In contrast to life expectancy (also named mean lifespan, and frequently and wrongly termed “longevity”) that may be modified depending on living conditions, maximum longevity (henceforth referred to as “longevity” for practical purposes) is a species-specific feature. For instance, for the world as a whole, human life expectancy has increased more than two-fold in the last century, from 30 years in 1900 to 65 years in 2000 and is estimated to increase to 81 by the end of the 21st century. However, our longevity has remained at approximately 120 years [1]. At the same time, longevity varies widely among animal species with differences of up to 5000 times occurring, for example, between some invertebrates and the mollusk Arctica islandica, which at 507 years has the record for animal longevity [2,3], demonstrating the evolutionary plasticity of longevity. Therefore, animal (and human) longevity is flexible, regulated, and evolves rapidly during animal species evolution.

Despite the importance of understanding the mechanisms involved in longevity, the existence and nature of the molecular mechanisms determining and controlling this biological trait remains unclear. Some mechanisms discovered from studies within and among animal species, but not exclusively, include oxidative stress-related pathways [4,5], insulin signaling pathways [6,7,8,9] and mechanistic targets of rapamycin pathway [10,11,12,13]. Notably, genetic, dietary and pharmacological manipulations targeting some of these pathways resulted in more than 10-fold longevity extension in the worm Caenorhabditis elegans [14], but only about 1.4-fold longevity extension in rodents [15,16].

Evidence accumulated during the last 25 years demonstrate that lipids are determinant players of animal (and human) longevity. Thus, genomic studies demonstrate that genetic changes in genes controlling lipid metabolism play a role in human longevity [17,18,19,20], as well as in longevity differences among animal species [21,22] and lipid studies on changes in membrane lipid unsaturation and lipidomic profiles among animal species clearly point to lipids as a good potential target for research on molecular mechanisms underlying differences in longevity among animal species.

Lipids possess an inherent ability to spontaneously self-organize to generate membranes [23]. This property of lipids was determinant for the origin and early evolution of life [24,25]. The relevance of lipids in life is found in the fact that, without exception, all organisms in the three domains of life (eukaryotes, bacteria, and archaea) have lipid membranes [26]. Indeed, this hydrophobic film of around 30 Å of thickness marks the limit between life and death for cells [27]. The unique trait of the first lipids to form membranes was expanded during evolution to cell signaling as new functional property, and finally to the energy storage [28]. This diversification of functional properties demanded and was supported by an enlargement in the structural and functional diversity of lipid species that conformed early organisms and that evolved toward more complex cell systems. This complexity is reflected in the genetic code, which assigns more than 5% of their genes for the biosynthesis of thousands of different lipids [29], which participate in multitude of pathways and mechanisms that are essential for cell physiology.

In agreement with the current classification system, lipids are categorized into eight groups: sterol (ST) and prenol lipids (PR), saccharolipids (SL) and polyketides (PK), glycerolipids (GL), sphingolipids (SP), glycerophospholipids (GP), and fatty acyls (FA). There are, at present, no consistent assessments on the number of discrete lipid compounds in nature, but the estimations of the cellular lipid profile comprise thousands of different molecular species [30]. This complexity is magnified when the membrane compositional diversity is considered. For instance, the lipid compositional profile varies between animal species, between tissues–organ, between subcellular organelles, between membrane domains, and between the two leaflets of the lipid bilayer [27,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46].

The importance of lipids in cell structure and physiology can also be applied to the pathological condition. Thus, a large number of human pathologies are either etiologically or physiopathologically linked to alterations in lipid homeostasis including genetic disorders [51], and common diseases such as neurodegenerative diseases [52,53], cancer [54,55], non-alcoholic fatty liver [56], cardiovascular disease [57], metabolic risk [58], and obesity and diabetes [53], among several others.

From these findings we can infer that the membrane lipidome (the comprehensive lipid profile) is a structured, complex, dynamic and flexible system, which demands functional plasticity, a location-specific recognition, internal controls to monitor changes and confer stability, and adaptive responses to preserve essential membrane biological properties and cellular functions within physiological limits [37,59]. The result is the formation and maintenance of a specific membrane lipidome. Current evidence from long-lived humans and model organisms demonstrate that the membrane lipidome is associated with longevity, and that it is possible to define a lipid profile of long life.

Membrane unsaturation was an early evolutionary adaptation of biological membranes of the first organisms to maintain the viscosity of the lipid bilayer [60]. Membranes need to stay in the liquid crystalline state to avoid rupture and to provide a homeoviscous environment for the insertion of membrane proteins over the temperature range that a cell can be exposed to [60,61,62]. The affected lipid molecules are essentially fatty acids, the core of most membrane lipids such as glycerophospholipids and sphingolipids, leading to the formation of unsaturated fatty acids. Membrane fatty acids have the dominant influence on the physical and chemical properties of the bilayer they constitute.

Unsaturation of fatty acids carries an inevitable physiological consequence of spontaneous chemical oxidation under aerobic conditions, a process also termed lipid peroxidation [63,64]. It is important to highlight that the degree of sensitivity to oxidation of unsaturated fatty acids is determined by the number of double bonds present in the acyl chain [64,65]. Thus, in saturated or monounsaturated fatty acids, the sensitivity to oxidation is non-existent or practically null, while in unsaturated ones with two or more double bonds their sensitivity to oxidation increases rapidly [66,67]. Thus, from the profile and susceptibility to oxidation of fatty acids of a certain membrane, the so-called peroxidizability or peroxidation index (PI) can be calculated. The peroxidizability index is calculated as PI = [(% monoenoic × 0.025) + (% dienoic × 1) + (% trienoic × 2) + (% tetraenoic × 4) + (% pentaenoic × 6) + (% hexaenoic × 8)]. The higher the PI value, the more susceptible to oxidation is the membrane, and the lower the PI, the greater the resistance of the lipid bilayer to oxidation [66,67].

The seminal finding of a link between membrane unsaturation and longevity was the work of Pamplona et al. (1996) [72], which revealed that the PI of liver mitochondria from humans, pigeons and rats was negatively correlated with their respective longevities. Later, it was evidenced that this was the case for a wide range of tissues and animal species including both invertebrates and vertebrates. All these outcomes obtained at tissue and mitochondrial level were also extended to plasma lipids of mammalian species, including humans [35], demonstrating that the greater the longevity of a species, the lower its plasma PI, reinforcing the idea that lipid unsaturation is a general adaptive trait associated with animal longevity.

While longevity can significantly differ within and between mammals and birds, there can also be important longevity differences within a particular species, as is the case, for instance, within both the wild-derived or the senescence-accelerated mouse (SAM) strains of mice. In both comparisons, the PI of the mouse strain displaying extended longevity was significantly lower that than of the short-lived strain [73,85,103]. Identical results were obtained when membrane unsaturation and PI was evaluated in exceptionally long-lived specimens of mice [95]. Additionally, the comparison between closely-related species with divergent longevity, such as the long-lived white-footed mouse P. leucopus and the short-lived M. musculus, also displayed a lower PI for the long-lived mouse when compared to the common laboratory mouse [97]. More importantly, two exceptionally long-living mammalian species (naked mole-rats and echidnas) also have membrane fatty acid profiles that are resistant to peroxidative damage as would be predicted from their longevities [32,84,89,104].

Honeybees (A. mellifera), flies (D. melanogaster), and worms (C. elegans) offer new examples of variations in longevity within species that expand previous observations in vertebrates to invertebrates. In honeybees, queens have a longevity that is an order-of-magnitude greater than that of workers [105]. In flies, a comparison is possible between long-living mutants [91] and wild type strains, differing in their longevities [100]. In worms, mutant strains of C. elegans that differ 10-fold in their longevities can be compared [92]. The results from all these approaches show that long-lived animals possess a lipid profile of their membranes that is more resistant to oxidation. Thereby, the greater the longevity of the honeybee (the queen), the mutant and strain of fly, and the mutant worm, the lower the membrane unsaturation and more protected are the other cellular components [86,91,92,100], supporting again the idea that membrane composition is an important feature in the determination of longevity.

Since correlation does not necessarily mean causation, it is important to know whether cellular components are protected in front of lipid oxidation when membrane unsaturation is modified. With this goal, different experimental studies, based on dietary interventions, were designed to force alterations in membrane lipid composition and to evaluate the changes in both PI and lipoxidative damage of proteins [112,113,114]. The results demonstrate that a low degree of lipid unsaturation of cellular membranes protects the membrane itself and other molecular components against lipid peroxidation-derived products and lipoxidation-derived damage. Importantly, the magnitude of the change in membrane unsaturation is tissue-dependent, and the magnitude of the lipoxidation-derived molecular damage is not proportional to the membrane unsaturation change [112,113,114].

Additional evidence that supports a relationship between membrane unsaturation and longevity comes from dietary interventions that extend longevity in experimental models. Thus, diverse dietary restrictions, such as caloric-, protein-, and methionine restrictions (CR, PR and MetR, respectively), were applied in different animal species (essentially rodents) inducing decreases in the degree of membrane unsaturation, the in vivo and in vitro lipid peroxidation, and the level of lipoxidation products in a diversity of tissues such as liver, heart, kidney and brain [115,116,117,118,119,120,121,122,123,124,125,126,127,128,129,130]. In this line, decreases in the levels of lipofuscin in tissues of C. elegans and rodents have also been reported after CR diets [110,131,132,133,134].

The molecular basis of the low PI presented by long-lived species is in the redefinition of the lipid profile that conforms membrane composition, with this redefinition being independent of the diet and according to a genotypic pattern [4,5,37,66]. Thereby, there is a general switch affecting the content and nature of unsaturated fatty acids present, from highly unsaturated fatty acids such as docosahexaenoic- (DHA, 22:6n-3), eicosapentaenoic- (EPA, 20:5n-3), and arachidonic (AA, 20:4n-6) acids in short-lived animals to the less unsaturated alpha-linolenic (ALA, 18:3n-3), linoleic (LA, 18:2n-6), and oleic (OA, 18:1n-9) acids in the long-lived ones. Remarkably, the shift produced between fatty acids in the membrane composition is carried out strictly by combining types of unsaturated fatty acids, and not by increasing the content of saturated fatty acids. The result is that the relationship between saturated and unsaturated fatty acids remains stable, the physical-chemical properties of the membrane are not affected, but its susceptibility to oxidation is clearly modified.

From these findings, it can be inferred that the support for the longevity-related differences in fatty acid profiles necessarily implies the unsaturated fatty acid biosynthesis pathways including desaturases, elongases, and peroxisomal beta-oxidation, as well as the diacylation–reacylation cycle. The estimation of elongase and desaturase (delta-5 and delta-6) activities from specific product/substrate ratios indicate that there is a substantial decrease in long-lived species when compared with short-lived ones [4,5,37,66,92]. This can explain why, for instance, LA and ALA decreases and DHA and AA increases from long- to short-lived animals. Therefore, elongation–desaturation pathways are responsible for the availability of unsaturated fatty acids for the de novo synthesis of lipids and later diacylation–reacylation cycle which, in turn, are responsible for the remodeling of glycerophospholipid acyl groups and, consequently, the determination of the lipid profile for a given membrane. In agreement with the proposed mechanism, the study from [21] using a phylogenomic approach certified that, at least in mammals, genes involved in fatty acid biosynthesis (lipoxidation repair, elongases, desaturases, and fatty acid synthase) have collectively undergone, in a specific and significant way, increased selective pressure in long-lived species. These findings were later confirmed [43].

Long-lived animal species (and humans) possess a low degree of membrane unsaturation based on the redefinition of the unsaturated fatty acid profile. This elegant evolutionary strategy allows decrease in the sensitivity to lipid peroxidation without altering fluidity, a basic property of cellular membranes. This occurs because membrane fluidity significantly increases with the incorporation of the first (usually around the center of the fatty acid) and less with the second double bond, whereas additional double bonds (usually toward the extreme of the acyl chain) cause few further variations in fluidity [138]. In contrast, peroxidation index increases with the number of double bonds, irrespective of its location [65]. Thus, the switch in the fatty acid profile from short- to long-lived animal achieves a lesser sensitivity to lipid peroxidation at the same time as maintaining the membrane fluidity. This idea, evocative of membrane acclimation to different environments at unsaturated fatty acid level in bacteria and poikilotherm animals, was defined as the homeoviscous longevity adaptation theory [66] and confirms the biological membranes as a dynamic structural adaptive system.

Fatty acids are essential components of the structural diversity of lipids of all cell membranes. Indeed, by permuting headgroups and fatty acids, more than 10,000 lipid molecular species can be built [139]. Recent developments in mass spectrometry- (MS)-based lipidomics offer the means to unambiguously and accurately detect, identify and quantify a huge number of lipid molecular species, as well as to profile large-scale changes in lipid composition. Current data also suggest that the lipidomic profile is an optimized trait associated with longevity.

The comprehensive lipidomic profiles and lipid concentrations identify the animal species, i.e., the lipidome is species-specific, as well as the long-lived animal species [35,41,43,140]. These comparative studies were carried out in a diversity of tissues such as cerebellum, cortex, heart, kidney, liver, muscle, and plasma, and exclusively in mammalian species including exceptionally long-lived animal species such as the naked mole-rat, bats and humans. Lipidomics studies also define exceptionally long-lived humans (centenarians) [48,96,141,142].

The findings from these studies reveal the presence of common lipidomic features, which accurately predict animal longevity, and the centenarian condition. More specifically, the relationship between lipids and longevity were ascribed to lower or higher lipid concentrations in species of specific lipid categories such as glycerolipids (diacylglycerols, DAGs), glycerophospholipids (GPs) (particularly ether lipids), sphingolipids (SPs), and long chain free fatty acids (LC-FFAs) [35,43,140]. Although the functional significance of most of these changes remains to be elucidated, some considerations can be exposed.

DAGs are precursors for the de novo GP biosynthesis, one of the main components of cell membranes.

Ether lipids are a subclass of GPs, mostly present as phosphatidylcholine (PC) and phosphatidylethanolamine (PE) species [143]. Ether lipids have an alkyl chain attached by an ether bond at the sn-1 position of the glycerol backbone. The sn-2 position of ether lipids has an ester-linked acyl chain, as in diacyl phospholipids. Some ether-linked phospholipids, called alkenyl-acylphospholipids, contain a cis or Z double bond adjacent to the ether linkage and are commonly referred to as plasmalogens [143]. Ether lipids represent about 20% of the total GP pool in mammals and have a tissue-dependent distribution [144]. Findings from studies in animal models of exceptional longevity such as A. islandica [3] and the naked mole-rat [32], as well as C. elegans [145], suggest an association between plasmalogen content and animal longevity. For instance, the naked mole-rat had no DHA-containing plasmalogens but had much higher levels of total PC plasmalogens. For bivalves, no correlation between plasmalogens content and longevity comparing five different bivalves’ species was found. For worms, C. elegans strains carrying loss-of-function mutations in genes encoding protein required for ether lipid biosynthesis demonstrated a shorter lifespan, and a decreased resistance to oxidative stress. In humans, lipidomic studies describe different plasma concentration of ether lipid species in centenarian [48,142] and in middle-aged offspring of nonagenarians [96], suggesting an ether lipid signature in long-lived humans. This signature is featured by higher level of alkyl forms derived from PC and decreased content in alkenyl forms from PE [142]. Remarkably, the compositional pattern in fatty acids of these ether lipids is specific, resulting in an ether lipid profile in centenarians that is more resistant to lipid peroxidation [142]. The physiological role of ether lipids, and specially plasmalogens, is associated with their function as membrane components [143]. However, interestingly, an antioxidant role has also been attributed to plasmalogens, which, similar to a scavenger, protects membrane unsaturation against oxidation [146]. Therefore, it was proposed that the ether lipid signature is a specific trait of long-lived humans.

Another important category of lipids highly conserved in all eukaryotic cells is sphingolipids (SPs). SPs possess a natural predisposition to the formation of lipid domains [147]. In the metabolism of sphingolipid, ceramide plays the role of a central metabolic hub [148] from which a wide diversity of structural and bioactive SP species is synthesized. Sphingolipids regulate a number of functions at membrane and cellular level. Its relationship with longevity proceeds from studies in model organisms such as yeast and flies. In these animal species, deletions or mutations of genes affecting ceramide biosynthesis or metabolism were able to significantly modify longevity [149,150,151]. Lipidomic studies in animal models, such as S. cerevisiae [152], C. elegans [153], the naked mole-rat [154], and mammals [35,43], also suggest an association between sphingolipid content and animal longevity. Specifically, a low sphingolipid content, particularly for sulfatides, ceramides and glycosphingolipids, seems to be a feature of long-lived animal species. Lipidomic studies in long-lived humans are limited and results are not conclusive [48,96,141,155,156,157].

Finally, the importance of lipid metabolism in determining longevity was also verified with respect to the plasma concentrations of LC-FFAs in different mammals, including humans [35]. The results reveal that long-lived species present lower plasma LC-FFA concentration, PI, and lipid peroxidation-derived products content, conferring to plasma, analogous to cell membranes, superior resistance to oxidation. Additionally, LC-FFAs are also essential by covering mammalian energy needs [158]. Interestingly, adipose tissue is the main source for the liberation of FFA into plasma. Consequently, the obtained results from the comparative approach suggest an evolutionary adaptation for adipose tissue regulating the kind of fatty acids stored and released to plasma. The result would be an adipose tissue with a composition and activity that is species-specific and designed to maintain a lower degree of unsaturation in long-lived species. Another important consequence derives from the fact that LC-FFAs also have signaling properties in several physiological processes [159,160,161]. Among them, a particularly relevant property is their direct effect on insulin secretion. Thus, it is plausible to hypothesize that long-lived mammals and human exhibit a lower LC-FFA concentration to maintain an attenuated activity of the insulin signaling pathway, which is in agreement with the consideration of insulin signaling pathway as an evolutionary conserved mechanism involved in the determination of animal longevity [162,163]. In this context, the lower concentration of LC-FFAs observed in humans could be considered an evolutionary adaptive response that could explain pathological processes like insulin resistance, which is linked to an increased LC-FFA plasma concentration [158] and, in turn, to a shortened longevity.