https://orcid.org/0000-0002-1938-5970
Figures
Abstract
The wolverine (Gulo gulo) in Finland has undergone significant population declines in the past. Since major histocompatibility complex (MHC) genes encode proteins involved in pathogen recognition, the diversity of these genes provides insights into the immunological fitness of regional populations. We sequenced 862 amplicons (242 bp) of MHC class II DRB exon 2 from 32 Finnish wolverines and identified 11 functional alleles and three pseudogenes. A molecular phylogenetic analysis indicated trans-species polymorphism, and PAML and MEME analyses indicated positive selection, suggesting that the Finnish wolverine DRB genes have evolved under balancing and positive selection. In contrast to DRB gene analyses in other species, allele frequencies in the Finnish wolverines clearly indicated the existence of two regional subpopulations, congruent with previous studies based on neutral genetic markers. In the Finnish wolverine, rapid population declines in the past have promoted genetic drift, resulting in a lower genetic diversity of DRB loci, including fewer alleles and positively selected sites, than other mustelid species analyzed previously. Our data suggest that the MHC region in the Finnish wolverine population was likely affected by a recent bottleneck.
Citation: Sugiyama Y, Nishita Y, Lansink GMJ, Holmala K, Aspi J, Masuda R (2022) Diversity of the MHC class II DRB gene in the wolverine (Carnivora: Mustelidae: Gulo gulo) in Finland. PLoS ONE 17(5): e0267609. https://doi.org/10.1371/journal.pone.0267609
Editor: Vikas Sharma, Sant Baba Bhag Singh University, INDIA
Received: October 20, 2021; Accepted: April 8, 2022; Published: May 10, 2022
Copyright: © 2022 Sugiyama et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All nucleotide sequences are available from the DNA Data Bank of Japan (DDBJ) (accession numbers LC631158-LC631171).
Funding: This study was supported in part by a Joint Research Program grant from the Japan Arctic Research Network Center. The funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Wild animals, particularly large carnivores, often cause damage to livestock through predation [1], and problematic animals are harassed or killed [2], resulting in reductions in population size [3]. Depleted populations, especially larger mammals with slow reproductive rates, can then become vulnerable to extinction due to their reduced reproductive capacity [4, 5].
The wolverine (Gulo gulo), the largest species in the family Mustelidae, is widely distributed in the northern hemisphere, i.e. in Norway, Sweden, Finland, Russia, Mongolia, northern China, the USA, and Canada [6]. Its long-life span and low reproductive rate put populations at risk of extinction, especially those with high mortality due to poaching and legal culling [7–9]. This animal varies physically with their habitat, leading to differences in traits like average cranial size among populations in northeastern Siberia, Norway, and Alaska [10]. Wolverines in the Finnish population, the focus of our study, ranges from 69 to 83 cm in body length (excluding the 16–25 cm bushy tail), 40 to 45 cm in height at the withers, and 8 to 28 kg in weight [11]. The wolverine is an opportunistic, facultative predator, with its diet in Finland varying from large ungulates to small rodents [12, 13]. The wolverine’s distribution in northern Finland overlaps considerably with the reindeer husbandry area, and the high costs of predation in this area has led to conflict between humans and wolverines [14, 15].
The Finnish wolverine population fluctuated between 300 and 600 individuals at the end of the 1800s, began to decline in the early 1900s [16], and was reduced to 50 to 80 individuals by the 1980s [17]. In 1978, the Finnish government initiated legal protection for wolverines in all areas except the reindeer husbandry area. In 1982, with the estimated wolverine population reduced to tens of individuals, protection was extended to the whole country [18]. Wide protection led gradually to population recovery, with a marked increase since 2010 [19].
The IUCN lists for the wolverine in Europe as ‘Vulnerable’ [20], whereas in Finland, it is in the ‘Endangered’ category [21]. Currently comprising around 385 to 390 individuals, the population in Finland comprises two geographical subpopulations: 135 to 140 individuals in the reindeer husbandry area (northern subpopulation) and around 250 individuals in the area south of the reindeer husbandry area (eastern subpopulation) [22]. Analyses of microsatellite and mtDNA markers [19] have shown these two subpopulations to be genetically distinct.
Population-genetic studies on wolverines in northern Europe have utilized a variety of neutral genetic markers, including microsatellites, mtDNA, and single nucleotide polymorphisms [19, 23–25]. Finland has proven to be critical for the conservation of wolverine genetic diversity in northern Europe, as it forms a bridge between Russian and Scandinavian populations [2, 19]. Some studies have detected signs of a recent population bottleneck, reflected in an overall low genetic diversity [19, 26].
One aspect of population-genetic diversity relevant to the future viability of the wolverine in Europe is the degree of polymorphism in the major histocompatibility complex (MHC), which is related to immune fitness [27]. The MHC, the most polymorphic region in the vertebrate genome, evolved under positive and balancing selection, as shown by trans-species polymorphism (TSP) evident in phylogenetic analyses [28–30]. TSP results when alleles in an ancestral species are retained by selection in descendant species, leading to high genetic diversity in multiple genes, such as in the MHC [31].
The MHC genes, which encode proteins essential for pathogen recognition, comprise two major subfamilies, class I and class II. Class I genes mainly encode proteins presenting cytoplasmic antigens to cytotoxic T cells, whereas class II genes mainly encode proteins that present extracellular antigens to helper and regulatory T cells, thereby promoting antibody production by B cells. The functional class II protein is a heterodimer consisting of an α chain and a β chain, in the case of the DR subclass, encoded by DRA and DRB genes, respectively [32]. The antigen binding groove is formed by the α1 domain of the α chain and the β1 domain in the β chain. An amino acid residue that directly binds to an antigen is called an antigen binding site (ABS). In vertebrates, MHC polymorphism occurs predominantly at ABSs [33].
The loss of genetic diversity in a population increases its risk of extinction [34, 35]. For example, the Tasmanian devil (Sarcophilus harrisii) has a very low population size and low genetic diversity due to island and bottleneck effects. In this species, diversity is low in loci important for immune defense and self-recognition, such as the MHC and Toll-like receptor genes, and this is thought to contribute to the spread of diseases such as transmissible cancers [36]. MHC gene diversity is thus an important indicator in wildlife conservation research [27, 37]. The goal of our study was to investigate the genetic diversity and molecular evolution of MHC class II DRBs in Finnish wolverines, including trans-species polymorphism, by focusing on exon 2, which encodes ABSs.
Materials and methods
DNA samples
This study utilized DNA previously extracted from tissue samples from 32 individuals (Fig 1) that Lansink et al. [19] used to investigate wolverine genetic diversity and population structure in Finland. These individuals had either been hunted legally through derogation licenses, or had died in accidents and were delivered to the Natural Resource Institute Finland (Luke) research station. All individuals were obtained between 2012 and 2018, except for one from 2008 and one from 1983. The samples were assigned to “North” (N = 11), “East” (N = 19), and “Admixed” (N = 2) groups based on scores (q>0.8) for microsatellite data previously determined by using Structure [38] at K = 2 (for additional details, see [19]).
Each small circle on the map indicates one individual. Based on microsatellite data [19], individuals were assigned to regional subpopulations in the north (blue, N = 11) or east (red, N = 19), or were considered to be admixed (white, N = 2). Inside each small circle is the sample number. The dark gray area indicates the approximate range of wolverines in Fennoscandia [19]. The pie charts show allele frequencies (in percent) in the two subpopulations, with a different color (key at the right) for each allele; admixed individuals are included in the northern subpopulation.
Molecular analyses
Part of MHC class II DRB exon 2 (242 bp, excluding the primer sequences) was amplified by PCR using forward primer Meme-DRBex2F (CGTCCCCACAGGACATTTC; [39]) and a reverse primer Mulu-DRBex2R (CTCGCCGCTGCACCGTGAAG; [40]). Reactions were performed in 25 μL volumes, each containing 5 μL of 5× PrimeSTAR GXL buffer (Takara Bio), 2 μL of dNTP mixture (2.5 mM each), 0.3 μL of phosphorylated forward and reverse primers (25 μM), 0.5 μL (1.25 U/μL) of PrimeSTAR GXL (Takara Bio), and 1–2 μL of DNA extract as the template. Thermal conditions in a Takara Thermal Cycler Dice® Touch were 94 °C for 2 min; 32 cycles of 98 °C for 10 s and 60 °C for 15 s; and 68 °C for 30 s. PCR products were confirmed by electrophoresis on 3% agarose gels and visualization with ethidium bromide under ultraviolet illumination.
For detection of DRB alleles, which were expected to represent multiple loci, PCR products were purified with a QIAquick PCR Purification Kit (Qiagen), ligated into pBluescript II SK+ plasmid vector (Agilent Technologies), and transformed into Escherichia coli (JM 109) competent cells. To select cells with plasmids, bacteria were grown on 2xYT agar plates containing 50 μM ampicillin. Bacteria containing plasmids with the target PCR product were identified by blue/white selection and direct-colony PCR amplification using M13 forward and reverse primers [41]. White colonies with plasmids containing an insert were cultured in 2xYT liquid media, and plasmids were isolated with a NucleoSpin® Plasmid EasyPure column (Macherey-Nagel). Inserts were sequenced in both directions with M13 forward and reverse primers by using a BigDye Terminator v 3.1 Cycle Sequencing Kit (Thermo Fisher Scientific) and an ABI 3730 automated DNA sequencer (Applied Biosystems).
Data analyses
Nucleotide sequences obtained were aligned by using MEGA 7 [42]. Final insert sequences were identified as potentially genuine DRB exon 2 sequences if they matched in the forward and reverse directions, and were detected at least twice (twice in one individual or once each from at least two individuals). Single, unique sequences were excluded, as they may have been PCR chimeras or due to other PCR errors. We confirmed candidate sequences with BLAST-N searches [43] against the National Center for Biotechnology Information (NCBI) GenBank database. Sequences verified as functional wolverine DRB genes or pseudogenes were named according to the MHC gene nomenclature rules for non-human species [44]. Sequences comprising an open reading frame (ORF) were considered to be functional genes; those with indels and/or a premature stop codon(s) were considered to be pseudogenes.
To determine whether the frequencies DRB alleles differed between the northern and eastern regional subpopulations, Fisher’s exact test implemented in R 4.0.2 [45] was used. To check whether the mean number of alleles differed between populations, two-sample t-tests assuming unequal variances were performed with Microsoft Excel. Two individuals in the admixed group were included in the northern subpopulation due to their geographic location (Fig 1).
As an indication of selective pressure on protein-coding genes or partial sequences, the ratio (ω) of nonsynonymous (dN) to synonymous (dS) substitution rates was calculated for ABSs, non-ABSs, and all sites. HyPhy [46] implemented in MEGA was used to detect signs of positive selection. ABSs in the wolverine MHC class II DR β-chain were inferred from those in human HLA, originally determined by a 3D structural analysis [47]. CodeML in PAML 4.8 [48] implemented in PAMLX [49] was used to investigate positive selection among all sites based on maximum likelihood methods. Four models were tested (M1a, nearly neutral; M2a, positive selection; M7, beta; and M8, beta and ω) with likelihood ratio tests (LRTs) to determine which model best fitted our data [50–52]. Two nested models (M1a vs. M2a; M7 vs. M8) were compared by using LRTs. Positively selected sites were identified by using Bayes Empirical Bayes inference (BEB) [53]. Finally, to detect codons under positive selection, a mixed-effects model of evolution (MEME) [54] analysis was performed with Datamonkey 2.0, a web-based server for the HyPhy Package [55].
A Bayesian phylogenetic analysis was performed to reconstruct the relationships among DRB alleles from wolverines. The data set analyzed included alleles from Canadian and eastern Russian wolverines [56, 57], other mustelid species, and canid and felid species. Allele sequences other than those we determined were obtained from the GenBank database. To simplify the phylogenetic tree, identical sequences, regardless of species, were grouped together as a single terminal taxon. Gblocks [58] was used to eliminate inadequately aligned positions. Optimal substitution models were selected by using Kakusan4 [59] under Bayesian information criterion 4 (sample size = number of sites). The models selected were HKY85_Gamma (first and third codon positions) and K80_Gamma (second position). The phylogenetic analysis was performed by using MrBayes v3.2.7 [60] with default settings (sampling every 500 generations, with 25% burn-in), as our dataset was sufficiently large [61]. To gauge completeness of the analysis, trace files generated by the MCMC run were analyzed with Tracer v 1.7.1 [62], which determined the run to be sufficient at 80 million generations.
Results
Diversity of MHC class II DRB alleles
In all, 862 plasmid clones were sequenced from the 32 Finnish wolverines (average, 26.9 plasmid clones/individual). Eleven DRB exon 2 (242 bp) sequences with an open reading frame (ORF) were detected that represented functional alleles (Gugu-DRB*01–11) (Table 1). Another three 241 bp sequences represented presumed pseudogenes (Gugu-DRB*PS01-03), as all three contained a one-base deletion (Table 1) leading to a frame shift and premature stop codon(s). Five functional alleles (Gugu-DRB*07–*11) and one pseudogene (Gugu-DRB*PS2) were detected from only one individual each. All the sequences we obtained were deposited in the DNA Data Bank of Japan (DDBJ) under accession numbers LC631158–LC631171.
BLAST searches showed that five functional alleles (Gugu-DRB*01–*05) and one presumed pseudogene (Gugu-DRB*PS01) from the Finnish wolverine population were identical in 185 bp of overlapping sequence to DRB alleles found in Canadian and eastern Russian populations (Table 2) (the nucleotide sequences differed in length: 242 bp in our study; 185 bp for Canadian/Russian wolverines [56, 57]). Our alleles Gugu-DRB*08 and *11 and presumed pseudogene Gugu-DRB*PS01 were identical to those from other mustelid species (Table 2). The maximum number of alleles we found in a single individual was five, indicating at least three functional DRB loci per haploid genome in the wolverine. The most frequent functional allele was Gugu-DRB*01, followed by *02 and *03 (Fig 1 and Table 1). These three alleles were also high in frequency in the Canadian and eastern Russian populations. Because Gugu-DRB*02 and *03 always occurred together in the same individuals, they may represent two linked loci. All 32 wolverines we examined in the Finnish population contained presumed pseudogene Gugu-DRB*PS01, which was also the case in the Canadian population (reported as Gugu-DRB3*03 in [56]).
DRB allele frequencies differed between the two regional subpopulations in Finland (Fig 1). Gugu-DRB*01 was detected in all individuals from the northern subpopulation, but in 13 of 19 (68.4%) individuals from the eastern subpopulation (Table 1). By contrast, both Gugu-DRB*02 and Gugu-DRB*03 were detected in all individuals in the eastern subpopulation, but only 4 of 13 (30.8%) individuals from the northern subpopulation. Individuals having both Gugu-DRB*02 and Gugu-DRB*03 were significantly more frequent in the eastern than in the northern subpopulation (Fisher’s exact test, P < 0.01). Four alleles (Gugu-DRB*07–*10) were detected only in the eastern subpopulation and two (Gugu-DRB*05 and *11) only in the northern subpopulation. Significantly more individuals (Fisher’s exact test, P < 0.01) contained only a single functional allele in the northern subpopulation (6 of 13, or 46.2%) than in the eastern subpopulation (0 of 19) individuals. Finally, there was a significant difference in the mean number of DRB functional alleles per individual between the eastern (3.1 alleles) and northern (2.0 alleles) subpopulations (t-test, P < 0.05).
Selection on DRB
The dN/dS ratio (ω) for ABS codons in Finnish wolverine DRBs was close to one. Values close to one indicate neutral evolution, whereas those less than one indicate negative selection and greater than one, positive selection. Our result thus indicates that the ABS codons were maintained by neutral evolution (Table 3). By contrast, ω for non-ABS codons was less than one, indicating that negative selection acted on these codons. However, PAML analyses under the M2a (positive selection) and M8 (beta and ω) models provided evidence of positive selection (Table 4). Likelihood ratio tests (LRT) showed that the M2a and M8 models provided significantly better fits to our data than models without selection (Table 5). The M2a and M8 models identified four and five positively selected sites (P < 0.05; “+” in Fig 2), respectively, with all sites occurring in presumed ABS codons. Finally, the MEME analysis showed two codons to be under positive selection (P < 0.05; “+” in Fig 2), both of which were presumed ABS codons.
Numbers above the amino acid sequences indicate positions in the β1-domain of the DR protein β-chain. Dots indicate amino acids identical to those in Gugu-DRB*01. Residues in presumed antigen binding sites (ABSs) predicted from human MHC-HLA [47] are shaded in gray. + signs at the bottom of the table indicate sites inferred to be under positive selection by MEME analysis and Bayes Empirical Bayes inference (BEB) using PAML. For the M2a and M8 models in BEB, only significant results are indicated by + (P > 95%).
Phylogeny of DRB alleles
In the Bayesian phylogenetic tree (Fig 3), mustelid DRBs comprise a monophyletic group at the family level, but not at the species or genus levels. Gugu-DRB*05 from Finland is located in a small, near-basal clade with several other Gugu alleles and two from the sable (Martes zibellina); this clade forms an unresolved, near-basal polytomy involving other alleles from the sable, some from the European badger (Meles meles) and marbled polecat (Vormela peregusna), and the large clade containing most other mustelid DRB sequences. The other alleles from Finland were scattered among three distinct clades also containing DRB alleles from one to several other mustelid species. The three pseudogenes from Finland grouped in a clade containing only pseudogenes, including those from two other mustelid species. Both the occurrence of identical sequences between wolverines and other mustelid species (Table 2), as well as Gugu-DRBs more closely related to DRBs from other species than to those from wolverines (e.g., Gugu-DRB*11 and Mazi-DRB*14), are manifestations of TSP.
Genbank accession numbers in parentheses follow the allele names. Gulo gulo DRB sequences detected in our study are in red font. Numbers near nodes are posterior probability values. The scale at the bottom indicates branch length in substitutions per site. Species abbreviations in allele names are: Cafa, Canis familiaris; Cala, Canis latrans; Calu, Canis lupus; Febi, Felis silvestris bieti; Feca, Felis catus; Fesi, Felis silvestris; Gugu, Gulo gulo; Mele, Meles leucurus; Mazi, Martes zibellina; Meme, Meles meles; Muni, Mustela nivalis; Tata, Taxidea taxus; Vope, Vormela peregusna.
Discussion
Differences in DRB allele frequencies between the northern and eastern regional subpopulations
Neutral genetic markers are useful for studying gene flow, migration, or dispersal [63], whereas MHC genes are adaptive genetic markers useful as indirect indicators of the immunological fitness of populations [27]. In highly habitat-adapted populations, MHC genes show high polymorphism, which is maintained by balancing selection [29]. Due to balancing selection, molecular phylogeographic studies based on MHC loci are expected to show less population differentiation than those based on neutral loci [64].
In previous researches on wolverines and other large carnivores, population-genetic structuring examined using MHC loci has been negligible compared to that detected through microsatellite loci [57, 65]. Nevertheless, the distribution of MHC class II DRB alleles that we observed revealed genetic differences between the northern and eastern subpopulations of Finnish wolverines. These results are congruent with genetic differentiation between these subpopulations detected through microsatellites and mtDNA [19], which is believed to be linked to the severe demographic bottleneck in the 20th century. Wolverines recovered slower from the effects of the bottleneck than other sympatric species [19]. The bottleneck effect may have also contributed to genetic differentiation of the DRB genes between the regional subpopulations. The allele frequencies may have changed, for example, due to genetic drift after habitat loss and reduced gene flow.
Diversity of MHC class II DRB alleles
We found the genetic diversity of MHC class II DRB loci to be lower in wolverines than in other mustelid species. In previous studies, 17 functional alleles were found among 26 Japanese martens (Martes melampus) [66]; 27 functional alleles among 35 least weasels (Mustela nivalis) [67]; and 22 functional alleles among 31 Japanese weasels (Mustela itatsi) [68]. By contrast, we detected only 11 functional alleles among 32 wolverines in Finland. Although different lengths of sequence were obtained, five of the 11 alleles in Finland were identical in the region of overlapping sequence to alleles in Canada [56] and eastern Russia [57]. Three of these five alleles were high in frequency in all three regions.
Using next generation sequencing, a method different from ours, Oomen et al. [56] and Rico et al. [57] detected an additional five DRB alleles in Canada and eastern Russia, yet across the entire region from Canada to Finland, only 16 alleles have been detected in the wolverine. Compared to other terrestrial mammals (e.g. 25 DRB alleles from 163 grey wolves [69], and 66 alleles from 256 raccoons [70]), including other mustelids, the wolverine shows low MHC diversity throughout its circumpolar range.
Our study showed higher genetic diversity in the eastern than in the northern subpopulation. All individuals in the eastern subpopulation had more than one functional DRB allele, whereas six of 13 (46.2%) individuals in the northern subpopulation carried only one functional allele, i.e. the proportion of homozygous individuals was higher in the northern subpopulation. These observations corroborate previous results from microsatellites, where the eastern subpopulation had higher genetic diversity overall than the northern [19]. In Fennoscandia, gene flow occurs between the eastern Finnish and western Russian wolverine populations via the Karelian Isthmus, whereas populations on the Scandinavian Peninsula remain fairly isolated, with no indications of gene flow—a pattern similar to that observed for European brown bears (Ursus arctos) [71]. Gene flow may be responsible for the higher genetic diversity in the eastern than in the northern subpopulation in Finland. Heterogeneity of MHC genes increases an individual’s fitness, allowing more antigens to be recognized and increasing resistance to pathogens, e.g. Salmonella [72]. Our results thus suggest that the wolverines in eastern Finland may be more resistant to pathogens, especially novel or mutated ones, than their conspecifics in northern Finland. More detailed studies would be necessary to confirm this observation.
Selection and phylogeny of DRB alleles
Our results were ambiguous with regard to how DRBs have evolved in Finnish wolverines. The ω value was around one at ABS codons but less than one at non-ABS codons. This could be interpreted to mean that the ABS codons evolved neutrally and non-ABS codons under purifying selection. However, the interpretation of ω values can be complicated. For example, a mixture of advantageous and disadvantageous mutations may also result in values less than one [51]. On the other hand, our other analyses suggested that Gugu-DRBs have evolved under positive selection: the PAML analysis found evidence of positive selection; the maximum likelihood codon-based selection models identified four and five positively selected sites; and the MEME analysis identified two positively selected sites. All the positively selected sites from the latter two analyses are consistent with ABSs. Nevertheless, the selection pressure has been weak, as the number of sites under selection was lower than in other mustelid species: six sites in Martes melampus [66] and nine in Mustela nivalis [67]. Finally, we should note that our study inferred ABS positions by analogy with structural studies on human ABSs, and that the exact locations of ABSs in the wolverine MHC class II DR β-chain may differ from human.
Some wolverine DRB alleles were identical in nucleotide sequence to those in other mustelid species, and several alleles were phylogenetically more closely related to DRBs of other mustelid species than to wolverine DRBs. This TSP among evolutionarily related species supports the conclusion that MHC genes in Finnish wolverines have been subject to balancing selection [73].
Rico et al. [57] found weak population-genetic structuring for MHC loci in Canadian wolverines and suggested that MHC variation was primarily influenced by balancing selection rather than neutral processes. However, we detected differences in genetic structure between the two regional subpopulations in Finland based on DRB genes. Even though population structuring is less likely to be detected at loci evolving under balancing selection [64], the frequency differences in DRB loci between the Finnish regional subpopulations suggests that they might have been affected by neutral evolution. In other words, both neutral genetic drift and balancing selection may have acted simultaneously and antagonistically. Other cases of such antagonism have been reported in which genetic drift exceeded balancing selection [74, 75], or the converse [37].
In conclusion, both the study by Lansink et al. [19] using microsatellites and mtDNA to estimate genetic diversity, and our study on the diversity of MHC class II DRB genes, have shown that Finnish wolverine populations underwent a bottleneck that reduced diversity. Maintaining and increasing genetic diversity is important for the future conservation of wolverines in Finland.
Acknowledgments
We thank the technical personnel at the Luke Taivalkoski Research Station for help in obtaining the DNA samples. We are grateful to Dr. Matthew Dick for giving helpful comments and editing the manuscript.
References
- 1. Treves A, Karanth KU. Human-carnivore conflict and perspectives on carnivore management worldwide. Conserv Biol. 2003; 17:1491–1499.
- 2.
Boitani L, Alvarez F, Anders O, Andren H, Avanzinelli E, Balys V, et al. Key actions for large carnivore populations in Europe. Bruxelles: Institute of Applied Ecology (Rome, Italy), Report to DG Environment, European Commission; 2015.
- 3. Ordiz A, Bischof R, Swenson JE. Saving large carnivores, but losing the apex predator? Biol Conserv. 2013; 168:128–133.
- 4. Liow LH, Fortelius M, Bingham E, Lintulaakso K, Mannila H, Flynn L, et al. Higher origination and extinction rates in larger mammals. Proc Natl Acad Sci USA. 2008; 105:6097–6102. pmid:18417455
- 5. Bouzat JL. Conservation genetics of population bottlenecks: the role of chance, selection, and history. Conserv Genet. 2010; 11:463–478.
- 6.
Abramov AV. Gulo gulo (Linnaeus, 1758). The IUCN Red List of Threatened Species 2016. 2016; e.T9561A45198537. https://doi.org/10.2305/IUCN.UK.2016-1.RLTS.T9561A45198537.en
- 7. Persson J, Landa A, Andersen R, Segerström P. Reproductive characteristics of female wolverines (Gulo gulo) in Scandinavia. J Mammal. 2006; 87:75–79.
- 8. Persson J, Ericsson G, Segerström P. Human caused mortality in the endangered Scandinavian wolverine population. Biol Conserv. 2009; 142:325–331.
- 9. Rauset GR, Low M, Persson J. Reproductive patterns result from age-related sensitivity to resources and reproductive costs in a mammalian carnivore. Ecology. 2015; 96:3153–3164. pmid:26909422
- 10. Pasitschniak-Arts M, Larivière S. Gulo gulo. Mamm Species. 1995; 499:1–10.
- 11.
Pulliainen E. Suomen suurpedot. Helsinki: Tammi; 1974. Finnish.
- 12. Koskela , Kojola I, Aspi J, Hyvärinen M. Effect of reproductive status on the diet composition of wolverines (Gulo gulo) in boreal forests of eastern Finland. Ann Zool Fenn. 2013a; 50(1–2):100–106.
- 13. Koskela A, Kojola I, Aspi J, Hyvärinen M. The diet of breeding female wolverines (Gulo gulo) in two areas of Finland. Acta Theriol. 2013b; 58:199–204.
- 14. Rasmus S, Kojola I, Turunen M, Norberg H, Kumpula J, Ollila T. Mission impossible? Pursuing the co-existence of viable predator populations and sustainable reindeer husbandry in Finland. J Rural Stud. 2020; 80:135–148.
- 15. Pekkarinen AJ, Kumpula J, Tahvonen O. Predation costs and compensations in reindeer husbandry. Wildlife Biol. 2020;2020: 1–14.
- 16. Mykrä S, Pohja-Mykrä M. Back-calculation of large carnivore populations in Finland in 1865–1915. Ann Zool Fenn. 2015; 52:285–300.
- 17.
Landa A, Lindén M, Kokjola I. Action plan for the conservation of wolverines (Gulo gulo) in Europe. Strasbourg: Council of Europe Publishing; 2000.
- 18.
Pohja-mykrä M, Kurki S. Asialistalla ahma (Wolverine Management and Public Attitudes in Finland) Seinäjoki: Ruralia Institute; 2008. Finnish with English abstract.
- 19. Lansink GMJ, Esparza-Salas R, Joensuu M, Koskela A, Bujnáková D, Kleven O, et al. Population genetics of the wolverine in Finland: the road to recovery? Conserv Genet. 2020; 21:481–499.
- 20.
Andrén H. Gulo gulo (errata version published in 2019). The IUCN red list of threatened species. 2018: e.T9561A144336120. https://www.iucnredlist.org/species/9561/45198537
- 21.
Hyvärinen E, Juslén A, Kemppainen E, Uddström A, Liukko U. Suomen lajien uhanalaisuus—Punainen kirja 2019 (The 2019 Red List of Finnish Species). Helsinki: Ympäristöministeriö; 2019.
- 22.
Kojola I, Heikkinen S, Mäntyniemi S, Ollila T. Ahmakannan kehitys ja ahmakanta Suomessa 2020. Luonnonvara- ja biotalouden tutkimus 68/2020. Helsinki: Natural Resources Institute Finland; 2020.
- 23. Walker C, Vilá C, Landa A, Lindén M, Ellegren H. Genetic variation and population structure in Scandinavian wolverine (Gulo gulo) populations. Mol Ecol. 2001; 10:53–63. pmid:11251787
- 24. Flagstad Ø, Hedmark E, Landa A, Brøseth H, Persson J, Andersen R, et al. Colonization history and noninvasive monitoring of a reestablished wolverine population. Conserv Biol. 2004; 18:676–688.
- 25. Ekblom R, Aronsson M, Elsner-Gearing F, Johansson M, Fountain T, Persson J. Sample identification and pedigree reconstruction in wolverine (Gulo gulo) using SNP genotyping of non-invasive samples. Conserv Genet Resour. 2021; 13:261–274.
- 26. Ekblom R, Brechlin B, Persson J, Smeds L, Johansson M, Magnusson J, et al. Genome sequencing and conservation genomics in the Scandinavian wolverine population. Conserv Biol. 2018; 32:1301–1312. pmid:29935028
- 27. Ujvari B, Belov K. Major histocompatibility complex (MHC) markers in conservation biology. Int J Mol Sci. 2011; 12:5168–5186. pmid:21954351
- 28. Hedrick PW. Balancing selection and MHC. Genetica. 1999; 104:207–214.
- 29. Bernatchez L, Landry C. MHC studies in nonmodel vertebrates: what have we learned about natural selection in 15 years? J Evol Biol. 2003; 16:363–377. pmid:14635837
- 30. Těšický M, Vinkler M. Trans-Species Polymorphism in Immune Genes: General Pattern or MHC-Restricted Phenomenon? J Immunol Res. 2015;2015. pmid:26090501
- 31. Klein J. Origin of major histocompatibility complex polymorphism: the trans-species hypothesis. Hum Immunol. 1987; 19:155–162. pmid:3305436
- 32.
Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P. Molecular biology of the cell. 5th Ed. New York: Garland Science; 2007.
- 33. Brown JH, Jardetzky TS, Gorga JC, Stern LJ, Urban RG, Strominger JL, et al. Three-dimensional structure of the human class II histocompatibility antigen HLA-DR1. Nature. 1993; 364:33–39. pmid:8316295
- 34. Keller LF, Waller DM. Inbreeding effects in wild populations. Trends Ecol Evol. 2002; 17:230–241.
- 35. Frankham R. Genetics and extinction. Biol Conserv. 2005; 126:131–140.
- 36. Morris KM, Wright B, Grueber CE, Hogg C, Belov K. Lack of genetic diversity across diverse immune genes in an endangered mammal, the Tasmanian devil (Sarcophilus harrisii). Mol Ecol. 2015; 24:3860–3872. pmid:26119928
- 37. Pearson SK, Bull CM, Gardner MG. Selection outweighs drift at a fine scale: lack of MHC differentiation within a family living lizard across geographically close but disconnected rocky outcrops. Mol Ecol. 2018; 27:2204–2214. pmid:29603473
- 38. Pritchard JK, Stephens M, Donnelly P. Inference of population structure using multilocus genotype data. Genetics. 2000; 155:945–959. pmid:10835412
- 39. Sin YW, Dugdale HL, Newman C, Macdonald DW, Burke T. MHC class II genes in the European badger (Meles meles): characterization, patterns of variation, and transcription analysis. Immunogenetics. 2012; 64:313–327. pmid:22038175
- 40. Becker L, Nieberg C, Jahreis K, Peters E. MHC class II variation in the endangered European mink Mustela lutreola (L. 1761)–consequences for species conservation. Immunogenetics. 2009; 61:281–288. pmid:19263000
- 41. Messing J. New M13 vectors for cloning. Method Enzymol. 1983; 101:20–78.
- 42. Kumar S, Stecher G, Tamura K. MEGA7: Molecular Evolutionary Genetics Analysis version 7.0 for bigger datasets. Mol Biol Evol. 2016; 33:1870–1874. pmid:27004904
- 43. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. Basic local alignment search tool. J Mol Biol. 1990; 215:403–410. pmid:2231712
- 44. Klein J, Bontrop RE, Dawkins RL, Erlich HA, Gyllensten UB, Heise ER, et al. Nomenclature for the major histocompatibility complexes of different species: a proposal. Immunogenetics. 1990; 31:217–219. pmid:2329006
- 45.
R Development Core Team. R: A language and environment for statistical computing. R Foundation for Statistical Computing; 2020. https://www.r-project.org/
- 46. Pond SLK, Frost SDW, Muse SV. HyPhy: hypothesis testing using phylogenies. Bioinformatics. 2005; 21:676–679. pmid:15509596
- 47. Reche PA, Reinherz EL. Sequence variability analysis of human class I and class II MHC molecules: functional and structural correlates of amino acid polymorphisms. J Mol Biol. 2003; 331:623–641. pmid:12899833
- 48. Yang Z. PAML 4: phylogenetic analysis by maximum likelihood. Mol Biol Evol. 2007; 24:1586–1591. pmid:17483113
- 49. Xu B, Yang Z. PamlX: a graphical user interface for PAML. Mol Biol Evol. 2013; 30:2723–2724. pmid:24105918
- 50. Nielsen R, Yang Z. Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene. Genetics. 1998; 148:929–936. pmid:9539414
- 51. Yang Z, Bielawski JP. Statistical methods for detecting molecular adaptation. Trends Ecol Evol. 2000; 15:496–503. pmid:11114436
- 52. Yang Z, Nielsen R. Codon-substitution models for detecting molecular adaptation at individual sites along specific lineages. Mol Biol Evol. 2002; 19:908–917. pmid:12032247
- 53. Yang Z, Wong WSW, Nielsen R. Bayes empirical Bayes inference of amino acid sites under positive selection. Mol Biol Evol. 2005;22: 1107–1118. pmid:15689528
- 54. Murrell B, Wertheim JO, Moola S, Weighill T, Scheffler K, Pond SLK. Detecting individual sites subject to episodic diversifying selection. PLoS Genet. 2012; 8:e1002764. pmid:22807683
- 55. Weaver S, Shank SD, Spielman SJ, Li M, Muse SV, Pond SLK. Datamonkey 2.0: a modern web application for characterizing selective and other evolutionary processes. Mol Biol Evol. 2018; 35:773–777. pmid:29301006
- 56. Oomen RA, Gillett RM, Kyle CJ. Comparison of 454 pyrosequencing methods for characterizing the major histocompatibility complex of nonmodel species and the advantages of ultra deep coverage. Mol Ecol Resour. 2013; 13:103–116. pmid:23095905
- 57. Rico Y, Morris-Pocock J, Zigouris J, Nocera J, Kyle CJ. Lack of spatial immunogenetic structure among wolverine (Gulo gulo) populations suggestive of broad scale balancing selection. PLoS One. 2015; 10:1–21. pmid:26448462
- 58. Talavera G, Castresana J. Improvement of phylogenies after removing divergent and ambiguously aligned blocks from protein sequence alignments. Syst Biol. 2007; 56:564–577. pmid:17654362
- 59. Tanabe AS. Kakusan4 and Aminosan: two programs for comparing nonpartitioned, proportional and separate models for combined molecular phylogenetic analyses of multilocus sequence data. Mol Ecol Resour. 2011; 11:914–921. pmid:21592310
- 60. Ronquist F, Teslenko M, Mark PVD, Ayres DL, Darling A, Höhna S, et al. Mrbayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst Biol. 2012; 61:539–542. pmid:22357727
- 61.
Ronquist F, Huelsenbeck J, Teslenko M, Nylander JAA. MrBayes version 3.2 manual: tutorials and model summaries. 2020. https://github.com/NBISweden/MrBayes/blob/develop/doc/manual/Manual_MrBayes_v3.2.pdf
- 62. Rambaut A, Drummond AJ, Xie D, Baele G, Suchard MA. Posterior summarization in Bayesian phylogenetics using Tracer 1.7. Syst Biol. 2018; 67:901–904. pmid:29718447
- 63. Holderegger R, Kamm U, Gugerli F. Adaptive vs. neutral genetic diversity: implications for landscape genetics. Landsc Ecol. 2006; 21:797–807.
- 64. Schierup MH, Vekemans X, Charlesworth D. The effect of subdivision on variation at multi-allelic loci under balancing selection. Genet Res. 2000; 76:51–62. pmid:11006634
- 65. Niskanen AK, Kennedy LJ, Ruokonen M, Kojola I, Lohi H, Isomursu M, et al. Balancing selection and heterozygote advantage in major histocompatibility complex loci of the bottlenecked Finnish wolf population. Mol Ecol. 2013; 23:875–889. pmid:24382313
- 66. Hosotani S, Nishita Y, Masuda R. Genetic diversity and evolution of the MHC class II DRB gene in the Japanese marten, Martes melampus (Carnivora: Mustelidae). Mammal Res. 2020; 65:573–582.
- 67. Nishita Y, Kosintsev PA, Haukisalmi V, Väinölä R, Raichev EG, Murakami T, et al. Diversity of MHC class II DRB alleles in the Eurasian population of the least weasel, Mustela nivalis (Mustelidae: Mammalia). Biol J Linn Soc. 2017; 121:28–37.
- 68. Nishita Y, Abramov A V., Kosintsev PA, Lin LK, Watanabe S, Yamazaki K, et al. Genetic variation of the MHC class II DRB genes in the Japanese weasel, Mustela itatsi, endemic to Japan, compared with the Siberian weasel, Mustela sibirica. Tissue Antigens. 2015; 86:431–442. pmid:26593752
- 69. Seddon JM, Ellegren H. MHC class II genes in European wolves: A comparison with dogs. Immunogenetics. 2002;54: 490–500. pmid:12389097
- 70. Srithayakumar V, Castillo S, Rosatte RC, Kyle CJ. MHC class II DRB diversity in raccoons (Procyon lotor) reveals associations with raccoon rabies virus (Lyssavirus). Immunogenetics. 2011;63: 103–113. pmid:20924575
- 71. Kopatz A, Eiken HG, Aspi J, Kojola I, Tobiassen C, Tirronen KF, et al. Admixture and gene flow from Russia in the recovering Northern European brown bear (Ursus arctos). PLoS One. 2014;9: 1–10. pmid:24839968
- 72. Penn DJ, Damjanovich K, Potts WK. MHC heterozygosity confers a selective advantage against multiple-strain infections. Proc Natl Acad Sci USA. 2002; 99:11260–11264. pmid:12177415
- 73. Azevedo L, Serrano C, Amorim A, Cooper DN. Trans-species polymorphism in humans and the great apes is generally maintained by balancing selection that modulates the host immune response. Hum Genomics. 2015;9: 4–9. pmid:26337052
- 74. Miller HC, Lambert DM. Genetic drift outweighs balancing selection in shaping post-bottleneck major histocompatibility complex variation in New Zealand robins (Petroicidae). Mol Ecol. 2004; 13:3709–3721. pmid:15548285
- 75. Ploshnitsa AI, Goltsman ME, Macdonald DW, Kennedy LJ, Sommer S. Impact of historical founder effects and a recent bottleneck on MHC variability in Commander Arctic foxes (Vulpes lagopus). Ecol Evol. 2011; 2:165–180. pmid:22408734