Abstract
Retinitis pigmentosa is a clinically heterogeneous disease affecting 1.6 million people worldwide. A growing number of identified disease-causing genes are associated with the spliceosome, but the molecular consequences that link defects in splicing factor genes to the aetiology of the disease remain to be elucidated. In this paper, we present a Drosophila model for Retinitis pigmentosa 11, a human disease caused by mutations in the splicing factor PRPF31. Here, we induced mutations in the Drosophila orthologue Prp31. Mutant flies are viable and show a normal eye phenotype when kept under regular light conditions. However, when exposed to constant light, photoreceptors of mutant flies degenerate, thus resembling the human disease phenotype. Degeneration could be shown to be associated with increased oxidative stress. This increase was in agreement with severe dysregulation of genes involved in oxidation/reduction processes, as revealed by high throughput transcriptome sequencing. In fact, light induced photoreceptor cell degeneration could be attenuated by experimentally reducing oxidative stress. A comparable decrease in retinal degeneration was achieved by raising mutant larvae on a vitamin A-depleted medium, thereby reducing rhodopsin levels. Finally, transcriptome data further uncovered an overall retention of introns in mRNAs. Among those, mRNAs of genes involved in synapse assembly, growth and stability were most prominent. These results point to a multifactorial genesis of light induced degeneration in retinae of Prp31 mutant flies, including transcriptional and splicing dysregulation, oxidative stress and defects in vitamin A metabolism.
Introduction
Retinitis pigmentosa (RP; OMIM 268000) is a clinically heterogeneous set of retinal dystrophies, which affects about 1.6 million people worldwide. It often starts with night blindness in early childhood due to the degeneration of rod photoreceptor cells (PRCs), continues with the loss of the peripheral visual field caused by degeneration of cones (tunnel vision), and progresses to complete blindness in later life. RP is a genetically heterogeneous disease, and can be inherited as autosomal dominant (adRP), autosomal recessive (arRP) or X-linked (xlRP) disease. So far >50 genes have been identified that are causally related to non-syndromic RP (Daiger et al., 2014) (see RetNet: http://www.sph.uth.tmc.edu/RetNet/disease.htm). Affected genes are functionally diverse. Some of them are expressed specifically in PRCs and encode, among others, transcription factors (e. g. CRX, an otx-like photoreceptor homeobox gene), components of the light-induced signalling cascade, including the visual pigment rhodopsin (Rho/RHO in Drosophila/human), or genes controlling vitamin A metabolism (e.g. RLBP-1, encoding Retinaldehyde-binding protein). Other genes are associated with the control of cellular homeostasis, for example CRB1, a gene required for the maintenance of polarity. Interestingly, the second-largest group of genes causing adRP, comprising 7 of 23 genes known, encodes regulators of the splicing machinery. So far, mutations in five PRPF (premRNA processing factor) genes, PRPF3, PRPF4, PRPF6, PRPF8 and PRPF31, have been linked to adRP, namely RP18, RP70, RP60, RP13 and RP11, respectively. PAP1 (Pim1-associated protein) and SNRNP200 (small nuclear ribonuclearprotein-200), two other genes involved in splicing, have been suggested to be associated with RP9 and RP33, respectively (Maita et al., 2004; Zhao et al., 2009) [reviewed in (Liu and Zack, 2013; Mordes et al., 2006; Poulos et al., 2011; Ruzickova and Stanek, 2016)]. The five PRPF genes encode components regulating the assembly of the U4/U6.U5 tri-snRNP, a major module of the pre-mRNA spliceosome machinery (Will and Luhrmann, 2011). Several hypotheses have been put forward to explain why mutations in ubiquitously expressed components of the general splicing machinery show a dominant phenotype only in PRCs. One hypothesis suggests that PRCs with half the copy number of genes encoding general splicing components cannot cope with the elevated demand of RNA-/protein synthesis required to maintain the exceptionally high metabolic rate of PRCs in comparison to other tissues. Hence, halving their gene dose eventually results in apoptosis. Although this model is currently favoured, other mechanisms, such as impaired splicing of PRC-specific mRNAs or toxic effects caused by accumulation of mutant proteins have been discussed and cannot be excluded to contribute to the disease phenotype [discussed in (Mordes et al., 2006; Scotti and Swanson, 2016; Tanackovic et al., 2011)].
The observation that all adRP-associated genes involved in splicing are highly conserved from yeast to human allows to use model organisms to unravel the genetic and cell biological functions of these genes, which ultimately will provide a mechanistic characterization of the origin of the diseases. In the case of RP11, the disease caused by mutations in PRPF31, three mouse models have been generated by knock-in and knock-out approaches. Unexpectedly, mutant mice did not show any sign of retinal degeneration (Bujakowska et al., 2009). Further analyses revealed that the retinal pigment epithelium, rather than the PRCs, is the primary tissue affected in Prpf31 heterozygous mice (Farkas et al., 2014; Graziotto et al., 2011). Morpholino-induced knock-down of zebrafish Prpf31 results in strong defects in PRC morphogenesis and survival (Linder et al., 2011). Defects obtained by retina-specific expression of zebrafish Prpf31 constructs that encode proteins with the same mutations as those mapped in RP11 patients (called AD5 and SP117, respectively) were explained to occur by either haplo-insufficiency or by a dominant-negative effect of the mutant protein (Yin et al., 2011). In Drosophila, no mutations in the orthologue Prp31 have been identified so far, but RNAi-mediated knock-down of Prpf31 in the developing Drosophila eye induced, besides strong developmental defects of the eye, signs of PRC degeneration (Ray et al., 2010)
In order to establish a meaningful Drosophila model for RP11-associated retinal degeneration, we isolated two mutant alleles of Prp31, Prp31P17 and Prp31P18, which carry missense mutations that result in exchanges of conserved amino acids. Flies heterozygous for either of these mutations are viable and develop normally. Strikingly, when exposed to constant light, mutant flies undergo retinal degeneration. Degeneration of mutant PRCs was associated with increased oxidative stress. Consistent with this, transcriptome analyses from heads of Prp31P18 homozygous flies showed transcriptional dysregulation of genes involved in oxidation/reduction processes. In addition, an overall retention of introns in mRNAs was observed. Retinal degeneration could be ameliorated by supplementing the food with NSC23766, a known inhibitor of NADPH (nicotinamide adenine dinucleotide phosphate)-oxidase activity, or by raising larvae on a vitamin A-depleted medium. From these results, we conclude a multifactorial genesis of Prp31-linked retinal degeneration in flies.
Results
Flies heterozygous for mutations in Prp31 undergo light-dependent retinal degeneration
It was recently shown that RNAi-mediated knockdown of Drosophila Prp31 in the eye using eyeless (ey)-Gal4 or GMR-Gal4 results in smaller eyes or no eyes at all and degeneration of photoreceptor cells (PRCs) and pigment cells (Ray et al., 2010). Since eyeless is expressed already in the early eye imaginal disc prior to PRC differentiation, some of the defects observed could be secondary, for example as a consequence of defective cell fate specification.
To establish a more meaningful Drosophila model for RP11-associated retinal degeneration, which would allow a deeper insight into the role of this splicing factor in the origin and progression of the disease, we set out to isolate specific mutations in Drosophila Prp31 by TILLING (Targeting Induced Local Lesions IN Genomes), following a protocol described recently (Spannl et al., 2017). In total, 2.400 genomes of EMS (ethyl methanesulfonate)- mutagenized flies were screened for sequence variants in two different amplicons of Prp31. Four sequence variants were identified, which were predicted to result in potentially deleterious missense mutations. Two of the four lines were recovered from the living fly library and crossed for three generations to wild-type (w1118) flies to reduce the number of accompanying sequence variants. In two of the lines, named Prp31P17 and Prp31P18, mutations in Prp31 could be verified. Prp31P18 was viable as homozygotes or in trans over Df(3L)Exel6262, which removes, among others, the Prp31 locus. In contrast, no homozygous Prp31P17 flies were obtained. However, Prp31P17 was viable in trans over Prp31P18 and over Df(3L)Exel6262. This suggests that the lethality was due to a second site mutation, which was not removed despite extensive out-crossing. The molecular lesions in the two alleles were mapped in the protein coding region of Prp31. Drosophila PRP31 is a protein of 501 amino acids, which contains a NOSIC domain (named after the central domain of Nop56/SIK1-like protein), a Nop (Nucleolar protein) domain required for RNA binding, a PRP31 _C-specific domain and a nuclear localization signal, NLS. Prp31P17 contained a point mutation that resulted in a non-conservative glutamine to arginine exchange (G90R) N-terminal to the NOSIC domain. Prp31P18 contained a non-conservative exchange of a proline to a leucine residue in the Nop domain (P277L) (Fig. 1A and Supplementary Fig. S1).
Homo- and hemizygous Prp3P18, hemizygous Prp3P17 flies as well as Prp31P17/Prp31P18 transheterozygous flies kept under normal light/dark cycles have eyes of normal size. Histological sections revealed normal numbers of PRCs (distinguished by the number of rhabdomeres) per ommatidium and a normal stereotypic arrangement of PRCs (Fig. 1B-F). This indicates that the development of the retina was not affected by these mutations. However, PRCs of Prp31P17/+, Prp31P18/+ and Prp31P18/ Prp31P18 flies showed severe signs of retinal degeneration when exposed to constant light for several days. The same phenotype was observed in PRCs of Prp31P17 or Prp31P18 hemizygous flies as well as in Df(3L)Exel6262/+ flies (Fig. 2B-C’ and data not shown). After seven days of light exposure, the majority of rhabdomeres of the six outer PRCs, R1-R6, were either completely gone or were strongly reduced in size, and many cell bodies exhibited the typical signs of degeneration, such as condensed chromatin and electron dense material with vesiculation and multivesicular bodies (Fig. 2B-C’; quantification in Fig. 2G). No significant differences were observed between the two alleles. R7 is preserved in most ommatidia. Strikingly, PRCs of transheterozygous Prp31P17/Prp31P18 flies showed less degeneration in comparison to those of flies heterozygous for either of these alleles (Fig. 2E, E’). Many intact rhabdomeres were present, although some of them showed first signs of degeneration. Only few PRCs showed the dark staining typical for degenerating cells (quantification in Fig. 2F). In contrast to other fly models of retinal degeneration, for example crb (Pocha et al., 2011), Prp31 mutant flies aged for 30 days under regular light/dark conditions did not show major signs of degeneration (data not shown). To summarise, the isolated mutant Prp31 alleles reveal dominant, light-dependent retinal degeneration, similar as mutations in RP11 patients, and show intragenic complementation.
Prp31P18 mutant flies exhibit increased oxidative stress signalling
Photoreceptors have an extraordinary oxygen consumption due to their high biosynthetic activity, which is required to continuously replenish the photosensitive apical membrane (Ng et al., 2015; Yu and Cringle, 2001). In addition, although PRCs are specialised for light reception to initiate phototransduction, light at the same time is a stress factor and induces increased production of reactive oxygen species (ROS) (German et al., 2015). Increased levels of cellular ROS, in turn, induce antioxidant responses, which include the expression of proteins against oxidative stress, e.g. superoxide dismutase (SOD) or glutathione Stransferase. Their activity can prevent the cell from the detrimental consequences of oxidative stress, such as increased lipid oxidation or damage of proteins and DNA (Tomanek, 2015). In photoreceptor cells, a failure of the antioxidant machinery to neutralise increased levels of ROS can lead to light-dependent retinal degeneration, for example in fly PRCs mutant for crb (Chartier et al., 2012).
This raised the question whether flies mutant for Prp31 are subject to increased oxidative stress. To address this question, we analysed Prp3P18 mutant flies that carried the GstD-GFP reporter transgene. This reporter expresses GFP under the control of upstream regulatory sequences of glutathione S-transferase (gstD1), one of the genes involved in detoxification, whose expression is activated by oxidative stress (Sykiotis and Bohmann, 2008). The expression of this reporter has been shown to correlate with ROS levels, as revealed by the ROS-sensitive dye Hydro-Cy3 in the midgut of adult flies stressed by feeding bacteria (Jones et al., 2013). GFP expression was determined by measuring fluorescence levels in lysates of flies kept for two days on standard food and of flies kept for three days on food containing 5% hydrogen peroxide (H2O2), an established oxidative stressor. Compared to control flies, Prp31P18 heterozygous flies kept on normal food showed a 25% increase in GstD-GFP expression (Fig. 3A). This suggests that Prp31P18 mutants are under increased levels of oxidative stress signalling already under standard/basal conditions. Upon exposure to 5% H2O2, lysates from control flies displayed an 18% increase in GstD-GFP expression, while lysates from Prp31P18 mutant flies exhibited a 47% higher level of GFP compared to control flies kept on standard food (Fig. 3A). To corroborate these findings, GFP expression was examined in-situ by immunostaining of adult eye tissue. In control eyes, GstD-GFP expression was high in the pigment and cone cells. Interestingly, no GstD-GFP expression was detected in the photoreceptor cells themselves (Fig. 3B-B’). In eyes of Prp31P18/+ flies GstD-GFP levels were strongly increased in cone and pigment cells (Fig. 3C-C’). Taken together, these data show that Prp31P18 mutant eyes exhibit increased oxidative stress signalling already under normal conditions, particularly in cone and pigment cells.
Prevention of light-induced degeneration of Prpf31 mutant photoreceptor cells
It is well documented that increased ROS levels in both vertebrate and invertebrate PRCs can lead to neuronal degeneration upon exposure to additional stress, such as light stress (Punzo et al., 2012). Therefore, we assumed that the observed increase in the antioxidant response in the retina of Prp31 mutant flies is the result of increased ROS levels, and that these could be the cause for light-dependent retinal degeneration. To test this assumption, we experimentally blocked one of the major sources of cellular ROS production, mediated by the NOX (NADPH oxidase) family (Bedard and Krause, 2007). One subunit of NOX is Rac1, a member of the small GTPase protein family, which is involved, besides cytoskeletal remodelling, in the generation of ROS (Hordijk, 2006). Previous reports suggested that light can stimulate Rac1 (Balasubramanian and Slepak, 2003; Belmonte et al., 2006), which in turn results in enhanced NADPH-oxidase activity and thus increased ROS levels. In mice, constitutively active Rac1 can promote photoreceptor degeneration (Song et al., 2016), while depletion of Rac1 in photoreceptor cells was able to protect the cells from photooxidative stress (Haruta et al., 2009; Song et al., 2016). Furthermore, inhibiting Rac1 activity in a fly model for RP12, caused by loss of crb function, reduced ROS production and prevented light-dependent retinal degeneration (Chartier et al., 2012). To test whether NOX is involved in light-dependent retinal degeneration in Prp31P18 homozygous mutants, we blocked Rac1 activation by feeding flies for two days before and during exposure to light with NSC23766, a selective inhibitor of the Rac1-GEF interaction and hence of Rac1 activation (Nassar et al., 2006). After light-exposure, treated animals exhibited a strongly reduced number of apoptotic PRCs (Fig. 3D). From these results we conclude, that increased NADPH-oxidase activity, which is likely followed by increased accumulation of ROS, is a major trigger of light-dependent PRC degeneration in Prp31P18 homozygous mutant flies.
It has been suggested that retinal degeneration in Prpf31 mutant mice is caused by accumulation of truncated, and hence toxic rhodopsin proteins, which is generated as a result of improper splicing of the rhodopsin pre-mRNA (Yuan et al., 2005). To find out whether rhodopsin accumulation may contribute to light-dependent degeneration in Prp31 mutant flies, we raised larvae in a food lacking vitamin A, the precursor for retinal, the cofactor covalently bound to opsin. This treatment was shown to prevent light-dependent retinal degeneration in crb mutant PRCs (Johnson et al., 2002). As shown in Fig. 4, lack of dietary vitamin A strongly suppressed light-dependent degeneration of PRCs in Prp31 mutant flies.
Prp31 mutants reveal mis-regulation of genes involved in oxidation-reduction
In mouse and zebrafish, PRC degeneration due to loss of PRPF31 has been associated with a general reduction in biosynthetic activity as well as with impaired splicing of PRC-specific mRNAs, such as rhodopsin, RDS (Peripherin) or Fascin (FSCN2) mRNA (Linder et al., 2011; Mordes et al., 2006; Tanackovic et al., 2011; Yin et al., 2011; Yuan et al., 2005). To get a deeper insight into the mechanisms by which Drosophila Prp31 prevents retinal degeneration, we performed whole transcriptome analysis from RNA isolated from heads of Prp31P18/Prp31P18 mutant and control (w) flies (2 days old, kept under normal light/dark conditions). For each genotype three biological replicates were analysed. Overall, 115 genes were significantly (q-value cut-off 0.01) and differentially expressed (2-fold change) in Prp31P18 homozygous fly heads compared to w control flies. Of these, 53 were up-regulated and 62 were down-regulated (Suppl. Table S1, S2). Differentially expressed genes were categorized using the Gene Ontology (GO)/PANTHER classification system (Mi et al., 2013; Thomas et al., 2003), based on the predicted protein class of their gene products. Within the groups of up- and down-regulated genes, the three most prominent categories comprised genes with hydrolase, oxidoreductase, and transporter activities (Fig. 5A, B).
We then focused on the 53 genes with increased expression in the mutants (Table S1). The largest group of those that match a GO term (n=9) fall into the GO term: oxidoreductase (Fig. 5A). This group includes genes of the Cytochrome-P450 family as well as the gene cinnabar (cn), which encodes a monoxygenase with a predicted NAD(P)H oxidase activity, and plays a role in the pigment biosynthetic pathway in the eye. We also observed an up-regulation of w, encoding a transporter, and Rh6, which encodes an opsin expressed in a subset of photoreceptors. Amongst the 62 genes with decreased expression in the mutant (Table S2) the three most abundant classes matching a GO term comprised genes encoding hydrolases, oxidoreductases, and transporters (Fig. 5B). The group of hydrolases contains a chitinase (Cht3). Relevant for eye function are genes encoding transporters, including the gene scarlet (st), which is known to be involved in pigment formation in the retina.
Since we observed overlapping categories in genes with increased and decreased expression in the mutants, we evaluated all 115 differentially expressed genes by applying statistical enrichment tests. Through this we identified enrichment for oxidoreductase activity, heme binding, tetrapyrrole binding, and lyase activity (Fig. 5C), corroborating our qualitative results. It should be noted that with the exception of Desat2 in the first group (oxido-reductase activity) the first four classes comprise the same 8 members of the Cytochrome-P450 (Cyp) gene family.
Prp31 mutants exhibit increased intron retention
Since Prp31 encodes a splicing factor, we asked whether a mutation in this gene impairs splicing on a genome-wide level, as has been described in zebrafish eyes with reduced Prpf31 activity (Linder et al., 2011), in Drosophila embryos mutant for Prp19 (Sauerwald et al., 2017), and in lymphoblasts derived from patients mutant for PRPF3, PRPF8 or PRPF31(Tanackovic et al., 2011). Our data reveal that from a total of 49345 introns encoded in the genome, present in 13917 genes, 131 were significantly retained (FDR<0.1) in the mutant transcriptome, representing 127 genes (Table S3). To further evaluate the genes showing intron retention, we applied statistical enrichment tests. Strikingly, introns derived from genes falling into the categories of synaptic vesicle cycle/localization/exocytosis/transmission were particularly retained in the mutant (Fig. 6).
Taken together, we established a fly model for RP11, a retinal disease caused by mutations in the highly conserved splicing factor PRPF31. We show that mutations in Drosophila Prp31 induce light-dependent retinal degeneration, thus mimicking the symptoms of the human disease. We further uncovered major dysregulation of the transcriptome of mutant fly heads. The nature of the mis-regulated genes as well as the result of rescuing the mutant phenotype let us to conclude that light-dependent PRC degeneration of Prp31 mutant eyes is of multifactorial origin, including increased oxidative stress, defects in rhodopsin metabolism, and splicing defects in genes involved synaptic transmission.
Discussion
Here we present a fly model for RP11, an autosomal-dominant human disease leading to blindness, which is caused by mutations in the splicing regulator PRPF31. Our results reveal that, similar as in humans, mutations in the Drosophila orthologue Prp31 lead to PRC degeneration under light stress, thus mimicking major features of RP11-associated symptoms. Similar as in human, mutations in Drosophila Prp31 lead to retinal degeneration when heterozygous. This is in stark contrast to a mouse heterozygous for Prpf31, which did not show any signs of retinal degeneration (Bujakowska et al., 2009). Different studies showed late-onset defects in the retinal pigment epithelium of Prpf31 mutant mice (Farkas et al., 2014; Graziotto et al., 2011). Most of the mutations in human PRPF31 linked with RP11 are associated with reduced PRPF31 mRNA, suggesting that these are loss-of-function alleles (Rio Frio et al., 2008; Ruzickova and Stanek, 2016).
Data presented here let us to conclude that the two missense mutations mapped in Prp31P17 and Prp31P18 represent hypomorphic conditions, which reduce, but do not abolish the function of the protein. First, unlike in humans, the two Drosophila alleles characterized here are hemizygous and homozygous (in the case of Prp31P18) viable and fertile. Second, Prp31P17/Prp31P18 flies showed intragenic complementation and exhibited only a mild degenerative phenotype. Intragenic (interallelic) complementation is a rare event, and is often explained by the fact that two or more defective proteins can form functional multimers, if their mutations reside in different domains. This has been shown, for example, for mutations affecting Drosophila Dynein (Gepner et al., 1996) or Posterior sex combs (Psc), a member of the homeotic Polycomb Group (PcG) proteins involved in epigenetic silencing (Wu and Howe, 1995). In fact, the mutations in the two established Prp31 fly lines reside in different parts of the protein, namely N-terminal to the NOSIC domain in Prp31P17 (G90R) and in the Nop domain in Prp31P18 (P277L) (see Fig. 1A). However, it still remains to be analysed whether Prp31 proteins dimerize. As shown in yeast, Prp31 is a component of the spliceosomal U4/U6 di-SNP, which contains, beside the base-paired U4 and U6 snRNAs, more than 10 other proteins, including Prp3 and Prp4. In this complex, Prp31 is required to stabilize a U4/U6 snRNA junction, which in turn is required for binding of Prp3/4 (Hardin et al., 2015). In human PRPF31, the Nop domain is involved in an essential step in the formation of the U4/U6-U5 tri-snRNP by building a complex of the U4 snRNA and a 15.5K protein. Consistent with this, many mutations in human PRPF31, which are linked to RP11, have been mapped to the Nop domain. Mutations in amino acid H270 in the Nop domain of human PRPF31 results in its reduced affinity to a complex formed by a stem-loop structure of the U4 snRNA and the 15.5K protein (Liu et al., 2007; Schultz et al., 2006). Interestingly, the mutated amino acid residue in Drosophila Prp31P18 (P277L) lies next to H278, which corresponds to amino acid H270 in the human protein. Therefore, it is tempting to speculate that the Drosophila P277L mutation could similarly weaken, but not abolish the corresponding interaction of the mutant Prp31 protein in flies, thus explaining the hypomorphic nature of this allele. Finally, PRC-specific RNAi-mediated knock-down of Prp31 induced a similar phenotype as the one observed in Prp31 heterozygous animals [data not shown and (Ray et al., 2010)]. The defects on eye morphogenesis detected upon more widespread knock-down of Prp31 (Ray et al., 2010) indicates that Prp31 may be important in other tissues as well, supporting our assumption that Prp31P17 and Prp31P18 are hypomorphic alleles. Further experiments are required to determine the functional consequence of the molecular lesions. Preliminary results make it unlikely that impaired nuclear localisation of the mutant proteins is responsible for the reduced function: the mutant proteins localise to nuclear speckles when expressed in HeLa cells (data not shown), and hence behave similar as the wild-type protein (Makarova et al., 2002).
Our results show that Prp31 heterozygous flies undergo retinal degeneration, a phenotype with striking similarity to human RP11 patients. This now allows to further unveil the cause of the aetiology of the disease in a model organism, which is easily accessible to genetic manipulations and molecular studies (Ugur et al., 2016). Our results propose a multifactorial genesis of Prp31-linked retinal degeneration in flies. i) We suggest that increased accumulation of intracellular rhodopsin may contribute to the degeneration in Prp31 mutant retinas. Immunostainings of retinae of hetero-, as well as homozygous Prp31P18 flies showed an accumulation of Rh1 in the photoreceptor cell body in comparison to w1118 control flies (data not shown). Accumulation of misfolded rhodopsin in the ER due to dominant mutations in the gene which impair the maturation of the protein, has been described to cause an overproduction of ER cisternae and eventually leads to degeneration (Colley et al., 1995). Interestingly, mis-localisation of rhodopsin in human PRCs to sites other than the outer segment is a common characteristic of various forms of RP and is considered to contribute to the pathological severity (Hollingsworth and Gross, 2012). Light-dependent PRC degeneration in Prp31-mutant flies could be prevented by raising mutant animals with food that lacks vitamin A, the precursor for retinal. This treatment reduces the amount of rhodopsin to about 3% of its normal content (Nichols and Pak, 1985), and thus strongly reduces rhodopsin accumulation in the cell body.
ii) Our data further suggest that besides accumulation of intracellular rhodopsin increased oxidative stress contributes to light-dependent PRC degeneration in Prp31 mutant flies, since the degree of degeneration could be reduced by feeding flies with NSC23766, a known inhibitor of NADPH oxidase (NOX) activity. NADPH oxidases are multi-subunit enzyme complexes, comprising two membrane-bound and three cytoplasmic components, and can be found in the plasma-membrane of many cell types. One of the cytoplasmic components is the small GTPase Rac1 (Hordijk, 2006). Recruitment of the cytoplasmic components to the membrane activates the complex, resulting in transfer of electrons from NADPH to molecular oxygen, thus producing superoxide. NSC23766 selectively inhibits the interaction between Rac1 and Rac1-specific guanine nucleotide exchange factors (GEFs), thereby preventing the activation of Rac1 and hence NOX activity (Katsuyama, 2010; Rastogi et al., 2016). The suppression of retinal degeneration in Prp31 mutant flies by NSC23766 suggests that increased ROS production is causally related to degeneration. This assumption is supported by the observation that mutant flies show enhanced expression of GstD1-GFP, a reporter that serves as a proxy of ROS levels (Sykiotis and Bohmann, 2008).
Data obtained from transcriptome analysis of Prp31 mutant heads strongly support the conclusion that oxidative stress contributes to the mutant phenotype. The genes most highly upregulated in heads of Prp31 mutant flies match GO terms that are known to be upregulated upon oxidative stress (Girardot et al., 2006; Landis et al., 2004). These are, for example genes in involved in oxidation/reduction processes, e.g. desat2, Cytochrome P450-6a17 (Cyp6a17) and Cyp9c1. Some of the genes, which are downregulated in Prp31 mutant fly heads are those involved in polysaccharide/chitin metabolisms, e. g. Cht3, encoding a chitinase, and Tweedle E (TwdlE), a cuticular protein (Cornman, 2009; Guan et al., 2006; Karouzou et al., 2007). Genes from this category were shown to be highly up-regulated in flies exposed to mild stress by increased atmospheric pressure (hyperbaric normoxia) (Yu et al., 2016). This stress has been suggested to induce cytoprotective responses, aimed to protect cells or organisms against the detrimental effects of stressors, such as senescence (Oh et al., 2008). Chitin oligosaccharides comprise a major constituent of the insect cuticle and have been suggested to function as antioxidants by scavenging ROS (Ngo and Kim, 2014) and to induce the immune response (Li et al., 2013). The downregulation of these genes in Prp31 mutant heads points to a reduced cytoprotective response and hence may contribute to the detrimental effects of light stress.
Taken together, the retina of Prp31 mutant flies exhibits increased oxidative stress response, suggesting that upon additional stress, provided by constant light exposure, the antioxidant response machinery is no longer able to prevent the damage induced by high levels of ROS. This may lead, among others, to oxidation of lipids, which are major constituents of the photosensitive organelle, the outer segments in vertebrates and the rhabdomeres in flies. Oxidised lipids may contribute to PRC degeneration, as shown for Age-related Macular Degeneration (AMD) (Handa et al., 2017).
iii) Data from the transcriptome analysis further reveal that impaired Prp31 function results in increased intron retention, suggesting defects in splicing. It is worth mentioning that with the exception of Adh, impaired splicing did not overlap with the down-regulated genes. From this we conclude that transcript down-regulation largely does not reflect nonsense-mediated mRNA decay, which often occurs in order to prevent translation of intron-containing mRNAs into non-functional and potentially detrimental proteins (Yap and Makeyev, 2013). Interestingly, genes with the most significant intron retention are those involved in regulating the function, localisation and maturation of synaptic vesicles. Whether this results in defective synaptic transmission, remains to be elucidated. Previous work has shown that mutations in genes required for proper synapse organization in PRCs may eventually result in retinal degeneration, e. g mutations in Drosophila Lin-7/veli (Soukup et al., 2013). Similarly, mislocalization of pre- and postsynaptic proteins, as observed in rd1 and rd10 mutant mice [reviewed in (Soto and Kerschensteiner, 2015)], or structural abnormalities in PRC synapses as observed in tulp mutant mice (Grossman et al., 2009) precede PRC degeneration.
Taken together, using the fly as a genetic model revealed a multifactorial genesis of light-dependent retinal degeneration of Prp31 mutant flies. Impaired Prp31 function impacts on various cellular processes and results in increased oxidative stress, defective rhodopsin transport/maturation and intron retention, predominantly in transcripts of genes involved in synapse formation and function. To what extend these dysfunctions influence each other has to be elucidated.
Materials and Methods
Fly maintenance and genetics
Flies and crosses were maintained at 25°C, on standard yeast-cornmeal-agar food, under 12 hours of light/12 hours of darkness, unless specified. Flies were placed for a total of 7 days under these conditions. Genetic control for all experiments was white (w*). Deficiency lines used here were obtained from the Bloomington Stock Centre and included Df(3L)Exel6262 (Parks et al., 2004), Df(3L)ED217and Df(3L)ED218 (Ryder et al., 2007). GstD-GFP flies (Sykiotis and Bohmann, 2008) (gift from D. Bohman) were used as an indicator of oxidative stress signalling by combining into the Prp3118 genetic background or into the genetic control (w*). For the light-stress experimental paradigm, flies were kept at 25°C for 7 days in a special incubator designed to have high intensity (1200-1300 lux), continuous light exposure (Johnson et al., 2002)
Anti-oxidant feeding
Female flies were collected immediately upon eclosion and divided into groups maintained on regular food, supplemented with a Whatman® filter paper (GE Healthcare) soaked in 5% sucrose (control) or with 500 μM NSC 23766 (Cayman Chemical). Food and paper were changed every second day.
Vitamin A depletion
For vitamin A depletion experiments, flies were raised from embryonic stages until adulthood and subsequently maintained on carotenoid free food (10% dry yeast, 10% sucrose, 0.02% cholesterol, and 2% agar) as described (Pocha et al., 2011).
Hydrogen Peroxide exposure
Flies were raised on standard yeast-food and upon eclosion, were transferred in groups of 10 onto standard food or food supplemented with 5% H2O2 (Sigma-Aldrich, Germany). After two days under 12h light and 12h dark conditions, 4-6 flies/genotype were used for further analyses
Quantification of GstD-GFP following H2O2 feeding
Flies were lysed in 200μl of phosphate buffer (pH 7.4-7.6) with 0.1% Tween-20 on ice. Lysates were centrifuged at 15.000 rpm for 10 minutes. Of this, 25μl was used to estimate protein content by BCA assay and 150μl was used for fluorescence measurements using a Plate reader (Perkin Elmer Envision) and 485nm excitation & 590-10nm Emission filters. To calculate percent change, fluorescence values were normalized after arbitrarily setting values of control flies (gstDGFP/+) at 100. Data from 5 biological replicates (standard food) and 2 biological replicates (5% H2O2) were used for the analyses. ANOVA and post-hoc Bonferroni test was used to compare different samples.
Isolation of Prp31 alleles by TILLING
To isolate point mutations in the Prp31 locus (FlyBase ID: FBgn0036487) a library, of 2.400 fly lines with isogenized third chromosomes which potentially carry point mutations caused by EMS treatment, was screened. Our approach targeted exon 1-3 of the Prp31 locus containing two thirds (67%) of the coding sequence and including several predicted functional domains (the NOSIC (IPRO012976), the Nop (IPRO002687) and parts of the Prp31_C terminal (IPRO019175) domain), making use of two different PCR amplicons. A nested PCR approach was used, where the inner primers contain universal M13 tails that serve as primer binding sites of the Sanger sequencing reaction:
amplicon1 (covers exon 1 and 2), outer primer, forward: TTCAATGAACCGCATGG, reverse: GTCGATCTTTGCCTTCTCC, inner / nested primer, forward: TGTAAAACGA CGGCCAGT-AGCAACGGTCACTTCAATTC, reverse: AGGAAACAGCTATGACCAT-GAAAGGGAATGGGATTCAG);
amplicon 2 (covers exon 3), outer primer, forward: ATCGTGGGTGAAATCGAG, reverse: TGGTCTTCTCATCCACCTG, inner / nested primer, forward: TGTAAAACGA CGGCCAGT-AAGCTGCAGGCTATTCTCAC, reverse: AGGAAACAGCTATGACCAT-TAGGCATCCTCTTCGATCTG.
PCR-reactions were performed in 10 μl volume and with an annealing temperature of 57 °C, in 384 well format, making use of automated liquid handling tools. PCR fragments were sequenced by Sanger sequencing optimized for amplicon re-sequencing in a large-scale format (Winkler et al., 2011; Winkler et al., 2005). Primary hits, resembling sequence variants, which upon translation result in potential nonsense and missense mutations or affect a predicted splice site, were verified in an independent PCR amplification and Sanger sequencing reaction.
Transmission electron microscopy
Fixation of adult eyes, semi-thin sections and ultra-thin sections for transmission electron microscopy was performed as described (Mishra and Knust, 2013). 2 μm semi-thin sections were stained with a toluidine blue 1% / sodium tetraborate dehydrate 0.5% solution and imaged with AxioImager.Z1 (Zeiss, Germany) with an AxioCamMRm and the AxioVision software (Release 4.7). 70nm ultrathin sections were imaged using a Morgagni 268 TEM (100kV) electron microscope (FEI Company), and images were taken using a Side-entry Morada CCD Camera (11 Megapixels, Olympus).
Quantification of Degeneration
Quantification of degeneration was performed as described in (Bulgakova et al., 2010). Briefly, from the semi-thin sections, the number of outer photoreceptor cells (R1-R6) with clearly detectable rhabdomeres in each ommatidium were recorded as surviving rhabdomeres. In each section, 50-60 ommatidia were counted and for each genotype, 6 eyes from different individuals were analysed, unless indicated. ANOVA and post-hoc Bonferroni correction and two-tailed Mann-Whitney U test, respectively, were used to compare different distributions of photoreceptor cell survival.
Cryosections of Drosophila eyes
Adult eyes were dissected and fixed in 4% formaldehyde. Following sucrose treatment and embedding of the tissues in Richard-Allan Scientific NEG50TM (Thermo Fisher Scientific, UK) tissue embedding medium, tissues were cryosectioned at 10μm thickness at −21°C. Sections were air-dried and then subjected to immunostaining as described previously (Spannl et al., 2017). Antibodies used were rabbit anti-GFP (1:500; A11122; Thermo Fisher Scientific, UK) and mouse anti-Na+-K+-ATPase (1:100; a5; Developmental Studies Hybridoma Bank, University of Iowa, USA). Alexa-Flour conjugated secondary antibodies (Thermo Fisher Scientific, UK) were used. F-actin was visualised with Alexa-Fluor-555–phalloidin (Thermo Fisher Scientific, UK). Images were taken on Olympus FV100 and processed using ImageJ/Fiji, Adobe Photoshop CS5.1 and Adobe Illustrator CS3 for image assembly.
RNA extraction
Whole RNA was extracted from fly heads (2 days old) with the RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. Each biological replicate consisted of 50 heads. A total of 3 biological replicates were used for further analyses.
RNA-seq and analyses
RNA-Seq was carried out in triplicate by the Deep Sequencing Group SFB 655 at Biotechnology Center, TU Dresden, with 76-bp single read sequencing on the Illumina HiSeq 2500. An average of 28.85 million reads per sample was obtained. Data analysis was carried out by the Scientific Computing Facility (MPI-CBG) by mapping to the *Drosophila* genome (BDGP6, Ensembl v81) using STAR (v2.5.1b). The differential expression analysis was performed using DESeq2 using a lfcThreshold parameter of 1.0 and Independent Hypothesis Weighting (Love et al., 2014). 115 differentially expressed genes were obtained with a q-value cutoff of 0.01. Detection of differentially spliced genes was carried out using the DEXSeq Bioconductor package (Anders et al., 2012) with a discovery rate (FDR) threshold of 0.1. To test for intron retention as well as for differential exon usage, the DEXSeq model used to estimate exon abundance was complemented with intron regions (similar to the method presented by (Sauerwald et al., 2017)). clusterProfiler (Yu et al., 2012) was used to identify enrichment on a 0.05 q-value level for biological processes and pathways associated with the up- and down-regulated genes and for genes whose introns are differentially retained. The transcriptome data have been submitted to GEO. The accession number is GSE99665.
Figure panel preparation
All figure panels were assembled using Adobe Illustrator CS3 and Inkscape. Statistical analyses and graphs were generated using GraphPad Prism (GraphPad Software, Inc, USA) and Microsoft Excel. For protein sequence visualization, Illustrator of Biological Sequences (IBS; (Liu et al., 2015)) software package was used.