Skip to main content
NCBI home page
Proceedings of the Royal Society B: Biological Sciences logoLink to Proceedings of the Royal Society B: Biological Sciences
. 2014 Jul 7;281(1786):20140021. doi: 10.1098/rspb.2014.0021

Adaptive phylogeography: functional divergence between haemoglobins derived from different glacial refugia in the bank vole

Petr Kotlík 1,✉, Silvia Marková 1, Libor Vojtek 3, Antonín Stratil 1, Vlastimil Šlechta 2, Pavel Hyršl 3, Jeremy B Searle 4
PMCID: PMC4046400  PMID: 24827438

Abstract

Over the years, researchers have used presumptively neutral molecular variation to infer the origins of current species' distributions in northern latitudes (especially Europe). However, several reported examples of genic and chromosomal replacements suggest that end-glacial colonizations of particular northern areas may have involved genetic input from different source populations at different times, coupled with competition and selection. We investigate the functional consequences of differences between two bank vole (Clethrionomys glareolus) haemoglobins deriving from different glacial refugia, one of which partially replaced the other in Britain during end-glacial climate warming. This allows us to examine their adaptive divergence and hence a possible role of selection in the replacement. We determine the amino acid substitution Ser52Cys in the major expressed β-globin gene as the allelic difference. We use structural modelling to reveal that the protein environment renders the 52Cys thiol a highly reactive functional group and we show its reactivity in vitro. We demonstrate that possessing the reactive thiol in haemoglobin increases the resistance of bank vole erythrocytes to oxidative stress. Our study thus provides striking evidence for physiological differences between products of genic variants that spread at the expense of one another during colonization of an area from different glacial refugia.

Keywords: adaptation, antioxidative capacity, climate change, cysteine, oxidative stress, redox

1. Introduction

Phylogeography has been one of the mainstays of evolutionary study over the 25 years since the seminal paper of Avise et al. [1] launched the field. Over 10 000 phylogeographic papers have been published and, for instance, the review article by Hewitt [2] has been cited more than 2000 times. The core aim of phylogeography is to use molecular markers to infer historical biogeographic scenarios at the within-species level. The ideal markers for such analysis would seem to be those that are selectively neutral, because these will purely reflect historical processes, revealing population origins and geographical spread [3]. However, colonization is not necessarily a selectively neutral process, and if we are properly to understand why individuals of a particular genetic constitution are in a particular area, we need to determine what aspect of that genetic constitution may have been favoured during the colonization process. This is what we mean by ‘adaptive phylogeography’.

Recently, phylogeographic data have been used to provide a historical context for adaptive divergence by local adaptation, for example to inform on how many independent times an adaptive phenotype evolved or a timeframe for the adaptation [4–6]. Such studies use phylogeographic data to study adaptation and also can qualify as adaptive phylogeography. However, we use the term primarily to refer to an inquiry into the role of adaptive differences between populations in shaping the phylogeographic patterns themselves.

One of the most persuasive situations where selection may be involved in colonization history is where data reflect some sort of genetic replacement event during colonization of an area, involving one source area at the cost of another. This can range from a selective sweep involving one or a few loci, to complete population replacement [7].

There are indications from studies incorporating analysis of ancient DNA that replacement may have been a not infrequent process during natural colonization of Eurasia and North America at the end of the last glaciation [8,9]. The most obvious driver of that process is climate change at that time, and as with replacements relating to recent climate change there is a strong likelihood that selection may have been involved [7]. An example of recent population replacement that can most reasonably be considered a selective response to changing climate is the twentieth century spread of the carrion crow at the expense of the hooded crow in northern Britain [10].

Within-species genetic structuring is believed to substantially relate to events at the end of the last glaciation [9,11]. Phylogeography thus largely aims to explain end-glacial colonization processes and has particularly been applied to the colonization of northern latitudes, Europe in most detail [12]. Therefore, it is clearly crucial to engage in the analysis of end-glacial replacement events. These may generate some of the starkest examples of selective processes influencing the colonization process. End-glacial replacements are likely to reflect the influence of changing climate. From the last glacial maximum 21 kyr ago until the beginning of the Holocene 11 kyr ago, there were periods of rapid warming and cooling, which undoubtedly had dramatic impacts on species distributions [13]. Given our increasing understanding of the genetics of climate adaptation [14], it should become possible to develop clear predictions on the loci involved in selectively driven replacement events at the end of the last glaciation.

Phylogeography is at its most persuasive when multiple species show like genetic patterns in a particular geographical area, indicating a similar progression and process of colonization, including the localization of glacial refugia and the manner of end-glacial spread from such refugia [15,16]. Multiple species within a geographical area may also show the genetic signature of comparable replacement events; multiple species responding to the same end-glacial climatic drivers in a particular geographical area. We here study replacement in just such a system, and provide evidence that supports the action of natural selection in explaining the replacement. Our analysis goes beyond the analysis of frequency and distribution of genetic markers that is typical in phylogeographic research and instead moves into the functional analysis of gene products. We suggest that this will be typical for studies of adaptive phylogeography. Studies of functional genes and their products are essential to understand selective processes acting during colonization.

On the basis of phylogeographic inference [7,17,18], we and others recently showed two-phase colonizations of Britain from continental Europe in each of five species of small mammals (three voles and two shrews), which involved initial colonization of the whole of Britain from one glacial refugium. The second colonization involved a partial genetic replacement originating in a different glacial refugium. Therefore for the genetic markers examined, there was a fringe around the north, west and south of Britain representing genotypes from the first refugium, and in central and eastern areas genotypes from the second refugium [7]. Interestingly, a much earlier report by Hall [19] showed a pattern of distribution of haemoglobin (Hb) in one of the species, the bank vole (Clethrionomys glareolus). Hall speculated that the clear-cut north–south separation between the two Hb variants (with a boundary running through northern England) may reflect the action of natural selection [19]. The two Hb variants coincide with the northern (‘fringe’) and southern mtDNA phylogroups of the bank vole [7]. The fact that Hb in the bank vole showed the same signature of replacement as the other markers in the different species gave a handle for examining the possibility of adaptive processes in explaining the replacements. Hb is one of only a handful of gene products for which there have been within-species studies in mammals with adequate functional analysis to demonstrate clear advantages and disadvantages of different naturally occurring variants [5,20,21]. Here, we use Hb to provide the support for an adaptive divergence involved in an end-glacial genetic replacement, a finding that may have wide significance for our understanding of the geographical redistribution of species under a changing climate.

2. Material and methods

(a). Samples

We sampled a total of 145 bank voles at 12 localities along an approximately north–south transect along Britain (figure 1; the electronic supplementary material, table S1). Bank voles were trapped and taken under Natural England (General) and Scottish Natural Heritage (10080) licences. Blood was obtained from dead voles by cardiac puncture and the heparinized red blood cells (RBCs) were collected by centrifugation (3000 r.p.m. in a microfuge, i.e. approx. 630g, 10 min), washed with phosphate-buffered saline, resuspended in water, and frozen in liquid nitrogen. Samples of bone marrow and spleen (an erythropoietic organ in rodents) [22] were stored in RNAlater (Qiagen, Valencia, CA, USA) and in 96% ethanol for DNA analyses.

Figure 1.

Figure 1.

Open in a new tab

Distribution map of the bank vole haemoglobin structural variants (Hb S and Hb F) in Britain. Pie diagrams on the left represent allele frequencies at the three amino acid polymorphisms in the two β-globin genes (HBB-T1 and HBB-T2) with the evolutionarily derived amino acids shown in white. Localities (see the electronic supplementary material, table S1) are grouped into a northern and southern group relative to the position of the Hb frequency cline (see the electronic supplementary material, figure S4).

(b). Gene isolation and sequencing

Genomic DNA was extracted using the DNeasy Tissue Kit (Qiagen). We used published PCR primers flanking globin genes of the house mouse (Mus musculus) [23] and deer mouse (Peromyscus maniculatus) [5] in an attempt to amplify orthologous bank vole genes. To isolate genes not amplified with these primers, we cloned and sequenced bank vole globin complementary DNA (cDNA). Total RNA of three voles was extracted using the RNeasy Mini Kit with DNase treatment (Qiagen) and was reverse transcribed using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, MA, USA). We first obtained partial bank vole cDNA sequences using primers matching conserved motifs in exons of mammalian globin genes [24], one pair for α-globin and another for β-globin (see the electronic supplementary material, table S2). We then designed new primers (see the electronic supplementary material, table S3) in regions of the bank vole sequences that did not show divergence (i.e. overlapped multiple peaks in chromatograms) between paralogous genes co-amplified with the conserved primers. These primers were used for Rapid Amplification of cDNA Ends (RACE) with the SMARTer RACE cDNA Amplification Kit and Advantage 2 PCR Kit (Clontech, Mountain View, CA, USA). This strategy ensured that transcripts of all bank vole adult α- and β-globin paralogues were amplified. The RACE products were cloned with the Qiagen PCR Cloning Plus Kit and sequenced from 55 to 103 clones per vole, and full-length cDNA sequence was assembled for each gene from the overlapping 5′- and 3′-RACE sequences. We then designed paralogue-specific primers in the 5′ and 3′untranslated regions (UTRs) to amplify and sequence each gene from start to stop codon from genomic DNA (see the electronic supplementary material, table S4). All genotypes heterozygous at multiple sites were resolved by cloning and sequencing of four to eight clones.

(c). RNA-seq and gene expression analysis

Library preparation and sequencing were performed for three voles with the use of the standard Illumina (San Diego, CA, USA) protocols on the Illumina HiSeq 2000. Briefly, after poly(A) enrichment and fragmentation, the RNA was size-selected to 250–400 bp, reverse transcribed into cDNA, end-repaired and PCR-enriched. The resulting libraries were sequenced using the 100-bp paired-end module. RNA-seq data from each vole were mapped to HBB-T1 and HBB-T2 transcript references matching the sequences of the alleles of that particular vole using the CLC Genomics Workbench v. 6.0.1 (CLC bio A/S, Aarhus, Denmark). No mismatches were allowed, which permitted RNA-seq data of each vole to be unambiguously assigned to a specific paralogue. Gene expression levels were quantified as reads per kilobase per million mapped reads (RPKM) and fold difference in expression between the paralogues was tested using the proportions-based Z-test [25].

(d). Protein separation

To relate the results of the sequence analysis to Hb protein variation, we carried out electrophoresis of RBC lysate on starch gel with a tris–EDTA–borate buffer (TEB, pH 8.9). To separate the α- and β-polypeptides, we dissociated the Hb tetramers with hydrochloric acid–acetone solution [26] and analysed them by cellulose–acetate membrane electrophoresis with a TEB buffer (pH 8.9) in 8 M urea. We predicted the relative net charge differences between the globin polypeptides using the CLC Genomics Workbench.

(e). In vitro cysteine reactivity

To assess Hb Cys reactivity in an oxidizing environment, we incubated RBC lysate with diazenedicarboxylic acid bis-N,N-dimethylamide (diamide; 2 mM for 15 min; Sigma-Aldrich, St Louis, MO, USA), a thiol-oxidizing reagent [27] inducing Hb polymerization by forming intermolecular disulfide bonds via surface-exposed Cys residues [28,29]. The involvement of disulfide-bond formation was verified by subsequent incubation with 2-mercaptoethanol (5% for 15 min), a disulfide-reducing agent [29].

(f). Antioxidative capacity assay

The capacity of RBCs to resist attacks by peroxyl radicals was measured by the total radical-trapping antioxidative potential (TRAP). The TRAP measurement was performed as previously described [30,31]. Peroxyl radicals produced by thermal decomposition of 2,2-azo-bis-2-amidinopropane hydrochloride (ABAP; Polyscience, Warrington, PA, USA) added to the RBC lysate were monitored by luminol-enhanced chemiluminescence using Chameleon V luminometer (Hidex Oy, Turku, Finland). Trolox (Aldrich Chemicals, Milwaukee, WI, USA) was used as a reference antioxidant. The results were expressed as nmol trolox g−1 protein; the protein content of RBC lysate was determined by DC Protein assay (Bio-Rad, Hercules, CA, USA).

(g). Structural modelling and electrostatics calculation

A homology-based model of bank vole Hb was predicted and built by SWISS-MODEL [32], using the experimentally determined structure of mouse Hb (3HRW) as the template. Visualizations and renderings were done with PyMOL Molecular Graphics System v. 1.5.0.3. (Schrödinger, LLC). The prediction and rationalization of protein pKa values (PROPKA) method [33] and the electrostatic model of Jacob et al. [34] were used to calculate Cys acid dissociation constant (pKa). The calculations were performed with PyMOL scripts propka and cyspka, respectively.

(h). Genetic data analysis

Signatures of gene conversion between the paralogous genes were assessed by the method of Betrán et al. [35] implemented in DnaSP v. 5.10.01 [36], and using the program GENECONV [37]. Genotypes at all pairs of non-synonymous nucleotide substitutions (excluding alleles with conversion tracts) were tested for linkage disequilibrium (LD) [38] by permuting genotypes within loci with GENETIX v. 4.05.2 [39]. The genotype frequencies were assessed for conformity to Hardy–Weinberg (HW) expectations using the exact test [40], and the differentiation across populations and between population pairs was tested with a log-likelihood-based G-test [41], both performed with GENEPOP v. 4.2 [42]. A Bonferroni correction was applied to adjust for multiple comparisons.

The shape of the geographical cline in Hb frequency was estimated using an asymmetric stepped model assuming a sinusoidal shape for the cline centre and an exponential tail on each side [43,44]. The likelihood model was fitted by a Metropolis–Hastings algorithm using ClineFit v. 0.2 [45].

A maximum-parsimony haplotype network was constructed for each gene using the program TCS v. 1.21 [46] and drawn with the help of Haploviewer [47]. The phylogenetic relationships of bank vole β-globin genes with those of other rodents were reconstructed using the maximum-likelihood optimality criterion and evolutionary models providing the best fit according to the Akaike information criterion. Because gene conversion in the coding sequence (CDS) may confound the phylogenetic signal [48], reliable inference of orthologous relationships required separate analysis of a sequence partition containing only the 3′ UTR (137 bp of aligned sequence). The analyses were performed by the nearest-neighbour interchanges approach implemented in MEGA v. 5 [49] and using the T92 and T92+G evolutionary model for CDS and 3′ UTR, respectively.

3. Results and discussion

(a). The five globin genes

The survey of tetrameric Hb among bank voles in Britain [19] identified two distinct Hb variants, one migrating faster during gel electrophoresis (Hb F) than the other (Hb S). To identify the underlying sequence polymorphism, we isolated bank vole globin genes from reverse-transcribed mRNA by RACE. The sequences of cloned 5′- and 3′-RACE PCR products assembled into full-length cDNA sequences of three α-globin and two β-globin genes. The sequence of one α-globin gene matched that obtained from bank vole genomic DNA with deer mouse HBA-T2 paralogue-specific primers [5] and the sequence of one β-globin gene matched that amplified with house mouse HBB-T2 paralogue-specific primers, indicating the orthology of these bank vole genes to, respectively, HBA-T2 and HBB-T2 of other rodents. No other deer mouse or house mouse primer pair yielded a specific amplicon in the bank vole, but the cDNA sequence of one bank vole α-globin gene showed high sequence similarity to bank vole HBA-T2, while the third α-globin gene was distinguished by 20 amino acid substitutions. By analogy with the divergence among α-globin paralogues in other rodents, we refer to these bank vole genes as HBA-T1 and HBA-T3 [50]. The phylogenetic relationships of the 3′ UTR sequences confirmed that the bank vole HBB-T2 gene is orthologous to the deer mouse and house mouse HBB-T2 genes and the other bank vole β-globin gene to HBB-T1 (see the electronic supplementary material, figure S1). The results thus show that the bank vole has five adult globin genes, at least three of which demonstrably have orthologues in other rodent species.

(b). The underlying amino acid change

Comparison of the coding DNA sequences of 145 bank voles from 12 localities across Britain (figure 1; the electronic supplementary material, table S1) revealed non-synonymous polymorphism in all five genes. However, the electrophoretic analysis of the globin chains dissociated from the Hb tetramer revealed the presence of two distinct β-globin polypeptides distinguishing Hb S and Hb F, but no difference in α-globins (see the electronic supplementary material, figure S2a). We identified two non-synonymous polymorphisms among bank vole β-globin sequences, Ser52Cys and Ala58Val. The Ala58Val polymorphism segregated at HBB-T1, but the Ser52Cys polymorphism segregated at both HBB-T1 and HBB-T2. Thus, three amino acid sites were polymorphic in the two genes, sites 52 and 58 in HBB-T1 and site 52 in HBB-T2. By cloning and sequencing the genes in the 145 bank voles, we found perfect association between the Ser52Cys polymorphism in HBB-T1 and the two Hb variants, with 41 voles showing Hb S being homozygous for Ser, 93 voles with Hb F being homozygous for Cys, and 11 voles heterozygous for Hb S and Hb F being heterozygotes Ser/Cys. The degree of spatial concordance (figure 1) produced a significant LD between the HBB-T1 Ser52Cys polymorphism and the HBB-T2 Ser52Cys polymorphism in the total sample (p < 0.001), but there was no complete association between the genotypes at the two loci, and only 28 of the voles with Hb F and one heterozygote carried Cys at HBB-T2 (figure 1; electronic supplementary material, table S5). Similarly, despite significant LD between the two HBB-T1 sites (p < 0.001), 58Val was present in only 59 out of the 197 sequences that contained 52Cys, and there is evidence of at least one recombination event between the two sites, resulting in a haplotype 52Ser/58Val (see the electronic supplementary material, table S5 and figure S3). By contrast, there was no significant LD between any of the three β-globin sites and the amino acid polymorphisms in the HBA genes, either at individual localities or in the total sample.

As expected from these results, the geographical distributions of the 52Ser and 52Cys alleles at HBB-T1 matched the distributions of Hb S and Hb F (figure 1). The sampling in a region in northern England revealed a steep north–south frequency cline (figure 1), with the 52Ser frequency decreasing in a southwards direction, from 28% at NYK to 3% at DNC (see the electronic supplementary material, table S5 and figure S4). Only the HBB-T1 52Ser allele was present at the four northern localities and the HBB-T1 52Cys allele at the five southern localities, except for two voles from DEV carrying the 52Ser allele (figure 1; the electronic supplementary material, table S5). The presence of Hb S in DEV was not detected in the original survey [19], but is consistent with the relict occurrence of the ‘fringe’ mtDNA in the south of England [7].

The observed geographical pattern produced highly significant genotypic and allele frequency differences between the northern and southern localities for all three amino acid sites in HBB-T1 and HBB-T2 (p < 0.001), and although the genotype frequencies within samples did not deviate from HW expectations, there was a highly significant excess of homozygotes at both HBB-T1 sites in the total sample (p < 0.001). By contrast, the HBB-T2 and HBA sites showed no such deviation from HW expectations and no significant frequency differences between the north and south were detected for the HBA genes.

Unlike Ser52Cys in HBB-T1, the polymorphism at the same site in HBB-T2 did not show association with Hb S and Hb F (figure 1). However, at the nucleotide level, the variation of both HBB genes showed fully concordant geographical partitioning, with no haplotypes shared between the northern and southern localities, except at DEV, in contrast to the HBA genes, which showed no such partitioning, with the majority of high- and medium-frequency haplotypes shared between the northern and southern localities (see the electronic supplementary material, figure S3). Therefore, we found unambiguous association between the Ser52Cys polymorphism at the HBB-T1 gene and the Hb variants S and F, while the other two HBB polymorphisms show no such association, despite the concordant geographical partitioning of the synonymous and noncoding variation at HBB-T2.

Gene conversion was detected in three of the five bank vole globin genes, HBA-T1, HBA-T2 and HBB-T1 (see the electronic supplementary material, figure S3). Ten conversions between the β-globin genes with a median tract length of 34 bp (range 2–177 bp) were detected, affecting five HBB-T1 haplotypes. None of the conversion tracts, however, included one or the other polymorphic amino acid site, and, therefore, unless there has been a parallel mutation at the site β52, the sharing of the Ser52Cys polymorphism by HBB-T1 and HBB-T2 must stem from an undetected conversion event.

The RNA-seq analysis of three voles possessing different combinations of protein alleles at HBB-T1 and HBB-T2 indicated a highly disproportionate contribution of the two genes to β-globin synthesis, with HBB-T1 showing on average a 24-fold higher expression level than HBB-T2 (see the electronic supplementary material, table S6). Hb S and Hb F therefore represent allelic variants of the major Hb isoform, with β-chains encoded by alternate protein alleles segregating at HBB-T1. Depending on the HBB-T2 genotype, bank voles showing the Hb F phenotype may additionally synthesize a minor Hb isoform incorporating β-chains structurally identical to those in Hb S. Some bank voles showing Hb F therefore synthesize major and minor Hb isoforms that differ in the presence of β52Cys. Only a slightly lower (a few per cent) Cys concentration is, however, expected in RBCs of such individuals due to the large expression difference between HBB-T1 and HBB-T2 (see the electronic supplementary material, table S6).

(c). The structural and functional consequences and their mechanistic basis

These findings mean that if there is a functionally important structural difference between Hb S and Hb F, it must relate to the replacement of Ser with Cys in the major β-chain. Structurally, Cys only differs from Ser in the presence of a sulfur atom in place of oxygen, but this change gives Cys dramatically different physical–chemical properties. The thiol (–SH) group of Cys ionizes readily to yield a thiolate anion, and under appropriate conditions, Cys thiol may turn into a highly reactive functional group [34]. The thiolate can react with other thiol groups in an oxidation reaction, yielding a disulfide bond. Intramolecular disulfide bonds are important determinants of the structure of many proteins, but physiologically important intermolecular disulfides are formed by reaction of thiols exposed on the surface of proteins with other protein and non-protein thiols (see below). As reviewed by Reischl et al. [51], Cys is infrequent in mammalian Hb, with only approximately 5% of β-chain Cys being surface exposed. Completely exposed Cys residues were present in β-globin sequences (not α-globin) of only seven out of 155 species of mammals compared [51], including a marsupial, a tenrec, a prosimian primate, a shrew and three rodent species. In only one of the species was the exposed Cys at the same site β52 as in the bank vole Hb F, in the lesser hedgehog tenrec (Echinops telfairi), and we have found another example in published β-globin sequences of Cook's mouse (Mus cookii) [48]. The functional consequences were not determined, but it was suggested that in tenrec the presence of the external Cys might relate to the large changes in body temperature and metabolic rate accompanying the daily switches from activity to torpor exhibited by this species [52]. Oxidative polymerization by intermolecular disulfide bonds, which is an indicator of the presence of an exposed and reactive Cys [51], has been observed for tenrec Hb, similar to house mouse d- and p-type Hbs [53] possessing a reactive exposed Cys at β13 [29].

No polymerization was shown by fresh RBC lysate of either Hb S or Hb F. However, after oxidation by diamide, a high-molecular-weight smear is rapidly formed by Hb F, which is accompanied by a loss of the major band corresponding to tetrameric Hb (see the electronic supplementary material, figure S2b), indicating Hb polymerization [52,53]. The involvement of disulfide bonds is confirmed by the finding that the polymers are dissociated by the reduction with 2-mercaptoethanol [29,51]. Therefore, it is evident that bank vole Hb F, but not Hb S, readily undergoes oxidation to form intermolecular disulfides. Because Hb F only differs from Hb S by β52Cys, this must be due to the β52Cys thiol oxidation, demonstrating reactivity of its sulfur.

The reactivity of Cys is determined by its exposed localization and its ionization state, which in turn depends on the pKa of the thiol group: the lower the pKa, the higher the fraction of the thiolate anion at a given pH [34]. The intrinsic pKa value of Cys (approx. 9) would mean that majority of Cys residues would be protonated under the physiological pH. However, Cys pKa can be shifted much below this value due to a stabilization of the anion form by electrostatic interaction with the other side chain and/or backbone atoms [34]. For example, low pKa values of the highly reactive β125Cys in Hbs of guinea pig (7.4) and rat (6.9) have been attributed to the formation of a hydrogen bond between the thiolate and side-chain atoms of a proximate residue [28,54]. The homology-based three-dimensional model of the bank vole Hb F confirms the exposed localization of β52Cys (figure 2) and the estimated 28.3 Å2 of solvent-accessible surface area of the sulfur atom compares with those of the highly reactive Cys residues in guinea pig (25.6 Å2) and mouse (19.5 Å2) Hbs [54]. The bank vole β52Cys therefore is a completely exposed Cys, such as has only been identified in a few mammalian Hbs [51]. We estimated its pKa using two methods. While PROPKA seeks to identify key pKa determinants for every atom in the static 3D structure [33], the method of Jacob et al. [34] specifically infers Cys pKa by averaging over all possible side-chain conformations that the residue can adopt. The average pKa of 6.0 was inferred for β52Cys, with a further shift down to 5.3 predicted by an interaction with the water dipole [34]. The PROPKA analysis pointed to a hydrogen bond with the β50Ser side-chain oxygen as the major contributor to the pKa shift. These results indicate that the pKa of the bank vole β52Cys is substantially lowered due to the electrostatic interactions with its molecular environment. The predicted pKa is close to pKa values of highly reactive Cys residues in Hbs of guinea pig and rat [28,54], and it appears that the reactivity has a similar basis in all these rodents. The pKa around 5–6 means that a majority fraction (94%) of the cellular Cys population would be present in the anion form, which explains its reactivity. Given the millimolar Hb levels, RBCs of bank voles with Hb F contain a high concentration of the reactive thiolate anion compared to the voles with Hb S. But can such a high RBC concentration of reactive thiols be of a selective advantage under particular situations?

Figure 2.

Figure 2.

Open in a new tab

Model of the bank vole haemoglobin Hb F showing locations of the β52Cys sulfur atoms on the molecular surface. The two α-subunits are in cyan, the two β-subunits in blue and the β52Cys sulfur atoms in yellow.

(d). The antioxidative effect

RBCs face a constant risk of oxidative damage due to their role as an oxygen carrier. Glutathione (GSH) is the most abundant non-protein thiol in mammalian RBCs and central regulator of redox balance and signalling, and a reduced GSH pool is thus essential for RBC protection against reactive oxygen species (ROS). Under oxidative stress, GSH reacts with ROS producing glutathione disulfide (GSSG), which is enzymatically reduced back to GSH [55]. However, the evidence suggests that the highly reactive Cys in rat and mouse Hb take part in the regeneration of GSH through a thiol-disulfide exchange with GSSG [28,29]. Furthermore, Hb thiols can react with ROS directly, and because the molar concentration of Hb exceeds that of GSH by several times, ROS interception by reactive Hb thiols is likely an important antioxidative mechanism [28]. For example, the presence of reactive Cys in mouse Hb confers increased resistance to ROS-induced haem oxidation [21].

To test whether β52Cys increases the tolerance of bank vole RBCs to oxidative stress, we compared the TRAP values between voles with Hb S (n = 10) and voles with Hb F (n = 10). TRAP was threefold higher, on average, in the Hb F group (44.0 ± 7.6 nmol g−1 protein) than in the Hb S group (15.7 ± 6.7 nmol g−1 protein), demonstrating that RBCs containing Hb F significantly (Student's t-test unpaired p < 0.001) delayed the peroxyl-radical-induced luminol oxidation compared with RBCs containing Hb S (figure 3). RBCs of voles carrying Hb F thus have demonstrably higher antioxidative capacity than RBCs of voles carrying Hb S. Although the assay measures the total peroxyl-radical-trapping capacity and the TRAP values thus reflect the net effect of all antioxidant systems in RBCs, the major role played by the thiol-based GSH redox system, and the fact that all voles with Hb F had larger TRAP than any vole with Hb S (figure 3), strongly corroborate that the presence of β52Cys in Hb F conveys an increased capacity of blood protection against oxidative stress.

Figure 3.

Figure 3.

Open in a new tab

Antioxidative capacity of RBCs of bank voles carrying Hb S and Hb F. Representative kinetic traces are shown illustrating the effect on the chemiluminiscence (CPS) emitted from solution containing luminol (1 mM) and ABAP (300 mM). The inset displays the derived TRAP values for bank voles with Hb S (left, n = 10) and Hb F (right, n = 10). Box plots represent the first and third interquartile range with line denoting median and whiskers encompassing the entire spread of the data.

(e). The evolutionary origin of the polymorphism

The HBB-T1 52Cys is fixed or nearly so at localities through much of England (figure 1), in an area where the first colonizing mtDNA phylogroup was replaced by the second [7]. By sequencing the site β52 in 20 bank voles from The Netherlands, we found that they were all homozygous for 52Cys at HBB-T1, except for one vole that was a heterozygote, while both amino acids segregated in nearly equal frequencies at HBB-T2 (52Cys frequency of 0.45). Taking the Dutch population as representative of the second wave of colonization before it arrived in Britain, β52Cys was fixed, or nearly so, at HBB-T1 in that population by the time it arrived. On the other hand, the first colonists only carried β52Ser, as indicated by the absence of β52Cys from 10 bank voles from the vicinity of Lund in southern Sweden, which belongs to the same mtDNA phylogroup as the British ‘fringe’ population [7]. Thus, there may have been two distinct bank vole populations that colonized Britain at the end of the glaciation, with different mtDNA and Hb characteristics, and the Hb cline in northern England (figure 1) would, therefore, have become established at secondary contact of those two populations. It appears unlikely that an exposed Cys with such a low pKa as the bank vole β52Cys would have surfed stochastically to fixation at the wavefront of the initial range expansion into Britain. Cys residues are significantly underrepresented on the molecular surface of proteins and, as argued by Marino & Gladyshev [56], negative selection would quickly eliminate any newly evolved exposed Cys that is not functionally advantageous in order to prevent unwanted reactions with its thiol. It also appears unlikely that such a narrow Hb cline (approx. 30 km as estimated by ClineFit) as seen in Britain would have been maintained without any kind of selection for over 10 000 generations (assuming two generations per year and the last land connection to Britain 7.5 kyr ago). Instead, the physical–chemical properties of β52Cys and its consequences for RBC physiology strongly suggest that the Hb divergence in the bank vole evolved by natural selection.

(f). Evolutionary significance of the polymorphism

If the structural change distinguishing Hb F from Hb S has consequences for the blood-mediated metabolism of ROS, possessing Hb F would likely be of selective advantage under conditions when an antioxidant is needed. The production of ROS markedly increases during energetically demanding physiological states, for example muscular activity, increased growth rate or reproduction and under environmental conditions such as thermal stress [57]. The role of oxidative stress as the principal mechanism underlying life-history trade-offs [58], particularly with respect to reproductive costs [59], is currently a hotly debated topic [60]. With regards to small mammals, a recent study on laboratory mice suggests a positive relationship between female reproductive effort and oxidative damage [61], while another study found no increase in oxidative damage in mouse females weaning larger litters [62], or in breeding compared with non-breeding female bank voles [63]. Taking into account methodological differences [60], the lack of agreement between studies points to variation in the rates of endogenous antioxidant depletion. Antioxidative capacity also relates to the ability to cope with climatic stress. Hypoxaemia elicited by mismatch between oxygen supply and demand during exposure to extreme cold or heat leads to elevated ROS production, which can narrow the limits of thermal tolerance, depending on the efficiency of an individual's antioxidative defence [57]. Oxidative stress is also an important cost of intraerythrocytic parasitic infection, which is common in wild small mammals including bank voles [64] and has been shown to elicit species-specific responses of the GSH/GSSG redox system [65]. Furthermore, impairment of antioxidative mechanisms in mice plays an important role in the pathogenesis of glucose metabolism and diabetes [66]. Interestingly, wild-caught and captive bank voles from some populations are known to develop diabetes symptoms [67,68], in some cases coincident with high population density [69].

Given the various life-history correlates of ROS production, the physiological differences between Hb S and Hb F might lead to different trade-offs in terms of the oxidative cost. Further study will establish the effect of the two phenotypes at the level of whole organism performance and fitness, but the evidence presented here clearly demonstrates that members of the second wave of colonization carried more oxidative stress-resistant RBCs than the first colonists. Our study is therefore a striking demonstration of a functional divergence between populations originating in different glacial refugia, which show different performance during different stages of the climate change at the end of the last glaciation.

(g). Adaptive phylogeography

The current distributions of species depend critically on the progression of colonization with climate warming at the end of the last glaciation, when species became able to occupy terrain that was previously climatically inhospitable, particularly in high latitude areas [12]. In phylogeographic studies over the past 25 years, it has generally been assumed that the populations with defining genetic characteristics that first colonized particular empty territory at the end of the last glaciation retain those characteristics until the present day. However, our published findings and the data of others [7,8] suggest that genetic replacements may have occurred frequently during end-glacial colonization and therefore selection may have had a place in determining distributions of current genetic lineages. This leaves open the possibility for a genetic lineage emanating from a particular glacial refugium to have a wide current distribution not because of geographical factors (e.g. because the refugium is closer to colonized areas than other refugia) but rather because conditions in the refugium selected for particular characteristics advantageous during colonization. Clearly, attempting to understand whether adaptive differentiation and, hence, selection does have influences like this is what we mean by ‘adaptive phylogeography’. We contend that the experimental approach we have adopted in this paper is precisely what is needed in other systems to establish whether selection did have an important role in end-glacial colonization history and, thus, in establishing the current distributions of species.

Acknowledgements

Stephen Hall shared his past research experience with the haemoglobin of British bank voles, Jana Kopecká assisted with protein electrophoresis, Maik Jacob advised on interpreting the results from his model, and Troels Linnet provided advice on his cyspka script. Pim Arntzen supported the fieldwork in The Netherlands and Lars Råberg and Martin Andersson in Sweden.

Data accessibility

DNA sequences were deposited in GenBank under accession nos KJ677123–KJ677213.

Funding statement

The study was carried out with the financial support from the Grant Agency of the Academy of Sciences of the Czech Republic (grant no. IAA600450901) and the Czech Science Foundation (grant no. P506-11-1872) and with the institutional support RVO: 67985904.

References

  • 1.Avise JC, Arnold J, Ball RM, Bermingham E, Lamb T, Neigel JE, Reeb CA, Saunders NC. 1987. Intraspecific phylogeography: the mitochondrial DNA bridge between population genetics and systematics. Annu. Rev. Ecol. Syst. 18, 489–522. ( 10.1146/annurev.ecolsys.18.1.489) [DOI] [Google Scholar]
  • 2.Hewitt GM. 2000. The genetic legacy of the Quaternary ice ages. Nature 405, 907–913. ( 10.1038/35016000) [DOI] [PubMed] [Google Scholar]
  • 3.Brito P, Edwards SV. 2009. Multilocus phylogeography and phylogenetics using sequence-based markers. Genetica 135, 439–455. ( 10.1007/s10709-008-9293-3) [DOI] [PubMed] [Google Scholar]
  • 4.Deagle BE, Jones FC, Absher DM, Kingsley DM, Reimchen TE. 2013. Phylogeography and adaptation genetics of stickleback from the Haida Gwaii archipelago revealed using genome-wide single nucleotide polymorphism genotyping. Mol. Ecol. 22, 1917–1932. ( 10.1111/mec.12215) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Storz JF, Sabatino SJ, Hoffmann FG, Gering EJ, Moriyama H, Ferrand N, Monteiro B, Nachman MW. 2007. The molecular basis of high-altitude adaptation in deer mice. PLoS Genet. 3, e45 ( 10.1371/journal.pgen.0030045) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Bulgarella M, Peters JL, Kopuchian C, Valqui T, Wilson RE, McCracken KG. 2012. Multilocus coalescent analysis of haemoglobin differentiation between low- and high-altitude populations of crested ducks (Lophonetta specularioides). Mol. Ecol. 21, 350–368. ( 10.1111/j.1365-294X.2011.05400.x) [DOI] [PubMed] [Google Scholar]
  • 7.Searle JB, Kotlík P, Rambau RV, Marková S, Herman JS, McDevitt AD. 2009. The Celtic fringe of Britain: insights from small mammal phylogeography. Proc. R. Soc. B 276, 4287–4294. ( 10.1098/rspb.2009.1422) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Barnes I, Matheus P, Shapiro B, Jensen D, Cooper A. 2002. Dynamics of Pleistocene population extinctions in Beringian brown bears. Science 295, 2267–2270. ( 10.1126/science.1067814) [DOI] [PubMed] [Google Scholar]
  • 9.Hofreiter M, Münzel S, Conard NJ, Pollack J, Slatkin M, Weiss G, Pääbo S. 2007. Sudden replacement of cave bear mitochondrial DNA in the late Pleistocene. Curr. Biol. 17, R122–R123. ( 10.1016/j.cub.2007.01.026) [DOI] [PubMed] [Google Scholar]
  • 10.Cook A. 1975. Changes in carrion/hooded crow hybrid zone and possible importance of climate. Bird Study 22, 165–168. ( 10.1080/00063657509476460) [DOI] [Google Scholar]
  • 11.Herman JS, Searle JB. 2011. Post-glacial partitioning of mitochondrial genetic variation in the field vole. Proc. R. Soc. B 278, 3601–3607. ( 10.1098/rspb.2011.0321) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Hewitt GM. 1999. Post-glacial re-colonization of European biota. Biol. J. Linn. Soc. 68, 87–112. ( 10.1111/j.1095-8312.1999.tb01160.x) [DOI] [Google Scholar]
  • 13.Sommer RS, Zachos FE, Street M, Jöris O, Skog A, Benecke N. 2008. Late Quaternary distribution dynamics and phylogeography of the red deer (Cervus elaphus) in Europe. Q. Sci. Rev. 27, 714–733. ( 10.1016/j.quascirev.2007.11.016) [DOI] [Google Scholar]
  • 14.Franks SJ, Hoffmann AA. 2012. Genetics of climate change adaptation. Annu. Rev. Genet. 46, 185–208. ( 10.1146/annurev-genet-110711-155511) [DOI] [PubMed] [Google Scholar]
  • 15.Taberlet P, Fumagalli L, Wust-Saucy AG, Cosson J-F. 1998. Comparative phylogeography and postglacial colonization routes in Europe. Mol. Ecol. 7, 453–464. ( 10.1046/j.1365-294x.1998.00289.x) [DOI] [PubMed] [Google Scholar]
  • 16.Soltis DE, Morris AB, McLachlan JS, Manos PS, Soltis PS. 2006. Comparative phylogeography of unglaciated eastern North America. Mol. Ecol. 15, 4261–4293. ( 10.1111/j.1365-294X.2006.03061.x) [DOI] [PubMed] [Google Scholar]
  • 17.Searle JB, Wilkinson PJ. 1987. Karyotypic variation in the common shrew (Sorex araneus) in Britain—a ‘Celtic Fringe’. Heredity 59, 345–351. ( 10.1038/hdy.1987.141) [DOI] [Google Scholar]
  • 18.Piertney SB, Stewart WA, Lambin X, Telfer S, Aars J, Dallas JF. 2005. Phylogeographic structure and postglacial evolutionary history of water voles (Arvicola terrestris) in the United Kingdom. Mol. Ecol. 14, 1435–1444. ( 10.1111/j.1365-294X.2005.02496.x) [DOI] [PubMed] [Google Scholar]
  • 19.Hall SJG. 1979. Haemoglobin polymorphism in the Bank vole, Clethrionomys glareolus, in Britain. J. Zool. 187, 153–160. ( 10.1111/j.1469-7998.1979.tb03939.x) [DOI] [Google Scholar]
  • 20.Storz JF, Runck AM, Sabatino SJ, Kelly JK, Ferrand N, Moriyama H, Weber RE, Fago A. 2009. Evolutionary and functional insights into the mechanism underlying high-altitude adaptation of deer mouse hemoglobin. Proc. Natl Acad. Sci. USA 106, 14 450–14 455. ( 10.1073/pnas.0905224106) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Storz JF, Weber RE, Fago A. 2012. Oxygenation properties and oxidation rates of mouse hemoglobins that differ in reactive cysteine content. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 161, 265–270. ( 10.1016/j.cbpa.2011.11.004) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Brodsky I, Dennis LH, Kahn SB, Brady LW. 1966. Normal mouse erythropoiesis: I. The role of spleen in mouse erythropoiesis. Cancer Res. 26, 198–201. [PubMed] [Google Scholar]
  • 23.Storz JF, Baze M, Waite JL, Hoffmann FG, Opazo JC, Hayes JP. 2007. Complex signatures of selection and gene conversion in the duplicated globin genes of house mice. Genetics 177, 481–500. ( 10.1534/genetics.107.078550) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Yingzhong Y, Yue C, Guoen J, Zhenzhong B, Lan M, Haixia Y, Rili G. 2007. Molecular cloning and characterization of hemoglobin alpha and beta chains from plateau pika (Ochotona curzoniae) living at high altitude. J. Biochem. Mol. Biol. 40, 426–431. ( 10.5483/BMBRep.2007.40.3.426) [DOI] [PubMed] [Google Scholar]
  • 25.Kal AJ, et al. 1999. Dynamics of gene expression revealed by comparison of serial analysis of gene expression transcript profiles from yeast grown on two different carbon sources. Mol. Biol. Cell. 10, 1859–1872. ( 10.1091/mbc.10.6.1859) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Ferrand N. 1989. Biochemical and genetic studies on rabbit hemoglobin. I. Electrophoretic polymorphism of the beta chain. Biochem. Genet. 27, 673–678. ( 10.1007/BF02396059) [DOI] [PubMed] [Google Scholar]
  • 27.Kosower NS, Kosower EM. 1995. Diamide: an oxidant probe for thiols. Methods Enzymol. 251, 123–133. ( 10.1016/0076-6879(95)51116-4) [DOI] [PubMed] [Google Scholar]
  • 28.Rossi R, Barra D, Bellelli A, Boumis G, Canofeni S, Di Simplicio P, Lusini L, Pascarella S, Amiconi G. 1998. Fast-reacting thiols in rat hemoglobins can intercept damaging species in erythrocytes more efficiently than glutathione. J. Biol. Chem. 273, 19 198–19 206. ( 10.1074/jbc.273.30.19198) [DOI] [PubMed] [Google Scholar]
  • 29.Giustarini D, Dalle-Donne I, Cavarra E, Fineschi S, Lungarella G, Milzani A, Rossi R. 2006. Metabolism of oxidants by blood from different mouse strains. Biochem. Pharmacol. 71, 1753–1764. ( 10.1016/j.bcp.2006.03.015) [DOI] [PubMed] [Google Scholar]
  • 30.Hyršl P, Číž M, Kubala L, Lojek A. 2004. Silkworm (Bombyx mori) hemocytes do not produce reactive oxygen metabolites as a part of defense mechanisms. Folia Microbiol. 49, 315–319. ( 10.1007/BF02931049) [DOI] [PubMed] [Google Scholar]
  • 31.Slavíková H, Lojek A, Hamar J, Dušková M, Kubala L, Vondráček J, Číž M. 1998. Total antioxidant capacity of serum increased in early but not late period after intestinal ischemia in rats. Free Radical Biol. Med. 25, 9–18. ( 10.1016/S0891-5849(98)00030-6) [DOI] [PubMed] [Google Scholar]
  • 32.Arnold K, Bordoli L, Kopp J, Schwede T. 2006. The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22, 195–201. ( 10.1093/bioinformatics/bti770) [DOI] [PubMed] [Google Scholar]
  • 33.Li H, Robertson AD, Jensen JH. 2005. Very fast empirical prediction and rationalization of protein pKa values. Proteins 61, 704–721. ( 10.1002/prot.20660) [DOI] [PubMed] [Google Scholar]
  • 34.Jacob MH, Amir D, Ratner V, Gussakowsky E, Haas E. 2005. Predicting reactivities of protein surface cysteines as part of a strategy for selective multiple labeling. Biochemistry 44, 13 664–13 672. ( 10.1021/bi051205t) [DOI] [PubMed] [Google Scholar]
  • 35.Betrán E, Rozas J, Navarro A, Barbadilla A. 1997. The estimation of the number and the length distribution of gene conversion tracts from population DNA sequence data. Genetics 146, 89–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Librado P, Rozas J. 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25, 1451–1452. ( 10.1093/bioinformatics/btp187) [DOI] [PubMed] [Google Scholar]
  • 37.Sawyer S. 1989. Statistical tests for detecting gene conversion. Mol. Biol. Evol. 6, 526–538. [DOI] [PubMed] [Google Scholar]
  • 38.Weir BS. 1979. Inferences about linkage disequilibrium. Biometrics 35, 235–254. ( 10.2307/2529947) [DOI] [PubMed] [Google Scholar]
  • 39.Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F. 2004. GENETIX 4.05, logiciel sous Windows pour la génétique des populations (Laboratoire Génome, Populations, Interactions, CNRS UMR 5000, Université de Montpellier II, Montpellier, France).
  • 40.Guo SW, Thompson EA. 1992. Performing the exact test of Hardy-Weinberg proportion for multiple alleles. Biometrics 48, 361–372. ( 10.2307/2532296) [DOI] [PubMed] [Google Scholar]
  • 41.Goudet J, Raymond M, de Meeus T, Rousset F. 1996. Testing differentiation in diploid populations. Genetics 144, 1933–1940. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Rousset F. 2008. GENEPOP'007: a complete re-implementation of the GENEPOP software for Windows and Linux. Mol. Ecol. Resour. 8, 103–106. ( 10.1111/j.1471-8286.2007.01931.x) [DOI] [PubMed] [Google Scholar]
  • 43.Barton NH. 1983. Multilocus clines. Evolution 37, 454–471. ( 10.2307/2408260) [DOI] [PubMed] [Google Scholar]
  • 44.Szymura JM, Barton NH. 1986. Genetic analysis of a hybrid zone between the fire-bellied toads, Bombina bombina and Bombina variegata, near Cracow in southern Poland. Evolution 40, 1141–1159. ( 10.2307/2408943) [DOI] [PubMed] [Google Scholar]
  • 45.Porter AH, Wenger R, Geiger H, Scholl A, Shapiro AM. 1997. The Pontia daplidice-edusa hybrid zone in northwestern Italy. Evolution 51, 1561–1573. ( 10.2307/2411208) [DOI] [PubMed] [Google Scholar]
  • 46.Clement M, Posada D, Crandall KA. 2000. TCS: a computer program to estimate gene genealogies. Mol. Ecol. 9, 1657–1659. ( 10.1046/j.1365-294x.2000.01020.x) [DOI] [PubMed] [Google Scholar]
  • 47.Salzburger W, Ewing GB, von Haeseler A. 2011. The performance of phylogenetic algorithms in estimating haplotype genealogies with migration. Mol. Ecol. 20, 1952–1963. ( 10.1111/j.1365-294X.2011.05066.x) [DOI] [PubMed] [Google Scholar]
  • 48.Runck AM, Moriyama H, Storz JF. 2009. Evolution of duplicated beta-globin genes and the structural basis of hemoglobin isoform differentiation in Mus. Mol. Biol. Evol. 26, 2521–2532. ( 10.1093/molbev/msp165) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S. 2011. MEGA5: molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28, 2731–2739. ( 10.1093/molbev/msr121) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Marková S, Searle JB, Kotlík P. 2014. Relaxed functional constraints on triplicate α-globin gene in the bank vole suggest different evolutionary history from other rodents. Heredity. ( 10.1038/hdy.2014.12) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.Reischl E, Dafre AL, Franco JL, Wilhelm D. 2007. Distribution, adaptation and physiological meaning of thiols from vertebrate hemoglobins. Comp. Biochem. Physiol. C Toxicol. Pharmacol. 146, 22–53. ( 10.1016/j.cbpc.2006.07.015) [DOI] [PubMed] [Google Scholar]
  • 52.Piccinini M, Kleinschmidt T, Gorr T, Weber RE, Künzle H, Braunitzer G. 1991. Primary structure and oxygen-binding properties of the hemoglobin from the lesser hedgehog tenrec (Echinops telfairi, Zalambdodonta). Evidence for phylogenetic isolation. Biol. Chem. H.-S. 372, 975–989. ( 10.1515/bchm3.1991.372.2.975) [DOI] [PubMed] [Google Scholar]
  • 53.Riggs A. 1965. Hemoglobin polymerization in mice. Science 147, 621–623. ( 10.1126/science.147.3658.621-a) [DOI] [PubMed] [Google Scholar]
  • 54.Miranda JJ. 2000. Highly reactive cysteine residues in rodent hemoglobins. Biochem. Biophys. Res. Commun. 275, 517–523. ( 10.1006/bbrc.2000.3326) [DOI] [PubMed] [Google Scholar]
  • 55.Filipovska A, Murphy MP. 2006. Overview of protein glutathionylation. Curr. Protoc. Toxicol. 28, 6.10.1–6.10.8. ( 10.1002/0471140856.tx0610s28) [DOI] [PubMed] [Google Scholar]
  • 56.Marino SM, Gladyshev VN. 2010. Cysteine function governs its conservation and degeneration and restricts its utilization on protein surfaces. J. Mol. Biol. 404, 902–916. ( 10.1016/j.jmb.2010.09.027) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Pörtner HO. 2002. Climate variations and the physiological basis of temperature dependent biogeography: systemic to molecular hierarchy of thermal tolerance in animals. Comp. Biochem. Physiol. A Mol. Integr. Physiol. 132, 739–761. ( 10.1016/S1095-6433(02)00045-4) [DOI] [PubMed] [Google Scholar]
  • 58.Dowling DK, Simmons LW. 2009. Reactive oxygen species as universal constraints in life-history evolution. Proc. R. Soc. B 276, 1737–1745. ( 10.1098/rspb.2008.1791) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Selman C, Blount JD, Nussey DH, Speakman JR. 2012. Oxidative damage, ageing, and life-history evolution: where now? Trends Ecol. Evol. 27, 570–577. ( 10.1016/j.tree.2012.06.006) [DOI] [PubMed] [Google Scholar]
  • 60.Metcalfe NB, Monaghan P. 2013. Does reproduction cause oxidative stress? An open question. Trends Ecol. Evol. 28, 347–350. ( 10.1016/j.tree.2013.01.015) [DOI] [PubMed] [Google Scholar]
  • 61.Stier A, Reichert S, Massemin S, Bize P, Criscuolo F. 2012. Constraint and cost of oxidative stress on reproduction: correlative evidence in laboratory mice and review of the literature. Front. Zool. 9, 37 ( 10.1186/1742-9994-9-37) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Garratt M, Pichaud N, King EDA, Brooks RC. 2013. Physiological adaptations to reproduction. I. Experimentally increasing litter size enhances aspects of antioxidant defence but does not cause oxidative damage in mice. J. Exp. Biol. 216, 2879–2888. ( 10.1242/jeb.082669) [DOI] [PubMed] [Google Scholar]
  • 63.Ołdakowski Ł, Piotrowska Ż, Chrząścik KM, Sadowska ET, Koteja P, Taylor JRE. 2012. Is reproduction costly? No increase of oxidative damage in breeding bank voles. J. Exp. Biol. 215, 1799–1805. ( 10.1242/jeb.068452) [DOI] [PubMed] [Google Scholar]
  • 64.Karbowiak G, Rychlik L, Nowakowski W, Wita I. 2005. Natural infections of small mammals with blood parasites on the borderland of boreal and temperate forest zones. Acta Theriol. 50, 31–42. ( 10.1007/BF03192616) [DOI] [Google Scholar]
  • 65.Isaksson C, Sepil I, Baramidze V, Sheldon B. 2013. Explaining variance of avian malaria infection in the wild: the importance of host density, habitat, individual life-history and oxidative stress. BMC Ecol. 13, 15 ( 10.1186/1472-6785-13-15) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Styskal J, Van Remmen H, Richardson A, Salmon AB. 2012. Oxidative stress and diabetes: what can we learn about insulin resistance from antioxidant mutant mouse models? Free Radical Biol. Med. 52, 46–58. ( 10.1016/j.freeradbiomed.2011.10.441) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Schønecker B, Freimanis T, Sorensen IV. 2011. Diabetes in Danish bank voles (M. glareolus): survivorship, influence on weight, and evaluation of polydipsia as a screening tool for hyperglycaemia. PLoS ONE 6, e22893 ( 10.1371/journal.pone.0022893) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Bartelik A, et al. 2013. Development of hyperglycemia and diabetes in captive Polish bank voles. Gen. Comp. Endocr. 183, 69–78. ( 10.1016/j.ygcen.2012.12.006) [DOI] [PubMed] [Google Scholar]
  • 69.Niklasson B, Hörnfeldt B, Nyholm E, Niedrig M, Donoso-Mantke O, Gelderblom HR, Lernmarkd Å. 2003. Type 1 diabetes in Swedish bank voles (Clethrionomys glareolus): signs of disease in both colonized and wild cyclic populations at peak density. Ann. NY Acad. Sci. 1005, 170–175. ( 10.1196/annals.1288.020) [DOI] [PubMed] [Google Scholar]

Associated Data

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

Data Availability Statement

DNA sequences were deposited in GenBank under accession nos KJ677123–KJ677213.

Articles from Proceedings of the Royal Society B: Biological Sciences are provided here courtesy of The Royal Society

RESOURCES