Why and how do termite kings and queens live so long?

Abstract

Lifespan varies greatly across the tree of life. Of the various explanations for this phenomenon, those that involve trade-offs between reproduction and longevity have gained considerable support. There is an important exception: social insect reproductives (queens and in termites, also kings) exhibit both high reproductive outputs and extraordinarily long lives. As both the ultimate and proximate mechanisms underlying the absence of the fecundity/longevity trade-off could shed light on the unexpected dynamics and molecular mechanisms of extended longevity, reproductives of social insects have attracted much attention in the field of ageing research. Here, we highlight current ecological and physiological studies on ageing and discuss the various possible evolutionary and molecular explanations of the extended lifespans of termite reproductives. We integrate these findings into a coherent framework revealing the evolution of longevity in these reproductives. Studies on termites may explain why and how ageing is shaped by natural selection.

This article is part of the theme issue ‘Ageing and sociality: why, when and how does sociality change ageing patterns?’

1. Introduction

Termites are one of the most abundant terrestrial animals on earth [1,2]. A key factor in terms of their ecological success is the reproductive division of labour that is a central feature of eusociality. In a colony, a limited number of male and female reproductives (kings and queens) produce all the offspring and a large number of non-reproductive individuals (workers and soldiers; but workers can become at least neotenic reproductives in most lower termites [3–5]) perform most of foraging, nest-building, brood care and nest-guarding [6,7]. Intriguingly, reproductives and non-reproductives exhibit up to a 100-fold difference in lifespans, and even more compelling is the fact that the reproductives live for an order of magnitude longer than solitary insects [8]. These features are also found in social Hymenoptera such as ants and honeybees, but the eusociality of termites has evolved independently of social Hymenoptera [9,10]. Therefore, unlike social Hymenoptera, termites have hemimetabolous development, bisexual societies and diplodiploid sex determination system [7]. Most notable is the continuous presence of sperm-producing kings in termites, which is a universal characteristic among termites, but completely absent in social Hymenoptera [1,7]. These differences result in an interesting feature in which only the queens are long-lived in social Hymenoptera, while in termites, not only the queens but also the kings exhibit both high reproductive outputs and extraordinarily long lives [5,7,11–16]. Although longevity is negatively correlated with reproduction in most organisms [17–19], the trade-off between fertility and longevity is apparently absent in termite reproductives [20]. Both the ultimate and proximate mechanisms underlying this unique characteristic could shed light on the unexpected dynamics and molecular mechanisms of extended longevity.

The aim of this review is to summarize recent ecological and physiological studies of termite ageing and longevity. First, we describe the evolutionary and ecological determinants of the long lifespans of termite reproductives. Then, we discuss the physiological and molecular mechanisms in play. Finally, we integrate these findings into a coherent framework explaining the evolution of longevity in such reproductives.

2. Evolutionary and ecological determinants of long lifespans

To explain why termite reproductives exhibit extended lifespans, it is necessary to understand the evolutionary aspects of longevity. The key idea underpinning the evolution of the lifespan is the ‘selection shadow’: the strength of natural selection declines after sexual maturation and with progressing age [17,21–25]. The selection shadow is shaped by age-independent random extrinsic mortality, which is in turn affected by external factors such as predation and early-life reproductive investment. But note exceptions if extrinsic mortality is condition-dependent, not random [26]. High extrinsic mortality reduces the probability of later-life reproduction, reducing selective forces at old age. Such decreases allow alleles with late-acting deleterious effects to accumulate over generations (the mutation accumulation theory) [27]. Moreover, genes associated with beneficial early-life effects would be favoured by selection even if they were deleterious in later life (the antagonistic pleiotropy theory) [21]. The deleterious effects cause physiological deterioration (ageing, also termed senescence) of an organism as age advances, reducing reproductive capacity and increasing the risk of death [28].

(a). Low extrinsic mortality

A major reason why termite reproductives live so long is that they experience low extrinsic mortality after the colony is established. The evolution of a reproductive division of labour is associated with strong social-level defence by non-reproductives [29], including nest defence, social immunity and self-sacrificial behaviours. These reduce the risks of predation, disease, starvation and desiccation, greatly limiting the extrinsic mortality of reproductives (figure 1 and table 1).

Figure 1.

Schematic of the evolution of extended longevity in termite kings and queens. The upper panel shows the susceptibilities to extrinsic mortality of solitary insects (left) and termite reproductives (right). Black solid circles indicate defence systems; termite reproductives implement both individual- and social-level defences (the latter by the evolution of eusociality). The red arrows indicate the effects of extrinsic mortality factors on individuals. Termite reproductives may be better protected than solitary insects against such attacks. The lower panel shows the evolutionary dynamics of longevity in short-lived solitary insects (left) and long-lived termite reproductives (right). The black solid curves indicate survival levels in the wild, and the grey areas the ‘selection shadows’. The reduced extrinsic mortality of termite reproductives extends the period available for longevity selection. (Online version in colour.)

Table 1.

Ecological and physiological characteristics in termites, and the predicted effect on extended longevity of the reproductives.

characteristics	expected effect on extended longevity of termite reproductives	references
Ecological characteristics
nest defence	nest defence, which involves nest structures in combination with the defensive behaviour of the non-reproductives, contributes to the decrease in the risks of predation, thermal stress and desiccation in the reproductives.	[30–32]
social immunity	social immunity, which is one adaptive mechanism that helps to alleviate the risks of pathogen infection in group-living termites, contributes to a reduced risk of mortality owing to diseases in the reproductives.	[33–37]
group foraging	group foraging by non-reproductives involves high foraging activity and new food discovery and storage, thus avoiding the risk of starvation in the reproductives.	[38–43]
hypoxic nest	hypoxic nest could protect the reproductives from oxygen toxicity.	[44,45]
Physiological characteristics
oxidative stress resistance	antioxidant system neutralizes the overproduction of reactive oxygen species associated with oxidative stress, which contributes to delayed ageing and long lifespan of the reproductives.	[46–48]
DNA repair	various repair pathways repairing DNA damage associated with long lifespan of the reproductives.	[49]
transposon defence	transposon defence such as PIWI-interacting RNA inhibits transposon activity, which contributes to improving genomic instability and slowing ageing in the reproductives.	[12]
downregulation of growth signalling	downregulation of insulin/IGF-1 signalling (IIS) and target of rapamycin (TOR) signalling prevents excessive biosynthesis resulting in the loss of homeostasis and contributes to longevity in the reproductives.	[50]

In general, termite reproductives live deep inside the nest of the central area (e.g. ‘royal chamber’ in a few species) of an enclosed and protected mound or nest (subterranean, epigeal or arboreal) safe from predation, thermal stress and desiccation, where tens of thousands of non-reproductives engage in social labour [30]. The most elaborate royal chamber is that of Macrotermes termites. It resembles a thick-walled protective bunker located in the most sheltered central part of the nest, often immediately below ground level [30,51,52]. The subterranean Reticulitermes termites also have royal chambers that are generally located deep inside the wood of the central areas [31,44]. Termite nests are multi-layered structures; many chambers are connected vertically and horizontally via very small openings. Thus, predators such as ants are required to break the multi-layered defence mounted by non-reproductives to access the royal chamber.

The social behaviour and group living of termites reduce the risks of disease and starvation in reproductives. Termites limit the spread of infection within the nest; the antennae recognize pathogenic conidia [53], followed by initiation and maintenance of allogrooming (social grooming) of contaminated body parts [33–35]. The additive effects of individual immune responses and their social amplification are also important in terms of anti-pathogen defence. Grouped termites cope better with disease than do isolated individuals [33,36], suggesting that resistance is socially enhanced [37,54]. Group foraging may reduce the risk of starvation in reproductives; especially in some foraging termite species, non-reproductives forage actively, discover new food sources and store foods [38–42]. If termite non-reproductives have residual nutrients to be shared via trophallaxis, the reproductives are prioritized; the reproductives thus survive for longer even within a starved colony [43].

In non-reproductive workers, there tend to be differences in the maximum lifespan among termite species [5], which may be dependent on their lifestyles. Termites can be divided into wood-dwelling and foraging termites (also called one-piece type and separate type, respectively) [3,55]. The wood-dwelling termites use a single piece of wood both as food source and shelter and produce simple and small colonies with totipotent workers that can become reproductives, where workers are as protected as reproductives. On the other hand, the foraging termites are characterized by multiple-pieces nesting and form large and complex societies where irreversible workers that usually cannot become reproductives perform risky tasks such as foraging outside of the nest. In foraging termites, the high mortality of workers is expected to result in earlier ageing and shorter lifespan compared to the workers of wood-dwelling termites. Currently, there is little evidence that their lifestyles have an effect on the longevity of non-reproductives (and also reproductives) as a general rule, but the reproductives of Macrotermes termites, foraging termites, are thought to reach 20 years of age while workers can live only a few months [41], whereas the lifespan of reproductives and workers is about the same 4–5 years in Zootermopsis termites, wood-dwelling termites [56] (an overview of longevities for termite reproductives and workers was reviewed in [5]).

3. The physiological and molecular mechanisms of extended longevity

An understanding of the physiological and molecular mechanisms in play would illuminate how termite reproductives achieve both high reproductive outputs and extraordinarily long lives. The evolutionary theory of ageing proposes two proximate theories; these are the ‘energy’ and ‘function’ trade-offs between growth, reproduction and longevity (the disposable soma theory and the developmental theory of ageing, respectively) [57]. The former theory considers that ageing is a trade-off between reproduction and somatic maintenance [24,58,59]. This suggests that delayed ageing and extended longevity can be achieved via greater allocation of resources to longevity assurance mechanisms, such as antioxidant and DNA repair systems, that slow the age-associated accumulation of physiological damage. In the alternative to damage accumulation, the latter theory suggests that ageing reflects suboptimal gene function in later life, mechanistically linked to the idea that superfluous nutrient-sensing growth pathways during adulthood can cause excessive biosynthesis that triggers functional decline (the hyperfunction hypothesis) [60–63]. Together, these trade-off theories are thought not to be mutually exclusive [64] and, broadly speaking, both sides suggest that the loss of homeostasis causes ageing.

(a). Oxidative stress resistance

The loss of homeostasis is caused, in part, by oxidative stress. Reactive oxygen species (ROS) that are by-products of mitochondrial aerobic energy metabolism [65] enhance immunity and cellular signalling when they are present at certain levels. However, when enzymatic and non-enzymatic antioxidants cannot fully neutralize overproduced ROS, the resulting imbalance seriously damages important biomolecules such as proteins, lipids and nucleic acids [66–71]. The basic homeostatic balance is compromised by either increased ROS production per se or reduction in the effectiveness of defences.

An efficient antioxidant system as a longevity assurance mechanism may explain why termite queens enjoy long lifespans. If mutations causing oxidative stress-induced DNA damage are not eliminated from the germline, they may pass to the offspring, reducing their fitness. Therefore, the germline must be highly protected against oxidative stress in long-lived organisms. The long-lived queens of Reticulitermes speratus show significantly higher activities of antioxidant enzymes (catalase and superoxide dismutase) than short-lived non-reproductives and suffer significantly less oxidative protein, lipid and DNA damage (figure 2) [46,47]. In particular, the catalase activities of termite queens were higher than those of certain solitary insects and social Hymenoptera, reflected by an increased expression of the catalase gene RsCAT1 [46]. In addition, a study on another subterranean termite R. chinensis revealed that differentiation of workers into neotenic reproductives was associated with increased catalase gene expression [48]. These studies partially support the hypothesis that transition to a reproductive state is associated with a gradual decrease in the extent of oxidative damage to body tissues (the oxidative shielding hypothesis) [72].

Figure 2.

Oxidative stress resistance in termite queens. (a) The ageing process associated with oxidative stress. (b) Comparisons of the catalase and Cu/Zn-superoxide dismutase (Cu/Zn-SOD) enzyme activities (antioxidant systems) of short-lived R. speratus workers and long-lived queens. (c) Comparisons of the oxidative damage levels of various biomolecules in R. speratus short-lived workers and long-lived queens after free radical formation by ultraviolet irradiation (left and right boxes, respectively). The error bars denote standard errors of the means. The statistical significances are *p < 0.05 and **p < 0.01. White and black bars indicate workers and queens, respectively. PC, protein carbonyl; MDA, malondialdehyde; 8-OHdG, 8-hydroxy-2′-deoxyguanosine. Modified from Tasaki et al. (2017) [46] and Tasaki et al. (2018) [47]. (Online version in colour.)

Lifestyles may influence the level of investment in their antioxidant system. Contrary to the results of the study using R. speratus [46], comparative studies of young and old individuals of the less socially complex (wood-dwelling) termite Cryptotermes secundus revealed that protein oxidative damage levels were lower in workers than in reproductives [73] and that workers increase their protection with age but not reproductives [74]. In addition, another study using C. secundus found a clear oxidative stress defence signal under stress conditions, inducing ageing, that was stronger in workers than in queens [75]. These findings suggest that workers that are totipotent to become reproductives, like in C. secundus, should invest more in longevity assurance mechanisms than sterile workers because the former can still reproduce and have not reached maturity yet [75].

(b). DNA integrity and transposon defence

Genomic instability and DNA damage can trigger ageing in multicellular organisms [76]. DNA damage is repaired via various pathways [77]. In R. speratus, reproductives expressed higher levels of DNA repair genes than did other castes [49]. Notably, the BRCA1 gene RsBRCA1 expression in somatic tissues was higher in long-lived kings than short-lived workers. Although BRCA1 is one of the best-studied DNA repair genes, particularly in the context of cancer research, any role of the gene in terms of organismal ageing or longevity remains unclear. The cited termite study proposed that the BRCA1-associated DNA repair pathway contributed to longevity.

DNA damage occurs during germline meiotic DNA replication. The DNA repair proteins BRCA1, MCPH1, XRCC3 and MLH1 are associated with meiotic progression and maintenance of genomic stability [78–81]. These genes RsBRCA1, RsMCPH1, RsXRCC3, and RsMLH1 were significantly upregulated in reproductive tissues compared to their expression in somatic tissues of R. speratus reproductives [49]. The expression pattern of these DNA repair genes suggests that they may protect the germline from the accumulation of progressive DNA damage. The strong evolutionary pressures placed on long-lived organisms with long reproductive periods may have enhanced the expression of DNA repair genes in reproductive tissues.

Transposon defence may also be involved in the exceptional longevity of termite reproductives. Transposable elements (TEs), also termed ‘jumping genes’, are DNA sequences that move from one location on the genome to another and thus enhance genomic instability [82]. Increased TE activation has been linked to ageing in the mouse [83], a fly [84] and a yeast [85]. A recent study on the termite Macrotermes bellicosus found that although TE expression increased with advancing age in short-lived workers, this was not the case in long-lived reproductives [12]. The study also performed age-estimates and used reproductives (around 9 years for old and 3–4 years for young) and non-reproductives (weeks to months) with large differences in age. Notably, the reproductives upregulated PIWI-interacting RNA (piRNA) biosynthesis; these RNAs silence TEs [86]. Thus, the extended lifespan of termite reproductives may be explained in part by TE suppression; this ensures longevity.

(c). Downregulation of growth signalling

The nutrient-sensing growth pathways (e.g. growth hormone signalling, insulin/IGF-1 signalling (IIS) and target of rapamycin (TOR) signalling) at the heart of the trade-off theories would be an important determinant of ageing in termite reproductives. Downregulation of these growth signalling pathways as well as dietary restriction are thought to reduce resource allocation to reproduction, growth and biosynthesis, thus prolonging lifespan in organisms [87–89]. A recent study using C. secundus to investigate transcriptome differences between young and old individuals revealed that aged primary reproductives (more than 7 years old) have lower expression levels of IMP-L2 and PRMT1 genes involved in IIS suppression and ATPsynD gene related to protein homeostasis under activated TOR signalling than young ones (1 year old) [74]. This suggests that lifespan determination and ageing processes might be modulated by the typical ageing pathways IIS and TOR in termites as in other organisms. In addition, 4-year-old reproductives of the termite R. chinensis expressed significantly lower levels of the mTOR, eIF4, and RPS6 genes involved in TOR signalling and IIS than did workers, suggesting that long-lived reproductives may downregulate these growth signals [50]. In general, these nutrient-sensing growth pathways effect juvenile hormone (JH) production followed by upregulation of vitellogenin (Vg), a yolk protein required for egg production [90,91]. Recent studies outlined the view that re-wiring of the IIS–JH–Vg circuit has occurred in social insects, which may explain the re-shaping of the fecundity/longevity trade-off [92]. However, most data have come from the honeybee, where JH has no further function in the maintenance of the reproductive status in queens [93,94]. The positive correlation between JH levels and vitellogenesis has been well studied in termites [95–99], but almost nothing is known about the details of the IIS–JH–Vg circuit. Further research on the relationship between the IIS–JH–Vg circuit in termites is needed.

There is also the possibility that termite reproductives, especially queens, may perhaps reduce this signalling without sacrificing high-level reproductive performance, by employing other signals to control reproduction. The transcription factor termed carbohydrate-responsive element-binding protein (ChREBP), a glucose sensor regulating the expression of genes that drive fatty acid biosynthesis, was highly expressed in mature queens of eight different termite species, compared to the levels in sterile workers and soldiers [100]. ChREBP critically mediates the glucose-dependent induction of glycolytic and lipogenic genes in metabolic tissues [101] and may thus provide essential precursors of oogenesis via redirection of a significant proportion of glucose carbons to de novo lipogenesis and nucleotide biosynthesis. Thus, in addition to growth hormone signalling, IIS and TOR signalling, the other signalling pathways such as ChREBP-mediated glucose signalling need to be further investigated.

4. Hypoxic adaptation and longevity evolution in termite reproductives

The physiological and ecological mechanisms of extended termite longevity may be evolutionarily linked via hypoxic adaptation. Termite reproductives are protected from the external environment because they live in closed nests, where the hypoxic condition can be viewed as a by-product of the nest structure. A recent study using R. speratus revealed that the royal chambers in termite nests were hypoxic (low in O₂; approximately 15%) and hypercapnic (high in CO₂; approximately 4%) [44]. Oxidative stress is influenced by the level of O₂ present, shown by the fact that increases in the level of O₂ in the atmosphere that the insects are breathing lead to enhanced rates of oxidative damage and reduced longevity [102]. Atmospheric O₂ can be toxic and the levels must be carefully regulated to avoid oxidative stress [103]. Indeed, the discontinuous gas-exchange cycle is known as a respiratory adaptation to avoid oxygen toxicity in insects [104] (also in a termite Zootermopsis nevadensis [105]). Hypoxic nests thus may protect reproductives from oxidative stress through interacting with their antioxidant system.

Termite reproductives exhibit higher-level reproductive activities, survival and expression of antioxidants and vitellogenin under hypoxia compared to normoxia, suggestive of hypoxic adaptation [44,45]. Intriguingly, during physiological adaptation to hypoxia, the development of anaerobic energy-producing systems operative in the conditions of low oxidative stress may extend the lifespans of termite reproductives. Glycolysis, which is a form of anaerobic metabolism, is considered to produce fewer ROS than mitochondrial aerobic metabolism; large amounts of glucose are required to create energy [106,107]. Termite queens exhibit glucose signalling [100]; workers may deliver glucose to reproductives via trophallaxis. The long lifespan of queens may be attributable in part to optimization and adaptation of their physiological responses to the hypoxia of their closed chambers, but this still needs to be further tested.

An evolutionary linkage between adaptation to hypoxia and longevity may also be in play in other long-lived animals. For example, the naked mole rat [108], parasitic nematodes [109] and the ocean quahog [110], adapted to protected habitats that are underground, inside the body of the host, and in muddy bottom sediments, respectively. Although adaptation to hypoxia may be physiologically useful, this does not necessarily delay ageing. When exploring the relationship between longevity and metabolic adaptation to hypoxia in termite reproductives, it is essential to define the pathways of energy metabolism, and the metabolic burdens imposed by (for example) oxidative stress during metabolism.

5. Perspectives

The ecological and physiological studies discussed here infer that the striking reproductive activity and lifespan of termite reproductives evolved under selection pressures that accompanied the development of eusociality. However, it remains unclear whether factors associated with such evolution increased termite longevity. The life-history traits (the reproductive outputs and longevities) of various termite species featuring different levels of social defence should be further compared. For instance, there are wood-dwelling termites in which all castes are protected against predators as workers do not forage, and in which all castes also stay in the nest and probably experience hypoxia. These direct comparisons would yield important insights. Probably, it is not a single factor but several traits that explain potential differences between wood-dwelling and foraging termite species.

Since termite reproductives have extraordinary longevity, it is difficult to track their entire life, and there is not so much information on the lifespan of termites [5,11–14]. Therefore, reliable and valid biomarkers of chronological and biological age are essential for future advanced studies of termites. This would allow us to take into account the effects of age, which is one of the difficulties in termite research. Moreover, molecular studies of termites have identified several genes that may be involved in longevity assurance, but their functions remain unclear. Since it is necessary to carefully assess how termite reproductives achieve extended lifespan, further studies are required to evaluate the biological function of these genes in the longevity of termites using genetic tools, such as RNA interference [111,112] and transgenic systems [113].

The quest for understanding the proximate mechanisms of extraordinary longevity in termite reproductives is only just beginning. We are now at the dawn of a big paradigm shift in the ageing study of termites, and have a great opportunity to uncover the unexplored longevity mechanisms that cannot be reached by studies using intrinsically short-lived model organisms. This study raises the possibility that termite reproductives achieve longevity without sacrificing reproduction, not only through efficient antioxidant systems, transposon defence and downregulation of growth signalling, but also through metabolic adaptation to hypoxic environment. These unexpected physiological and molecular pathways explaining the remarkable lifespan extension of termite reproductives may be identified in the future.

Data accessibility

This review has no additional data.

Authors' contributions

E.T., M.T. and K.M. wrote the paper together.

Competing interests

We declare that we have no competing interests.

Funding

This work was supported by the Japan Society for the Promotion of Science (Kiban Kenkyu S: 18H05268) to K.M.