Immunogenetics DOI 10.1007/s00251-011-0588-7
ORIGINAL PAPER
Evidence for evolutionary convergence at MHC in two broadly distributed mesocarnivores Vythegi Srithayakumar & Sarrah Castillo & Julien Mainguy & Christopher J. Kyle
Received: 2 August 2011 / Accepted: 31 October 2011 # Springer-Verlag 2011
Abstract Variation within major histocompatibility complex (MHC) genes is important in recognizing pathogens and initiating an immune response. These genes are relevant in enhancing our understanding of how species cope with rapid environmental changes and concomitant fluctuations in selective pressures such as invasive, infectious diseases. Disease-based models suggest that diversity at MHC is maintained through balancing selection arising from the coevolution of hosts and pathogens. Despite intensive balancing selection, sequence motifs or even identical MHC alleles can be shared across multiple species; three potential mechanisms have been put forth to explain this phenomenon: common ancestry, convergent evolution, and random chance. To understand the processes that maintain MHC similarity across divergent species, we examined the variation at two orthologous MHC-DRB genes in widespread North American Musteloid species, striped skunks (Mephitis mephitis), and raccoons (Procyon lotor). These species are often sympatric and exposed to a similar suite of diseases (e.g., rabies, canine distemper, and parvovirus). Given their exposure to similar selective pressures from pathogens, we postulated that similar DRB alleles may be present in both species. Our results indicated that similar motifs are present within both species, at functionally relevant polymorphic sites. However, Electronic supplementary material The online version of this article (doi:10.1007/s00251-011-0588-7) contains supplementary material, which is available to authorized users. V. Srithayakumar (*) : S. Castillo : C. J. Kyle Trent University, Peterborough, ON, Canada e-mail:
[email protected] J. Mainguy Ministère des Ressources naturelles et de la Faune, Direction de l’expertise sur la faune et ses habitats, Service de la biodiversité et des maladies de la faune, 880, chemin Sainte-Foy, Québec, QC, Canada
based on phylogenetic analyses that included previously published MHC sequences of several closely related carnivores, the respective MHC-DRB alleles do not appear to have been maintained through common ancestry and unlikely through random chance. Instead, the similarities observed between the two mesocarnivore species may rather be due to evolutionary convergence. Keywords MHC class II-DRB . Trans-species polymorphism . Convergent evolution . Procyon lotor . Mephitis mephitis . Wildlife disease
Introduction Environmental and anthropogenic factors have been associated with increases in pathogen infectivity, virulence, and immunogenicity in the wild (Harvell et al. 2002; Patz et al. 2005; Daszak et al. 2000; Rachowicz et al. 2005). Examples include several broadly distributed, invasive wildlife diseases such as chronic wasting disease, West Nile virus, parvovirus, and rabies (Daszak et al. 2000). Similar factors have also facilitated demographic and distribution range expansions in many wildlife species (Parmesan 2006; Pearson and Dawson 2003; Carroll et al. 2004); therefore, increasing risks of disease spread within and among species. In this context, rapid evolutionary adaptation is critical for wildlife to respond to these changing selective pressures. Understanding the mechanisms that drive adaptation is central in both evolutionary biology and conservation and imparts knowledge that may assist in stemming the spread of these diseases. Given their functional and evolutionary significance, major histocompatibility complex (MHC) genes are ideal candidates for immunogenetic studies (Bernatchez and Landry 2003). MHC is involved in the initiation and regulation of the immune response by peptide presentation
Immunogenetics
to the T lymphocytes (T cells) (Bernatchez and Landry 2003) and comprises two major subsets, classes I and II genes that are tightly linked together to form a single complex (Alberts et al. 2002; Hughes and Yeager 1998). In vertebrates, MHC consists of the most polymorphic genes of the genome with most of the polymorphism concentrated at peptide-binding region (PBR: regions involved in pathogen binding) (Bernatchez and Landry 2003; Hughes and Yeager 1998). The origin and maintenance of extensive diversity at MHC genes in vertebrates has been the subject of considerable speculation and controversy (as reviewed in Klein 1986; Klein et al. 2007). Despite extensive MHC polymorphism within species, sharing of similar or even identical sequence motifs across multiple mammalian species has been documented (Klein 1986; Gustafsson and Andersson 1994; Otting et al. 2002; Wei et al. 2010). Three potential mechanisms have been proposed to explain this phenomenon: (1) common ancestry, (2) convergent evolution, and (3) similarity by chance. Under common ancestry, allelic lineages would be passed from one species to the next along an evolutionary line, sometimes extending across taxonomic orders (e.g., MHC similarity between human and mouse who shared a common ancestor 110 mya) (Klein 1986; Klein et al. 1993; Lundberg and McDevitt 1992; Yeager and Hughes 1999). This trans-species polymorphism has been documented in primates (Otting et al. 2002; Huchard et al. 2006), rodents (Cutrera and Lacey 2007), ungulates (Sena et al. 2003; Janova et al. 2009), Sphenisciformes (Kikkawa et al. 2009), lagomorphs (Goüy de Bellocq et al. 2009), carnivores (Wei et al. 2010), elephants (Archie et al. 2010), and anurans (Kiemnec-Tyburczy et al. 2010). Unlike shared ancestry, convergent evolution is reflected in the emergence of similar traits with similar functions, evolving independently in two or more lineages from different ancestral states (Gustafsson and Andersson 1994; Hughes 1999). The similarity of MHC alleles has thus also been explained via convergent evolution arising as a result of adaptation to similar selective pressures (Gustafsson and Andersson 1994; Kriener et al. 2000; Xu et al. 2008). Under the third potential mechanism, allelic similarities between species would simply arise by chance (Kriener et al. 2000). In order to gain more insight into the processes that maintain similar sequence motifs between species, we examined the allelic diversity of orthologous MHC class II-DRB gene in two wildlife species, striped skunks (Mephitis mephitis) and raccoons (Procyon lotor). These two mesocarnivores share similar demographic and life history traits (Rosatte et al. 2001), and are from the same superfamily, Musteloids, sharing a common ancestor approximately 33 mya (Yonezawa et al. 2007; Fulton and Strobeck 2007). Historically, skunks and raccoons occupied different ecological niches and distributions, with skunks
associated with dry habitats and raccoons more closely associated with marshlands and wet environments (Hall and Kelson 1959). The advent of human settlement and agriculture resulted in distributional expansions and adaptation to human-associated food sources for these animals (Broadfoot et al. 2001; Rosatte 2000). As such, the recent proximity and sympatry of these species has led to an increase in exposure to a common suite of diseases (i.e., rabies, canine distemper, infectious canine hepatitis, leptospirosis, tularaemia, trypanosomiasis, tuberculosis, listeriosis, histoplasmosis, and various forms of encephalitis) (Sanderson 1987; Rosatte 1987; Zeveloff 2002). Hence, these species can present a serious threat to human and ecosystem health as vectors of these diseases (Broadfoot et al. 2001; Rosatte 2000). In addition to the infectious diseases, these species share a similar fauna of internal and external parasites, which can further compromise their immune system and affect reproduction (Sanderson 1987; Rosatte 1987; Zeveloff 2002; Hall and Kelson 1959). Among the different diseases that are known to affect both skunks and raccoons, the raccoon variant of rabies (RRV) is presumed to be a strong selective force on the survival of these two mesocarnivores, especially along the eastern seaboard of North America, where they are often sympatric. RRV is a RNA virus (Lyssavirus genus) that can infect mammals, including humans, via the central nervous system through the peripheral nerves, causing acute encephalitis (Jackson and Wunner 2007). Although raccoons are the main reservoir species of RRV, the frequency of infection in skunks suggests that this vector also plays a vital role in disease maintenance and transmission (Rosatte et al. 1992). Furthermore, according to past control activities the immune response of skunks likely differs from that of raccoons as rabies positive skunks often appear clinically normal prior to euthanasia (Rosatte 1988) and they do not seroconvert equally when administered rabies vaccines (Rupprecht et al. 1988). Studies focusing on immunoprotection against rabies, have found that apart from T cell-mediated immunity (through MHC class I), the virus neutralizing antibodies (through MHC class II) play a major role in successful clearance of this virus (Dietzschold et al. 1992). MHC class II DRB-exon 2 has been previously characterized in raccoons (Prlo-DRB) and was found to be duplicated in raccoons, with high levels of polymorphism and sequence diversity between alleles (Castillo et al. 2010). Associations between specific Prlo-DRB alleles and RRV has also been established by Srithayakumar et al. (2011), highlighting the fact that variation at MHC genes play an important role in how raccoons respond to pathogens. In the present study, sequences of MHC class II DRBexon 2 were elucidated in skunks (Meme-DRB) and compared with those of raccoons. Given the similarity of
Immunogenetics
diseases encountered by these animals, we hypothesized that high levels of amino acid similarity should exist between these species, particularly at sites involved in pathogen binding as per the crucial role of MHC in initiating an immune response. In order to assess whether similarities present are driven by their phylogenetic relationship or similar selective pressures, we included previously published MHC class II DRB-exon 2 sequences of European mink (Mustela lutreola; Mulu-DRB) (Becker et al. 2009). These three species are from the same superfamily, Musteloids, and cluster together on phylogenetic trees based on mitochondrial and nuclear markers (Yonezawa et al. 2007); however, they are presumably subjected to different selective pressures as the mink included in this study are from a different continent. Given the phylogenetic relationship between these animals, we expect trans-species polymorphism to be maintaining similarities at DRB gene which would be evident through intermixing of the DRB sequences on the phylogenetic tree. Alternatively, the recent exposure to similar selective pressures in raccoons and skunks (Hall and Kelson 1959) may favor convergent evolution which would be evident through high degree of similarity in amino acid sequences and differential codon usage between raccoons and skunks than mink. The presence similar MHC class II-DRB alleles across divergent species living in different environments suggests that there may be structural constraints on the evolution of MHC polymorphism and certain sequence motifs are particularly favored (Gustafsson and Andersson 1994; Andersson et al. 1991). In order to assess this, we included sequences from the well-characterized Canids (DLA-DRB) (Sarmiento et al. 1990; Wagner et al. 1996, 1998; Francino et al. 1997; Kennedy et al. 1998, 2000). If similarity exists between Musteloids and Canids, this indicates that structural constraints on the MHC molecules are one of the leading causes of convergence in these species, rather than selective pressures. This broad-scale study was therefore designed to examine MHC diversity in common wildlife species and has the potential to deepen our understanding of the processes (e.g., variation of immunological functions, evolutionary constraints and selective forces) that affect MHC variation within and between species.
Materials and methods Sample collection For the purpose of this study, it would have been ideal to include samples that were infected with several diseases known to occur in both raccoons and skunks; however, as that was unattainable, we attempted to control for this variable by
including samples from regions where the raccoon strain of rabies is presumed to be the strongest selective force (where epizootic RRV results in ∼85% mortality of population (Blanton et al. 2007)). Samples of brain tissue were obtained from 15 skunks that were known to be infected by RRV (confirmed by direct fluorescent antibody testing of brain tissue) from Ontario and southern Quebec, Canada. DNA was extracted from these samples using DNeasy blood and tissue kit (Qiagen) and quantified using PicoGreen® (Invitrogen, Burlington, Canada) following manufacturer’s protocols. Molecular cloning, sequencing, and nomenclature A 228-bp fragment of MHC class II DRB exon 2 of the sampled skunks was amplified from genomic DNA using Taq DNA polymerase enzyme (Invitrogen) and primers DRB-5c, TCAATGGGACGGAGCGGGTGC (Gillett 2009) and DRB-3c, CCGCTGCACAGTGAAACTCTC (Murray and White 1998) using the conditions described previously in Castillo et al. (2010). Two microliters of the amplified product was subjected to agarose gel electrophoresis to confirm the amplification of the desired size fragment. MHC-DRB alleles were characterized by cloning and direct sequencing DNA from individual clones. A TOPO TA cloning kit (Invitrogen) was utilized for cloning the amplified DRB gene. The manufacturer’s instructions were followed with the exception of 30-min incubation at room temperature instead of 5 min, as it improved the transformation efficiency. Approximately 26–30 clones were picked per sample and boiled for 10 min at 100°C to lyse cells. A 10-μl PCR was prepared and amplified using M13 forward, CAGGAAACAGCTATGAC, and reverse, CTGGCCGTCGTTTTAC, (Invitrogen) primers. The PCR components were in the following concentrations: 1× PCR buffer, 0.04 mM of each dNTP, 1.5 mM MgCl2, 0.2 μM of forward primer, 0.2 μM of reverse primer, 0.05 U/μL of Taq DNA polymerase, 10 ng of DNA and ddH2O for a total volume of 10 μl. Two microliters of the amplified product was subjected to agarose gel electrophoresis to confirm the insertion of the PCR fragment into the vector. As MHC-DRB gene was duplicated in skunks (between two and four alleles/individual, similar to the findings in raccoons (Castillo et al. 2010)), a homogenous discretetime Markov Chain was used to determine that 16 clones needed to be sequenced per individual to provide 95% confidence that all alleles within each individual were screened (Breuer and Baum 2005). PCR products were purified using ExoSAP (NEB) following the manufacturer’s instructions. BigDye® Terminator v3.1 Cycle Sequencing Kit (ABI) and the M13 forward primer were used to sequence the transformed clones. Sequenced PCR products were ethanol-precipitated, then electrophoresed and visualized on an ABI™ 3730 DNA Analyser. Direct sequencing
Immunogenetics
of PCR products from genomic DNA was also performed in the same manner to cross-validate the location of the variable sites relative to cloned products. Due to the high amount of variation, and to account for sampling error, sequences were only considered as alleles if the same sequence was identified in more than one individual or in two independent PCRs. These strict criteria were followed to validate DRB alleles as in vitro recombination during PCR, and heteroduplex mismatch repair during cloning, commonly occur when characterizing DRB alleles by cloning (Longeri et al. 2002). All skunk MHC-DRB sequences (Meme-DRB) were named following the nomenclature set out by Klein et al. (1990). Functionality of Meme-DRB To evaluate the functionality of the amplified locus, we screened for the expression of this gene in two individuals. Total RNA was isolated from lymph nodes using a Trizol-LS (Invitrogen) following the manufacturer’s instructions. Extracted RNA was further treated with DNase enzyme (NEB) according to the manufacturer’s protocol to remove any residual DNA and cleaned using an isopropanol precipitation. cDNA was constructed using a ThermoScript RT-PCR system (Invitrogen) following the manufacturer’s instructions. cDNA was used as template for PCR, and sequenced following the parameters previously outlined. Expression was assessed through gel electrophoresis of the PCR and as well as the sequenced products. Data analysis All nucleotide sequences were edited and aligned to the DRB gene of other species (human, cat, dog, raccoon) using MEGA™ 4.1 (Tamura et al. 2007). Average pairwise nucleotide distances (Kimura 2 parameter model or K2P), Poisson-corrected amino acid distances and average rate of synonymous (dS; silent) and nonsynonymous (dN; nonsilent) substitutions per site were computed in MEGA 4 (Tamura et al. 2007) using the modified Nei–Gojobori method with the Jukes-Cantor correction for multiple substitutions (Nei and Gojobori 1986). In order to reveal evidence for selection acting to maintain polymorphism, the presence of positively selected codon sites (PSS) were investigated in Meme-DRB amino acid sequences. PSS are characterized by a ratio ω=dn/ds >1. A null model, model M7 (ω1), were compared by a likelihood ratio test (Yang et al. 2000). These models were chosen as they have been found to be more robust against recombination in the sequences than other models (Anisimova et al. 2003). In the instance where M8 fits the dataset better than
M7, PSS were subsequently identified using an empirical Bayesian approach. These analyses were performed using the software CODEML, implemented in the package PAML version 3.14 (Yang 1997). The rates of synonymous and nonsynonymous substitutions were calculated separately for putative peptide-binding region (PBR) and non-PBR of all identified alleles assuming concordance with human PBR (Brown et al. 1988, 1993; Stern et al. 1994) and then for PSS and non-PSS. The relative rates of synonymous and nonsynonymous substitutions were determined following modified Nei–Gojobori approach and Jukes–Cantor correction (to account for multiple hits) using MEGA™ 4.1 (Tamura et al. 2007). A codon-based Z test of selection was also performed using MEGA™ 4.1 (Tamura et al. 2007). Molecular evidence for trans-species polymorphism was examined through phylogenetic analysis. We included the 67 previously characterized DRB alleles (GenBank accession numbers: GU388312–GU388377; HM589210; Castillo et al. 2010; Srithayakumar et al. 2011) of raccoons, and 9 European mink DRB alleles (GenBank accession numbers: EU263550EU263558; Becker et al. 2009); giant panda (EF125965) and sea lion (AY491456) were included as outgroups based on sequence identity. In order to avoid relying on single best-fit substitution models, jModelTest (Guindon and Gascuel 2003; Posada 2008) was used to determine the model of molecular evolution that best fit our data (HKY+I+G). Mr. Bayes (Huelsenbeck and Ronquist 2001; Ronquist and Huelsenbeck 2003) was used to construct the phylogenetic tree, and the analyses were run for 6×107 generations, sampling every 2,000 generations. Branch support was assessed through Markov Chain Monte Carlo methodology (Mau et al. 1999). Nucleotide and amino acid sequences of skunk, raccoons, and mink DRB alleles were then compared to examine molecular signatures of convergent evolution. In order to assess if DRB polymorphism was restricted by structural constraints we included the 93 DLA-DRB alleles (n>800, Calu-DRB) (Sarmiento et al. 1990; Wagner et al. 1996, 1998; Francino et al. 1997; Kennedy et al. 1998, 2000; 2008). Furthermore, amino acid residues in DRBexon 2 involved in forming particular pockets were determined, again assuming concordance with human DRB structure (Kaufman et al. 1984; Stern et al. 1994; Suarez et al. 2006). Variability at the particular pockets was evaluated as to whether similar motifs at particular pockets were shared between species.
Results Characterization of Meme-DRB A 228 bp fragment of MHC class II DRB-exon 2 was amplified from 15 skunks. A total of 245 clones were
Immunogenetics
sequenced and 12 unique sequences (Meme-DRB*01Meme-DRB*06 and Meme-DRB*08- Meme-DRB*13) were identified (GenBank accession: JN193535- JN193546) (see Electronic supplementary material). Two pairs of alleles encoded for the same amino acid sequence (Meme-DRB*8 & Meme-DRB*10; Meme-DRB*9 & Meme-DRB*11); however, these alleles were present in more than one individual and in multiple clones. The number of pairwise nucleotide differences between pairs of Meme-DRB alleles were 1 to 18 and the number of amino acid differences ranged from 0 to 11 (Table 1). Meme-DRB is likely duplicated in skunks, as two to four alleles were detected per individual (specifically 33% with two alleles, 27% with three alleles, and 40% with four alleles); however, we were unable to assign alleles to an individual locus as interlocus exchange is known to occur at MHC loci (Andersson et al. 1991). Therefore, we considered all alleles to be representative of the DRB locus for subsequent analyses. Duplication has been documented in other mammalian species (e.g., raccoons (Castillo et al. 2010), sea lion (Bowen et al. 2004), and cats (Kennedy et al. 2002)) with majority of loci being functional. None of the alleles from these amplified loci exhibited features compatible with it being a pseudogene. Furthermore, the functionality was assessed by confirming the expression of the gene using both DNA and RNA from two individuals. Same MHC-DRB sequences were derived from both DNA and RNA, suggesting the data presented using gDNA relates to transcribed genes. Evidence for selection process at Meme-DRB Evidence for selection maintaining polymorphism at MemeDRB was investigated using the two approaches outlined in Srithayakumar et al. (2011). First, maximum likelihood analysis was employed to identify species-specific codon sites affected by positive selection (PSS). The likelihood ratio test statistic comparing M7 and M8 indicated that our data best fitted the M8 model, thereby rejecting the null hypothesis (neutral evolution; χ2 =15.62, df=2, p
