|Molecular Vision 2005;
Received 26 January 2005 | Accepted 29 March 2005 | Published 27 May 2005
Constant light-induced retinal damage and the RPE65-MET450 variant: Assessment of the NZW/LacJ mouse
Michael Danciger,1 Haidong
Yang,2 Lisa Handschumacher,1 Matthew M. LaVail2
1Department of Biology, Loyola Marymount University, Los Angeles, CA; 2Beckman Vision Center, UCSF School of Medicine, San Francisco, CA
Correspondence to: Michael Danciger, Department of Biology, Research Annex, Loyola Marymount University, Los Angeles, CA, 90045-2659; Phone: (310) 338-7337; FAX: (310) 338-7833; email: firstname.lastname@example.org
Purpose: In a previous constant light-induced retinal damage (CLD) quantitative genetics study between the albino C57BL/6J-c2J (B6al) and BALB/c mouse strains, we identified a very strong and highly significant quantitative trait locus (QTL) on distal Chr 3 that we associated with a variant of the Rpe65 gene. The B6al strain carries the MET450 variant of RPE65 and is resistant to CLD while the BALB/c strain carries the LEU450 variant and is sensitive. Since then, we have discovered that the NZW/LacJ (NZW) albino mouse strain is sensitive to CLD but carries the MET450 variant of RPE65. The purpose of this study was to determine if the NZW mouse disproves the hypothesis that the MET450 variant of RPE65 protects the mouse retina against constant light-induced retinal damage.
Methods: F2 progeny were bred from an intercross between the NZW/LacJ and B6al mouse strains. After a prolonged exposure to moderate constant light, F2 mice were phenotyped for retinal outer nuclear layer thickness as the quantitative trait. A subset of 156 of the 201 F2 mice was genotyped for a set of markers spanning the genome, and any marker with a significant association with the quantitative trait was genotyped in the remaining 45 F2s. Data were analyzed for QTL by the Map Manager QTX software.
Results: No QTL was identified at distal Chr 3, although several QTL on Chrs 1 (two), 10, 13, 14, 16, and X were detected. One QTL on middle Chr 1 (LOD 5.22) mapped to the same location of a QTL (LOD 6.8) in a previous intense, short exposure light-induced retinal damage study conducted with an intercross between the 129S1/SvImJ and BALB/c strains. QTL on Chrs 1 (distal), 10, and 14 also appeared in other retinal damage quantitative genetics studies. Three pairs of genes exhibited significant epistatic effects. Two of the pairs involved synergistic interactions between NZW and B6al alleles, and the third between two B6al alleles.
Conclusions: If another gene besides Rpe65 was responsible for the QTL in the original BALB/c x B6al study, and the NZW mouse carried a light sensitive allele of this gene, a QTL should have been present in this study. Since a QTL on Chr 3 was not found, the hypothesis that RPE65-MET450 protects the retina from constant light-induced damage is left intact. The explanation for the NZW mouse being sensitive to constant light while carrying the RPE65-MET450 variant is that other light sensitive QTL (gene alleles) negate the protective effect.
In a previous quantitative genetics study of constant light-induced retinal damage (CLD) between the BALB/c and C57BL/6J-c2J (B6al) albino strains , a highly significant and strong quantitative trait locus (QTL) was identified on distal Chr 3. A trinucleotide repeat in the Rpe65 gene polymorphic between the two strains was typed and found to segregate at the peak of the QTL. Sequencing of Rpe65 revealed a difference in the coding region predicting a leucine at residue 450 for BALB/c and a methionine for B6al. Our hypothesis was that the MET450 was converted to methionine sulfoxide under the oxidizing conditions of constant light, rendering the RPE65 protein nonfunctional. Since RPE65 is integral to the recycling of the chromophore 11-cis-retinal, rhodopsin would not be regenerated and its absence would protect the rods from damage by shutting down the phototransduction cascade. The finding that photoreceptors in Rpe65 -/- mice were greatly protected from light-induced structural damage  supported the hypothesis. LEU450 is present in RPE65 in all the other animals we found in the NCBI sequence database including the African green monkey, chicken, dog, cow, rat, human, tiger salamander, zebra fish, and Japanese fire belly newt. In addition, we sequenced the variant in several albino mouse strains shown to be sensitive to CLD previously  and found LEU450 in all but one of them, the NZW/LacJ strain, which carries the MET450 variant of RPE65 yet is sensitive to CLD.
There are at least three explanations for this "NZW phenomenon". The first is that the Rpe65 gene variant is not responsible for the retinal light damage difference between BALB/c and B6al but is only a polymorphism segregating with a nearby gene that is responsible for the effect. The NZW and BALB/c alleles of this gene would contribute relative susceptibility to CLD while the B6 allele would contribute relative resistance. The second explanation is that there may be an additional change in the NZW Rpe65 gene that suppresses the protective effect of the MET450 variant. The third explanation is that there may be other NZW susceptibility gene alleles that overcome the protective effect of the MET450 variant.
To test these possibilities, we performed a constant light-induced retinal damage quantitative genetics study based on a small intercross between the B6al and NZW albino strains. If either the first or second explanation were true, we would expect a strong and highly significant QTL at distal Chr 3 with a B6al protective allele. We found no such distal Chr 3 QTL.
NZW/LacJ and C57BL/6J-c2J albino mice were originally purchased from Jackson Laboratories (Bar Harbor, ME) and maintained through several generations in our vivarium before study. C57BL/6J-c2J mice are derived from a C57BL/6J strain that underwent a mutation that inactivated the tyrosinase gene (c) to make it albino. Therefore, the strain is coisogenic with C57BL/6J. By convention, C57BL/6J mice are abbreviated with "B6" or "B" and NZW/LacJ mice with "NZW" or "N". We have abbreviated C57BL/6J-c2J as "B6al". All mice were kept under a 12-12 h light cycle with an in-cage illuminance of 2 to 7 foot candles. The temperature of the vivarium was maintained between 20 °C and 22 °C. Cages were kept on four shelves of five-shelf racks (never on the top shelf). Each week, the cages were rotated by shelf, by side of the rack (left/right) and by position on the shelf (seven positions from front to back looking from the front end of the rack). Mice were maintained on a low fat diet (Number 15001 Rodent Lab Chow, Newco Distributors, Rancho Cucamonga, CA) with chow and water ad libitum.
For the quantitative genetics study, a non-reciprocal (NZW x B6al)F2 cross was made and 201 F2 progeny were aged to 92 to 105 days along with 22 NZW, 19 B6al, and 38 F1 controls before being exposed to constant light of 20 to 70 foot candles for four weeks. All mothers of the F1 mice were B6al because of breeding problems with the NZW females.
Immediately after light exposure, eyes were enucleated from mice after sacrifice and fixed in a mixture of 2% formaldehyde and 2.5% glutaraldehyde in phosphate buffer. The fixed eyes were embedded in an Epon-Araldite mixture and bisected along the vertical meridian through the optic nerve head. A single 1 μm section was taken from the cut surface of one of the eye hemispheres from each mouse and stained with toluidine blue as described previously . On this section, measurements of the thickness of the outer nuclear layer (ONL) were made; three measurements each spaced 50 μm apart were taken at four 0.25-mm intervals in the superior posterior region of the retina (Figure 1). This was the same technique used in the previous B6al x BALB/c retinal light damage study . The mean of the 12 measurements from each eye was used to score the mice for the quantitative trait. All procedures involving the mice adhered to the ARVO Resolution on the Use of Animals in Research and the guidelines of the Loyola Marymount University Committee on Animal Research.
Genotyping services for an initial set of 156 F2 mice were provided by the Center for Inherited Disease Research (CIDR). CIDR is fully funded through a federal contract from the National Institutes of Health to The Johns Hopkins University, Contract Number N01-HG-65403. For each chromosome, the most proximal marker genotyped was within 15 cM of the centromere, internal markers were no more than 30 cM apart and the most distal markers were within 15 cM of the telomere. The average spacing was 16.5 cM. The exceptions to the above were Chrs 2, 11, 13, and 17 where the most proximal markers were 28, 17, 16, and 23 cM from the centromere, respectively, and Chrs 9, 12, and 17 where the most distal markers were 30, 21, and 28 cM from the telomere, respectively. A list of markers used in crosses between NZW and B6al mice is available on the CIDR website. A few additional dinucleotide repeat markers were analyzed in our laboratory. These were amplified by standard PCR methods and electrophoresed in 4% agarose gels for allele determination by size. All map positions were based on the map of the mouse genome from the Jackson Laboratories web site Mouse Genome Informatics (JAX).
Mouse genomic DNAs
Genomic DNAs were isolated from livers with the Puregene DNA isolation kit (Gentra systems, Minneapolis, MN).
Genotypes and quantitative traits for the first set of 156 F2 progeny were analyzed with the Map Manager QTX, version b17 (Map Manager) . With this program, a likelihood ratio statistic (LRS) was calculated for each of the marker genotypes with a probability inclusion level for further study of 0.05. Any markers with a greater than or equal to 95% probability of being associated with the phenotype and any pair of markers with a greater than or equal to 95% probability of interactive effects and an interaction likelihood ratio score of greater than or equal to 20 (recommended for significance by Map Manager QTX) were further genotyped in an additional 45 F2 progeny. Each set of significant markers in a chromosomal region (starting with the set with the highest significance) was studied by interval mapping of all the markers on that chromosome. To determine significance levels for these LRS scores, a test of 1,000 permutations of all marker genotypes together from the set of 201 F2 progeny was performed. p<0.001 was 21.3 (highly significant or HS); p<0.05 was 13.6 (significant or S), and p<0.67 was 7.3 (suggestive or sugg). A highly significant LRS of greater than or equal to 21.3 would only occur by chance in 1 of 1,000 genome scans such as this one. The LRS was converted to a LOD score by dividing by 4.6 (2 x the natural log of 10). The only QTL that was highly significant (middle Chr 1) was placed in the background function of Map Manager for all other QTL.
Figure 1 shows a significant difference in ONL thickness of the retina between NZW and B6al control mice exposed to constant light in both the superior- and inferior-posterior retina. However, for the purposes of comparison with our previous CLD study between B6al and BALB/c, we used the average of only positions 2 to 5 in the superior-posterior retina for the quantitative trait (Figure 1).
Preliminary Map Manager QTX analysis of a genome-wide scan of 156 F2 progeny revealed potential QTL on Chrs 1 (two), 2, 6, 9, 10, 13, 14, and 16. After genotyping all markers on these chromosomes for the additional 45 F2 progeny, QTL on Chrs 1 (two), 10, 13, 14, and 16 remained. QTL on Chrs 2 and 9 disappeared and the peak LRS for the Chr 6 QTL was 6.5, less than the minimal 7.3 for a suggestive QTL. When we analyzed the controls in Map Manager QTX as if they were the results of a single marker with genotype BB for the B6al controls (n=19), NN for the NZW controls (n=22), and H for the F1 heterozygous controls (n=38), the LOD score was nearly 18 and the total variance due to inherent, genetic strain differences 65%. Therefore, an estimate of 35% of the total variance was due to environmental differences. We divided the total genetic variance of the controls into the variance for each QTL to get the percent genetic effect. Table 1 shows the LRS, LOD, percent effect, percent genetic effect, inheritance pattern, and which allele is protective against constant light-induced retinal damage. The total effect of all the QTL detected was 36%. Therefore, nearly half (29%) of the genetic variance between NZW and B6al was not accounted for with individual QTL. There was specifically no QTL at distal Chr 3. The most distal Chr 3 marker genotyped in the first set of 156 F2 progeny was D3Mit128. The locus of this marker is only 3 cM from the Rpe65 gene. D3Mit128 did not meet the p less than or equal to 0.05 criterion for further study nor did any of the other markers genotyped on Chr 3. Nevertheless, we typed it in the additional 45 mice. The LRS for this marker with appropriate background was 4.9, far less than the cut-off of 7.3 for a suggestive QTL.
Some of the remaining undetected 29% variance can be accounted for by an X-linked effect. The average ONL thickness of 100 light-exposed male F2 mice (36.21 μm) was greater than that of 101 females (34.01 μm; Student's unpaired t test, p=0.018) revealing a gender effect that could be due to an X-linked gene. Since the Map Manager QTX program does not distinguish between hemizygous male and homozygous female genotypes for the LRS function in intercrosses, we evaluated the influence on CLD of the loci on the X chromosome by other means. The average ONL thickness of male N and B6al hemizygote genotypes and female homozygous B6al and heterozygous genotypes was calculated for each of the four markers typed on the X chromosome. There were no homozygous N females as the intercross was non-reciprocal. Table 2 shows the ONL values and a comparison of males and females with only the B6al allele to males and females with the N allele or both alleles, respectively. Only the most proximal marker, DXMit192, showed a significantly greater ONL for the B6al allele relative to the N allele revealing a QTL on the proximal X chromosome.
In addition to the X-linked effect, there were three pairs of loci with significant interactions (i.e., two genes acting together to influence the light-induced retinal degeneration in a significant, synergistic way). The interaction function of Map Manager QTX for an intercross tests every marker as an additive and dominant allele against every other marker as additive and dominant (four interactions per pair of markers). The interaction likelihood ratio statistic (IX) needed for significance is about 20 (LOD score of 4.35) for an intercross. When this function was performed with an exclusion probability of p<0.05 on genotype data from the first set of 156 progeny, three interactions of two markers each were found with IX scores of greater than or equal to 20. These three interactions remained significant when all 201 progeny were tested with a p<10-5 (as Map Manager recommends). They were (1) between markers of the middle Chr 1 QTL and markers from the Chr 14 QTL (LRS 24.5, LOD 5.32), (2) between the same Chr 1 markers and markers from the Chr 16 QTL (LRS 23.7, LOD 5.15), and (3) between markers from the Chr 10 QTL and the Chr 6 region where the LRS score was not quite high enough to be a suggestive QTL (LRS 24.1, LOD 5.24).
The first two interactions were protective for a combination of NZW and B6al alleles, and the third interaction was protective for a combination of B6al alleles.
In the Introduction, we identified three possibilities to explain why the NZW/LacJ albino mouse is highly susceptible to constant light-induced retinal damage while carrying the considered highly protective MET450 variant of RPE65. One explanation was that the Rpe65 gene variant did not influence CLD but is only a polymorphism that segregates with a nearby gene. The NZW allele of the gene would be susceptible to CLD and the B6al allele resistant. A second explanation was that the NZW allele of Rpe65 has an intragenic suppressor of the protective MET450 variant. If either possibility were true, there would have been a QTL at the Rpe65 locus in this intercross between B6al and NZW. There was none. The third explanation was that there are other gene alleles contributing to the CLD susceptibility of NZW that overcome the protective effect of RPE65 MET450. We did find other QTL, but the total percentage of their effects is not enough to account for the total variance in the intercross. Thus, QTL with B6al protective alleles on Chrs 1 (distal), 10, 14, and 16 add up to 21% of the total variance, and those on Chrs 1 (middle) and 13 with NZW protective alleles equal 15% (Table 1). Therefore, 29% of the total variance (65%) is not accounted for by single QTL. There was a significant protective effect of an X-linked B6al allele and there was a significant interaction between loci on Chrs 6 and 10 that add to the B6al protective effect, but both of these cannot be measured in percent effect by the Map Manager program. It may be that these are strong enough to account for the rest of the difference in CLD between the two strains. However, a more likely possibility is that the total number of progeny studied was not large enough to uncover a number of relatively weak QTL that together add up to the unaccounted for difference. Nevertheless, the main result of this study was that we did not detect the strong QTL on distal Chr 3 that was present in the retinal light damage study between B6al and BALB/c . Thus, even though NZW carries the MET450 variant and is highly susceptible to CLD, the hypothesis that the MET450 variant of RPE65 protects rod photoreceptors from constant light-induced damage is left intact.
The degree of influence of the RPE65-MET/LEU450 variant on intense bright light-induced retinal damage and constant light-induced retinal damage may be quite different. Wenzel et al.  showed that the amount of RPE65 protein was much greater in BALB/c eyecups relative to those of C57BL/6J (B6), that the regeneration of rhodopsin after bleach was much faster and that retinal light damage susceptibility was much greater in BALB/c mice than in B6 mice when exposed to intense bright light after pupil dilation. In a study of the rate of recovery of a normal electroretinogram after bright flash, we showed that the b-wave returned to normal significantly more rapidly in BALB/c retinas than in those of B6al . However, neither of these studies measures the effect on photoreceptors under conditions of constant light. In the Wenzel study , the same measurements of RPE65 protein, rhodopsin regeneration, and retinal light damage were made after intense bright light exposure of a B6:129S(N2) mouse that was homozygous for the MET450 variant but of a mixed B6/129S background. In this case, the level of RPE65 protein and the rate of rhodopsin regeneration were the same as in pure B6, but the degree of protection was only about 10% of that of the B6 mouse. As the authors pointed out, there must be other factors besides the RPE65 variant that significantly protect the B6 mouse retina from this type of intense, short exposure light insult. On the other hand, in our previous quantitative genetics study between the B6al and BALB/c strains, the QTL locus on distal Chr 3 containing the RPE65-MET450 variant accounted for nearly 50% of the genetic effect of constant light-induced retinal damage (about 40% of the total effect), and the other QTL found had only minor effects . The two types of retinal light damage, bright light-induced and constant light-induced, are mediated by different intracellular mechanisms . This difference may account for the different degree of protection provided by the RPE65 variant. In other words, the MET450 variant is protective in both instances, but to a much greater extent in constant light-induced retinal damage.
Even though the total difference in CLD susceptibility between NZW and B6al was not completely accounted for by the QTL discovered in this study, presumably due to a number of QTL too small to be detected, several of the QTL that were detected have appeared in other quantitative genetics studies. The QTL on Chrs 1 (distal) and 14 were present in the constant light-induced retinal damage study between B6al and BALB/c ; the QTL on Chr 10 was present in an age-related retinal degeneration study between the same two strains ; and the QTL on middle Chr 1 was present in a short exposure, intense light-induced retinal degeneration study conducted between the strains 129S1/SvImJ and BALB/c . A comparison of these two QTL showed them to be remarkably similar in location (Figure 2).
Although the middle Chr 1 QTL occupies a large area, there are at least two excellent candidates in this locus, the genes encoding arrestin (Sag) and the soluble δ subunit of cGMP-phosphodiesterase (Pde6d). Differences in the abundance or efficacy of either of these genes might influence the rate of the phototransduction cascade and therefore the amount of light damage occurring after constant light exposure. However, there are several hundred additional genes in this QTL as it, along with most other QTL, covers a broad chromosomal region. Therefore, the QTL must be refined by testing for phenotype in congenic lines that have been bred in such a way as to introgress in tiled subdivisions the QTL region from one strain onto the background of the other. Such refinement will substantially reduce the number of genes to be considered. There is one small shortcut that is related to QTL appearing in more than one study with more than two strains of mice such as this middle Chr 1 QTL. This allows a quick screening method for variants in candidate genes, because in order for a gene variant to merit further consideration it must segregate with phenotypic effect. Thus, a candidate gene in the middle Chr 1 QTL should have one variant that is shared by 129S1/SvImJ and NZW (protective QTL alleles) and a second variant that is shared between B6al and BALB/c (susceptible QTL alleles). The fact that there was no middle Chr 1 QTL in the CLD study between B6al and BALB/c , both susceptible in this QTL in separate studies, supports the feasibility of this approach.
The identification of modifying genes that influence the severity of light-induced retinal degenerations may be important for human retinal degenerative diseases that are exacerbated by light exposure. Work such as this may also provide clues to the physiological processes involved in retinal degenerations, and suggest explanations for the phenomena of variable degrees of disease severity in individuals carrying a common genetic mutation. Understanding how mutations in one gene are modified by the variations in other "background" genes will provide new avenues of study for the development of effective therapeutic interventions.
This study was supported in part by NIH grants EY13280 (MD), EY01919, and EY02162 (MML), the Foundation Fighting Blindness (MD and MML), Research to Prevent Blindness, That Man May See, Inc., and the Macula Vision Research Foundation (MML). MML is a Research to Prevent Blindness Senior Scientist Investigator.
1. Danciger M, Matthes MT, Yasamura D, Akhmedov NB, Rickabaugh T, Gentleman S, Redmond TM, La Vail MM, Farber DB. A QTL on distal chromosome 3 that influences the severity of light-induced damage to mouse photoreceptors. Mamm Genome 2000; 11:422-7.
2. Grimm C, Wenzel A, Hafezi F, Yu S, Redmond TM, Reme CE. Protection of Rpe65-deficient mice identifies rhodopsin as a mediator of light-induced retinal degeneration. Nat Genet 2000; 25:63-6.
3. LaVail MM, Gorrin GM, Repaci MA, Thomas LA, Ginsberg HM. Genetic regulation of light damage to photoreceptors. Invest Ophthalmol Vis Sci 1987; 28:1043-8.
4. Manly KF, Cudmore RH Jr, Meer JM. Map Manager QTX, cross-platform software for genetic mapping. Mamm Genome 2001; 12:930-2.
5. Wenzel A, Reme CE, Williams TP, Hafezi F, Grimm C. The Rpe65 Leu450Met variation increases retinal resistance against light-induced degeneration by slowing rhodopsin regeneration. J Neurosci 2001; 21:53-8.
6. Nusinowitz S, Nguyen L, Radu R, Kashani Z, Farber D, Danciger M. Electroretinographic evidence for altered phototransduction gain and slowed recovery from photobleaches in albino mice with a MET450 variant in RPE65. Exp Eye Res 2003; 77:627-38.
7. Hao W, Wenzel A, Obin MS, Chen CK, Brill E, Krasnoperova NV, Eversole-Cire P, Kleyner Y, Taylor A, Simon MI, Grimm C, Reme CE, Lem J. Evidence for two apoptotic pathways in light-induced retinal degeneration. Nat Genet 2002; 32:254-60.
8. Danciger M, Lyon J, Worrill D, LaVail MM, Yang H. A strong and highly significant QTL on chromosome 6 that protects the mouse from age-related retinal degeneration. Invest Ophthalmol Vis Sci 2003; 44:2442-9.
9. Danciger M, Lyon J, Worrill D, Hoffman S, Lem J, Reme CE, Wenzel A, Grimm C. New retinal light damage QTL in mice with the light-sensitive RPE65 LEU variant. Mamm Genome 2004; 15:277-83.