Therapeutic potential of archaeal unfoldase PANet and the gateless T20S proteasome in P23H rhodopsin retinitis pigmentosa mice

Abstract

Neurodegenerative diseases are characterized by the presence of misfolded and aggregated proteins which are thought to contribute to the development of the disease. In one form of inherited blinding disease, retinitis pigmentosa, a P23H mutation in the light-sensing receptor, rhodopsin causes rhodopsin misfolding resulting in complete vision loss. We investigated whether a xenogeneic protein-unfolding ATPase (unfoldase) from thermophilic Archaea, termed PANet, could counteract the proteotoxicity of P23H rhodopsin. We found that PANet increased the number of surviving photoreceptors in P23H rhodopsin mice and recognized rhodopsin as a substate in vitro. This data supports the feasibility and efficacy of using a xenogeneic unfoldase as a therapeutic approach in mouse models of human neurodegenerative diseases. We also showed that an archaeal proteasome, called the T20S can degrade rhodopsin in vitro and demonstrated that it is feasible and safe to express gateless T20S proteasomes in vivo in mouse rod photoreceptors. Expression of archaeal proteasomes may be an effective therapeutic approach to stimulate protein degradation in retinopathies and neurodegenerative diseases with protein-misfolding etiology.

Fig 1. PAN^et increases rod survival in P23H rhodopsin mice.

In all plots, PAN^et (-) Rho^P23H/WT (black circles) and PAN^et (+) Rho^P23H/WT (white circles) were compared at the indicated ages. Significance was determined using t-tests, showing two-tailed p-value or asterisk for P <0.05. A, Visual responses of dark-adapted rods were recorded using a Celeris (Diagnosis LLC) rodent ERG system. The amplitude of elicited ERG a-wave is shown as a function of flash intensity (SEM, n = 8–10). Responses of wild type 129E mice at 60 days of age (open triangles) are shown for comparison. Each dataset was fitted with a simple rectangular hyperbola with two parameters using SigmaPlot 13 software. B, Panel: Ocular cross-section stained with hematoxylin and eosin showing areas 1–7; Graphs: The number of stacked photoreceptors’ nuclei (ONL thickness) is plotted against the area number (SEM, n = 3). C, Representative images of retinal cross-sections showing a portion of photoreceptor nuclei identified and color-coded by ONLyzer at different stages of retinal degeneration (top to bottom: age 30, 120 and 240 days); RGC − retinal ganglion cells, INL − inner nuclear layer, ONL − outer nuclear layer, RPE–retinal pigment epithelium. D, Photoreceptor nuclei counts in the entire retinal cross-sections measured by ONLyzer. Morphologically intact cross-sections adjacent to the optic nerve, such as the one shown in B, were analyzed, generating the total nuclei count within the ONL. Distribution of numeric data with the 50% range, median value, and SD. E, Normalized levels of rhodopsin transcript determine by quantitative real-time PCR at 21 days of age (SEM, n = 4).

Fig 2. PAN^et increases proteasome load in P23H rhodopsin mice.

A, GFP pulldowns from retinas of wild-type mice expressing Ub^G76V-GFP reporter and negative controls were analyzed by Western blotting. Bands corresponding to Ub-GFP and its GFP fragment are indicated. B, Consecutive GFP and anti-FLAG pulldowns from the retinas of mice of the indicated genotype were analyzed by Western blotting. In all blots, 37kD and 28kD bands corresponding to Ub-GFP, and GFP fragment, respectively, are indicated. PAN^et containing FLAG and HA epitope tags runs as 55kD band. C, Relative abundance of the combined Ub-GFP and GFP signals in retinas of PAN^et(+) mice of P23H rhodopsin and wild-type backgrounds. Bar height shows mean percent value in PAN^et(+) retinas compared to PAN^et(-) retinas, with a combined Ub-GFP and GFP signal of 124.6±3.3% for P23H rhodopsin mice and 83.5±4.0% for wild-type mice (SEM, n = 6). Shown two-tailed p-values are from paired t-tests between PAN^et(-) and PAN^et(+) groups. D, Normalized levels of Ub^G76V-GFP transcript determined by quantitative real-time PCR (SEM, n = 6). In all plots, significance was determined using paired t-tests, showing two-tailed p-value or asterisk for P <0.05.

Fig 3. Proteolytic degradation of rhodopsin by recombinant archaeal PAN-T20S.

A, GFP-ssrA protein is degraded by recombinant PAN-T20S resulting in decreasing fluorescence over time. Data were normalized to control without PAN-T20S (see Materials and Methods for details). Error bars indicate SEM (n = 3). B, Representative Western blotting showing degradation of rhodopsin isolated from mouse retina by PAN-T20S. C, Time curve of rhodopsin degradation. Density of monomeric rhodopsin band (~35kD) at each time point was quantified. Each dataset was fitted with a single exponential decay function with three parameters using global curve fit. Error bars indicate SEM (n = 6). Data were compared using paired t-test. Asterisk indicates two-tailed p-value <0.05. D, Spread of the data used in C with black and white circles representing normalized rhodopsin signal in the absence- and presence of PAN-T20S, respectively. Bar graphs show mean value with standard error at the indicated time points. Data at 3 min used for normalization is omitted from the plot.

Fig 4. Expression of gateless archaeal 20S proteasome in HEK293 cells.

A, Side- and top-view of T20S proteasome from thermophilic archaea, Thermoplasma acidophilum comprised of two α-rings and two β-rings. Seven identical α or β subunits within each ring are shown in different colors (PDB ID 8F7K) B, T20S-αΔN construct and the encoded sequences of the epitope-tagged β subunit (PsmB-6x-His) and the N-terminally truncated α subunit (^Δ2-11PsmA). C, Ni-NTA and anti-6x-His tag pulldowns from HEK293 cells, transfected with empty vector (1) or CMV-T20S-αΔN (2), were analyzed by Western blotting with indicated antibodies. D, Chymotrypsin-like peptidase activity of T20S-αΔN, captured with Ni-NTA resin from HEK293 cells, transfected with empty vector (1) or CMV-T20S-αΔN (2), was determined with Suc-LLVY-AMC, a fluorogenic peptide substrate (SEM, n = 3). E, Negative stain TEM of 6x-His tag antibody pulldowns from HEK293 cells transfected with empty vector (1) or CMV- T20S-αΔN (2) and global average side- and top-view of T20S-αΔN.

Fig 5. T20S-αΔN is nontoxic for rod photoreceptors.

A, Experimental design: 129E and PAN^et (+) mice were subretinally injected with AAV-T20S-αΔN vector, allowed to recover for 1- and 3 months, and analyzed by ERG and Western blotting. B, T20S-αΔN detected by anti-6x-His pulldown and Western blotting in retinal extracts from noninjected and AAV-T20S-αΔN-injected 129E mice 1-month post-injection. C, T20S-αΔN detected by anti-6x-His pulldown and Western blotting (left) and PAN^et detected by anti-FLAG pulldown and Western blotting (right) in retinal extracts from noninjected 129E mice and AAV-T20S-αΔN-injected PAN^et mice 3-month post-injection. D, Comparison of visual responses one month and three months post-injection. Electroretinographic responses were recorded using a UTAS BigShot (LKC Technologies) rodent ERG system. The amplitude of elicited ERG a-wave is plotted as a function of flash intensity (SEM, n = 12). Each dataset was fitted with a simple rectangular hyperbola with two parameters. Inset: Distribution of numeric data with the 50% range, median value, and SD for the criterion flash of 0.4 cd s^-1 m^-2. The significance was determined by paired t-tests, showing two-tailed P value.