DNA extracts were obtained using the HotSHOT (Truett et al., 2000) or salt extraction protocols (Talbot et al., 2011). In most cases, we used blood samples, except for a few instances: in the case of peregrine falcons, four of the DNA samples were extracted from muscle tissue, four from museum skins (skin and feather; Aleutian Islands), one from eggshell membrane (Aleutian Islands) and one from an egg (Fiji). Two of the Merlin (F. columbarius) samples were from feathers and one hobby (F. subbuteo) from museum skin (toe pad). Total RNA from the spleen of a freshly dead Eleonora’s falcon F. eleonorae was isolated following the procedure described by Chomczynski & Sacchi (1987). Tissue was homogenized into a solution containing 4 m guanidine-isothiocyanate, 25 mm sodium citrate, 0.5% sodium dodecyl sulphate and 100 mmβ-mercaptoethanol. After an organic extraction based on the addition of phenol and chloroform–isoamyl alcohol (24 : 1), the pellet was washed twice with 70% ethanol and resuspended in RNAase-free Milli-Q water.
MHC typing and sequence analyses
We amplified the third exon of a single MHC class I gene using the primers MHCI-int2F and MHCIEx4Rv (Alcaide et al., 2009) or MHCI-INT18 (5′-CAGGGGCTCACACAATACAG-3′) and MHC-ex395R (5′-GGCAGTACAAGGTCAGCGTCCC-3′). The second exon of a single MHC class II B gene was amplified according to Alcaide et al. (2007) using the primers Fal2FC and Fal2RC. Genomic fragments spanning exon 2 to exon 3 were amplified using the primers Fal2FC and RapEx3CR (Alcaide et al., 2007). PCR products were sequenced according to the Big Dye technology (Applied Biosystems) and resolved into an ABI3130xl (Applied Biosystems, Foster City, CA, USA), or universal tailed simultaneous bidirectional cycle sequencing (SBS, LI-COR, Inc., 1999; see Steffens et al., 1993; Oetting et al., 1995), using procedures similar to those reported in the study by Talbot et al. (2011) and resolved on a LI-COR 4200 or 4300 automated sequencer (LI-COR, Inc., 1999).
Resolving the gametic phase of the MHC class II locus was straightforward, as the majority of the falcons analysed were homozygous and alleles in heterozygotes differed in no more than two point mutations (see Table 1). However, the MHC class I locus was slightly more polymorphic in some species, such as the gyrfalcon and peregrine falcon. Here, heterozygotes were inferred from the sequence data but confirmed through single-strand conformational polymorphism (SSCP; Sunnucks et al., 2000) using automated procedures modified from Dahse et al. (1998). We employed the same forward and reverse universal tailed primers used in the sequencing reactions. PCR amplifications of the SSCP product were carried out in a final volume of 10 μL reaction mixture containing 2–100 ng genomic DNA, 0.2 mm dNTPs, 5 pmole unlabelled primer, 1.5 pmole IRD-labelled universal primer, 0.1 μg BSA, 1× PCR buffer (Perkin Elmer Cetus I) and 0.3 units Taq polymerase (Promega, Madison, Wisconsin, USA). PCRs were began at 94 °C for 2 min and continued with 40 cycles each of 94 °C for 30 s; 50 °C for 30 s; 72 °C for 60 s and concluded with a 30-min extension at 72 °C. PCR-amplified SSCP products were diluted approximately five-fold (2 μL of PCR product to 9 μL standard formamide-loading dye) and denatured for 4 min at 94 °C. The fluorescently labelled PCR products were electrophoresed on a 48-well 0.5× mutation detection enhancement (MDE) gel (Lonza) containing 0.5× MDE gel solution, 0.6× TBE, 0.005% to 10% APS and 0.0005% TEMED. Gel electrophoresis was carried out with 0.6× TBE at room temperature (22 °C; motor speed 1; power settings: voltage 2000 V, current 30 mA, power 6 W) for 12 h on a LI-COR 4200 automated sequencer. Two individuals homozygous for different MHC class I alleles, based on sequence data, were included in all SSCP gels to facilitate allele identification and augment quality control standards.
Table 1. Polymorphism statistics at MHC class I and MHC class II genes across different species of (a) kestrels and (b) falcons. See Fig. S1 for the phylogenetic relationships among species. Na, number of different alleles at a given locus (the number of different amino acid sequences is indicated in parentheses); k, average number of nucleotide differences between alleles; and N, number of individuals genotyped.
|Species||MHC class I (exon 3)|
|Na||K||N||Populations sampled||References||GenBank Acc. Nos.|
| Falco naumanni||> 80||9.15||> 80||Portugal, Spain, France, Italy, Greece, Israel||Alcaide et al., 2010; A. Rodríguez & M. Alcaide, unpublished data||JF831086-JF831120|
| Falco tinnunculus (continental)||23 (23)||10.99||25||Spain||Alcaide et al., 2010||EU120696- EU120722|
| Falco tinnunculus (insular)||6 (6)||8.45||25||Canary Islands||Alcaide et al., 2010||EU120696- EU120722|
| Falco punctatus||1 (1)||0||4||Mauritius Islands||This study||JN613279|
| Falco peregrinus||5 (2)||1.00||30||Fiji, Tasmania, Australia, Alaska, Greenland, Russia, Argentina, Chile, Falkland Islands, Spain, Northern Africa||This study||JN613264, JN613269-72|
| Falco eleonorae||3 (2)||4.66||32||Canary Islands, Greece||This study||JN613263, JN613265-66|
| Falco rusticolus||4 (4)||3.83||8||Alaska, Canada||This study||JN613273-76|
| Falco cherrug||2 (2)||2.00||3||United Arab Emirates||This study||JN613277|
| Falco fasciinucha||1 (1)||0||2||Zimbabwe||This study||JN613278|
| Falco subbuteo||2 (2)||2.00||1||Spain||This study||JN613267-68|
| Falco biarmicus||NA||NA||NA||NA||NA||NA|
| Falco columbarius||NA||NA||NA||NA||NA||NA|
| Falco concolor||NA||NA||NA||NA||NA||NA|
| Falco femoralis||NA||NA||NA||NA||NA||NA|
|MHC class II B (exon 2)|
| Falco naumanni||>100||22.68||>100||Spain, France, Italy, Greece, Israel, Kazakhstan||Alcaide et al., 2008, 2010||EF370839370864; EU10767-EU107746; HQ418344-HQ402921|
| Falco tinnunculus (continental)||41 (41)||24.31||25||Spain||Alcaide et al., 2010||EU118314-EU118359|
| Falco tinnunculus (insular)||10 (10)||25.78||25||Canary Islands||Alcaide et al., 2010||EU118314-EU118359|
| Falco punctatus||1 (1)||0||5||Mauritius Islands||This study|| |
| Falco peregrinus||3 (3)||1.50||63||Fiji, Tasmania, Australia, Alaska, Greenland, Russia, Canada, Northern Africa, South American migrants||Alcaide et al., 2007; this study||EF370947|
| || ||1||Spain||Alcaide et al., 2007; this study||EF370948|
| || ||22||Northern Africa, South American residents (Argentina, Chile, Falkland Islands)||This study||JN613255|
| Falco eleonorae||2 (1)||1.00||32||Canary Islands, Greece||This study||JN613254, JN613256|
| Falco rusticolus||1 (1)||0||12||Alaska, Canada||This study||JN613259|
| Falco cherrug||1 (1)||0||3||United Arab Emirates||This study||JN613262|
| Falco fasciinucha||1 (1)||0||2||Zimbabwe||This study||JN613261|
| Falco subbuteo||1 (1)||0||1||Spain||This study||JN613258|
| Falco biarmicus||2 (2)||2.00||1||Unknown||Alcaide et al., 2007||EF370949370950|
| Falco columbarius||1 (1)||0||3||Alaska||This study||JN613260|
| Falco concolor||1 (1)||0||1||Bahrain||This study||JN613257|
| Falco femoralis||1 (1)||0||1||Unknown||Alcaide et al., 2007||EF370988|
Unique alleles identified in the SSCP analysis were reamplified in an independent PCR and resequenced. For quality control purposes, we extracted, amplified and sequenced in duplicate 20% of the individuals typed using SSCP.
We verified that resulting SSCP allele configurations and quality control tests were consistent with sequence data; no inconsistencies between the SSCP and the sequence scores were identified, and there were no failures in quality control comparisons. Similar to the MHC class II locus, the high frequency of homozygous individuals for a particular MHC class I allele and the high sequence similarity between the alleles isolated from the same species (see Table 1), and verification using SSCP procedures, made cloning of individual alleles unnecessary.
The phylogenetic relationships among MHC sequences were visualized through neighbour-net networks built using SplitsTree 4.0 (Huson & Bryant, 2006) and based on Kimura 2-parameter distances (Kimura, 1980). Rates of positive diversifying selection at putative peptide-binding regions (PBR) and non-PBR codons were inferred by comparing nonsynonymous (dN) and synonymous (dS) substitution rates. Codons were classified as PBR and non-PBR in accordance with the predicted PBR of humans (see Bjorkman et al., 1987; Brown et al., 1993) and previous analyses of positive selection in birds, including kestrel MHC genes (see Alcaide et al., 2007, 2009; Balakrishnan et al., 2010). For the MHC class I, codons 4, 6–8, 22, 24, 37, 59–61, 64–65, 67 and 72 (exon 3) were labelled as PBR sites. For the MHC class II, codons 5, 7, 9, 13, 24, 26, 28, 32–35, 40, 43, 49, 52–54, 56, 57, 61, 63, 64, 66, 67, 70, 74, 77, 78, 81, 82, 84, 85 and 88 (exon 2) were labelled as PBR sites. Calculations were made in mega 5.0 (Tamura et al., 2011) using the modified distance-based Nei & Gojobori (1986) method with Jukes & Cantor (1969) correction and 10 000 bootstrap replicates. Two set of analyses were performed, one including all kestrel sequences and the other including all falcon sequences (see Fig. 1).
Figure 1. Phylogenetic network of MHC class I and MHC class II alleles from different species of falcons and kestrels. Rates of diversifying selection (± SD) at putative PBR and non-PBR codons are also indicated. Species codes are as follows: Fana (Falco naumanni, Lesser kestrel); Fati (Falco tinnunculus, Eurasian kestrel); Fapu (Falco punctatus, Mauritius kestrel); Fafe (Falco femoralis, Aplomado falcon); Facol (Falco columbarius, Merlin); Facon (Falco concolor, Sooty falcon); Fasu (Falco subbuteo, Eurasian hobby); Fape (Falco peregrinus, Peregrine falcon); Faru (Falco rusticolus, Gyrfalcon); Fabi (Falco biarmicus, Lanner falcon); Fach (Falco cherrug, Saker falcon); Fafa (Falco fasciinucha, Taita falcon); Fael (Falco eleonorae, Eleonora’s falcon). Numbers reflect allele codes (see also Table S1 for GenBank accession numbers).
Download figure to PowerPoint