Retinal degeneration mutation in Sftpa1tm1Kor/J and Sftpd -/- targeted mice

Surfactant proteins are important collectin immune molecules with a wide distribution throughout the body, including the ocular system. Mice with gene deletions for the surfactant protein genes Sftpa1 and Sftpd were observed to have visual impairment and thinning of the outer nuclear layers of the retina. We hypothesized that gene deletion of Sftpa1 and Sftpd (Sftpa1tm1Kor/J and Sftpd-/-) results in early retinal degeneration in these mice. Sftpa1tm1Kor/J and Sftpd-/- retinas were evaluated by histopathology and optical coherence tomography (OCT). Retinas from Sftpa1tm1Kor/J and Sftpd -/- mice showed early retinal degeneration with loss of the outer nuclear layer. After screening of mice for known retinal degeneration mutations, the mice were found to carry a previously unrecognized Pde6brd1 genotype which resulted from earlier breeding of the strain with Black Swiss mice during their generation. The mutation was outbred and the genotype of Sftpa1tm1Kor/J and Sftpd-/- was confirmed. Outbreeding of the Pde6brd1 mutation resulted in restoration of normal retinal architecture confirmed by in vivo and in vitro examination. We can therefore conclude that loss of Sftpa1 and Sftpd do not result in retinal degeneration. We have now generated retinal Sftpa1 and Sftpd targeted mice that exhibit normal retinal histology.


Introduction
The surfactant proteins A and D (SP-A and SP-D) belong to the super family of collectins or "c-type lectins. Other collectins include mannose binding lectin (Mbl), pulmonary surfactant proteins (SP) SP-A and SP-D, and the conglutinins CL-L1 and CL-P1 [1]. The pulmonary surfactant proteins SP-A and SP-D play a key role in pulmonary immunity [2]. They are important in recognizing and mediating neutralization of viruses, bacteria and fungi, clearance of apoptotic and necrotic cells, and resolution of inflammation [3][4][5]. Recently, these molecules have been found in other organ systems aside from the pulmonary system [6] including vaginal and amniotic fluid, the gastrointestinal tract, renal system and the ocular system [7][8][9][10][11][12][13][14][15].
Surfactant protein targeted mice, Sftpa1 tm1Kor/J [4] and Sftpd targeted mice (Sftpd -/-) [5] were developed by the Whitsett /Korfhagen group at the University of Cincinnati Medical Center in order to study the effect of these proteins in the pulmonary system. Both have been extensively characterized regarding pulmonary phenotype and experiments using these mouse strains have been widely published and cited in the literature. Since their development in the early 1980's, pulmonary research groups had observed that these mice appeared to have visual disturbances and had blindness (unpublished data), which was attributed to deletion of the surfactant genes. This has never been comprehensively evaluated in the literature. These proteins are very closely associated with the fatty acid pool of surfactant in the pulmonary system. SP-A and D have both been shown to be important in the metabolism of surfactant in that they inhibit of secretion of surfactant from alveolar type II cells, enhance the uptake of diphosphatidylcholine (DPPC) by alveolar cells, they increase the adsorption of phospholipids and help to decrease surface tension in surfactant mixtures [2]. As the rods and cones in the retina have an abundance of fatty acids involved in the signal transduction process, SP-A and or D may impact fatty acids and/or lipase activity in rods and cones of the ONL.
We hypothesized that deletion of Sftpa1 and Sftpd results in early retinal degeneration. Both mouse strains were found to be express the Pde6b retinal degeneration mutation, Pde6b rd1 in addition to the targeting of Sftpa1 tm1Kor/J and Sftpd. Our goal is to report the previously unrecognized Pde6b rd1 genotype and phenotype in Sftpa1 tm1Kor/J and Sftpd -/mice. We report how the baseline retinal phenotype was examined, how the Pde6b rd1 genotype was identified and the strategy employed to outbreed this mutation from the mice. We then detail the confirmation of a "wild type" normal retinal phenotype by application of the same rigorous tests.

Methods
Our experimental approach is summarized here with details following. We first examined and compared the morphology of retinas harvested from Sftpa1 tm1Kor/J and Sftpd -/mice to those from wild type C57BL/6J mice by H&E stained sections and in vivo using optical coherence topography (OCT). PCR genotyping was performed in order to confirm Sftpa1 and Sftpd knock-down. Genotyping also identified the Pde6b rd1 mutation in both strains, Sftpa1 tm1Kor/J and Sftpd -/-. The Pde6b rd1 mutation was outbred and retinal morphology was again examined and compared by OCT evaluation as well as by histological examination of retinal cross sections.

Animals
C57BL/6J wild type mice (WT), and SP-A targeted Sftpa1 tm1Kor/J mice were purchased from Jackson Laboratories (Bar Harbor ME). Sftpd -/mice were a kind gift from the Whitsett /Korfhagen lab in Cincinnati, OH. All mice were maintained on a 12-hour light/dark cycle and fed standard mouse chow. The University of Oklahoma Health Sciences Center Institutional Animal Care and Use Committee specifically approved the research protocols associated with this study (IACUC #101341-16-033 and 101005-14-081).

Examination of microscopic morphology of the retina
Mice (N = 6) were euthanized by CO 2 asphyxiation. The eyecup was enucleated whole and fixed in PreFer fixative (Anatech Ltd, Battlecreek, MI) for 15 min at room temperature. It was then embedded in paraffin and sectioned at 5 μm onto glass slides. Retinal tissue sections on glass slides were deparaffinized and then incubated with hematoxylin, blue buffer and then incubated with eosin. Microscopic examination was performed using the Olympus IX71microscope (Melville, NY).

Examination of gross morphology of the retina in vivo using OCT
Mice from each strain (N = 6 per strain) were anesthetized with an intraperitoneal injection of a ketamine/xylazine (50-75/1-2 mg/kg) cocktail. Eyes were dilated with tropicamide 1% and phenylephrine 2.5% drops and lubricated with methylcellulose. Spectral-domain optical coherence tomography images were taken with the Envisu R2000 (Bioptigen, Morrisville, NC, USA) at 6 weeks' postnatal age. Total retinal thickness was measured and compared for all strains by Students t test with a significant p value < 0.05.

Identification of genetic mutation
Mouse genotyping was performed in order to confirm Sftpa1 and Sftpd deletion. Next, a literature review was performed to identify retina mutations known to cause early degeneration in mice. The mutations, retina degeneration 1 (Pde6b rd1 ) and retina degeneration 8 (rd8 -/-) and Rpe65 gene (Rpe65(450Leu) or Rpe65(450Met)) [16] were identified.

PCR genotyping
Genotyping was performed for the mutations of interest using OneTaq Quick-Load 2X Master Mix (M0486) from New England Biolabs (Ipswich, MA) and oligonucleotides ordered from Integrated DNA Technologies (Coralville, IA). DNA was extracted from ear tissue by boiling it with 50 μL 50 mM NaOH for 60 min at 95˚C, followed by neutralization at room temperature with 25 μL of 100 mM Tris/HCl buffer pH 8.0 containing 10 mM EDTA. 0.5 μL of the supernatant after centrifugation served as DNA template for a 12.5 μL PCR mix. A typical PCR protocol involved an initial denaturing step at 94˚C for 30 seconds followed by 30 cycles of 30 seconds at 94˚C, 30 seconds at 64˚C and 60 seconds at 68˚C. The protocol concluded with a final extension at 68˚C for 5 minutes. Primer sequences are found in the S1 Table. The primer sequences for Sftpa1 were taken from the Jackson lab protocols "Sftpa1tm1Kor" (www.jax. org). Integrities of Pde6b rd1 and rd8 were verified using the genotyping methods Blazek et al. [17] and of Mattapallil et al. [18]. The RPE65 genotype was confirmed to be solely homozygous Rpe65 450 Met by the method of Grimm et al. [19] using the primers RPE65For/RPE65Rev. The mannose lectin binding (Mbl) gene lies in between both the Sftpa1 and Sftpd gene and is also an important collectin protein. In order to confirm that Sftpa1 and Sftpd gene targeting did not disrupt immediately adjacent sequences, the integrity of the Mbl1 gene and the 3'ends of the Sftpa1 and Sftpd genes were further verified by PCR using primers MmMbl1F1/ MmMbl1Rev, SPAex6For/SPAex6Rev and SPDex8For/SPDex8Rev.

Outbreeding the Pde6b rd1 mutation
The Sftpa1 and Sftpd mice were mated with C57BL/6J rd1 +/+ mice (WT mice) in order to outbreed Pde6b rd1 from both strains. The F1 generation was genotyped and rd1 +/+ genotypes of SP -/as well as SP +/were selected and mated again for both mutants. Sftpa1 -/-rd1 +/+ was confirmed by F4, Sftpd -/-rd1 +/+ was confirmed by F2. The mice were then bred for further six consecutive generations. The resulting genotype was validated for each generation by PCR as well as by histology with cross sections of retina tissues.

Retinal morphology in Sftpa1 tm1Kor/J and Sftpd -/mice
We began our phenotypic assessment by performing microscopic H&E evaluation of retinal cross sections from each group. Retinas from Sftpa1 tm1Kor/J and Sftpd -/at 3 months of age showed diminished retinal thickness as compared to WT C57BL/6J mice. Fig 1 shows the histological morphology as compared by H&E stains. At 3 months of age, loss of the outer nuclear layers (ONL) is observed in both Sftpa1 tm1Kor/J and Sftpd -/strains as compared to WT. We then performed in vivo comparison of retinal morphology by OCT, as shown in Fig 2  panel a). OCT evaluation showed a loss of the ONL as well as a significant decrease in total  retinal thickness (WT 0.2μm vs Sftpa1 -/-, Pde6b rd1 0.12, p = 0.001) as depicted in the bar graph in Fig 2, panel b).

Genotyping of mouse strains to confirm mutations
The phenotype of the loss of ONL was reminiscent of that seen in mouse strains with early retinal degeneration mutations i.e. those targeting the Pde6b gene. The Pde6b rd1 mutation had been reported to occur in C57Bl/6J mice as had the rd8 mutation, therefore, these genetic variants were targeted first. PCR genotyping confirmed that the mouse strains contained the gene manipulation of interest, i.e. Sftpa1tm1Kor/J and Sftpd -/-. PCR transcripts targeting the 5' and 3'-ends of the genes confirmed integrity of the beginning and end of the gene segments as shown in Fig 3a) for Sftpa1 and Fig 3b) for Sftpd. The gene between the two surfactant protein  4 and 6), respectively. Primer sequences are found in the S1 Table. https://doi.org/10.1371/journal.pone.0199824.g003 genes encodes the lectin, mannose binding lectin (Mbl) known to have similar properties as Sftpa1 and Sftpd, therefore, the 5'-and 3'-ends of the Mbl gene were confirmed to be intact as shown Fig 3 panel c). Genotyping for Pde6b rd1 is shown in Fig 4. As the rd8 mutation (retinal degeneration 8) and the variant of Rpe65 are also of concern to scientists studying retinal biology, confirmation of genotype was performed. All used mice were found to have the rd8 +/+ genotype and carry the Rpe65 protein variant with Rpe65 450 Met (methionine at position 450). Representative PCR transcripts are shown in Fig 4. On successful outbreeding of the Pde6b rd1 mutation, final genotypes were [Sftpa1 -/-, rd1 +/+ , rd8 +/+ , Mbl1 +/+, Rpe65 450 Me] and [Sftpd -/-, rd1 +/+ , rd8 +/+ , Mbl1 +/+, Rpe65 450 Me].

Retinal morphology in mice with rd1 +/+ genotype
Once Pde6b rd1 was outbred, we again analyzed the histological archetype of the retinas by H&E stained retinal cross sections in order to confirm if the loss of ONL had been secondary to the Pde6b rd1 mutation or if it was secondary to absence of the surfactant proteins. Retinal histological morphology was identical and comparable in [Sftpa1 -/-, rd1 +/+ , rd8 +/+ , Mbl1 +/+, Rpe65 450 Me] and [Sftpd -/-, rd1 +/+ , rd8 +/+ , Mbl1 +/+, Rpe65 450 Me] with that of C57BL/6J mice as shown in Fig 5. All strains show normal retinal layer differentiation with a healthy ganglion cell body layer and appropriate nuclear bodies in the inner and outer nuclear layers. The RPE is also of normal appearance. Similarly, OCT also showed restoration of total retinal thickness.

Discussion
Pulmonary biologists had anecdotally observed that the vision of Sftpa1 -/and Sftpd -/mice appeared compromised. It was therefore hypothesized that the surfactant gene deletion resulted in retinal abnormalities/degeneration and resultant blindness. We can now confirm that the absence of SP-A and SP-D proteins in Sftpa1 tm1Kor/J and Sftpd gene targeted mice, do not lead to early retinal degeneration but is a result of the Pde6b rd1 mutation on the Pde6b gene. This retinal degeneration causing mutation was previously unknown to the authors who first described the gene targeted mice, nor to the commercial vendors who now carry these mouse lines. The detection, outbreeding of the mutation, and generation of mice with the said Sftpa1 and Sftpd gene deletion but with intact Pde6b gene expression (rd1 +/+ ) is reported here. This important finding bears reporting as these mouse strains may be used for study of disease processes unrelated to the pathology for which they were first created.
Our initial hypothesis regarding the potential impact of surfactant proteins A and D on the ONL and rod and cone function was based on the premise that pulmonary SP's are known to bind to fatty acids including saturated and unsaturated fatty acids of various chain lengths [20]. Furthermore, SP-A has been shown to inhibit phospholipase activity in vitro [21]. Interestingly, in studies employing Sftpa1 -/mice (Sftpa1 -/-rd1 +/+ generated by us), we showed that when retinal architecture is preserved that absence of SP-A protein impacts retinal neovascularization in a mouse model of oxygen-induced retinopathy (OIR) [7]. Thus, it appears that the SP-A protein plays a key role of directing early angiogenic/vascular processes but may not impact the neuronal function of the retina. We were successful in outbreeding the Pde6b rd1 mutation, restoring normal retinal architecture that allows us to study the impact of SP-A and SP-D on retinal vascular disease.
Several mouse strains with deletions in pulmonary surfactant proteins have been described in the literature, which are listed in the Knockout Mouse Project Repository (KOMP) and/or in the Jackson Labs Repository. These include the SP-A targeting Sftpa1 tm1Kor/J mouse, the Sftpa1 tm2Haw mouse, the Sftpa1 tm1(KOMP)Vlcg mouse, the conditional cre/lox targeted Sftpa1 tm1a(KOMP)Wtsi . Additionally, a double knockout B6.Cg-Sftpa1 tm2Haw Sftpd tm2Haw /J has also been documented. Many studies have been identified in which the Sftpa1 tm1Kor/J mouse strain was used in organ systems other than the lungs and airways [3,10,15,22]. When generated, these gene-targeted mice are typically characterized regarding the organ system for which they were created. The performance of body wide phenotypic or genotypic analysis is usually not feasible.
In order to determine how the Pde6b rd1 mutation may have been introduced in the mouse strains of interest, we reviewed the methods employed by Korfhagen and colleagues in their initial generation of these mice. The SP-A targeted Sftpa1 tm1Kor/J mouse strain was developed by Korfhagen and colleagues in 1996. The gene targeting strategies of the SP-A -/and SPD -/mice have previously been described, in detail [4]. In short, a part of the Sftpa1 gene was cloned into pUC18 vector and blunt-end ligated with a neo cassette from the vector pGKneoBPA, so that parts of exons 3 and 4 were disrupted. Herpes simplex virus-thymidine kinase (HSV-TK) was cloned into the Puc18 vector downstream of the insert. Embryonic Stem (ES) cells generated from 129-Ola mice were electroporated with the Sftpa1 targeting construct. Targeted ES cells were selected by culture. Clone 87, a rapidly proliferative, undifferentiated ES cell clone, was microinjected into C57B1/6J host blastocysts. Chimeric males were then bred to NIH Swiss black females. A similar approach was used for generation of Sftpd -/mice. The Sftpd gene was cloned into the vector with a neo cassette from pGKneoBPA, so that parts of exon 2 and intron 2 were disrupted. These sequences were electroporated onto ES cells and microinjected into C57/B16 host blastocysts. Chimeric males were then bred to NIH Swiss black females. Subsequent breeding was all performed on a C57Bl/6J background, which have also been used as control animals for experiments. Of note, the commercial website (Jax.org) on which these mice are now carried, notes the mouse strain of origin as 129/OlaHsd. Black Swiss mice are not mentioned in the description of the genetics of the mouse. In 2005, it was reported that the Black Swiss and NIH Swiss mouse both had visual and cognitive disturbances which was shown to be secondary to homozygous Pde6b rd1 mutation [23]. Several other commonly used mouse strains have subsequently been found to also carry the homozygous rd1 mutation [24,25].
Mice, homozygous for Pde6b rd1 are characterized by early onset, severe retinal degeneration [26,27]. The Pde6 gene is important for amplification of visual signals via normal function of the photoreceptors in the retina. The beta subunit (Pde6b) is crucial for proper signaling of the gene [28]. The Pde6b rd1 mutation occurs as a result of a murine viral insert [28] and a second nonsense mutation in exon 7 of Pde6b [25]. Yu et al showed that the nonsense mutation, Y347X, was responsible for the clinical presentation of the disease [29].
The importance of thorough analysis of the background of mouse strains and potential for unwanted/unknown mutations is present and risky, as it may compromise the results of all downstream experiments. Unintended and surprise mutations can often be missed. This can sometimes be phenotypically obvious but may be subtle and only apparent at a biochemical level. Some may only appear during developmental stages of the organism (mouse). While genome wide sequencing prior to the use of mice in experiments would be ideal, this is not practical or possible due to the prohibitive high cost. This leaves researchers with a dilemma as there is a risk that experimental results may be tainted due to a mutation or gene variant that has not yet been detected in the strain or has not yet been found to play a role in their field of interest. Caution should be exercised to refrain from over ascribing a phenotype to a particular allele when that phenotype extends beyond the known function of the gene.
In conclusion, the Pde6b rd1 mutation is present in Sftpa1 tm1Kor/J and Sftpd -/mice resulting in early, complete retinal degeneration. We were successful in outbreeding of the mutation and were able to confirm normal phenotype in the retinas. The generation of Sftpa1 and Sftpd targeted mice with normal retinal histology will allow us to better characterize the impact of these surfactant proteins on retinal/visual pathways.
Supporting information S1 Table. Sequences of used oligonucleotides. All the primers (forward and reverse) used for genotyping are listed here with the descriptions of targeted sequences and details of transcripts. (DOCX)