Introduction

An organism’s lifespan is stated as the sum of damaging changes that occur in its body and the repair and maintenance mechanisms that respond to damage (Johnson et al. 1999). Both environmental and genetic components determine the lifespan of an organism (Paaby and Schmidt 2009). After discoveries were made in the worm Caenorhabditis elegans (Kenyon et al. 1993), it was found that many genetic pathways that govern lifespan are highly conserved from nematode to human (Garofalo 2002; Holzenberger et al. 2003; Nakae et al. 2002). Loss or gain of function mutations were employed for the identification of gerontogenes (that affect aging and lifespan) and hundreds of these genes were identified (Guarente and Kenyon 2000). Insulin/IGF-1(IIS), PI3K, TOR, MAPK, AMPK, PKC, NF-κB, TGF-β, Notch and WNT are some of the signalling pathways that regulate aging (Moskalev et al. 2014). They maintain energy levels, cellular pliability, homeostasis, growth and reproduction of an organism under favourable conditions (Barzilai et al. 2012). Under stressful conditions, these pathways repress growth and activate stress-resistance proteins (Koubova and Guarente 2003; Tatar et al. 2003). Insulin/insulin-like growth factor-1 (IGF-1) signalling (IIS) pathway was the first pathway to be established for aging and was identified in C. elegans through mutations of the genes age-1 encoding phosphoinositide 3-kinase (PI3K) and daf-2 encoding IGF-1 receptor (Kenyon et al. 1993). This pathway is greatly conserved among C. elegans, Saccharomyces cerevisiae, Drosophila melanogaster, rodents and humans and regulates nutrient homeostasis, growth, aging and longevity (Barbieri et al. 2003).

The common denominators of aging in different organisms include genomic instability (Moskalev et al. 2013), telomere attrition (Blackburn et al. 2006), mitochondrial dysfunction (Harman 1965), stem cell depletion (Shaw et al. 2010), loss of proteostasis (Powers et al. 2009), cellular senescence (Campisi and di Fagagna 2007), altered intercellular communications (Russell and Kahn 2007), deregulated IIS pathway (Barzilai et al. 2012) and epigenetic alterations (Talens et al. 2012). Epigenetic modifications play numerous key roles in regulatory events that alter gene expression and hence impact other hallmarks of aging (Jung and Pfeifer 2015). The inherited epigenetic status undergoes numerous changes by environmental stimuli or stochastic errors that either benefit the organism or accelerate aging (Bennett-Baker et al. 2003; Cedar and Bergman 2012; Salpea et al. 2012).

In this report, we discuss the role of IIS pathway in aging of worm, fly, rodent and human, and its epigenetic regulation through different epigenetic machineries including non-coding RNA. We focus on non-coding RNA and the chromatin for reasons that regulatory RNA and epigenomic modulations are involved in transcriptional changes related to aging.

Insulin/insulin-like growth factor-1 signalling circuit and longevity

IIS pathway acts through the PI3-kinase (PI3K)/AKT kinase signalling cascade (explained in detail by (De Meyts 2016)) and is initiated by insulin-like peptides. These peptides are secreted in response to food or the sensory perception of food (Fig. 1) and bind to insulin/IGF tyrosine kinase receptor (Wolkow et al. 2002). Activated receptor transduces the signal to the phosphatidylinositol 3-kinase (PI-3K) which in turn converts phosphatidylinositol 4,5-biphosphate (PIP2) into phosphatidylinositol 3,4,5-biphosphate (PIP3) (Morris et al. 1996). Increased levels of PIP3 activate the protein kinases, leading to phosphorylation and retention of the forkhead box O (FOXO) transcription factors in the cytoplasm (Ogg et al. 1997). Under attenuated insulin signalling conditions, unphosphorylated FOXO gets transported to the nucleus to aid in the transcription of genes involved in organism longevity, of which shall be discussed in the next section.

Fig. 1
figure 1

Pathways that regulate aging in C. elegans. Activation of insulin/IGF signalling (IIS) pathway (see text for names of genetic homologs found in D. melanogaster, rodents and human) prevents translocation of DAF-16/FOXO transcription factor to the nucleus. Blocking of IIS pathway causes FOXO activation or repression of target genes that promote longevity. FOXO is also regulated by AMPK, JNK and SIRT1. Dietary restriction (DR)obstructs IIS pathway and TOR pathway whereas it activates SIRT1. TOR pathway results in an increase in autophagy and suppression of translation. Stress stimulates JNK pathway, which involves the activation of FOXO. Chromatin modifier zinc finger protein 1 (ZFP-1) acts as a common mediator of IIS and DR mediated lifespan regulation. ZFP-1 interacts with glioma amplified sequence-41 (GFL-1 in C. elegans). DAF-16/FOXO directly binds and regulates the expression of the different isoforms of zfp-1 as well as gfl-1. ZFP-1/GFL-1 in turn negatively regulates the expression of DAF-16/FOXO target genes such as sod-3, gpd-2, mtl-1, dod-11, by forming feed-forward loops that control the amplitude and duration of target gene expression.

Full size image

A few other pathways also interact with IIS pathway to regulate lifespan. They are the target of rapamycin (TOR) signalling pathway which responds to nutrient levels (Hafen 2004; Partridge et al. 2011), the Jun-NH2-terminal kinase (JNK) pathway (Karpac and Jasper 2009; Wang et al. 2005) which responds to stress levels and the transforming growth factor-β (TGF-β) signalling pathway (Narasimhan et al. 2011). In addition, the IIS pathway also controls lifespan through secondary hormones such as the Juvenile hormone (JH) and Ecdysone hormone (Richard et al. 2005). Central nervous system (CNS) plays a significant part in the longevity of varied organisms. Though lowered insulin signalling in the CNS can extend lifespan, IIS has been documented as being neuroprotective and vital for the development and survival of neurons. Hence, manipulation of the right signalling component in the right tissue is required to extend organism lifespan with less damaging effects or with even better health and function (Broughton and Partridge 2009).

Forkhead box O (FOXO)

FOXO proteins play the most pivotal role in IIS/PI3K/Akt signalling pathway (Brunet et al. 1999; Dong et al. 2008). They act as “master-switch” for cells to adapt and show metabolic stability under conditions of food shortage and stress (Horst and Burgering 2007). They belong to the Forkhead family of transcription factors, specified by a conserved DNA binding domain called the ‘Forkhead box’ (FOX). Forkhead gene was initially discovered in fly whose alteration lead to abnormal head-like structures that resembled a fork (Carter and Brunet 2007). Currently, they are much researched for their many roles involved in organism survival and lifespan and have been reviewed in detail by (Eijkelenboom and Burgering 2013; Kim et al. 2015; Martins et al. 2015; Proshkina et al. 2015). FOXO proteins respond to both nutritional and stress signals. Dietary restriction (DR) or stress conditions cause nuclear translocation of FOXO, to promote the expression of target genes that induce stress resistance, damage repair, and cell cycle arrest (Greer and Brunet 2008; Shimokawa et al. 2015).

The mechanisms by which the amplified activity of FOXO leads to lifespan extension are still unclear (Papatheodorou et al. 2014). FOXO increases insulin sensitivity by prompting the expression of insulin receptor and insulin receptor substrate 2 (IRS2) through feedback control (Puig and Tjian 2005). In addition, FOXO activation enables cell survival under conditions of food shortage by induction of cell cycle arrest and quiescence, reminiscent of the dauer phase induced in C. elegans (van der Horst and Burgering 2007). FOXO also delivers a metabolic shift from glucose to lipid oxidation, suppression of inflammation and enhanced mitochondrial biogenesis (van Heemst 2010). Identification of FOXO transcriptional targets has revealed a second tier of transcription factors that regulate a variety of other downstream responses (Alic et al. 2011). However, depending on the cell type, quality and strength of the stress, FOXO activity can also shift the cellular response from survival towards apoptosis (van Heemst 2010). FOXO activity in response to stress prevents aging and age-dependent impairments such as cancer, neurodegenerative diseases and diabetes (Calnan and Brunet 2008) and advocate organism longevity (Demontis and Perrimon 2010; Soerensen et al. 2015; Sun et al. 2015; Zeng et al. 2015).

Homologs of FOXO are DAF-16 in worm, dFOXO in fly (Puig et al. 2003; Tatar et al. 2003) (discussed in the next section) and FOXO1 (FKHR), FOXO3 (FKHRL1), FOXO4 (AFX) and FOXO6 in mammals (Papanicolaou et al. 2008). In mammals, a higher FOXO1 activity is observed in organs that control glucose homeostasis such as liver, pancreas, adipose tissue and skeletal muscle where, FOXO1 gets activated under low glucose conditions (Eijkelenboom and Burgering 2013; Zhang et al. 2011). High FOXO4 levels are exhibited in skeletal muscles, while FOXO3 is expressed abundantly in the brain, heart, kidney and spleen (Morris et al. 2015b). FOXO6 is observed in the brain, both during development and in adulthood (Eijkelenboom and Burgering 2013; Webb and Brunet 2014). FOXO3 presents an exciting avenue for clinical investigation of aging in human as it has been identified as the strongest example of longevity gene (Morris et al. 2015b). The association of FOXO3 with human longevity has been researched by many individual groups on different long-lived subjects (Anselmi et al. 2009). Studies of single-nucleotide polymorphisms (SNPs) associated with human FOXO3 gene in a group of American men of Japanese ancestry (known to live long) as well in other diverse human populations revealed the association of FOXO3 SNPs with longer lifespan (Murabito et al. 2012; Willcox et al. 2008).

Along with organism aging, FOXOs are also highly investigated for other intrinsic roles such as modulation of neuronal activities under low-insulin signalling conditions (Kaletsky et al. 2016), maintenance of germline progenitor cells in C. elegans (Qin and Hubbard 2015) and the maintenance of lower blood pressure and essential hypertension (Morris et al. 2015a).

IIS circuit across species

In Caenorhabditis elegans

C. elegans was among the first and the most intensively researched organisms to study longevity. It has 38 insulin-like ligands (Broughton and Partridge 2009) that initiate insulin signalling. IIS pathway in C. elegans includes proteins coded by daf-2, age-1, akt-2, daf-16 and daf-18 and they mediate regulation of dauer diapause, reproduction and adult lifespan (Claeys et al. 2002). Aging in C. elegans is influenced in a varied manner by signals from different tissues, with the most important signal from the CNS (Wolkow et al. 2000). These signals coordinate the aging rate between different tissues to create homeostasis, thereby affecting general aging of the whole organism. Communication between the tissues is intermediated by insulin-like peptide INS-7, which is regulated by daf-16 activity in the intestine. The daf-16 activity in the intestine in turn influences daf-16 activity in other tissues (Murphy et al. 2007). Mutations in the daf-2 gene extend worm lifespan along with a high release of enzymes involved in antioxidant activity (Barbieri et al. 2003). Working synergistically with TGF-β signalling pathway, daf-16 also cooperates with nematode SMAD proteins to mediate the activation of genes that impact nematode metabolism and development (Ogg et al. 1997). Reduced IIS also delays C. elegans aging through the activation of dauer-related processes during adulthood (McElwee et al. 2004; Partridge and Harvey 1993). However, certain reduced IIS conditions have been found to confer robust lifespan extension without dauer formation (Arantes-Oliveira et al. 2003; Gems et al. 1998) but through the involvement of NF-E2-related factor orthologue SKN-1. SKN-1 acts similar to DAF-16. It upturns the expression of collagens and other extracellular matrix genes that mediate adulthood extracellular matrix remodelling, which is an essential feature for longevity assurance (Ewald et al. 2015).

Of various environmental methods, DR is the most effective in increasing longevity through insulin signalling pathway. Various DR methods such as genetic mutations that lower pharyngeal pumping rate, changes in dilution of bacteria in liquid culture, changes in components of liquid media and the use of ‘DR mimetics’ have been employed to increase survivorship in C. elegans (Greer and Brunet 2009; Mair and Dillin 2008). In worms, daf-16 has been shown to be required for lifespan extension through DR (Greer and Brunet 2009). However, the use of many other model organisms have exposed DR to also affect longevity independent of daf-16 (Kenyon 2010). Along with DAF-16, transcription factor PHA-4 also responds to certain DR methods in C. elegans (Mair and Dillin 2008).

IIS pathway is also regulated by sirtuins (Burnett et al. 2011). In worm, overexpression of sirt-2.1 activates daf-16 by interacting with 14-3-3 proteins where, SIRT-2.1 and 14-3-3 act in parallel to IIS pathway to activate daf-16 and extend lifespan (Berdichevsky et al. 2006). sirt-2.1 can also turn on daf-16 directly by deacetylation (Kenyon 2005). The finding that sirtuins can deacetylate FOXO proteins directly, and the fact that IIS pathway mutants do not require sirt-2.1 for longevity, demonstrates that sirtuins in worm can impact FOXO and lifespan independent of IIS pathway (Kenyon 2010).

In Drosophila melanogaster

The IIS pathway in Drosophila is similar to that of C. elegans and consists of insulin/IGF receptor (INR), insulin receptor substrate (CHICO), PI3K and PI3K target protein PKB (Weinkove and Leevers 2000). There are eight insulin-like peptides (DILPs) in Drosophila (Brogiolo et al. 2001; Garelli et al. 2012; Gronke et al. 2010). The dilp genes show divergent temporal expression patterns (Broughton et al. 2005) and are expressed in varied sites (Nassel et al. 2015). DILPs are majorly secreted in the median neurosecretory cells of the brain called as insulin-producing cells (IPCs) (Ikeya et al. 2002). Removal of neurosecretory cells from Drosophila brain was found to increase organism lifespan and exhibit resistance to oxidative stress and starvation (Broughton et al. 2005). Among the eight DILPs encoded, DILP2 is homologous to human insulin and has been widely studied for lifespan regulation (Broughton et al. 2008). The dilp2 null-mutant flies live longer (Gronke et al. 2010). DILP6, which is largely produced in the fat body, also plays an important role as lifespan regulator (Okamoto et al. 2009). Overexpression of dilp6 in the abdominal fat body represses dilp2 in the brain, suggesting synergistic effects of dInR antagonist DILP6 and agonist DILP2 for longevity (Bai et al. 2012). On the whole, cross talk between DILPs of the brain and fat body influence IIS mediated longevity. Hypomorphic mutation (teq f01792) or downregulation of the gene Tequila in fly neurons, drastically increases longevity through reduced circulation of DILP2, reduced Akt phosphorylation, small body size and changed glucose homeostasis (Huang et al. 2015). Genetic inhibition of negative regulators of dFOXO such as DILPs (Broughton et al. 2005), dInR (Tatar et al. 2001) and CHICO (Clancy et al. 2001) extend the lifespan of Drosophila. On the other hand, overexpression of IIS negative regulators such as dPTEN or dFOXO extend the fly lifespan (Demontis and Perrimon 2010). Mutation of chico regulates aging by interacting with eukaryotic translation initiation factor 4E binding protein (4E-BP) (Bai et al. 2015). Tissue-specific overexpression of IIS components also regulate the fly lifespan in a cell non-autonomous way such as, through upregulation of dFOXO in the adult head fat body (Hwangbo et al. 2004), muscle-specific overexpression of either dPTEN or dFOXO (Demontis and Perrimon 2010) and the ablation of IPCs (Haselton et al. 2010). Stress responsive JNK pathway also requires FOXO to promote longevity in Drosophila. JNK counteracts IIS and translocates FOXO to the nucleus (Wang et al. 2005). Fly adult diapause results in a drastically prolonged lifespan with no senescence, increased stress resistance, somatic maintenance, developmental arrest and reallocated energy resources. IIS is one among the many pathways to get downregulated during diapause (Kucerova et al. 2016).

In response to DR, FOXO does not play any essential role for lifespan extension as flies knocked out of foxo live long under DR. But, it is suggested that the presence of an active FOXO affects the normal response to DR by changing the expression of FOXO target genes (Giannakou et al. 2008). Overexpression of foxo under DR is also associated with a reduction in DILP2 (Min et al. 2008).

Studies on dsirt2 (Drosophila sirtuin 2) show that SIRT2 and its orthologs play a crucial role in metabolic homeostasis and IIS (Liang et al. 2009; Schenk et al. 2011). dSIRT2 present in the fat body of fly is crucial for maintaining the metabolic regulatory network across the organism and for survival during starvation (Banerjee et al. 2012a, b).

In rodents

Altering genes involved in IIS pathway and the growth hormone (GH) cause longevity in mice (Bartke et al. 2013; Longo and Finch 2003). Deficits in GH and IGF-1 levels show reduced body size, delay in sexual maturation and age-related diseases, whereas elevated plasma levels of GH lead to increased body mass, early onset of reproductive maturity and senescence, high frequency of cancer occurrence and diminished lifespan in mice (Bartke et al. 2013). The GH-deficient mice also show high levels of antioxidant enzymes and increased stress resistance in muscle cells and fibroblasts (Brown-Borg 2009). Similar to GH-deficient mice, mice that do not have GH receptor-binding protein (GHR-BP) show 50% lifespan extension with high insulin sensitivity and lesser incidence of age-dependent diseases (Ikeno et al. 2009). Reduced occurrence of tumours in kidney and small intestine and protection against β-amyloid toxicity were found in GH-deficient mice (Cohen et al. 2009; Garcia et al. 2008; Patel et al. 2005). Decrease in IGF-1 delays the onset of cancer and immune decline and helps to maintain youthful cognitive ability (Anisimov and Bartke 2013). Mice carrying fat-specific insulin receptor knockout (FIRKO) genotype live 18% longer (Bluher et al. 2003). Reduction of insulin receptor signalling in the mouse brain extends lifespan by 18% (Taguchi et al. 2007).

DR in rodents reduced the levels of GH and IGF-1 (Lee and Longo 2011), but the link between DR, GH and aging is yet to be uncovered. The long-lived GH receptor knock-out (GHRKO) mice did not show extended lifespan or the health benefits associated with DR (Bonkowski et al. 2006; Westbrook et al. 2014), whereas the long-lived GH-deficient Ames mice responded to DR (Bartke et al. 2001). This suggests a complex involvement of the GH/IGF-1 axis and periphery pathways in response to DR. Starvation also protects mice from high dosage chemotherapy treatments by decreasing IGF-1 signalling in the serum. This suggests that starvation or severe DR can be used for disease treatment to a certain extent. However, this does not help in the treatment of cancer cells as the oncogenes block turning on of stress resistance during decline of nutrient levels (Raffaghello et al. 2008).

In worms and flies, a positive role for sirtuins in prolonging lifespan is well established. In mammals, certain sirtuins have been identified which are associated with different aging pathways. Alterations in their balance have proved beneficial in different models of age-associated diseases. Experiments with rodents are yet to examine lifespan extension from overexpression of the sirt gene, although resveratrol mediated activation of sirt1 was found to reverse partially the life-shortening effect of high fat diet (Kenyon 2010). Under DR, SIRT1 helps in stress tolerance by deacetylating FOXO1 (Brunet et al. 2004). SIRT1 also deacetylates FOXO4 along with p53 for stress response and cell survival (Outeiro et al. 2008) and FOXO3 to promote DR induced renal protection against hypoxia (Kume et al. 2010). Sirt6 global knockout mice were reported dead by 4 weeks of age due to severe hypoglycemia (Mostoslavsky et al. 2006). This attributes to cell-autonomous roles of SIRT6 in promoting mitochondrial respiration via suppression of HIF-1α (hypoxia-inducible factor 1-α) function and in reducing cellular insulin signalling, through mechanisms unknown (Xiao et al. 2010; Zhong et al. 2010).

In human

Human aging is more complicated than worm, fly and rodents and the relevance of IIS pathway to human longevity has generated more controversy. In humans, insulin insensitivity increases with age and this constitutes a risk factor for various conditions like hypertension, atherosclerosis, obesity that cause ill health, frailty and death among the elderly (Barbieri et al. 2003). Although the role of IIS pathway in humans is poorly understood, the genetic variants and/or combinations of small-nucleotide polymorphisms (SNPs) in the human components of the IIS pathway, including various FOXO SNPs have been found to correlate with low IGF-1 plasma levels present in centenarians (van Heemst et al. 2005). Overexpression of phosphatase and tensin homolog (PTEN) which dephosphorylates PIP3 to PIP2, causes a high antioxidant activity and longevity in human umbilical vein endothelial cell (HUVEC) cultures (Tait et al. 2015). Low IGF-1 levels have been predicted as the cause for long-lived individuals (Milman et al. 2014; Vitale et al. 2012). Female Ashkenazi Jewish centenarians show overrepresentation of loss-of-function mutations in the IGF1R (Suh et al. 2008) while, Laron dwarves, who are normally overweight and long lived, show reduced levels of IGF-1 and reduced incidence of type-2 diabetes and cancer (Guevara-Aguirre et al. 2011; Steuerman et al. 2011). Studies conducted on people with acromegaly (caused by elevated levels of GH and IGF-1) showed increased mortality rates (Clayton 2003). These findings suggest that “IGF-related longevity genes” do exist in the human population. When trying to establish a relationship between IIS signalling, tallness and longevity in humans, women having lower IIS scores exhibited shorter height with increased old age survival. Genetically induced IIS inactivation can thus be advantageous to women than men for survival at old age (van Heemst et al. 2005).

DR offers a likely route for longevity in humans (Cava and Fontana 2013; Zhao et al. 2014). The discovery of the positive effects of DR in various animal models led to the search for possibilities in middle aged humans. In humans, suppression of IGF-1 depends more on reducing protein consumption rather than decreasing calorie intake (Redman et al. 2010). Long-term DR in humans inhibits IIS pathway in skeletal muscle, with a transcriptional profile resembling those who are young in age. This indicates that humans, like rats, share similar transcriptional responses to DR for enhanced health and survival. However, this association cannot be implied as the exact causative for increased longevity (Mercken et al. 2013). It is yet to be resolved whether DR can extend lifespan in humans in addition to decreasing disease specific mortality risk.

The life-prolonging effects of sirt genes and the emerging evidences of link between sirtuins and DR in lower organisms and rodents prompted a number of investigations in humans to identify sirt genes associated with exceptional longevity. Human genome encodes seven sirt genes of which, sirt1 and sirt3 have been investigated for their role in longevity (Polito et al. 2010). sirt1 variation cannot be linked to reduced human mortality rate (Flachsbart et al. 2006; Kuningas et al. 2007; Willcox et al. 2008) but sirt3 has been found to play a role in modulating human longevity (Bellizzi et al. 2005; Lescai et al. 2009).

Epigenetic mechanisms of aging

Various cellular processes face age-dependent changes in gene expression regulated by different epigenetic machineries. Epigenetic regulation of aging is mediated through methylation of regulatory DNA sequences, histone protein modification and the expression of regulatory non-coding RNAs (Esteller 2002). With time, regulation of epigenetic marks affect the expression of specific genes involved in longevity (Han and Brunet 2012; Kouzarides 2007) and create physiological changes associated with aging (Thompson et al. 2010). While we discuss the role of different epigenetic mechanisms that impact aging, we also examine evidences for the influence of epigenetics on aging through IIS pathway.

DNA methylation

The extent of DNA methylation is inversely proportional to the activation of a gene. As an organism ages, DNA methylation affects in two ways. One, it brings a general decrease in global gene expression mediated by changes in the distribution of 5-methylcytosine (5-mC) across the genome (Waki et al. 2003) and two, it causes promoters of specific genes to likely shift from unmethylated to methylated status, leading to gene turnoff (examples—promoters of tumor, aging related genes such as runx3 and tig1) (Kim et al. 2004; Waki et al. 2003). A link has been observed in humans among 5-mC DNA methylation, chronological and biological age suggesting that DNA methylation patterns could be used as an aging clock (Benayoun et al. 2015; Hannum et al. 2013; Horvath 2013; Johnson et al. 2012; Mitteldorf 2016; Weidner et al. 2014).

There are many reports that establish DNA methylation regulation of aging (Ben-Avraham et al. 2012; Bennett-Baker et al. 2003; Lin et al. 2005; Tra et al. 2002). DR influences organism aging through ways other than IIS pathway. DR causes reversal of aberrant DNA methylation by aiming at specific loci rather than on a global level (Munoz-Najar and Sedivy 2011). Studies show that DR increases methylation levels of proto-oncogene ras (Hass et al. 1993), transcription factor RUNX3 (Kim et al. 2004) and tumor suppressor gene p16 INK4a (Lichtenberg 2011) (Fig. 2). Hypermethylated gene promoters are recognized by transcriptional repressors, leading to silencing of the gene. This contributes to DR mediated cancer prevention (Li et al. 2011c). Global decrease of DNA methylation with aging causes a decrease in DNMT1 and DNMT3a activity significantly (Casillas et al. 2003). DR contributes to a significant increase in DNMT1 activity to correct the decreased methylation level caused by aging (Li et al. 2010).

Fig. 2
figure 2

DR contributes to extended lifespan by delay of aging and aging-related diseases. It regulates aging through metabolic pathway and epigenetic processes of DNA methylation and histone modification. Reduced food intake affects IIS pathway to cause nuclear translocation and transcription of FOXO. FOXO activates or represses genes important for promoting longevity. DNA methylation regulation during dietary restriction involves DNMT activation and silencing of target genes such as ras. DR-induced histone modification involves up and downregulation of target genes through histone acetylation and methylation. DR-associated epigenetic regulation reverses aberrant gene expression and maintains chromatin stability.

Full size image

Effect of DNA methylation on aging through IIS pathway remains uncovered. However, a recent study shows that the expression of Dnmt1 and Dnmt3a is influenced by GH, suggesting an association between DNA methylation and the IGF-1/FOXO pathway (Armstrong et al. 2014).

Histone modification

Gene expression is also controlled by various histone modifications catalyzed by acetyltransferases, deacetylases, methyltransferases and demethylases (Kouzarides 2007; Mosammaparast and Shi 2010). These enzymes regulate the expression of several gerontogenes and pose as therapeutic targets that help avert or treat age-dependent impairments. Histone methylation, facilitated by histone methyl transferases and histone demethylases, impacts organism aging (Greer et al. 2011; Pu et al. 2015; Sen et al. 2015; Siebold et al. 2010). Histone acetylation, mediated by histone acetyltransferases (HATs) and histone deacetylases (HDACs) influences metabolic health and longevity in a conserved manner and its pattern changes during normal aging (Satoh et al. 2013; Toiber et al. 2013). Changes in activities of HATs and HDACs by genetic manipulation or by targeting with chemicals, natural polyamines (Eisenberg et al. 2009) and drugs show extensive changes in longevity across taxa (Benayoun et al. 2015).

DR-associated epigenetic regulation of lifespan extension also includes histone remodelling (Fig. 2). DR-induced stimulation of SIRT1 and HDAC1 leads to deacetylation changes in the expression of key genes such as foxo, p53, ku70, pgc-1a and p16 INK4a and DR-induced histone methylation regulates genes such as hTERT and p16 INK4a (Li et al. 2011b).

Various studies conducted prove the involvement of histone modifications in IIS pathway regulation of aging. Both DNA methylation and histone modification in regions of IGF-1 promoter alter the expression of Igf-1 gene (Yang et al. 2015). High glucose (HG) induced epigenetic modifications of IGF1R is found during development of diabetic cardiomyopathy (Yu et al. 2010). HG also increases the connection between p53 and HDAC subtype 1 and reduces the association between acetylated histone 4 and IGF1R promoter. This shows that HG induced suppression of IGF1R is arbitrated by the association of p53 with the IGF1R promoter, and also by the involvement of HDAC1 to the IGF1R promoter-p53 complex (Yu et al. 2010). Activation of methionine synthase by IGF-1 and dopamine has shown to fuel PI3K and MAPK-dependent pathways that amplify DNA methylation (Waly et al. 2004). UTX-1 has been identified as one of the epigenetic regulators of the IIS pathway (Jin et al. 2011; Maures et al. 2011) (Fig. 3). UTX-1 is the structural homolog of mammalian demethylases that target H3 K27me3 (Jin et al. 2011). In C. elegans, age-related increase of UTX-1 level leads to deletion of the repressive H3 K27me3 mark on genes daf-2, akt-1, and akt-2 in the IIS pathway. This removal leads to increased transcription of these genes and downregulation of DAF-16 activity leading to cell aging (Jin et al. 2011). FOXO/DAF-16 transcription factor also acts through the use of SWI/SNF (SWItch/Sucrose Non-Fermentable) chromatin remodeler (Riedel et al. 2013). While SWI/SNF modifies chromatin in an ATP-dependent manner and not through epigenetic change, the recruitment of SWI/SNF is not totally dependent on DAF-16 suggesting the possibility of involvement of epigenetic mechanism (Riedel et al. 2013). Chromatin modifier zinc finger protein 1 (ZFP-1) acts as a common facilitator of IIS and DR mediated lifespan regulation (Fig. 1). ZFP-1 interacts with Glioma amplified sequence-41 (GFL-1 in C. elegans) and DAF-16/FOXO regulates the expression of the different isoforms of zfp-1 and gfl-1. ZFP-1/GFL-1 negatively regulates the expression of DAF-16/FOXO target genes through feed-forward loops. This suggests that the knock down of zfp-1 or gfl-1 would lead to suppression of DAF-16 target gene expression under low insulin signalling conditions (Singh et al. 2016).

Fig. 3
figure 3

Schematic representation of the role of miRNAs and UTX-1 in IIS pathway to regulate C. elegans aging. miR-71, 238, 239 and 246 regulate worm lifespan. Targets have been established for some of the miRNAs. miRNA lin-4 downregulates LIN-14 expression by forming RNA duplex with 3′-UTR of lin-14 mRNA. LIN-14 causes direct repression of DAF-16 or indirect repression of DAF-16 through DAF-2 activation. miR-71 and miR-239 have opposite effects on AGE-1 and PDK-1 to regulate lifespan. During aging, UTX-1 levels increase and lead to the removal of the repressive H3K27me3 mark on genes daf-2, akt-1, and akt-2 leading to increased transcription of these genes and in turn cause downregulation of DAF-16 activity.

Full size image

Non-coding RNA surveillance of organism longevity

Non-coding RNAs include small regulatory RNAs such as microRNAs (miRNAs), siRNAs, piwi-interacting RNAs, QDE-2-interacting RNAs (qiRNAs) and a wide range of long non-coding RNAs (lncRNAs). They influence varied biological events by controlling gene expression and maintaining stability of the genome. Although the role of these regulatory RNAs is well studied in different biological processes, their role in aging mediated through IIS pathway remains largely unknown.

MicroRNAs

MicroRNA (miRNA) maturation and function requires argonaute like gene-1 (alg-1). Loss of alg-1 in adult worm has been found to affect lifespan, thus signifying the need of miRNAs in adulthood for proper aging (Kato et al. 2011). lin-4 of C. elegans was the first miRNA to be discovered, followed by the successive identification of other miRNAs (Lee et al. 1993). Many miRNAs have now been discovered in plants, animals and human, with over 2588 human miRNAs noted in the miRBase (http://www.mirbase.org) (Kozomara and Griffiths-Jones 2014). In C. elegans, miRNAs affect lifespan in both positive and negative manner and significant expression changes are observed in multiple miRNAs during aging (de Lencastre et al. 2010; Grillari et al. 2010; Kato et al. 2011). miRNA lin-4 targets 3′-UTR of lin-14 mRNA to form RNA duplex and suppresses the expression of lin-14 in C. elegans. LIN-14 blocks the activity of daf-16 and promotes the activity of daf-2 (Fig. 3). Loss of function of lin-4 or gain of function of lin-14 reduces the worm lifespan. Contrariwise, overexpression of lin-4 or knockdown of lin-14 results in extended lifespan (Baugh and Sternberg 2006; Boehm and Slack 2005). miR-71, miR-238, and miR-246 are the most overexpressed in aged animal and hence considered to increase longevity in C. elegans. miR-239 limits worm lifespan (de Lencastre et al. 2010). miR-71 and miR-246 act as biomarkers of aging in C. elegans (Pincus et al. 2011). miR-71 and miR-239 of C. elegans function through IIS pathway and show opposite effects on AGE-1 and PDK-1 to regulate lifespan. In addition to it, miR-71 also interacts with the DNA damage response pathway by knocking down two check point proteins, CDC-25.1 and CHK-1 (de Lencastre et al. 2010). miR-34 gets upregulated during aging, dauer phase and early dormancy of C. elegans (de Lencastre et al. 2010; Karp et al. 2011) and target the aging genes (Zhao et al. 2010) when compared to other miRNAs that get normally downregulated with time (de Lencastre et al. 2010; Ibanez-Ventoso et al. 2006). This is in contrary to the general upregulation of miRNA expression reported in mice during aging. Thus, variances can be seen in global and tissue-specific miRNA expression between worm and rodents (Li et al. 2011b).

Fly miR-8 (homolog of human miR-200) targets USH (ZFPM2 or FOG2 in humans), which obstructs PI3K of IIS pathway (Hyun et al. 2009). Also, miR-8 null flies were observed small in size with faulty insulin signalling in the fat body (Hwangbo et al. 2004; Hyun et al. 2009). This suggests miR-8/miR-200 to be vital for governing fly and human aging respectively (Hyun et al. 2009).

Genotype-by-age interaction (GbA) refers to the genotype-specific changes found in circulating levels of 21 miRNAs of young and old Ames dwarf mice during aging. Genotype-by-age miRNAs (GbAmiRNAs) target transcripts involved in tumor suppression, anti-inflammatory response and modulate Wnt, insulin, mTOR, and MAPK signaling pathways. Comparison of circulating GbAmiRNAs in Ames mice with DR-mediated circulating miRNAs in another long-lived mouse proved that both DR-independent and DR-dependent mechanisms increase lifespan in the Ames mouse (Victoria et al. 2015). Mouse model of Hutchinson-Gilford progeria syndrome (Enhanced aging due to improper maturation of lamin A in the nuclear envelope) shows higher levels of miR-1 in the liver, kidney and muscle when compared to wild-type animals (Marino et al. 2010). miR-1 suppresses IGF1 in progeroid mice, indicating that it functions through the IIS pathway (Marino et al. 2010). Aging mammals show miRNA expression to be tissue-specific such as in liver (Bates et al. 2010; Li et al. 2011a), brain (Khanna et al. 2011; Li et al. 2011b; Liang et al. 2011) and skeletal muscles (Drummond et al. 2011; Hamrick et al. 2010). This suggests for aging related tissue-specific functions of signalling pathways. miR-470, miR-669b and miR-681 in mammals suppress the expression of IGF1R and AKT to result in diminished phosphorylation of FOXO3 (Liang et al. 2011). The miR-17-92 cluster along with its paralogous clusters miR-106a-363 and miR-106b-25 that regulate cellular senescence, are reduced in several aging models and upregulated in some cancers (Faraonio et al. 2012; Grillari et al. 2010). These miRNAs generally target PTEN and their downregulation during aging elevates PTEN levels, which in turn suppresses the IIS pathway (Grillari et al. 2010).

Expression of most sirtuins is regulated by multiple miRNAs. Interaction between them helps in the regulation of multiple pathways. Interplay between SIRT1 and miR-181a regulates hepatic insulin sensitivity (Zhou et al. 2012) and the crosstalk between SIRT6 and miR-33a/b aids in fatty acid metabolism and insulin signalling (Davalos et al. 2011).

Long non-coding RNAs

Long non-coding RNAs (lncRNAs) control different facets of aging. LncRNAs are involved in various cellular activities, proliferation, and differentiation, quiescence, senescence and stress response related to aging (Grammatikakis et al. 2014; Jain et al. 2016). These include TERC and TERRA that control telomere length, Airn, PTENpg1-AS and H19 involved in epigenetic changes, lncRNA-p21 involved in translation repression process and MALAT1, ANRIL, eRNAs and 7SL that control cell division (Grammatikakis et al. 2014; Kim et al. 2016).

lncRNA transcribed telomeric sequence 1 (tts-1) are found on ribosomes of the worm daf-2 mutants. Deletion of tts-1 in daf-2 mutants increases ribosome levels and significantly decreases their lifespan. This suggests the expression of tts-1 in different longevity affecting signalling pathways to reduce ribosome levels and promote life extension (Essers et al. 2015). The H19 long intergenic non-coding RNA (lincRNA) is one of the most abundant and conserved transcripts expressed during mammalian development. Overexpression of miR-675 which is found in H19’s first exon upregulates Igf1r (growth-promoting insulin-like growth factor 1 receptor) gene. H19’s main physiological role is chartered by the controlled processing of miR-675. The controlled release of miR-675 from H19 might inhibit cell proliferation in response to cellular stress or oncogenic signals (Keniry et al. 2012).

Other regulatory non-coding RNAs

Endo-siRNAs are a class of siRNAs first identified in the germline and soma of flies (Chung et al. 2008; Czech et al. 2008; Ghildiyal and Zamore 2009). They were later identified in worm and mammals (Okamura and Lai 2008). Endo-siRNAs of Drosophila are produced in a Dicer-2 (Dcr-2)-dependent manner. Loss of Dcr-2 function reduced fly lifespan, making them more vulnerable to oxidative, endoplasmic reticulum, starvation and cold stresses along with abnormal lipid and carbohydrate metabolism (Lim et al. 2011). This suggests the involvement of endo-siRNA pathway in defence against stress and aging. However, the mechanism of involvement of endo-siRNAs in age-related processes is yet to be uncovered.

Piwi-interacting RNAs (piRNAs) were first discovered in germ line cells of organisms ranging from worm (Batista et al. 2008; Wang and Reinke 2008) to mammals (Aravin and Hannon 2008). They were identified to stabilise germ line cells by silencing transposons. Recent findings have identified piRNAs to be operative other than in germline cells (Malone et al. 2009; Rajasethupathy et al. 2012). The association of aging with changes in chromatin structure and methylation status (Liu et al. 2009) and the discovery of piRNAs to form complexes with RecQ1, ( DNA helicase necessary for genomic stability) (Lau et al. 2006; Sharma and Brosh 2007), it opens ways for piRNAs to be explored in the context of aging.

QDE-2-interacting RNAs (qiRNAs), first reported in the fungus Neurospora crassa, interact with the Argonoute protein QDE-2 and hence their name. DNA damage initiates the transcription of qiRNAs for repair mechanisms (Lee et al. 2009). Dicer, QDE-1 (an RNA-dependent RNA polymerase) and QDE-3 are needed for qiRNA transcription. QDE-3 is the homologue of Werner (WRN) and Bloom (BLM) DNA helicase of human RecQ family (Lee et al. 2009). The discovery of alterations in human WRN or BLM genes leading to premature aging (Martin and Oshima 2000) and the identification of telomerase as the mammalian RNA-dependant RNA polymerase (Maida et al. 2009), it offers a wide area of research in the field of biogerontology and qiRNA biogenesis in mammals.

Concluding remarks

Although evolutionarily conserved, IIS pathway is more complicated in higher organisms. IIS pathway controls life expectancy in many organisms, but despite recent research, its molecular mechanisms remain largely unknown. Disruption of IIS pathway in worm or fly causes significant increase in lifespan but it does not do so in higher animals as mammals have different receptors for insulin and IGFs with distinct pathways and diverse functions. Mammals have insulin/IGF-1 receptors in many organs whereas lower organisms have this signalling mainly through nervous system. IIS pathway in humans is responsible for approximately 5% of lifespan, but the kinds of dramatic increases seen from IIS in worms and flies are unlikely from similar manipulations in humans. On a molecular level, human longevity is mysterious. Further research is needed to throw light on the distinct molecular mechanisms of human longevity and their evolutionary trends. FOXO transcriptional factors play a crucial and diversified role in the IIS pathway. Activation of FOXO under low IIS activity leads to diverse effects on various tissues to aid in lifespan increase. Studies on FOXO in both lower and higher organisms open broader avenues for the investigation of lifespan extension. Interaction between diet and nutrient-sensing pathways plays a vital role in affecting organism physiology and health.

DR has attracted much attention as it extends lifespan and prevents or postpones a number of age-related diseases without causing irreversible developmental or reproductive flaws. DR increases lifespan by reducing IIS. DR also increases lifespan by stimulation of sirtuins, some of which in turn act through IIS. Exploration of DR coupled with IIS pathway and sirtuins holds a promising future for organism longevity.

Additionally, we show epigenomic and non-coding RNA-mediated manipulation of aging. DNA methylation and histone modifications are known to influence aging. Recent research shows their role in aging through IIS. Regulatory RNA-based epigenomics appears to be a novel understanding of aging regulatory pathway. Non-coding RNA induced epigenetic instability during lifespan forms an integral driver of the aging process. Among non-coding RNAs, the role of miRNAs in aging has been most widely researched. Components of IIS pathway are involved in miRNA-mediated regulation of aging in C. elegans. However, with both beneficial and antagonising roles of miRNAs on the lifespan of C. elegans, it opens a whole new arena of research both in lower and developed organisms. LncRNAs show diverse roles in regulating lifespan and age-related diseases. Their involvement in aging through IIS pathway is yet to be established. We suggest the importance of different regulatory non-coding RNAs in the pursuit of various factors for healthy aging.

Despite recent progress, a complete mapping of the insulin signalling pathway, role of DR in relation to IIS and the multiple roles of regulatory RNAs including microRNAs and lncRNAs that impact aging through nutrient sensing pathway is still elusive. It will be of great interest to further explore the association between epigenetic and non-epigenetic factors that influence the lifespan of an organism. This will help in the advancement of strategies to reverse different aging mechanisms in humans.