3.2.5. Prph2N229S/+ and Prph2N229S/N229

The knockin mouse that alters the N-linked glycosylation at asparagine 229 (N229S) in the D2 loop was used to study if Prph2 glycosylation plays a role in its interaction with Rom1. Heterozygous mice (*Prph2N229S*/+*)* did not have any significant changes in structure or function of the OS [75]. However, homozygous mice (*Prph2N229S*/*N229S*) displayed a late onset thinning of the outer nuclear layer (ONL) and occasional abnormal disc staking in cones and a slightly reduced photopic ERG at P180. Since Prph2 could not be glycosylated, higher order complexes were decreased and there was an increase of Prph2 and Rom-1 in the intermediate complexes [75]. Therefore, it was concluded that the glycosylation plays a major part in regulating the interaction between Prph2 and Rom1, which is critical for cone health.

Tables 1 and 2 summarize the phenotypes associated with the Prph2 knockin models, highlighting rod and cone structure, Scotopic and Photopic ERG, complex formation and protein localization. It also summarizes fundus observations, and patient's phenotypes whether mainly rod- or cone-specific defects or the combination of the two. Figures 1 and 2 show structural and functional di fferences between Prph2 heterozygous knockin mutations, while Figure 3 is a representative IF showing retinal localization of Prph2 among the models.


**Table 1.** *Prph2* mutations and correlating phenotypes.


**Table 1.** *Cont.*


**Table 2.** *Prph2* mutations and correlating phenotypes (*continued*).

### **4. Gene Therapy of** *PRPH2* **Mutations**

There is a vast variety of *PRPH2* mutations associated with autosomal dominant retinal degenerative diseases, including RP, several forms of macular dystrophy and cone rod dystrophies [29]. Gene therapy seems to be a promising approach in the treatment of those *PRPH2* associated diseases. The small size of the *Prph2* cDNA, which is approximately 1.1 kb, is advantageous since many vectors used for gene therapy can only carry DNA of a limited size. Despite the variety of feasible approaches available for the gene therapy of *Prph2*, thus far no treatment ready for clinical trials has been developed.

The expression of the WT Prph2 in the background of pathogenic *Prph2* mutations represents one of the most promising approaches. Pioneer studies using transgenic mice expressing WT Prph2 under di fferent promoters in a *Prph2*+/− or *Prph2*−/− background revealed a significant increase in both structural and functional properties of rods and cones [105,112]. Replacing Prph2 in *Prph2*−/− mice utilizing adeno-associated virus (AAV) carrying *Prph2* regulated under the rhodopsin promoter provided promising results [113]. Here the subretinal injection of the AAV resulted in a partial rescue of ROS structure as well as scotopic ERG response. A follow up study was able to show that repeated injections with the AAV resulted in an even more pronounced rescue of the phenotype with an increase in the scotopic b-wave response observed in the injected mice [114]. In this study, several time points after the injections were analyzed in order to validate whether the observed rescue was long lasting. This analyses revealed that the rescue achieved by the AAV injection was lost after 15 weeks post injection [114]. In addition to the loss of the rescue with time, the amplitude of the scotopic b-wave was significantly lower than in WT mice and the scotopic a-wave was not improved by the injection of the AAV. A reason for this is that the injection of the AAV in mice only resulted in a low transduction rate of roughly 30% [114]. An improvement in the transduction rate will thus be necessary in order to develop a viable AAV based gene therapy.

Nanoparticles (NP) represent a second promising approach in transferring WT Prph2. These particles were found to be well tolerated by the retina, even after multiple injections, and have a high DNA capacity up to 14 kb, as tested in the eye [115–120]. NP carrying full-length murine *Prph2* cDNA under the control of either rod or cone specific IRBP promoters or ubiquitous chicken beta actin promoter was injected in the retina of *Prph2*+/− mice [121]. The injection resulted in a partial rescue of OS structure and also prevented the thinning of the ONL. These rescue e ffects lasted 15 months post injection and were most pronounced near the site of injection [121]. While the NP treatment could overcome the decline in rescue overtime, the e ffect on the photoreceptor function remained limited. NP injection resulted in a small ye<sup>t</sup> insignificant improvement in the scotopic a-wave but a significant improvement in the photopic b-wave. The results obtained with the di fferent promoters used were comparable. The small benefit of the NP injection observed in the functional tests might be due to an incomplete distribution of the NPs in the retina. In line with this, the structural improvements observed after NP injection were best close to the site of injection. Improving on the distribution and uptake of the used vector in the retina, regardless if AAVs or NPs are used, seems to be a necessary next step in order to achieve a more pronounced functional rescue.

The studies above described the replacement of Prph2 in either *Prph2*+/− or *Prph2*−/− mice, thus in a scenario whereby Prph2 is either absent or haploinsu fficient. However, most patients su ffer from a dominant mutation in *PRPH2*. In order to analyze the e fficiency of treatment in a scenario closer to actual *PRPH2* related diseases, the knockin mouse models carrying a disease related *Prph2* mutation were analyzed. Both *Prph2K153*Δ/+ and *Prph2C213Y*/+ mice were crossed with a normal-Prph2-overexpressing mouse line (NMP) [76,79]. The resulting *Prph2K153*<sup>Δ</sup>/+/*NMP*+/− and *Prph2C213Y*/+/*NMP*+/− mice displayed a rescue in OS structure, protein expression levels and tra fficking, but no rescue in both rod and cone functions [76]. In both cases, the presence of the mutant protein continues to a ffect the photoreceptor function. This is due to dominant-gain-of-function e ffects caused by these mutations. These results show that, in addition to gene augmentation therapy, silencing the mutant *Prph2* allele is essential for ultimate rescue. A combined approach of short hairpin (sh)-RNA mediated knockdown and expression of a shRNA resistant protein has been performed

in models for ADRP carrying pathogenic mutations of rhodopsin [122,123]. Here, both WT and mutant Rhodopsin were knocked down by sh-RNA carried in AAV together with a sh-RNA resistant rhodopsin. The expression of the sh-RNA resistant rhodopsin following the knockdown rescued the phenotype. Studies combining the knockdown of mutant Prph2 with the expression of WT Prph2 provided first promising results thus far showing partial functional and structural rescue in mice expressing pathogenic mutations of Prph2 [124,125].

The examples above show that while progress was made in the gene therapy of *PRPH2* related diseases, there are many factors, which need to be considered for the development of a successful therapy. Further complications might arise from the fact that the functional role of Prph2 seems to vary in rods and cones. In addition to that, secondary effects of *PRPH2* mutations on the RPE and the choroid could be observed [29–31]. A high variability in the clinical phenotypes displayed by patients, even when carrying the same mutation, represents another challenge, which has to be overcome in order to treat *PRPH2* related diseases.
