Mitochondria are essential subcellular organelles for cellular energy production [reviewed in (1)]. Mitochondria also play important functions in a wide array of other cellular processes, ranging from cellular signaling to apoptosis. In addition, mitochondria play crucial roles in organismal aging, and functional declines in mitochondria are associated with age-related diseases [reviewed in (2)]. However, mild inhibition of mitochondrial respiration has been shown to promote longevity in multiple species [reviewed in (3)]. In Caenorhabditis elegans, the genetic inhibition of mitochondrial respiration genes, including isp-1 (Rieske iron-sulfur protein in complex III) and clk-1 (coenzyme Q biosynthesis enzyme crucial for electron transport), prolongs life span. Inhibition of mitochondrial respiration also increases life span in Drosophila and mammals. Therefore, life-span extension by reduced mitochondrial respiration is conserved, and the elucidation of the molecular mechanism in C. elegans may enhance our understanding of human aging and longevity.
Adenosine 5′-monophosphate (AMP)–activated protein kinase (AMPK), a critical cellular energy sensor that increases life span in multiple species [reviewed in (4)], is one of the factors required for the enhanced longevity caused by inhibition of mitochondrial respiration in C. elegans (5–7). AMPK is activated by an elevated AMP/adenosine 5′-diphosphate (ADP)–to–adenosine 5′-triphosphate (ATP) ratio that results from reduced cellular energy, leading to increases in catabolic processes and inhibition of anabolic processes [reviewed in (8)]. Activation of AMPK results from phosphorylation at residue Thr172 in its catalytic α subunit by upstream kinases. Liver kinase B1 (LKB1) and Ca2+/calmodulin-dependent protein kinase kinase β (CaMKKβ) are two kinases that function to activate AMPK in mammals [reviewed in (9)]. Transforming growth factor–β–activated kinase 1 (TAK1) has also been shown to phosphorylate AMPK in vitro [reviewed in (9)]. However, the upstream regulators of AMPK that facilitate its life span–extending effects remain incompletely understood.
The vaccinia virus–related kinase (VRK) family consists of three serine-threonine protein kinases (VRK1 to VRK3) in mammals, which are related to casein kinases (10–12). Among these three, the best characterized is VRK1, a cell cycle regulator that is abundant in proliferative tissues (13). Unlike mammals, C. elegans has a single VRK ortholog, VRK-1, whose function in cell proliferation is relatively well established (14–16). Strong loss-of-function mutations of the C. elegans vrk-1 gene results in sterility, and reduced vrk-1 function leads to mislocalization of barrier-to-autointegration factor 1 (BAF-1), a phosphorylation target of VRK-1, resulting in severe mitotic defects (15, 17, 18). vrk-1 is also required for germ cell proliferation, likely through its ability to regulate the p53-like protein, C. elegans p53-like-1 (CEP-1) (16), and plays important roles in the development of vulva and uterus (17, 18). However, it remains unknown whether VRK-1 acts in postmitotic cells or has a role in adult life span.
In this study, we sought to elucidate the role of VRK-1 in regulation of adult life span in C. elegans. We found that overexpression of VRK-1::GFP (green fluorescent protein), which was detected in the nuclei of cells in multiple somatic tissues, including the intestine, increased life span. Conversely, genetic inhibition of vrk-1 decreased life span. We further showed that vrk-1 was essential for the increased life span of mitochondrial respiratory mutants. We demonstrated that VRK-1 was responsible for increasing the level of active and phosphorylated form of AMPK (p-AMPK). In addition, we found that mammalian VRK1 directly phosphorylated AMPK at Thr172, resulting in its increased activity. Together, these data indicate that the nuclear protein kinase, VRK-1, which acts in somatic cells, promotes longevity by increasing the activity of AMPK through phosphorylation.
Nuclear expression of VRK-1 in somatic tissues extends C. elegans life span
To determine the role of vrk-1 in regulating adult life span, we generated transgenic C. elegans overexpressing vrk-1 fused with the GFP gene (vrk-1::GFP). The VRK-1::GFP was predominantly localized to the nuclei of many cells, including neural, intestinal, and hypodermal cells (Fig. 1A and fig. S1, A and B) (16–18). In addition, we found that VRK-1::GFP was highly expressed throughout the C. elegans life cycle, in the soma of larvae (Fig. 1A and fig. S1, A and B) and fully grown adult worms (Fig. 1B and fig. S1C), comprising postmitotic cells after the cells stopped dividing. Notably, extrachromosomal transgenes that overexpress vrk-1::GFP, which is limited to somatic cells because of transgene silencing in germ cells (19), substantially increased adult life span (Fig. 1, C and D, and fig. S2A). We then confirmed the longevity conferred by overexpression of vrk-1::GFP using an integrated transgenic line (Fig. 1, E and F, and fig. S2B). These data suggest that vrk-1 overexpression specifically in somatic cells promotes the longevity of C. elegans.
We next investigated the effect of vrk-1 inhibition on the life span of C. elegans. vrk-1 RNA interference (RNAi) and strong loss-of-function mutation vrk-1(ok1181) (fig. S2G) (15–17) shortened C. elegans life span (Fig. 1, G and H, and fig. S2, E and F). These results indicate that vrk-1 is required for the maintenance of normal life span. Together, these data indicate that VRK-1 is necessary and sufficient for longevity.
VRK-1 contributes to the longevity of mitochondrial respiration mutants
We then asked whether VRK-1 played a role in the increased life span conferred by various mutations in C. elegans. RNAi knockdown of vrk-1 largely suppressed the longevity conferred by mutations in mitochondrial respiration genes, isp-1 (Fig. 2A and fig. S3, A to C) and clk-1 (Fig. 2B). Similarly, the vrk-1(ok1181) mutation suppressed the long life span of isp-1 mutants (Fig. 2C). We then found that genetic inhibition of vrk-1 by RNAi or by the deletion mutation, vrk-1(ok1181), decreased the longevity of insulin/insulin-like growth factor 1 receptor daf-2(−) mutants (Fig. 2, D and E). In addition, vrk-1 RNAi shortened the life span of dietary restriction mimetic eat-2(−) (Fig. 2F) and hypoxia-inducible factor 1 (HIF-1)–hyperactive vhl-1(−) (Fig. 2G) mutants as well as RNAi-mediated electron transport chain–inhibited cco-1(RNAi) animals (fig. S3, F to H; see Fig. 2 legends for discussion). Together, these data suggest that vrk-1 contributes to longevity conferred by various longevity interventions and, in particular, is essential for that caused by the inhibition of mitochondrial respiration.
VRK-1 regulates expression of AMPK target genes
We next investigated the mechanism by which VRK-1 contributed to the longevity of mitochondrial respiration mutants. To this end, we first determined whether inhibition of nuclear protein VRK-1 affected gene expression in isp-1(−) mutants by performing mRNA sequencing (mRNA-seq) analysis. We identified 1589 up-regulated and 199 down-regulated genes in isp-1(−) mutants versus wild-type worms (fold change > 2 and < 0.5, P < 0.001; Fig. 3A and data file S1). Among these differentially expressed genes (DEGs), the expression of 328 up-regulated genes and 22 down-regulated genes was dependent on vrk-1 (Fig. 3A and data file S1). Gene Ontology (GO) analysis indicated that genes involved in diverse biological processes were enriched among vrk-1–dependent up-regulated genes in isp-1(−) mutants, whereas those involved in lipid transport were enriched among the down-regulated genes (Fig. 3B).
We then asked which factors functioned in conjunction with VRK-1 to affect gene expression changes in mitochondrial respiration-defective mutants. We compared the vrk-1–dependent DEGs in isp-1(−) mutants with those regulated by respiration mutation-mediated longevity factors, including AMP-activated kinase 2 (AAK-2)/AMPK, activating transcription factor associated with stress–1 (ATFS-1)/basic leucine zipper transcription factor required for mitochondrial unfolded protein response, HIF-1, SKN-1, C. elegans homeobox 23 (CEH-23), and CEP-1 [WormExp (20)]. Genes whose induction was dependent on vrk-1 in isp-1(−) mutants displayed the most significant overlap with those up-regulated by AMPK (Fig. 3, C to F). In contrast, transcriptomic comparison with ATFS-1, HIF-1, SKN-1, CEH-23, or CEP-1 did not yield a substantial overlap (Fig. 3C and fig. S4).
We then performed quantitative reverse transcription polymerase chain reaction (qRT-PCR) analysis to measure the expression levels of vrk-1–dependent genes induced in isp-1(−) mutants (21–23) and positively regulated by AMPK (24–26) (Fig. 4, A and B). Of the 14 genes that we analyzed, most were up-regulated by the isp-1 mutation in an AMPK-dependent manner (Fig. 4A). Consistent with our RNA sequencing (RNA-seq) results (fig. S5A), we also found that expression of these putative AMPK target genes was decreased by vrk-1 RNAi in isp-1 mutants (Fig. 4B). Moreover, down-regulation of the tested isp-1(−)–dependent AMPK target genes by aak-2(−) mutation was not additive to that by vrk-1 RNAi (fig. S5B). These results support the hypothesis that vrk-1 mediates the up-regulation of AMPK-dependent genes. In contrast, RNAi knockdown of vrk-1 did not suppress the induction of hsp-6p::GFP (ATFS-1 reporter), nhr-57p::GFP (HIF-1 reporter), gst-4p::GFP (SKN-1 reporter), or ceh-23 in isp-1(−) mutants (Fig. 4, C to F, and fig. S5, C to E). Together, these data suggest that VRK-1 induces a large subset of genes that are regulated by AMPK in long-lived mitochondrial respiration mutants.
VRK-1 promotes life extension by up-regulating AMPK, thereby down-regulating CRTC-1
We further determined the relationship between VRK-1 and AMPK in the regulation of C. elegans longevity. We first confirmed that genetic inhibition of aak-2/AMPK partially suppressed the long life span of isp-1 mutants (Fig. 5A and fig. S6A), consistent with previous reports (5, 6). We found that vrk-1 RNAi slightly decreased the life span of isp-1(−); aak-2(−) mutants (Fig. 5B and fig. S6, B and C). These data raise the possibility that vrk-1 has an additional life span–regulatory function that is independent of AMPK. We then performed the converse experiment using vrk-1::GFP-overexpressing animals and showed that aak-2/AMPK was required for the long life span of vrk-1::GFP-overexpressing animals (Fig. 5C and fig. S6D). Thus, VRK-1 appears to promote longevity in mitochondrial respiration mutants by acting at least partially through AMPK.
Previous reports have shown that AMPK promotes longevity by phosphorylating CRTC-1 (cyclic AMP response element–binding protein–regulated transcriptional coactivator 1) and subsequently promoting its nuclear exclusion (24, 26). We found that vrk-1::GFP O/E increased the nuclear exclusion of CRTC-1::RFP (Fig. 5, D and E, and fig. S6J), raising the possibility that VRK-1 overexpression promotes longevity through regulating CRTC-1 via AMPK activation. However, converse experiments using vrk-1 RNAi were negative (Fig. 5, D and E, and fig. S6J), and therefore, inhibition of vrk-1 does not seem to be sufficient to affect the CRTC-1 activity. Overall, up-regulation of VRK-1 may promote longevity, at least in part, acting via the AMPK–CRTC-1 signaling axis.
VRK1 increases the activity of AMPK via phosphorylation
We next asked how VRK-1 up-regulated AMPK target gene expression to promote longevity. Because both VRK-1 and AMPK are protein kinases, we tested whether VRK-1 regulated AMPK activity through phosphorylation. Using cultured human cells, we found that the levels of human phospho-AMPK (p-AMPKα) were increased by overexpression of human VRK1 (Fig. 6A) and reduced by small interfering RNA (siRNA)–mediated VRK1 knockdown (Fig. 6B). By using coimmunoprecipitation assays, we showed the physical interaction between endogenous VRK1 and AMPKα2 (Fig. 6C) as well as Flag-AMPK and VRK1 (Fig. 6D and fig. S8D). Overexpression of VRK1 also increased the level of phosphorylated AMPKα (p-AMPKα) in the immunoprecipitated fraction (Fig. 6D). These data suggest a weak and transient interaction between the kinase (VRK1) and its substrate (AMPK).
We then performed in vitro kinase assays with recombinant AMPKα2 and VRK1 to determine whether VRK1 can directly phosphorylate AMPK (Fig. 6, E and F). We found that VRK1 increased the level of AMPKα phosphorylation at Thr172, a known phosphorylation site for activation, in a dose-dependent manner (Fig. 6E). Because we detected substantial autophosphorylation of AMPKα2 (Fig. 6E, lane 2, and Fig. 6F, lane 2), we used a kinase-dead (KD) form of AMPKα2 (AMPKα2-KD) and found that VRK1 phosphorylated AMPKα2–KD (Fig. 6F, lane 5). In contrast, AMPKα2 did not phosphorylate a KD form of VRK1 (VRK1-KD) in vitro (Fig. 6G). These data indicate that VRK1 can directly phosphorylate AMPK but reverse does not occur. By using in vitro kinase assay, we showed that C. elegans VRK-1 phosphorylated AAK-2 isoform b at Thr243, an equivalent residue for Thr172 in the mammalian AMPK (Fig. 6H). We also found that the aak-2(T181A)::GFP, which alters the Thr181 to Ala in the isoform c, homologous to Thr172 to Ala change in the mammalian AMPK, did not restore the longevity of vrk-1::GFP O/E in an aak-2(−) mutant background (fig. S9, E and F). Thus, C. elegans VRK-1 appears to phosphorylate AAK-2 for life span extension. Last, we demonstrated that vrk-1 overexpression in C. elegans increased the level of active p-AMPK (Fig. 6I and fig. S8K). Conversely, genetic inhibition of vrk-1 by RNAi or by a reduction of function mutation, vrk-1(x1), decreased p-AMPK levels that were elevated by isp-1 mutations (Fig. 6I and fig. S8K). These data suggest that VRK-1/VRK1 functions as an upstream kinase of AMPK. Together, VRK1 functions to increase the activity of life span–extending AMPK via phosphorylation in an evolutionarily conserved manner.
Somatic nuclear expression of vrk-1 increases life span by activating AMPK through phosphorylation in C. elegans
The evolutionarily conserved protein kinase, VRK-1, has been known to function in cell division in C. elegans. However, its role in postmitotic somatic cells has remained elusive. Here, we showed that vrk-1 was highly expressed in various somatic tissues in adult C. elegans. Moreover, VRK-1 was essential and sufficient for the longevity of adult C. elegans. We further found that VRK-1 extended life span by increasing the level of active AMPK through phosphorylation. We demonstrated that both C. elegans VRK-1 and human VRK1 activated AMPK via direct phosphorylation. Thus, our data provide mechanisms by which VRK-1, a nuclear protein kinase known to regulate cell division, promotes longevity by functioning in postmitotic somatic cells.
Expression of vrk-1 in the soma of adult animals is critical for life span extension
Previous reports have revealed that vrk-1 is required for germ cell proliferation and proper cell division. In this study, we show that vrk-1 also functions in postmitotic somatic cells to promote organismal longevity. Several cell cycle–regulatory or DNA damage checkpoint proteins play additional roles in postmitotic somatic cells of C. elegans. For example, genetic inhibition of the cell cycle checkpoint genes, caffeine-induced death 1 (cid-1), checkpoint kinase 1 (chk-1), and cell division cycle 25.1 (cdc-25.1), increases endoplasmic reticulum protein homeostasis in postmitotic cells and increases life span, thus indicating postmitotic functions for these cell cycle regulators (27). Similarly, knockdown of cyclin E (cye-1) or cyclin-dependent kinase 2 (cdk-2) increases the life span of adult C. elegans and enhances stress resistance (28). Our data indicate that VRK-1 also plays key roles in the soma of adult animals in addition to regulating germ cell division in developing larvae.
VRK-1 is an AMPK-activating kinase acting in the nucleus
AMPK and its upstream regulators have been extensively studied as key factors that control cell metabolism. LKB1 and CaMKKβ, two established AMPK upstream kinases, directly phosphorylate AMPK via distinct signaling pathways. LKB1 is considered to be constitutively active and phosphorylates AMPK, whereas CaMKKβ phosphorylates AMPK when Ca2+ levels are increased [reviewed in (9)]. TAK1 may also function as an AMPK upstream kinase [reviewed in (9, 29)]. In our current work, we identified VRK1 as a novel upstream activating kinase of AMPK and further demonstrated the physiological importance of VRK-1–to–AMPK signaling for organismal longevity. Our data indicate that VRK-1 mediates the activation and the phosphorylation of AMPK upon inhibition of mitochondrial respiration. Therefore, we speculate that the increased AMP/ADP-to-ATP ratio resulting from impaired mitochondrial respiration enhances VRK-1–mediated phosphorylation of AMPK to promote longevity in C. elegans. In future studies, it will be important to further dissect the molecular mechanisms by which VRK-1 phosphorylates AMPK under mitochondria-defective conditions in both C. elegans and mammals.
We also propose that VRK-1 has functions that are distinct from other known AMPK kinases, mainly because of the fact that VRK-1 is a nuclear kinase, whereas LKB1 and CaMKKβ function in the cytoplasm. In C. elegans, AMPK is mostly localized in the nucleus (26). Mammalian AMPKα2 is also preferentially localized to the nucleus (30), and AMPKα1 displays circadian rhythm–dependent nuclear localization in mice (31). Therefore, our study proposes a possibility that the nuclear kinase, VRK-1, may phosphorylate AMPKα1 and/or AMPKα2 in the nucleus and thus contribute to the transcriptional changes elicited by factors acting downstream of AMPK.
Mammalian VRKs may play beneficial roles in animal physiology, including longevity
In mammals, VRK1 plays pivotal roles in various biological processes, and loss of VRK1 function leads to severe defects. Life span is often associated with telomere length in various organisms, including humans (32, 33). Mouse VRK1 phosphorylates and activates heterogeneous nuclear ribonucleoprotein A1, a single-stranded oligonucleotide-binding protein that plays an important role in telomere maintenance (34). Loss-of-function VRK1 mutant mice exhibit shortening of telomeres in male germ cells (34). In addition, the genetic inhibition of VRK1 results in defective cell proliferation, whereas overexpression of VRK1 stabilizes p53 and increases its transcriptional activity (35). These findings suggest that VRK1 plays an integral role in maintaining normal cellular function and physiology in animals.
The functions and phosphorylation targets of VRKs are highly conserved across species. Genetic and/or biochemical interaction between VRK-1 and its targets, BAF-1 and p53, has been shown in both mammalian cells and C. elegans, suggesting that these organisms share other common targets of VRK-1 (14–16, 35, 36). In particular, mutations in BAF result in aberrant nuclear envelope structure and are associated with progeria syndrome in humans (37). It is therefore tempting to speculate that the loss of VRK1 may lead to impairment in BAF function, which disrupts its interaction with LEM (LAP2, emerin, and MAN1) domain proteins and triggers progeria-like phenotypes. In our study, we demonstrated that VRK-1 activated AMPK by phosphorylation to extend life span in C. elegans. Considering the conserved role of reduced mitochondrial respiration in life span, it will not be surprising to find that VRK1 also contributes to longevity in mammals, likely through activation of AMPK.
MATERIALS AND METHODS
Worms were grown on Escherichia coli OP50-seeded plates following standard laboratory culture conditions. Some of the strains used in this study were obtained from Caenorhabditis Genetics Center, which is funded by the National Institutes of Health (NIH) Office of Research Infrastructure (P40 OD010440). The C. elegans strains used in this study are as follows: wild-type Bristol strain N2, IJ848 yhEx432[vrk-1p::vrk-1::gfp; odr-1p::rfp], IJ849 yhEx433[vrk-1p::vrk-1::gfp; odr-1p::rfp], IJ288 yhEx63[vrk-1p::vrk-1::gfp; rol-6D], IJ289 yhEx534[vrk-1p::vrk-1::gfp; rol-6D], CF1290 N2 carrying pRF4(rol-6(su1006)) marker, YL279 unc-119(ed3); vrIs18[pie-1p::gfp:vrk-1:pie-1 3′UTR; unc-119(+)], IJ391 vrk-1(x1) II, CF2172 isp-1(qm150) IV, CF2354 clk-1(qm30) III, IJ401 vrk-1(x1) II; clk-1(qm30) III, IJ402 vrk-1(x1) II; isp-1(qm150) IV, CF1041 daf-2(e1370) III, IJ173 eat-2(ad1116) II, CF2553 osm-5(p813) X, IJ7 vhl-1(ok161) X, IJ1376 vrk-1(x1) eat-2(ad1116) II, IJ1377 vrk-1(x1) II; daf-2(e1370) III, IJ1378 vrk-1(x1) II; osm-5(p813) X, IJ32 Iszc[hsp-6p::gfp], IJ31 isp-1(qm150) IV; Iszc[hsp-6p::gfp], ZG120 iaIs7[nhr-57p::GFP; unc-119(+)], IJ2 isp-1(qm150) IV; iaIs7[nhr-57p::GFP; unc-119(+)], CL2166 dvIs19[pAF15(gst-4::GFP::NLS)], IJ945 isp-1(qm150) IV; dvIs19[pAF15(gst-4::GFP::NLS)], IJ1670 yhIs90[vrk-1p::vrk-1::GFP; odr-1p::RFP], IJ1721 aak-2(ok524) IV; yhIs90[vrk-1p::vrk-1::GFP; odr-1p::RFP], CF2725 aak-2(ok524) X, IJ259 isp-1(qm150) IV; aak-2(ok524) X, AGD418 uthIs205[crtc-1p::crtc-1::RFP::unc-54 3′UTR; rol-6D], IJ1945 uthIs205[crtc-1p::crtc-1::RFP::unc-54 3′UTR; rol-6D]; yhIs90[vrk-1p::vrk-1cDNA::GFP; odr-1p::RFP], IJ249 yhEx53[aak-2p::aak-2(T181A)::GFP; rol-6D], IJ1942 aak-2(ok524) X; yhEx53[aak-2p::aak-2(T181A)::GFP; rol-6D], IJ1943 aak-2(ok524) X; yhIs90[vrk-1p::vrk-1cDNA::GFP; odr-1p::RFP]; yhEx53[aak-2p::aak-2(T181A)::GFP; rol-6D], IJ2017 vrk-1(ok1181)/mIn1[mIs14 dpy-10(e128)], IJ2018 vrk-1(ok1181)/mIn1[mIs14 dpy-10(e128)]; daf-2(e1370), and IJ2019 vrk-1(ok1181)/mIn1[mIs14 dpy-10(e128)]; isp-1(qm150).
Life-span assays were conducted at 20°C on nematode growth medium (NGM) plates seeded with E. coli OP50 or HT115 for RNAi experiments, starting at day 1 of adulthood. Worm strains that were used for experiments were kept at 20°C for at least three generations before life-span assays. Briefly, gravid adult worms were allowed to lay eggs on plates seeded with OP50 or HT115, and the worms that developed to prefertile young adults were transferred to new plates. For life-span assays without chemical 5-fluoro-2′-deoxyuridine (FUDR), which prevents progeny from hatching, young adults were placed on new plates every 1 or 2 days until they ceased to lay eggs. For life-span assays with FUDR, synchronized young adult worms were transferred to new plates containing FUDR (5 μM; Sigma-Aldrich, MO, USA). For RNAi experiments, 1 mM isopropylthiogalactoside (Gold Biotechnology, St. Louis, MO, USA) was added to plates seeded with HT115 that express specific double-stranded RNA (dsRNA) and incubated at room temperature for ~24 hours. For double RNAi experiment, HT115 RNAi bacteria were grown in liquid LB media containing ampicillin (50 μg/ml; USB, Santa Clara, CA, USA) until the media display the optical density measured at a wavelength 600 nm of 0.9. Two different HT115 targeting different genes were mixed at the ratio 1:1 for double RNAi. HT115 that express an empty RNAi plasmid (L4440) were mixed with specific RNAi bacteria for control. For all the life-span assays, a minimum of five plates with worms were used for each condition except for the ones that included discarded plates due to contamination. All the life-span assays were conducted by at least two independent researchers at least twice per condition. All the mutant animals that were used for life-span assays were outcrossed to Lee laboratory wild-type N2 at least three times before the experiments. Animals that displayed internal hatching or ruptured vulvae or crawled off the plates were censored but included in the life-span analysis as censored worms. Online application for the survival analysis 2 (http://sbi.postech.ac.kr/oasis2) was used for statistical analysis of the data, and P values were calculated using the log-rank (Mantel-Cox) test.
DNA staining with 4′,6-diamidino-2-phenylindole
Stage-synchronized worms were harvested with M9 buffer with polyethylene glycol 4000 (PEG 4000) (0.01%; Tokyo Chemical Industry, Japan) and washed three times with M9 buffer. Worms were fixed with 4% paraformaldehyde (158127, Sigma-Aldrich, MO, USA) solution in phosphate-buffered saline (PBS) (137 mM NaCl, 2.7 mM KCl, 8 mM Na2HPO4, and 2 mM KH2PO4; AM9624, Thermo Fisher Scientific, MA, USA) for 45 min with gentle agitation. After fixation, worms were washed with PBS containing 0.01% PEG 4000 twice with gentle agitation for 5 min for each washing step. Fixed worms were stored in 70% ethanol at 4°C overnight. Worms were placed on a 2% agarose pad on a slide glass, and the worms were soaked in a drop of 4′,6-diamidino-2-phenylindole (DAPI) solution (2 ng/μl) (62248, Thermo Fisher Scientific, MA, USA) in VECTASHIELD antifade mounting media (Vector Laboratories, CA, USA). After covering the agar pad with a coverslip, worms were incubated for at least 30 min in the dark before imaging.
Microscopy and quantification of fluorescence
Confocal fluorescence images of VRK-1::GFP and CRTC-1::RFP were acquired using an inverted LSM 880 laser scanning microscope (Zeiss Corporation, Germany) with Plan-Apochromat 20× 0.8 M27, C-Apochromat 40× 1.2 water Korr FCS M27, or Plan-Apochromat 63× 1.4 oil differential interference contrast (DIC) M27 objectives. Green fluorescence was detected under the excitation wavelength at 488 nm and emission wavelength at 526 nm. Red fluorescence was detected under the excitation wavelength at 561 nm and emission wavelength at 668 nm. DAPI signal was detected under the excitation wavelength at 405 nm and emission wavelength at 498 nm. Confocal Z stack fluorescence images shown in Fig. 1B were acquired using an inverted Leica (Wetzlar, Germany) TCS SP8 laser scanning confocal microscope with a 40× 1.3 numerical aperture oil or 63× 1.2 water objective. Green fluorescence was detected under the excitation wavelength at 488 nm and emission wavelength at 520 nm. Microphotographs used in Fig. 1B were obtained using Z-project (average intensity) function in ImageJ software (Rasband, W. S., ImageJ, U.S. NIH, Bethesda, MD, USA; http://rsbweb.nih.gov/ij/, 1997–2018). Fluorescence images of animals expressing hsp-6p::GFP, nhr-57p::GFP, or gst-4p::GFP were captured by using Axiocam (Zeiss Corporation, Germany) mounted to a HRc Zeiss Axioscope A.1 (Zeiss Corporation, Germany) equipped with EC Plan-Neofluar (Zeiss Corporation, Germany) objective lens. Green fluorescence was detected under Zeiss filter set 38 HE emission filter (Zeiss Corporation, Germany). Fluorescence images of CRTC-1::RFP used for quantification shown in Fig. 5E were captured by two researchers independently, using DS-Qi1Mc (Nikon, Japan) mounted on Nikon ECLIPSE Ni microscope (Nikon, Japan) with Plan Fluor 40× DIC M N2 objective (Nikon, Japan) and NEO scientific complementary metal-oxide semiconductor camera (Andor, Belfast, UK) mounted on Nikon Eclipse Ti2 inverted confocal microscope (Nikon, Japan) with a Niji light-emitting diode light source (Bluebox Optics, Huntingdon, UK) and a 10× objective (Nikon, Japan); the red fluorescence was detected under tetramethyl rhodamine isothiocyanate (TRITC) band-pass filter (Nikon, Japan) in the Eclipse Ni and ET-DsRed (TRITC/Cy3) filter (Chroma Technology, VT, USA) in the Eclipse Ti2 microscopes, respectively. Image planes that showed distinct nuclear morphology assessed by DAPI signals were chosen for imaging in Figs. 1A and 5D and figs. S1, S6J, and S7B. Image planes that showed distinct foci on nuclei with green fluorescence in the intestine for Fig. 1B were chosen. Image planes were chosen for the regions where outlines of most worms were visible for fig. S5 (C to E). ImageJ software was used to quantify GFP intensity with arbitrary unit by measuring average pixel intensity. Fluorescence intensity of individual worms was normalized by subtracting background signals, and mean fluorescence intensity was calculated by averaging the normalized values of >60 individuals. Levamisole (2 mM) was used to immobilize the animals on 2% agarose pads before imaging. P values were calculated using the unpaired Student’s t test (two-tailed) by comparing mean fluorescence intensity of experimental group with that of control group.
Synchronized L4 or young adult animals grown at 20°C were used for RNA extraction, cDNA synthesis, and qRT-PCR analysis. For qRT-PCR analysis of food-deprived animals shown in fig. S7A, day 3 adult worms were used. Briefly, RNAiso (Takara, Japan) was used to extract total RNA, and cDNA was obtained using ImProm-II Reverse Transcriptase (Promega, WI, USA). A 7300 Real-Time PCR System (Applied Biosystems) was used to perform quantitative PCR experiments, and the results were analyzed by using the comparative CT method described in the manufacturer’s manual. The average values of the mRNA levels of ama-1, tba-1, or pmp-3 gene were used for normalization. The average of at least two technical repeats was applied for each biological data point.
Food deprivation of C. elegans
For qRT-PCR analysis of food-deprived worms, wild-type embryos were obtained by using a bleaching method (7) and placed on control RNAi or vrk-1 RNAi bacteria-seeded NGM plates. Worms were then grown until reaching day 2 adults, harvested with M9 buffer, and washed three times to remove remaining bacteria. For a fed condition, worms were placed on control RNAi bacteria– or vrk-1 RNAi bacteria–seeded NGM plates. For a food deprivation condition, worms were placed on NGM plates containing streptomycin (10 μg/ml) to prevent E. coli (HT115) that expresses dsRNA from growth, harvested with M9 buffer, and used for RNA extraction and qRT-PCR analysis after 24 hours of food deprivation. FUDR was applied to prevent progeny from hatching at L4 stage. For assaying the subcellular localization of CRTC-1::RFP, worms were treated with control RNAi from hatching until they reached L2 or L3 stages. The animals were then collected with M9 buffer, washed three times, and placed on NGM plates containing streptomycin (50 μg/ml) and kanamycin (50 μg/ml) for food deprivation. After 24 hours of food deprivation, the worms were used for micrograph imaging.
In vitro kinase assay
Recombinant His-VRK1, GST-AMPKα2, GST–AMPKα2-KD (K45R), His-VRK-1, GST–AAK-2 isoform b, and GST–AAK-2 isoform b-KD (K116R) proteins were expressed in E. coli BL21 (DE3). GST-AMPKα2 and GST–AAK-2 were purified using glutathione Sepharose 4B (GE Healthcare Life Sciences, MA, USA). His-VRK1 and His-VRK-1 were purified with Ni–nitrilotriacetic acid agarose (Invitrogen, CA, USA) following the manufacturers’ instructions. For nonradioactive in vitro kinase assays using human VRK1 and AMPK, purified His-VRK1 was incubated with recombinant GST-AMPKα2 in the kinase assay buffer [60 mM HEPES (pH 7.4), 10 mM MgCl2, and 10 mM MnCl2] and 200 μM ATP for 30 min at 30°C. The enzyme reaction was stopped by the addition of 5× SDS loading buffer by boiling at 95°C for 5 min. Samples were immunoblotted with antibodies against GST (91G1, Cell Signaling Technology, MA, USA), His (2365S, Cell Signaling Technology, MA, USA), or p-AMPKα (Thr172) (2531S, Cell Signaling Technology, MA, USA). The nonradioactive in vitro kinase assays for C. elegans VRK-1 and AAK-2 was performed similarly to those for the human VRK1 and AMPK. For radioactive in vitro kinase assay, purified His-VRK1 was incubated with recombinant GST-AMPKα2 in kinase assay buffer [60 mM HEPES (pH 7.4), 10 mM MgCl2, and 10 mM MnCl2] and [γ-32P]ATP for 30 min at 30°C. The reaction was stopped by the addition of 5× SDS loading buffer by boiling at 95°C for 5 min. Radioactive signal was detected by using an x-ray film (Agfa, Antwerp, Belgium).
For immunoprecipitation of endogenous AMPK, human embryonic kidney (HEK) 293A cells were resuspended with cell lysis buffer [20 mM Tris, 150 mM NaCl, 1 mM EDTA, 0.5% Triton X-100, and protease inhibitor cocktail (F. Hoffmann–La Roche, Basel, Switzerland)] and disrupted by using sonication. One milligram of cell lysates was incubated with Protein A agarose bead (F. Hoffmann–La Roche, Basel, Switzerland) and normal rabbit immunoglobulin G (IgG; 2729S, Cell Signaling Technology, MA, USA) or anti-AMPK antibody (2532S, Cell Signaling Technology, MA, USA) overnight at 4°C on a rotator. After incubation, beads were washed four times with cell lysis buffer. Bound proteins were eluted with 2× SDS sample buffer and 10 mM dithiothreitol. Immunoprecipitated proteins were analyzed by immunoblotting. VeriBlot for IP Detection Reagent (AB131366, Abcam, Cambridge, UK) was used to detect the VRK1 band without interference from IgG heavy chain. Immunoprecipitation experiments using overexpression of Flag-AMPK and/or hemagglutinin (HA)–VRK1 was performed using 3 × 106 HEK293A cells in cell lysis buffer [50 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, and 1% Triton X-100] supplemented with protease inhibitors (A32953, Thermo Fisher Scientific, MA, USA). Cell lysates were incubated overnight with an anti-Flag antibody (F1804, Sigma-Aldrich, MO, USA) and protein G Sepharose beads (GE Healthcare Life Sciences, MA, USA) at 4°C on a rotator. After incubation, the beads were washed three times with the cell lysis buffer and boiled for 5 min at 95°C. Samples were immunoblotted with antibodies against Flag (2368S, Cell Signaling Technology, MA, USA), VRK1 (38), p-AMPKα (Thr172) (2531S, Cell Signaling Technology, MA, USA), HA (A190-108A, Bethyl Laboratories, TX, USA), or glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (sc-47724, Santa Cruz Biotechnology, TX, USA).
Western blot analysis
Western blot analysis using C. elegans samples was performed with synchronized L4 or young adult animals, which were harvested, washed three times using M9 buffer, and centrifuged at 1400 rpm for 30 s. The worms were shock-frozen in liquid nitrogen, mixed with SDS sample buffer, and boiled for 10 min. The samples were subsequently vortexed for 10 min and centrifuged for 30 min at 15,700g at 4°C, and the supernatant was collected. The worm protein samples were electrophoresed using 8% SDS–polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membrane. The membrane was blocked with TBS-T [20 mM Tris-HCl (pH 7.6), 140 mM NaCl, and 0.1% Tween 20] buffer containing 5% bovine serum albumin for 1 hour at room temperature and incubated with primary antibodies against p-AMPKα (1:2000; 2531S, Cell Signaling Technology, MA, USA) or α-tubulin (1:1000; sc-32293, Santa Cruz Biotechnology, TX, USA) overnight at 4°C. Anti-rabbit (1:15,000; Thermo Fisher Scientific, MA, USA) or anti-mouse (1:5000; Thermo Fisher Scientific, MA, USA) secondary antibodies conjugated with horseradish peroxidase were used to detect anti-AMPKα or anti–α-tubulin primary antibodies, respectively. The membrane was incubated in the chemiluminescent horseradish peroxidase substrate (Thermo Fisher Scientific, MA, USA), and the signal was detected by using x-ray film (Agfa, Antwerp, Belgium). Western blot analysis using cultured human cells was performed as previously described (38). Western blot data were quantified using ImageJ software.
Generation of transgenic animals
To generate transgenic animals expressing GFP-fused VRK-1, a promoter (~1 kb) and the genomic region of vrk-1 (~2.5 kb) were PCR-amplified from C. elegans genomic DNA. The PCR products were fused to Gateway donor vectors using BP Clonase (Invitrogen) followed by recombination with a Gateway destination vector containing GFP and unc-54 3′UTR using LR Clonase (Invitrogen). vrk-1p::vrk-1::gfp expression vector (25 ng/μl) was coinjected with an injection marker odr-1p::rfp (75 ng/μl) or pRF4(rol-6(su1006)) (75 ng/μl) into the gonad of day 1 adult wild-type animals. Transgenic animals expressing T181A mutant AAK-2 were generated by microinjection of PWM23 aak-2(T181A)::GFP expression vector (a gift from W. B. Mair) (25 ng/μl) and pRF4(rol-6(su1006)) (75 ng/μl).
mRNA library preparation and sequencing
Wild-type N2 and isp-1(qm150) mutant animals were synchronized by using an embryo bleaching method (7), and synchronized larvae were fed with either control or vrk-1 RNAi bacteria until the worms reached young adult stage. Worms were harvested with M9 buffer, washed to remove remaining bacteria, and frozen in liquid nitrogen. Three independent biological repeats for each condition were used for mRNA-seq experiments. Total RNAs were isolated using RNAiso plus (Takara, Shiga, Japan), followed by ethanol precipitation. Paired-end mRNA-seq library was prepared and sequenced by Illumina platform at Macrogen (Macrogen Inc., Seoul, South Korea).
Reads were aligned to the C. elegans genome ce11 (Ensembl WBcel235.87) by HISAT2 (version 2.0.5) with default parameters. Aligned reads were assembled to transcripts and levels of these transcripts were quantified by using SeqMonk (www.bioinformatics.babraham.ac.uk/projects/seqmonk/). FPKM (fragments per kilobase of transcript per million mapped reads) values were calculated by using RNA-seq. Quantitation pipeline and P values were calculated using EdgeR. Protein-coding genes whose FPKM values in wild type treated with control RNAi were greater than 1 were used for DEG analysis. Overlapping genes between our mRNA-seq data and published transcriptome data were analyzed by using WormExp (version 1.0). Venn diagrams were made by using Venn Diagram Plotter (http://omics.pnl.gov/software/venn-diagram-plotter), and representation factors for overlaps between two different gene sets were calculated (http://nemates.org/MA/progs/overlap_stats.html).
GO analysis for genes up- or down-regulated by isp-1 mutations in a vrk-1–dependent manner was performed by using DAVID. Enriched biological process GO terms with P < 0.05 and false discovery rate < 0.05 were listed.
Primers used for cloning
vrk-1 promoter Gateway. Forward: GGGGACAACTTTGTATAGAAAAGTTGAAATCGCGAATCAAGAAA
vrk-1 coding region Gateway. Forward: GGGGACAAGTTTGTACAAAAAAGCAGGCTTGCCACCGAAAAAAGCTCCCGCCAAAA
Scrambled siRNA: CCUACGCCACCAAUUUCGU(dTdT)
VRK1 siRNA: CAAGGAACCTGGTGTTGAA(dTdT)
ama-1. Forward: TGGAACTCTGGAGTCACACC
tba-1. Forward: GTACACTCCACTGATCTCTGCTGACAAG
pmp-3. Forward: GTTCCCGTGTTCATCACTCAT
vrk-1. Forward: GCCATGGATGGCTCTCGAATCGT
vrk-1-3′UTR. Forward: GATGATTATCAAAGAGACCAG
cco-1. Forward: TTGCTGGAGATGATCGTTAC
ceh-23. Forward: CAATTCCAGCACAGCCTTTTATTCC
col-176. Forward: CAGCTGGACAACCAGGAAC
col-90. Forward: ACACCAGGATCAAATGGAAG
rol-6. Forward: AATCGCCAACTTCAGAAGTC
R11G11.6. Forward: CAGACCAATAATCTTTCGGATC
col-17. Forward: GTTTGCTTCTTATTTTGGGTAC
bli-2. Forward: ATGGACGAGAAGGAACTGAA
dpy-10. Forward: TGTGTTGCTCTCCCAATTATG
col-73. Forward: AGAGCCAGGACAAGATGGAG
E01G4.6. Forward: GCTCGATTTCCCATCATGAG
col-14. Forward: CTTGGAGAGTGTCACGTTTC
F35B3.4. Forward: TCAATACTGTGCTGCTCAAG
Y47D7A.13. Forward: GACAAGGATACCAACAACAGC
wrt-6. Forward: GTACTAGATCCGACTTAAAATTG
Y11D7A.9. Forward: TTACGGAAATTGTCAGGCTG
tax-6. Forward: TGTGGGAGAAAAAGTAACGG
par-5. Forward: CGCCATGAAGAAAGTGACC
ftt-2. Forward: GACCTCGGACGCTGCC
Acknowledgments: We thank I. Mattaj, V. Reinke, W. B. Mair, and A. Dillin for providing C. elegans strains and constructs. We thank D. Lee for helping generate aak-2(T181A)::GFP transgenic strain. We also thank Lee laboratory members and K. Kim for helpful discussions and valuable comments on the manuscript. We apologize to the authors whose papers were not cited in this manuscript because of the number limit. Funding: This research was supported by grants NRF-2019R1A3B2067745 and NRF-2017R1A5A1015366 funded by the Korean Government (MSIT) through the National Research Foundation (NRF) of Korea to S.-J.V.L. and by grant Basic Science Research Program (No. 2019R1A2C2009440) funded by the Korean Government (MSIT) through the NRF of Korea to K.-T.K. Author contributions: Conceptualization: M.A., S.P., S.H.H., K.-T.K., and S.-J.V.L. Investigation: M.A., S.P., S.H.H., H.-E.H.P., Y.J., A.B.H., W.S.S., and S.-J.V.L. Writing: M.A., S.P., and S.-J.V.L. Funding acquisition: S.-J.V.L. Supervision: K.-T.K. and S.-J.V.L. Competing interests: The authors declare that they have no competing interests. Data and materials availability: RNA-seq data have been deposited in Gene Expression Omnibus under accession code GSE138129. All data needed to evaluate the conclusions in the paper are present in the paper and/or the Supplementary Materials. Additional data related to this paper may be requested from the authors.