Loss of muscleblind splicing factor shortens Caenorhabditis elegans lifespan by reducing the activity of p38 MAPK/PMK-1 and transcription factors ATF-7 and Nrf/SKN-1

Abstract Muscleblind-like splicing regulators (MBNLs) are RNA-binding factors that have an important role in developmental processes. Dysfunction of these factors is a key contributor of different neuromuscular degenerative disorders, including Myotonic Dystrophy type 1 (DM1). Since DM1 is a multisystemic disease characterized by symptoms resembling accelerated aging, we asked which cellular processes do MBNLs regulate that make them necessary for normal lifespan. By utilizing the model organism Caenorhabditis elegans, we found that loss of MBL-1 (the sole ortholog of mammalian MBNLs), which is known to be required for normal lifespan, shortens lifespan by decreasing the activity of p38 MAPK/PMK-1 as well as the function of transcription factors ATF-7 and SKN-1. Furthermore, we show that mitochondrial stress caused by the knockdown of mitochondrial electron transport chain components promotes the longevity of mbl-1 mutants in a partially PMK-1-dependent manner. Together, the data establish a mechanism of how DM1-associated loss of muscleblind affects lifespan. Furthermore, this study suggests that mitochondrial stress could alleviate symptoms caused by the dysfunction of muscleblind splicing factor, creating a potential approach to investigate for therapy.


Introduction
The human Muscleblind-like (MBNL) protein family consists of MBNL1, MBNL2, and MBNL3. As they regulate alternative splicing programs, polyadenylation, mRNA turnover, and localization (Pascual et al. 2006;Masuda et al. 2012;Wang et al. 2012;Batra et al. 2014;Xu et al. 2021), dysfunction of MBNLs has been found to be implicated in many degenerative diseases such as Huntington's disease-like 2, spinocerebellar ataxia type 8, and polyglutamine disorders (Fernandez-Costa et al. 2011). However, MBNL dysfunction is best known for its role in myotonic dystrophies.
Myotonic dystrophy type 1 (DM1) and type 2 (DM2) are multisystemic diseases caused by 50-4000 CUG trinucleotide repeat expansions in the 3 0 untranslated region (UTR) of the myotonic dystrophy protein kinase (DMPK) gene and 55-11,000 CCUG repeats in the first intron of the cellular nucleic acid-binding protein (CNBP) gene, respectively. DM1 is the most prevalent muscular dystrophy in adults, and its clinical features are more severe compared to DM2 (Meola and Cardani 2015). At the molecular level, repeat-containing RNAs form imperfect double-stranded structures that sequester, and thereby disrupt the function of MBNLs and several other RNA-binding factors (Miller et al. 2000;Fardaei et al. 2001), which leads to splicing defects (Ranum and Cooper 2006). Importantly, Mbnl1 and Mbnl2 knockout mice develop RNA splicing abnormalities that are characteristic of DM1 (Kanadia et al. 2003;Charizanis et al. 2012;Suenaga et al. 2012), demonstrating that the loss of MBNL splicing factors is a central contributor of DM1 pathogenesis. Since DM1 manifests with phenotypes commonly associated with aging (e.g. cataracts, muscular weakness, atrophy, insulin resistance, and metabolic dysfunction), it has also been considered as a disorder of accelerated aging (Mateos-Aierdi et al. 2015). Thus, the function of MBNL splicing factors may play an important regulatory role in aging.
MBNLs are well-conserved factors across animals. In Caenorhabditis elegans, MBL-1, the sole ortholog of mammalian MBNL proteins binds CUG and CCUG repeats (Sasagawa et al. 2009;Garcia et al. 2014) and is required for normal mRNA splicing (Norris et al. 2017; Thompson et al. 2019). MBL-1 appears to be required for normal muscle structure and function as well as for the proper formation of neuromuscular junction synapses and dendrite morphogenesis (Wang et al. 2008;Spilker et al. 2012; Pseudomonas aeruginosa assay 3.5-cm NG plates were seeded with 40 mL of overnight grown Pseudomonas aeruginosa (PA14) suspension and incubated at 37 C for 24 h (Tan et al. 1999). Twenty microliters of SDS 2% was added to the edges of the plate to prevent the escape of the animals. Caenorhabditis elegans were transferred to PA14 plates at L4 stage and incubated at 25 C. Animals were scored daily for survival based on their ability to respond to touch. Animals dehydrated on the wall of the plate were censored from the analysis. For Western blot upon PA14 treatment, animals were transferred to PA14 plates at L4 stage and collected 24 h later as day 1 adults.

RNA interference
EV refers to "Empty Vector," which was used as a control in RNA interference (RNAi) experiments. tir-1, pmk-1, atf-7, skn-1, and cco-1 RNAi clones were taken from either Ahringer or Vidal RNAi library. daf-16 (#34833) and daf-2 RNAi (#34834) clones were obtained from Addgene. mbl-1 RNAi was cloned from C. elegans cDNA by using 5 0 -CTTGAGCTCTTCGACGAAAACAGTAATGCCGC-3 0 and 5 0 -CTTCTCGAGCTAGAATGGTGGTGGCTGCATG-3 0 primers. RNAi was performed using the feeding protocol as described earlier (Timmons et al. 2001). Lifespan experiments were initiated by letting gravid hermaphrodites (P0 generation) to lay eggs on RNAi/NGM plates, and P1 generation was scored for lifespan. In some lifespan experiments, animals were bleached and let to hatch overnight in M9 before plating P1 generation as L1 larvae to experimental plates. These two alternative ways to initiate lifespan did not affect the conclusions made from the experiments. For qRT-PCR and Western blot animals were bleached and let to hatch overnight in M9 before plating L1 larvae for experimental plates.

RNA sequencing
N2 and mbl-1(tm1563) strains were synchronized by bleaching and plated as L1 larvae on RNAi plates seeded with HT115 bacteria carrying EV ( control vector for RNAi). To prevent the hatching of the progeny, animals were transferred to plates containing 10 lM of 5-fluorouracil (Sigma) at the L4 stage. Animals were collected on day 2 of adulthood (three biological replicates for both strains) and frozen in liquid nitrogen. Total RNA was extracted with TRIzol Reagent (Ambion) and assessed for degradation using Agilent 2100 Bioanalyzer. Illumina Truseq stranded polyA-mRNA library was prepared and sequenced for 86 cycles at the DNA Sequencing and Genomics Laboratory (Institute of Biotechnology, University of Helsinki). The six samples were multiplexed and sequenced on one lane of Illumina NextSeq 500, yielding circa 18-20 million reads per sample. Sequencing reads were mapped to the C. elegans reference transcriptome (WBcel235 assembly) using STAR (Dobin et al. 2013). Read counts over transcripts were calculated using HTSeq (Anders et al. 2015). For differential expression analysis, we used the edgeR method (Robinson et al. 2010) and classified genes as differentially expressed based on the cutoffs of twofold change in expression value and false discovery rates (FDR) below 0.05. For differential splicing analysis, we used the diffSpliceDGE function in the edgeR package. Splicing isoforms were detected by applying Simes-statistical method within diffSpliceDGE function with an FDR cutoff of 0.05.

Lifespan analysis
Caenorhabditis elegans lifespan experiments were done at 20 C. At the L4 larval stage, animals were transferred to plates containing 10 mM of 5-Fluorouracil (Sigma) to prevent progeny production. Animals that had exploded vulva or that crawled off the plate were censored. Animals were counted as dead if there was no movement after poking with a platinum wire. Lifespans were checked every 1-3 days. For lifespan data, mean lifespan 6 standard error (SE) is reported (see Supplementary Tables S1  and S2).

Quantitative RT-PCR
Animals were collected at indicated age and frozen in liquid nitrogen. TRIzol Reagent (Ambion) was used to extract RNA. cDNA synthesis was done with QuantiTect Reverse Transcription Kit (Qiagen) and quantitative RT-PCR (qRT-PCR) reactions were run with the SYBR Green reagent (Roche) using Lightcycler 480 (Roche). qRT-PCR data were normalized to the expression of cdc-42 and pmp-3. qRT-PCR oligos used in this study are provided in Supplementary Table S9. qRT-PCR experiments were performed with three biological replicates (see Supplementary Table S8 for raw qRT-PCR data) and repeated at least twice with similar results. Statistical significances were analyzed by using Student's t-test or two-way ANOVA.
Agarose gel-based splicing assay Animals were collected at indicated age and frozen in liquid nitrogen. TRIzol Reagent (Ambion) was used to extract RNA. cDNA synthesis was done with QuantiTect Reverse Transcription Kit (Qiagen). Splicing of unc-43 and unc-104 mRNAs was analyzed by running PCR products in agarose gels. Oligos used for PCR are provided in Supplementary Table S10. Splicing assays were performed with three biological replicates. Agarose gels were quantified by using Fiji. Statistical significances were analyzed by using two-way ANOVA.
Fluorescent imaging and oxidative stress assay gst-4p::GFP reporter strains (wild-type and mbl-1(tm1563) background) were grown on EV RNAi plates and imaged as day 1 adults with a standard stereomicroscope with fluorescent light source (Zeiss). For induction of oxidative stress, 5% H 2 O 2 (in H 2 O, H 2 O used as a control) was pipetted to EV plates containing 1-day adult gst-4p::gfp reporter animals. Worms were washed after 20 min, and let to recover on EV plates for 5 h before imaging. GFP fluorescence was quantified by using Fiji (see Supplementary  Table S8 for GFP quantifications).

Statistical analysis
Statistical analyses for qRT-PCR data were carried out in GraphPad Prism or Excel, and the data represent the mean of three biological replicates 6 SD. Statistical analyses for Western blot data, splicing assays (agarose gels), and GFP fluorescence were carried out in GraphPad Prism. Statistical details can be found in the figures and figure legends. Statistical analyses for lifespan experiments were carried out in R by using the Cox proportional hazard regression. Statistical details for the lifespan data can be found in Supplementary Tables S1  and S2. Supplementary Table S1 shows calculations from combined independent lifespan experiments (from Supplementary  Table S2).

Data availability
Strains and plasmids are available upon request. The Gene Expression Omnibus (GEO) accession number for the RNA-seq data originating from this study is GSE146801.
Supplementary material is available at GENETICS online.

Results
Loss of MBL-1 leads to shortened lifespan and reduced expression of p38 MAPK/PMK-1-regulated genes To examine how the loss of MBL-1 affects C. elegans lifespan, we utilized a strain carrying the mbl-1(tm1563) allele (513-bp deletion) that eliminates an exon shared among all mbl-1 isoforms and creates a putative null allele. This deletion has been shown to shorten lifespan (Sasagawa et al. 2009). We repeated this experiment, performing lifespan assays on two Escherichia coli strains, HT115 and OP50, which are both widely used in C. elegans experiments. We found that mbl-1(tm1563) mutant animals have significantly shorter lifespan on both bacterial strains compared to wild-type C. elegans (N2) ( Figure 1A and Supplementary Tables S1 and S2). Interestingly, loss of MBL-1 has a more drastic decrease in lifespan on HT115 compared to OP50. One potential explanation is that the microbiome could modulate lifespan partly through MBL-1. As many experiments in this study take advantage of RNAi, all following experiments (with the exception of the P. aeruginosa resistance assay) were performed using HT115 bacteria. After confirming the earlier finding that mbl-1(tm1563) mutants have shortened lifespan (Sasagawa et al. 2009) ( Figure  1A and Supplementary Tables S1 and S2), we asked whether knockdown of mbl-1 produces a similar phenotype. It has been shown previously that mbl-1 RNAi shortens lifespan (Wang et al. 2008). Likewise, we also found that mbl-1 RNAi shortens N2 lifespan compared to treatment with EV (control RNAi) ( Figure 1B and Supplementary Tables S1 and S2), thus further demonstrating that disruptions to MBL-1 function affect lifespan. Interestingly, simultaneous overexpression of three mCherrytagged mbl-1 isoforms (Supplementary Figure S1, A and B) under endogenous mbl-1 promoters also shortens lifespan compared to N2 ( Figure 1C and Supplementary Tables S1 and S2), indicating that an optimal level of this developmentally regulated splicing factor is required for normal lifespan. Next, we asked whether MBL-1 activity decreases with aging. For this purpose, we examined the unc-43 and unc-104 genes, known targets of MBL-1mediated splicing (Norris et al. 2017), for age-related changes in their alternative splicing patterns (Supplementary Figure S2A). PCR analysis revealed that MBL-1 is required for exon exclusion events in both unc-43 and unc-104 ( Supplementary Fig S2, B-E and Table S3). Day 4 adult N2 animals show enrichment of longer unc-43 isoform, which is observed as early as L4 larval stage with mbl-1(tm1563) mutants (Supplementary Figure S2, B and C and Table S3), whereas unc-104 splicing shows a similar enrichment already in day 2 adult N2 (Supplementary Figure S2, D and E and Table S3). These data suggest that MBL-1 activity decreases upon aging.
Since the reduced activity of MBNLs is a hallmark of DM1, we focused on the mechanism by which the loss of MBL-1 shortens lifespan. As an initial step, we performed RNA-seq of day 2 adult (day 5 from hatch) N2 and mbl-1(tm1563) mutants. Our analysis revealed that 135 genes are up-and 235 downregulated (at least 2-fold change in expression, FDR < 0.05) in mbl-1(tm1563) mutants compared to N2 (Supplementary Table S4). In order to identify aging-associated signaling pathways linking the loss of MBL-1 with shortened lifespan, we utilized the WormExp database, which collates nearly all published C. elegans expression data sets from public databases (Yang et al. 2016). We found that genes differentially regulated in mbl-1(tm1563) mutants are modulated by factors such as PQM-1, ELT-2, DAF-16, and PMK-1 ( Figure 1, D and E and Supplementary Table S5). Interestingly, PMK-1, which encodes the conserved p38 MAPK (p38 MAPK), was the only factor associated exclusively with downregulated genes ( Figure 1E and Supplementary Table S5). More specifically, PMK-1-regulated genes, whose expression is reduced in day 15 vs. day 6 old C. elegans (Youngman et al. 2011), are enriched among the downregulated genes in mbl-1(tm1563) mutants ( Figure 1E and Supplementary Table S5). Due to this enrichment, we focused on the link between p38 MAPK/PMK-1 and MBL-1 to regulate lifespan.

Loss of MBL-1 does not impair innate immune response
When subjecting up-and downregulated genes to gene ontology (GO) analysis by using the DAVID database (Huang da et al. 2009), we found that innate immune response-related GO terms are enriched among genes upregulated in mbl-1(tm1563) mutants ( Figure 2A). This is interesting since the WormExp analysis indicated that PMK-1, a central factor of innate immunity-regulating p38 MAPK signaling (Kim et al. 2002), has reduced activity in mbl-1(tm1563) mutants ( Figure 1E and Supplementary Table S5). On the other hand, WormBase GO enrichment analysis of 34 PMK-1regulated genes whose expression is decreased in mbl-1(tm1563) mutants ( Figure 1E and Supplementary Table S5) reveals that only six genes are associated with innate immunity (Supplementary Table S5). These data indicate that the loss of MBL-1 does not impair the expression of PMK-1-regulated innate immunity genes. Nevertheless, we asked whether the loss of MBL-1 affects pathogen resistance, and for this experiment, we investigated the survival of mbl-1(tm1563) mutants on pathogenic P. aeruginosa (PA14) in a slow-killing assay. In this assay, we cultured PA14 on normal NGM plates where the killing occurs over the course of several days (Tan et al. 1999). mbl-1(tm1563) mutants were found to have similar survival on PA14 when compared to  Supplementary Tables S4 and S5 for differentially expressed genes between N2 and mbl-1(tm1563) mutants and WormExp statistics, respectively.
N2. This is in contrast to pmk-1(km25) mutants, which are extremely susceptible to PA14 ( Figure 2B and Supplementary Tables S1 and S2). Furthermore, mbl-1(tm1563) mutants show significantly elevated levels of phosphorylated p38 MAPK/PMK-1 upon PA14 infection (Figure 2, C and D and Supplementary Figure  S3 and Table S6). The data demonstrate that MBL-1 dysfunction does not impair the ability to mount a p38 MAPK/PMK-1mediated innate immune response upon infection.
mbl-1(tm1563) mutants have reduced level of activated p38 MAPK/PMK-1, which contributes to their short lifespan Since our RNA-seq data suggested that mbl-1(tm1563) mutants exhibit decreased PMK-1 activity ( Figure 1E and Supplementary  Table S5), we analyzed whether the level of phosphorylated p38 MAPK/PMK-1, which is traditionally used to assess the activity of this kinase, is altered in these mutants. Western blot analysis showed that day 2 old mbl-1(tm1563) mutants have a reduced level of phosphorylated PMK-1 (Figure 3, A and B, Supplementary  Figures S3 and S4 and Table S6), thus demonstrating that basal p38 MAPK/PMK-1 activity is decreased upon disrupted MBL-1 function. mbl-1 RNAi does not significantly affect PMK-1 phosphorylation ( Figure 3, C and D, Supplementary Figure S3 and Table S6), which could be due to insufficient knockdown efficiency. However, the quantified mean of phosphorylated PMK-1 levels from three independent experiments ( Figure 3C) indicates that, in addition to mbl-1 mutation, mbl-1 knockdown also modulates PMK-1 phosphorylation.
To examine whether MBL-1-regulated splicing is implicated in modulating the activity of p38 MAPK/PMK-1, we analyzed our RNA-seq data for alternative splicing (Supplementary Table S7).
However, this analysis, or previously published investigation of MBL-1-dependent splicing in L4 larvae (Norris et al. 2017), did not reveal any obvious splicing event that could cause the decline in the activity of p38 MAPK/PMK-1 in mbl-1 mutant animals. In addition to alternative splicing, MBNL1 regulates also mRNA stability (Masuda et al. 2012). Therefore, we asked whether mRNA levels of C. elegans p38 MAPK pathway ( Figure 3E) genes are altered in mbl-1(tm1563) mutants. We found that in L4 larvae the expression of tir-1 is upregulated in mbl-1(tm1563) mutants, whereas the expression of nsy-1, sek-1, and pmk-1 is unchanged ( Figure 3F and Supplementary Table S8). In addition to p38 MAPK pathway genes, we examined the expression of mek-1 MAPK kinase and vhp-1 MAPK phosphatase, which both regulate PMK-1 activity (Kim et al. 2004). Interestingly, whereas mek-1 expression is unchanged in mbl-1(tm1563) mutants, we found that the expression of vhp-1 is increased in these animals ( Figure 3F and Supplementary Table S8). Additionally, we found that the expression of nsy-1 and vhp-1 is decreased and increased, respectively, in day 2 adult mbl-1(tm1563) mutants ( Figure 3G and Supplementary Table S8). Since VHP-1 negatively regulates p38 MAPK/PMK-1phosphorylation (Kim et al. 2004), it is likely that the elevated expression of vhp-1 contributes to the reduced p38 MAPK/PMK-1 activity in mbl-1(tm1563) mutants.
To determine whether the decreased p38 MAPK/PMK-1 activity contributes to the short lifespan of mbl-1(tm1563) mutants, we studied how these mutants are affected by the knockdown of pmk-1. RNAi of pmk-1 reduces N2 lifespan but does not affect the lifespan of mbl-1(tm1563) mutants ( Figure 3H and Supplementary  Tables S1 and S2). Next, we tested whether the shortened lifespan upon loss of MBL-1 is affected by the depletion of other components of the p38 MAPK pathway ( Figure 3E). We found that   Figure S3 for Western blot repeats and Supplementary Table S6 for Western blot quantifications.
To study a role for mitochondrial stress on the mbl-1(tm1563) mutant lifespan, we used RNAi of mitochondrial respiratory chain complex IV subunit cox-5B (also known as cco-1), which is known to extend lifespan (Dillin et al. 2002). As cox-5B RNAi also promotes innate immunity (Matilainen et al. 2017), we hypothesized that it leads to activation of p38 MAPK/PMK-1 kinase. Strikingly, we found that cox-5B RNAi leads to a significantly larger (P < 0.01, Cox proportional hazard regression analysis) lifespan extension in mbl-1(tm1563) mutants (þ31%) compared to N2 animals (þ15.3%) [HR compared to N2 EV: N2 cox-5B RNAi 0.42, mbl-1(tm1563) EV 3.426, mbl-1(tm1563) cox-5B RNAi 0.302; HR compared to mbl-1(tm1563) EV: mbl-1(tm1563) cox-5B RNAi 0.088] ( Figure 5A and Supplementary Tables S1 and S2). To test whether cox-5B RNAi affects the p38 MAPK pathway, we performed Western blot to analyze p38 MAPK/PMK-1 phosphorylation. We found that cox-5B knockdown does not increase the level of phosphorylated p38 MAPK/PMK-1 in L3 larvae ( Figure 5, B and C and Supplementary Figure S4 and Table S6). At L4 larval stage, cox-5B RNAi increases the level of phosphorylated p38 MAPK/PMK-1 in both N2 and mbl-1(tm1563) mutants in all individual experiments ( Figure 5B and Supplementary Figure S4 and Table S6), but due to the variation between experiments, the increase is not statistically significant ( Figure 5D). Also in day 2 adults, the effect of cox-5B RNAi on p38 MAPK/PMK-1 phosphorylation varies between experiments, and the increase in p38 MAPK/PMK-1 phosphorylation is not statistically significant ( Figure 5, E and F and Supplementary Figure S4 and Table S6). However, in day 2 adult, mbl-1(tm1563) mutants cox-5B RNAi increases the level of phosphorylated p38 MAPK/PMK-1 in four out of five independent experiments (Supplementary Figure S4 and Table S6), thus indicating that similar to L4 larvae, cox-5B RNAi may modulate the activation p38 MAPK/PMK-1 in day 2 adults. Interestingly, it has been shown previously that RNAi of mitochondrial respiratory chain complex IV subunit cox-4 (W09C5.8) does not promote the phosphorylation of p38 MAPK/PMK-1 (Chikka et al. 2016). Therefore, it is possible that knockdown of some, but not all, complex IV subunits activate p38 MAPK/PMK-1, which could be due to their role in complex IV function. Additionally, RNAi of separate complex IV subunits may lead to different knockdown efficiencies, which may also explain the divergence between our observations and the previously published data (Chikka et al. 2016).

Discussion
The data presented here support the hypothesis that loss of MBL-1 disrupts the mechanism maintaining basal p38 MAPK/PMK-1 activity in normal conditions, which leads to reduced activity of transcription factors ATF-7 and SKN-1 and shortened lifespan ( Figure 7C). Since the decline in basal p38 MAPK/PMK-1 activity is an aging-associated process (Youngman et al. 2011), it can be hypothesized that the aberrant p38 MAPK signaling contributes to the premature aging-phenotype observed with MBNL deficiency in DM1 patients. Although mbl-1(tm1563) mutants have a decreased level of activated p38 MAPK/PMK-1 at nonpathogenic conditions, they do not show increased susceptibility to pathogens ( Figure 2B and Supplementary Tables S1 and S2). This demonstrates that p38 MAPK/PMK-1 regulates lifespan independently of its role in innate immunity. In this context, chronic stress caused by aging differs from acute stress caused by pathogen infection. Therefore, we propose that moderate reduction in p38 MAPK/PMK-1 activity shortens mbl-1(tm1563) mutant lifespan, but it retains sufficient activity to respond to acute pathogenic bacterial stress. On the other hand, upregulation of innate immunity-associated genes in mbl-1(tm1563) mutants (Figure 2A) suggests that other pathogen surveillance-mechanisms may compensate the reduced p38 MAPK/PMK-1 activity, thus ensuring the normal survival of mbl-1(tm1563) mutants on PA14 ( Figure  2B).
Importantly, the role of p38 MAPK/PMK-1 as a regulator of lifespan under nonpathogenic conditions is controversial, as it has been shown that pmk-1(km25) mutants have either normal    Figure S4 for Western blot repeats and Supplementary Table S6 for Western blot quantifications.
Since PMK-1 is required for response to biotic stimulus, one possible explanation for differences between laboratories could be due to bacterial growth conditions, which in turn may be affected by brands of peptone and agar used for plate media. In this context, since C. elegans lifespan is regulated by their adaptive capacity to different diets (Pang and Curran 2014), lifespan analyses performed on HT115 and OP50 may result in different outcomes. Interestingly, although we found that both pmk-1(km25) mutants and pmk-1 RNAi-treated animals have shortened lifespans, the HR is bigger for pmk-1 RNAi-treated animals in Figure 3H (HR  In (A) and (B), bars represent mRNA levels relative to N2 EV with error bars indicating mean 6 SD of three biological replicates, each with three technical replicates (*P < 0.05, **P < 0.01, unpaired Student's t-test). (C) The expression of PMK-1 regulated genes in L4 larvae N2 and mbl-1(tm1563) mutants upon cox-5B RNAi. The EV bars are from the same experiment presented in (A). (D) The expression of selected PMK-1 targets in L4 larvae mbl-1(tm1563) mutants and pmk-1(km25); mbl-1(tm1563) double mutants. In (C) and (D), bars represent mRNA levels relative to (C) N2 EV or (D) mbl-1(tm1563) EV with error bars indicating mean 6 SD of three biological replicates, each with three technical replicates (*P < 0.05, **P < 0.01, two-way ANOVA with Tukey's test). See Supplementary Table S8 for raw qRT-PCR data.  A and B) cox-5B RNAitreated mbl-1(tm1563) mutants have extended longevity compared to cox-5B RNAi-treated N2 (P < 0.01), pmk-1(km25) (P < 0.01) and pmk-1(km25); mbl-1(tm1563) (P < 0.01) mutants. Panel (B) shows the subset of lifespan curves presented in (A). See Supplementary Tables S1 and S2 for lifespan statistics.
(C) Model based on the data presented here. compared to N2 EV: 5.062) than for pmk-1(km25) mutants in Figure 7A (HR compared to N2 EV: 2.406) and in Supplementary Figure S7A (HR compared to N2 EV: 2.769). Although conclusions should not be made by comparing independent lifespan experiments, these data suggest that acute loss of PMK-1 could be more detrimental relative to chronic PMK-1 depletion. Although demonstrating that the MBL-1 alternative splicing factor is required to maintain the basal p38 MAPK/PMK-1 activity and normal lifespan, this study leaves an open question about the molecular mechanism behind this phenotype. However, our data support that increased vhp-1 expression (Figure 3, F and G and Supplementary Table S8) contributes to the reduced p38 MAPK/PMK-1 activity in mbl-1(tm1563) mutants. Interestingly, in addition to alternative splicing, MBNL also regulates mRNA stability (Masuda et al. 2012;Xu et al. 2021). Hence, it is possible that MBL-1 affects the activity of p38 MAPK/PMK-1 by promoting vhp-1 mRNA decay ( Figure 7C). Furthermore, p38 MAPK can also be activated by mechanical stimuli (Nguyen et al. 2000;Zhang et al. 2000;Sawada et al. 2001;Wang et al. 2005;Lewthwaite et al. 2006;Chaudhuri and Smith 2008;Dolinay et al. 2008;Hoffman et al. 2017). Since mbl-1 mutants display the improper formation of neuromuscular junction synapses and altered locomotion (Spilker et al. 2012), one intriguing hypothesis is that dysfunctional neuromuscular communication and the consequent abnormal muscle function leads to lower level of mechanical stimuli, and thus to decreased basal p38 MAPK/PMK-1 activity. In this context, one interesting experiment would be to address whether electrical stimulation promotes basal p38 MAPK/PMK-1 activity and increases mbl-1 mutant lifespan.
In addition to the model that MBL-1 promotes normal lifespan by maintaining the basal p38 MAPK/PMK-1 signaling ( Figure 7C), our data show that mitochondrial stress promotes the longevity of mbl-1(tm1563) mutants in a partially PMK-1-dependent manner (Figure 7, A Table S8) indicate that cox-5B knockdown modulates the activity of p38 MAPK/PMK-1, it must be taken into account that p38 MAPK can also be activated by acetylation (Pillai et al. 2011), raising the possibility that mitochondrial stress promotes p38 MAPK/PMK-1 function through different mechanisms. Importantly, overactive p38 MAPK signaling is toxic and shortens lifespan (Cheesman et al. 2016), and therefore, it can be assumed that p38 MAPK signaling is only moderately activated during noninfectious conditions such as mitochondrial stress. Research in mammalian systems has demonstrated that p38 MAPK functions as a signaling hub that converts signals from multiple upstream factors into cellular phenotypes by modulating the function of several downstream effectors (Zarubin and Han 2005). Hence, it is possible that MBL-1 and mitochondrial function are among multiple factors fine-tuning the activity of p38 MAPK signaling, which determine its effect on gene expression, and consequently, on longevity. Since p38 MAPK/PMK-1 is only partially required for the increased mbl-1(tm1563) mutant lifespan upon mitochondrial stress (Figure 7, A and B, Supplementary Figure S7, A and B and Tables S1 and S2), there are also other mechanisms involved. Since DM1, which is associated with reduced MBNL1 activity, has been liked with mitochondrial changes (Ono et al. 1986;Gramegna et al. 2018), and MBNL1 is required for normal mitochondrial function in C2C12 cells (Yokoyama et al. 2020), it is likely that the loss of MBL-1 disrupts mitochondrial function [although mitochondrial genes are not enriched among differentially expressed genes in mbl-1(tm1563) mutants, Supplementary   Table S4]. Therefore, mitochondrial perturbation-activated stress responses may restore the homeostasis in this organelle, thus leading to strikingly increased lifespan.
Taken together, these data elucidate an important mechanism of how the loss of DM1-associated muscleblind splicing factor affects lifespan. In terms of quality of life, it has been proposed that targeting aging may be a more efficient approach compared to targeting individual diseases (Kaeberlein et al. 2015). Therefore, since DM1 is a multisystemic disease with similarities to aging (Mateos-Aierdi et al. 2015;Meinke et al. 2018), this study suggests that targeting the aging process could provide a powerful complementary therapeutic approach for this severe disorder.