Knocking Down Snrnp200 Initiates Demorphogenesis of Rod Photoreceptors in Zebrafish

Purpose. The small nuclear ribonucleoprotein 200 kDa (SNRNP200) gene is a fundamental component for precursor message RNA (pre-mRNA) splicing and has been implicated in the etiology of autosomal dominant retinitis pigmentosa (adRP). This study aims to determine the consequences of knocking down Snrnp200 in zebrafish. Methods. Expression of the Snrnp200 transcript in zebrafish was determined via whole mount in situ hybridization. Morpholino oligonucleotide (MO) aiming to knock down the expression of Snrnp200 was injected into zebrafish embryos, followed by analyses of aberrant splicing and expression of the U4/U6-U5 tri-small nuclear ribonucleoproteins (snRNPs) components and retina-specific transcripts. Systemic changes and retinal phenotypes were further characterized by histological study and immunofluorescence staining. Results. Snrnp200 was ubiquitously expressed in zebrafish. Knocking down Snrnp200 in zebrafish triggered aberrant splicing of the cbln1 gene, upregulation of other U4/U6-U5 tri-snRNP components, and downregulation of a panel of retina-specific transcripts. Systemic defects were found correlated with knockdown of Snrnp200 in zebrafish. Only demorphogenesis of rod photoreceptors was detected in the initial stage, mimicking the disease characteristics of RP. Conclusions. We conclude that knocking down Snrnp200 in zebrafish could alter regular splicing and expression of a panel of genes, which may eventually trigger rod defects.

Recent genetic and functional studies have revealed the important role of pre-mRNA splicing in RP etiology and has shed light on the splicing process itself, a fundamental biological process [9,[12][13][14]. Pre-mRNA splicing is predominantly regulated by the spliceosome, a large complex that recognizes the pre-mRNA splice sites, removes the introns, and ligates the flanking exons accurately [15]. Notably, six of the seven identified adRP causative pre-mRNA splicing genes encode proteins embodied in the U4/U6-U5 tri-small nuclear ribonucleoproteins (snRNPs), a dynamic entity critical for the assembly and catalytic activation of the spliceosome, suggesting the important role of U4/U6-U5 tri-snRNP defects in the etiology and pathogenesis of RP [13]. The six genes can be further divided into two groups: U4/U6-specific genes including PRPF3, PRPF4, and PRPF31 and U5-specific genes including PRPF6, PRPF8, and SNRNP200. How RP mutations in these ubiquitously expressed genes lead to retinopathy is currently under debate [12].
SNRNP200, located on 2q11.2 (RP33 locus), encodes the RNA helicase hBrr2, a U5-specific protein that contains 2136 amino acids and catalyzes the U4/U6 unwinding [16]. We have previously established that SNRNP200 mutations do not compromise triple snRNP assembly but do compromise U4/U6 unwinding [9]. However, the retinal phenotypes and mRNA metabolism induced by knocking down SNRNP200 in animal models have never been characterized. Therefore, to reveal the potential link between knockdown of SNRNP200 and its relevant consequences, we studied the biochemical and morphological changes caused by knockdown of Snrnp200 in a zebrafish model.

Whole Mount In Situ Hybridization in Zebrafish.
All animal experiments were conducted in accordance with the Institutional Animal Care and Use Committee (IACUC) approved protocol and were approved and reviewed by the Institutional Committee of Nanjing Medical University. Zebrafish rearing and husbandry were maintained at 28.5 ∘ C with a 14-hour (hr) light/10 hr dark cycle. Digoxigeninlabeled antisense RNA probes were generated with cDNA of Snrnp200 (NM 001123257.1), analogous of human Snrnp200 in zebrafish, per the manufacturer's directions (Roche Applied Science, Mannheim, Germany). Whole mount in situ hybridization was conducted on 6 zebrafish for each developmental stage using a previously described modified protocol [17,18]. Primer information is detailed in Supplementary Table S1 (in Supplementary Material available online at http://dx.doi.org/10.1155/2015/816329).

Morpholino Oligos and Knockdown of Snrnp200 in
Zebrafish. Standard control morpholino oligos (control-MO: 5 -CCTCTTACCTCAGTTACAATTTATA-3 ) and zebrafish Snrnp200 splicing-blocking MO targeting exon 25 (Snrnp200-MO: 5 -TCAACATCAAGACAACTCACATCC-T-3 ) were purchased from Gene Tools (Philomath, OR, USA). MO has been widely used to interfere in the transcription or translation of a targeted gene resulting in loss-of-function of the gene [19,20]. Zebrafish embryos were injected at the 1-or 2-cell stage (0 day postfertilization (dpf)) with ∼1 nL of purified MOs dissolved in water. To get the best adjusted dosage for Snrnp200-MO, zebrafish embryos were divided into three groups and injected with 2 ( = 67), 4 ( = 68), and 8 ng ( = 65) of MOs, respectively. Another two groups were further obtained and were injected with control-MO (4 ng; = 72) and Snrnp200-MO (4 ng; = 68), respectively. The counting and the percentage calculation of deformation and death in each injected group were conducted from 2 to 4 dpf as described previously [12]. Embryos that died within 24 hr postfertilization were excluded because such death likely resulted from unspecific causes. At 4 dpf, embryos with relatively normal appearance were collected from each injected group for further investigations.

RT-PCR and Q-PCR.
RNA was isolated from 10 zebrafish embryos from each injection group, followed by reversetranscriptase polymerase chain reaction (RT-PCR) to generate cDNA templates [9,21]. PCR was subsequently conducted on the obtained cDNA to verify the effectiveness of Snrnp200-MO and amplify the targeted region of the cbln1 gene with primers detailed in Supplementary Table S1 using a previously defined protocol [9,21]. Quantitative real-time PCR (Q-PCR) was further performed using FastStart Universal SYBR Green Master (ROX; Roche, Basel, Switzerland) with the StepOnePlus Real-Time PCR System (Applied Biosystems, Darmstadt, Germany) per manufacturer's protocol. Expression analyses were conducted as described previously [22]. Primer information for Q-PCR was detailed in Supplementary Table S1.

Immunofluorescence Staining and Antibodies.
Twelve embryos from the uninjected group, 11 from the control-MO injected group, and 14 from the Snrnp200-MO injected group were harvested and cryopreserved per standard procedures. They were fixed in 4% paraformaldehyde, incubated with 30% sucrose, embedded with optimal cutting temperature solution, and frozen in liquid nitrogen for sectioning at 5 m using a Leica CM1900 cryostat (Leica, Wetzlar, Germany). Rod and cone photoreceptors were further visualized through immunofluorescence staining as indicated previously [12]. Briefly, cryosections were incubated with designated primary antibodies, including antirhodopsin (Mouse, 1 : 250; Abcam, Cambridge, UK) and zpr-1 antibodies (Mouse, 1 : 250; ZRIC, USA), to label rod and cone photoreceptors, respectively. The cryosections were then treated with fluorescence-conjugated secondary antibodies (Invitrogen, Carlsbad, CA, USA) for another 1 hr at room temperature and finally counterstained by 4 ,6-diamidino-2-phenylindole (DAPI; Sigma, USA) for cell nuclei staining. Images were taken with an Olympus IX70 confocal laser-scanning microscope (Olympus, Tokyo, Japan).

Statistics.
GraphPad Prism (version 4.0; GraphPad Software, San Diego, CA, USA) was applied for statistical analysis. We also use one-way ANOVA or Student's -test for comparisons among different groups. Data was presented as mean ± standard deviation (SD), and < 0.05 was taken as statistically significant.

Ubiquitous Expression of Snrnp200 in Zebrafish.
We have previously showed a ubiquitous expression of SNRNP200 in multiple human and murine tissues [9]. Whole mount in situ hybridization revealed overall expression of Snrnp200 in zebrafish among all developmental stages (Figure 1).

Knocking Down Snrnp200 Induces Aberrant Splicing in cbln1.
The efficacy of Snrnp200-MO was confirmed by RT-PCR ( Figure S1A) and the optimized condition for tissuespecific effects by Snrnp200-MO injection was defined with graded levels of concentration ( Figures S1B and S1C). Injection of Snrnp200-MO at 4 ng in zebrafish was proved as the optimization dosage for an effective knockdown of the Snrnp200 gene with moderate mortality (18%) and aberration (41.5%) rates (Figures 2(a), 3(a), and S1C). Previous studies demonstrated that patients with mutations in PRPF31, PRPF3, PRPF8, and PRPF6 failed to correctly remove the intron between the first and the second exons of the cbln1 gene [11]. Herein, the aberrantly spliced product (589 bp) containing the first intron of the cbln1 gene was also detected in zebrafish injected with Snrnp200-MO (Figure 2(b)).

Knocking Down Snrnp200 Triggers Upregulation of U4/ U6-U5 Tri-snRNP Components and Downregulation of Retinal Transcripts.
Knocking down a pre-mRNA splicing gene, like sart3, prpf31, or prpf4, in zebrafish can trigger compensatory responses by upregulating itself and other relevant splice components [12,14,23]. To test whether knocking down Snrnp200 will have similar effects, we performed Q-PCR on the cDNA templates obtained from zebrafish injected with control-MO or Snrnp200-MO, respectively. Increased expressions of prpf3 (NM 205748.1), prpf31 (NM 200504.1), prpf6 (NM 212655.1), and prpf8 (NM 200976.2) were found in embryos injected with Snrnp200-MO when compared with those injected with control-MO, suggesting that other splicing components will show compensatory responses to Snrnp200 deficiency in zebrafish (Figure 2(c)). In addition, consistent with previous findings in zebrafish with prpf31 and prpf4 knocked down [12,14], Q-PCR revealed that the expression levels of several important retinal transcripts including opn1lw1 (NM 131175.1), gnat2 (NM 131869.2), rs1 (NM 001003438.2), and rho (NM 131084.1) were decreased in Snrnp200 morphants when compared with the control group (Figure 2(d)).

Demorphogenesis of Rod Photoreceptors in Snrnp200
Morphants. We further analyzed whether knocking down Snrnp200 in zebrafish will lead to primary damages on retina, especially defects in rod and cone photoreceptors. Retinal phenotypes were investigated on zebrafish injected with control-MO or Snrnp200-MO at 4 dpf. To confirm the integrity of the eyeballs, thus minimizing the possibility for false positive results, we only include larvae with relatively normal systemic appearances from each injected group to visualize the morphology of cone and rod photoreceptors. Larvae with severely general morphological defects, including malformed brains, short trunks, cardiac edema, and curved body axis, were excluded. The rhodopsin activity was significantly reduced in Snrnp200-morphants (Figures 3(b)-3(d)), while no significant divergence was shown concerning the morphology of the cone photoreceptors between the control and the morphant eyes (Figures 3(e)-3(g)).

Discussion
Mutations in the ubiquitously expressed pre-mRNA splicing genes have long been implicated in the etiology of RP, while the specific mechanisms underlying how such mutations would cause retina-specific phenotypes have not been fully elucidated. Previous studies suggest that reduced expression levels of prpf31 and prpf4 in zebrafish would selectively affect gene expression, in particular, retina-specific genes [12,14]. However, both prpf31 and prpf4 are U4/U6-specific genes, and whether knocking down U5-specific genes will have similar impacts is still unknown. In the present study, we for the first time used a zebrafish model to characterize the retinal phenotypes and mRNA metabolism induced by knockdown of Snrnp200. Our findings in zebrafish indicate that knocking down of Snrnp200 would cause systemic defects, retinal phenotypes, and splicing anomalies. Similar to findings in zebrafish with sart3, prpf31, or prpf4 knocked down [12,14,23], upregulation of other U4/U6-U5 tri-snRNP components was detected in Snrnp200 morphants suggesting compensatory responses of other splicing components. Retinal phenotypes in the Snrnp200 morphants were also similar to previous findings in prpf31 and prpf4 morphants [12,14]. The expression levels of retina-specific transcripts were significantly reduced, and rod loss/demorphogenesis was predominantly detected in Snrnp200 morphants, mimicking the RP phenotypes in patients. In addition, consistent with previous finding in patients carrying mutations in PRPF31, PRPF3, PRPF8, and PRPF6, aberrant splicing of the cbln1 gene with retaining of its first intron was revealed in zebrafish injected with Snrnp200-MO [11]. Our result implies that defects in SNRNP200 and other pre-mRNA splicing genes would share common splicing abnormalities.
The hBrr2 protein is critical for the proofreading of pre-mRNA splicing [9]. Thus, SNRNP200 defects may interfere with the fidelity of retinal transcripts and further generate aberrantly spliced products toxic to the retina. Therefore, transcriptome analyses are needed for the detection of downstream target genes, particularly, retina-specific genes, of SNRNP200 mutations and mutations in other adRPassociated splicing genes. Identification of such target genes will definitely provide us with a better insight into the molecular mechanisms for pre-mRNA splicing defects and could lead to discovery of new pathways for RP, which would assist in genetic counseling and direct future gene therapy for the patients with pre-mRNA splicing deficiency.
Taken together, our data suggest the potential pathogenesis shared by defects in pre-mRNA splicing genes, in particular, defects in SNRNP200, PRPF31, and PRPF4. Mutations in PRPF31 have been reported to induce RP through haploinsufficiency [12], while PRPF4 mutations are found as RP causative in a dominant negative manner [12]. Therefore, we highly hypothesize that SNRNP200 mutations will cause RP eventually via loss-of-function, and investigations on the definite pathogenesis underlying SNRNP200 mutations will be part of our future work. Elucidating the pathogenesis for SNRNP200 mutations would aid in the development of novel therapeutics for retinitis pigmentosa. If haploinsufficiency mechanism is responsible for SNRNP200 defects, the primary aim should be restoration of the normal levels of wild-type hBrr2 proteins. On the other hand, if the dominant negative gain-of-function toxicity mechanism plays an important role in photoreceptor loss, then enhancing the clearance of the toxic mutant hBrr2 protein would be of therapeutic value if implemented early.