Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2025 Oct 13;34(22):e70147. doi: 10.1111/mec.70147

Sex‐Chromosome‐Dependent Ageing in Female Heterogametic Methylomes

Marianthi Tangili 1,✉, Joanna Sudyka 1,2, Fabricio Furni 1, Per J Palsbøll 1,3, Simon Verhulst 1,✉
PMCID: PMC12617039  PMID: 41078303

ABSTRACT

Recent research in humans and both model and non‐model animals has shown that DNA methylation (DNAm), an epigenetic modification, is one of the mechanisms underlying the ageing process. DNAm‐based indices predict mortality and provide valuable insights into biological ageing mechanisms. Although sex‐dependent differences in lifespan are ubiquitous and sex chromosomes are thought to play an important role in sex‐specific ageing, they have been largely ignored in epigenetic ageing studies. We characterised the genome‐wide distribution of age‐related CpG (Cytosine‐phosphate‐Guanine) sites from longitudinal samples in two avian species (zebra finch and jackdaw), including for the first time the avian sex chromosomes (Z and the female‐specific, haploid W). In both species, we find a small fraction of the CpG sites to show age‐related changes in DNAm with the majority of them being located on the haploid, female‐specific W chromosome, where DNAm levels predominantly decrease with age. Age‐related CpG sites were over‐represented on the zebra finch but under‐represented on the jackdaw Z chromosome. Our results highlight distinct age‐related changes in sex chromosome DNAm compared to the rest of the genome in two avian species, suggesting this previously understudied feature of sex chromosomes may be instrumental in sex‐dependent ageing. Moreover, studying the DNAm of sex chromosomes might be particularly useful in ageing research, facilitating the identification of shared (sex‐dependent) age‐related pathways and processes between phylogenetically diverse organisms.

Keywords: avian sex chromosomes, DNA methylation, female heterogametic, sex‐dependent ageing

1. Introduction

The underlying causes of ageing, the decline in biological functioning over time, remain an enigma. Phenotypic and molecular biomarkers that serve as proxies of biological age (Xia et al. 2017) have been instrumental in unravelling ageing mechanisms. In this context, indices based on age‐dependent epigenetic marks have shown an unrivalled capacity to predict chronological and biological age (Pal and Tyler 2016). Such indices are mostly based on age‐related changes in DNA methylation (DNAm) at CpG sites: sites in the genome where a 5′ cytosine is followed by a guanine (Cytosine‐phosphate‐Guanine, Moore et al. 2013). Indices of chronological age based upon the degree of DNAm at certain CpG loci (‘epigenetic clocks') have opened novel, pivotal windows into the complex landscape of age‐associated phenotypes. Numerous epigenetic clocks have been developed in recent years in humans (Horvath 2013; Levine et al. 2018; Marioni et al. 2015), as well as in model (Han et al. 2020) and non‐model animals (Lu et al. 2023; Tangili et al. 2023), and even plants (Gardner et al. 2023). DNAm contributes to the regulation of gene expression (Moore et al. 2013) and is known as one of the most stable epigenetic modifications, possibly due to its important roles in embryonic development, cellular identity and genomic stability (Bird 2002). However, age‐related changes in DNAm still occur but are restricted to a relatively small subset of CpG sites across the genome (Horvath 2013). The study of changes in the level of DNAm with age across a broad range of species holds the potential to identify fundamental sites, sequences and processes related to ageing conserved across diverse taxa. Casting the net widely will enable identifying ageing pathways shared among species, or classes, that likely are fundamental to ageing (Moqri et al. 2024).

Sex differences in age‐related pathology and lifespan are ubiquitous across the animal kingdom, where the homogametic sex in female heterogametic species lives on average 7.1% longer (Xirocostas et al. 2020), suggesting a link between sex chromosomes and ageing. Despite the link between sex and age, most epigenetic ageing studies have so far routinely ignored the sex chromosomes, with a few notable exceptions only in humans (G. Li et al. 2022; Lund et al. 2020). Employing a novel, robust pipeline, we identified age‐related CpG (AR‐CpG) sites where the degree of DNAm correlates with age among blood samples from captive zebra finches (Taeniopygia castanotis) and wild jackdaws ( Corvus monedula ). We employed longitudinal data, that is two samples collected from each individual during its lifespan, to characterise the distribution of AR‐CpG sites across the genome. We included, for the first time, the avian sex chromosomes (Z and the female‐specific haploid W) to assess if the observed distribution of AR‐CpG sites across chromosomes deviates from null expectations. Our expectation was that if sex chromosomes play a key role in ageing, this would be evident in the prevalence of AR‐CpG sites on the sex chromosomes.

2. Materials and Methods

2.1. Sample Origin

Zebra finch blood samples were collected during a long‐term experiment where birds were housed in outdoor aviaries (320 × 150 × 210 cm), each containing single‐sex groups of 18–24 adults (Briga et al. 2017). Twenty longitudinal samples were collected between June 2008 and December 2014 from ten known‐age individuals (six males and four females) sampled twice during their lifetime at an average interval of 1470 days between the two sampling points. Jackdaw blood samples were collected from individuals of a free‐ranging population breeding in nest boxes south of Groningen, the Netherlands (53.1708° N, 6.6064° E). We analysed 22 longitudinal blood samples collected during 2007–2021 from 11 known‐age adults (five males and six females) with an average sampling interval of 2429 days. For both species, the precise age of sampled individuals was known because individuals were followed since birth (Table S1). Samples were collected in EDTA buffer and stored at −80°C in glycerol buffer.

2.2. Whole Genome Bisulphite Sequencing

We extracted total‐cell DNA using the innuPREP DNA Mini Kit (Analytik Jena GmbH+Co) from 3 μL (nucleated) red blood cells according to the manufacturer's protocol. Whole genome sequencing was performed by The Hospital for Sick Children (Toronto, Canada), where paired‐end Illumina next‐generation sequencing (150 bp) was carried out on either an Illumina HiSeqX (12 zebra finch samples) or an Illumina NovaSeq sequencer (eight zebra finch samples and 22 jackdaw samples). Libraries were prepared using the Swift Biosciences Inc. Accel NGS Methyl Seq kit (part no. 30024 and 30096), and the DNA was bisulphite converted using the EZ‐96 DNA Methylation‐Gold kit from Zymo Research Inc. (part no. D5005) as per the manufacturer's protocol, and libraries were subsequently PCR amplified.

2.3. Bioinformatic Processing

Sequences were trimmed using Trim Galore! v. 0.6.10 (Krueger et al. 2023) in paired‐end mode. Visual quality controls of the data were carried out before and after trimming using FastQC v. 0.11.9 (Andrews 2010) and MultiQC v. 1.11.14 (Ewels et al. 2016). Because the Swift Biosciences Inc. Accel NGS Methyl Seq kit was used for library preparation, the first ~10 bp showed extreme biases in sequence composition and M‐bias, so after checking the M‐bias plots, the first 10 base pairs (bp) were further trimmed from each sequence.

Alignments were performed using Bismark v. 0.14.433 (Krueger and Andrews 2011) using the Bowtie 2 v. 2.4.5 alignment algorithm (Langmead and Salzberg 2012) for both in silico bisulphite conversion of the reference genomes and alignments. For the zebra finch, trimmed reads were aligned against the in silico bisulphite‐converted zebra finch ( Taeniopygia guttata ) reference genome (GCA_003957565.4, Rhie et al. 2021), and the average mapping efficiency was 64.5% (SD: 3.22). Because the jackdaw reference genome did not contain an assembled W chromosome, the jackdaw bisulphite sequencing data were aligned to an in silico bisulphite converted Hawaiian crow genome ( Corvus hawaiiensis , GCA_020740725.1, Rhie et al. 2021) – the closest relative species with an assembled W chromosome; the average mapping efficiency was 64.7% (SD: 0.98).

2.4. Functional Annotation of AR‐CpG Sites

The GFT‐formatted annotation files for the zebra finch (GCF_003957565.2) and the Hawaiian crow (GCF_020740725.1) were converted into BED12 format using the University of California, Santa Cruz (UCSC) utilities gtfToGenePred and genePredToBed (available at https://hgdownload.soe.ucsc.edu/downloads.html#utilities_downloads). AR‐CpG sites were annotated using the tool annotateWithGeneParts from the R 4.1.2 package genomation 1.4.1 (Akalin et al. 2015). This tool hierarchically classifies the sites into pre‐defined functional regions, that is promoter, exon, intron or intergenic, herein referred to as annotation categories. The predefined annotation categories were based on the annotation information present in the BED12 files accessed with the genomation tool readTranscriptFeatures. Annotations were performed in a hierarchical manner (promoter > exon). Subsequently, a customised R script was employed to integrate the annotation results of AR‐CpG sites with their respective annotation category information.

2.5. Gene Ontology Enrichment Analysis

The gene ontology (GO) enrichment analysis for the genes in or near which we identified AR‐CpG sites was conducted using the ClueGo plug‐in v. 2.5.10 (Bindea et al. 2009) for Cytoscape v. 3.10.2 (Shannon et al. 2003) and was conditioned on including at minimum three associated genes. We applied a two‐sided enrichment‐depletion test employing the Benjamini‐Hochberg correction for multiple testing (Benjamini and Hochberg 1995), and GO terms were considered significantly enriched if they had a corrected p‐value of 0.05 or less.

2.6. Statistical Assessment

We conducted χ 2 tests using the R package janitor v.2.2.0 (Firke 2023) to compare the observed and expected number of AR‐CpG sites in each species and applied exact binomial tests using the R package stats v.4.1.2 (R Core Team 2023) to compare the observed and expected number of AR‐CpG sites for each chromosome. The expected number of AR‐CpG sites for each chromosome was inferred as the product of the observed number of CpG sites on each chromosome and the species‐specific frequency of AR‐CpG sites for the whole genome, which was 0.008% for the zebra finch and 0.006% for the jackdaw. We applied a Benjamini‐Hochberg correction to control for the false discovery rate to account for multiple hypothesis testing (Benjamini and Hochberg 1995).

We assessed the observed and expected number of AR‐CpG sites per annotation category using a χ 2 test. The proportion of expected AR‐CpG sites per annotation category was estimated as the proportion of all CpG sites that passed the initial filters (Steps 1, 2 and 3 in our pipeline, Results) located in each annotation category. For all χ 2 tests, we used an exact binomial test as a post hoc test to assess observed vs. expected occurrences. All statistical tests were conducted using the CRAN R (4.1.2).

To assess whether age‐related CpG sites were non‐randomly distributed across genomic regions, we performed an enrichment analysis for both species (separately) based on the genomic coverage of each region. We first calculated the total length (in base pairs) of each region type per chromosome type (autosomes, W and Z). The observed number of AR‐CpGs in each region was then compared to the expected number based on that region's relative genomic coverage. Enrichment or depletion was tested using Fisher's exact test by comparing observed and expected CpG counts in the focal region versus all other regions. The results were expressed as odds ratios with p‐values, with enrichment defined as an odds ratio > 1 and depletion as < 1.

The terms ‘over‐represented’ and ‘underrepresented’ are used to refer to sites that are more or less numerous than expected by chance on either specific chromosomes or in annotation categories if AR‐CpG sites were uniformly distributed across all CpG sites in the genome.

3. Results

Here we present the steps for the development of the pipeline employed to characterise the distribution of AR‐CpG sites (Figure 1a).

FIGURE 1.

FIGURE 1

Open in a new tab

Graphical representation of the pipeline for identification of AR‐CpG sites from longitudinal, whole‐genome bisulphite sequencing (WGBS) data for zebra finches and jackdaws (A). Number of CpG sites remaining (blue panels) and lost (red panels) after each step in the pipeline for the zebra finch (B) and jackdaw (C).

  • Step 1: We applied a minimum mean coverage per site of ≥ 20 for zebra finches and ≥ 25 for jackdaws. Since the optimal solution with respect to the coverage filtering has to do with the distribution of the initial coverage, these thresholds were chosen based on the distribution of the initial coverage, with median values of 6 for zebra finches and 7 for jackdaws. This step depends on the quality of sequencing data and can be tailored to the coverage distribution of a particular dataset.

  • Step 2: We then selected CpG sites with an average DNAm across all samples between 40% and 60% that are likely to have a higher absolute change in DNAm over time.

  • Step 3: We selected only the sites that were shared by the majority of the samples to increase the site detection probability across the whole population. In our datasets, that was ≥ 15 samples for autosomes (75% of the samples in the zebra finch, 68% in the jackdaw) and ≥ 6 samples for the W chromosome for the zebra finch (75% of female samples) and ≥ 8 for the jackdaw (67% of female samples), as it is present only in females in both species.

  • Step 4: For each sequenced CpG site, we generated a Pearson's correlation coefficient between DNAm at the focal sampling point and delta age (the difference between an individual's chronological age at the time of sampling and its average age calculated for both samplings, van de Pol and Wright 2009) to account for between‐ and within subject effects and correlated the Pearson's correlation values with the average sequencing coverage per CpG site over all samples. To include the sites with the highest correlations between DNAm at the focal sampling point and delta age (negative or positive, as we want to include methylation increases and decreases with age across all ranges of coverage), we selected the top 5% sites for every 20 units of coverage.

Since the mitochondrial DNA was not found to contain any AR‐CpG sites when included in the initial analysis, it was excluded from subsequent analyses.

This newly developed pipeline to identify AR‐CpG sites across the genome can be applied to (i) an array of different species, (ii) on DNAm data of varying qualities (in terms of sequencing coverage) and (iii) for which a limited number of individuals sampled longitudinally is sufficient. The robustness of our pipeline was confirmed by the high correlation between the chromosome‐wise number of AR‐CpG sites generated with different cutoff parameters (Figures S2, S3). This robustness stems from numerous factors. Firstly, we analysed whole‐genome DNAm data, meaning that we had the opportunity to investigate the change in DNAm of CpG sites throughout the entirety of all of the chromosomes, including the sex chromosomes. Moreover, as previously stated, the longitudinal nature of our data makes it ideal to identify within‐individual changes in DNAm with age. Another strength of our method is that we take extra steps to ensure that we not only identify CpG sites whose DNAm changes with age, but CpG sites that have extreme changes (in the same direction) in DNAm that are shared by the majority of the samples in the population. Additionally, we use the metric of DNAm correlated with Δage (the difference between an individual's chronological age at the time of sampling and its average age calculated for both samplings) instead of chronological age, meaning that individuals with larger timespans between sampling points have a higher influence on the identification of AR‐CpG sites. A larger time interval between samplings exerts more pronounced changes in DNAm patterns over time, thereby enhancing the detectability of AR‐CpG sites. Lastly, we observe comparable yet not identical patterns of AR‐CpG site location in the genomes of our two study species, which, is to be expected as they belong to the same class but are not very closely related phylogenetically. Because of all the above, we maintain confidence in the strength and robustness of our results.

In zebra finches, 0.008% of CpG sites (1,468/18,906,252) and in jackdaws 0.006% (1,218/21,835,349) showed age‐related DNA methylation changes (Figure 1). We did not detect AR‐CpG sites in the mitochondrial genome, but AR‐CpG sites were present on most nuclear chromosomes (Figure 2).

FIGURE 2.

FIGURE 2

Open in a new tab

Location of AR‐CpG sites across chromosomes and direction of change in DNAm. Position of each AR‐CpG site per chromosome for (A) the zebra finch and (B) the jackdaw. Counts of AR‐CpG sites on each chromosome whose level of DNAm increased (grey) or decreased (black) with age in the zebra finch (C) and the jackdaw (D). No AR‐CpG sites were detected on zebra finch chromosomes 15, 16, 33 and 36 and jackdaw chromosomes 23, 34, 36 and 38. The actual AR‐CpG site positions for both species can be found in (Tangili et al. 2024).

The majority of AR‐CpG sites were located on the W chromosome and occurred at significantly higher frequencies than expected by chance in both species (zebra finch: Figure 3A, 17% of all AR‐CpG sites, p < 0.0001, Table S2; jackdaw: Figure 3B, 15% of all AR‐CpG sites, p < 0.0001, Table S3). The AR‐CpG site density on the Z chromosome was elevated in the zebra finch (12% of all age‐related sites; χ 2 = 2,622, df = 36, p < 0.001, Figure 3) and depressed in the jackdaw (3% of all AR‐CpG sites; χ 2 = 2,243, df = 35, p < 0.001, Figure 3). The density of AR‐CpG sites also varied among the autosomal chromosomes. In the zebra finch, AR‐CpG sites were over‐represented on six chromosomes and under‐represented on 22 chromosomes (Table S2). The corresponding numbers in the jackdaw were six and 10, respectively (Table S3). In the zebra finch autosomes, DNAm level increased with age on 43% of AR‐CpG sites (n = 632; Figure 2C), which contrasts with the jackdaw, where DNAm increased with age on 24.5% of AR‐CpG sites (n = 298; Figure 2D).

FIGURE 3.

FIGURE 3

Open in a new tab

Number of AR‐CpG sites in relation to total CpG sites per chromosome. (A) Zebra finch and (B) jackdaw. The diameter of each data point is proportional to chromosome length in kilobase pairs (kbp). Black data points denote autosomal chromosomes with an AR‐CpG density that deviated significantly from the null expectations (dotted line).

AR‐CpG sites were predominantly located in intergenic regions in both species, followed by introns, promoters and exons in the zebra finch and introns, exons and promoters in the jackdaw (Figure 4A,D). The observed distribution of AR‐CpGs deviated significantly from null expectations (zebra finch: χ 2 = 5,655, df = 3, p < 0.001; jackdaw: χ 2 = 7.98, df = 3, p = 0.04). For the zebra finch, the AR‐CpG site distribution deviated significantly from null expectations also when considered separately for each annotation category (Table S4A), whereas in the jackdaw, only AR‐CpG sites in promoters were significantly more numerous than expected (Table S4B). AR‐CpG sites were absent from exons in the W chromosome in the jackdaw, while there were three AR‐CpGs located in exons in the zebra finch (Figure 4B,F and Figure S1).

FIGURE 4.

FIGURE 4

Open in a new tab

Functional annotation of AR‐CpG sites. Count of identified AR‐CpG sites per 1 million base pairs in each annotation category (A, E), raw proportion (not normalised) of identified AR‐CpG sites in each annotation category by chromosome type (B, F), and observed vs. expected in the whole genome (C, G) for the zebra finch and jackdaw, respectively. The count of AR‐CpG sites increasing or decreasing in the degree of DNAm in promoters per chromosome (ordered by decreasing length) for the zebra finch and jackdaw respectively (D, H).

Significant enrichment for AR‐CpG sites was observed in both species across annotation categories and chromosome types (Figure 5, Table S5). In the zebra finch, AR‐CpG sites were enriched in promoter regions and depleted in introns across all chromosome types (Figure 5A, Table S5A), while in the jackdaw, AR‐CpG sites were depleted in introns for all chromosome types (Figure 5B, Table S5B). Importantly, in both species, the W chromosome showed the strongest divergence, and AR‐CpG sites were enriched in intergenic regions, with the highest observed odds ratios (Figure 5, Table S5).

FIGURE 5.

FIGURE 5

Open in a new tab

Results of the enrichment analysis across genomic locations and chromosome types in the zebra finch (A) and the jackdaw (B). Shapes represent the enrichment status of each category, while the colour indicates the associated p‐value of Fisher's exact test.

DNAm within promoter regions is known to block gene expression (Moore et al. 2013; Steyaert et al. 2016), driving phenotypic variation in development, ageing and disease and making it particularly relevant for ageing research and a wide range of other biological studies. On the zebra finch sex chromosome promoters, DNAm increased with age on 12 sites and decreased on 21 sites (Figure 4D). In the jackdaw, all AR‐CpG sites in promoters of sex chromosomes increased with age (n = 4; Figure 4H). In the zebra finch autosomal chromosomes, 80 AR‐CpG sites were detected in promoter regions, with DNAm increasing with age at 44 sites and decreasing with age at 36 sites (Figure 4D). In the jackdaw autosomal chromosomes, 43 AR‐CpG sites were found in promoters, among which 15 and 28 increased and decreased in DNAm with age, respectively (Figure 4H). For a small number of AR‐CpG sites in promoters, the specific genes could be identified (Table S6). In both species, DNAm decreased with age on the majority of intergenic W chromosome AR‐CpG sites (Figure S1). Details of the direction of change in the degree of DNAm at AR‐CpG sites per annotation category per chromosome type in each species are depicted in Figure S1.

The GO enrichment analysis of the genes whose promoters contained AR‐CpG sites was not involved in shared biological processes or molecular functions for either species.

The number of AR‐CpG sites identified by our pipeline depends on the stringency of the selection criteria, raising the question of whether this biases our findings. We therefore repeated the analysis with less stringent criteria, resulting in an increase in identified AR‐CpG sites (Figures S2, S3). However, the overall trends regarding the distribution of AR‐CpG sites across the chromosomes remained the same in both species (Figures S2, S3), leading us to assume that our results were robust with respect to the specific choice of pipeline criteria.

4. Discussion

Our finding of a high density of AR‐CpG sites on the W chromosome in two avian species highlights the potential of sex chromosome DNAm to drive sex‐specific ageing (G. Li et al. 2022; Lund et al. 2020). A unique strength of our study was the longitudinal character of the data, combined with whole‐genome methylome data, which provided us with a high‐resolution view of the entire methylome. Consequently, our whole‐genome, longitudinal approach enabled us to detect previously unidentified sex‐chromosome‐specific DNAm patterns, unlike earlier attempts that were restricted to targeting specific genomic regions (Meyer et al. 2023) or relied on cross‐sectional analyses (Sun et al. 2021). We hypothesise this pattern may contribute to sex‐dependent ageing (e.g., Briga et al. 2017). The similarity of our findings between jackdaws and zebra finches, such as the comparable overall number of AR‐CpG sites, their increase and decrease in DNAm over time, as well as their distribution within each annotation category on the W chromosome suggests our findings may apply generally to passerine birds.

Whether DNAm increases (hypermethylation) or decreases (hypomethylation) with age typically varies among CpG sites (S. Li et al. 2017). On the W chromosome in both species, DNAm decreased with age in the majority of AR‐CpG sites. The avian W chromosome contains a high density of transposable elements (Peona et al. 2021), whose expression has negative fitness consequences (Warmuth et al. 2022), which may explain why DNAm is substantially higher on W compared to the autosomal chromosomes (Tone et al. 1984, also in our study species, data not shown). Decreasing DNAm with age likely leads to declining suppression of transposable element expression, and thereby increased genomic instability and deregulation of gene expression, both drivers of senescence (Gaudet et al. 2003; López‐Gil et al. 2023). On the W chromosomes of both species, AR‐CpG sites were significantly enriched in intergenic regions. Besides transposons, intergenic regions also harbour microRNAs (Ambros 2004) that modulate longevity and the ageing processes (Kinser and Pincus 2020) while playing a significant role in cell proliferation, cell death and haematopoiesis (Ambros 2004). Dysregulation of microRNA gene expression can lead to malignancy (Costinean et al. 2006) and disrupted immune cell homeostasis (Rose et al. 2021). Decreased levels of DNAm with age in intergenic regions of the W chromosome suggest loss of the ability to inhibit detrimental genetic elements such as deleterious mutations leading to genomic instability, which is one of the hallmarks of ageing. Our findings suggest that age‐related changes in DNAm on the W chromosome may correlate with the shorter lifespans observed in female birds compared to males (Tower and Arbeitman 2009). However, further longitudinal studies are needed to disentangle developmental and ageing‐related epigenetic dynamics in this context.

The density of AR‐CpG sites was higher than expected by chance on the zebra finch Z chromosome, whereas there were slightly fewer AR‐CpG sites than expected on the jackdaw Z chromosome. This difference matches the sex difference in lifespan in the two species: in zebra finches, females have a shorter lifespan (Briga et al. 2017), whereas there is no difference in lifespan between the sexes in jackdaws (Boonekamp et al. 2014). The density of AR‐CpG sites deviated significantly from the null expectations on most autosomes, but most deviations were small. The notable exceptions were chromosomes 32 and 35 in the zebra finch, where the density of AR‐CpG sites was higher than expected. Whether variation in the AR‐CpG density among chromosomes (other than the W chromosome) reflects the frequency of transposable elements or other functional components among chromosomes remains to be investigated.

Our finding that AR‐CpG sites on the W chromosome mostly decreased in DNAm with age and that most such sites were located in intergenic regions contrasts with research on the human haploid, heterogametic Y chromosome, where the majority of AR‐CpG sites were found to become hypermethylated with age and were located in genes (Lund et al. 2020). On the other hand, a single early study found the mammalian X chromosome to harbour relatively few AR‐CpG sites (Acevedo et al. 2015), as we found for the jackdaw Z chromosome. The avian sex chromosomes ZW are not homologous to the mammalian XY (Nada et al. 1999), and it is therefore unsurprising that DNAm patterns differ from male heterogametic (XY) species. Our findings suggest that DNAm on sex chromosomes likely plays different roles in male and female heterogametic systems. Furthermore, our results underscore the notion that some of the key mechanisms driving sex‐specific differences in ageing might be fundamental and thus conserved among closely related species but not across taxa. Consequently, efforts should be directed towards deciphering the role of sex chromosomes in ageing across a broader range of species as well as understanding the mechanisms underlying this process in species with various sex determination systems.

5. Conclusions

Utilising longitudinal, whole‐genome DNAm data from two avian species, we were able to characterise the distribution of CpG sites whose DNAm significantly changes with age. The location, direction of change in DNAm and functional significance of those lead us to believe that avian sex chromosomes – particularly the haploid, female‐specific W – are subject to distinct age‐related epigenetic regulation. Furthermore, our findings suggest that the observed significant changes in DNAm can lead to genomic instability, potentially contributing to the observed shorter lifespans of female birds compared to males. While these findings are a first step towards deciphering the role that epigenetic changes of sex chromosomes play in ageing, further research across diverse taxa will be essential to uncover conserved mechanisms underlying the ageing process and sex‐dependent ageing.

Author Contributions

M.T., J.S. and S.V. conceived and designed the study. M.T. and J.S. performed the bioinformatics and data analysis under the supervision of S.V. and P.J.P. F.F. performed the genomic annotation. M.T. and S.V. wrote the manuscript with input from all authors. All authors approved the final version of the manuscript.

Ethics Statement

All methods and experiments detailed in this manuscript were performed under the approval of the Central Committee for Animal Experiments (Centrale Commissie Dierproeven) of the Netherlands, under licences AVD1050020174344 and AVD1050020184967.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Proportion of AR‐CpG sites increasing and decreasing in DNAm with age in each annotation category in the zebra finch (A) and the jackdaw (B).

Figure S2: Comparison of ‘lenient’ (A) and ‘standard’ (B) pipelines for the zebra finch. The lenient pipeline selects positions with average coverage of ≥ 17 (Step 1), sites with average % DNAm between 30% and 70% (Step 2) and the top 8% of the correlations between % DNAm and Δage per 20 coverage (Step 4). As a result, the total number of sites finally selected increases to 7593 in the zebra finch. Note that the numbers of sites per chromosome are highly correlated between the two pipelines (r(35) = 0.95, p < 0.001), pointing to the high repeatability and robustness of the pipeline in response to threshold modifications, making it applicable to the whole range of methylation data.

Figure S3: Comparison of ‘lenient’ (A) and ‘standard’ (B) pipelines for the jackdaw. The lenient pipeline selects positions with average coverage of ≥ 20 (Step 1), sites with average % DNAm between 30% and 70% (Step 2), and at the top 8% of the correlations between % DNAm and Δage per 20 coverage (Step 4). As a result, the total number of sites finally selected increases to 9 in the zebra finch. Note that the numbers of sites per chromosome are highly correlated between the two pipelines (r(35) = 0.95, p < 0.001), pointing to the high repeatability and robustness of the pipeline in response to threshold modifications, making it applicable to the whole range of methylation data.

Table S1: Chronological age at collection and sex of all the individuals in our dataset.

Table S2: Proportions of observed and expected distribution of age‐related CpG sites in each chromosome and results of two‐sided exact binomial tests of observed versus expected proportions of AR‐CpG sites in each chromosome in the zebra finch. Significant differences (p < 0.05) are marked in bold.

Table S3: Proportions of observed and expected distribution of age‐related CpG sites in each chromosome and results of two‐sided exact binomial tests of observed versus expected proportions of AR‐CpG sites in each chromosome in the jackdaw. Significant differences (p < 0.05) are marked in bold.

Table S4: Proportions of observed and expected distribution of age‐related CpG sites in each annotation category and results of two‐sided exact binomial tests of observed versus expected proportions of AR‐CpG sites in each annotation category for the zebra finch (A) and the jackdaw (B). Expected proportions were calculated from the location of all CpG sites captured by our analysis in each annotation category.

Table S5: Enrichment and depletion of AR‐CpG sites across the genomic relative to total genomic coverage in the autosomes grouped together and the W and Z chromosomes of the zebra finch (A) and jackdaw (B). Enrichment statistics were calculated using a Fisher's exact test. Here enrichment is defined as an odds ratio > 1 and depletion as < 1.

Table S6: Identified genes in whose promoters we located AR‐CpG sites and whether DNAm increased or decreased with time at that AR‐CpG site in the zebra finch (A) and jackdaw (B). The 88 and 43 remaining AR‐CpG sites in promoters for the zebra finch and jackdaw, respectively were located in genes with unknown function.

MEC-34-e70147-s001.docx (2MB, docx)

Acknowledgements

We thank the Center for Information Technology at the University of Groningen for support and access to the Peregrine and Hábrók high‐performance computing clusters. We express our gratitude to Ellis Mulder for her help in the laboratory, Berber Maarsingh for illustrations and Raphaël Scherrer for assistance with the statistical analyses. We are also indebted to the animal caretakers at the University of Groningen, as well as numerous students whose invaluable help made this project possible.

Handling Editor: Maren Wellenreuther

Funding: M.T. was supported by an Adaptive Life Scholarship awarded by the University of Groningen. J.S. received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska‐Curie grant agreement no. 101025890. Contributions by F.B. and P.J.P. were supported by the European Union's Horizon 2020 Research and Innovation Programme under the Marie Skłodowska‐Curie grant agreement no. 813383 and the University of Groningen.

Marianthi Tangili and Joanna Sudyka contributed equally to this study.

Contributor Information

Marianthi Tangili, Email: tangilim@gmail.com.

Simon Verhulst, Email: s.verhulst@rug.nl.

Data Availability Statement

The dataset supporting the conclusions of this article is available in the the National Center for Biotechnology Information under BioProject ID PRJNA1108628. The bioinformatics code and R code, as well as intermediate and final data files used in this study can be found at https://doi.org/10.5061/dryad.wm37pvmw8.

References

  1. Acevedo, N. , Reinius L. E., Vitezic M., et al. 2015. “Age‐Associated DNA Methylation Changes in Immune Genes, Histone Modifiers and Chromatin Remodeling Factors Within 5 Years After Birth in Human Blood Leukocytes.” Clinical Epigenetics 7, no. 34: 1–20. 10.1186/s13148-015-0064-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Akalin, A. , Franke V., Vlahovic K., Mason C. E., and Schu D.. 2015. “Genome Analysis Genomation: A Toolkit to Summarize, Annotate and Visualize Genomic Intervals.” Bioinformatics 31: 1127–1129. 10.1093/bioinformatics/btu775. [DOI] [PubMed] [Google Scholar]
  3. Ambros, V. R. 2004. “The Function of Animal MicroRNAs.” Nature 431: 350–355. 10.1038/nature02871. [DOI] [PubMed] [Google Scholar]
  4. Andrews, S. 2010. “FastQC: A Quality Control Tool for High Throughput Sequence Data.” https://www.bioinformatics.babraham.ac.uk/projects/fastqc/.
  5. Benjamini, Y. , and Hochberg Y.. 1995. “Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing.” Journal of the Royal Statistical Society 57, no. 1: 289–300. [Google Scholar]
  6. Bindea, G. , Mlecnik B., Hackl H., et al. 2009. “ClueGO: A Cytoscape Plug‐in to Decipher Functionally Grouped Gene Ontology and Pathway Annotation Networks.” Bioinformatics 25, no. 8: 1091–1093. 10.1093/bioinformatics/btp101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Bird, A. 2002. “DNA Methylation Patterns and Epigenetic Memory.” Genes and Development 16, no. 1: 6–21. 10.1101/gad.947102. [DOI] [PubMed] [Google Scholar]
  8. Boonekamp, J. J. , Salomons M., Bouwhuis S., Dijkstra C., and Verhulst S.. 2014. “Reproductive Effort Accelerates Actuarial Senescence in Wild Birds: An Experimental Study.” Ecology Letters 17, no. 5: 599–605. 10.1111/ele.12263. [DOI] [PubMed] [Google Scholar]
  9. Briga, M. , Koetsier E., Boonekamp J. J., Jimeno B., and Verhulst S.. 2017. “Food Availability Affects Adult Survival Trajectories Depending on Early Developmental Conditions.” Proceedings of the Royal Society B: Biological Sciences 284, no. 1846: 20162287. 10.1098/rspb.2016.2287. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Costinean, S. , Zanesi N., Pekarsky Y., et al. 2006. “Pre‐B Cell Proliferation and Lymphoblastic Leukemia/High‐Grade Lymphoma in Eμ‐miR155 Transgenic Mice.” Proceedings of the National Academy of Sciences 103, no. 18: 7024–7029. 10.1073/pnas.0602266103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Ewels, P. , Magnusson M., Lundin S., and Käller M.. 2016. “MultiQC: Summarize Analysis Results for Multiple Tools and Samples in a Single Report.” Bioinformatics 32, no. 19: 3047–3048. 10.1093/bioinformatics/btw354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Firke, S. 2023. “janitor: Simple Tools for Examining and Cleaning Dirty Data.” R Package Version 2.2.0. https://cran.r‐project.org/package=janitor.
  13. Gardner, S. T. , Bertucci E. M., Sutton R., Horcher A., Aubrey D., and Parrott B. B.. 2023. “Development of DNA Methylation‐Based Epigenetic Age Predictors in Loblolly Pine ( Pinus taeda ).” Molecular Ecology Resources 23, no. 1: 131–144. 10.1111/1755-0998.13698. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Gaudet, F. , Hodgson J. G., Eden A., et al. 2003. “Induction of Tumors in Mice by Genomic Hypomethylation.” Science 300, no. 5618: 489–492. [DOI] [PubMed] [Google Scholar]
  15. Han, Y. , Nikolić M., Gobs M., et al. 2020. “Targeted Methods for Epigenetic Age Predictions in Mice.” Scientific Reports 10: 22439. 10.1038/s41598-020-79509-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Horvath, S. 2013. “DNA Methylation Age of Human Tissues and Cell Types.” Genome Biology 16, no. 1: 6. 10.1186/s13059-015-0649-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Kinser, H. E. , and Pincus Z.. 2020. “MicroRNAs as Modulators of Longevity and the Aging Process.” Human Genetics 139, no. 3: 291–308. 10.1007/s00439-019-02046-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Krueger, F. , and Andrews S. R.. 2011. “Bismark: A Flexible Aligner and Methylation Caller for Bisulfite‐Seq Applications.” Bioinformatics 27, no. 11: 1571–1572. 10.1093/bioinformatics/btr167. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Krueger, F. , James F., Ewels P., et al. 2023. “FelixKrueger/TrimGalore: v0.6.10.” Zenodo. 10.5281/zenodo.7598955. [DOI]
  20. Langmead, B. , and Salzberg S. L.. 2012. “Fast Gapped‐Read Alignment With Bowtie 2.” Nature Methods 9, no. 4: 357–359. 10.1038/nmeth.1923. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Levine, M. E. , Lu A. T., Quach A., et al. 2018. “An Epigenetic Biomarker of Aging for Lifespan and Healthspan.” Aging 10, no. 4: 573–591. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Li, G. , Wang C., Guan X., et al. 2022. “Age‐Related DNA Methylation on Y Chromosome and Their Associations With Total Mortality Among Chinese Males.” Aging Cell 21, no. 3: 1–12. 10.1111/acel.13563. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Li, S. , Christiansen L., Christensen K., et al. 2017. “Identification, Replication and Characterization of Epigenetic Remodelling in the Aging Genome: A Cross Population Analysis.” Scientific Reports 7, no. 1: 1–8. 10.1038/s41598-017-08346-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. López‐Gil, L. , Pascual‐Ahuir A., and Proft M.. 2023. “Genomic Instability and Epigenetic Changes During Aging.” International Journal of Molecular Sciences 24, no. 18: 1–28. 10.3390/ijms241814279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Lu, A. , Fei Z., Haghani A., Robeck T. R., Zoller J. A., and Li C.. 2023. “Universal DNA Methylation Age Across Mammalian Tissues.” Nature Aging 3: 1144–1166. 10.1038/s43587-023-00462-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lund, J. B. , Li S., Christensen K., et al. 2020. “Age‐Dependent DNA Methylation Patterns on the Y Chromosome in Elderly Males.” Aging Cell 19, no. 2: 1–8. 10.1111/acel.12907. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Marioni, R. E. , Shah S., Mcrae A. F., et al. 2015. “DNA Methylation Age of Blood Predicts All‐Cause Mortality in Later Life.” Genome Biology 16, no. 25: 1–12. 10.1186/s13059-015-0584-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Meyer, B. S. , Moiron M., Caswara C., et al. 2023. “Sex‐Specific Changes in Autosomal Methylation Rate in Ageing Common Terns.” Frontiers in Ecology and Evolution 11: 1–9. 10.3389/fevo.2023.982443.38516293 [DOI] [Google Scholar]
  29. Moore, L. D. , Le T., and Fan G.. 2013. “DNA Methylation and Its Basic Function.” Neuropsychopharmacology 38, no. 1: 23–38. 10.1038/npp.2012.112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Moqri, M. , Herzog C., Poganik J. R., et al. 2024. “Validation of Biomarkers of Aging.” Nature Medicine 30: 360–372. 10.1038/s41591-023-02784-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Nada, I. , Shan Z., Schartl M., et al. 1999. “300 Million Years of Conserved Synteny Between Chicken Z and Human Chromosome 9.” Nature Genetics 21: 258–259. [DOI] [PubMed] [Google Scholar]
  32. Pal, S. , and Tyler J. K.. 2016. “Epigenetics and Aging.” Science Advances 2: 1–19. 10.1007/978-3-319-55530-0_123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Peona, V. , Palacios‐Gimenez O. M., Blommaert J., et al. 2021. “The Avian W Chromosome Is a Refugium for Endogenous Retroviruses With Likely Effects on Female‐Biased Mutational Load and Genetic Incompatibilities.” Philosophical Transactions of the Royal Society, B: Biological Sciences 376, no. 1833: 20200186. 10.1098/rstb.2020.0186. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. R Core Team . 2023. “R: A Language and Environment for Statistical Computing.” (RStudio 2021.09.0 + 351 “Ghost Orchid” Release). https://www.r‐project.org/.
  35. Rhie, A. , McCarthy S. A., Fedrigo O., et al. 2021. “Towards Complete and Error‐Free Genome Assemblies of All Vertebrate Species.” Nature 592, no. 7856: 737–746. 10.1038/s41586-021-03451-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Rose, S. A. , Wroblewska A., Dhainaut M., et al. 2021. “A microRNA Expression and Regulatory Element Activity Atlas of the Mouse Immune System.” Nature Immunology 22: 914–927. 10.1038/s41590-021-00944-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Shannon, P. , Markiel A., Ozier O., et al. 2003. “Cytoscape: A Software Environment for Integrated Models of Biomolecular Interaction Networks.” Genome Research 13, no. 11: 2498–2504. 10.1101/gr.1239303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Steyaert, S. , Diddens J., Galle J., De Meest E., and De Keulenaer S.. 2016. “A Genome‐Wide Search for Epigenetically Regulated Genes in Zebra Finch Using MethylCap‐Seq and RNA‐Seq.” Scientific Reports 6, no. 20957: 1–17. 10.1038/srep20957. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Sun, D. , Layman T. S., Jeong H., et al. 2021. “Genome‐Wide Variation in DNA Methylation Linked to Developmental Stage and Chromosomal Suppression of Recombination in White‐Throated Sparrows.” Molecular Ecology 30, no. 14: 3453–3467. 10.1111/mec.15793. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Tangili, M. , Slettenhaar A. J., Sudyka J., et al. 2023. “DNA Methylation Markers of Age(Ing) in Non‐Model Animals.” Molecular Ecology 32, no. 17: 4725–4741. 10.1111/mec.17065. [DOI] [PubMed] [Google Scholar]
  41. Tangili, M. , Sudyka J., Furni F., Palsbøll P. J., and Verhulst S.. 2024. “Dataset—Identification of Age‐Related CpG Sites From Longitudinal Avian Methylomes.” Dryad. 10.5061/dryad.wm37pvmw8. [DOI]
  42. Tone, M. , Sakaki Y., Hashiguchi T., and Mizuno S.. 1984. “Genus Specificity and Extensive Methylation of the W Chromosome‐Specific Repetitive DNA Sequences From the Domestic Fowl Gallus gallus Domesticus.” Chromosoma 89, no. 3: 228–237. 10.1007/BF00295004. [DOI] [PubMed] [Google Scholar]
  43. Tower, J. , and Arbeitman M.. 2009. “The Genetics of Gender and Life Span.” Journal of Biology 8, no. 4: 8–10. 10.1186/jbiol141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. van de Pol, M. , and Wright J.. 2009. “A Simple Method for Distinguishing Within‐ Versus Between‐Subject Effects Using Mixed Models.” Animal Behaviour 77, no. 3: 753–758. 10.1016/j.anbehav.2008.11.006. [DOI] [Google Scholar]
  45. Warmuth, V. M. , Weissensteiner M. H., and Wolf J. B. W.. 2022. “Accumulation and Ineffective Silencing of Transposable Elements on an Avian W Chromosome.” Genome Research 32, no. 4: 671–681. 10.1101/gr.275465.121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Xia, X. , Chen W., McDermott J., and Han J. D. J.. 2017. “Molecular and Phenotypic Biomarkers of Aging.” F1000Research 6: 1. 10.12688/f1000research.10692.1. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Xirocostas, Z. A. , Everingham S. E., and Moles A. T.. 2020. “The Sex With the Reduced Sex Chromosome Dies Earlier: A Comparison Across the Tree of Life.” Biology Letters 16, no. 3: 1–6. 10.1098/rsbl.2019.0867. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Figure S1: Proportion of AR‐CpG sites increasing and decreasing in DNAm with age in each annotation category in the zebra finch (A) and the jackdaw (B).

Figure S2: Comparison of ‘lenient’ (A) and ‘standard’ (B) pipelines for the zebra finch. The lenient pipeline selects positions with average coverage of ≥ 17 (Step 1), sites with average % DNAm between 30% and 70% (Step 2) and the top 8% of the correlations between % DNAm and Δage per 20 coverage (Step 4). As a result, the total number of sites finally selected increases to 7593 in the zebra finch. Note that the numbers of sites per chromosome are highly correlated between the two pipelines (r(35) = 0.95, p < 0.001), pointing to the high repeatability and robustness of the pipeline in response to threshold modifications, making it applicable to the whole range of methylation data.

Figure S3: Comparison of ‘lenient’ (A) and ‘standard’ (B) pipelines for the jackdaw. The lenient pipeline selects positions with average coverage of ≥ 20 (Step 1), sites with average % DNAm between 30% and 70% (Step 2), and at the top 8% of the correlations between % DNAm and Δage per 20 coverage (Step 4). As a result, the total number of sites finally selected increases to 9 in the zebra finch. Note that the numbers of sites per chromosome are highly correlated between the two pipelines (r(35) = 0.95, p < 0.001), pointing to the high repeatability and robustness of the pipeline in response to threshold modifications, making it applicable to the whole range of methylation data.

Table S1: Chronological age at collection and sex of all the individuals in our dataset.

Table S2: Proportions of observed and expected distribution of age‐related CpG sites in each chromosome and results of two‐sided exact binomial tests of observed versus expected proportions of AR‐CpG sites in each chromosome in the zebra finch. Significant differences (p < 0.05) are marked in bold.

Table S3: Proportions of observed and expected distribution of age‐related CpG sites in each chromosome and results of two‐sided exact binomial tests of observed versus expected proportions of AR‐CpG sites in each chromosome in the jackdaw. Significant differences (p < 0.05) are marked in bold.

Table S4: Proportions of observed and expected distribution of age‐related CpG sites in each annotation category and results of two‐sided exact binomial tests of observed versus expected proportions of AR‐CpG sites in each annotation category for the zebra finch (A) and the jackdaw (B). Expected proportions were calculated from the location of all CpG sites captured by our analysis in each annotation category.

Table S5: Enrichment and depletion of AR‐CpG sites across the genomic relative to total genomic coverage in the autosomes grouped together and the W and Z chromosomes of the zebra finch (A) and jackdaw (B). Enrichment statistics were calculated using a Fisher's exact test. Here enrichment is defined as an odds ratio > 1 and depletion as < 1.

Table S6: Identified genes in whose promoters we located AR‐CpG sites and whether DNAm increased or decreased with time at that AR‐CpG site in the zebra finch (A) and jackdaw (B). The 88 and 43 remaining AR‐CpG sites in promoters for the zebra finch and jackdaw, respectively were located in genes with unknown function.

MEC-34-e70147-s001.docx (2MB, docx)

Data Availability Statement

The dataset supporting the conclusions of this article is available in the the National Center for Biotechnology Information under BioProject ID PRJNA1108628. The bioinformatics code and R code, as well as intermediate and final data files used in this study can be found at https://doi.org/10.5061/dryad.wm37pvmw8.

Articles from Molecular Ecology are provided here courtesy of Wiley

RESOURCES