3.2.2. Prph2C213Y/+ and Prph2C213Y/C213Y

The C213Y mutation is located in the D2 loop of Prph2, in the motif that mediates intramolecular and intermolecular disulfide bonds, thus responsible for the formation of Prph2/Prph2 and Prph2/Rom1 tetramers as well as for the formation of intermediate and higher order complexes [77]. The heterozygous mice (*Prph2C213Y*/+) displayed a shortened and disorganized ROS. When compared to *Prph2*+/− mice, the ROS of *Prph2C213Y*/+ mice showed a slight improvement in the stacking and alignment of the discs (Figure 1A,B) [76].

**Figure 1.** Mutations in the mouse *Prph2* gene lead to varying degrees of photoreceptor degeneration. (**A**) Representative light microscopic images from hematoxylin and eosin stained retinal sections at P30 aligned at the upper edge of the retinal pigment epithelium (RPE). (**B**) Transmission electron microscopic (TEM) images of the interface between the IS and OS of photoreceptors of the indicated genotypes. OS, outer segments; IS, inner segment; ONL, outer nuclear layer; OPL, outer plexiform layer. Scale bars: 20 μm for A and 2 μm for B. Eyes used in this study were dissected, fixed and embedded as previously described [79]. Images were captured at 40× and converted to black and white using ZEN Image Analysis software. The plastic-embedding and TEM methods were as described previously [85]. Images were adjusted and cropped using Adobe Photoshop CS5.

This is evident from the well stacked OS discs seen in some photoreceptors while others looked like whorls similar to those seen in the *Prph2*+/− (arrows in Figure 1). Scotopic ERG responses of the *Prph2C213Y*/+ mice were significantly reduced starting at P30 (Figure 2) and persist all the way to P365. The scotopic response was nearly completely abolished in *Prph2C213Y*/*C213Y* animals and the OS was almost non-existent. The photopic responses were significantly decreased in both heterozygous and homozygous animals at P30. At later time points, the photopic response decreased further in the *Prph2C213Y*/+ mice and was completely absent in the *Prph2C213Y*/*C213Y* animals. A key finding in this study was that the *Prph2C213Y*/+ mice, that represent the genotype present in patients, showed better retinal structure despite the fact that both of rod and cone ERG responses were reduced when compared to *Prph2*+/− animals.

**Figure 2.** Mutations in *Prph2* hinder OS function assessed via scotopic and photopic electroretinograms (ERGs) at P30. Full-field ERGs were recorded under scotopic and photopic conditions. Shown are representative ERG waveforms from the indicated genotypes at P30. Full-field ERG tests were performed as previously described [85]. After overnight dark adaptation, mice were anesthetized and their pupils dilated. ERGs were recorded with a UTAS system (LKC, Gaithersburg, MD, USA). Waveforms were exported into GraphPad software to obtain wave traces and then exported into Photoshop using a uniform scale.

A key requirement for the function of Prph2 is its ability to interact with Rom1 and to form oligomeric complexes [18,27,74]. Co-immunoprecipitation (co-IP) experiments using retinal lysate from *Prph2C213Y*/*C213Y* mice revealed the inability of Prph2C213Y to interact with Rom1. The homomeric interaction of Prph2 was also reduced in these mice as evident from reduced ability of Prph2C213Y to form intermediate and higher order complexes as determined by sucrose gradient velocity sedimentation [76]. The reduction in oligomeric complexes offers an explanation for the functional decline observed in both *Prph2C213Y*/+ and *Prph2C213Y*/*C213Y* mice and no abnormal high molecular weight aggregates were observed under non-reducing conditions. *Prph2C213Y*/+ retina retains some Prph2 in the IS, shown by immunofluorescence (IF) staining of Prph2 (green and arrows in Figure 3) and IS maker syntaxin 3B (STX3B) (red, Figure 3). However, *Prph2C213Y*/*C213Y* lacks the ability to form complexes which leads to

complete retention of Prph2 in the inner segments (IS), and perinuclear region while a smaller amount of Rom1 was retained in the IS [76].

Patients carrying the C213Y mutation in *PRPH2* display a butterfly-shaped pattern/macular dystrophy, while a rod dominant RP phenotype is absent [29,108–110]. The knockin mouse model for the C213Y displays functional defects in both rods and cones. While the defects in the cones are in line with the defects observed in the cone rich macula of the patients, the defects in the rods displayed by the model are not. The di fference between phenotypes in patients and the mouse model may be due to the fact that the mouse retina consists almost exclusively of rod photoreceptors. Since the murine retina lacks a macula, it was not possible to reproduce the butterfly shaped macular dystrophy in the mouse model. One interesting finding in the Prph2C213Y mouse model was the observation of a yellow flecking in the fundus of both *Prph2C213Y*/+ and *Prph2C213Y*/*C213Y* mice at P180. This phenotype mimics funduscopic anomalies found in patients carrying the C213Y mutation [110].

Gene supplementation was performed by crossing a WT Prph2 overexpressing mouse line (NMP) onto hetero- (*Prph2C213Y*/+) or homozygous (*Prph2C213Y*/*C213Y*) mice. The defects observed in the protein tra fficking as well as in OS structure were rescued in *Prph2C213Y* mice. A rescue on the functional level however could not be observed [76]. These results show that the presence of the mutant protein has a detrimental e ffect on photoreceptor function.

**Figure 3.** Mutated Prph2 protein tra ffics to the OS while a small pool is retained in the inner segment. Retinal sections at P30 from the indicated genotypes were probed with antibodies against Prph2 (green) and syntaxin 3B (STX3B) (red). Arrows indicate regions of mislocalization of Prph2. OS, outer segments; IS, inner segment; ONL, outer nuclear layer. Scale bar: 20 μm. Primary antibodies used for immunostaining were polyclonal antibody against Prph2 C-terminus (Prph2-CT) [22] and monoclonal antibody against STX3B [111] (inner segmen<sup>t</sup> marker) diluted at (1:1000). AlexaFluor conjugated secondary antibodies (Alexa 488 Rabbit and Alexa 555 Mouse, Life Technologies/ThermoFisher) were used at a dilution of 1:1000 for 2 hours at room temperature. Images were captured on a ZEISS Confocal LSM 900 microscope equipped with a Zeiss Axiocam (Zeiss, Jena, Germany) using a 63× (oil, 1.4 NA) objective. Images were then processed using ZEN Image Analysis software (Zeiss, Jena, Germany). All images shown are orthogonally projected from an eight slice confocal z-stack.
