INTRODUCTION
Telomeres are highly conserved, non-coding DNA sequences that form protective caps at the end of eukaryotic chromosomes (Blackburn,1991, Blackburn & Epel 2012). In the absence of telomerase (a reverse transcriptase that adds telomeric repeats de novo after each cell division), telomeres shorten with each round of cell division (Harley, Futcher, & Greider, 1990). When a critical length is lost, telomeres become dysfunctional and cells enter a state of replicative senescence (Hornsby, 2003; Verdun & Karlseder, 2007, see also Victorelli & Passos, (2017) for length independent damage to telomeres triggering cell senescence). The accumulation of senescent cells is known to contribute to age-related declines in tissue and organ function (Wong et al., 2003). Accordingly, within species, relatively short telomeres and accelerated rates of telomere shortening have been associated with fitness costs, predominantly via reduced lifespan, at the individual level (Monaghan, 2010; Eastwood et al., 2019; Boonekamp, Mulder, Salomons, Dijkstra, & Verhulst, 2014, Wilbourn et al., 2018). Given the relationship with lifespan, telomeres have been widely used as biomarkers of biological age (Jylhava, Pedersen, & Hagg, 2017). This has also led to a recent focus in understanding how inter-individual variation in telomere length arises. Early life telomere length is partly determined by genetic factors (Olsson et al., 2011; Dugdale & Richardson 2018), however, accumulating evidence suggests that environmental cues also impact telomere length dynamics across the life course (Kotrschal, Ilmonen, Penn, 2007; Vedder, Verhulst, Zuidersma, & Bouwhuis, 2018; Dupoue et al., 2017; Boonekamp, et al., 2014). Information on the relationship between environmental and telomere length variation is steadily increasing, yet surprisingly little is known about whether other DNA modifications mirror telomere dynamics. Insight on how telomere length can vary in concert with other genetic traits can expand our understanding of DNA based biological markers of aging, and of physiological reactions to environmental change.
DNA methylation (DNAm) is an epigenetic mechanism that appears to be an important component of telomere length regulation (Blasco, 2007), and is also considered a promising biological clock (Horvath & Raj 2018). Yet, the relationship between DNAm and telomere length has not yet been explored in an ecological context in any species. DNAm usually refers to the addition of a methyl group to a cytosine base at a CG dinucleotide (a ‘CpG’ site) on the DNA sequence, but can occur at other sites in different taxa (Anglers et al., 2010). When DNAm occurs at a CpG site close to a gene regulatory region, it can modulate phenotypic variation through its effects on gene expression. Evidence suggests that DNAm could be involved in two, key telomere regulatory processes; those involving telomerase (Buxton et al., 2014), and ‘alternative mechanisms’ relying on homologous recombination between telomeric sequences such as alternative lengthening of telomeres (Gonzalo et al., 2006). Indeed, while telomeres themselves do not contain CpG sites, decreases in global and subtelomeric DNAm are known to be concomitant with increased homologous recombination between telomeric sequences and dramatically elongated telomeres in mouse cells (Gonzalo et al., 2006). Additionally, a negative relationship between telomere length and genome-wide (Lee et al., 2019) and gene-specific (Lee et al., 2019; Buxton et al., 2014) DNAm has been described in humans (however, this relationship may be complicated as a positive relationship has been detected in a different correlative study (Dong et al., 2018)). Evidence also suggests that short telomeres have specific epigenetic marks that may facilitate their preferential elongation (Hemann, Strong, Hao, & Greider, 2001; de Lange, 2005). These studies suggest the existence of a direct relationship between DNAm and telomere length. However, there is also plenty of scope for indirect relationships between DNAm and telomere length to occur given that they are associated with a range of the same biological and ecological factors (Figure 1) and are also both independently identified as molecular measures of age and ageing (Lu et al., 2019; Banszerus, Vetter, Salewsky, König & Demuth, 2019).
To date, surprisingly little work has explored the potential links between telomere length and DNAm, although a few studies have indicated some conceptual connections (Horvath et al 2018; Blasco, 2007; Figure 1). DNAm and telomere length are highly responsive to environmental cues (Feil & Fraga, 2012; Monaghan, 2014), particularly during early life (Watson, Bolton & Monaghan, 2019; Boonekamp et al., 2014). For example, studies have detected associations between DNAm or telomere length and clutch/brood size (Noguera & Velando, 2020; Jimeno, Hau, Gomez-Diaz, & Verhulst, 2019; Sheldon, Schrey, Ragsdale & Griffith 2018; Nettle et al., 2016; Reichert et al., 2014; Costanzo et al., 2016; Boonekamp et al., 2014); ambient temperature (Stier, Metcalfe & Monaghan, 2020; Sheldon, Schrey, Hurley, & Griffith, 2020; Yan et al., 2015); and body size/growth rate (Young et al., 2017; Vedder et al., 2018). It is therefore important to account for these influences when testing the association between DNAm and telomere length. The present study is the first examination, to date, of the relationship between DNAm and telomere length dynamics in a species in the wild in which individuals are exposed to natural variation in early life conditions. Specifically, we collected longitudinal measures of telomere length (using qPCR) and genome-wide levels of DNAm (using methylation sensitive- amplification fragment length polymorphisms (MS-AFLP)) across the early life of wild zebra finches (Taeniopygia guttata ). Although MS-AFLP analyses do not provide inference on the functional, gene regulatory consequences of DNAm differences, they do provide a useful tool to compare the relationship between telomere length dynamics and changes in the percentage of a consistent subset of CpG loci that are methylated among individuals (Schrey et al., 2013). Our study aims to provide an initial exploration into the relationship between early life DNAm changes and telomere length changes in wild animals as well as potential associations with ambient temperature and life history effects. This can establish a base for future research to link and/or disentangle the relationship between early life conditions, DNA based biomarkers of age, and individual fitness parameters.
In addition to considering the links between telomere length and DNAm, our study is also of value in providing the first investigation of telomere length dynamics in wild zebra finches Taenopygia guttata . Characterisations of telomere dynamics in the wild have become increasingly common for a wide range of bird species (Supplementary Table 1a and 1b). In contrast, laboratory studies on birds are largely performed using the zebra finch, in which paradoxically, telomere dynamics have yet to be explored in the wild (Supplementary Table 1a). Studies of ecologically relevant populations of wild zebra finches are thus of value in helping to contextualize and interpret the controlled laboratory tests that have made such a significant contribution to our understanding of telomere biology in birds.