Introduction:
Most species have narrow geographic ranges, but a few are distributed
across much of the globe. Many such species have maintained these broad
ranges for long periods, but others have only recently spread into new
areas, oftentimes through accidental or intentional human activities
(Jeschke and Strayer 2006). In addition to anthropogenic range
expansions, some species have altered their distributions as the
environment changed around them (Du et al. 2024). Habitats that were
historically natural became farms, cities, suburbs, or other
human-modified landscapes, and some species thrived in these new
locations (Ducatez et al. 2018, Polaina et al. 2021). Whether range
expansion is a natural process or anthropogenic, there is an enduring
question: how, mechanistically, do individuals endure such conditions
and establish new populations? One favored explanation is phenotypic
plasticity (Usui et al. 2023, Chown and McGeoch 2023): organisms that
best match their behavioral, morphological, and/or physiological traits
to prevailing conditions comprise founder populations (Kilvitis et al.
2017).
Whereas the evidence for a role of plasticity in range expansions is
strong (Davidson et al. 2011), the molecular processes whereby it is
realized are less known. A promising area of study entails the
epigenetic processes that alter how genetic variation is expressed
(Marin et al. 2020, Mounger et al. 2021). DNA methylation, histone
acetylation, small non-coding RNAs activity, and other processes alter
the accessibility of the genome to transcription factors (Vogt 2021,
Husby 2022b). More importantly, the interplay among these elements and
genetic sequence variation partly underlies heterogeneity in phenotype
among organisms (Vogt 2017b, Vogt 2021, Bogan and Yi 2024). Many
molecular epigenetic processes are also sensitive to current and past
environments, such that some organisms will adjust their gene expression
contingently, releasing adaptive (or non-adaptive) plasticity when a
particular environmental signal induces it (Zhang et al. 2020, Sepers et
al. 2019).
For practical reasons, DNA methylation has to date been the molecular
epigenetic mechanism that has garnered the most research attention
(Husby 2022a, Laine et al. 2023). DNA methylation tends to reduce gene
expression by impeding interactions between transcription factors and
regulatory regions of the genome, namely promoters (Vogt 2021), but gene
bodies, enhancers and other regions of the genome can be methylated,
too. Whereas methyl marks also strongly mitigate transposon activity
(Marin et al. 2019, Vogt 2021), at the organismal level, DNA methylation
plays a critical role during cell differentiation, which underlies many
forms of phenotypic plasticity. Indeed, methylation is involved in
polyphenisms in honeybees (Lyko et al. 2010), thermal plasticity in
zebrafish (Loughland et al. 2021), biorhythms and reproductive
phenotypes in mammals (Stevenson 2018), and differentiation and
activation of various leukocytes in vertebrates (Hong and Medzhitov
2023) among many other traits.
In the context of range expansions, DNA methylation and the genetic
substrates on which it acts (i.e., CpG motifs) seem to determine which
and by what means certain individuals come to comprise new populations
(Chen et al. 2024, Kilvitis et al. 2017). Our focal species in the
present study, the house sparrow (Passer domesticus ), is one of
world’s most broadly distributed birds (Hanson et al. 2020b). It is also
among the strongest examples of the role of DNA methylation in
vertebrate range expansions (Hanson et al. 2022, Hanson et al. 2020a,
Sheldon et al. 2018). In Kenya, for instance, where the species arrived
probably via human shipping activity, we found a strong inverse
correlation between genetic and epigenetic variation among populations
(Liebl et al. 2013). In this study, it was speculated that the pattern
arose because DNA methylation rescued some populations from extinction
by enabling some individuals to mitigate the effects of lethal alleles
or inbreeding depression, which are common in small populations. A
subsequent study revealed more direct support for DNA methylation in
this range expansion: CpG count in the genomes of individual birds (what
was termed epigenetic potential ) declined from the vanguard
population (i.e., the border of Uganda where the species probably
arrived no earlier than 2015) to the site of initial introduction (i.e.,
the city of Mombasa in ~1950) (Hanson et al. 2022). The
rationale was that fewer CpG sites would mean fewer opportunities for
methylation and hence reduced potential for plasticity. Subsequent
analysis revealed that high CpG counts were in fact under directional
selection (low Tajima’s D) at the range edge but not the core of the
Kenyan invasion (Hanson et al. 2022).
These findings partly motivate the present study: to compare expression
of the enzymes that regulate DNA methylation among populations of house
sparrows with different introduction histories. Our interest
specifically was to test the hypothesis that introduced populations
would express more of the enzymes important to the maintenance, addition
and subtraction of methyl marks on the genome than native populations
(Hanson and Liebl 2022, Vogt 2017a). DNA methylation in vertebrates is
coordinated by two methyltransferases, DNA methyltransferase 1 and 3
(DNMT1 and 3), and one ten-eleven translocation methylcytosine
dioxygenase, TET2 (Robertson and Wolffe 2000). DNMT1 is largely
responsible for the maintenance of methyl marks set down during
development; once established, cell identity becomes lost if methyl
marks disappear (Law and Jacobsen 2010). The main role of DNMT1 is thus
to keep methyl marks intact, imbuing cells with a sort of memory
transfer between cell generations, an indispensable trait for species
with high cell turnover rates (Regev et al. 1998). DNMT3, by contrast,
is important in de novo methylation. DNMT3 sets down the marks
during the blastocyst stage, with marks later being maintained by DNMT1
(Wu and Zhang 2010). DNMT3 can also methylate the genome later in life,
but contingent on environmental exposures to a variety of factors
(Hanson and Liebl 2022). Finally, TET2 is responsible for inactivating
methyl marks, functionally eliminating methylation from a previously
methylated region (Wang et al. 2018). Historically, methylation was
thought to be a quite stable epigenetic mark, but extensive recent work
has revealed methylation can be quite labile (see also Schrey et al.,
this issue). In some tissues and contexts, it is even reversible (Wu and
Zhang 2010, Stevenson 2018), especially during early phases of
development (Vogt 2021, Vogt 2017a).
We tested our hypothesis by comparing DNMT/TET gene expression among 9
different populations of house sparrows (Table 1), choosing specific
populations depending on whether they: i) had independent introduction
histories, ii) are definitively native or not, and iii) are sufficiently
large to enable sample collection (Hanson et al. 2020b). We also
selected these populations to test whether the effects of introduction
history were smaller or larger than other factors relevant to
plasticity, specifically latitude, altitude, and climatic
predictability. All three factors should relate to plasticity, as they
represent different forms of the rate of environmental change;
phenotypic plasticity cannot be adaptive if environmental change happens
too fast. High latitude or altitude environments, for instance, are much
more dynamic and unpredictable than near-equatorial and low elevation
ones (Hau 2001). Extremes of climatic unpredictability exist generally
across the globe, and these conditions are well described by Colwell’s
indexes (Colwell 1974). Colwell’s indexes represent day-to-day
predictability of temperature and precipitation relative to the local
annual cycles; high values represent comparatively predictable
conditions and low values represent unpredictable ones. Recently,
Colwell’s indices have been calculated at a 0.5ospatial resolution across the world using climate data from 1901-2012
(Jiang et al. 2017, Harris et al. 2014). We predicted both DNMTs and TET
gene expression would be higher in more dynamic environments. We also
expected that human-habitat modification could influence expression of
these genes. Urbanization is evolutionarily novel, so birds dwelling in
or near cities should express more DNMTs or TET2 to help birds adjust
their phenotypes adequately to unnatural conditions. Finally,
individual-level factors such as sex and genetic ancestry could also
affect DNMT/TET expression. In the case of sex, methylation is a major
means by which heterogametic genes are differentially silenced or
enhanced (Vogt 2021, Vogt 2017a). In the case of genetic ancestry,
similarity in gene expression could come from shared history. There
appears to be two major lineages of house sparrows from which all extant
populations derive (Ravinet et al. 2018). Overall, we predicted that
introduction history, climatic predictability, and tissue identity would
be the strongest predictors of variation in enzyme expression (Mishra et
al. 2020, Coyle et al. 2020). Tissues are comprised of many cell types,
each of which derives from different types and degrees of epigenetic
activity that various cells experience during differentiation. We
expected that the effects of other factors such as individual sex,
genetic ancestry, urbanization, latitude, and altitude would be
detectable but comparatively weaker than the above forces.