The Rd8 Mutation of the Crb1 Gene Is Present in Vendor Lines of C57BL/6N Mice and Embryonic Stem Cells, and Confounds Ocular Induced Mutant Phenotypes

Abstract

Purpose.

We noted an unexpected inheritance pattern of lesions in several strains of gene-manipulated mice with ocular phenotypes. The lesions, which appeared at various stages of backcross to C57BL/6, bore resemblance to the rd8 retinal degeneration phenotype. We set out to examine the prevalence of this mutation in induced mutant mouse lines, vendor C57BL/6 mice and in widely used embryonic stem cells.

Methods.

Ocular lesions were evaluated by fundus examination and histopathology. Detection of the rd8 mutation at the genetic level was performed by PCR with appropriate primers. Data were confirmed by DNA sequencing in selected cases.

Results.

Analysis of several induced mutant mouse lines with ocular disease phenotypes revealed that the disease was associated 100% with the presence of the rd8 mutation in the Crb1 gene rather than with the gene of interest. DNA analysis of C57BL/6 mice from common commercial vendors demonstrated the presence of the rd8 mutation in homozygous form in all C57BL/6N substrains, but not in the C57BL/6J substrain. A series of commercially available embryonic stem cells of C57BL/6N origin and C57BL/6N mouse lines used to generate ES cells also contained the rd8 mutation. Affected mice displayed ocular lesions typical of rd8, which were detectable by funduscopy and histopathology as early as 6 weeks of age.

Conclusions.

These findings identify the presence of the rd8 mutation in the C57BL/6N mouse substrain used widely to produce transgenic and knockout mice. The results have grave implications for the vision research community who develop mouse lines to study eye disease, as presence of rd8 can produce significant disease phenotypes unrelated to the gene or genes of interest. It is suggested that researchers screen for rd8 if their mouse lines were generated on the C57BL/6N background, bear resemblance to the rd8 phenotype, or are of indeterminate origin.

The rd8 mutation of the Crb1 gene, which can cause significant retinal degeneration, was found in homozygous form in the C57BL/6N mouse substrain. It is present in stocks from most major mouse vendors and in ES cells derived from the C57BL/6N substrain. Its presence can confound interpretation of ocular mutant phenotypes.

Introduction

The rd8 mutation is a single nucleotide deletion in the Crb1 gene, which results in a form of retinal degeneration having a distinct clinical appearance, including multiple light-colored spots in the fundus of the eye that correspond histologically to retinal folds, pseudorosettes, as well as focal retinal dysplasia and degeneration.¹ The trait is autosomal recessive and has been identified in several mouse lines.² C57BL/6 mice, from which many transgenic strains are generated and which are the source of increasingly popular lines of embryonic stem (ES) cells, were not known to carry the mutation. Furthermore, the C57BL/6 background was reported to suppress expression of the rd8 phenotype.²

Recent data with induced ocular mutants generated by the authors, or derived from strains available from collaborators, revealed an unexpected inheritance pattern of ocular lesions in several lines of gene-manipulated mice. Specifically, while crossing mice carrying a gene deletion or a transgene designed to confer an ocular phenotype onto the C57BL/6 background, the phenotype being studied was noted in some littermate controls. These findings raised the possibility that the observed phenotypes were not due to the genes of interest, but were the result of other unidentified factors. On close examination, the lesions were reminiscent of rd8 associated retinal degeneration.² Therefore, we conducted a PCR screen for rd8 and found 100% of mice with disease were homozygous for rd8. Since the rd8-positive mice examined were derived in different institutions, and the ocular phenotype was discovered during backcross to C57BL/6 in littermates that did not carry the gene of interest, it was decided to examine C57BL/6 mice from major commercial vendors and commercially available C57BL/6 ES cells for the rd8 mutation. These data demonstrated the presence of the rd8 mutation in homozygous form in all C57BL/6N substrains regardless of the commercial source, but not in the C57BL/6J substrain. Embryonic stem cells of C57BL/6N origin also contained the rd8 mutation. Affected mice displayed ocular lesions typical of rd8, which were detectable by fundus and histological examination as early as at 6 weeks of age. These findings have serious implications for the ocular community who develop transgenic and knockout mice using C57BL/6N embryos or C57BL/6N-derived ES cells.

Materials and Methods

Mice

HLA-A29 construct was provided by Dr. Jacques Cohen (INSERM, Paris, France).³ Founder mice transgenically expressing this construct were generated by the NEI genetic core facility using embryos derived from female C57BL/6N mice purchased from DCT (Frederick, MD). Ccl2/Cx3cr1 double-knockout mice were bred from the respective single knockouts obtained from collaborators.^4,5 Wild type C57BL/6 mice (6–8 weeks old) were purchased from several vendors, including DCT (stock number 01CC55), Taconic (TAC, Hudson, NY, stock number B6), Harlan Sprague Dawley (HSD, Indianapolis, IN, stock number 044), Charles River Laboratories (CRL, Frederick, MD, stock number 027), and Jackson Labs (JAX, Bar Harbor, ME, stock number 000664). (Note that DCT, HSD, CRL and TAC supplied C57BL/6N mice and JAX supplied C57BL/6J mice.) B6Smn. C3-FasL^gld/J (B6-gld) mice were obtained from JAX, C57BL/6-Bid^−/− (B6-Bid^−/−) mice were obtained from Dr. Richard Hotchkiss (Washington University School of Medicine, St. Louis, MO), and the C57BL/6-Apo-2L^−/− (C57BL/6-Trail^−/− or B6-Trail^−/−) mice were provided by Dr. Thomas Griffith (University of Minnesota, Minneapolis, MN). These lines were crossed to the Ccl2/Cx3cr1 double knockout mice to obtain triple knockout mice. Studies conform to the ARVO statement for the Use of Animals in Ophthalmic and Vision Research and to Institutional guidelines.

Cells and Reagents

Mouse ES lines R1, W4 and JM8.N4 were obtained from the NEI Genetics Core. The ES cells characterized in our study were cultured in the absence of feeder cells for several months (at least 3 passages). Mouse strains used to derive the ES cell lines SCC1O, B6-Blu, B6-GFP, EDJ22, R1, and SCC10 were tested at the Siteman Cancer Center at Washington University School of Medicine. DNA was prepared by standard protocols and samples were genotyped for the rd8 mutation by PCR.

Detection of rd8 Mutation by PCR

DNA was isolated from mouse tail biopsy samples using a Qiagen Biosprint instrument, Qiagen reagents (Qiagen Inc., Valencia, CA) and the protocol suggested by the manufacturer. DNA samples isolated from tail biopsies were amplified separately for wild type (Wt) allele and mutant rd8 allele using primers specified by Mehalow et al.² Primer sequences included mCrb1 mF1: GTGAAGACAGCTACAGTTCTGATC; mCrb1 mF2: GCCCCTGTTTGCATGGAGGAAACTTGGAAGACAGCTACAGTTCTTCTG; and mCrb1 mR: GCCCCATTTGCACACTGATGAC. For PCR amplification approximately 25 ng DNA were used in a 25 μL reaction volume containing 1.5 mM MgCl₂, 100 μM of each dNTP, 1.6 μM each of forward and reverse primer for Wt allele, and 0.8 μM of forward and 1.6 μM of reverse primer for rd8 mutant allele and 0.025U AmpliTaq DNA polymerase. Reactions initially were denatured at 94°C for 5 minutes followed by 35 cycles at 94°C for 30 seconds, 65°C for 30 seconds, 72°C for 30 seconds and a final extension at 72°C for 7 minutes. Amplicons were separated using 3% Nusieve 3:1 (Lonza, Rockland, ME) agarose gel and visualized under UV light after staining with ethidium bromide. Amplicon sizes are Wt allele = 220 bp and rd8 allele = 244 bp. Genotyping of ES cells at the NEI Genetic Core was performed by a TaqMan allelic discrimination assay with common PCR primer sequences CCCTGTTTGCATGGAGGAAA (forward) and CCTGACCATCCCGAGAGACA (reverse), and the allele-specific probe sequences were [6FAM]AGCTACAGTTCTTATCGG[MGBNFQ] (wild type) and [VIC]CAGCTACAGTTCTTATGGT[MGBNFQ] (rd8). The allelic discrimination assay was run on an Applied Biosystems (Carlsbad, CA) 7900HT running SDS 2.3 software using the protocols specified by the manufacturer. Please note that primer sequences for the two methods are different.

DNA Sequencing for Detection of rd8-Associated Nucleotide Deletion

Mouse tail DNA from a C57BL/6J (JAX) and C57BL/6N (DCT) mouse was PCR amplified with the primers GGTGACCAATCTGTTGACAATCC (forward) and GCCCCATTTGCACACTGATGAC (reverse), and the product was cleaned up using a QIAquick PCR Purification Kit (Qiagen). The two samples and two PCR primers were sent to ACGT Inc. (Wheeling, IL) for sequencing in both directions. Portions of the sequences were aligned using the MegAlign program (DNAStar, Madison, WI).

Fundus Photography and Optic Coherence Tomography (OCT)

Mice were anesthetized systemically with 200–300 μL of a mixture of Ketamine (7.7 mg/mL) and Xylazine (0.45 mg/mL), and topically with 0.5% Proparacaine HCl (Alcon Labs, Inc., Fort Worth, TX). The pupil was dilated with 0.5% tropicamide ophthalmic solution (Bausch & Lomb, Rochester, NY) and 2.5% phenylephrine HCl (Bausch & Lomb) drops, and the corneas were kept moist with regular application of sterile lubricant GenTeal severe (Novartis, New York, NY). Fundus photographs were taken with a Micron-2 camera (Phoenix Research Laboratories, Inc., Pleasanton, CA) using StreamPix 5 software.

An ultra-high resolution spectral-domain optic coherence tomography (SD-OCT) system designed for small animal imaging (Bioptigen, Research Triangle Park, NC) was used for in vivo imaging of mouse retina. A 1.4 mm linear scan was obtained with 1000 A-Scan per B-scan.

Histopathology

Eyes were enucleated, fixed in 4% phosphate buffered glutaraldehyde for 1 hour and stored in 10% phosphate buffered formaldehyde until processing. Tissues were embedded in methacrylate. Sections 6–8 μm thick were stained with hematoxylin and eosin. Eight to 10 sections cut at different planes were examined for each eye in a masked fashion by one of us (CCC), an ophthalmic pathologist.

Results

Detection of rd8 in HLA-A29 Transgenic Mice

The human uveitic disease known as birdshot chorioretinopathy (BC, also known as birdshot retinochoroidopathy) is associated very strongly with the HLA-A29 gene.⁶ A model for this disease had been generated in transgenic mice by Szpak et al.³ using an HLA-A29 gene sequence cloned from a BC patient, in whom BC-like ocular lesions developed between 8 and 12 months of age. The strain had been lost, so one of us (RRC) obtained the original HLA-A29 DNA construct and generated founder mice by microinjection into embryos of C57BL/6N females purchased from DCT. The founders were bred to C57BL/6N mice from the same source to produce F1 mice. F1 mice either were intercrossed or bred to C57BL/10J (in preparation for subsequent crossing to human β2m transgenic mice, which are on the C57BL/10J background). All founders and progeny were examined periodically for ocular lesions by fundus examination. Ocular lesions were observed in all HLA-A29 founders and in many of their progeny, suggesting the original findings of Szpak et al.³ may have been recapitulated.

However, some inconsistencies became apparent that put in question the association of the observed lesions with the HLA-A29 transgene: First, unlike BC in patients and the original mouse line of Szpak et al.,³ the ocular lesions in HLA-A29 transgenic mice appeared to be degenerative, not inflammatory, and appeared soon after weaning age. Second, mice did not have the human β2 microglobulin gene needed to associate with the A29 chain for expression on the cell surface. Therefore, expression of HLA-A29 molecules in these mice was so low as to be undetectable by conventional assays. This suggested that the HLA-A29 may not be responsible for the observed phenotype. Third, and most significant, similar lesions were observed in the littermates not carrying the A29 transgene (Table 1 and Fig. 1).

Table 1.

Association of Ocular Phenotype with rd8 Genotype in HLA-A29 Tg Mice

Mouse Line	No. Examined	No. with Disease (%)	rd8Genotype*
HLA-A29 Tg founders (B6N)†	5	5 (100%)	rd8/rd8
HLA-A29 Tg (B6NxB6N)	18	17 (94%)	rd8/rd8
HLA-A29 WT littermates (B6Nx B6N)	9	8 (89%)	rd8/rd8
HLA-A29 Tg (B6NxB10J)	20	5 (25%)	rd8/wt
HLA-A29 WT littermates (B6NxB10J)	25	5 (20%)	rd8/wt

^* B6N, C57BL/6N; B10J, C57BL/10J.

^† Genotype shared by all mice within each category (phenotype-positive and negative).

Figure 1.

Ocular findings in Wt littermates of HLA A29 Tg mice compared to healthy eyes. Fundus photographs were taken at 9–10 weeks of age (A) and SD-OCT at 18 weeks of age (B). Healthy retinas are depicted on the right. (A) Fundus lesions appeared as multiple bright spots in the retina, coalescing to form diffuse lesions or large patches. (The three bright spots visible in the same position on all fundi and the central black spot are camera artifacts.) (B) SD-OCT of eyes are shown in panel A: linear scans of retina showed localized areas of hyperreflectivity and thin retina with loss of inner and outer segment layers, or loss of normal architecture of the inner and outer nuclear layers at focal areas away from the optic nerve; note similarity of these lesions to the ones described by Aleman et al.¹⁰ in mouse Crb1 associated retinal degeneration. (C) Histology – note foci of total photoreceptor atrophy, with loss of outer and inner segments, outer nuclear and outer plexiform layers (left), multiple focal photoreceptor dystrophy without retinal pigment epithelial (RPE) abnormalities (center). (D). Sample gel showing resolution of the rd8 homozygous (244bp), rd8 heterozygous and Wt (220bp) genotypes. GC, ganglion cells; INL, inner nuclear layer; ONL, outer nuclear layer; PIS/OS, photoreceptor inner segments/outer segments.

Careful genotyping of phenotype-positive vs. -negative mice, performed independently by one of us (TAF) and by Dr. Bo Chang at JAX, and confirmed by the NEI Genetic Core Facility, revealed that (1) all founders were homozygous for the rd8 mutation, and (2) the various progeny displayed a complete correlation between the ocular phenotype and presence of homozygous Crb1^rd8 allele (Table 1). Notably, of the F2 progeny that were crossed to C57BL/10J and were heterozygous for rd8, about 20% also displayed similar retinal lesions whether or not they carried the A29 transgene. Since disease was independent of the A29 genotype, we conclude that the observed phenotype either was caused by the rd8 mutation or was influenced heavily by its presence.

The Presence of the rd8 Mutation in Mice with Advanced Macular Degeneration (AMD)-Like Phenotypes

In 2007, one of us (CCC) published an AMD-like model in Ccl2^−/−Cx3cr1^−/− mice.⁷ These mice were produced by crossing the single knockout Ccl2^−/−4 and Cx3cr1^−/−5 strains, both with reported AMD-like features that developed by 1 year of age. The Ccl2^−/−Cx3cr1^−/− double knockout mice develop ocular spots observed by funduscopic examination (Fig. 2A) and retinal lesions upon histological examination as early as 6–8 weeks of age (Fig. 2B). In the process of exploring the mechanisms of disease in the Ccl2^−/−Cx3cr1^−/− mice by one of us (TAF) the Ccl2^−/−Cx3cr1^−/− line was crossed to several other mouse lines (B6-gld, B6-Bid^−/− and B6-Trail^−/−) to develop triple knockout mice. During this process a number of heterozygous littermate controls were generated (Table 2). Significant disease (Fig. 2A, 2B) was noted that was independent of the genotype. A more complete list of mouse lines is shown in Table 2, where disease correlated 100% with the presence of the rd8 mutation. It should be noted that backcrossing the Ccl2^−/−Cx3cr1^−/− mice onto the C57BL/6J background (Wt at the Crb1 locus) resulted in mice that were disease-free when examined at 12 weeks.

Figure 2.

Ocular finding in AMD-like mice. (A) Fundus photographs from selected mice displaying multiple spots typical of the rd8 phenotype. All mice with the genotype rd8/rd8 had a similar appearance, while Wt/Wt mice were identical to the Ccl2^−/-Cx3Cr1^−/-gld mice shown. (B) Histological finding in selected mice rd8/rd8 and Wt/Wt mice. Note the retinal folding, dysplasia of the nuclear layers, retinal degeneration, and RPE vacuolation. Lesions of this type typically were focal in the eyes examined.

Table 2.

Association of Ocular Phenotype with rd8 Genotype in AMD-Like Model Mice

Mouse Line	No. Examined	Disease*	rd8Genotype†
Ccl2−/− Cx3cr1−/−	80	Yes	rd8/rd8
Ccl2−/−Cx3cr1+/−	5	Yes	rd8/rd8
Ccl2−/−	4	No	Wt/rd8
Ccl2−/− (Jax)‡	3	No	Wt/Wt
Cx3cr1−/−	6	Yes	rd8/rd8
Cx3cr1+/−	2	No	Wt/rd8
Ccl2+/−Cx3cr1+/−	4	No	Wt/rd8
Ccl2+/− Cx3cr1+/−	5	No	Wt/Wt
Ccl2−/− Cx3cr1−/− Bid−/−	5	Yes	rd8/rd8
Ccl2−/− Cx3cr1−/− Bid−/−	4	No	Wt/Wt
Ccl2−/− Cx3cr1+/− Bid+/−	3	Yes	rd8/rd8
Ccl2−/− Cx3cr1−/− Trail−/−	5	Yes	rd8/rd8
Ccl2−/− Cx3cr1−/− Trail−/−	5	No	Wt/Wt
Ccl2−/− Cx3cr1−/− gld/gld	50	No	Wt/Wt
Ccl2+/− Cx3cr1+/− gld/+	3	No	Wt/rd8
B6-gld (Jax)‡	15	No	Wt/Wt
B6-Trail−/−	10	No	Wt/Wt
B6-Bid−/−	5	No	Wt/Wt
Ccl2−/− Cx3cr1−/− (C57BL/6J)§	10	No	Wt/Wt

^* Disease criteria included 2 or more of the following: spots on fundus, retinal folds, photoreceptor degeneration, focal retinal lesions such as retinal angiomatous proliferation, photoreceptor dystrophy, RPE vacuolation, RPE degeneration.

^† Genotype shared by all mice within each category.

^‡ Purchased from Jackson labs.

^§ Backcrossed to C57BL/6J and determined to be 100% C57BL/B6J by microsatellite (SNP) analysis.

The Ccl2^−/−Cx3cr1^−/− mice were derived from Ccl2^−/− (obtained from Bao Lu and Barrett J. Rollins of Children's Hospital, Harvard Medical School, Boston, MA)⁴ and Cx3cr1^−/− (obtained from Philip Murphy of the National Institute of Allergy and Infectious Diseases, National Institutes of Health, Bethesda, MD).⁵ Notably, the founder Cx3cr1^−/− mice had been bred to C57BL/6N mice from DCT for 13 generations, and the Ccl2^−/− founder mice had been bred to C57BL/6N mice obtained from CRL for 6–7 generations before we received them (Philip Murphy and Bao Lu, personal communication). Analysis of these parental lines (Table 2) revealed that the Ccl2^−/− mice were either heterozygous for the rd8 mutation or wild type at the Crb1 locus. All Cx3cr1^−/− parental mice were homozygous rd8. Interestingly, only 50% Cx3cr1^−/− mice had retinal lesions at 6 weeks and no Ccl2^−/− mice had retinal lesions at 6 weeks of age. This analysis suggests that the rd8 mutation is likely to have entered these mouse lines during the creation of the single knockout mice. Furthermore, as with the B6-HLA-A-29 transgenic mice discussed above, the pathological features in single knockout Ccl2^−/−, Cx3cr1^−/−, and in double knockout Ccl2^−/−Cx3cr1^−/− mice, either depended on rd8 or were modified by the presence of this mutation.

The rd8 Mutation Is Present in C57BL/6 Vendor Lines

On the basis of the data described above, it seemed unlikely that rd8 appeared spontaneously in the unrelated HLA-A29 transgenic and Ccl2^−/−Cx3cr1^−/− double knockout mouse lines. Rather, the source was more likely a mouse strain common to their genetic history. Although it was not possible to trace the complete lineage of all strains involved, we did have the full pedigree of the HLA-A29 Tg line, which was developed in house, and detected presence of rd8 in the founders. Therefore, we thought it reasonable to examine strains common to the ancestry of many transgenic and knockout mice used in vision research. We obtained C57BL/6 mice from the major vendors, including HSD, DCT, CRL, and TAC. These suppliers all maintain the C57BL/6N substrain, which was derived after 1951 from C57BL/6J that had been maintained at JAX since the early part of the 20^th century.⁸ We also obtained C57BL/6J and C57BL/10J mice from JAX. Genotyping revealed that the C57BL/6N substrain, regardless of the vendor, uniformly carried the rd8 gene (Fig. 3A), while C57BL/6J and C57BL/10J were wild type at the Crb1 locus. The location and identity of the mutation found in C57BL/6N were confirmed by sequencing through the Crb1 gene, demonstrating the expected deletion in the C57BL/6N mouse line (Fig. 3B). Fundus and histological examination revealed that C57BL/6N mice from all four suppliers demonstrated typical rd8 lesions to a variable extent, which were well developed as early as 6 weeks of age (Figs. 4A–D). Notably, C57BL/6J mice appear normal, as did control rd8 negative C57BL/10J mice (Figs. 4E, 4F). Histology was performed on affected eyes demonstrating typical lesions with non-inflammatory retinal degeneration involving the inner and outer nuclear layers, the outer plexiform layer, as well as the photoreceptor outer segments (Fig. 4G).

Figure 3.

Genotyping for rd8. (A) Genotyping of vendor mice by PCR for rd8. Bands were resolved by capillary electrophoresis. Mice are from HSD, DCT, CRL, TAC, and JAX. Genotyping for rd8 was performed by PCR. Two representative mice out of 7 obtained from each vendor in 2 separate shipments are shown. (B) Sequencing of Crb1 gene shows a single base deletion in the N substrain at the expected position in the aligned sequences (MegAlign program). N-MF, N substrain sequenced with forward primer; N-MR, N substrain sequenced with reverse primer; J-MF, J substrain sequenced with forward primer; J-MR, J substrain sequenced with reverse primer.

Figure 4.

Vendor C57BL/6N mice, but not C57BL/6J mice, have fundus lesions. (A–E) Fundus photographs of mice whose genotyping is shown in Figure 3A that were obtained from the specified vendors. Shown is one representative mouse out of 7 obtained from each vendor in 2 separate shipments. Note spotty lesions of varying extent in all mice except C57BL/6J. (F) C57BL/10J mice with normal fundi serve as control. (G) Histopathology. Shown are eye sections from affected DCT and CRL C57BL/6N mice compared to normal JAX C57BL/6J mice. Note retinal lesions marked by arrows. Histology of other lesion-positive mice looked essentially identical.

The rd8 Mutation Is Present in ES cells of C57BL/6N Origin

C57BL/6N mice are the source of many ES cell lines used to generate gene targeted mice. Consequently, we obtained a number of ES cell lines from the NEI Genetic Engineering Core as well as the ES cell core lab in the Siteman Cancer Center at Washington University in St. Louis, Missouri. PCR analysis revealed that all ES cells derived from C57BL/6N were uniformly homozygous for the rd8 mutation, in keeping with the vendor C57BL/6N results. ES lines derived from C57BL/6J or the 129 strain had the wild type Crb1 allele (Tables 3 and 4).

Table 3.

Genotyping of C57BL/6N-derived Embryonic Stem (ES) Cells

Mouse	ES Cell Line	Rd8 Status
129Sv hybrid	R1	Wt/Wt
129S6/SvEvTac	W4	Wt/Wt
C57BL/6N	JM8.N4 (KOMP)	rd8/rd8

ES cells tested at NEI genetic core facility.

Table 4.

Genotyping of C57BL/6N-derived Embryonic Stem (ES) Cells

Mouse	ES Cell Line	Rd8 Status
129Sv/J*	SCC10	Wt/Wt
C57BL/6N*	B6-Blu	rd8/rd8
C57BL/6N	EUCOMM	rd8/rd8
C57BL/6J*	B6-GFP	Wt/Wt
129svEv/6J*	EDJ22	Wt/Wt
129x129 hybrid*	R1	Wt/Wt
MJC (129)*	SCC10	Wt/Wt

ES cell lines tested at the Siteman Cancer Center (Washington University).

^* Mice used to generate ES cell lines for the Siteman Cancer Center ES Core were screened for the rd8 mutation.

Discussion

Crb1^rd8 is a recessive mutation and the C57BL/6 background was reported by Mehalow et al. to suppress expression of the phenotype.² These authors demonstrated that rd8-positive mice on a mixed background lost the phenotype upon extensive backcrossing of the mutation to C57BL/6J. Interestingly, these investigators noted what they called “founder effects” in mice that expressed the phenotype and considered the gene to have originated from C57BL/6 (of unspecified substrain). Our findings indicated that the C57BL/6N background not only harbors the mutation in its homozygous state, but also allows for its full expression. The C57BL/6N substrain is offered by many major vendors and it seems that neither the vendors, nor the vision research community, are aware of this issue. The fact that C57BL/6N from all of the major United States mouse vendors that were examined carry the rd8 allele suggests that this mutation occurred and became fixed in the C57BL/6N substrain after 1951, when C57BL/6 was originally transferred from JAX to NIH.⁸ Furthermore, it remained in this mouse substrain when the C57BL/6N line was transferred from NIH to CRL labs in 1974 and subsequently was disseminated to additional mouse vendors.⁸ Notably, the C57BL/6NJ strain (JAX #005304), maintained at JAX from NIH embryos that had been cryopreserved in 1984, also carries the rd8 mutation (B. Chang, Jackson Labs, personal communication). While C57BL/6N mice available from vendors in Europe have not yet been examined, numerous substrains of C57BL/6N available in Japan, including the C57BL/6ByJ strain (derived from C57BL/6N around 1960 by JAX), were found to be homozygous for Crb1^rd8. C57BL/6J mice available in Japan did not harbor rd8 (Atsushi Yoshiki, RIKEN BioResource Center, Ibaraki, Japan, personal communication). The wide prevalence of the rd8 mutation seems reminiscent of the previously described rd1 mutation that was identified in many common laboratory mouse strains.⁹ However, in contrast to rd8, the rd1 mutation causes complete photoreceptor loss, so it is unlikely to be confused with any other retinal phenotype.

Importantly, ES cells used to generate gene-targeted mice typically are derived from the C57BL/6N strain, including the widely used Knockout Mouse Project (KOMP) and European Conditional Mouse Mutagenesis Program (EUCOMM) cell lines. We found the rd8 mutation in KOMP as well as EUCOMM ES cells that we tested. Although we did not examine ES cells from Texas A&M Institute for Genomic Medicine (TIGM, College Station, TX) or Canadian Mouse Mutant Repository (NorCOMM, Toronto, Ontario, Canada), these lines all were derived from C57BL/6N mice and likely harbor the rd8 mutation. This raises the possibility that the undiscovered presence of rd8 in induced mutant strains might account for some of the published ocular phenotypes. We currently are reexamining all our mutant lines for presence of the rd8 gene. Notably, unlike C57BL/6J, at least some investigator-generated mice on C57BL/6 background distributed by JAX carry the rd8 mutation (e.g. IFN-γ^−/−, IL-10^−/−, data not shown). It is likely that these lines were generated using C57BL/6N ES cells and/or were crossed onto the B6 background using C57BL/6N mice by the investigator who deposited them at JAX.

It must be taken into account that, although full expression of an rd8 phenotype was reported to require homozygosity, it is difficult to exclude the possibility that a heterozygous rd8 might modify expression of other ocular phenotype genes, such that a phenotype might be observed that would not be present without rd8. Moreover, because retinal lesions were seen in rd8-heterozygous animals with C57BL/6N x C57BL/10J parentage (Table 1), it must be considered that some genetic backgrounds might allow heterozygous expression of the rd8 phenotype. In addition, since ocular phenotypes are not completely uniform between mutant mouse lines examined, we also cannot exclude the possibility that background genes (or even the genes of interest) might modify the rd8 phenotype. Similarly, it is possible that a mutation in the Crb1 gene, which is involved in regulating cell polarity in epithelial cells, might influence phenotypes in inflammatory or injury-induced ocular disease models. This will require further analysis. However, on the basis of the data presented here, it is highly recommended that researchers working with knockout or transgenic mice with retinal phenotypes should genotype their mice for the presence of the rd8 mutation.