**3. Results**

### *3.1. Loss of Functional ERG in Early-Onset RD*

Patients expressing the Ter349Glu mutant rhodopsin experience a loss of photoreceptor function earlier in life, resulting in early and rapid central vision loss, when compared to other mutants of rhodopsin [9]. We tested if the Ter349Glu knock-in mouse displayed a similar early-onset (4 weeks of age) loss of visual capacity by ERG (Figure 1). Compared to +/+ mice (*n* = 5), the *Ter349Glu*/*Ter349Glu* mice (*n* = 5) showed an increase in the threshold of the dark-adapted b-wave by three orders of magnitude, with a maximum amplitude about 25% that of +/+ mice, while the maximum a-wave amplitude was only about 6% that of +/+ mice. This indicated a drastic loss of rod photoreceptor function. Interestingly, when compared to +/+ and *Ter349Glu*/*Ter349Glu* mice, *Ter349Glu*/+ mice (*n* = 4) appeared to have a gain of function with an increase in sensitivity and a decrease in response latency without significant changes in amplitudes.

### *3.2. E*ff*ects of RD-Associated Photoreceptor Loss on Retinal Vasculature and Laminar Architecture*

In cases of RP, retinal degeneration exerts e ffects on both retinal vasculature and laminar architecture in the forms of attenuated vessels and outer nuclear layer (ONL) thinning, respectively [10]. To examine the *Ter349Glu*/*Ter349Glu* retina for such abnormalities, FA and OCT were performed on 4-week-old +/+ and *Ter349Glu*/*Ter349Glu* animals in triplicate (Figure 2). Using FA, when compared to +/+ mice, *Ter349Glu*/*Ter349Glu* mice exhibited heterogeneous types and degrees of vascular abnormalities. The most common anomalies included attenuated vessels, tortuous vessels indicating hyperoxia, and reduced retinal venous and arterial vessel numbers. Using OCT, *Ter349Glu*/*Ter349Glu* mice exhibited the expected thinning of the ONL; however, at the interfaces between choroid, RPE, and rod outer segments (ROS)/ rod inner segments (RIS) layers the mice also exhibited a possible edema, likely due to loss of contacts between the RPE and the photoreceptors, associated with the lack of ROS. This edema was observed extending to varying degrees both inferiorly and superiorly to the optic nerve. Representative images were taken from the retina inferior to the optic nerve.

**Figure 1.** Electroretinogram (ERG) responses decline in mice expressing Ter349Glu rhodopsin. Using ERG to record extracellular potential differences across the retina, the electrophysiological function of +/+, *Ter349Glu*/+, and *Ter349Glu*/*Ter349Glu* mice was monitored by measuring a-, b-waves, and response latencies (Time-to-Peak, TTP) under increasing stimulus intensities (**A**,**C**,**E**). Graphs compare maximum average wave amplitudes and TTP (**B**,**D**,**F**) under dark-adapted conditions. Data analyzed using two-tailed T-test and expressed as the mean ± S.E.M. \*, *p* < 0.05; \*\*, *p* < 0.01; \*\*\*, *p* < 0.001; ns, not significant.

**Figure 2.** Ter349Glu rhodopsin knock-in mouse retina exhibits both vascular and laminar abnormalities. (**A**–**D**) Utilizing fluorescein angiography (FA), the state of the retinal vasculature of 4-week-old +/+ (**A**,**B**) and *Ter349Glu*/*Ter349Glu* (**C**,**D**) mice was examined. Abnormal phenotypes varied in severity among mice, with overall attenuated retinal vessels and tortuous retinal vessels (arrowheads) being commonplace amongs<sup>t</sup> all mice examined. (**E**,**F**) Optical coherence tomography (OCT) was used to examine the retinas of 4-week-old +/+ (**E**) and *Ter349Glu*/*Ter349Glu* mice (**F**) for architectural abnormalities. *Ter349Glu*/*Ter349Glu* mice exhibited thinning of the outer nuclear layer (ONL) and patches of varying degrees of separation among the choroid, retinal pigment epithelium (RPE), and photoreceptors (block arrow), indicative of edema. Retinal thickness (red calipers) = 240 μm (**E**) and 180 μm (**F**); ONL (green calipers) = 60 μm (**E**) and 50 μm (**F**).

### *3.3. Activated Monocytes Are Present in RD Retinas from Rhodopsin Mutant Knock-in Mice*

The retina contains resident macrophages similarly to the cortex, known as microglia, and these cells remain in the inner retinal layers under normal physiological conditions. Here, they remain in an inactivated state unless triggered by cytokine signaling or apoptotic signals [11]. Other types of leukocytes are typically not resident in the retina, and evidence of these cells in ocular tissues is indicative of retinal inflammation. Damage to retinal and choroidal vessels can allow leakage of not only blood-borne macrophages into the retina, but also cytokines, antibodies, and a plethora of other inflammatory factors [12]. Unfortunately, no method exists to differentiate between microglia and blood-borne macrophages; however, due to the observation of abnormal vasculature and retinal edemas, we chose to monitor the +/+ and *Ter349Glu*/*Ter349Glu* retinas for activated macrophages as a whole, using an antibody against F4/80 antigen, a cell surface glycoprotein expressed upon macrophage maturation (Figure 3). F4/80-positive macrophages were found in multiple animals from the *Ter349Glu*/*Ter349Glu* cohort at 12 weeks of age, with the most labeling observed in sections from *Ter349Glu*/*Ter349Glu* animals where nearly the whole retina was degenerated. Macrophages remained in the outer retina after almost total rod cell death (12 weeks). These macrophages were not observed in +/+ sections from any animal (*n* = 3 at all ages).

**Figure 3.** Retinas from retinal dystrophy (RD) mice exhibit monocyte activation. Using fluorescent immunohistochemistry, the presence of activated macrophages was examined in +/+ (**A**–**C**) and *Ter349Glu*/*Ter349Glu* (**D**–**F**) mice at 12 weeks of age. Retinal sections were labeled with anti-F4/80 antigen (green) and K62-82 (rhodopsin, red) antibodies. Nuclei were labeled with DAPI (blue). Autofluorescence in the choroid was observed in the red channel. CC, choriocapillaris; ROS, rod outer segments; RIS, rod inner segments; OPL, outer plexiform layer. Scale bars = 20 μm.

### *3.4. Activation of the Pro-Inflammatory JAK*/*STAT Pathway and Its Inhibitor SOCS3 in RD*

STAT3 is a downstream signaling partner of JAK. In inflammatory conditions, STAT3 is activated when cytokines such as IL-6, ciliary neurotrophic factor, leukemia inhibitory factor, and others bind to and activate the glycoprotein 130 (gp130) receptor [13]. Gp130 activates JAK, which in turn phosphorylates STAT proteins (pSTAT). Upon phosphorylation, pSTATs form both homo- and heterodimers, allowing nuclear entry and activation of gene transcription. The protein SOCS3 works as a negative regulator of the JAK/STAT pathway by binding the gp130 receptor and JAK together and blocking active sites involved in phosphorylation of STAT3, thus prohibiting further signal transduction [14]. In instances where inflammatory signaling needs to be slowed or stopped, SOCS3 works to perform this task. To examine the Ter349Glu knock-in mouse retina for inflammatory cytokine signaling, retinas from both +/+, *Ter349Glu*/+*,* and *Ter349Glu*/*Ter349Glu* mice were labeled with an antibody against pSTAT3 (Figure 4). We found that +/+ retinas showed no STAT3 activation across all ages, while the *Ter349Glu*/+ and *Ter349Glu*/*Ter349Glu* mice exhibited increasing activation with age, beginning at 8 weeks and 4 weeks, respectively (*n* = 3 for all ages).

Early pSTAT3 activation began in the inner nuclear layer (INL) and, by 12 weeks, extended to the RPE. The location of the activated nuclei in the INL suggested the nuclei belonged to Müller cells. Indeed, co-labeling for GFAP, a marker of gliosis in Müller cells, showed Müller cells to be pSTAT3-positive (Figure 5).

**Figure 4.** JAK/STAT pathway is activated in Ter349Glu rhodopsin knock-in mouse retina. Activation of the JAK/STAT pathway was examined using fluorescent immunohistochemistry on retinas from WT (+/+*,* **A**–**C**), Ter349Glu heterozygous (*Ter349Glu*/+, **D**–**F**), and Ter349Glu homozygous (*Ter349Glu*/*Ter349Glu*, **G**–**H**) mice. Retinal sections were treated for antigen retrieval and labeled for phosphorylated STAT3 (pSTAT3, red) and rhodopsin (green). Nuclei were labeled with DAPI (blue). Arrowheads (>), areas of JAK/STAT activation; OS, outer segments; IS, inner segments; INL, inner nuclear layer. Scale bar = 20 μm.

Since Müller cells maintain retinal homeostasis, this finding suggests they may act to respond to inflammatory cytokines, as STAT3 activation indicates the presence of pro-inflammatory cytokines in the neural retina, likely being released from activated macrophages within the retina. It should be noted that in these images, labeling for rhodopsin can be seen in the ONL even though rhodopsin is not normally localized in large amounts in this region. This is likely an artifact due to a heightened number of anti-rhodopsin epitopes following the antigen retrieval process. Labeling for SOCS3 revealed minimal expression in +/+ retinas while, similarly to pSTAT3, *Ter349Glu*/+ and *Ter349Glu*/*Ter349Glu* mice began expressing SOCS3 in the neural retina and RPE at 4 weeks of age; with time, the expression increased in *Ter349Glu*/*Ter349Glu* mice and decreased in *Ter349Glu*/+ mice (Figure 6, *n* = 3 at all ages).

**Figure 5.** Identification of Müller cell nuclei as INL centers for STAT3 activation. To assess which retinal cells exhibited STAT3 phosphorylation, labeling was performed on +/+ (**A**,**B**) and *Ter349Glu*/*Ter349Glu* (**C**,**D**) animals for glial fibrillary acidic protein (purple), a marker of astrocytes and gliotic Müller cells, rhodopsin (green), and pSTAT3 (red). Nuclei were labeled with DAPI (blue). Arrows show red nuclei surrounded by purple. Scale bar = 20 μm.

**Figure 6.** The JAK/STAT antagonist SOCS3 is expressed in Ter349Glu rhodopsin knock-in mouse retina. Using fluorescent immunohistochemistry, the expression of SOCS3, an antagonist to the JAK/STAT pathway, was examined in retinas from wild-type (WT, +/+*,* **A**–**C**), Ter349Glu heterozygous (*Ter349Glu*/+*,* **D**–**F**), and Ter349Glu homozygous (*Ter349Glu*/*Ter349Glu*, **G**–**I**) mice. Retinal sections were treated for antigen retrieval and labeled for SOCS3 (red) and rhodopsin (green). Nuclei were labeled with DAPI (blue). GCL, ganglion cell layer; arrowheads (<), SOCS3 labeling. Scale bar = 20 μm.

### *3.5. Cell-Specific Expression of PAD4 and Heightened Citrullination in Early-Onset RD*

Recent studies have shown an increase in the expression of the deiminating enzyme PAD4 and increased citrullination in the event of ocular insult of the anterior segmen<sup>t</sup> [15]. Due to the inherent ability of citrullinated proteins to cause autoimmunity [16–22] and our previous finding that PAD4 is the primary retinal PAD in mouse [23], we tested the Ter349Glu retina for changes in PAD4 expression and citrullination compared with WT retina at 10 to 12 weeks of age (Figure 7). WT retina exhibited expression of PAD4 and exhibited INL nuclear citrullination, as shown previously [23]; however, the Ter349Glu retina exhibited higher levels of citrullination, much of it spanning the entire retina. This was observed in parallel with PAD4 expression increases, especially in the photoreceptors.

**Figure 7.** Expression of PAD4 and citrullination of retinal proteins in normal and degenerated states. WT (+/+*,* **A**–**C**) and *Ter349Glu*/*Ter349Glu* (**D**–**F**) mice at 10 to 12 weeks of age were labeled for PAD4 (green) and citrullinated peptides (red). Nuclei were labeled with DAPI (blue). Arrowheads (<), areas of increased citrullination in ONL; OS/IS, photoreceptor outer and inner segments; IPL, inner plexiform layer; Scale bar = 20 μm.
