intact and positive controls. (c) siCASP2 enhanced RGC survival by 15% after ONT compared with the protection afforded by a control siGFP at 14 days after ONT. (b) and (d) Representative FG back-labelled RGC in retinal whole mounts demonstrate the neuroprotection promoted by siCASP2 treatment after ONC and ONT, respectively (scale bars (b) and (d)=100 μm). Note: the differences in FG labelling between (b) and (d) are due to the methodology used in the two different laboratories, please see Materials and Methods for full details