Splicing factor PRP-19 regulates mitochondrial stress response

Animals respond to mitochondrial perturbation by activating the mitochondrial unfolded protein response (UPR mt ) to induce the transcription of mitochondrial stress response genes. In C. elegans , activation of UPR mt allows the animals to maintain organismal homeostasis, activate the innate immune response and promote lifespan extension. Here we show that splicing factors such as PRP19 are required for the induction of UPR mt in C. elegans . PRP-19 also modulates mitochondrial perturbation-induced innate immune response and lifespan extension. Knockdown of PRP-19 in mammalian cells suppresses UPR mt activation and disrupts the mitochondrial network. These findings reveal an evolutionarily conserved mechanism that maintains mitochondrial homeostasis and controls innate immunity and lifespan through splicing factors. a affecting and chromatin remodeling RNA In this study, we explore the effect of splicing factors on UPR mt in a physiologically relevant setting. We show that upon mitochondrial perturbation, splicing factors such as PRP-19 are required for the induction of UPR mt in C. elegans . PRP-19 also plays an essential role in promoting innate immunity and lifespan extension under mitochondrial stress conditions. Moreover, PRPF19, the mammalian ortholog of PRP-19, regulates mitochondrial homeostasis in mammals. In summary, our results reveal an essential function of splicing factors in UPR mt signaling. salt SDS, 2 EDTA mM Tris-HCl pH 8.0, 150 NaCl), once high salt (1% SDS, Triton-X100, 2 mM EDTA 8.0, 20 mM Tris-HCl 8.0, NaCl), once with LiCl (0.25 1% deoxycholate, EDTA, Tris 8.1) and with TE centrifugation at 1000× g for the beads were mixed with 230 μL freshly prepared elution (1% SDS and 0.1 M sodium bicarbonate) and incubated at room for 20 min. The mixture was centrifuged at 20,000 × g for 15 min and the supernatant was saved. 8 μL of 5 M NaCl was added to the supernatant and de-crosslinked at 60 °C for 8 h. For input samples, 1 μl of 5 M NaCl solution was added for every 20 μL sample. 4 μL of 0.5 M EDTA, 8 μL of 1 M Tris-HCl, pH 6.5, and 1 μL of proteinase K were added to the immunoprecipitation samples after de-crosslinking. 2 μL of 0.5 M EDTA, 4 μL of 1 M Tris-HCl, pH 6.5, and 1 μL of proteinase K were added to the input samples, and digested at 45 °C for 2 h. The DNA fragments were recovered using a ChIP DNA Clean & Concentrator kit.

A c c e p t e d M a n u s c r i p t 5 localization and the expression level of PRP-19 were not affected by mitochondrial perturbation ( Fig.   1f and g). Notably, prp-19 RNAi also affected the induction of the endoplasmic reticulum (ER) stress reporter hsp-4p::gfp when the worms were challenged with hsp-4 RNAi or tunicamycin ( Fig. 1h and i). Conversely, the heat shock response and the induction of hsp- 16.2p::gfp reporter were not affected by knockdown of prp-19 (Fig. 1j).
PRP-19 is a core component of the large Prp19C/NTC complex mentioned above. This protein complex consists of eight core proteins and more than 30 other proteins in mammals [23]. The best characterized function of Prp19C/NTC is its role in the catalytic activation of the spliceosome [17,24]. Prp19C/NTC has also been shown to play a role in transcription elongation and the maintenance of genome stability [25][26][27]. To see if other splicing factors similarly modulate UPR mt , we individually knocked down a broad spectrum of splicing factors in different spliceosome subunits, including many splicing factors functioning outside of the Prp19C/NTC complex (Fig. 2a). RNAi of most of the splicing factors strongly suppressed the induction of UPR mt and UPR ER , but not the heat shock response ( Fig. 2b-d). These results suggest that PRP-19 regulates UPR mt activation through its function in splicing.
Alternative splicing is a post-transcriptional process in eukaryotes that produces different isoforms of mRNAs and increases the diversity of gene expression. Alternative splicing events are divided into five main patterns: use of alternative 3′ (acceptor) splice sites (A3SS), use of alternative 5′ (donor) splice sites (A5SS), mutually exclusive exon usage (MXE), retention of introns (RI), and skipping of exons (SE) [28]. To test if any alternative splicing events are required for the production of factors that play a role in UPR mt regulation, we performed RNA-seq experiments to detect global transcriptome alteration. Mitochondrial perturbation induced by atp-2 RNAi only caused 8 alternatively spliced transcripts in total, whereas knockdown of prp-19 alone or knockdown of prp- 19 together with atp-2 resulted in 117 or 102 alternative splicing events, respectively (Supplementary Fig. S1a and b, and Table S1). We validated some of the alternative splicing events A c c e p t e d M a n u s c r i p t 6 and confirmed the RNA-seq results ( Supplementary Fig. S1c-f). However, UPR mt genes were not alternatively spliced (Supplementary Table S1). We also found no evidence that alternatively spliced genes play a role in UPR mt activation.

PRP-19 functions downstream of ATFS-1 to modulate mitochondrial stress response
To explore the molecular mechanism by which PRP-19 modulates UPR mt , we tested if knockdown of prp-19 affected the functions of known components in the UPR mt pathway. In C. elegans, the transcriptional induction of UPR mt is mainly governed by two transcription factors, ATFS-1 and DVE-1 [10,12,13]. In addition, a histone deacetylase HDA-1 functions in concert with DVE-1 to activate the transcription of UPR mt genes [11]. Unlike hda-1 RNAi, which reduces the DVE-1 protein level [11], knockdown of prp-19 actually promoted the nuclear accumulation of DVE-1 and elevated the DVE-1 protein level ( Supplementary Fig. S2a). Consistent with this, prp-19 RNAi also promoted the nuclear accumulation of HDA-1 and elevated the HDA-1 protein level (Supplementary Fig. S2b and c).
We then tested if PRP-19 modulates UPR mt gene expression via ATFS-1. The transcription factor ATFS-1 harbors an amino-terminal mitochondrial targeting sequence and a carboxy-terminal nuclear localization sequence. Upon mitochondrial perturbation, the import efficiency of mitochondria is decreased, leading to the nuclear accumulation of ATFS-1 [12]. Deletion of the N-terminal amino acids 1-32 of ATFS-1 disrupts its mitochondrial targeting signal, resulting in constitutive nuclear accumulation of ATFS-1 and activation of UPR mt gene expression [12]. We employed a transgenic strain that allows the expression of ATFS-1 Δ1-32.myc upon heat shock and noticed that expression of the UPR mt gene hsp-60 also requires prp-19 (Fig. 3a). Knockdown of prp-19 did not affect the A c c e p t e d M a n u s c r i p t 7 expression and nuclear accumulation of ATFS-1 Δ1-32 after heat shock (Fig. 3a), suggesting that PRP-19 acts downstream of ATFS-1 once it enters the nucleus.
After entering the nucleus, ATFS-1 binds to the promoter of UPR mt genes and activates their transcription. A group of ATFS-1-dependent UPR mt genes, including mitochondrial chaperone genes hsp-6 and hsp-60, mitochondrial import complex genes tomm-7 and timm-23, and innate immune response genes cyp-14A4 and ugt-61, also require the presence of PRP-19 for their induction during mitochondrial stress (Fig. 3b). We therefore sought to test if lack of PRP-19 affects the binding of ATFS-1 to the promoters of UPR mt genes. Again, we employed a transgenic strain which expresses ATFS-1 Δ1-32.myc ::GFP upon heat shock induction. Chromatin immunoprecipitation (ChIP) coupled with quantitative PCR (qPCR) was performed using anti-GFP antibody to pull-down ATFS-1 Δ1-32.myc ::GFP ( Fig. 3c). A significant amount of ATFS-1 Δ1-32.myc was associated with the promoters of hsp-6 and hsp-60 after heat shock induction. However, knockdown of prp-19 did not impair, but rather promoted the binding of ATFS-1 to the promoters of hsp-6 ( Fig. 3d). Therefore, PRP-19 and spliceosome seem to function in a step after ATFS-1 associates with the promoter of mitochondrial stress response genes to modulate UPR mt activation.
Interestingly, a recent study has shown that spliceosomal repression directly affects gene transcription in mouse embryonic stem cells (ESCs), resulting in the decreased expression of pluripotent genes, but not affecting the expression of totipotent genes [29]. This study proposed that the greater length and increased number of introns in pluripotent genes relative to totipotent genes may explain their transcriptional regulation by spliceosomal repression. However, we analyzed the number and length of introns in the UPR mt genes and noticed no difference compared with other genes not regulated by the mitochondrial stress response (data not shown). Therefore, at this stage, the detailed mechanism of how the transcription of UPR mt genes is specifically regulated by the splicing factors remains elusive.

PRP-19 is essential for UPR mt -mediated innate immunity and lifespan extension
Mitochondrial function is actively challenged by wild microbes in the natural habitats of C. elegans [1]. As a defense mechanism, UPR mt also initiates the innate immune response [1,7] and promotes lifespan extension [8,10,11]. Therefore, we further tested the function of PRP-19 in innate immunity and lifespan regulation. irg-1p::gfp has been used as a reporter strain for the induction of innate immune response [1,30]. We challenged the irg-1p::gfp transgenic worms with a Pseudomonas aeruginosa strain isolated from natural habitats of C. elegans. This pathogen has been shown to disrupt mitochondrial function and activate UPR mt [1]. Knockdown of prp-19 suppressed the induction of irg-1p::gfp upon pathogen infection (Fig. 4a). Deficiency of prp-19 also suppressed the endogenous induction of several other immune response genes and reduced the survival rate of C. elegans when they were exposed to Pseudomonas aeruginosa ( Fig. 4b and c). Taken together, these results indicate that PRP-19 plays a critical role in mediating the innate immune response.
The physiological function of PRP-19 was assessed by examining lifespan regulation in the presence or absence of mitochondrial stress. Under normal conditions, prp-19 RNAi had a minor shortening effect on worm lifespan (Fig. 5a). However, knockdown of prp-19 greatly suppressed the lifespan extension in atp-2 RNAi-treated worms (Fig. 5a). To test the knockdown efficiency of double RNAi, we performed qPCR experiments and showed that the reduction in atp-2 expression achieved with a mix of atp-2 RNAi plus prp- 19 RNAi is similar to that with atp-2 RNAi plus the control RNAi ( Fig. 5b). In addition, knockdown of prp-19 suppressed the lifespan extension of worms carrying a mutation in the mitochondrial gene isp-1 (Fig. 5c). Age-related decline of protein homeostasis results in the toxic accumulation of protein aggregates, including aggregates of proteins containing polyglutamine (polyQ) repeats [2,11]. Consistent with the role of PRP-19 in lifespan regulation, knockdown of prp-19 impaired the mobility of worms expressing polyQ repeats (Fig. 5d) and significantly increased the number of polyQ aggregates in C. elegans ( Fig. 5e and f). Collectively, A c c e p t e d M a n u s c r i p t 9 these results suggest that PRP-19 is essential for the UPR mt -mediated beneficial impact to alleviate age-related pathology.

PRP-19 modulates mitochondrial stress response in mammals
We further examined if PRP-19 plays a conserved role in modulating mitochondrial homeostasis in higher eukaryotes. Interestingly, the expression levels of PRPF19 (mammalian ortholog of PRP- 19) strongly correlate with the mitochondrial proteases YME1L1 and LONP1, the mitochondrial import inner membrane translocase TIMM17A, the asparagine synthetase ASNS, and the mitochondrial chaperones HSPE1, HSPD1 and HSPA9 in various human tissues (Fig. 6a). To validate the function of PRPF19 in mediating mitochondrial homeostasis in mammals, we used shRNA to knock down PRPF19 in HEK293T cells and examined the expression levels of mitochondrial stress response genes.
Deficiency of PRPF19 suppressed the induction of mitochondrial stress response genes in cells treated with FCCP, a potent uncoupler of mitochondrial oxidative phosphorylation (Fig. 6b).
Knockdown of PRPF19 impaired both basal and ATP-linked respiration in the presence or absence of mitochondrial perturbation ( Supplementary Fig. S3). In the presence of Antimycin A, mitochondrial morphology was disrupted more extensively in PRPF19 knockdown cells than in wild-type cells ( Fig.   6c and d). Taken together, these results revealed an evolutionarily conserved role of PRPF19 in mediating mitochondrial stress response.

Discussion
Upon mitochondrial perturbation, UPR mt is activated to initiate a mitochondrion-to-nucleus crosstalk to activate the transcription of mitochondrial stress response genes, as well as innate immune response genes. Here we report that splicing factors such as PRP-19 are required for the activation A c c e p t e d M a n u s c r i p t 10 of UPR mt in C. elegans. In addition, we show that PRPF19 (mammalian ortholog of PRP-19) also mediates mitochondrial homeostasis in mammals.
prp-19 is an essential gene. prp-19 RNAi causes sterility and embryonic lethality. Therefore, in several experiments (Figs. 1c, 4c and Supplementary Fig. S1), we used the temperature-sensitive sterile glp-4 mutant strain to avoid collecting the embryos generated in the control RNAi group, which may affect the results. Knockdown of prp-19 suppresses the induction of both UPR mt and UPR ER , but not the induction of the heat shock response ( Fig. 1a-c, h-j), which reveals the specificity of PRP-19 in regulating stress responses. Knockdown of prp-19 promotes the accumulation of polyQ35::YFP aggregates in muscles. We speculate that the effect of PRP-19 on age-associated polyQ accumulation is due to UPR mt but not UPR ER based on the following two reasons: 1) It has been shown that the accumulation of polyQ35::YFP aggregates in muscles is unaffected by the expression of xbp-1s, which activates UPR ER [31]. 2) Our previous studies showed that knockdown of the histone deacetylase HDA-1, which specifically regulates UPR mt but not UPR ER , promotes polyQ35::YFP aggregate accumulation in muscles [11].
It has been reported that PRP-19 has both pre-mRNA processing factor and E3 ubiquitin ligase activities [32]. The E3 ligase activity of PRP-19 is critical for the catalytic activation of the spliceosome [23]. On the other side, Prp19 has been reported to require components of the Prp19C/NTC complex for the E3 activity [33]. Taking these findings together, it is very difficult to separate the putative E3 activity of PRP-19 from the splicing activity. We knocked down several splicing factors that function outside the Prp19C/NTC complex ( Fig. 2b-d), and found that many of those factors also affect UPR mt . These results suggest that PRP-19 may modulate UPR mt through its function in splicing.
Splicing has been shown to occur and function co-transcriptionally. Although PRP-19 and other splicing factors modulate the activation of UPR mt , we did not observe extensive alternative splicing events when C. elegans was challenged with mitochondrial perturbation (Supplementary Fig. S1a).

Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the Lead Contact, Ying Liu (ying.liu@pku.edu.cn).

EXPERIMENTAL MODEL AND SUBJECT DETAILS
Worms were maintained on NGM plates seeded with OP50 bacteria under normal conditions and grown on RNAi plates supplemented with 1.2 mg/mL IPTG in NGM plates seeded with RNAi bacteria.
HEK293T cells and HeLa cells were obtained from ATCC. Cells were cultured in DMEM high glucose medium supplemented with 10% (v/v) fetal bovine serum at 37 °C.

RNA Interference
All the RNAi bacteria were obtained from the Ahringer library. RNAi bacteria were grown in liquid LB

Chromatin immunoprecipitation (ChIP)
ChIP assay was performed using the hsp-16.

Mitochondrial respiration analysis
Mitochondrial respiration was measured using an Agilent Seahorse XFe24 analyzer and Seahorse XF Worm ATFS-1 ChIP-seq processed data (Nargund et al., 2015) were obtained under accession GEO: GSE63803. The ATFS-1 binding genes were defined as the nearest genes to ATFS-1 ChIP-seq peaks.

Expression correlation analyses
Expression correlation analysis was performed using the expression data of PRPF19 and UPR mtrelated genes in various human tissue samples from the GTEX database (https://www.gtexportal.org/home/datasets). Gene expression levels in the above samples were compared by calculating Pearson's correlation. Heat map presentation was performed by the ggplot2 function package in R language.

Quantification and statistical analysis
The experiments in this paper were all repeated at least three times unless otherwise indicated.
Statistical analysis was performed using GraphPad software and Student's t-test (two-tailed, unpaired) to calculate the P values. IGV software was used to visualize the RNA-seq sequencing results.
A c c e p t e d M a n u s c r i p t 26 Figure.1 prp-19     RNAi. n = 3 biological replicates. n >100 worms for each condition.