Figure 1.
Rhodopsin expression in mice with various numbers of rhodopsin genes.
(A) Northern blot analysis of rhodopsin mRNA from various mouse lines. Copy numbers of wild type rhodopsin and P23H-rhodopsin genes are indicated below each lane, where mRho 0 is mRho−/−, mRho 1 is mRho+/−, mRho 2 is mRho+/+, and mRho 0 NHRE 2 is mRho−/−NHRE+/+. Quantification of the 18S and 26S rRNAs bands from the agarose gels prior to transfer (shown as a negative below the Northerns) served as loading controls. Six retinas were used to prepare each mRNA sample. Northern images were obtained on a PhosphorImager. Sizes are indicated in kb.
Table 1.
Rhodopsin expression from mRho and NHRE alleles.
Table 2.
Crosses to generate mice with different numbers of rhodopsin genes.
Figure 2.
Hematoxylin and eosin stained images of retinal sections from mice with different numbers of wild type rhodopsin genes.
Genotypes are indicated at left, where 1 Rho is mRho+/−, 2 Rho is mRho+/+, and 3 Rho is mRho+/+NHRE+/0. Ages in months are indicated at the top. Retinal layers are abbreviated at right, where ROS is rod outer segment, RIS is rod inner segment, ONL is outer nuclear layer, OPL is outer plexiform layer, INL is inner nuclear layer, and IPL is inner plexiform layer.
Figure 3.
Measurements of retinal degeneration.
(A) Counts of nuclei per column of nuclei in the outer nuclear layers (ONL) of retinas from mice with different numbers of wild type rhodopsin genes. Genotypes are abbreviated on the curves: 0 Rho is mRho−/− (filled squares), 1 Rho is mRho+/− (open circles), 2 Rho is mRho+/+ (filled circles), 3 Rho is mRho+/+NHRE+/0 (open triangles), and 4 Rho is mRho+/+NHRE+/+ (open squares). (B) Counts of nuclei per column of nuclei in the ONL of retinas from mice with the P23H-rhodpsin transgene. Genotypes are abbreviated on the curves: 1 Rho+P23H is mRho+/−P23H+/0 (open circles), 2 Rho+P23H is mRho+/+P23H+/0 (filled circles), and 3 Rho+P23H is mRho+/+NHRE+/0P23H (open triangles). We counted 60–100 columns of nuclei for multiple areas within each retina (3 eyes from 3 individual mice per genotype per timepoint) and averaged them for each time point. Error bars indicate standard error of the mean. Curves were fit to an exponential decay curve, allowing for a plateau value. Exponentials are from non-linear curve fitting using the Marquardt-Levenberg method in Origin, with weighting by 1/variance. The 3 Rho curve-fitting also included data from 3 mice at 8.5 months, which had 6.7 nuclei, and the 4 Rho curve-fitting included data from 1 mouse at 10 months, which had 1.5 nuclei (not shown).
Figure 4.
Electron micrographs of rod cells from 1-month old mice with different numbers of wild type rhodopsin genes.
Genotypes are abbreviated at the top: 1 Rho is mRho+/−, 2 Rho is mRho+/+, 3 Rho is mRho+/+NHRE+/0, and 4 Rho is mRho+/+NHRE+/+. Arrows indicate incisures. The top row of images is at 700X magnification, the middle row is at 4400X, and the bottom row is at 12000X, with scale bars indicating 11 µm, 1.7 µm, and 0.6 µm, respectively.
Table 3.
Measurements of rod outer segments in P30 mice with increasing rhodopsin expression.
Figure 5.
Volumes of rod outer segments from 1-month old mice.
Volumes were calculated from measurements of width and length (Table 3), assuming that the rod outer segments are cylinders (V = πr2h). Genotypes are abbreviated below the figure: 1 Rho is mRho+/−, 2 Rho is mRho+/+, 3 Rho is mRho+/+NHRE+/0, 4 Rho is mRho+/+NHRE+/+, 1Rho+P23H is mRho+/−P23H+/0, 2 Rho+P23H is mRho+/+P23H+/0, and 3 Rho is mRho+/+NHRE+/0P23H+/0. Error bars represent the propagated error from the width and length measurements in Table 3, which were used in the calculation of volume. P values were determined for mice that differed by 1 copy of rhodopsin gene, using Student’s t-test. ** indicates P<0.001.
Figure 6.
Hematoxylin and eosin stained images of retinal sections from mice with the P23H-rhodopsn transgene in combination with different numbers of wild type rhodopsin genes.
Genotypes are indicated at left, where 1Rho+P23H is mRho+/−P23H+/0, 2 Rho+P23H is mRho+/+P23H+/0, and 3 Rho is mRho+/+NHRE+/0P23H+/0. Ages in months are indicated at the top. Retinal layers are abbreviated at right, where ROS is rod outer segment, RIS is rod inner segment, ONL is outer nuclear layer, OPL is outer plexiform layer, INL is inner nuclear layer, and IPL is inner plexiform layer.
Figure 7.
Electron micrographs of rod cells from 1-month old mice with the P23H transgene in combination with different numbers of wild type rhodopsin genes.
Genotypes are abbreviated at the top: 1 Rho+P23H is mRho+/−P23H+/0, 2 Rho+P23H is mRho+/+P23H+/0, and 3 Rho+P23H is mRho+/+NHRE+/0P23H+/0. The top row of images is at 700X magnification, the middle row is at 4400X, and the bottom row is at 12000X, with scale bars indicating 11 µm, 1.7 µm, and 0.6 µm, respectively.
Figure 8.
Distribution and quantification of GFP fluorescence in mice expressing hRho-GFP or P23H-hRho-GFP in the presence of 1 or 2 copies of wild type rhodopsin.
(A) Retinas from age-matched hRho-GFP and P23H-hRho-GFP mice were sectioned and stained with DAPI. Sections were imaged using the heterozygous (hRho-GFP/mRho) sections to establish gain and laser power settings that gave limited numbers of saturated pixels in the region of the rod outer segments. These settings are referred to as “Low” and were also used to image retinas from hRho-GFP/mRho NHRE+/0 mice. Due to the lower amounts of P23H-hRho-GFP rhodopsin, retinas from P23H-hRho-GFP/mRho and P23H-hRho-GFP/mRho NHRE+/0 were imaged by increasing the laser power 5-fold, which is referred to as “High.” Immunostaining with 1D4 antibodies, shown in red, identifies the location of wild type rhodopsin. The scale bars are 10 µm. (B) Quantification of average GFP fluorescence intensity. The first and third images from the top and bottom rows in (A) were quantified using ImageJ and expressed relative to hRho-GFP/mRho retinas, which were defined as 100%. (C) Quantification of GFP localization. The amounts of GFP fluorescence the rod outer segments (ROS), rod inner segments (RIS), and outer nuclear layer (ONL) in images like those in the first and third columns of (A) were quantified using ImageJ. For the results displayed in (B) and (C), 5 different sections were examined from each of 4 hRho-GFP/mRho mice, each of 4 hRho-GFP/mRho NHRE+/0 mice, each of 8 P23H-hRho-GFP/mRho mice, and each of 8 P23H-hRho-GFP/mRho NHRE+/0 mice. Retinal sections were derived from various regions across the retina–excluding the periphery and the optic nerve. In all cases, error bars indicate standard deviations. We tested for statistical significance using a multivariate ANOVA with a Least Squares Difference post-hoc analysis. NS, not significant; *, P<0.05; **, P<0.001; ***, P<0.001.