Telomerase activity can mediate the effects of growth on telomeres during post-natal development in a wild bird

Telomeres are terminal DNA–protein complexes placed at end of linear chromosomes that play a key role in preventing genome degradation (O'Sullivan and Karlseder, 2010). Although there are some exceptions, telomeres decrease with age in the majority of somatic tissues in many endothermic animals, especially during postnatal development when most of the somatic growth takes place (Monaghan and Ozanne, 2018), and evidence indicates that individuals with shorter telomeres have reduced phenotypic quality and lifespan (Blackburn et al., 2015; Heidinger et al., 2012; Whittemore et al., 2019). Because of its role in ageing and survival, the study of telomere dynamics is important to diverse biological disciplines and telomere length and attrition are now recognised as potential factors underpinning life-history trade-offs (reviewed in Monaghan, 2010; Young, 2018). Identifying the mechanisms governing the loss and/or restoration of telomere length in vivo is therefore of great interest, because it may contribute to better understand the costs and constraints at play in life-history evolution and ageing trajectories in wild animal populations.

Most previous studies that have investigated the mechanisms shaping telomere dynamics in vivo have focused on factors that can potentially lead to telomeres loss. In this regard, it is often assumed that telomere attrition is a cost associated with rapid somatic growth (Monaghan and Ozanne, 2018). Although the overproduction of reactive oxygen species (ROS) during growth may be involved in the loss of telomere length (see e.g. Reichert and Stier, 2017; Von Zglinicki, 2002), the mechanism linking growth and telomere dynamics is still not well understood, especially in non-model organisms (Monaghan and Ozanne, 2018). Notably, telomere length and dynamics depend not only on shortening but also on maintenance. Although several recombination-based telomere restoration mechanisms have been described (e.g. alternative lengthening mechanisms), at least as far as is currently known in vertebrates, telomeres are often restored by the action of telomerase, a ribonucleoprotein that counteracts telomere shortening by adding repetitive DNA sequences to the end of chromosomes (Blackburn, 2005). Despite its important role in telomere restoration, the study of telomerase dynamics has largely been overlooked out of the biomedical disciplines (but see Haussmann et al., 2004, 2007). This is probably because most studies on telomere length and dynamics in non-model vertebrates often use blood cells as target tissue (e.g. red blood cells or leucocytes). Telomerase activity is often present in stem cells, gonads and cancer cells through adulthood but previous data from biomedical studies suggest that telomerase is downregulated in most somatic cells once embryonic development is complete (Blackburn, 2005; Criscuolo et al., 2018; Forsyth et al., 2002). However, telomerase dynamics vary among and within-species and recent evidence has challenged this view by showing that telomerase may be active even after birth in some taxa. For instance, we have recently shown that in the yellow-legged gull (Larus michahellis), hatchlings show detectable and presumably functional levels of telomerase activity in the red blood cells (RBCs) (Noguera et al., 2020), supporting previous data on other long-lived bird species (Haussmann et al., 2004, 2007). While these studies suggest that long-lived birds may maintain some telomerase activity during early development there are no longitudinal studies in vivo that have investigated the links between age, somatic growth, telomerase activity and telomere length.

Here, we have performed such a longitudinal study and measured telomerase activity and telomere length at different time points between hatching and fledging (i.e. 0, 8 and 30 days of age) in a free-living yellow-legged gull population. We first examined whether telomerase activity is present during the full postnatal period and relates to telomere dynamics. Additionally, because a faster growth is often related to telomere attrition, we further used structural equation models (SEMs) to explore the causal conditional relationship between growth, telomerase activity and telomere length.

We carried out the study in a large breeding colony of yellow-legged gulls (Larus michahellis Naumann 1840) at Parque Nacional de las Islas Atlánticas (Sálvora Island, Spain). In this species, females normally lay the eggs at 1- to 3-day intervals until the (modal) three-egg clutch is completed. For the study, we surveyed three breeding subcolonies (>50 nests each) once daily during egg-laying and marked nests with numbered sticks. We visited each nest every day until clutch completion to mark the eggs of 52 three-egg clutches and register egg-laying order. After clutch completion, we randomly selected one of the two senior (A or B) eggs in each nest and measured it (±0.01 mm) to calculate its volume, as previously described (Noguera et al., 2017). We focused on one (senior) egg per clutch to reduce the number of times each experimental nest had to be visited during the study period and so avoiding any adverse effect on chicks’ physiology and telomere dynamics as a result of human-induced stress responses (Noguera et al., 2017; Noguera and Velando, 2019a,b). Moreover, because senior eggs have higher hatching success and postnatal survival than their younger (C) sibling (Noguera and Velando, 2020), by focusing on the senior (A or B) eggs we also maximised our final sample sizes (Noguera and Velando, 2020).

At hatching, we blood sampled, weighed (±1 g) and marked all experimental chicks (i.e. A or B chicks) with numbered leg flags for their identification, and because gull chicks are very mobile, we also fenced each nest to keep chicks in their territory (see Kim et al., 2013 for further details). When the experimental chicks were 8 days of age, we weighed them again (±1 g) and collected a second blood sample (N=46). Moreover, to avoid the stress of reduced territory size, we removed the enclosures around the nests and then marked them with a numbered plastic ring with an individual two-digit combination to facilitate their long-term identification. At 30 days of age, when chicks were fully grown and near fledging, we came back to the colony and searched for them around their marked nests. For all marked chicks we found, we took a third blood sample and weighed them (±5 g). During the experiment, blood samples (≈90 µl) were always collected with heparinised capillary tubes and kept refrigerated on ice blocks until they were centrifuged (within 3 h after collection). After centrifugation, we carefully removed the plasma and buffy coat layer (i.e. white blood cells and platelets) by aspiration and stored the RBC fraction in liquid nitrogen for later laboratory analyses.

The study complied with the standards of animal experimentation and animal welfare established under current Spanish law (RD53/2013), and permissions were granted by the authorities of Parque Nacional de las Islas Atlánticas and approved by the Xunta de Galicia review board (45/RX97704 and 263/RX583146).

Telomerase activity in RBCs was measured using the quantitative-telomeric repeat amplification protocol (Q-TRAP) assay described by Herbert et al. (2006), with some minor modifications described for gull samples (Noguera et al., 2020). For the assay, RBCs cells were lysed in CHAPS lysis buffer and incubated on ice for 30 min. The lysate was centrifuged at 12,000 g for 30 min at 4°C, and the supernatant was collected. Protein concentration in the supernatant was then measured using Pierce BCA Protein Assay Kit (Thermo Fisher Scientific) according to the manufacturer's protocol. The Q-TRAP was optimized using the PCR reagent PERFECTA^® SYBR^® GREEN FASTMIX (Quanta Bioscience, Gaithersburg, MD), containing 2 µl extract, 12.5 µl 1× SYBR Green Master Mix, 0.1 µg TS primer (5′-AATCCGTCGAGCAGAGTT-3′), 0.1 µg ACX primer (5′-GCGCGGCTTACCCTTACCCTTACCCTAACC-3′) and adjusted to 25 µl using sterile H₂O. Samples were then run on a StepOnePlus qPCR (Applied Biosystems) and each plate included measurements of telomerase-positive and telomerase-negative controls. The average efficiency of the assay was 88.3% (range 83.5–93.1) and the relative telomerase activity (RTA) of each sample was quantified based on the linear equation of the standard curve derived from a serially diluted positive control sample (R²>0.99 in all cases). All samples were run in triplicate and the repeatability [intraclass correlation coefficient (ICC); Lessells and Boag, 1987] of the assay based on triplicates was 0.9 and the intra- and inter-plate coefficient of variation was 1.17 and 1.21%, respectively.

Telomere length was measured in RBC DNA samples using the same qPCR device as described above and following a previously established protocol for birds (Criscuolo et al., 2009) but with some minor modifications described for yellow-legged gull samples (Kim and Velando, 2015). The qPCR method ‘normalizes’ the quantity of telomere product (T) to a single-copy gene (S) to provide a mean telomere length for cell population (T/S ratio). The GAPDH gene was used as a single-copy gene in all analyses (Kim and Velando, 2015). The efficiency of each amplicon (TEL and GAPDH) was estimated from the slopes of the amplification curves for each qPCR reaction using LinRegPCR software (TEL: range 88.4–89.8%; GAPDH: range 89.8–91.0%) (Ruijter et al., 2009). All DNA samples were run in triplicate and the mean intra- and inter-plate variation of the C_t values was 1.05 and 1.09% for the telomere reactions and 0.49 and 1.84% for the GAPDH reactions, respectively. The within-individual ICC of T/S ratios was 0.8.

Finally, gull chicks were also sexed following the methodology and primer sequences described by (Fridolfsson and Ellegren, 1999). The method is based on polymerase chain reaction (PCR) to amplify part of the W-linked avian CHD gene (CHD-W) in females, and its non-W-linked homologue (CHD-Z) in both sexes. The DNA products were run on a 2% agarose gel and stained with Greensafe Premium (NZYtech, Portugal).

First, we investigated age-related changes in telomerase activity and telomere length using linear mixed effect models (LMM). These two models included chick age, sex, order and the two-way interactions between age and chick sex and order as fixed factors, and chick identity (ID) as a random effect. Because egg quality and the location of the nests within the breeding colony may potentially influence chick telomerase activity and telomere length (e.g. via differences in egg quality and/or stress exposition; Noguera et al., 2020; Noguera and Velando, 2019b), in all the above models we also accounted for egg volume and sub-colony of origin (three-level fixed factor).

Because any change in telomere length during postnatal development may be related to age-related changes in growth and telomerase activity, we further used SEMs to investigate the causal relationships between these variables. SEM extends the basic correlation approach to path analysis by directly testing the goodness of fit of the model to the data, calculates correlation coefficients, and separates total effects into direct and indirect effects. Specifically, we started building an initial (saturated) model that included all biologically meaningful connections among traits according to previous literature and our knowledge. Because the initial model did not provide a good fit to the data (χ²<0.001), the SEM was simplified and evaluated again (Byrne, 2010). We assessed the suitability of the final model using several comparative fit indexes [i.e. the χ²-test, comparative fit index (CFI) and the root square mean error of approximation (RMSEA) (Byrne, 2010)], and calculated the standardised regression coefficients between pair of variables and the R² of all response variables.

Before running the analyses, telomerase activity and telomere length were natural logarithm(+1)-transformed to improve data distribution. The constant was introduced to avoid negative values and thus ease of interpretation. Note that differences in sample sizes in some analyses are due to the death or loss of some chicks or because samples failed to amplify (see Table S1 for a detailed description of sample sizes). For LMMs, we report results for full models after removing non-significant interactions (Engqvist, 2005). All analyses were conducted in IBM SPSS statistics v.24 using Satterthwaite's degrees of freedom, and Amos 18 for Windows (IBM SPSS Inc., Chicago, USA). The significance level was always set at α=0.05 and all statistical tests were two-tailed.

Gull chicks RBCs showed telomerase activity during the entire postnatal developmental period but activity levels decreased with age, especially between 8 and 30 days of age (Table 1; Fig. 1A and Fig. S1). Thus, although it is generally assumed that telomerase is downregulated in all postmitotic somatic tissues in long-lived vertebrates, our yellow-legged gull chicks showed detectable levels of telomerase activity, which supports previous studies in this and other long-lived birds (Haussmann et al., 2004; Noguera et al., 2020). Interestingly, we found that RBC telomere length also changed with age, with chicks experiencing a significant reduction in telomere length from 8 to 30 days of age (Table 1, Fig. 1B and Fig. S1). Telomere length also negatively covaried with egg volume (Table 1) but the other variables included in the models were non-significant (all P>0.092; Table 1). When added to the model, we found that telomerase activity was positively related to telomere length, especially at hatching (telomerase activity: F_1,108=5.944, P=0.016; chick age×telomerase activity: F_1,108=4.781, P=0.010; Fig. S2). Because telomerase is highly upregulated during embryonic development, it is likely that soon after hatching, the higher covariation between these two traits may reflect long-lasting changes that occurred during earlier stages of embryonic development. Importantly, although at a lower level, telomerase activity was present until near fledging in RBCs (i.e. 30 days), suggesting that telomerase may still play a repair (i.e. lengthening) function during the postnatal life in some somatic tissues. Indeed, this may explain why telomere studies commonly report telomere lengthening in some individuals (e.g. Bateson and Nettle, 2017; Tricola et al., 2018 and references therein), an effect that cannot completely be attributed to measurement errors (Bateson and Nettle, 2017). Telomere length was also found to be repeatable within individuals based on measures of chicks sampled at hatching, day 8 and 30 of age (r=0.21, P=0.038), but that was not the case for telomerase activity (r=0.01, P=0.460).

Body mass growth was not directly related to telomere length change in our initial SEM (standardized coefficient: −0.16, P=0.375; Fig. 2A). After deleting this non-significant pathway, the final SEM predicting telomere length change between hatching and fledging provided an adequate fit to the data (χ²=0.337; CFI=0.99, RMSEA<0.001; Fig. 2A). This model showed that body mass gain (i.e. growth) had an indirect effect on telomere length change (standardize coefficient=−0.174), via an effect on telomerase activity (standardize coefficient=−0.4). Thus, the more body mass the chicks gained during postnatal development, the greater the reduction in telomerase activity (Fig. 2A) which in turn, related to a faster loss of telomere length (Fig. 2C). Hence, our results suggest that the change in body mass and telomere length between hatching and fledging was (indirectly) related. This is in agreement with previous studies in other vertebrates and adds further support to the hypothesis that telomeres play a significant role in driving the growth–lifespan trade-off (reviewed in Monaghan and Ozanne, 2018). Our results, however, suggest that the effects of growth on telomere length are indirectly mediated by the decline in telomerase activity, and the direct effect of growth on telomere length (if any) was minor. In this regard, previous studies have shown that yellow-legged gull chicks endure increased oxidative damage (e.g. DNA damage) during their postnatal development as a result of their fast growth (Noguera et al., 2011a,b) and recent studies have demonstrated striking inhibitory effects of oxidative damage on telomerase activity (see e.g. Ahmed and Lingner, 2018). In our experiment, telomerase activity declined as body mass growth increased; an effect that may have favoured a higher loss of telomeres in the chicks that showed a faster growth rate. The potential inhibitory effect of oxidative stress on telomerase activity as well as its effects on telomere dynamics still needs to be confirmed in future experimental studies. Irrespective of the mechanism, the suppressive effects of fast somatic growth on telomerase activity during postnatal development may potentially explain why in several bird species telomeres are mostly eroded during early postnatal life (e.g. Boonekamp et al., 2014; Heidinger et al., 2012; Noguera et al., 2016), a period where young birds are also exposed to a variety of environmental stressors, and increased oxidative stress levels.

Although it is unclear why some bird species maintain telomerase activity in some somatic cell lines like RBCs, telomerase may confer some advantages to this cell type. In contrast to mammals, RBCs in birds are fully functional cells with a nucleus, active transcription/translation (Velando et al., 2019; Watson et al., 2017) and functional mitochondria (Stier et al., 2013). Because mitochondrial biogenesis is under the control of nuclear genes (Passos et al., 2007), it is plausible that RBC telomerase may help to prevent a premature loss of important cellular functions in bird such as cellular respiration capacity and for extension, stress resistance of RBC metabolism (Stier et al., 2013). Future work should experimentally assess the adaptive function of maintaining telomerase activity in RBCs and other critical tissues/organs (e.g. liver, intestines, etc.) in long-lived birds and the costs and benefits associated to this regulation.

In conclusion, this study indicates that telomerase activity is not completely downregulated during postnatal life and follows a similar dynamic as that observed in telomere length. Moreover, we provide evidence suggesting that somatic growth may lead to telomere shortening by its inhibitory effect on telomerase activity. Overall, our study highlights the potential role of telomerase in mediating telomere dynamics during postnatal development and offers a mechanism through which fast growth rates may negatively affect telomere dynamics and potentially life-history trajectories (Monaghan, 2010; Young, 2018). More longitudinal studies are needed to confirm whether telomerase activity is maintained in other somatic tissues rather than RBCs and whether oxidative damage accelerates the loss of telomerase activity in vivo.

We are grateful to the staff at the Atlantic Islands of Galicia National Park, especially to X. Arca, P. Mallo and R. Castiñeira and to two anonymous reviewers for their constructive comments on earlier versions of the manuscript. We also thank A. da Silva for helping with the telomerase analyses, and N. Alvarez-Quintero and B. Otero for their assistance during the fieldwork.