Conclusion
Clearly, our study raises more questions than it answers. Several drivers of expression were identified, but in a direction opposite of predictions (e.g., tissue, temperature predictability, native/non status, urbanization). Likewise, factors expected to have some explanatory power on expression (e.g., sex of bird, precipitation predictability) had little to none. There are a few next steps that could be particularly enlightening to understand the role of DNMTs and TET2 in this or related systems. First, experiments involving transcriptional repression of DNMT expression or exclusion of DNMTs from nuclei so methyl marks naturally degrade (Law and Jacobsen 2010) could reveal both the role of DNMTs in phenotypic plasticity in various cells of house sparrows but also the success of individual birds in new ecological contexts (Luo et al. 2012). Genetically-modified animals cannot ethically be released from captivity, but creative approaches to studying sparrow domestication or the colonization of captive spaces could be insightful. Relatedly, DNMT/TET expression could be studied in embryonic and nestling birds (Wilks et al. 2023, Siller Wilks et al. 2024). Our bias to studying adults here probably missed important among and within population differences, but the study of adults is justified from other work (Sun et al. 2021). It will also be important to consider alternative, functional roles for DNMTs besides plasticity. InDaphnia magna (Agrelius et al. 2023), DNMT3 expression was higher in calorically-restricted individuals, altering the life history trajectories faced by individuals depending on the environments in which they were reared (Nguyen et al. 2021).
A final lens through which to view DNMT/TET expression relates to the concept of epigenetic potential (Kilvitis et al. 2017), genetic variation among individuals in the propensity for their genomes to be methylated. Originally, we proposed genetic polymorphisms in DNMTs as a possible, ecologically relevant form of epigenetic potential, but neither we nor others have yet to search for DNMT genetic variants in invaders. One recent study of great tits (Sepers et al. 2023) revealed nine SNPs in DNMT3a, two of which were associated with methylation of two distant CpGs. One of these CpGs occurred in an exon of the gene, SELPLG, and another intronically in CTNNA3. Whether these SNPs affect expression of these genes, the expression of DNMT3 itself, or the physiological functions any of these genes was not considered. Nevertheless, in future work, it would be valuable to determine whether different DNMT forms are directionally selected, just as the CpG content of gene promoters was in the Kenyan house sparrow range expansion (Kilvitis et al. 2017).
It is an exciting time for ecological epigenetics, as the technical toolkit it requires is expanding rapidly (Loughland et al. 2021). We are also becoming better able to ‘iteratively measure plastic traits’ (Dupont et al. 2024) and distinguish plasticity via epigenetic processes as the outcome of ‘directional induction or bet-hedging stochasticity’ (Vogt 2021). We are only just beginning to appreciate, though, that plasticity is probably important to so many biological processes because it underpins organismal agency (Mitchell 2023, Kirchhoff et al. 2018, Friston 2010, Ball 2023). Strong eco-evolutionary roles of plasticity are well known across biological systems (Wade and Sultan 2023), but as we come to understand how methylating enzymes help sculpt the epigenotype in nature, we could be taking a small but important step to revealing how organisms use a variety of entangled, cognitive plasticities (Watson and Szathmáry 2016) to achieve resilience through antifragility (Taleb 2014) and thus endure natural and ongoing anthropogenic change.