Further Extension of Lifespan by Unc-43/CaMKII and Egl-8/PLCβ Mutations in Germline-Deficient Caenorhabditis elegans

Abstract

Reduction of insulin/insulin-like growth factor 1 (IGF1) signaling (IIS) promotes longevity across species. In the nematode Caenorhabditis elegans, ablation of germline stem cells (GSCs) and activity changes of the conserved signaling mediators unc-43/CaMKII (calcium/calmodulin-dependent kinase type II) and egl-8/PLCβ (phospholipase Cβ) also increase lifespan. Like IIS, these pathways depend on the conserved transcription factor daf-16/FOXO for lifespan extension, but how they functionally interact is unknown. Here, we show that altered unc-43/egl-8 activity further increases the lifespan of long-lived GSC-deficient worms, but not of worms that are long-lived due to a strong reduction-of-function mutation in the insulin/IGF1-like receptor daf-2. Additionally, we provide evidence for unc-43 and, to a lesser extent, egl-8 modulating the expression of certain collagen genes, which were reported to be dispensable for longevity of these particular daf-2 mutant worms, but not for other forms of longevity. Together, these results provide new insights into the conditions and potential mechanisms by which CaMKII- and PLCβ-signals modulate C. elegans lifespan.

1. Introduction

Studies in invertebrate model organisms, such as the roundworm Caenorhabditis elegans and the fruit fly Drosophila melanogaster, provided key insights into the biology of lifespan regulation that apparently are also applicable to human aging [1]. A prototypic example is provided by reduced insulin/insulin-like growth factor 1 (IGF1) signaling (IIS). IIS was first recognized in C. elegans to cause a dramatic, ~two-fold increase in lifespan and, subsequently, was confirmed to promote longevity in mammals [1]. In C. elegans, lifespan extension by reduced IIS, e.g., by reduction-of-function mutations in the daf-2 gene (the common ortholog of mammalian insulin and IGF1 receptors), is dependent on daf-16 (the ortholog of mammalian FOXO transcription factors) [1]. Of note, genetic variants in FOXO3 have been repeatedly and robustly associated with lifespan in humans [2].

Beyond daf-2 mutation, additional mechanisms for lifespan extension have been described in C. elegans, including two pathways that can even further extend the extraordinary long lifespan of daf-2 mutant worms: the absence of germline stem cells (GSCs) [3,4], and a gain-of-function mutation in the calcium/calmodulin-dependent kinase type II (CaMKII), unc-43 [5]. This further extension of daf-2 lifespan raises the possibility that the germline- and the unc-43 signaling pathways are at least in part mechanistically different from the IIS/daf-2 pathway. Interestingly, both the germline- and the unc-43 pathway, just as the daf-2 pathway, depend on daf-16 for lifespan extension [3,5]. Whether other factors that promote longevity in GSC- and in daf-2 deficient animals, such as hlh-30 (TFEB), hsf-1 (HSF1), and skn-1 (NRF2) [6,7,8,9], are also shared with the unc-43 pathway has not yet been investigated.

Several lines of evidence indicate that activation and function of the FOXO-transcription factor DAF-16 differ between GSC(−) and daf-2(−) C. elegans. For example, only GSC(−) worms require the conserved adaptor protein KRI-1 and the translation elongation factor TCER-1 for lifespan extension [10,11]. Moreover, whilst DAF-16 acts in the intestine to ensure GSC(−) longevity, it acts in the intestine and neurons to ensure daf-2(−) longevity [12]. DAF-16/FOXO function is strongly regulated by multiple protein kinases [13], including, but not limited to, the CaMKII-ortholog UNC-43. UNC-43 promotes DAF-16 nuclear localization by direct phosphorylation at a site different from the sites at which DAF-16 is phosphorylated and inhibited by the kinase AKT downstream of IIS [5]. CaMKII, in turn, is activated upon a sudden increase in intracellular calcium ion (Ca²⁺) levels by binding to Ca²⁺-calmodulin complexes [14]. Many additional mediators of Ca²⁺-signaling are conserved across metazoans and present in C. elegans, including multiple types of sarcoplasmic/endoplasmic reticulum or plasma membrane channels, G-protein coupled receptors (GPCRs), phospholipase C (PLC) enzymes, and Ca²⁺-pumps [15]. Among these, the G-protein α-subunit egl-30, the γ-aminobutyric acid type B (GABA_B) receptor gbb-1, and the PLCβ-ortholog egl-8, through their function in neurons, also contribute to intestinal daf-16 activation and longevity [16,17,18].

Given that both, germline-ablation and unc-43 gain-of-function mutation, activate DAF-16 and further extend daf-2(−) lifespan, we hypothesized that the two pathways are largely overlapping. We tested this hypothesis in the present study and provide evidence for a novel longevity-promoting mechanism triggered by unc-43(gf) and also, albeit more weakly, by unc-43 and egl-8 loss, which may explain their different effects on various lifespan-extending pathways in C. elegans.

2. Materials and Methods

2.1. C. elegans Strains and Culture

Strains used in this study are listed in Supplementary Table S1. Worms were cultured following standard protocols on NG agar plates seeded with E. coli OP50 [19]. Worms carrying the glp-1(e2144ts) mutation served as a genetic model for germline ablation [4,9]. To eliminate germ cells, glp-1(ts) strains (referred to as GSC(−) in Results/Discussion), and corresponding glp-1(+) (i.e., GSC(+)) control strains were incubated at 25 °C for the first 24 h of postembryonic development and subsequently shifted to 20 °C for the remainder of the experiment. daf-2(e1370) worms and corresponding daf-2(+) control worms were continuously cultured at 20 °C. For experiments, all strains were routinely cultured on plates containing 20 µM 5-Fluoro-2′-deoxyuridine (FUDR; Sigma-Aldrich/Merck, Munich, Germany) from L4 onward to avoid progeny development and internal hatching in fertile strains carrying the unc-43(gf) mutation (allele n498).

2.2. Lifespan Analysis

Worms were synchronized by hypochlorite treatment, cultured under the appropriate temperature regimen (cf. above), and scored for survival every other day starting on days 8–10 of adulthood. From the late L4 stage onward, animals were maintained on 6 cm plates at a density of 40 worms/plate and every 10 days, they were transferred to fresh OP50-seeded FUDR-containing NG agar plates to prevent progeny development and desiccation, respectively. Worms were considered dead if they did not respond to gentle touching with a worm pick. Animals that showed a protruding vulva or had ruptured, died from internal progeny hatching (bagging), or escaped from the plate were censored.

2.3. Stress Resistance Assays

Worms were synchronized by hypochlorite treatment, cultured under the appropriate temperature regimen (cf. above), and transferred to assay plates on day 2 of adulthood (20–30 worms per 3 cm plate for heat stress experiments, 50–60 worms per 3 cm plate for oxidative stress experiments). Survival was scored every 1–2 h. Oxidative stress assay plates contained 15.4 mM tert-butyl hydroperoxide (TBHP) and were prepared 12 h before starting the experiment.

2.4. Growing Worms for RNA-Extraction

To obtain synchronized populations, gravid adults were treated with hypochlorite, and eggs were allowed to hatch in M9 overnight. ~700 L1 larvae per strain were plated on 10 cm NG agar plates seeded with concentrated E. coli OP50 (for RNA-seq: 1600 worms/2 plates) and cultured at the required temperatures (cf. above). At the L4-stage, 500 worms per strain were manually transferred to two E. coli OP50-seeded 6 cm NG agar plates supplemented with 20 µM FUDR to inhibit germ cell proliferation and progeny production (for RNA-seq: 1500 worms to two 10 cm plates). At day 1 of adulthood, worms were harvested by washing them off their plates with M9 buffer. After additional washing with M9 and RNAse-free water, worms were suspended in 1 mL Trizol, snap-frozen in liquid nitrogen, and stored at −80 °C until RNA extraction.

2.5. RNA-Extraction

RNA was extracted using Trizol and cleaned up with the Monarch Total RNA Miniprep Kit or RNA Cleanup Kit (New England Biolabs, Ipswich, MA, USA) according to the manufacturer’s instructions. Equal amounts of RNA extracted from three biological replicates were pooled for RNA-sequencing to mitigate batch effects.

2.6. RNA-Sequencing

RNA quality control, library preparation, and 50-bp paired-end RNA-sequencing on the Illumina HiSeq4000 platform were performed at Eurofins Genomics, Ebersberg, Germany. For each sample, a minimum of 31.6 × 10⁶ reads (maximum 39.1 × 10⁶, median 37.4 × 10⁶ reads) were obtained, resulting in at least 15.7× genome coverage (maximum 19.5×, median 18.7×). A minimum of 94.5% of reads (maximum 96.3%, median 95.7%) could be mapped to the C. elegans reference genome (cf. below).

2.7. RNA-Seq Data Analysis

The following tools provided by the European Galaxy Server (Freiburg Galaxy Team, Freiburg, Germany), at https://usegalaxy.eu (accessed on 30 May 2020) were used for the initial analysis steps: FastQC (v0.72) and MultiQC (v1.7) for initial quality control; Trimmomatic (v0.36.5; ILLUMINACLIP with default settings, LEADING:20, TRAILING:20, SLIDINGWINDOW:5:20, MINLENGTH:20) for gentle trimming of reads; STAR (v2.7.2b) for alignment to the C. elegans reference genome (WS276 release); and featureCounts (v1.6.4) for read summarization [20,21,22,23]. All subsequent analysis steps were performed using Bioconductor (v3.12; [24]) and custom R scripts (v4.0.4; R core team, Vienna, Austria). Differential expression analysis was performed using NOIseq (v2.34.0; [25,26]). Low counts (<5) were filtered by the CPM method, and replicate simulation was performed using TMM- or UQUA-normalization [27], and otherwise default settings of NOIseq-sim. Differentially expressed genes (DEGs) were defined as having a probability of differential expression >95%. Only DEGs that were detected in 20 repetitions of replicate simulation and that displayed an expression-fold chance ≥1.5 with each of the two normalization methods in each of the three repetitions of the entire NOIseq-analysis were considered further. PCA-analysis of CPM-filtered, TMM/UQUA-normalized RNA-seq count data was performed using the NOIseq PCA function. Heatmaps of filtered, normalized, log-transformed count data or of sample-to-sample distances were generated using the R package pheatmap (v1.0.12). Complete NOIseq-data is provided in Supplementary Table S14. Supplementary Tables S15 and S16 list NOIseq M and probability values for all DEGs, along with gene ontology (GO) term information.

2.8. Overlap of DEG-Lists

The statistical significance of overlaps between gene lists was calculated as the hypergeometric probability of detecting at least as many common genes as observed in the two lists using the phyper function in R. Representation factors were calculated as the number of overlapping genes divided by the expected number of overlapping genes in the two lists (http://nemates.org/MA/progs/overlap_stats.html; accessed on 31 August 2020). For all calculations, the number of genes in the genome was set to 10,602, i.e., the number of genes in Wormbase WS276 that passed CPM-filtering during NOIseq-analysis. DEG-lists from published studies were, if necessary, converted to WBGene-IDs using WormMine (http://intermine.wormbase.org/tools/wormmine/begin.do; accessed on 10 August 2020) and adjusted to genes in WS276 using custom R-scripts and manual curation.

2.9. GO Term Enrichment Analysis and Collagen/Matrisome Classification of DEG-Lists

GO term enrichment analysis was performed using the enrichment tool provided by Wormbase (https://wormbase.org/tools/enrichment/tea/tea.cgi; accessed on 27 May 2021) [28]. The q-value threshold was set to <0.05. GO term categories were retrieved using REViGO ([29]; http://revigo.irb.hr/; accessed on 27 May 2021). For WormCat analysis [30], the respective R package was downloaded and installed as described on www.wormcat.com (accessed on 9 January 2022). Collagen and matrisome classification [31] was performed using the lists provided at http://CeColDB.permalink.cc/ (accessed on 3 January 2022) and http://ce-matrisome-annotator.permalink.cc/ (accessed on 4 January 2022) and custom R scripts.

2.10. Dauer Analysis

Synchronized (hypochlorite treatment) L1 larvae were plated at a density of ~50 larvae/6 cm plate (2 plates per strain), incubated at 20 °C or 25 °C, and inspected for the presence of dauers every 12 h. At the 60 h (20 °C) or 48 h (25 °C) timepoint, when wildtype worms reached adulthood, worms were morphologically classified as “dauer” or “non-dauer”. Daf-2(−) strains, which all formed 100% dauers when incubated at 25 °C, were allowed to recover from dauer by incubation at 15 °C for 120 h before classification. Experiments in daf-2(+) strains were either conducted in parallel with experiments in daf-2(−) strains or included daf-2(−) control plates to provide a reference for the typical dauer morphology.

2.11. qPCR

A total of 1 µg total RNA was reverse transcribed using LunaScript RT SuperMix (New England Biolabs, Ipswich, MA, USA). qPCR-reactions were performed in duplicates or triplicates in a 20 µL reaction volume on an CFX Connect Real-Time PCR Detection System (Bio-Rad Laboratories, Hercules, CA, USA) with iTaq Universal SYBR Green Supermix (Bio-Rad Laboratories). The thermal cycling protocol comprised one activation step at 95 °C for 3 min, followed by 40 cycles of denaturation at 95 °C for 10 s and combined annealing/extension at 60 °C for 30 s. Melting curve analysis was performed from 65 °C to 95 °C with 0.5 °C increments at 5 s per step. Data were analyzed by the ΔΔCt method, and target gene expression levels were normalized to the geometric means of cdc-42, tba-1, and Y45F10D.4 expression [32,33]. Primer sequences are listed in Supplementary Table S2.

2.12. Statistical Analysis

Statistical analysis was performed using Prism 5 or 9 (GraphPad Software, San Diego, CA, USA). Details on the particular tests used are specified in the figure legends.

3. Results

3.1. Unc-43(gf) and Unc-43(−) Further Extend Lifespan of Germline-Deficient C. elegans

To test the hypothesis that germline ablation extends C. elegans lifespan at least in part by activating unc-43/CaMKII, we generated a set of relevant double mutants and measured their lifespans. Thereby, we took advantage of a widely used genetic model for germline stem cell deficiency, glp-1(e2144ts) (hereafter referred to as GSC(−); [4]) and combined this allele with the previously studied [5] unc-43 alleles n498 and n498n1186 (referred to as unc-43(gf) and unc-43(−), respectively). All experiments were conducted in the presence of FUDR (cf. Materials and Methods), to avoid internal progeny hatching in fertile unc-43(gf) worms and to facilitate comparison with similar previous studies that also took advantage of FUDR to prevent progeny development in fertile strains [5,8,17,18,34]. In GSC(+), i.e., otherwise wildtype worms, unc-43(gf) robustly extended lifespan (36–87%; Figure 1A, Supplementary Table S3), consistent with a previous report [5]. However, in contrast to this previous report, unc-43(−) also consistently produced a small to moderate lifespan increase in otherwise wildtype worms (6–21%; Figure 1A, Supplementary Table S3). Similarly, in GSC(−) worms, unc-43(gf) substantially extended lifespan (67–92%), while unc-43(−) had a more modest effect (2/3 experiments; 22–24%; Figure 1B, Supplementary Table S3). Collectively, these data cannot easily be reconciled with our original hypothesis and instead indicate that unc-43(gf) and unc-43(−) both trigger longevity-promoting mechanisms. Moreover, these mechanisms apparently are not, or are not maximally, triggered by germline ablation.

3.2. Loss of Egl-8 Also Extends Lifespan of Germline-Deficient C. elegans

In parallel with the unc-43 alleles, we investigated how loss of egl-8/PLCβ influences the lifespan of GSC(−) worms. As unc-43(−), egl-8(md1971), a putative loss-of-function allele, extended the lifespan of otherwise wildtype worms (13–38%; Figure 1A, Supplementary Table S3), in agreement with published work [16,17,18]. Moreover, egl-8(md1971) further extended lifespan in GSC(−) worms (30–43%, Figure 1B, Supplementary Table S3). Similar results were obtained with another putative loss-of-function allele, egl-8(e2917) (Supplementary Figure S1A,B, Supplementary Table S3). Thus, egl-8(−), similar to unc-43(−) and unc-43(gf) appears to induce lifespan-extending programs that are not, or not fully, active in GSC(−) worms.

3.3. Unc-43(gf), Unc-43(−) and Egl-8(−) Do Not Further Extend Lifespan of Daf-2(−) Worms

In addition to our lifespan analyses on unc-43(gf), unc-43(−) and egl-8(−) in intact and in GSC-deficient animals, we examined the effect of these alleles on the lifespan of worms harboring a strong reduction-of-function mutation in the insulin/IGF-1 like receptor daf-2 (allele e1370) [35]. As mentioned above, in wildtype (daf-2(+)) animals, unc-43(gf) substantially and unc-43(−) more modestly extended lifespan (34–82% and 5–23%; Figure 1C, Supplementary Table S3). In daf-2(−) worms, however, unexpectedly in the light of published work by others [5], unc-43(gf) did not further extend but rather slightly decrease lifespan (9–13%; Figure 1D, Supplementary Table S3). Moreover, unc-43(−) did not significantly decrease daf-2(−) lifespan. Consistent with our previous observations (Figure 1A), both egl-8(−) alleles extended lifespan in daf-2(+) worms (27–39%; Figure 1C, Supplementary Figure S1C, Supplementary Table S3). Finally, egl-8(−), similar to unc-43(−), had essentially no effect on daf-2(−) lifespan (Figure 1D, Supplementary Figure S1D, Supplementary Table S3). In summary, irrespective of the discrepancies with previously published work [5], our results suggest that unc-43(gf), unc-43(−) and egl-8(−) all trigger longevity-promoting pathways that are already highly active, or that are dispensable for longevity, in daf-2(−) worms.

3.4. Unc-43(gf), Unc-43(−) and Egl-8(−) Differentially Affect Stress Resistance of Wildtype, GSC(−) and Daf-2(−) Worms

Longevity frequently correlates with stress resistance [1]. Therefore, we examined how our gain- and loss-of-function alleles in unc-43 and egl-8 affect the resistance of GSC(−), daf-2(−) and otherwise wildtype worms to selected forms of physical stress. Specifically, we examined heat and TBHP-induced oxidative stress. Only in daf-2(−) background, unc-43(gf) slightly increased heat stress resistance, while only in otherwise wildtype worms, unc-43(−) conferred heat sensitivity (Figure 2, Supplementary Table S4). On the other hand, upon oxidative stress, unc-43(gf) and unc-43(−) both consistently caused at least trends towards sensitivity in most or all backgrounds, with only the GSC(−); unc-43(gf) strains being clearly not sensitive (Figure 3 and Supplementary Table S4). All egl-8(−) strains consistently tended to be or were more resistant than controls to oxidative stress but more sensitive to heat stress, with the exception of daf-2(−); egl-8(−) strains, which did not consistently show heat sensitivity (Figure 2 and Figure 3, Supplementary Figures S2 and S3, Supplementary Table S4). In summary, in all backgrounds in which they (further) extended lifespan, unc-43(gf), unc-43(−) and egl-8(−) did not confer broad stress resistance but rather sensitivity to at least one of the two physical stressors examined.

3.5. Unc-43(gf) and Egl-8(−) Globally Shift Gene Expression towards Daf-2(−)

In our hands, unc-43(−) and egl-8(−) extended C. elegans lifespan more modestly, but in the same genetic background-dependent manner as unc-43(gf). To identify novel factors and pathways that may mediate unc-43′s and/or egl-8′s lifespan-regulatory functions in addition to daf-16, we performed an exploratory RNA-sequencing (RNA-seq) study in daf-2(−) and otherwise wildtype (daf-2(+)) worms with normal or altered unc-43/egl-8 activity. GSC(−) worms were omitted from this study since unc-43/egl-8 mutations changed their lifespan in essentially the same way as the lifespan of otherwise wildtype worms. Out of the two egl-8 loss-of-function alleles, md1971 was chosen, and hereafter, egl-8(−) refers to md1971 unless otherwise stated. Initial comparison of gene expression between daf-2(−) and wildtype worms revealed highly significant overlaps of differentially expressed genes (DEGs) from our study with published work [36,37], thus validating our data (Supplementary Tables S5 and S6). Hierarchical clustering and principal component analysis clearly separated the wildtype and daf-2(−) control strains from each other, and strikingly, egl-8(−) and especially unc-43(gf) single mutant worms appeared closer to daf-2(−) in these analyses (Figure 4A–C). In line with this result, genes induced by unc-43(gf) and egl-8(−) in otherwise wildtype worms significantly overlapped with genes induced by daf-2(−) in our and in published studies (Supplementary Table S7). Similarly, unc-43(gf)/egl-8(−)- and most to all lists of daf-2(−) repressed genes significantly overlapped with each other. Interestingly, in otherwise wildtype worms, unc-43(−) induced and repressed genes also significantly overlapped with most to all lists of daf-2(−) induced and -repressed genes (Supplementary Table S7). In daf-2(−) worms, unc-43(gf), unc-43(−) and egl-8(−) seemed to support some gene expression changes already induced by daf-2(−), while opposing others (Supplementary Table S7). Similar overlap patterns resulted when comparing our lists of unc-43(gf), unc-43(−) and egl-8(−)-regulated DEGs to genes regulated by daf-2(−) in a daf-16 dependent manner [38,39,40,41] (Supplementary Table S7). Taken together, our RNA-seq results indicate that unc-43(gf) and egl-8(−) globally shift gene expression towards daf-2(−), and they are consistent with previously reported roles of unc-43(gf) and egl-8(−) as positive regulators of daf-16 [5,16,17,18].

3.6. Unc-43 and Egl-8 Regulated Genes Are Enriched for Collagen-Related Processes

To functionally categorize unc-43(gf), unc-43(−) and egl-8(−) DEGs in otherwise wildtype and in daf-2(−) worms, we performed gene ontology (GO) term enrichment analysis [28] (Figure 4D,E, Supplementary Table S8). This analysis revealed overrepresentation of GO terms reminiscent of known functions of unc-43 and egl-8 in neurons, muscle, and immunity, respectively [16,42]. Moreover, multiple GO terms enriched among DEGs from unc-43(gf), unc-43(−) or egl-8(−) strains were also enriched among DEGs between daf-2(−) and wildtype worms, consistent with the previously observed overlaps between the respective DEG-lists (Supplementary Table S9). Strikingly, the most significantly enriched GO term among upregulated DEGs in all three daf-2(+) strains was “structural component of cuticle”, followed by additional GO terms containing many collagen genes (e.g., “collagen trimer”, “cuticle development”, and “molting cycle”), and by multiple GO terms related to pathogen defense. “Structural component of cuticle” and pathogen defense-related GO terms were also enriched among upregulated DEGs of most to all daf-2(−) strains. On the other hand, unc-43(gf), unc-43(−) and egl-8(−) downregulated DEGs were most significantly and commonly enriched for pathogen defense related GO terms in all daf-2(+)- and in the daf-2(−); egl-8(−) strain, and for organic acid-metabolism-related GO terms in daf-2(−); unc-43(gf) and daf-2(−); unc-43(−) worms. Enrichment of similar functional categories was detected by a different analysis tool, WormCat [30], including, most prominently among unc-43(gf), unc-43(−) and egl-8(−)-induced genes, “extracellular material: collagen” (Supplementary Figure S4 and Table S9). Given the conditional requirement of collagens for C. elegans longevity [34], we focused on these genes for subsequent analyses.

3.7. Unc-43 and Egl-8 Modulate the Expression of Non-Dauer Longevity-Associated Collagen Genes

Certain cuticle-forming collagens have been shown to be required for longevity, specifically under conditions that do not induce processes related to dauer, an extremely stress-resistant alternative developmental stage, during adulthood [34]. Expression of these collagens appears to be indirectly controlled by the conserved transcription factor SKN-1/NRF2 [34]. Non-dauer-associated, collagen-dependent lifespan-extending conditions include germline deficiency and certain means of reducing IIS, but not the daf-2(−) allele analyzed by us, daf-2(e1370), at least not at the standard culture temperature of 20 °C at which we conducted our study [34]. Unc-43(gf), unc-43(−), and egl-8(−) do not sensitize C. elegans to dauer entry at 20 °C or 25 °C (Supplementary Figure S5 and Table S9). Therefore, we examined the effect of these alleles on the expression of “SKN-1 upregulated daf-2(−) collagens” ([34], subsequently referred to as “non-dauer longevity-associated collagens”). Significant numbers of these collagens were present in our lists of unc-43(−), egl-8(−) and especially unc-43(gf) upregulated DEGs from daf-2(+), as well as among unc-43(gf) upregulated DEGs from daf-2(−) worms (Supplementary Tables S10–S12). Up- and downregulated genes from these strains also significantly overlapped with the complete lists of genes up/downregulated by skn-1 under non-dauer longevity conditions. Yet, only egl-8(−) appeared to up- and downregulate significant numbers of genes that are up- or downregulated by skn-1 under normal conditions (Supplementary Table S11). Classification of collagen-genes [31] induced by unc-43(gf), unc-43(−) and egl-8(−), especially in daf-2(+) backgrounds, revealed strong overrepresentation of cuticular collagens, as also seen for skn-1 upregulated (non-dauer longevity-associated) collagens ([34], Supplementary Table S12). When more broadly examining genes of the extracellular matrix, i.e., the matrisome [31], similar patterns were observed as well for unc-43(gf), unc-43(−), egl-8(−) and skn-1(−) induced DEG-sets, all of which showed particular enrichment of nematode-specific core matrisome genes (Supplementary Table S12). Of note, collagen or core-matrisome genes were not overrepresented in 5/6 published sets of daf-16-induced genes (Supplementary Table S12). qPCR analysis of selected non-dauer longevity-associated collagens [34] confirmed the picture suggested by our RNA-seq data: col-120, col-133, col-141, and col-176 all were strongly upregulated in unc-43(gf) relative to otherwise wildtype worms, while induction in unc-43(−) and egl-8(−) was less consistent and/or weaker (Figure 5A,C, Supplementary Table S13). In GSC(−) and daf-2(−) worms, col-120, col-141, and col-176 upregulation by unc-43(gf) was repeatedly observed, while unc-43(−) did not trigger strong alterations (Figure 5B,D, Supplementary Table S13). Additionally, in daf-2(−) worms, egl-8(−) appeared to at least mildly suppress all four collagen genes examined (Figure 5D, Supplementary Table S13). Relative to wildtype, both GSC(−) and daf-2(−) worms displayed at least trends towards elevated mRNA levels for most of the collagen genes examined (Supplementary Figure S6). Collectively, these results suggest that unc-43(gf), and to a smaller extent, unc-43(−) and egl-8(−), promote the expression of non-dauer longevity-associated collagens. Moreover, the effects of the three alleles appear strongest in GSC(+)/daf-2(+), and weakest, or even lost, in daf-2(−) background.

4. Discussion

In this study, we report lifespan-regulatory functions of the conserved signaling mediators unc-43 and egl-8 in long-lived GSC-deficient C. elegans. Moreover, our experiments confirm previously published positive roles of unc-43 hyperactivation and egl-8 loss for longevity of otherwise wildtype animals and further support the concept that unc-43 and egl-8 modulate the activity of the conserved lifespan-regulatory key transcription factor DAF-16 [5,16,17,18]. In addition, gene expression analyses indicate, to at least some extent, the induction of specific non-dauer longevity-associated collagens in response to unc-43 and egl-8 mutations. Currently available data do not support a prominent role for daf-16 in promoting expression of these collagens, raising the interesting possibility that they constitute a second, largely independent mechanism through which unc-43 and egl-8 regulate C. elegans lifespan (Figure 6).

A key finding of our study consists in the observation that altered unc-43 and egl-8 activities further extend the lifespan of wildtype and GSC-deficient but not of daf-2(−) (i.e., daf-2(e1370) at 20 °C) worms. This result may be explained by at least two, not mutually exclusive, models. First, unc-43(gf), unc-43(−) and egl-8(−) all may trigger processes that are already maximally induced by daf-2(−), but not by GSC(−) and wildtype worms. This model is supported by the fact that all three mutations induce similar gene expression changes than daf-2(−). Alternatively, processes triggered by unc-43(gf), unc-43(−) and egl-8(−) may not be critical to daf-2(e1370) longevity. Consistent with this possibility, unc-43(gf) and, to a lesser extent, unc-43(−) and egl-8(−), induce the expression of certain collagen genes, which are required for longevity under conditions that do not induce dauer-like traits in adults. These conditions include germline deficiency but not the daf-2(e1370) allele under our culture conditions of 20 °C [34]. Thus, it is tempting to speculate that these collagens constitute a second mechanism besides the regulation of daf-16, by which in particular unc-43(gf), but also unc-43(−) and egl-8(−) modulate lifespan. Of note, overexpression of a single key collagen, such as col-120, also examined by us by qPCR, is sufficient for lifespan extension [34]. Conversely, in daf-2(e1370), mild induction of collagen expression, which is still observed, at least in response to unc-43(gf), may not be functionally relevant. Clearly, it will be extremely interesting to directly examine the role of non-dauer longevity-associated collagens in unc-43(gf), unc-43(−) and egl-8(−) dependent C. elegans lifespan regulation and, eventually, to define the mechanisms that link unc-43/egl-8 to expression of these collagens. Although we have not yet experimentally confirmed or excluded the possibility that the known mediator of unc-43(gf)/egl-8(−) longevity, daf-16, regulates non-dauer longevity-associated collagen expression in the specific context of unc-43/egl-8 mutation, even though it does not do so in other contexts, an alternative candidate is emerging from the literature: skn-1 [34]. In support of this hypothesis, we already showed that unc-43(gf), unc-43(−) and egl-8(−) modulate the expression of similar gene sets, including similar collagens, than skn-1.

Yet, our study and a previous report [5] on the longevity effects of unc-43(−) and unc-43(gf) differ from each other in several details despite the use of the same alleles. In the case of unc-43(−), these discrepancies may be explained by different experimental conditions, i.e., different temperatures, and by the use of FUDR, both of which have been shown to affect lifespan in certain genetic backgrounds [43,44]. Specifically, we conducted the respective assays at 20 °C rather than 25 °C [5] and analyzed the unc-43(−) and unc-43(gf) alleles within the same experiments, which required the addition of FUDR not just to unc-43(gf) and corresponding control [5], but also to unc-43(−) strains. Importantly, FUDR does not affect wildtype lifespan nor daf-2(−)’s and GSC(−)’s basic ability to promote longevity [8,34,43].

Beyond different assay temperatures, our choice of daf-2(e1370), rather than daf-2 RNAi [5] to inhibit daf-2 activity may account for unc-43(gf) in our hands not extending the lifespan of worms with reduced IIS. Differential interactions of “strong” (e.g., e1370) and “weak” (e.g., RNAi) daf-2 inhibition with other lifespan-regulatory mechanisms are not unprecedented in the literature, e.g., with somatic gonad signaling and collagen expression [3,34,45]. Moreover, in contrast to genetic mutations, daf-2 RNAi does not efficiently reduce daf-2 activity in neurons [12,46,47]. Yet, activity of the common daf-2(−)- and unc-43(gf) lifespan-extending factors daf-16 in neurons accounts for a fraction of daf-2(−) longevity [12]. Thus, unc-43(gf) may trigger daf-16 regulated processes in neurons that are already active in daf-2(e1370), but not in daf-2 RNAi worms, which may explain why it further extends daf-2 RNAi- but not daf-2(e1370) lifespan.

The observation that both, unc-43(−) and unc-43(gf) extend lifespan, at least in certain genetic backgrounds under certain experimental conditions, appears paradoxical at the first glance. On the other hand, the moderate to strong longevity increase of unc-43(−) and unc-43(gf) worms can be expected, based on the nature of the—relatively moderate to relatively large—changes in gene expression triggered by these two mutations. Indeed, as discussed above, unc-43(−) and unc-43(gf) regulated gene sets are similar to each other and to daf-2- and daf-16-regulated gene sets. In addition, both mutations, although to different extents, induce non-dauer longevity-associated collagens. How unc-43(−) and unc-43(gf) affect collagen expression and/or its indirect regulator skn-1 is currently unclear [34]. Previous work [5] suggested UNC-43 as a direct activator of DAF-16, implying that the induction of some daf-16 targets upon unc-43(−) loss, as observed in our RNA-seq study, represents additional, indirect effects of unc-43 on daf-16. Moreover, UNC-43′s direct contribution to DAF-16 activity appears to be dispensable for the maintenance of a normal/normally long lifespan in all genetic backgrounds tested, as unc-43(−), GSC(−); unc-43(−) and daf-2(−); unc-43(−) worms are long-lived, in contrast to daf-16(-), GSC(−); daf-16(-) and daf-2(−); daf-16(-) worms, which are short- to normal-lived [3,48]. As proposed previously for reduced IIS [5], GSC loss and other mechanisms induced by unc-43 loss may “overpower” (direct) DAF-16 regulation by UNC-43.

In agreement with published studies [16,17,18], we found that loss of egl-8 further extends wildtype lifespan. However, in contrast to previous work [16], we did not observe decreased oxidative stress resistance upon egl-8 loss. Again, this difference may be explained by different experimental conditions, such as different egl-8(−) alleles (md1971 vs. n488), different assay temperatures (20 °C vs. 25 °C) and treatment with different chemicals to impose oxidative stress (TBHP vs. arsenite). Indeed, TBHP and arsenite have already been shown to trigger different transcriptional responses [49]. In a current model, egl-8 modulates lifespan cell non-autonomously through its function in neurons: egl-8 loss reduces neuronal secretion of insulin-like peptides, which in turn increases daf-16 activity in the intestine; in addition, evidence suggests that this pathway is further mediated by the EGL-8-generated second messenger diacylglycerol (DAG) and the DAG-dependent kinase dkf-2 [16,17,18,50]. This model predicts that worms with compromised insulin/IGF1-like receptor function in the intestine will not display increased longevity upon an egl-8(−) mutation, while GSC(−) worms may display further intestinal daf-16 activation and lifespan extension. Results from our lifespan analyses are in line with this prediction. Conceptually, egl-8 loss also decreases the levels of another second messenger, inositol-1, 4, 5-trisphosphate and subsequently, intracellular Ca²⁺ levels, and CaMKII/UNC-43 activity [15]. Yet, additional studies are necessary to determine how egl-8(−) and unc-43(−) interact to modulate daf-16 activity and C. elegans lifespan.

In summary, our study identified new genetic backgrounds in which unc-43 and egl-8 can/cannot modulate lifespan in C. elegans and suggests collagen expression as a second lifespan-regulatory mechanism functioning downstream of unc-43 and egl-8, in addition to the known regulator daf-16.