Data on the generation of two Nr2e3 mouse models by CRISPR / Cas9D10A nickase

NR2E3 encodes an orphan nuclear receptor that plays a dual function as both transcriptional activator and repressor in photoreceptors, being necessary for cone fate inhibition as well as rod differentiation and homeostasis. Mutations in this gene cause retinitis pigmentosa (RP), enhanced S cone syndrome (ESCS) and Goldmann-Favre syndrome (GFS). There is one reported Nr2e3 isoform that contains all 8 exons and a second –previously unreported– shorter isoform, which only spans the first 7 exons and whose function is still unknown. In this data article, we designed and generated two new mouse models by targeting exon 8 of Nr2e3 using the CRISPR/Cas9-D10A nickase in order to dissect the role of the two isoforms in Nr2e3 function and elucidate the different disease mechanisms caused by NR2E3 mutations. This strategy generated several modified alleles that altered the coding sequence of the last exon thereby affecting functional domains of the transcription factor. Allele Δ27 is an in-frame deletion of 27 bp that ablates the dimerization domain, whereas allele ΔE8 (full deletion of exon 8), produces only the short isoform that lacks the dimerization and repressor domains. Morphological and functional alterations of both Δ27 and ΔE8 mutants are reported in the associated research article “Nr2e3 functional domain ablation by CRISPR-Cas9D10A identifies a new isoform and generated Retinitis Pigmentosa and Enhanced S-cone Syndrome models” (Aísa-Marín et al., 2020).

a b s t r a c t NR2E3 encodes an orphan nuclear receptor that plays a dual function as both transcriptional activator and repressor in photoreceptors, being necessary for cone fate inhibition as well as rod differentiation and homeostasis. Mutations in this gene cause retinitis pigmentosa (RP), enhanced S cone syndrome (ESCS) and Goldmann-Favre syndrome (GFS). There is one reported Nr2e3 isoform that contains all 8 exons and a second -previously unreported-shorter isoform, which only spans the first 7 exons and whose function is still unknown. In this data article, we designed and generated two new mouse models by targeting exon 8 of Nr2e3 using the CRISPR/Cas9-D10A nickase in order to dissect the role of the two isoforms in Nr2e3 function and elucidate the different disease mechanisms caused by NR2E3 mutations. This strategy generated several modified alleles that altered the coding sequence of the last exon thereby affecting functional domains of the transcription factor. Allele 27 is an inframe deletion of 27 bp that ablates the dimerization domain, whereas allele E8 (full deletion of exon 8), produces only the short isoform that lacks the dimerization and repressor domains. Morphological and functional alterations of both

Value of the Data
• The presented data give a detailed account of the generation of two mouse models caused by mutations in exon 8 of Nr2e3 using the CRISPR/Cas9 system. The Cas9-D10A nickase (requiring the use of 4 different guide RNAS) was used to prevent potential off-target effects. • These data may provide useful information for researchers who seek better understanding of CRISPR/Cas9-D10A modifications (using the nickase mutant enzyme) and the generation of new mouse models, as well as researchers interested in using the novel 27 and the E8 Nr2e3 mouse models because of their resemblance to human-related phenotypes. • These data may stimulate further investigations comparing the molecular mechanisms of the two diseases caused by mutations in NR2E3 , Retinitis Pigmentosa and Enhanced S-cone Syndrome, in humans and mice. The novel mice models may also provide an instrumental tool for evaluating disease progression and therapeutic efficacy since the E8 mutant is the first model of Retinitis Pigmentosa caused by mutations in Nr2e3 . From the technical point of view, other researchers may be interested in comparing methods for minimizing potential off-target effects of the full Cas9, or in the efficiency of the Cas9 D10A nickase for the generation of medium-large size deletions.

Data Description
In this data article, we first designed the deletion of exon 8 of Nr2e3 using the CRISPR-Cas9 system. The CRISPR-Cas9 system, which is widely used to target genes and generate modifications in the genome, has two components: a cis proto-space adjacent motif (PAM) next to the target site, and a trans guide RNA (gRNA) complementary to the target DNA. gRNAs are normally designed to be specific for the desired target site, however, in large genomes such as those of rodents, there are often similar sequences with a few mismatches that may be recognized by Cas9, thereby representing potential off-target effects. Actually, off-target effects are the main concern of the CRISPR-Cas9 system usage, as it may negatively affect experimental results, especially in clinical applications [2][3][4][5][6][7][8][9][10][11][12] . We thus explored the use of the mutant variant Cas9 D10A, which produces nicks on one strand of the DNA. In this case and to induce one double strand break (DSB), two gRNAs closely located in the genome target site (each with its PAM sequence motif nearby) have to be designed. If the gene-editing design intends a medium/large size deletion, then four different gRNAs have to be designed (two per each flanking DSB), which strongly minimize the probability of undesired off-target site cuts [13][14][15][16][17] .
To generate the genetically modified mice, embryos from pregnant donor females were extracted, manipulated in vitro and then transferred to pseudo-pregnant receptor females ( Fig. 1 ). As aforementioned, we opted for the use of the D10A Cas9 nickase and therefore, we used a total of 4 guide RNAs to delete exon 8 (guide design and position in Fig. 2 ). These four gRNAs were microinjected together with the mRNA encoding the Cas9D10A nickase into in vitro fertilized zygotes. In order to identify potentially modified alleles, we first used the T7 endonuclease assay, but immediately designed a highly specific and discriminative PCR test (specific primers and electrophoresis gels with high percentage agarose, 3%), which allowed the detection of different Fig. 1. Generation of gene-edited mice using the CRISPR/Cas9-D10A system. Embryos from pregnant donor females were extracted in a one-or two-cell state and were microinjected with the four gRNAs and the Cas8-D10A RNA. Receptor females were mated with vasectomized males, which present an escision of the testis vas deferens, to achieve the state of pseudo-pregnancy. The modified embryos were transferred to the pseudo-pregnant receptor females, thus obtaining mosaic mice bearing different deleted alleles. size alleles and heteroduplexes. The presence of gene-edited alleles in mosaic mice (shown in Figs. 3 and 4 ) were detected by PCR and confirmed by Sanger sequencing. Most of the geneedited alleles were only modified at the junction of intron 7 with exon 8 ( Fig. 3 A and 3 B ). Only one allele was modified containing the whole deletion ( Fig. 4 A and 4 B ) and no modifications were found only affecting in the 3 UTR region ( Fig. 4 C ), which strongly suggest that the 4 gRNAs were not equally efficient in targeting their sites.   From all the edited alleles, we selected alleles 27 (a short in-frame deletion at the 5 target site) and E8 (the designed medium size deletion of exon 8) for further analysis. By subsequent crosses, we generated heterozygous and homozygous strains of these two selected alleles, whose phenotypic and molecular characterization is described in the related research article [1] . In addition, potential off-target effects were analyzed using the Zhang Lab Tools for Guide Design ( https://zlab.bio/guide-design-resources ) ( Fig. 5 ). Specific primers were designed and assayed in all the gene-edited animals, but no additional off-target site editing was detected.

Generation of gene-edited mice using the CRISPR/Cas9 system
The CRIPSR/Cas9 system was used to generate a Nr2e3 mouse model by deleting the exon 8 of the locus. To generate this Nr2e3 mutant mice, zygotes from pregnant donor females were extracted, then modified and finally transferred to pseudo-pregnant receptor females ( Fig. 1 ). Superovulation of donor females was achieved by administering 5IU PMSG (Pregnant Mare's Serum Gonadotropin) and, 47 h later, 5IU hCG (human Chorionic Gonadotropin) via intraperitoneal injection. Immediately after hGC administration, females were mated with males [18] . 24-48 h post-mating, several murine zygotes (in a one or two-cell state) from pregnant donor females were extracted [19] and microinjected with a number of guide-RNA and the endonuclease Cas9-D10A mRNA. To minimize potential off-targets, D10A Cas9, one of the nickase mutants of Cas9, was used. Four guides were designed, two guides per deletion site, in such a way to ensure single strand breaks in the targeted acceptor site of intron 7 and 3 UTR region of Nr2e3 locus ( Fig. 2 ) and minimize potential off-target effects. A schematic representation of the gRNA location in the Nr2e3 locus can be found in the related research article [1] ). Microinjected zygotes were subsequently transferred to pseudo-pregnant receptor females [ 20 , 21 ]. To induce the pseudo-pregnant state in the receptor females, they were crossed with vasectomized males, which presented an excision of the testis vas deferens [22] and were capable to mate but unable to fertilize the female. The copulation induced a pseudo-pregnancy state in the receptor females, optimal for the embryo implantation and gestation. All embryonic procedures up to the generation of the chimaera founder mice ( Fig. 1 ) were performed at the Mouse Mutant Core Facility, Institute for Research in Biomedicine (Barcelona, Spain).

Characterization of the offspring modified alleles
The offspring obtained was genotyped to characterize the modified alleles as well as to detect off-targets, if any, generated by the system. PCR products were electrophoresed in a resolutive high concentration agarose gel (3% agarose, composed of 1.5% normal agarose and 1.5% low melting agarose) that allowed to resolve different size alleles without the need to perform T7 Endonuclease I Assay. When PCR bands for CRISPR-edited alleles were identified, the deletion was validated by Sanger sequencing. To confirm that the CRISPR-Cas9 technique did not introduce any off-target deletion/mutation, all the potential off-target sequences (up to three mismatches with sgRNA) determined by a prediction software ( https://zlab.bio/guide-design-resources ) were analyzed ( Fig. 5 ). Subsequently, the mice bearing alleles of interest were selected and crossed to obtain murine heterozygous and homozygous lines for the different mutations [23] . Mutants with a partial or complete deletion ( Fig. 3 and Fig. 4 ) of the exon 8 and 3 -UTR regions were selected.

Genotyping
Mouse genomic DNA was isolated from ear biopsies following overnight digestion at 55 °C in a lysis buffer with proteinase K. DNA amplification by PCR was used to genotype the mouse colony. Primer pairs were used to discern between Nr2e3 27 allele ( Nr2e3 intron 7 Fw and exon 8 Rv), E8 allele ( Nr2e3 intron 7 FW and down Rv), and WT alleles (primer sequences are detailed in the related research article [1] ). Founder mice bearing the selected alleles were subsequently mated for heterozygous F 1 offspring. After genotyping, animals carrying the selected allele were further mated until obtaining an F 2 or F 3 homozygous lineage to perform phenotypic studies.

Ethics Statment
Animal handling, euthanasia and surgical dissection was performed according to the ARVO statement for the use of animals in ophthalmic and vision research, following the guidelines for animal care of the University of Barcelona and with the approval of the Bioethics Committee of the University of Barcelona (File references FUE-2019-00965313, ID 2MDLDY4WZ).

Declaration of Competing Interest
The authors declare that they have no conflict of interest.