Increased proteasomal activity supports photoreceptor survival in inherited retinal degeneration

Abstract

Inherited retinal degenerations, affecting more than 2 million people worldwide, are caused by mutations in over 200 genes. This suggests that the most efficient therapeutic strategies would be mutation independent, i.e., targeting common pathological conditions arising from many disease-causing mutations. Previous studies revealed that one such condition is an insufficiency of the ubiquitin-proteasome system to process misfolded or mistargeted proteins in affected photoreceptor cells. We now report that retinal degeneration in mice can be significantly delayed by increasing photoreceptor proteasomal activity. The largest effect is observed upon overexpression of the 11S proteasome cap subunit, PA28α, which enhanced ubiquitin-independent protein degradation in photoreceptors. Applying this strategy to mice bearing one copy of the P23H rhodopsin mutant, a mutation frequently encountered in human patients, quadruples the number of surviving photoreceptors in the inferior retina of 6-month-old mice. This striking therapeutic effect demonstrates that proteasomes are an attractive target for fighting inherited blindness.

Fig. 1

Proteasome composition of the mouse retina. a The molar ratio among 20S, 19S, and 11S proteasomal components determined by quantitative mass spectrometry. Data are shown as mean ± SEM; n = 3. b Fractionation of proteasome components in retinal extracts from 1-month-old mice (200 μg total protein) by size-exclusion chromatography on a Superose-6 column. Proteins in 0.5 ml fractions were probed by western blotting using antibodies against β1 subunit of the 20S proteasome core, PSMD11 subunit of the 19S proteasome cap, and PA28α subunit of the 11S cap. Data are taken from one of the four similar experiments. c The distribution of β1, PSMD11, and PA28α in 20 μm serial tangential sections throughout the entire WT mouse retina. Each section was solubilized in 30 μl SDS-PAGE sample buffer for analysis. Proteins were visualized by western blotting using the ECL technique. Rhodopsin was used as a photoreceptor outer segment marker; phosducin was used as a marker of the entire photoreceptor layer. Data are taken from one of two similar experiments. A representative retinal cross-section is shown below western blot panes; the corresponding position of the photoreceptor cells is illustrated by a cartoon

Fig. 2

Characterization of PA28α and PSMD11 overexpressing (OE) mice. a Western blots of proteasomal subunits in retinal lysates containing 30 μg total protein. Bands were visualized using the LiCor Odyssey imaging system. Each protein was analyzed in at least 3 pairs of 1-month-old WT and overexpressing animals. b Retinal morphology of 3-month-old overexpressing and WT mice. Retinas were embedded in plastic, 1 μm cross-sections were stained by toluidine blue and analyzed by light microscopy. Data are taken from one of the five similar experiments; scale bar: 20 μm. c Chymotrypsin-like proteasomal activity in retinal extracts from 1-month-old overexpressing and WT mice; measurements were performed in the presence or absence of ATP, as indicated. The number of measurements was 10, 7, and 5 for WT, PA28α overexpressing, and PSMD11 overexpressing mice, respectively. The data are shown as mean ± SEM; p values determined across individual preparations are indicated in the text. d Fractionation of proteasomal components in retinal extracts from 2-month-old overexpressing and WT mice by size-exclusion chromatography on a Superose-6 Increase column. Proteins in 0.5 ml fractions were probed by western blotting using antibodies against the β1 subunit of the 20S proteasome core, PSMD11 subunit of the 19S proteasome cap, and PA28α subunit of the11S cap. Data are taken from one of the three similar experiments

Fig. 3

Overexpression of PA28α or PSMD11 does not affect ubiquitin-dependent proteasome activity. The decrease in fluorescence polarization representing the degradation of polyubiquitinated protein (left) or tetraubiquitinated peptide (right) substrates was monitored in retinal extracts from 1-month-old mice of indicated genotypes. Data points represent individual measurements collected every 20 s in three technical replicates performed with the same pair of retinal extracts. Data were fitted with single exponents (solid lines); shaded areas show 95% confidence intervals of the fits. The data are taken from one of the three similar independent experiments. In all cases, R² values for exponential fits were at least 0.96 for the polyubiquitin and 0.99 for the peptide substrates

Fig. 4

Overexpression of PA28α or PSMD11 improves photoreceptor survival in P23H mice. a Spider diagrams representing the number of photoreceptor nuclei in 100 μm segments of the inferior and superior retina counted at various distances from the optic nerve head. Data collected from 3- and 6-months-old mice are shown as mean ± SD. The number of eyes analyzed at 3 months was: P23H—6, P23H/PA28α OE—6, P23H/PSMD11 OE—6, WT—4; the number of mice analyzed at 6 months was: P23H—5, P23H/PA28α OE—4, P23H/PSMD11 OE—6, WT—4. Data points for which the difference in the number of nuclei between treated and untreated animals was statistically significant (p < 0.05) are shown as filled circles and those for which statistical significance was not achieved as open circles. b Representative images of inferior retina cross-sections at the 1 mm distance from the optic nerve head. Mouse genotypes are indicted above the panels. Scale bar: 50 μm. See Supplementary Figs. 6 and 7 for images of representative cross-sections through the entire retinas. c The total number of nuclei in all eight 100 μm retinal segments presented in the spider diagrams of a. Data are shown as mean ± SEM

Fig. 5

Overexpression of PA28α improves retinal function in P23H mice. a, b Response amplitudes of ERG a- and b-waves evoked by light flashes of increasing intensity were measured in PA28α overexpressing P23H mice, their control P23H littermates, and WT animals at the ages of 3 (a) and 6 (b) months. The number of eyes analyzed at 3 months was: P23H—6, P23H/PA28α OE—6, WT—3; the number of mice analyzed at 6 months was: P23H—3, P23H/PA28α OE—4, WT—4. Data were averaged and fitted using a single (for the a-wave) or double (for the b-wave) hyperbolic function. Error bars represent SEM. On the right are shown representative ERG recordings from animals of each genotype evoked by flashes of three indicated intensities representing scotopic, mesopic, and photopic light intensity ranges

Fig. 6

Overexpression of PA28α improves photoreceptor survival in Gγ₁^−/− mice. a Spider diagrams representing the number of photoreceptor nuclei in 100 μm segments of the inferior and superior retina at various distances from the optic nerve head. Data were collected from five mice at 3 months of age and shown as mean ± SD. Data points for which the difference in the number of nuclei between treated and untreated animals was statistically significant (p < 0.05) are shown as filled circles and those for which statistical significance was not achieved as open circles. b The total number of nuclei in all eight 100 μm retinal segments represented in spider diagrams of a. Data are shown as mean ± SEM. c Representative images of inferior retina cross-sections at the 1 mm distance from the optic nerve head; scale bar: 50 μm. Mouse genotypes are indicted above the panels. See Supplementary Fig. 9 for images of representative cross-sections through the entire retinas

Fig. 7

Overexpression of PA28α or PSMD11 does not affect accumulation of the Ub^G76V-GFP reporter. The Ub^G76V-GFP reporter was detected in retinal lysates from 1-month-old mice of indicated genotypes (30 μg total protein/lane) using an anti-GFP antibody; Hsc-70 was used as a loading control. The band representing the non-proteolyzed non-fluorescent GFP product co-accumulating with this reporter in cells suffering from proteasomal insufficiency^, is labeled as xGFP. Data are taken from one of the four similar experiments