Seasonal Adaptation: Geographic Photoperiod-Temperature Patterns Explain Genetic Variation in the Common Vole Tsh Receptor

Abstract

The vertebrate photoperiodic neuroendocrine system uses the photoperiod as a proxy to time the annual rhythms in reproduction. The thyrotropin receptor (TSHR) is a key protein in the mammalian seasonal reproduction pathway. Its abundance and function can tune sensitivity to the photoperiod. To investigate seasonal adaptation in mammals, the hinge region and the first part of the transmembrane domain of the Tshr gene were sequenced for 278 common vole (Microtus arvalis) specimens from 15 localities in Western Europe and 28 localities in Eastern Europe. Forty-nine single nucleotide polymorphisms (SNPs; twenty-two intronic and twenty-seven exonic) were found, with a weak or lack of correlation with pairwise geographical distance, latitude, longitude, and altitude. By applying a temperature threshold to the local photoperiod-temperature ellipsoid, we obtained a predicted critical photoperiod (pCPP) as a proxy for the spring onset of local primary food production (grass). The obtained pCPP explains the distribution of the genetic variation in Tshr in Western Europe through highly significant correlations with five intronic and seven exonic SNPs. The relationship between pCPP and SNPs was lacking in Eastern Europe. Thus, Tshr, which plays a pivotal role in the sensitivity of the mammalian photoperiodic neuroendocrine system, was targeted by natural selection in Western European vole populations, resulting in the optimized timing of seasonal reproduction.

Keywords: Microtus arvalis; Tsh receptor; climate change; common vole; natural selection; seasonal reproduction; temperature–photoperiod ellipsoid.

Figure 1

Sample locations and local annual photoperiod–temperature ellipsoids. (A) Distribution map of localities of common vole samples used in this study. Location numbers with corresponding environmental proxies are listed in Table S1. (B) Annual patterns (counter-clockwise) of photoperiod and 10-year average monthly ambient temperature for each sample location in Western Europe and (C) in Eastern Europe. Temperature data were obtained from the closest weather station (always within 110 km of sample location) obtained from Wunderground (https://www.wunderground.com/ (accessed on 7 December 2022)). Photoperiod was obtained from https://www.timeanddate.com (accessed on 7 December 2022) and is based on civil twilight times at dawn and dusk, which is the timing at which log light intensities change most rapidly [5,48]. Civil twilight incorporates geographical and seasonal variation in the duration of twilight [49] and is, therefore, considered as the moment of ‘lights on’ and ‘lights off’ for biological systems [5]. Dotted lines indicate a temperature threshold at 6.6 °C, from which the corresponding predicted critical photoperiod (pCPP) in spring (ellipse crossing the 6.6 °C line for the second time) is used as a proxy for onset of grass growth and, consequently, as a proxy for onset of the favorable season. Regional pCPPs are listed in Table S1. Colors indicate latitude, ranging from 42° N (yellow) to 59° N (purple).

Figure 2

The Tshr gene and predicted protein for the common vole. (A) The Tshr gene for the common vole, consisting of 8 introns and 9 exons. (B) The magnified region, including the last part of intron 8 and the first part of exon 9, were sequenced in this study. Green arrows indicate location of forward primers, and red arrows indicate location of reverse primers for sequencing. (C) The predicted TSHR protein and (D) its 3D structure. All mutations found in the current study are labeled and listed in Tables S4–S7. SP = signal peptide, LRR = leucine-rich repeat, TMD = transmembrane domain, ICD = intracellular domain, ECD = extracellular domain, CDS = coding sequence.

Figure 3

Pairwise multilocus F_ST heatmap for Tshr haplotypes. F_ST values were calculated using the original method for estimation of genetic distance with correction for sample size bias [70]. Colors indicate pairwise multilocus F_ST values ranging from 0 (yellow) to 1 (dark blue). Western and Eastern European populations are ordered from long to short predicted critical photoperiods (pCPPs).

Figure 4

Tshr mutations in Western European common vole populations. Manhattan-type plots (−log₁₀(p)) for the sequenced Tshr region for (A) geographical distance, (D) latitude, (G) longitude, (J) altitude, and (M) predicted critical photoperiod (pCPP). Gray bars indicate Benjamini–Hochberg adjusted p-values, and black bars indicate R²-values. SNPs that meet the threshold for significant correlations (p < 0.05) cross the red dashed line. Pairwise differences in SNP frequency for two representative mutations (SNP-158 and SNP126) related to (B,C) pairwise geographical distance, (E,F) pairwise latitudinal difference, (H,I) pairwise longitudinal difference, (K,L) pairwise altitudinal difference, and (N,O) pairwise difference in pCPP. Significant correlations are indicated by linear regression lines. All statistic results of linear models for SNP frequency related to environmental proxies are found in Table S6.

Figure 5

Tshr mutations in Eastern European common vole populations. Manhattan-type plots (−log₁₀(p)) for the sequenced Tshr region for (A) geographical distance, (D) latitude, (G) longitude, (J) altitude, and (M) predicted critical photoperiod (pCPP). Gray bars indicate Benjamini–Hochberg adjusted p-values, and black bars indicate R²-values. SNPs that meet the threshold for significant correlations (p < 0.05) cross the red dashed line. Pairwise differences in SNP frequency for two representative mutations (SNP-158 and SNP729) related to (B,C) pairwise geographical distance, (E,F) pairwise latitudinal difference, (H,I) pairwise longitudinal difference, (K,L) pairwise altitudinal difference, and (N,O) pairwise difference in pCPP. Significant correlations are indicated by linear regression lines. All statistic results of linear models for SNP frequency related to environmental proxies are found in Table S7.