Nrl:CreERT2 mouse model to induce mosaic gene expression in rod photoreceptors

Abstract

Photoreceptors are sensory neurons that capture light within their outer segment, a narrow cylindrical organelle stacked with disc-shaped membranes housing the visual pigment. Photoreceptors are the most abundant neurons in the retina and are tightly packed to maximize the capture of incoming light. As a result, it is challenging to visualize an individual cell within a crowded photoreceptor population. To address this limitation, we developed a rod-specific mouse model that expresses tamoxifen-inducible cre recombinase under the control of the Nrl promoter. We characterized this mouse using a farnyslated GFP (GFPf) reporter mouse and found mosaic rod expression throughout the retina. The number of GFPf-expressing rods stabilized within 3 days post tamoxifen injection. At that time, the GFPf reporter began to accumulate in basal disc membranes. Using this new reporter mouse, we attempted to quantify the time course of photoreceptor disc renewal in WT and Rd9 mice, a model of X-linked retinitis pigmentosa previously proposed to have a reduced disc renewal rate. We measured GFPf accumulation in individual outer segments at 3 and 6 days post-induction and found that basal accumulation of the GFPf reporter was unchanged between WT and Rd9 mice. However, rates of renewal based on the GFPf measurements were inconsistent with historical calculations from radiolabeled pulse-chase experiments. By extending GFPf reporter accumulation to 10 and 13 days we found that this reporter had an unexpected distribution pattern that preferentially labeled the basal region of the outer segment. For these reasons the GFPf reporter cannot be used for measuring rates of disc renewal. Therefore, we used an alternative method that labels newly forming discs with a fluorescent dye to measure disc renewal rates directly in the Rd9 model and found it was not significantly different from WT. Our study finds that the Rd9 mouse has normal rates of disc renewal and introduces a novel Nrl:CreERT2 mouse for gene manipulation of individual rods.

Keywords: NRL; RD9; disc renewal; inducible; outer segment; photoreceptor; retinal degeneration.

Figure 1

Tamoxifen-inducible Nrl:CreERT2 mouse. (A) Gene diagram for CreERT2 insertion within the Nrl locus. (B) Agarose gel showing bands produced by genotyping PCRs on isolated genomic DNA from Nrl:CreERT2^+/+, Nrl:CreERT2^+/−, and Nrl:CreERT2^−/− mice.

Figure 2

Nrl:CreERT2 mouse displays a mosaic pattern of reporter gene expression in rod photoreceptors. (A) Representative retinal cross-section from an iCre75/GFPf mouse stained with anti-GFP antibody and WGA. Nuclei are counterstained with Hoechst. (B) Representative image of a retinal whole-mount from an iCre75/GFPf mouse stained with anti-GFP antibody and WGA. (C) Membrane and soluble retinal fractions from C57Bl6/J and iCre75/GFPf mice immunoblotted (IB) for GFP. (D,E) Nrl:CreERT2/GFPf mice were injected 3 consecutive days with either tamoxifen or corn oil. Eyes were collected 2 weeks post-injection. (D) Retinal cross-sections stained with anti-GFP antibody and WGA. Nuclei are counterstained with Hoechst. (E) Retinal whole-mount stained with anti-GFP antibody and WGA. Scale Bars, 10 μm. Here and in all figures: outer segment (OS), inner segment (IS), nuclei (N), and synapse (S).

Figure 3

The number of rod photoreceptors expressing GFPf in the Nrl:CreERT2 mouse stabilizes 3 days post tamoxifen induction. (A) Representative retinal whole mounts from Nrl:CreERT2/GFPf stained with anti-GFP antibody showing GFP expression in the rod synapse at 2, 3, and 5 days post-injection (DPI). Scale Bars, 10 μm. (B) Bar graph plotting the number of GFP positive rods in a 2,500 μm² retinal area. Mean for each timepoint is displayed above the bar. Error bars represent S.D. A one-way ANOVA was preformed: ****p < 0.0001, ***p = 0.0003, ns p = 0.2172. (C) Representative retinal cross-sections from Nrl:CreERT2/GFPf mice at 2, 3, and 5 DPI stained with anti-GFP antibody and WGA. Nuclei are counterstained with Hoechst. Scale Bars, 10 μm.

Figure 4

Investigating whether the Nrl:CreERT2/GFPf model can be used to measure rates of outer segment renewal. (A) Retinal cross-sections from WT and Rd9 mice on the Nrl:CreERT2/GFPf background at 3 and 6 DPI stained with anti-GFP antibody and WGA. Nuclei are counterstained with Hoechst. Scale bars, 10 μm. (B) Representative images showing the analysis used to measure the length of GFPf accumulation in outer segments. (C) Bar graph plotting stratified measurements of GFPf accumulation in outer segments. Mean values are displayed above the bars. Error bars represent S.D. Linear mixed modeling: 3 DPI ns p = 0.531 and 6 DPI ns p = 0.896. (D) Retinal cross-sections from WT mice on the Nrl:CreERT2/GFPf background at 10 and 13 DPI stained with anti-GFP antibody and WGA. Nuclei are counterstained with Hoechst. Scale bars, 10 μm. (E) Bar graph plotting stratified measurements of GFPf accumulation in outer segments. Mean values are displayed above the bars. Error bars represent S.D. Linear mixed modeling: ns p = 0.704.

Figure 5

Rate of new disc addition is not affected by loss of RPGR-ORF15. (A) Images showing representative isolated outer segments from WT and Rd9 mice 5 days after CF-568 injection. Differential interference contrast (DIC) overlay on left and CF-568 alone on right. Scale bar, 5 μm. (B) Bar graph plotting the distance between the outer segment base and the band of incorporated CF-568. Mean values are displayed above the bar. Error bars represent S.D. Unpaired Student’s two-tailed t-test: ns p = 0.9543.