Introduction

Most cellular processes are controlled by protein-protein interactions and networks of proteins. Therefore, the inter-relationships between proteins, rather than the individual protein, eventually determine the behavior of multiple coordinated processes in a biological system. Given the availability of large scale high-throughput proteomic and genomic interaction data, the computational methods provide us an opportunity to utilize this multiomics data for prediction of new functionality of the essential genes in these biomolecular networks. Aging happens in nearly all living organisms including humans by genetic programs and biochemical accidents such as oxidation. It implies a reduction in all provided functions over time and an increased risk for disease such as cancer. There is a clear association between aging and cancer by programmed pathways as well as by biochemical accidents. Although the intricate association between these two biological processes has been described in several studies1,2,3, many genetic details and intricacies are still unknown. There are a number of reported signaling molecules, oncoproteins and tumor suppressors including IGF1, FOXO, mTOR, and others that are well-known common targets of cancer and ageing. Interestingly, few drugs such as metformin have shown promising anti-ageing and tumor-inhibition potentials4,5.

Caenorhabditis elegans (C. elegans) was used as a model system to investigate unexpected new links between these two entities is a powerful approach for this question as many genes that affect tumors and ageing in mammals have orthologs representatives in C. elegans. The worm bearing 60–80% human gene counterparts, has already contributed enormously to our understanding of both ageing and cancer6,7,8. In fact, the tumor protective nature of longevity mutations or phytomolecules that mediate longevity has been revealed in this model system6,9. Therefore, finding common targets that underlie longevity and tumor suppression is an important starting point for developing novel drug molecules with therapeutic potential for both ageing and cancer. RNAi inhibition can point to the function of such targets. However, diverse cellular processes may be carried out by molecular targets disrupted by the RNAi. The complex action and effect of the knockout happens by complex networks of protein interactions. Therefore, the inter-relationships between the implied proteins, rather than the individual mRNA knockout itself, determine the behavior of the various integrated processes in a biological system.

Taking these considerations into account, we studied the protein-protein interactions (PPIs) of C. elegans using several computational approaches to predict genes that serve dual functions, behaving as tumor suppressor genes (TSG) and also influencing the lifespan in C. elegans. For validation, we combined the in silico predictions with RNAi tests in C. elegans. Our results clearly indicate the influence of top predicted targets such as, nhr-6, vab-3 and gst-23, on the lifespan of C. elegans as knockdown of these genes significantly modified the C. elegans longevity. Moreover, orthologs of these genes are already reported as tumor suppressor genes in humans however we suggest their additional involvement in complex network processes implied in aging thereby sharpening the pleiotropic effects of these genes.

Material and Methods

Network reconstruction

The experimentally determined physical and genetic PPIs of C. elegans present in the BioGRID database version 3.2.110. (https://thebiogrid.org/) were used to parse PPI of C. elegans genes associated with aging and tumor suppression. The BioGRID ver. 3.2.110 consist 8459 PPIs of C. elegans. To reconstruct the aging network, we used the C. elegans longevity genes reported in GenAge database (http://genomics.senescence.info/genes/)as query to mine their curated PPIs from BioGRID. Further, to create a TSG network in C. elegans, first we used OrthoMCL version 2.0.9 (http://orthomcl.org/orthomcl/) to identify C. elegans orthologs of human TSGs present in TSGene database (https://bioinfo.uth.edu/TSGene/), with stringent BlastP cutoff of 1e-5 and MCL inflation parameter of 1.5. OrthoMCL uses the reciprocal best hits strategy and then exploits similarity measures to identify clusters of orthologs and paralogs, using a Markov clustering (MCL) algorithm10. The prepared list of TSG in C. elegans was manually curated, mapped to gene symbols and the genes specifically reported as TSG in C. elegans were added to the list. To avoid the possibility of missing interactions during offline parsing of BioGRIDdata we used all the synonyms of genes in the TSG query list. Both the aging and TSG networks were reconstructed and merged with Cytoscape software version 2.8.111.

Network analysis

The graphical properties of the aging, TSG and merged aging-TSG PPI networks were analyzed using Network Analyzer plugin (ver. 2.7) of Cytoscape. The graph theoretical analysis was performed on each network to determine the network diameter, the mean path length, degree and centrality of nodes. The topologically important nodes for the integrity of networks were defined based on degree distribution and betweenness values. The aging and TSG network were compared to identify the C. elegans TSG present in aging network, which were probable candidates for affecting lifespan on C. elegans. The TSG that were not previously reported for involvement in aging were ranked based on the degree and betweeness rank in the merged aging-TSG network. The genes were further reranked by Rankmin score.

$$Ran{k}_{\min }(i)=Ran{k}_{de}(i)+Ran{k}_{be}(i)$$
(1)

where ‘i’ indicates the ith protein in the reconstructed PPI, Rankde is the ranking of protein ‘i’ based on degree score and Rankbe is the ranking of proteins ‘i’ based on betweenness score estimated by cyto-Hubba plugin of Cytoscape12.

Nematode strains and RNAi knockdown

The C. elegans strains used in this work include wild-type Bristol N2, GR1307: daf-16 (mgDf50), DA1116: eat-2 (ad1116), EU1: skn-1 (zu67), EU31: skn-1(zu135), VC199: sir-2.1(ok434) IV, PS3551: hsf-1 (sy441), PS746: let-23(sy97) II.

RNAi analysis of vab-3, nhr-6 and gst-23 mutants was performed as described previously13. Briefly, RNAi clones were grown overnight in LB with 50 µg/ml ampicillin and then seeded into the NGM agar plates containing 1 mM isopropylthiogalactoside (IPTG) and 25 ug/ml carbenicillin. Plates were kept for 7 days before feeding and then L4-stage larvae were transferred to the NGM plates seeded with RNAi induced bacteria. The worms were allowed to lay eggs at 20 °C and the progeny was transferred onto another plate seeded with the same bacteria. The process was repeated till the 6th generation L4-stage larvae.

Lifespan analysis

Age-synchronized RNAi treated L4 molts were transferred to RNAi agar plates containing 50 mM 5-fluorodeoxyuridine (FudR). Worms were kept at 20 °C and transferred to fresh agar plates, containing their respective RNAi bacteria, every 2–3 days to avoid contamination and to ensure the continued efficacy of RNAi knockdown. Worms feeding on bacteria carrying the empty vector (L4440) was used as control. The survival of worms was scored on the basis of body movement using touch-provoked method.

Oxidative stress assay

For oxidative stress assay, worms were maintained for 6 days until progeny production ceased. Subsequently, 90 nematodes were transferred to fresh NGM plates containing 10 mM paraquat and checked daily for viability14.

Quantification of ROS formation

The ROS assay was performed using around 300 young (day 1) adult worms as described15. The Cary Eclipse fluorescence spectrophotometer (Agilent Technologies) was used to capture the fluorescence spectral measurements at the excitation and emission wavelength of 485 nm and 535 nm, respectively.

Real-time qPCR analysis

The worms grown at 20oC were washed thrice and RNA was isolated using RNAzol RT (Molecular research Centre, Cincinnati, Ohio). The cDNA was synthesized using cDNA synthesis kit (Invitrogen, USA) according to manual instructions. Primers for specific amplification of genes were designed using the NCBI primer designer. The mRNA expression of target genes was quantified in comparison to housekeeping gene β-actin (act-1) using SYBR green (Takara Bio SYBR Premix Ex Taq, DSS Takara Bio India Pvt. Ltd.) detection method on a fast real-time PCR system (Applied Biosystems 7900 HT). qPCR data was analysed using the ΔΔCt relative quantitation method16.

Statistics

Statistical analyses of all data relevant to lifespan and stress resistance assays was carried out by log rank test using the Kaplan–Meier survival assay in MedCalc software. ASSISTAT 7.7 beta statistical assistance software was used to perform ANOVA (Analysis of Variance). The data was considered statistically significant at p value less than 0.05. The number of asterisks represents the following: * denotes p < 0.05 and ** denotes p < 0.01.

Results and Discussion

Protein-protein interactions are essential for life. Studying such interactions primarily relies on sophisticated experimental methods such as RNAi. RNAi has been used widely to decipher unreported functions. Advances in the in silico network biology have made it possible to use protein-protein interaction networks for better understanding of cellular functions in both physiologic and pathologic conditions. These in silico omics methods can help us in making functional predictions based on the principle that a protein with unreported function may interact with previously characterized proteins and accordingly we can assign its function. These functional predictions complement approaches such as RNAi that probe gene function directly by interfering on the transcript level.

In silico PPI analysis reveals involvement of vab-3, gst-23 and nhr-6 in C. elegans longevity

With the aforesaid considerations in mind, we utilized experimentally validated, physically interacting protein pairs obtained from the BioGRID database to analyze the mediators between TSG and aging networks in C. elegans. The BioGRID database (https://thebiogrid.org/) has collaborated with WormBase (http://www.wormbase.org/) for data sharing and well-curated protein-protein interaction (PPI) information. We queried 741 aging related (lifespan increasing and/or decreasing) genes of C. elegans from the GenAge database (http://genomics.senescence.info/genes/) and retrieved 2279 interactions associated with 1399 unique interactor proteins to reconstruct the aging relevant network. Furthermore, to determine the PPIs associated with C. elegans TSG (tumor suppressor gene) homologs, we first identified 464 worm orthologs of 221 human TSG in 196 ortholog clusters, out of 716 human TSG present in the TSGene database (https://bioinfo.uth.edu/TSGene/) by the robust orthology detection pipeline OrthoMCL (http://orthomcl.org/orthomcl/). The list was updated manually, enriched with literature based C. elegans TSG, and queried with the BioGRID database, which resulted in 1388 interactions mediated by 1159 interactors in the reconstructed TSG network. The merged aging-TSG network hosts a total of 2017 unique C. elegans genes mediating 3249 PPIs (Fig. 1, Supplementary File 1).

Figure 1
figure 1

Protein-protein interaction network. (a) Venn diagram showing overlap between aging network and tumor supressor gene (TSG) network. Combined network is splitted into clusters. (b) Number of genes into each cluster. (c) Protein-protein interaction network of aging and human tumor supressor gene orthologs of C. elegans. Only the primary connectors of nhr-6 are highlighted. The annotation of the nodes can be accessed from the XGMML (extensible graph markup and modelling language) network Supplementary File 1. Network is organized into eight clusters. Cluster1 - proteins associated with aging. Cluster2- proteins exclusively interacts with aging proteins but not with the TSG orthologs. Cluster3- connector proteins annotated to be involved in aging. Cluster4- proteins included exclusively in intersection of aging and TSG network but connectors are annotated as both aging and TSG. Cluster5- proteins included exclusively in intersection of aging and TSG network but conntectors are not annotated as aging or TSG. Cluster6- connector proteins annotated as human TSG orthologs. Cluster 7- proteins exclusively interacts with tumor supressor proteins but not with the aging associated proteins. Cluster8- proteins associated with tumor supression.

Full size image

A number of cancer genes affect aging, however only few reports that TSG may also affect aging are available17. The topological properties of a PPI network give a comprehensive view of the network. This helps to identify the central components important for the network connectivity. Some of our identified TSG orthologs are already known to be involved in aging and had many network connections with aging genes in the aging network. Therefore, we hypothesized TSG may also impact the lifespan of C. elegans. Firstly, we manually organized the network into eight functional clusters (all proteins for each cluster are given in Supplementary Table 1). Second, within the aging gene network, we annotated 74 TSGs (Fig. 1c, Cluster 4 and Cluster 6) of which 25 already have a reported role in aging as curated from the literature. Next, we ranked the rest of 49 genes according to their overall importance in the merged aging-TSG network using important network parameters such as their node degree distribution and their “betweenness centrality”. Network analysis and re-ranking based on Rankmin score (see online Methods and materials) revealed the genes vab-3, gst-23, and nhr-6 as the top three TSG candidates that should impact C. elegans aging and be tested by RNAi.

A priori probabilities to hit a longevity gene

Our approach was systematically built up to identify by a guilt-by-association approach combined with semantic similarity filters the high scoring genes with a high functional score to really be connected to longevity or functionally similar to lifespan reduction genes (LRGs). The Supplementary File 2 gives a detailed analysis, all network raw data are given in Supplementary File 3. In the following the key results of this analysis are given so that the reader can better appreciate the power of the in silico approach used to identify longevity genes:

In particular, the probability of any randomly picked C. elegans gene to reduce lifespan upon inhibition is low (0,35%) and hence picking three random genes each having a reduction in lifespan has an a priori very low probability of (0.0035)3 = 0.000000042875 so 4.2 × 10−8. The probability of any randomly picked C. elegans gene with high network importance to reduce lifespan upon inhibition is also low, the fraction of LRGs among hub nodes is about 5% (see Supplementary File 2) and to just randomly point at hub nodes and to get in a row three LRGs is again quite low, in this case 0.00014, so 14 in 100.000. Moreover, the guilt-by-association approach shows that we actively enriched specific functions in each cluster and the clusters have highly significant different gene functions enriched (see the cross-correlation table in Supplementary File 2). Furthermore, to subclassify genes further we used a semantic similarity score according to gene ontology. Hence, the genes picked come from clusters enriched in longevity genes were further differentiated by being those genes with the highest scores to be good targets, beyond 0.9 (Supplementary File 2).

RNAi confirm pro-longevity effects of vab-3, gst-23 and nhr-6 in C. elegans

We reasoned that if the three predicted genes (vab-3, gst-23, and nhr-6) are involved in aging, the RNAi inhibition of these genes should affect the lifespan of C. elegans. The transcription factor VAB-3, an ortholog of vertebrate Pax-6, is a member of the paired domain containing pax-6 gene family. Pax-6 is involved in various developmental processes, functions as a tumor suppressor in glioblastoma and also limits prostate cancer development18,19. Likewise, GSTP1 (a member of the glutathione S-transferase family), an ortholog of gst-23, is a promising cancer biomarker and its hyper-methylation was seen in multiple types of cancers such as breast, prostate, and hepatocellular carcinoma20. The loss of function of GSTP1 was found to increase the susceptibility to chemically-induced skin and lung cancers21. Similarly, Nr4a1 (Nur77), an ortholog of nhr-6, has been reported to regulate tumorigenesis by maintaining the homeostasis of proliferation, apoptosis, and differentiation22. The involvement of nhr-6 in cell proliferation and cell differentiation during the development of the spermatheca and spermatheca–uterine valve of C. elegans has also been observed23.

The silencing of gene expression by RNAi has evolved as a powerful approach to delineate the gene function. Although, not all organisms are RNAi-capable but C. elegans has been widely used as an exemplary model organism for RNAi experiments due to its rapid take up and spread of triggering dsRNAs24. The short generation time and ease of culture makes it an ideal model organism for an RNAi experiment, particularly for lifespan experiments. Further, the availability of large libraries of engineered ‘RNAi foods’ has significantly increased the robustness of RNAi in C. elegans. Since, the ‘feeding RNAi technique’ has the potential to provide large number of desired RNAi worms required for lifespan analysis; we performed an RNAi screen of the three predicted genes (nhr-6, gst-23, and vab-3) in C. elegans by feeding the respective RNAi clones using the well-established RNAi protocol13. As shown in Fig. 2a and Table 1, the lifespan of C. elegans was greatly reduced by 23.46, 15.27, and 13.29%, after feeding the RNAi of nhr-6, gst-23, and vab-3, respectively. These findings validate the in silico predictions and show that all the three identified genes act as potential pro-longevity genes and are involved in the C. elegans lifespan maintenance. Since nhr-6 showed great impact on the C. elegans lifespan, we were interested to further analyze the role of nhr-6 in C. elegans longevity.

Figure 2
figure 2

Effect of nhr-6 RNAi on lifespan and stress level in wild type C. elegans. (a) Lifespan curves of wild type N2 worms fed with nhr-6 RNAi. Complete lifespan data are presented in Table 1 including numbers of worms and S.E. measured. The lifespan data were statistically analyzed by log rank test using the Kaplan–Meirsurvival analysis in Medcalc 12.7.7.0 software and Graphpad prism 5. Differences between the data were considered significant at P ≤ 0.05. (b) Effect of nhr-6 RNAi on paraquat induced oxidative stress level in wild type worms (n = 90). Bars represent means ± SE. **P < 0.01. (c) ROS levels of treated (nhr-6 RNAi) and untreated worms. *P < 0.05.

Full size image
Table 1 Lifespan data: effects of nhr-6, vab-3 and gst-23 RNAi on C. elegans.
Full size table

nhr-6 silencing promotes stress tolerance in C. elegans

nhr-6 belongs to the family of nuclear receptors (NRs) that are the most abundant ligand gated transcriptional regulators in metazoans, conserved across species and are known to regulate metabolism, reproduction, homeostasis, and key developmental signalling pathways. Three complete metazoan genome sequences compared revealed different numbers of predicted NR genes which include 270 for the nematode (C. elegans), 21 for the fruit fly (Drosophila melanogaster), and ~50 for the human25. The ability of NRs to regulate transcription is associated with and based on their susceptibility to the influence of specific small binding ligands. Some of the known NR binding ligands include thyroid, steroid, xenobiotic and metabolic intermediates23. Several studies have taken the advantage of the free living nematode model C. elegans and revealed role of few NRs in the lifespan regulation, metabolism, stem cell proliferation, development, nutrient sensing, and longevity26,27.

In PPINs, the topological characteristics of PPIs reflect the functionality of the interacting genes and the important genes are more likely to be well connected and globally centered in the PPIs. To get functional insights of nhr-6, we analyzed the network up to secondary connections by considering nhr-6 as seed and then performed the Gene Ontology (GO) annotations. Statistically significantly over-represented GO slim terms were obtained by Fisher’s exact test followed by correcting the multiple testing errors using an FDR (false discovery rate) cutoff 0.5. The over-represented terms include MAPK cascade, signal transduction, cell communication, DNA repair, metabolic and developmental process, and response to stress. This is in line with the fact that aging is a multifactorial process which affects several biological functions. The detailed GO results of nhr-6 primary network and nhr-6 secondary network is listed in the Supplementary Table 2. Notably; none of the toxicity related terms were over-represented which discards the possibility of induced toxicity by gene RNAi inference.

Besides the in silico predictions, multiple studies have reported the major role of stress in lifespan regulation28,29. Therefore, we first examined whether nhr-6 contributes to stress resistance in worms. The oxidative stress was induced in worms using reactive oxidants generator paraquat and the effect of nhr-6 on the C. elegans oxidative stress resistance was analyzed by calculating the viability of worms. Next, we measured the total ROS production in C. elegans with the help of a dye (the dye H2DCF-DA). Our measurements showed that RNAi of nhr-6 increases resistance to paraquat induced oxidative stress by 27.8% (Fig. 2b). As shown in Fig. 2c, the ROS production of worms was also decreased by 33.4%. These results suggest that nhr-6 regulates oxidative stress resistance in worms. The findings also demonstrate that stress resistance is not intimately linked to increased lifespan and their overlap is inexact. An inverse correlation between lifespan and ROS production across a variety of species was also reported by Lambert and Brand while conducting study on mitochondria and aging30. In fact, researchers were unable to find any lifespan extension after supplementation of some anti-oxidants such as N-acetylcysteine and vitamin C31,32. These findings make it clear that increased resistance to oxidative stress is not sufficient for longevity assurance.

nhr-6 modulates SKN-1 and HSF-1 activity in C. elegans

The skn-1/nrf-2 pathway and the heat shock response (hsf-1) are cellular defense systems which are directly influenced by the transcription factors SKN-1 and HSF-133,34. In a situation of cellular stress, primarily the skn-1/nrf-2 comes to rescue, followed by hsf-1, which neutralizes proteotoxicity thus saving the cellular proteome. HSF-1, which encodes a heat shock factor ortholog, regulates the expression of stress induced genes required for the maintenance of protein homeostasis and development in organisms34.

The gene silencing by feeding RNAi has been widely used by the researchers to silence a gene of interest in a mutant strain (Minois et al.35; Sutphin et al.36). In this study, nhr-6 RNAi feeding was given to skn-1 (zu67) and hsf-1 (sy441) mutant animals to trigger the RNAi knock-down of the candidate nhr-6 gene in these mutants. The mean lifespan of the nhr-6 RNAi fed animals (skn-1 and hsf-1) was then compared with the control worms to elucidate the effect of nhr-6 knock down on skn-1 and hsf-1. The nhr-6 failed to augment the mean lifespan in skn-1 (zu67) and hsf-1 (sy441) mutants which suggests interaction of nhr-6 with the stress response pathways regulated by skn-1 and hsf-1 (Fig. 3a,b; Table 2).

Figure 3
figure 3

Lifespan analysis on RNAi knockdown of nhr-6 in different mutant strains of C. elegans relevant to lifespan and stress resistance (detailed data including numbers of worms and S.E. are given in Table 2). The lifespan data were statistically analyzed by log rank test using the Kaplan–Meir survival analysis in Medcalc 12.7.7.0 software. Differences between the data were considered significant at P ≤ 0.05.

Full size image
Table 2 Lifespan analysis of C. elegans mutant strains fed with nhr-6 RNAi.
Full size table

qRT-PCR is one of the most common tools used to measure gene expression changes in living organisms. We used this technique to determine the gene expression of skn-1 and hsf-1 in nhr-6 RNAi treated worms. It was observed that nhr-6 knock down up-regulated 90-fold and 11-fold the expression of skn-1 and hsf-1 in N2 wild-type worms in comparison to empty vector fed worms (Fig. 4), thus validating the role of nhr-6 in regulating expressions of skn-1 and hsf-1 that are key regulators of stress response in C. elegans. We hypothesize that in the absentia of nhr-6, skn-1 and hsf-1 comes into action to maintain normal homeostasis and development in worms. We assume that the increased expression of these genes might be one of the probable reasons for the increased oxidative stress resistance seen in nhr-6 RNAi treated worms.

Figure 4
figure 4

Real time quantification of mRNA expression of gerontogenes on RNAi knockdown of nhr-6. The nhr-6 knockdown in N2 wild-type C. elegans up regulated the expression of skn-1 and hsf-1 in comparison to control worms fed on empty vector. However, the expression of sir-2.1 was seen unaffected. The β-actin gene served as endogenous control.

Full size image

nhr-6 acts in a pathway with let-23 to regulate lifespan in C. elegans

Since MAPK cascade was recognized as one of the over-represented terms in GO annotations, we evaluated the possible interaction of nhr-6 with MAPK by analyzing the impact of nhr-6 RNAi on the lifespan of let-23 gene mutant let-23(sy97) EV. let-23 receptor/mpk-1 MAP kinase signalling pathway is responsible for vulva development in C. elegans9. The nhr-6 knock down failed to prolong the lifespan of let-23 gene mutant (Fig. 3c, Table 2), validating the in silico predictions and suggesting that nhr-6 may influence lifespan, at least in part, by interacting with let-23.

nhr-6 extends C. elegans lifespan independently of dietary restriction

Dietary restriction (DR) is one of the major phenomenons that regulate aging in a variety of species including C. elegans37. C. elegans eat-2 encodes a ligand gate ion channel subunit similar to nicotinic acetylcholine receptors (nAChR), which functions in the pharyngeal muscle and regulate the pharyngeal pumping rate38. The eat-2 gene plays a key role in the regulation of normal lifespan, feeding behavior, and defecation39. eat-2 mutants have defective pharyngeal pumping that limits their food intake and therefore these animals are considered a genetic model of DR. We asked if eat-2 is necessary for nhr-6 mediated longevity in C. elegans and examined the lifespan of eat-2 (ad1116) mutants supplemented with nhr-6 RNAi. nhr-6 knock down showed an increment of 9.54% in the mean lifespan of eat-2 mutants in comparison to worms fed on an empty vector (control), suggesting that eat-2 does not participate in the nhr-6 mediated longevity in C. elegans (Fig. 3d, Table 2). In addition, several studies have reported sir-2.1 (C. elegans homolog of Sir2) to function similar to eat-2 and its overexpression has been shown to account for C. elegans dietary restriction-induced longevity40,41,42. We therefore examined the effect of nhr-6 RNAi on the lifespan of sir-2.1(ok434) mutant and found that nhr-6 knock down also augmented the mean lifespan in sir-2.1 mutant worms (Fig. 3e, Table 2). As the increment in the sir-2.1 mutant lifespan was marginal, we tested the sir-2.1 gene expression in wild-type worms treated with nhr-6 RNAi. nhr-6 RNAi knock down worms failed to demonstrate any significant up-regulation or down-regulation of sir-2.1 expression, indicating that sir-2.1 is not necessary for nhr-6 mediated longevity in C. elegans. The cap ‘n’ collar transcription factor SKN-1 is known to encode a bZip transcription factor orthologous to the mammalian Nrf (Nuclear factor-erythroid-related factor) transcription factors, which are known to regulate development in organisms (Tullet et al., 2008). SKN-1 located in the ASI neurons sense food availability and has been shown necessary for lifespan extension in response to a variety of DR methods33. To test this hypothesis, we used a loss of function mutant strain of skn-1, skn-1(zu135). Interestingly, we find that nhr-6 RNAi significantly extended the lifespan of this mutant strain (Fig. 3f, Table 2), indicating that skn-1 is not required for nhr-6 mediated longevity in C. elegans. These observations signify that nhr-6 did not affect the feeding behavior of worms to mediate lifespan.

Altogether, our results confirm the role of some tumor suppressor genes such as nhr-6, vab-3 and gst-23 in aging and demonstrate the potential of omics methods, particularly PPI network analysis, to uncover the new functionality of human candidate genes. The conserved molecular architecture of these genes raises the prospective to understand the molecular mechanism that governs the expression of these genes and to regulate them by some exogenous drug or plant-based molecules for human health benefits as these genes are already reported as tumor suppressors in humans.