Genomic Signatures of Climate Adaptation in Anolis cybotes

Anolis cybotes, female from Barahona, Dominican Republic

Katharina Wollenberg Valero & Ariel Rodríguez

Thermal adaptation is the evolution of the ability to persist in novel thermal environments. Phenotypic characters that allow such adaptation, as well as the resulting shifts in the geographic distributions of species, are an emerging field of study in the midst of a changing global climate. Yet, the genomic basis of such phenotypic adaptation is less well understood, so recent efforts of evolutionary biologists are now aiming at one emerging question: Which genes determine thermal adaptation, and are these the same across different populations and species? Luckily, Anolis is yet again at the forefront of novel discoveries being made in this field (see Campbell-Staton et al., 2017).

Many studies have independently identified genes that are responding to changes in the thermal environment, be it through change of expression under an acute stress, or through changes in the DNA sequence as evolutionary response. In 2014, we gathered information on such thermal adaptation candidate genes from Drosophila to Homo sapiens from the literature.

From the published evidence, we extracted a set of gene functions that potentially underlie climatic adaptation. We were able to match these with functions that are known from phenotypic thermal adaptation (Wollenberg Valero et al., 2014). Interestingly, the products of these genes (Proteins, RNAs) were found to be functionally related with each other thus forming gene networks within the cellular environment.

The Caribbean Anolis cybotes is widely distributed across Hispaniola, and thrives in hot, xeric environments just as well as in cooler and more humid montane environments. The rift valley of Lago Enriquillo heats up to 40.5 °C (104.9 °F), and a few instances of frost were reported at the highest peak (Pico Duarte at 3,098m elevation) – so population survival across these climatic extremes does not seem to be a trivial endeavor.

Populations of this species show pronounced differences between montane and lowland forms in morphology, physiology, behavior, and perch use (Wollenberg et al., 2013; Muñoz et al., 2014), which led us to expect that at least some of this variation should have a genetic basis. Thus, we set up to test whether Anolis cybotes displays any signatures of genomic adaptation to the diverse kinds of environments it inhabits, and whether any genes showing evidence for selection can also be subsumed under the candidate functions we defined previously.

We sampled tissue of these lizards from several high and low elevations (the specimens being the same as in Wollenberg et al., 2013), and looked for variation according to climatic differences via RAD sequencing and subsequent analysis with LFMM. RAD sequencing generates a reduced representation of the target genome, producing thousands of short sequences representing the distribution of the restriction enzyme’s cutting sites throughout the genome. Owing to this property, it cannot be expected that this type of data will necessarily contain “the total set of adaptation genes”; to this effect, detailed genome sequencing is required and such studies have been done in some model organisms (stickleback fish, beech mice, Drosophila, etc.). With our study design, however, we could trace signatures of selection as climate-related changes in the allelic frequencies in the fragments that were sequenced. We identified a total of 84 SNPs with statistical signatures of selection and 14 of these matched protein-coding genes on different chromosomes of the Anolis carolinensis genome (the best available reference). Not surprisingly, our data set and analysis did not “hit” any major known candidate genes for thermal adaptation, but we made another discovery. Most of the genes that we did identify as having adapted to the different climatic environments in Anolis cybotes populations perform the set of previously predicted gene functions that we had predicted in 2014 (Rodríguez et al., 2017).

Moreover, the newly identified genes were also in close functional connection with each other, and with many of the previously predicted genes, forming a tightly knit functional network. Some forays into tissue expression databases further revealed that many of these genes are also expressed in brain, and during early development.

The figure shows candidate gene functions for thermal adaptation that we predicted in 2014 and now verified in Anolis cybotes

Due to the fact that we couldn’t scan the entire genome of Anolis cybotes (so far unknown), we may have missed part of the story. However, it is encouraging that at least one of the genes adapting to different climatic environments in the green anole, A. carolinensis, also has one of the functions we predicted (vasodilation, constriction and regulation of blood pressure, Campbell-Staton et al., 2016).

Phenotypic adaptation to climate can happen in several different ways and we are yet only scratching the surface of the genetic basis of this phenomenon. For reptiles, their thermoregulatory behavior, water balance in hot climates, freeze tolerance, and anti-oxidative strategies seem to be the most important eco-physiological challenges (Storey and Storey, 2017), and we deem it likely that changes in a multitude of genes are contributing to these adaptive responses. Our study provides evidence that climate adaptation on the genomic level is constrained to specific organismal functions and biochemical pathways, which may underpin the observed molecular and phenotypic differences. The definitive answer to the question on whether the same set of genes underlie climatic adaptation across populations and species, is yet to be found; but the study of the functional connections between genes can be very informative for this endeavor.

You can read our full paper here (open access)