Abstract
purpose. FVB/N is considered an ideal inbred mouse strain for transgenic mouse production because of the ease of pronuclear microinjection and its overall fecundity. It is well established that vertebrate lens fiber cells normally express a modified intermediate filament network consisting of the proteins filensin and CP49, and it was recently reported that the mouse strain 129 harbors mutations in CP49 that have the potential to confound the interpretation of gene knockout studies of the lens. The purpose of this study was to evaluate the status of the CP49/Bfsp2 gene in the FVB/N strain.
methods. PCR analysis of genomic DNA was used to evaluate the status of the CP49 gene in FVB/N mice procured from the four major US distributors of these animals—Harlan Laboratories, Taconic Farms, Jackson Laboratory, and the NIH/NCI/DCT production facility run by Charles River Laboratories. The structure of the CP49 transcript was evaluated by RT-PCR, and the presence of CP49 protein in the lens was evaluated by immunofluorescence.
results. FVB/N mice obtained from all four US distributors were shown to harbor a 6-kb deletion of the CP49 gene identical with that previously reported in mouse strain 129; C57BL/6 mice did not have this modification. Immunofluorescence demonstrated that FVB/N mice do not have detectable CP49 or filensin protein in the lens, whereas C57BL/6 mice have the expected protein distribution.
conclusions. In humans, mutations in the CP49/BFSP2 gene have been linked to familial, congenital cataract, demonstrating an important role of this gene in lens transparency. The demonstration that FVB/N mice lack CP49 protein in the lens suggests that it may be necessary to reevaluate the mechanisms underlying lens phenotypes obtained as a result of transgenic manipulation of this strain.
The ability to manipulate the mouse genome by the addition of new genes or the deletion of endogenous genes has greatly advanced our understanding of ocular biology.
1 2 3 The interpretation of these manipulations is greatly aided by the use of inbred mouse strains to remove the influence of genetic variability on any resultant phenotypes.
4 However, a major disadvantage of inbred mice is that they often exhibit reduced fecundity and poor parental care, which makes the propagation of these strains problematic. Since the mid 1980s, the mouse strain FVB/N has been the preferred inbred strain for transgenic studies because of its excellent reproductive performance
4 and the presence of unusually large pronuclei in newly fertilized FVB/N oocytes, which makes pronuclear microinjection easier in this strain than in others.
5
However, the usefulness of FVB/N mice to the study of ocular biology has been limited by the presence of the retinal degeneration-1 mutation in the phosphodiesterase 6B gene (
Pde6b rd1 ) in this strain
6 and the resultant loss of rod photoreceptors from the retina by 36 days after birth.
7 In spite of this, a large percentage of transgenic mouse studies of lens biology published in the past 20 years use FVB/N mice and thus make all comparisons to the normal lens within this strain.
It was recently reported that the inbred mouse strain most commonly used for gene knockout studies, 129, carries a spontaneous 6-kb deletion in the
CP49/Bfsp2 gene. This mutation removes a portion of intron 1 and the 5′ splice acceptor of exon 2 and results in a lack of CP49 protein production. The consequence of this mutation is that lenses from 129 mice lack beaded intermediate filaments, experience increased light scatter, and exhibit alterations in lens optical quality.
8 9 Here, we show that FVB/N mice also harbor this same mutation, with the consequence that CP49 and filensin protein are lacking from the FVB/N lens.
All animal experiments described in this article conform to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. FVB/NTac mice were obtained from Taconic Laboratories (Hudson, NY), FVB/NHar and C57Bl/6Har mice were obtained from Harlan Laboratories (Indianapolis, IN), FVB/NCr and C57BL/6NCr mice were obtained from the NIH/NCI/DCT production facility (Frederick, MD) run by Charles Rivers Laboratories, and 129 mice were bred in-house and were a mixed background of 129S4/SvJae and 129S6/SvEvTac or 129 × 1/SvJ. Genomic DNA was prepared from tail biopsy specimens unless otherwise stated. Genomic DNA was prepared from the livers of FVB/NCr mice by cesium chloride centrifugation. Genomic DNA from FVB/NJ mice was purchased from the Jackson Laboratory Mouse DNA Resource (Bar Harbor, ME).
Genotyping for the previously reported 6303-bp deletion in the
Bfsp2/CP49 gene was performed using two different methods. PCR analysis was performed for the intact allele using primers f (5′-AGT GCT TAC AGA GGC CAG AAG AAG G-3′) and d′ (5′-CCT CTG ACA AAG TCT TGA GCT CTC-3′) and for the mutant allele using primers c (5′-TGG GGT TGG GCT AGA AAT CTC AGA-3′) and e′ (5′-AGC CCC TAC GAC CTG ATTTTT GAG-3′), as previously described.
9 Products of these reactions were sequenced to confirm the structure of the mutant allele. Competitive PCR was used to independently confirm the genotype results using reverse primer, intron 2 (5′-CAG CAT TAT CTA CCG TGG TCT GGA G-3′) and forward primer, intron 1 C57BL/6J (5′-CGC TCT GGG TCT CGC ATG AG-3′), giving a 205-bp product, and forward primer, intron 1 129 × 1/SvJ (5′-CAG TCA TGT GGT TCT GGA AG-3′) giving a 347-bp product. PCR reactions were performed (Hot Start with FastStart
Taq DNA Polymerase; Roche Applied Science, Indianapolis, IN) by denaturation at 95°C for 4 minutes; 35 cycles of 94°C for 30 seconds, 50°C for 30 seconds, and 72°C for 1 minute; and a final extension of 72°C for 5 minutes.
Eyes were isolated from C57Bl/6Har and FVB/NHar mice, and 16-μm sections were prepared and immunostained as previously described.
10 Briefly, sections were fixed with acetone-methanol, blocked with 1% bovine serum albumen (BSA), and incubated for 1 hour at room temperature with either anti-CP49 (a gift from Paul Fitzgerald, University of California, Davis) or anti-filensin (a gift from Roy Quinlan, Durham University, Durham, UK) in 1% BSA/PBS. The primary antibodies were detected with labeled goat anti-rabbit IgG (Alexa-Fluor 568; Molecular Probes, Eugene, OR), and the sections were counterstained with a nucleic acid stain (Draq-5; Biostatus Ltd., Leicestershire, UK). Images were captured on a confocal microscope (LSM 510; Carl Zeiss Inc., Göttingen, Germany) configured with an argon/krypton laser (488-nm and 568-nm excitation lines) and a helium neon laser (633-nm excitation line; Carl Zeiss Inc.).
During analysis of animals derived from a cross between 129 strain mice carrying a floxed gene and FVB/N mice overexpressing CRE recombinase in the lens,
11 we genotyped the animals for the deletion in the
Bfsp2 gene
9 previously reported in the 129 strain (data not shown). Surprisingly, all animals from this cross were homozygous for the deleted allele; hence, we sought to determine how widespread the
Bfsp2 deletion was in FVB/N substrains. PCR analysis for the wild-type allele of
Bfsp2 only detected its presence in C57BL/6Har mice, whereas no amplification was detected in 129 × 1/SvJ or FVB/N mice from four different US distributors
(Fig. 1A) . PCR analysis using primers flanking the 6196-bp deletion previously reported in 129 substrains detected the deleted allele in 129 × 1/SvJ and all four FVB/N substrains, but, as expected, no amplification was detected in C57BL/6Har mice. Sequence analysis of the resultant products demonstrated that the 129 and FVB/N strains harbored exactly the same 6303-bp deletion (not shown). DNA from mice heterozygous for the FVB/N and C57BL/6 genetic backgrounds amplified with both primer sets
(Figs. 1A 1B) . To independently confirm this result, the same FVB/N substrains were assayed at the Transgenic Animal and Genome Manipulation Section of the National Eye Institute using a competitive PCR strategy, and all FVB/N animals were found to harbor the
Bfsp2 deletion (
Fig. 1C , gray arrow).
It was previously reported that the
Bfsp2 deletion mutation results in a loss of CP49 and a drastic reduction of filensin protein expression from the lens.
8 9 Thus, we performed immunostaining on lenses from adult FVB/N and C57BL/6 mice using anti-CP49 and anti-filensin antibodies. C57BL/6 lenses yielded intense immunostaining in the lens fibers for both proteins
(Figs. 2A 2C) , whereas no CP49 and greatly reduced filensin immunoreactivity were detected in the FVB/N lens
(Figs. 2B 2D) .
The mouse has long been a valuable model for vision research because many spontaneous mutants affecting visual function
12 are available and additional mutations influencing the visual system can be easily isolated in genetic screens.
13 14 15 The advantages of studies in the mouse have increased with the development of transgenic mouse technology
16 and the ability to completely or conditionally delete any gene of interest.
17 18 These manipulations are usually performed with inbred strains of mice that are essentially homozygous at every genetic locus because this allows study of the manipulation in a uniform genetic background.
19 However, it is essential that unmanipulated mice from the chosen inbred strain reflect the so-called normal biology of the system tested.
Numerous inbred mouse strains harbor endogenous defects in the visual system. Many genetically isolated inbred mouse strains, including FVB/N, CBA/J, C3H/HeJ, and SWR/J, harbor the rd1 allele of the
Pde6b gene, which results in rod photoreceptor degeneration.
6 The inbred strain DBA/2J develops pigment dispersion glaucoma
20 as a result of at least two mutant loci in these animals,
21 and C57BL/6 mice are susceptible to the development of congenital microphthalmia.
22 .
Since the earliest days of the study of lens biology with transgenic mice,
23 24 FVB/N—with its fecundity and ease of oocyte microinjection
5 —has been used extensively despite the compromise in its visual function attributed to the
Pde6b rd1 mutation.
6 Here we demonstrate that inbred FVB/N mice also harbor a
Bfsp2 deletion mutation identical to that in the inbred strains 129, 101, and CBA.
8 Notably, these inbred strains do not have a known common genetic ancestor. Rather, FVB/N mice were established as an inbred strain between 1966 and 1986 at the NIH from Swiss Webster stock, which does not share any known ancestry with the 101, 129, or CBA strains that arose from ancestors maintained in William Castle’s laboratory sometime at the beginning of the 20th century.
19 This suggests that the
Bfsp2 deletion mutation arose even earlier than previously proposed
8 and that it may, like the
Pde6b mutation,
6 be present in diverse wild mouse stocks. Alternatively, this locus may be a hot spot for mutation, with the same mutation arising spontaneously multiple times in mice.
Detailed analysis of lenses from 129 strains of mice mutant for
Bfsp2 have demonstrated that they lack CP49 protein and that filensin protein levels are greatly reduced, resulting in an absence of beaded intermediate filaments.
8 9 Similarly, we demonstrate here that FVB/N lenses lack CP49 immunoreactivity and that filensin immunoreactivity is greatly reduced. Given that both the in-frame deletion of a single amino acid and the missense mutation of the
BFSP2 gene result in congenital and juvenile-onset cataract in humans,
25 26 it is apparent that beaded filament proteins are essential for lens clarity, though the lens phenotype of the functionally null
Bfsp2 mutation in mice is more subtle.
8 9 Overall, the widespread use of FVB/N in the production of transgenic mice to study lens biology suggests that some of the phenotypes reported in the literature may result from interactions between the transgenic manipulation and the underlying lack of beaded filaments.
The best inbred mouse strain for studying ocular biology is far from clear since many of the most common strains harbor mutations in genes important for ocular function. Thus, it is important to keep these underlying defects in visual function in mind when choosing which inbred strain should be used for the study of any particular research question.