ORIGINAL PAPER
Major histocompatibility complex variation and evolution
at a single, expressed DQA locus in two genera of elephants
Elizabeth A. Archie & Tammy Henry & Jesus E. Maldonado & Cynthia J. Moss &
Joyce H. Poole & Virginia R. Pearson & Suzan Murray & Susan C. Alberts &
Robert C. Fleischer
Received: 9 November 2009 /Accepted: 12 November 2009
# Springer-Verlag 2009
Abstract Genes of the vertebrate major histocompatibility
complex (MHC) are crucial to defense against infectious
disease, provide an important measure of functional genetic
diversity, and have been implicated in mate choice and kin
recognition. As a result, MHC loci have been characterized
for a number of vertebrate species, especially mammals;
however, elephants are a notable exception. Our study is the
first to characterize patterns of genetic diversity and natural
selection in the elephant MHC. We did so using DNA
sequences from a single, expressed DQA locus in ele-
phants. We characterized six alleles in 30 African elephants
(Loxodonta africana) and four alleles in three Asian
elephants (Elephas maximus). In addition, for two of the
African alleles and three of the Asian alleles, we characterized
complete coding sequences (exons 1?5) and nearly complete
non-coding sequences (introns 2?4) for the class II DQA
loci. Compared to DQA in other wild mammals, we
found moderate polymorphism and allelic diversity and
similar patterns of selection; patterns of non-synonymous
and synonymous substitutions were consistent with
balancing selection acting on the peptides involved in antigen
binding in the second exon. In addition, balancing selection
has led to strong trans-species allelism that has maintained
multiple allelic lineages across both genera of extant elephants
for at least 6 million years. We discuss our results in the
context of MHC diversity in other mammals and patterns of
evolution in elephants.
Keywords African elephant . Asian elephant . DQA .
Major histocompatibility complex . Coding sequence .
Molecular evolution
Introduction
Crucial to an animal?s ability to resist disease, genes of the
major histocompatibility complex (MHC) encode cell-
surface glycoproteins that bind to and present foreign
antigens for immune recognition (Klein 1986). MHC genes,
especially those in class II, are among the most diverse in
vertebrate genomes (Garrigan and Hedrick 2003; Gaudier et
al. 2000). Each MHC allele is thought to respond to a class
of potential antigens, and individuals and populations with
E. A. Archie : T. Henry : J. E. Maldonado : R. C. Fleischer
Center for Conservation and Evolutionary Genetics,
National Zoological Park & National Museum of Natural History,
Smithsonian Institution,
Washington, DC, USA
E. A. Archie (*)
Department of Biological Sciences, University of Notre Dame,
Notre Dame, IN 46556, USA
e-mail: earchie@nd.edu
T. Henry
Department of Environmental Science and Policy,
George Mason University,
Fairfax, VA, USA
C. J. Moss : J. H. Poole
Amboseli Trust for Elephants,
Nairobi, Kenya
V. R. Pearson
Philadelphia Zoo,
3400 Girard Avenue,
Philadelphia, PA, USA
S. Murray
Animal Health, National Zoological Park,
Washington, DC, USA
S. C. Alberts
Department of Biology, Duke University,
Durham, NC, USA
Immunogenetics
DOI 10.1007/s00251-009-0413-8
more diverse MHC loci are better able to cope with
multiple infections (Meyer-Lucht and Sommer 2005;
Paterson et al. 1998; Penn et al. 2002; Westerdahl et al.
2005). Genetic diversity at the MHC is likely to be
maintained by positive balancing selection (reviewed in
Garrigan and Hedrick 2003; Hughes 1999; Hughes and
Yeager 1998; Piertney and Olivier 2006). In support, allelic
lineages may be maintained over long evolutionary time
scales (i.e., trans-species allelism; Klein 1980; Takahata
and Nei 1990), and the codons responsible for antigen
binding?located in the second exon in the case of class II
loci?often have a higher rate of non-synonymous than
synonymous nucleotide substitutions (e.g., Hughes and Nei
1989). However, patterns of diversity differ somewhat
across exons in the same gene; for instance, the third exon
of class II loci encodes an extracellular domain close to the
transmembrane region, which may experience purifying
selection (Hughes and Nei 1989). Hence, comparing
different gene regions may inform us about evolutionary
relationships and different patterns of selection (Hughes
and Nei 1989).
Here, we characterize a DQA-like MHC locus in 30
African (Loxodonta africana) and three Asian elephants
(Elephas maximus). The MHC is completely uncharacter-
ized in elephants, yet these genes are important for
understanding elephant evolution and conservation for three
reasons. First, elephants are members of a distinct lineage
of mammals, the superorder Afrotheria, which diverged
from other placental mammals approximately 100 MYA
(Kriegs et al. 2006; Murphy et al. 2001; Scally et al. 2001;
Waddell and Shelley 2003). The MHC has yet to be
characterized in any Afrotherian mammal, and understand-
ing the evolutionary relationships between MHC loci in
multiple species of elephants and across mammals may
shed light on the evolution of mammalian MHC. Second,
evidence suggests that elephants might recognize and avoid
inbreeding with kin, including those that are apparently
socially unfamiliar to them (Archie et al. 2007). MHC
variation has been implicated in kin recognition in a few
vertebrate species (Brown and Eklund 1994; Manning et al.
1992; Rajakaruna et al. 2006; Zelano and Edwards 2002;
but see Cheetham et al. 2007), and characterizing MHC
diversity is a first step toward understanding whether
similar mechanisms operate in elephants. Third, the MHC
is a particularly informative measure of functional genetic
diversity; such genetic diversity is a conservation concern
for both wild and captive elephant populations (Armbruster
and Lande 1993; Caughley et al. 1990). Characterizing
MHC diversity is important for predicting how captive and
wild elephant populations will respond to disease threats. In
the last decade, herpes viruses killed at least 25 captive
elephants (Richman et al. 1999; Richman et al. 2000; Ryan
and Thompson 2001), and encephalomyocarditis, salmo-
nellosis, and anthrax have all threatened wild African
elephants (Grobler et al. 1995; Lindique and Turnbull
1994; Mbise et al. 1998). In addition, some diseases affect
Asian and African elephants differently, and these species-
level differences might be explained by differences in the
MHC.
Our objectives were to describe patterns of allelic
variation and confirm transcription in a DQA locus in
African and Asian elephants. We did this by characterizing
variation in a region spanning DQA?s second exon, second
intron, and third exon and confirming transcription by
amplifying this locus from RNA. We then used cDNA from
RNA transcripts to generate coding sequence and then
designed primers to obtain non-coding sequence from
genomic DNA. Finally, we tested for evidence of selection
by examining the ratios of synonymous and non-
synonymous changes and patterns of trans-species allelism.
We discuss our results in the context of DQA diversity in
other mammals and evolution in elephants.
Methods
Study subjects
We characterized DQA variation in 30 African elephants
and three Asian elephants. Of the 30 African elephants, 23
were wild elephants living in and around Amboseli
National Park, Kenya, and seven were captive elephants
living in North American zoos. The 23 wild African
elephants included seven parent?offspring pairs (mothers
or fathers) that were previously confirmed through micro-
satellite genotyping (Archie et al. 2006, 2008; Hollister-
Smith et al. 2007). The seven captive African elephants
included one individual who was born in captivity (the
offspring of two other captive elephants in our sample) and
housed at the Six Flags Wild Safari Park in Jackson, NJ.
The six other captive African elephants were born in the
wild but housed in captivity; one was born in Zimbabwe
and housed at the Philadelphia Zoo in Philadelphia, PA,
another was born in Uganda and housed at the Gladys
Porter Zoo in Brownsville, TX, and four were born in
Uganda and housed at the Great Adventure Safari Park in
Jackson, NJ. Of the three Asian elephants in our sample, all
were born in the wild?one each in Sri Lanka, India, and
Thailand?and all were housed in captivity at the
Smithsonian National Zoological Park.
Because it can be difficult to obtain fresh blood samples
for RNA isolation from elephants, the study subjects we
used to confirm transcription of the DQA locus differed
from those used for the genomic DNA analysis. Animals
used to confirm transcription were three captive African
elephants (two from the Baltimore Zoo and one from
Immunogenetics
Disney World?s Animal Park) and one of the three Asian
elephants from the Smithsonian National Zoo included in
the sample above. RNA isolated from that same Asian
elephant was used to amplify the full-length coding
sequence. Internal primers were then designed to generate
full sequences from genomic DNA samples from the other
two Asian elephants from the National Zoo and the African
Elephant from the Philadelphia zoo (see details below).
Genetic methods to characterize DQA variation across
individuals
In order to characterize allelic diversity across the 30
African elephants and three Asian elephants in our sample,
we amplified an 818 base pair region of DQA from
genomic DNA spanning the second exon, second intron,
and third exon. DNA extraction was carried out in a room
physically separate from the lab in which PCR amplifica-
tion or post-PCR analyses were conducted. We extracted
DNA from 35 samples (for two elephants we extracted
DNA from two samples); 29 of the samples (from 27
individuals, three Asian elephants and 24 African ele-
phants) were either whole blood, tissue collected from
biopsy darts, or buccal cells collected from cheek swabs.
The remaining six samples, all from African elephants,
were dung. DNA was extracted from blood, tissue, and
buccal samples using Qiagen?s DNeasy Tissue Kits
according to the manufacturer?s instructions (Qiagen,
Valencia, CA, USA). DNA was extracted from feces
according to the methods described in Archie et al.
(2003, 2006). Briefly, feces were collected within 10 min
of defecation, preserved in ethanol, and DNA was
extracted using a modified protocol (Archie et al. 2003)
for the QIAmp DNA Stool Kit (Qiagen).
Each DNA extract was amplified via PCR using the
primers MDQA1 and MDQA2 (Table 1; Slade et al. 1993;
Fig. 1). Each 10 ?l reaction contained 1 ?l of DNA extract,
0.25 ?l of each 10 ?M primer, 1 ?l of 2 mM dNTP mix
(Invitrogen, Carlsbad, CA, USA), 1 ?l of 100 mg ml?1
BSA, 1 ?l 10? PCR buffer without MgCl2, 1.0 ?l of
1.5 mM MgCl2, 0.15 ?l of AmpliTaq Gold DNA
polymerase (Applied Biosystems, Foster City, CA, USA),
and 4.35 ?l of water. A negative control (with water
replacing template DNA) was run with all PCR reactions.
Reactions were amplified in a MJ Research PTC-200
Thermal cycler (MJ Research, Waltham, MA, USA).
Amplification was preceded by a 10-min denaturation and
polymerase activation step at 95?C, followed by 40 cycles
of 1 min each at 56?C annealing, 72?C extension, and 95?C
denaturation. These cycles were followed by a 5-min
extension step at 72?C.
To assess the total number of alleles amplified by
MDQA1 and MDQA2, ligation and cloning were con-
ducted using the TOPO-TA Cloning? Kit for Sequencing
(Invitrogen). Positive PCR products were ligated into a
pCR?4-TOPO cloning vector and transformed into compe-
tent Mach1?-T1R Escherichia coli cells by standard
procedures. Colonies were cultured for 6 h in LB broth
and 1 ?l of each culture was used as template in a PCR
reaction using T3 and T7 universal primers. Each 20 ?l
reaction contained 1 ?l of cultured E. coli, 0.5 ?l of each
10 ?M primer, 2 ?l of 2 mM dNTP mix (Invitrogen), 2 ?l
10? PCR buffer without MgCl2, 1.6 ?l of 1.5 mM MgCl2,
0.2 ?l of AmpliTaq Gold DNA polymerase (Applied
Biosystems), and 12.2 ?l of water. Amplification was
preceded by 10 min at 95?C, followed by 35 cycles of
1 min each at 56?C annealing, 72?C extension, and 95?C
denaturation. These cycles were followed by 5 min at 72?C.
Positive PCR reactions of the expected length (~800 base
pairs) were sequenced in both the 3? and 5? directions using
an ABI PRISM? 3100 DNA Analyzer using Dye Termi-
nator Cycle Sequencing (Applied Biosystems).
To control for the mutation errors inherent in sequencing
cloned products, the risk of heteroduplex formation in PCR
(Borriello and Krauter 1990; L?Abbe et al. 1992), and the
expected high heterozygosity at MHC loci, we cloned, on
average, 2.0 independently generated PCR products (range=
1?3 PCR products) and sequenced 28.8 clones (range=16?56
clones) per individual. As reported by previous studies that
used the primers DQA1 and DQA2, our cloned products each
contained one of two very divergent sets of sequences that
could not be aligned with each other (Decker et al. 2002;
Primer name Primer sequence (5? to 3?) Ta (?C) Source
E1F CACAACCCTGGACAGCAAC 60?58a This study
E1F77 TGAGCAACTGTGGAGGTGAA 60?58a This study
E1R AGCACAGCTATGTTCCTCAGTC 60?58a This study
MDQA1 CCGGATCCCAGTACACCCATGAATTTGATGG 56 Slade et al. 1993
MDQA2 CCGGATCCCCAGTGCTCCACCTTGCAGTC 56 Slade et al. 1993
E45F ACTTCCTCCCCAAGGATGAT 60?58a This study
E5R TGGGAAATTTATTGCTTCCA 60?58a This study
Table 1 Primer sets, sequences,
and annealing temperatures used
in this study
a 60?58 indicates a touchdown
protocol starting at 60?C and
ending at 58?C (see text for
details).
Immunogenetics
Lehman et al. 2004). Comparison with sequences on NCBI
GenBank revealed that the first set of sequences (519 clones,
55%) aligned closely to MHC sequences from the class
II locus DOA, rather than DQA. The DOA locus is not
directly involved in antigen binding (Decker et al. 2002;
Naruse et al. 1999), and these sequences will be the focus
of a separate analysis not discussed here. The second set
of sequences (429 clones, 45%) aligned closely with MHC
sequences from the class II locus DQA, and these
sequences are the focus of our analysis in this paper. On
average, we sequenced 12.6 DQA-like clones (range=
2?42 clones) per individual. A given DQA sequence was
defined as an allele when copies with identical nucleotide
sequences occurred in at least three clones (Table 2). In
addition, all alleles occurred either in multiple individuals
or in independently amplified PCR reactions from the
same individual (Table 2).
Genetic methods to confirm transcription
In order to confirm that the DQA locus we amplified was
transcribed, we used the same primers as above (MDQA1
and MDQA2; Table 1) to amplify DQA from cDNA
prepared from purified RNA. Specifically, blood from three
African elephants and one Asian elephant was collected
into EDTA vacutainer tubes. For the African elephants,
2 ml of blood in EDTA was added to 5 ml of RNALater
(Qiagen) and stored at ?80?C. The same procedure was
followed for the Asian elephant, but the blood/EDTA
mixture was preserved in 5 ml of Trizol (Invitrogen)
instead of RNALater. For the African elephant samples,
800 ?l of each sample was centrifuged, and 1 ml of Trizol
with ?-mercaptoethanol was added to the pellet. For the
Asian elephant sample, 800 ?l of the sample was added to
400 ?l of Trizol with ?-mercaptoethanol. RNA was
extracted from all four samples using a standard phenol?
chloroform separation followed by isopropanol precipita-
tion and ethanol washes. The RNA pellet was resuspended
in RNase-free water and stored at ?20?C. To confirm
transcription, the RNA extract was reverse transcribed into
cDNA using an iScript cDNA Synthesis Kit (Bio-Rad,
Hercules, CA, USA), and the primers MDQA1 and
MDQA2 were used to amplify the second and third exons
using the same amplification conditions as previously
described. PCR products were run on an agarose gel and
the presence of two bands constituted evidence for
transcription: one 818-bp band, amplified from genomic
DNA, and one 460-bp band, amplified from cDNA.
Genetic methods to generate full-length coding sequences
Using the RNA extracted from the Asian elephant sample
above, we amplified the full-length DQA coding sequence
using Invitrogen?s GeneRacer Kit (Invitrogen). The resulting
products were cloned using the TOPO-TA Cloning? Kit for
Sequencing (Invitrogen) according to the methods described
above and sequenced using the primers provided with the
GeneRacer kit and MDQA1 and MDQA2.
From the resulting full-length coding sequence, we used
Primer3 software (Rozen and Skaletsky 2000) to design
Fig. 1 Diagram of the MHC DQA locus characterized in this study.
Black boxes indicate exons, lines indicate introns, and arrows indicate
the relative locations of primers used in this study. From genomic
DNA, primer pair E1F and E1R amplifies an approximately 2,400-bp
product; primers E1F77 and E1R amplify an approximately 2,300-bp
product (with cleaner sequences than E1F and E1R); primers MDQA1
and MDQA2 amplify an 818-bp product; primers E45F and E5R
amplify a 1,050-bp product
Table 2 MHC DQA alleles from two species of elephants
Allele No. of elephants
with allele (no. of
independent PCR
reactions, no. of
clones)
Frequencya Genbank
accession
no.
Loxodonta africana
LoafDQA*01 20 (20, 235) 0.611 GU369694
LoafDQA*02 4 (4, 4) 0.056 GU369695
LoafDQA*03 1 (2, 5) 0.028 GU369696
LoafDQA*04 10 (14, 25) 0.167 GU369697
LoafDQA*05 2 (4, 13) 0.028 GU369698
LoafDQA*06 7 (11, 51) 0.111 GU369699
Elephas maximus
ElmaDQA*01 1 (2, 23) ? GU369700
ElmaDQA*02 1 (2, 4) ? GU369701
ElmaDQA*03 1 (3, 35) ? GU369702
ElmaDQA*04 1 (4, 5) ? GU369703
a For African elephants, allele frequencies are calculated from a subset
of 18 individuals that were not genotyped from feces and for which
we sequenced ?6 clones. We do not present allele frequencies for
Asian elephants because of the small sample size (N=3 individuals)
Immunogenetics
new primers to amplify complete coding and nearly
complete non-coding sequences from genomic DNA
(Table 1 and Fig. 1). We then used these primers to
characterize exons 1?5 and introns 2?4 for five alleles from
four animals from our main sample: the African elephant at
the Philadelphia Zoo and the three Asian elephants housed
at the Smithsonian National Zoo. Reaction conditions were
identical for all primer sets; each 20 ?l reaction contained
2 ?l of template DNA, 0.5 ?l of each 10 ?M primer, 2 ?l of
2 mM dNTP mix (Invitrogen), 2 ?l 10? PCR buffer without
MgCl2, 2.0 ?l of 1.5 mM MgCl2, 2 ?l of 100 mg ml?1
BSA, 0.3 ?l of AmpliTaq Gold DNA polymerase (Applied
Biosystems), and 8.7 ?l of water. Amplification was
preceded by 10 min at 95?C, followed by five cycles of
30 s at 95?C denaturation, 30 s at 60?C annealing, and
1 min at 72?C extension, followed by 30 cycles of 30 s
at 95?C denaturation, 30 s at 58?C annealing, and 1 min at
72?C extension. These cycles were followed by 10 min at
72?C. The resulting PCR products were sequenced directly
in both the 3? and 5? directions using an ABI PRISM? 3100
DNA Analyzer using Dye Terminator Cycle Sequencing
(Applied Biosystems). DNA sequences were inspected and
aligned using Sequencher software version 4.1 (Gene
Codes Corporation). Polymorphism was identified by
double peaks, and haplotypes were assigned using PHASE
2.1.1 software (Harrigan and Mazza 2008).
Data analyses
In order to characterize patterns of allelic diversity, we
calculated nucleotide diversity (pi) across all exons and
introns as the average proportion of nucleotide differences
between sequences using MEGA software version 2.1
(Kumar et al. 2001). For African elephants, allele frequen-
cies and tests of Hardy?Weinberg equilibrium were carried
out in Genepop (version 4.0; these analyses were not
performed on Asian elephant genotypes because of small
sample size). These analyses were performed on a subset of
African elephant genotypes that met two conditions: their
genotypes were all derived from tissue samples (as opposed
to DNA extracted from feces in which dropout of alleles is
more likely), and they were all from individuals for which
we sequenced more than six clones, which is the minimum
number of clones needed to detect two alleles at a locus
with 95% certainty.
To investigate patterns of selection, we calculated the
rates of non-synonymous (dN) and synonymous (dS)
substitutions using the Nei and Gojobori (1986) method,
with Jukes?Cantor correction using MEGA software
version 2.1 (Kumar et al. 2001). We did this for all exons
and for second exon codons in the putative antigen-binding
region (ABR; defined according to Reche and Reinherz
2003). In addition, we used MEGA to perform Z tests of
positive, neutral and purifying selection on the entire
second exon, ABR codons, and the third exon, which
encodes an extracellular domain close to the transmem-
brane region.
In order to investigate evolutionary relationships between
DQA in elephants (Afrotheria) and other mammals
(Boreoeutheria), we constructed phylogenetic trees using
maximum parsimony (MP), maximum likelihood (ML),
and neighbor joining (NJ) methods from PAUP v. 4.0b
(Swofford 2002). For all analyses, we used a concatenated
data set of the second and third exons. These exons were
chosen because we were able to find representative
sequences for the widest range of mammal species.
Specifically, we included: (a) the African and Asian
elephant DQA sequences from our study, (b) a represen-
tative selection of mammalian DQA sequences from
GenBank (species and accession numbers=Mirounga
leonine, MLUO3583; Zalophus californianus, AF502564;
Alopex lagopus, Z26591; Canis lupus, NM_001011726;
Canis familiaris, AJ311099; Bos taurus, D50454; Ovis
avies, EE803522; Sus scrofa, AY285934; Equus caballus,
L33909; Mus musculus, BC019721; Aotus nancymaae,
AF201296; Macacca mulatta, EF362438; Pan troglo-
dytes, AY663401; Homo sapiens, AY375903 and
AK130811), and (c) the marsupial mammal, Monodelphis
domesticus (XM_001376727), as an outgroup. The best
model of evolution was chosen using the analyses
generated with ModelTest version 3.7 (Posada and
Crandall 1998). According to the results of ModelTest,
ML trees were generated using the K80+G, Kimura two-
parameter model with rate variation among sites. The ML
analysis was conducted using a heuristic search under
likelihood criteria, obtained via random stepwise addition
with ten replicates starting from random trees, including
tree bisection reconnection with multiple tree swapping.
Nodal support was assessed with bootstrap resampling by
using GARLI to create 1,000 ML replicates under the
same search conditions as above, after which PAUP was
used to calculate majority rule consensus.
Results
MHC DQA alleles in Asian and African elephants
From 429 cloned sequences of the second exon, second
intron, and third exon of DQA (MDQA1 and MDQA2
amplicons), we identified ten unique alleles: six alleles in
30 African elephants and four alleles in three Asian elephants
(Table 2). We chose five of these alleles, two derived from
African elephants (LoafDQA*01, LoafDQA*02) and three
from Asian elephants (ElmaDQA*01, ElmaDQA*02, and
ElmaDQA*03) to characterize complete coding and nearly
Immunogenetics
complete non-coding sequences. Intron one sequences are
not included here because intron one contained two
microsatellites and several poly-T and poly-A regions,
making it difficult to reliably sequence and align across
individuals.
None of the nucleotide (Fig. 2) or amino acid sequences
(Fig. 3) for any of the ten alleles have been previously
described. Evidence supports the hypothesis that this locus
is transcribed and expressed; cDNA amplification with
MDQA1 and MDQA2 amplified the expected 460-bp
fragment, none of the alleles contained stop codons in their
expected coding regions, and the observed patterns of
selection were inconsistent with neutral evolution (see
below). These alleles are likely part of a single locus as
no more than two alleles ever occurred in a single
individual, and we observed Mendelian inheritance in six
pairs of known parents and offspring whose genotypes were
derived from tissue or blood samples (parentage confirmed
by microsatellite genotyping; Archie et al. 2006, 2008;
Hollister-Smith et al. 2007). We did not observe Mendelian
inheritance in three additional pairs of parents and offspring
that were genotyped from feces. This lack of Mendelian
inheritance was probably due to allelic dropout, as all
genotypes derived from feces were homozygous. Such
allelic dropout is not surprising given that that DNA
derived from fecal samples is often degraded into short
fragments and even the short allele we amplified was still
relatively long (818 base pairs). Primer sets that amplify
smaller overlapping pieces may reduce the rate of allelic
dropout in genotypes from fecal-derived DNA samples.
Allelic variation within species
African elephants In the six African elephant alleles, there
were 70 variable nucleotide sites across all 1,893 coding
and non-coding nucleotides (nucleotide diversity, pi=0.027;
Fig. 2). Across the entire coding sequence, 21 out of 255
amino acid residues were variable (8.24% variable amino
acids; Fig. 3). Allelic diversity was highest in the second
exon, with the highest nucleotide diversity of any region
(pi=0.039), and each second exon allele had a unique amino
acid sequence, with 15.66% (13 of 83) variable amino acid
residues. By comparison, nucleotide diversity was lower in all
other exons and introns (exon 1 pi=0; intron 2 pi=0.026; exon
3 pi=0.019; intron 3 pi=0.008; exon 4 pi=0.029; intron 4 pi=0;
exon 5 pi=0.015). In addition, amino acid variability was
lower in all other exons; no amino acids varied in the first
exon, in the third exon 5.25% (5 of 95) of amino acids varied,
while 5.88% (3 of 51) of amino acids varied in the fourth exon
(the fifth exon is non-coding).
Asian elephants In the four alleles found in Asian ele-
phants, there were 81 variable nucleotides across all 1,893
coding and non-coding bases (pi=0.026; Fig. 2). Across the
entire coding sequence, there were 18 variable amino acid
residues in 255 amino acids (7.06% variable amino acids;
Fig. 3). Allelic diversity was highest in the second exon;
nucleotide diversity (pi) was 0.039, and each second exon
allele had a unique amino acid sequence, with 13.25% (11
of 83) of amino acid residues varying across the exon. By
comparison, amino acid variability was lower in all other
exons; one amino acid varied in the first exon (3.70%; one
of 27), in the third exon 4.21% (four of 95) of amino acids
varied, while 5.88% (three of 51) of amino acids varied in
the fourth exon. Compared to the second exon, nucleotide
diversity was also lower in all other exons and introns,
except exon 4 (exon 1 pi=0.008; intron 2 pi=0.025; exon 3
pi=0.017; intron 3 pi=0.011; exon 4 pi=0.040; intron 4 pi=
0.031; exon 5 pi=0.023).
Allele frequencies and heterozygosity in African elephants
We calculated DQA allele frequencies from the 18 African
elephants (15 wild and three captive) that were genotyped
from six or more clones using only tissue-derived (no fecal-
derived) DNA (Table 2). Among these individuals, mean
heterozygosity was 0.611, and we observed one very
common allele, LoafDQA1*01, which occurred in 16 of
18 samples (frequency of 0.611). Two other alleles, Loaf-
DQA1*04 and LoafDQA1*06, were also relatively common
(LoafDQA1*04 frequency=0.167 and LoafDQA1*06
frequency=0.111), while LoafDQA1*02, LoafDQA1*03,
and LoafDQA1*05 were more rare (LoafDQA1*02
frequency=0.056 and LoafDQA1*05 frequency=0.028).
It is fairly unusual to observe a single common allele at
an MHC locus, and LoafDQA1*01 might be common in
multiple populations of African elephants; out of all 30
African elephants we genotyped, LoafDQA1*01 occurred
in 20 individuals, including three of seven zoo elephants
(one from Zimbabwe and two from Uganda), and 17 of 23
individuals from Amboseli National Park, Kenya. The high
frequency of LoafDQA1*01 in Amboseli was not due to
the fact that our sample contained close kin (parent?
offspring pairs). Rather, with parent?offspring pairs re-
moved, the frequency of LoafDQA1*01 in Amboseli was
Fig. 2 Alignment of MHC DQA sequences in Asian and African
elephants relative to the most common allele in African elephants,
LoafDQA1*01. Identities are shown by dashes and deletions are shown
by back slashes. Alleles LoafDQA1*01, LoafDQA*02, ElmaDQA*01,
ElmaDQA*02, and ElmaDQA*03 depict complete coding and non-
coding sequences, except intron 1. Alleles LoafDQA1*03, Loaf-
DQA*04, LoafDQA1*05, LoafDQA*06, and ElmaDQA*04 were
amplified using the primers MDQA1 and MDQA2, which span the
second exon, second intron, and third exon
?
Immunogenetics
Immunogenetics
Fig. 2 (continued)
Immunogenetics
Fig. 2 (continued)
Immunogenetics
Fig. 2 (continued)
Immunogenetics
still 0.58. The high frequency of LoafDQA1*01 in
Amboseli occurred in the context of Hardy?Weinberg
equilibrium; heterozygosity in these 15 individuals was
0.6, which was not significantly different from expected
heterozygosity (expected heterozygosity=0.58; exact test,
P=0.71).
dN/dS supports balancing selection in the ABR of exon 2
and purifying selection in exon 3
The relative proportions of non-synonymous (dN) and
synonymous (dS) substitutions reveal historical patterns of
selection; a larger number of non-synonymous than
synonymous changes constitutes evidence for balancing
selection, while dN/dS ratios less than one are evidence for
purifying selection (Hughes and Nei 1989). Consistent with
the hypothesis that balancing selection operates on the ABR
of the second exon, the dN/dS ratio in the putative ABR was
3.18 across all ten alleles (Table 3). Z tests of selection (Nei
and Kumar 2000) indicated that this ratio was significantly
different from neutrality (Z=2.17, P=0.032) and supported
balancing selection (positive selection; Z=2.00, P=0.024).
In contrast to the ABR, the third exon is expected to
be under purifying selection (Hughes and Yeager 1998).
In support of this prediction, the rate of synonymous
substitutions was similar across both the second and third
exon (African elephants, second exon dS=0.034, third
exon dS=0.042; Asian elephants, second exon dS=0.034,
third exon dS=0.037; Table 3), but the rate of non-
synonymous substitutions was much lower in the third
exon as compared to the second exon (African elephants,
second exon dN=0.042, third exon dN=0.013; Asian
elephants, second exon dN=0.039, third exon dN=0.012;
Table 3). As a result, the dN/dS ratio is markedly lower in
the third exon (dN/dS=0.32). Z tests allowed us to exclude
the possibility that the third exon experiences balancing
selection (Z=?1.44, P=1.00); however, a test of purifying
selection was not significant (Z=1.45, P=0.075), and Z
Fig. 2 (continued)
Immunogenetics
tests indicated that the dN/dS ratio does not differ from
neutrality (Z=?1.55, P=0.12).
Phylogenetic patterns: trans-generic allelism supports
balancing selection
An additional prediction of the hypothesis that balancing
selection acts on this DQA locus is that allelic lineages
should be maintained across species over long evolutionary
time frames (Klein 1980; Takahata and Nei 1990).
Topologies estimated from MP, ML, and NJ analyses of
concatenated second and third exon data set were highly
concordant and we found that ML produced the most
representative relationships (Fig. 4). The phylogenetic
analysis resolves the basal branching between the Afro-
therian (e.g., elephants, aardvarks) and Boreoeutherian
mammals (e.g. primates, carnivores, rodents, ruminants),
and indicates that elephants do not share second exon
alleles with any major groups of Boreoeutherian mammals.
However, within the two extant species of elephants, we
found strong evidence of trans-generic allelism (Fig. 4). In
particular, the entire second exon, second intron, and third
exon sequence of each Asian allele was closely matched to
an allele found in African elephants. The nucleotide
sequences at the second exon, second intron, and third
exon for the African allele, LoafDQA1*03, and the Asian
allele, ElmaDQA1*03, have identical nucleotide sequences
across the second exon, second intron, with only a single
Fig. 3 Alignment of MHC DQA predicted amino acid sequences
from Asian and African elephants and five other Boreoeutherian
mammals. Amino acid positions included in the peptide binding
region are shaded gray and were defined according to Reche and
Reinherz (2003). Genbank accession numbers are in parentheses. Bars
indicate divisions between exons; identities are shown by dots
Immunogenetics
mutation leading to a single amino acid change in the third
exon (Figs. 2 and 3). ElmaDQA1*02 and LoafDQA1*02
had identical nucleotide sequences across all exons, except
for two amino acid changes in non-ABR regions of the
second exon (Fig. 3). ElmaDQA1*01 and LoafDQA1*01
had identical amino acid sequences across all exons, except
for two amino acid changes in the third exon (Fig. 3).
ElmaDQA*04 and LoafDQA*04 had identical amino acid
sequences except for a single amino acid change in the
second exon (Fig. 3).
Discussion
Our characterization of elephant MHC loci represents the
first description of an MHC locus in the superorder of
Afrotherian mammals. Beyond the phylogenetic value of
the analysis, it represents a first step towards characterizing
genetic diversity at a set of loci that may be involved in the
kin discrimination that we have previously documented in
this species (Archie et al. 2007). In addition, MHC loci
might provide an important measure of functional genetic
Table 3 Estimates of dN and dS (?SE) in all DQA exons including the hypothesized antigen-binding region (ABR) of the second exon
Region No. of codons Loxodonta africana Elephas maximus Species combined
dN dS dN/dS dN dS dN/dS dN/dS
Exon 1 27 0?0 0?0 0 0.017?0.016 0?0 NA NA
Exon 2 83 0.042?0.015 0.034?0.018 1.24 0.043?0.016 0.034?0.021 1.26 1.15
Non-ABR 64 0.018?0.008 0.038?0.022 0.47 0.021?0.010 0.038?0.028 0.55 0.49
ABR 19 0.116?0.040 0.039?0.039 2.97 0.109?0.059 0.024?0.025 4.54 3.18
Exon 3 89 0.013?0.006 0.042?0.016 0.31 0.012?0.006 0.037?0.019 0.32 0.32
Exon 4 52 0.025?0.014 0.043?0.029 0.58 0.028?0.012 0.078?0.031 0.036 0.38
Fig. 4 Maximum likelihood
phylogram depicting evolution-
ary relationships among MHC
DQA haplotypes across Bor-
eoeutherian mammals and ele-
phants (Afrotheria) for
concatenated exons 2 and 3.
Numbers above internodes are
bootstrap support values from
ML analyses. See text for tree
construction methods and se-
quence accession numbers
Immunogenetics
diversity in threatened elephant populations, especially with
regard to response to infectious disease. The locus we
characterized is most closely related to MHC DQA1 in
other mammals.
Comparative genetic diversity
Patterns of MHC diversity across different species can
reveal how differences in demography, life history, or
disease threats may shape MHC evolution. Elephants have
several life history traits that differ from those of typical
mammals (e.g., extremely large body size and long life
span), and these traits, as well as the complex sociality that
characterizes elephants and a number of other mammal
species, might influence their exposure to disease. For
instance, because elephants are unusually long-lived, we
might expect them to encounter relatively more infectious
agents over the course of their lives and consequently
require relatively high MHC diversity. However, the
patterns of genetic diversity at the DQA locus in our study
were?for the most part?qualitatively similar to DQA
diversity in other wild mammals. For instance, compared to
elephants, more alleles per study subject were observed in
wild baboons and Weddell seals (Alberts 1999; Lehman et
al. 2004), while fewer alleles were found in Ross seals,
leopard seals, elephant seals, marmosets, and tuco tucos
(Antunes et al. 1998; Cutrera and Lacey 2006; Lehman et
al. 2004; Weber et al. 2004).
However, the African DQA locus in our study differed
from other studies of wild mammals in that one allele was
especially common. This allele, LoafDQA*01, occurred in
over half of our samples and may be common in multiple
populations. Such high frequencies of a single DQA
allele have previously been reported only in species with
low polymorphism; for instance, only species with two
or fewer alleles have produced frequencies for a single
DQA allele that are greater than 0.5 (Antunes et al.
1998; Lehman et al. 2004; Weber et al. 2004). One
speculation is that LoafDQA*01 is common because it
confers a selective advantage to individuals in resisting a
disease that was or is highly prevalent. However, a larger
sample size of individuals and populations is necessary to
confirm that this allele is really as common as it appears in
our sample. In addition, we might expect to observe
evidence of a selective sweep in the region of the genome
surrounding this allele.
Evidence for balancing and purifying selection
Genetic diversity at MHC loci is likely to be maintained by
balancing selection (reviewed in Garrigan and Hedrick
2003; Hughes 1999; Hughes and Yeager 1998; Piertney and
Olivier 2006). Such selection leaves a characteristic
mutational signature in coding regions; in particular, codons
that have experienced balancing selection should have more
non-synonymous changes than synonymous changes, while
codons that experienced purifying selection will have very
few non-synonymous changes relative to the number of
synonymous changes. We found evidence that is consistent
with both balancing and purifying selection in elephants;
the dN/dS ratios in the ABR of the second exon were
significantly greater than one, indicating that positive
evolution has acted on the ABR. In contrast, the third
exon, which is not involved in antigen recognition, appears
to have experienced purifying selection, as we found two or
three times as many synonymous changes than non-
synonymous changes in both African and Asian elephants.
However, for the third exon, we were not able to reject the
null hypothesis of neutral evolution. This might be because
of our small sample size of alleles (and therefore high
variance in our estimates of dN and dS); further study is
needed to conclusively distinguish between the forces of
selection and neutrality at these loci.
Balancing selection can also act to maintain lineages of
alleles across species over long evolutionary time scales.
While we did not find any evidence that Asian or African
elephants share alleles with the analyzed Boreoeutherian
mammals, we did find strong evidence of trans-species
allelism acting across the two extant elephant genera,
Loxodonta and Elephas. This pattern was clearly evident
in the second exon, where the amino acid sequence of all
four Asian elephant alleles matched very closely the amino
acid sequences of four African elephant alleles. The trans-
species allelism we observed in the second exon extended
to the intron and third exon of the same pairs of allele, but
was less extreme. These trans-species allelic lineages may
have been maintained over relatively long evolutionary
time frames, as Asian and African elephants are thought to
have diverged around 6 million years ago (Krause et al.
2006). However, this phenomenon has been observed over
even longer evolutionary time frames. For instance, in
primates, trans-species allelic lineages have been conserved
for over 30 million years (Geluk et al. 1993). As more
information on the MHC of other mammal species becomes
available, it will be interesting to know if trans-species
allelism operates across Afrotheria.
Conclusions and implications
Our results provide the first view of sequence structure and
diversity of MHC in elephants and Afrotheria. They
provide full transcript sequences that should prove useful
for further development of markers to assess genetic
diversity at this locus, and indicate that diversity at the
DQA locus in elephants is similar to that seen in other
mammals. This diversity is likely to be a result of the
Immunogenetics
historical action of balancing selection; in particular, the
proportions of synonymous to non-synonymous changes
and the strong evidence for trans-species allelism suggests
that balancing selection has acted to maintain multiple,
diverse lineages of DQA alleles in Asian and African
elephants over at least 6 million years. However, our study
did not reveal the underlying mechanisms driving balancing
selection at this locus. A more thorough characterization of
diversity at this locus, and especially other loci, are needed
to fully understand the evolutionary ecology of the MHC in
elephants.
Acknowledgments We thank the Office of the President of Kenya
for permission to work in Amboseli National Park under permit
number MOES&T 13/001/30C 72/7. We thank the Kenya Wildlife
Service for local sponsorship. We thank the Amboseli Elephant
Research project for invaluable scientific and logistical support,
particularly the team of N. Njiraini, K. Sayialel, and S. Sayialel who
contributed greatly to the collection of genetic samples. We thank the
Smithsonian National Zoo, the Philadelphia Zoo in Philadelphia, PA,
the Gladys Porter Zoo in Brownsville, TX, and the Six Flags Wild
Safari Park in Jackson, NJ for their support and cooperation in sample
collection. This research was supported by the Smithsonian Institution
Abbott Endowment Fund, the National Zoo?s Institution Center for
Conservation and Evolutionary Genetics, the Friends of the National
Zoo, the National Science Foundation (IBN0091612 to SCA), the
Amboseli Trust for Elephants, the Amboseli Elephant Research
Project, and Duke University.
References
Alberts SC (1999) Thirteen Mhc-DQA1 alleles from two populations
of baboons. Immunogenetics 49:825?827
Antunes SG, de Groot NG, Brok H, Doxiadis G, Menezes AAL,
Otting N, Bontrop RE (1998) The common marmoset: a new
world primate species with limited MHC class II variability. Proc
Natl Acad Sci U S A 95:11745?11750
Archie EA, Moss CJ, Alberts SC (2003) Characterization of
tetranucleotide microsatellite loci in the African savannah
elephant (Loxodonta africana africana). Mol Ecol Notes 3:244?
246
Archie EA, Moss CJ, Alberts SC (2006) The ties that bind: genetic
relatedness predicts the fission and fusion of groups in wild
African elephants (Loxodonta africana). Proc Roy Soc London
273:513?522
Archie EA, Hollister-Smith JA, Poole JH, Lee PC, Moss CJ, Maldonado
JE, Fleischer RC, Alberts SC (2007) Behavioral inbreeding
avoidance in wild African elephants. Mol Ecol 16:4138?4148
Archie EA, Maldonado JE, Hollister-Smith JA, Poole JH, Moss CJ,
Fleischer RC, Alberts SC (2008) Fine-scale population genetic
structure in a fission?fusion society. Mol Ecol 17:2666?2679
Armbruster P, Lande R (1993) A population viability analysis for
African elephant (Loxodonta africana): how big should reserves
be? Conserv Biol 7:602?610
Borriello F, Krauter KS (1990) Reactive site polymorphism in the
murine protease inhibitor gene family is delineated using a
modification of the PCR reaction. Nucleic Acids Res 18:5481?
5487
Brown JL, Eklund A (1994) Kin recognition and the major
histocompatibility complex: an integrative review. Am Nat
143:435?461
Caughley G, Dublin HT, Parker ISC (1990) Projected decline of the
African elephant. Biol Conserv 54:157?164
Cheetham SA, Thom MD, Jury F, Ollier WER, Beynon RJ, Hurst JL
(2007) The genetic basis of individual-recognition signals in the
mouse. Curr Biol 17:1771?1777
Cutrera AP, Lacey EA (2006) Major histocompatibility complex
variation in talas tuco-tucos: the influence of demography on
selection. J Mammal 87:706?716
Decker DJ, Stewart BS, Lehman N (2002) Major histocompatibility
complex class II DOA sequences from three Antarctic seal
species verify stabilizing selection on the DO locus. Tissue
Antigens 60:534?538
Garrigan D, Hedrick PW (2003) Perspective: detecting adaptive
molecular polymorphism: lessons from the MHC. Evolution
57:1707?1722
Gaudier S, Dawkins RL, Habara K, Kulski JK, Gojobori T (2000)
SNP profile within the human major histocompatibility complex
reveals and extreme and interrupted level of nucleotide diversity.
Genome Res 10:1579?1586
Geluk A, Elferink DG, Slierendregt BL, Vanmeijgaarden KE, de Vries
RRP, Ottenhoff THM, Bontrop RE (1993) Evolutionary conser-
vation of major histocompatibility complex-DR/peptide/T cell
interactions in primates. J Exp Med 177:979?987
Grobler DG, Raath JP, Braak LE, Keet DF, Gerdes GH, Barnard BJ,
Kriek NP, Jardine J, Swanepoel R (1995) An outbreak of
encephalomyocarditis-virus infection in free-ranging African
elephants in the Kruger National Park. Onderstepoort J Vet Res
62:97?108
Harrigan RJ, Mazza ME (2008) Computation vs. cloning: evaluation
of two methods for haplotype determination. Mol Ecol Resour
8:1239?1248
Hollister-Smith JA, Poole JH, Archie EA, Vance EA, Georgiadis NJ,
Moss CJ, Alberts SC (2007) Age, musth and paternity in wild
male African elephants, Loxodonta africana. Anim Behav
74:287?296
Hughes AL (1999) Adaptive evolution of genes and genomes. Oxford
University Press, New York
Hughes AL, Nei M (1989) Nucleotide substitution at major
histocompatibility complex class II loci evidence for overdominant
selection. Proc Natl Acad Sci 86:958?962
Hughes AL, Yeager M (1998) Natural selection at major histocom-
patibility complex loci of vertebrates. Annu Rev Genet 32:415?
435
Klein J (1980) Generation of diversity at MHC loci: implications for
T-cell receptor repertoires. In: Fougereau M, Dausset J (eds)
Immunology 80. Academic, London, pp 239?235
Klein J (1986) Natural history of the major histocompatibility
complex, 1st edn. Wiley, New York
Krause J, Dear PH, Pollack JL, Slatkin M, Spriggs H, Barnes I, Lister
AM, Ebersberger I, Paabo S, Hofreiter M (2006) Multiplex
amplification of the mammoth mitochondrial genome and the
evolution of elephantidae. Nature 439:724?727
Kriegs JO, Churakov G, Kiefmann M, Jordan U, Brosius J, Schmitz J
(2006) Retroposed elements as archives for the evolutionary
history of placental mammals. PLoS Biol 4(4):e91
Kumar S, Tamura K, Jakobsen IB, Nei M (2001) MEGA2: molecular
evolutionary genetics analysis software. Bioinformatics 17:1244?
1245
L?Abbe D, Belmaaza A, Decary F, Chartrand P (1992) Elimination
of heteroduplex artifacts when sequencing HLA genes ampli-
fied by polymerase chain reaction. Immunogenetics 35:395?
397
Lehman N, Decker DJ, Stewart BS (2004) Divergent patterns
of variation in major histocompatibility complex class II
alleles among Antarctic phocid pinnipeds. J Mammal
85:1215?1224
Immunogenetics
Lindique PM, Turnbull PCB (1994) Ecology and epidemiology of
anthrax in the Etosha National Park, Namibia. Onderstepoort J
Vet Res 61:71?83
Manning CJ, Wakeland EK, Potts WK (1992) Communal nesting
patterns in mice implicate MHC genes in kin recognition. Nature
360:581?583
Mbise AN, Mlengeya TDK, Mollel JO (1998) Septicaemic
salmonellosis of elephants in Tanzania. Bull Anim Health
Prod Afr 46:95?100
Meyer-Lucht Y, Sommer S (2005) MHC diversity and the association
to nematode parasitism in the yellow-necked mouse (Apodemus
flavicollis). Mol Ecol 14:2233?2243
Murphy WJ, Eizrik E, O?Brien SJ, Madsen O, Scally M, Douady CJ,
Teeling E, Ryder OA, Stanhope MJ, de Jong WW, Springer MS
(2001) Resolution of the early placental mammal radiation using
Bayesian phylogenetics. Science 294:2348?2351
Naruse TK, Kawata H, Anzai T, Takashige N, Kagiya M, Nose Y,
Nabeya N, Isshiki G, Tatsumi N, Inoko H (1999) Limited
polymorphism in the HLA-DOAgene. Tissue Antigens 53:359?365
Nei M, Gojobori T (1986) Simple methods for estimating the numbers
of synonymous and nonsynonymous nucleotide substitutions.
Mol Biol Evol 3:418?426
Nei M, Kumar S (2000) Molecular evolution and phylogenetics.
Oxford University Press, New York
Paterson S, Wilson K, Pemberton JM (1998) Major histocompatibility
complex variation associated with juvenile survival and parasite
resistance in a large unmanaged ungulate population (Ovis aries
L.). Proc Natl Acad Sci U S A 95:3714?3719
Penn DJ, Damjanovich K, Potts WK (2002) MHC heterozygosity
confers a selective advantage against multiple-strain infections.
Proc Natl Acad Sci U S A 99:11260?11264
Piertney SB, Olivier MK (2006) The evolutionary ecology of the
major histocompatibility complex. Heredity 96:7?21
Posada D, Crandall KA (1998) MODEL TEST: testing the model of
DNA substitution. Bioinformatics 14:817?818
Rajakaruna RS, Brown A, Kaukinen KH, Miller KM (2006) Major
histocompatibility complex and kin discrimination in Atlantic
salmon and brook trout. Mol Ecol 15:4569?4575
Reche PA, Reinherz EL (2003) Sequence variability analysis of human
class I and class II molecules: functional and structural correlates
of amino acid polymorphisms. J Mol Biol 331:623?641
Richman LK, Montali RJ, Garber RL, Kennedy MA, Lehnhardt J,
Hildebrandt T, Schmitt D, Hardy D, Alcendor DJ, Hayward GS
(1999) Novel endotheliotropic herpesviruses fatal for Asian and
African elephants. Science 283:1171?1176
Richman LK, Montali RJ, Hayward GS (2000) Review of a newly
recognized disease of elephants caused by endotheliotropic
herpesviruses. Zoo Biol 19:383?392
Rozen S, Skaletsky HJ (2000) Primer3 on the WWW for general users
and for biologist programmers. In: Krawetz S, Misener S (eds)
Bioinformatics methods and protocols: methods in molecular
biology. Humana Press, Totowa, NJ, pp 365?386 Source code
available at http://fokker.wi.mit.edu/primer3/.
Ryan SJ, Thompson SD (2001) Disease risk and inter-institutional
transfer of specimens in cooperative breeding programs: herpes
and the elephant species survival plan. Zoo Biol 20:89?101
Scally M, Madsen O, Douady CJ, de Jong WW, Stanhope MJ (2001)
Molecular evidence for the major clades of placental mammals. J
Mamm Evol 8:239?277
Slade RW, Moritz C, Heideman A, Hale PT (1993) Rapid assessment
of single-copy nuclear DNA variation in diverse species. Mol
Ecol 2:359?373
Swofford DL (2002) Phylogenetic analysis using parsimony (*and
other methods). Sinauer, Sunderland
Takahata N, Nei M (1990) Allelic genealogy under overdominant and
frequency-dependent selection and polymorphism at the major
histocompatibility complex loci. Genetics 124:967?978
Waddell PJ, Shelley S (2003) Evaluating placental inter-ordinal
phylogenies with novel sequences including RAG1, [gamma]-
fibrinogen, ND6, and mt-tRNA, plus MCMC-driven nucleotide,
amino acid, and codon models. Mol Phylogenet Evol 28:197?
224
Weber DS, Brent SS, Schienman J, Lehman N (2004) Major
histocompatibility complex variation at three class II loci in the
northern elephant seal. Mol Ecol 13:711?718
Westerdahl H, Walenstrom J, Hansson B, Hasselquist D, von Schantz
T, and Bensch S (2005) Associations between malaria and MHC
genes in a migratory songbird. Proceedings of the Royal Society
of London 272:1511?1518
Zelano B, Edwards SV (2002) An MHC component to kin recognition
and mate choice in birds: predictions, progress, and prospects.
Am Nat 160:S225?S237
Immunogenetics