High frequencies of alleles MICA*020 and MICA*027 in Amerindians and evidence of positive selection on exon 3.

Abstract

MICA is a nonclassical polymorphic MHC molecule. We investigated MICA allelic frequencies and MICA-HLA-B-HLA-C haplotypes in Brazilian Amerindians to describe the polymorphism and to extract information about the evolution of MICA gene. Kaingang is the first population described to have a high frequency of MICA*020, found associated with HLA-B*3505-HLA-Cw*0401. Allele MICA*020 probably originated de novo in South America. The Guarani population had high frequencies of MICA*027. Allele MICA*00801 is common worldwide but rare among Amerindians, occurring only because of gene flow. The analysis of the 64 described MICA alleles revealed that in exons 2 and 4, synonymous substitutions are in excess, a result compatible with purifying selection. The opposite was observed for exons 3 and 6 and the excess of nonsynonymous substitutions was significant for exon 3, indicating positive selection. Few of the alleles described so far had exon 6 sequenced, impeding conclusions for the corresponding portion of the molecule. The analysis of the entire gene is required for a better understanding of the evolution of MICA's polymorphism and its functional consequences. This knowledge is of prime importance in view of the increasing awareness of the functional and medical implications of MICA gene variability.

Full text

Page 1

ORIGINAL ARTICLE
High frequencies of alleles MICA*020 and MICA*027 in
Amerindians and evidence of positive selection on exon 3
LA Oliveira1, F Ribas2, MG Bicalho2, LT Tsuneto1and ML Petzl-Erler1
1Laborato ´rio de Gene ´tica Molecular Humana, Departamento de Gene ´tica, Universidade Federal do Parana ´, Curitiba, Parana ´, Brazil and
2Laborato ´rio de Imunogene ´tica e Histocompatibilidade, Departamento de Gene ´tica, Universidade Federal do Parana ´, Curitiba, Parana ´, Brazil
MICA is a nonclassical polymorphic MHC molecule. We investigated MICA allelic frequencies and MICA–HLA-B–HLA-C
haplotypes in Brazilian Amerindians to describe the polymorphism and to extract information about the evolution of MICA gene.
Kaingang is the first population described to have a high frequency of MICA*020, found associated with HLA-B*3505–HLA-
Cw*0401. Allele MICA*020 probably originated de novo in South America. The Guarani population had high frequencies of
MICA*027. Allele MICA*00801 is common worldwide but rare among Amerindians, occurring only because of gene flow. The
analysis of the 64 described MICA alleles revealed that in exons 2 and 4, synonymous substitutions are in excess, a result
compatible with purifying selection. The opposite was observed for exons 3 and 6 and the excess of nonsynonymous
substitutions was significant for exon 3, indicating positive selection. Few of the alleles described so far had exon 6 sequenced,
impeding conclusions for the corresponding portion of the molecule. The analysis of the entire gene is required for a better
understanding of the evolution of MICA’s polymorphism and its functional consequences. This knowledge is of prime
importance in view of the increasing awareness of the functional and medical implications of MICA gene variability.
Genes and Immunity (2008) 9, 697–705; doi:10.1038/gene.2008.65; published online 4 September 2008
Keywords: MICA; MHC; positive selection; Amerindians
Introduction
The major histocompatibility complex (MHC) is a 4Mb
region located in chromosome 6 (6p21.31). It is histori-
cally divided into three regions: class II (the most
centromeric), class III and class I (the most telomeric).
The classical genes HLA-A, -B and -C are located in the
class I region, together with some nonclassical genes,
such as HLA-E, -F and -G.1The most divergent class I
genes are the MIC (MHC class I chain-related) genes.2
The MIC gene family consists of seven genes, MICA to
MICG, of which only MICA and MICB are expressed.
MICA and MICB sequences are very similar to each
other, with an identity of 91% for the coding sequence.
However, the similarity of MICA extracellular domains
a1, a2 and a3 to other MHC class I genes is only 18, 25
and 30%, respectively.2MICA is located about 40kb from
HLA-B3and linkage disequilibrium (LD) has been
described between these two loci.4
MICA was thought to be normally expressed only in
the gastrointestinal epithelium,3but it is possible that it
occurs in every organ, except for the central nervous
system.5Its expression is stress induced, similar to that of
heat-shock protein 70, because of a putative heat-shock
element in its 50-untranslated region,6and it is commonly
observed in tumors.7,8It binds to an activating receptor
of natural killer (NK) cells, NKG2D, and interacts with gd
T cells, NK cells and CD8 ab T cells.9,10Differently from
other class I molecules, its expression does not depend
on the presence of b2-microglobulin or ATP-binding
cassette transporter TAP.6,11It may have an important
role as a first line of defense in the epithelium signaling
stressed cells to intraepithelial lymphocytes, but still
much has to be learned about the functions of this gene.
Another feature that distinguishes MICA from most
nonclassical class I genes is its high level of polymorph-
ism. The gene has many single nucleotide polymorphisms
(SNP) and also a short tandem repeat (STR) polymor-
phism that is located in exon 5, the region encoding the
transmembrane domain. So far, 64 MICA alleles have been
described (http://www.anthonynolan.org.uk/HIG/lists/
otherlist.html) and only one amino-acid substitution was
found to have functional implications: the proteins with
methionine in position 129 have a higher affinity to the
NKG2D receptor than the proteins with valine at this
position.12
In addition,in
MICA*025, there is a nucleotide substitution that causes
a proline-for-arginine replacement at position six in the a1
domain, affecting the folding of the molecule and
abolishing its expression on the cell surface.13
The STR consists of (GCT)n repeats, corresponding to
alanines in the protein.14
described, A4, A5, A6, A7, A8, A9 and A10, with 4–10
GCT repetitions, and A5.1, which has a G insertion,
causing a frame shift mutation that creates a premature
alleles
MICA*010
and
Eight alleles have been
Received 10 June 2008; revised 15 July 2008; accepted 17 July 2008;
published online 4 September 2008
Correspondence: Professor ML Petzl-Erler, Laborato ´rio de Gene ´tica
Molecular Humana, Departamento de Gene ´tica, Universidade
Federal do Parana ´, Curitiba, Parana ´, Brazil.
E-mail: perler@ufpr.br
Genes and Immunity (2008) 9, 697–705
& 2008 Macmillan Publishers Limited All rights reserved 1466-4879/08 $32.00
www.nature.com/gene

Page 2

Table 1
MICA allelic frequencies (%) in Kaingang and Guarani Amerindians
AlleleCommentPosition
129
KIV
(n)
KRC
(n)
GRC
(n)
GKW
(n)
GND
(n)
TOBA27WIC27TER27JAP20JAP28JAP19KOR29THAI30WEL31BRAaEUR-
AM32
EUR-
AM33
AFR-
AM34
AFR-
AM32
MOR35
MICA*001
MICA*00201
MICA*00202
MICA*004
MICA*006
MICA*00701
MICA*00702
MICA*00801bSTR*A5.1
MICA*00802 STR*A5.1
MICA*00901
MICA*00902
MICA*010
MICA*011
MICA*01201
MICA*015
MICA*016
MICA*017
MICA*018c
MICA*019
MICA*020
MICA*021
MICA*024
MICA*026
MICA*027
MICA*029
MICA*030
MICA*035
MICA*041
MICA*043
MICA*044
MICA*045
MICA*046
MICA*048
MICA*052
deld
met
met
met
val
val
met
met
val
val
val
val
0.4 (1)
35.5 (93) 44.3 (101) 21.5 (43) 64.1 (223) 37.5 (66)
0.5
34.6
0.9
9.6
0.3
4.5
0.3
4.2
1.0
17.7
1.1
14.1
0.2
7.5
0.3
4.6
0.5
43.1
0.3
7.1
1.4
5.8
2.9
1.9
0.2
2.3
1.1
3.9
1.2
2.9
12.6
1.2
27.9
1.4
25.4
3.0
7.347.6 44.512.5 14.69.9 17.817.6
2.7 (7)1.3 (3) 0.3 (1)1.1 11.19.28.212.33.5 11.5
0.7
3.2
5.818.7 19.1
0.1
0.7
0.6
26.6
5.1
1.7
2.1
1.2
2.6
1.2
5.6
0.1
0.6
3.0
0.8
23.2
1.1 (2)1.20.9 1.21.33.31.90.72.4
4.4 (10) 1.5 (3)0.6 (2)3.4 (6)1.00.825.230.8 14.510.821.453.025.3
1.2
13.7
54.526.9
5.5
4.2
1.5
0.7
1.2
1.5
2.7
26.8
0.4 (1) 2.0 (4)8.9 (31) 1.7 (3)1.6
1.6
27.7
0.5
1.2 2.318.416.519.7 10.62.4
0.4
18.2
7.82.9
1.0
5.0
1.7
1.9
8.5
14.0
No expression val27.5 (72) 25.4 (58) 33.0 (66) 14.1 (49)34.1 (60)
0.6 (1)
19.0 17.212.5 10.816.118.34.2
1.5
1.8
7.1
4.7
1.0
0.7
1.7
1.7
3.9
2.5
met
met
met
val
met
met
val
met
met
val
met
val
met
met
met
met
met
val
met
met
val
met
1.8 (4)3.0
0.4 (1)
1.1 (3)
3.0 (8)
10.912.3 14.511.13.1
del exon 42.2 (5)0.6 (2)0.6 (1)1.8
1.5
1.8
2.4
1.5
0.5 (1)3.00.2
2.2
6.1
15.3
0.6
4.8
3.3
2.1
1.9
3.9
1.9
2.3
del exon 4 0.5
2.5
1.0
5.7 (10)
0.6 (1)1.83.5
20.6 (54) 11.8 (27) 0.5 (1)
0.20.3
0.1
0.2
0.1
1.4
0.4 (1) 1.3 (3) 39.5 (79) 11.2 (39)13.6 (24)30.931.035.2 13.60.6
0.3
1.70.50.6
0.20.50.3
0.6
2.2 0.5
0.3
0.1
0.3
0.21.2
11.8
8.2
del gene8.4 (22) 7.0 (16)1.5 (3)0.3 (1) 1.1 (2)6.7 1.2 0.5
Abbreviations: AFR-AM, Afro-Americans; BRA, Brazilian; EUR-AM, Euro-Americans; GKW, Guarani Kaiowa ´; GND, Guarani N˜andeva; GRC, Guarani M’bya ´; JAP, Japanese; KIV, Kaingang
from Ivaı ´; KOR, Korean; KRC, Kaingang from Rio das Cobras; MOR, Morocco; TER, Terena; WEL, Welsh; WIC, Wichi.
Other previously analyzed populations were included for comparison.
n: number of chromosomes.
aRibas et al. MICA polymorphism and linkage disequilibrium with HLA-B alleles in Euro-Brazilians (Tissue Antigens, in press).
bIn the populations from this study it could also be *00804.
cIn the populations from this study it is MICA*01801.
dDeletion of the entire MICA gene.
MICA polymorphisms in Amerindians
LA Oliveira et al
698 Genes and Immunity

Page 3

stop codon, leaving the protein with no cytoplasmic tail
and a different positioning in the cell membrane.15A7,
A8 and A10 are exclusively found in alleles MICA*050,
MICA*055 and MICA*020, respectively, and are very rare:
A7 and A8 have been described in one individual each16
(http://www.ebi.ac.uk/imgt/hla/) and have not been
found in the populations studied so far. A10, described
by Perez-Rodriguez et al.,17has been found only once,
in a single haplotype together with HLA-B*5801 in the
Hunan Han population of Southern China.18
Another frame shift mutation is caused by a G deletion
at the 30end of exon 4, in alleles MICA*015 and
MICA*017. Consequently, exon 5 encodes a novel
polypeptide containing 12 leucine residues, as a result
of the (GCT)9STR allele, and a premature stop codon is
created at the beginning of exon 6.19
Furthermore, a deletion of the entire MICA gene has
been described and is relatively common in some
populations, such as Japanese, Koreans and Angaite
Amerindians. This deletion is found in LD with the allele
HLA-B*4801.20–22
There are several population studies of MICA gene but in
some of them, exons 5 (that includes the STR) and 6 were
not analyzed.23A greater number of population studies
concerned only the STR in the transmembrane region.
In this study, we analyzed the Guarani N˜andeva,
Guarani Kaiowa ´, Guarani M’bya ´ and Kaingang Amer-
indian populations. The Guarani population is from the
linguistic family Tupi-Guarani belonging to the Tupi group
and the Kaingang population belongs to the Macro-Je ˆ
group. Guarani and Kaingang live close to each other but
have remarkable genetic and cultural differences.24,25
Another important feature of these populations is that
gene flow from non-Amerindian and other Amerindian
populations is low, with the exception of Guarani
N˜andeva, whose admixture was estimated as 14%.25,26
The aim of this study was to describe STR_MICA and
MICA allelic frequencies and MICA–HLA-B–HLA-C
associations in the Guarani and Kaingang populations
from Brazil, to compare the diversity found with that of
other populations and to extract information about the
evolution of MICA polymorphism.
Results
The most common alleles in the Guarani population
were MICA*00201, MICA*010 and MICA*027. Their
cumulative frequency corresponds to approximately
85–94% in the three Guarani groups studied (Table 1).
For Guarani N˜andeva and Kaiowa ´, MICA*00201 was the
most frequent (38 and 64%, respectively), whereas for
Guarani M’bya ´, it was MICA*027 (40%). These three
alleles were also the most frequent in the Amerindian
populations Toba, Wichi and Terena27and are part of the
Eastern Asian MICA allele set, in which the alleles
MICA*004, MICA*00801, MICA*00901 and MICA*01201
are also frequent (Table 1).
For the Kaingang populations, the highest frequencies
were observedfor
MICA*00201,
MICA*020, representing 84% of the alleles (Table 1).
MICA*00201 was the most frequent in both Rio das
Cobras (44%) and Ivaı ´ (36%). Interestingly, MICA*020,
which differs from MICA*00201 only by the transmem-
brane STR, was almost twice more common in Ivaı ´ (21%)
than in Rio das Cobras (12%), compensating the
frequency difference of allele MICA*00201 (Table 1 and
above). We emphasize that this is the first report of a
population with a high frequency of MICA*020. The A10
STR allele (Table 2), which is associated exclusively with
allele MICA*020, has thus far been found in just a single
chromosome in a Hunan Han population of Southern
China.18
The STR_MICA alleles A5 and A9 occurred at high
frequencies (420%) in all five Amerindian populations
(Table 2). In addition, A10 and the deletion of the MICA
gene were frequent (47%) only in Kaingang, whereas A6
was frequent (9%) only in Guarani Kaiowa ´.
Haplotypic frequencies and linkage disequilibrium
(LD) between the alleles of the STR in exon 5 of MICA
and HLA-B gene were calculated. MICA and HLA-C
alleles associated with each of the haplotypes are also
listed (Table 3). LD was very high, especially in the
Guarani M’bya ´ population. The deletion of the MICA
gene was found in the same haplotype as HLA-B48 in all
populations. Other haplotypes at high LD were A5–HLA-
B40 and A9–HLA-B39. The allele HLA-B*5104, initially
described as -B53G,26occurred exclusively with A5 and
probably with MICA*010 and HLA-Cw15. Conversely,
MICA*010 and MICA*00201 appeared with more than
one HLA-B allele.
In the Guarani N˜andeva, the most common haplotype
(20%) was MICA*010–HLA-B15–HLA-Cw3, which had
a high frequency (27%) also in the Guarani M’bya ´
population,whosemost
MICA*027–HLA-B40–HLA-Cw3 (38%), whereas for Guar-
ani Kaiowa ´, haplotype MICA*00201–HLA-B39–HLA-Cw7
MICA*010
and
frequenthaplotype was
Table 2
Allelic frequencies of STR_MICA and the respective MICA alleles including these STR variants
STR alleleGND % (n)GKW % (n) GRC % (n)KRC % (n) KIV % (n)MICA alleles
A4
A5
A5.1
A6
A9
A10
dela
6.8 (12)
48.3 (85)
3.4 (6)
2.3 (4)
38.1 (67)
0.8 (2)
30.9 (81)
001/00701/01201/01801
010/016/019/027
00801(4)
004/00901/011
00201/015
020
—
25.3 (88)
0.6 (2)
9.2 (32)
64.7 (225)
73.0 (146)
1.5 (3)
2.0 (4)
21.5 (43)
0.5 (1)
1.5 (3)
26.8 (61)
4.8 (11)
3.5 (8)
46.1 (105)
11.8 (27)
7.0 (16)
2.7 (7)
36.2 (94)
20.6 (54)
8.4 (22) 1.1 (2)0.3 (1)
Abbreviations: GKW, Guarani Kaiowa ´; GND, Guarani N˜andeva; GRC, Guarani M’bya ´; KIV, Kaingang from Ivaı ´; KRC, Kaingang from
Rio das Cobras; STR, short tandem repeat.
n: number of chromosomes.
aDeletion of the entire MICA gene.
MICA polymorphisms in Amerindians
LA Oliveira et al
699
Genes and Immunity

Page 4

was the most common (34%). Interestingly, the only
difference between the MICA alleles of the haplotypes
MICA*00201–HLA-B35–HLA-Cw4 and MICA*020–HLA-
B35–HLA-Cw4, which were the most common in the
Kaingang populations from Rio das Cobras (25 and 10%)
and from Ivaı ´ (16 and 20%), respectively, is the STR in
exon 5 of MICA gene, A9 in allele MICA*00201 and A10
in MICA*020. The haplotype MICA*010–HLA-B*510101
had a high frequency (23%) in both the Kaingang
populations and was observed with different HLA-C
alleles (Cw1, Cw3 or Cw4).
The dendrogram based on MICA genetic distances
showed a correlation of allelic frequencies with geo-
graphic distribution and ethnic characteristics of the
populations. Amerindians integrated a cluster, apart
from populations of other continents and closer to
Eastern Asian populations. The Kaingang were sepa-
rated from the other Amerindians (Figure 1).
The analysis of molecular variance (AMOVA) showed
that 90% of the variation was found among individuals,
Table 3
Two-loci haplotypic frequencies and linkage disequilibrium between STR_MICA and HLA-B
STR
HLA-B
Guarani N˜andeva
Guarani Kaiowa ´
Guarani M0bya ´
Kaingang Rio das Cobras
Kaingang Ivaı ´
MICA
HLA-C
Freq %
D0
P
Freq %
D0
P
Freq %
D0
P
Freq %
D0
P
Freq %
D0
P
A4
18
5.8
0.902
o10?9
0.4
1
o10?9
001/01801
02/07/12/18
A5
40
14.4
1
2.2?10?9
10.8
0.931
o10?9
39.3
1
o10?9
0.4
1
0.141
027
03/04
A5
51
5.2
0.807
0.007
0.6
?0.399
0.390
23.7
0.884
o10?9
23.1
0.929
o10?9
010
01/03/04/05/15
A5
5104
4.0
1
0.005
6.6
1
0.027
010
15
A5
62
22.4
0.862
o10?9
13.7
0.972
o10?9
27.0
0.794
3?10?5
2.2
0.610
0.007
4.3
0.878
5?10?6
010
01/03/04/07/11/15
A9
35
21.8
0.959
o10?9
26.7
0.970
o10?9
19.4
0.937
o10?9
24.6
0.428
2.1?10?8
15.6
0.064
0.328
00201
03/04/15
A9
39
13.8
1
o10?9
36.9
1
o10?9
1.5
1
6.9?10?4
19.7
0.921
o10?9
18.0
1
o10?9
00201
04/07
A10
35
10.1
0.770
9?10?9
19.9
0.968
o10?9
020
04
dela
48
1.2
1
o10?9
0.3
1
o10?9
1.5
1
o10?9
7.0
1
o10?9
8.6
1
o10?9
del
03/04/07/08
Others
11.5
11.0
4.6
12.7
9.8
Abbreviation: STR, short tandem repeat.
MICA and HLA-C alleles associated with these haplotypes are also shown. Only haplotypes with frequencies of at least 5% in at least one population are shown.
D0: relative value of linkage disequilibrium; Freq %: haplotypic frequencies; P: probability value relative to the linkage disequilibrium.
aDeletion of the entire MICA gene.
MOROCCO
AFR-AM-G
AFR-AM-Z
EURO-AM-G
EURO-AM-P
WELSH
BRAZ
JAP-OB
JAP-KW
JAP-KAT
THAI
GND
GKW
TERENA
WICHI
TOBA
GRC
KIV
0.01
KRC
Figure 1
genetic distances for the MICA gene. Morocco;35AFR-AM: Afro-
Americans (G;32Z34); EURO-AM: Euro-Americans (G;32P33);
Welsh;31BRAZ: Brazilian (Ribas et al. MICA polymorphism and
linkage disequilibrium with HLA-B alleles in Euro-Brazilians Tissue
Antigens, in press); JAPAN: Japanese (OB;19KAT;28KW20); THAI:
Thailand;30Terena, Wichi and Toba;27GND: Guarani N˜andeva;
GKW: Guarani Kaiowa ´; GRC: Guarani M’bya ´; KIV: Kaingang from
Ivaı ´ and KRC: Kaingang from Rio das Cobras.
Dendrogram based on Cavalli-Sforza’s chord measure of
MICA polymorphisms in Amerindians
LA Oliveira et al
700
Genes and Immunity

Page 5

9% among populations within groups (that is, among the
different Guarani and between the different Kaingang
populations) and only 1% of the variation occurred
among groups (that is, between Guarani and Kaingang
populations).
When comparing the number of synonymous sub-
stitutions per synonymous site and the number of
nonsynonymous substitutions per nonsynonymous site,
synonymous substitutions were in excess for exons 2 and
4 and nonsynonymous substitutions were in excess for
exons 3 and 6, indicating deviation from selective
neutrality (Table 4). The codon-based test of selective
neutrality showed significant deviation only for exon 3
(P-value of 0.007), revealing positive selection.
Discussion
The distribution of MICA alleles in Amerindians differs
clearly from that of populations in other continents.
Noticeable differences are seen also between the Amer-
indian groups. Each Kaingang and Guarani population
has only three to four common MICA alleles of which
only two are frequent in all of the examined Amerindian
populations. These are MICA*00201 and MICA*010,
which occur also in populations of the other continents.
Allele MICA*020 is extremely rare in all populations thus
far analyzed, but occurs at elevated frequency in the
Kaingang population, whereas MICA*027 is common in
most Amerindians and in Eastern Asians, but not in the
Kaingang population, and MICA*00901 is common only
in the Guarani Kaiowa ´ population and in non-Amer-
indians. The distinctiveness of the Guarani and Kain-
gang populations is in line with the results of earlier
studies.25,26Among the Amerindians, Guarani N˜andeva
had the highest genetic diversity, which agrees with the
relatively high amount of gene flow (14%) from non-
Amerindians into this population.25On the other hand,
Guarani M’bya ´ had a lower diversity and the highest
levels of LD, possibly reflecting their genetic isolation
and the effect of random genetic drift.
A series of alleles at frequencies typically lower than
5%, observed in the Kaingang and Guarani, was most
probably introduced into these populations by post-
Columbian gene flow from Europeans and Africans:
MICA*001, -*004, -*00701, -*00801, -*011, -*01201, -*015,
-*01801 and -*019. This is evidenced by the HLA alleles of
the corresponding haplotypes, which do not occur in
isolated Amerindian populations but are common in
Europe or Africa, such as HLA-B7, -B8, -B13, -B14, -B18,
-B37 and -B45.
It seems likely that MICA*020 originated de novo in
South America, in a haplotype MICA*00201–HLA-
B*3505–HLA-Cw*0401. MICA*020 was described in a cell
line of Caucasian origin that had the HLA-B alleles -B35
and-B7.17
Itdiffers from
MICA*00201 because of the presence of the STR allele
A10 instead of A9. The STR allele A10 has been found in
a Hunan Han population of Southern China in a single
chromosome, along with HLA-B*5801.18In the Kaingang
population, the allele was found in a different haplotype,
MICA*020–HLA-B*3505–HLA-Cw*0401.
HLA-B35 alleles were -B*3501, -B*3506 or -B*3505, the
latter two being exclusively found in Amerindians from
South America and thought to have originated in this
continent by gene conversion with HLA-B*3501 as the
founding allele.36In the Kaingang population, HLA-
B*3501 occurred together with MICA*00201. In the
Guarani population, HLA-B*3501 has not been detected
and -B*3505 occurred in cis with MICA*00201. Thus, it
seems likely that MICA*020 originated de novo in South
America in a haplotype MICA*00201–HLA-B*3505–HLA-
Cw*0401. The frequency of this new haplotype may have
been increased by positive selection of the HLA-B allele
or the MICA allele. Alternatively, random genetic drift
might have been responsible for increasing its frequency.
For MICA*020 to be originated, the only mutation
required is an addition of one repetition unit in the
STR of exon 5. So, it is not unlikely that this allele was
originated de novo. Microsatellites (STR) have a high
mutation rate and the addition of one repetition unit is
usually the most frequent one.37Therefore, the only
factor necessary to explain this convergent evolution is
the mechanism of microsatellite mutation, by which the
same allele may arise independently more than once.
Although MICA*00801 is the most frequent allele in
almost all populations studied so far except Amerindians
(Table 1), MICA*027 is common in Eastern Asians28,29and
in all Amerindians analyzed,27except in the Kaingang
population (Table 1). The widespread distribution of
allele MICA*00801 and its occurrence at high frequency
and in many HLA haplotypes have led to the suggestion
that it might be the ancestral MICA allele.33Interestingly,
MICA*027 differs from MICA*00801 only by the STR in
exon 5. MICA*00801 has A5.1, the STR allele with a G
insertion that causes a frame shift mutation, and a
premature stop codon, whereas MICA*027 has A5, which
lacks the insertion. We remind that the microsatellite was
not analyzed in all studies, so MICA*027 may have
remained undetected in other populations. In this study,
MICA*027 was associated with HLA-B40 as in Toba and
Wichi, and not with -B48, a combination also observed in
these other Amerindian populations27and in Japanese.28
In the Korean population, it was associated mainly with
HLA-B61(40).29MICA*00801 has been observed in hap-
lotypes with many HLA-B alleles, including -B40.4
MICA*027 may have originated from MICA*00801 by a
deletion of G in the STR region or, alternatively, by
recombination or gene conversion with an allele with
five STR repetitions and no insertion. The high frequen-
cies of MICA*00801 may result from positive selection. It
has been shown that a protein of the human cytomega-
lovirus is able to downregulate surface expression of
MICA, but not of MICA*00801, suggesting that this allele
thewidespreadallele
TheKaingang
Table 4
site and non-synonymous substitutions per non-synonymous site to
test the hypothesis of selective neutrality
Analysis of synonymous substitutions per synonymous
MICA
Ns
Ms
Ms/Ns
Na
Ma
Ma/Na
Exon 2
Exon 3
Exon 4
Exon 5
Exon 6
57.1
64.1
64.0
32.2
32.1
3
1
6
1
1
0.0526
0.0156
0.0938
0.0311
0.0312
194.9
223.9
215.0
99.8
93.9
50.0257
0.0715
0.0465
0.0301
0.0426
16
10
3
4
Ns: number of synonymous sites (average for all sequences); Ms:
number of synonymous substitutions; Na: number of non-synon-
ymous sites (average for all sequences); Ma: number of non-
synonymous substitutions.
MICA polymorphisms in Amerindians
LA Oliveira et al
701
Genes and Immunity

Page 6

may be protective against cytomegalovirus infections.38
Concerning the hypothesis of MICA*00801 being the
ancestral allele, it is more likely that the origin of MICA
alleles had beenpolyphyletic.39
MICA*00801 seems to be one of the oldest alleles, and
its association with HLA-B40 in populations from many
continents4indicates that MICA*00801/027-HLA-B40 hap-
lotypes may have been conserved.
The only allele observed at high frequency (X7%)
worldwide is MICA*00201. In this study, MICA*00201
was associated with HLA-B35 and -B39. Association with
HLA-B35 has been observed in populations from all
continents, whereas association with HLA-B39 has been
observed in Japanese and Amerindian.4Interestingly, the
frequency of MICA*00201 differed between the Kaingang
population from Ivaı ´ and that from Rio das Cobras, as
did the frequency of MICA*020. However, the sum of the
frequencies of these two alleles was very similar (56%) in
the two Kaingang populations. It should be noticed that
these two populations did not differ significantly for
most genetic polymorphisms so far analyzed. The cause
of the difference in the frequencies of the alleles
MICA*00201 and MICA*020 between the Kaingang
populations is not evident. The two groups live ca.
180km apart but maintain similar habits and living
conditions.
Another example of apparent allele substitution is
MICA*027 replacing MICA*00801 in all Amerindian
populations analyzed, apart from the Kaingang. As
commented above, this allele differs from MICA*00801
only by the STR in exon 5, which is not analyzed in most
studies. Another allele with high frequency among
Guarani and Kaingang populations is MICA*010, found
worldwide, but common only in Eastern Asians and
Amerindians. This allele encodes a protein with no
surface expression.13In this study, MICA*010 was found
associated with HLA-B15 and -B51. The association with
HLA-B15 is common in other populations, whereas HLA-
B51 occurs mostly with MICA*004 and MICA*009.4
The Kaingang population had a high frequency of
MICA gene deletion. This deletion occurs in Eastern
Asians21and presented a high frequency (38%) in the
Angaite Amerindians from Paraguay.22It was associated
with allele HLA-B*4801 and, in most cases, with a null
allele of MICB. Apparently, individuals homozygous for
the deletion are normal and it has been hypothesized
that this haplotype is maintained by natural selection.21,40
Surprisingly, it seems that all populations have high
frequencies of MICA alleles that may have an altered
function or no function at all. If we sum the frequencies
(Table 1) of the deletion and of alleles MICA*00801,
MICA*00802,
MICA*00804,
MICA*053 (alleles with the G insertion in the STR of
exon 5), MICA*010, MICA*025, MICA*054 (alleles that
have no surface expression because of a proline-for-
arginine substitution in the a1 domain), MICA*015 and
MICA*017 (alleles with a G deletion at the end of exon 4,
encoding a polypeptide with a different amino-acid
sequence in the transmembrane region and with a short
cytoplasmic tail), the total frequency is as high as 50–63%
in Euro-Americans32,33and Welsh31and varies from 30 to
44% in other non-Amerindian populations. In Amerin-
dians, this summed frequency is approximately 18% in
Wichi, Terena27and Guarani Kaiowa ´ and 30–40% in the
other populations. In conclusion, it seems that MICA
Nevertheless,
MICA*023,
MICA*028,
alleles with no or limited function are not prejudicial to
its carriers and may even have been subjected to positive
selection or had their frequency raised by hitchhiking
resulting from selection of the closely linked HLA-B or
other MHC genes. In addition, the high frequency of
these alleles leads us to conclude that much is still to be
learned about the functions of MICA.
Position 129 of the MICA protein has been related to
the affinity of binding to NKG2D receptor.12If we add
the alleles with low affinity, the ones with valine at
position 129, the summed frequency of alleles with
altered function mentioned above achieves 75–80% in the
Morocco population,35Japanese,20Welsh31and Euro-
Americans33and is of at least 36% (in Guarani Kaiowa ´
population). However, the functional implications of the
binding affinity between MICA and its receptor are still
unknown.
Interestingly, 38 of the 52 SNP in MICA cause an
amino-acid substitution, which could indicate that
selection may play a role in maintaining MICA diversity.
However, the tests for selective neutrality showed
evidence of positive selection only for exon 3 of the
MICA gene, which codes for the a2 domain on the
protein. A similar result was previously observed in an
analysis of exons 2–4 of only 51 MICA alleles.41However,
we observed that for exons 2 and 4, which code for a1
and a3 domains, the tendency is the opposite, pointing to
purifying selection, but the results are not statistically
significant. It has been suggested that the high variability
in exon 3 could be reminiscent of the peptide groove of
HLA molecules.41However, this groove is formed by
both exons 2 and 3, and the former did not show an
excess of nonsynonymous substitutions. For exon 6,
which codes for the cytoplasmic domain, the results
point to positive selection, but P-values were of 0.1, thus
not significant. However, only 17 sequences could be
used for this analysis, because most alleles are not
sequenced for exon 6. It is possible that, with more
sequences available, interesting results could be found
for this portion of the molecule.
It may be advantageous for a population to have
MICA molecules with different affinities to its receptor,
as has been suggested for HLA class I molecules,40which
could probably explain the high level of polymorphism
and the high frequencies of MICA alleles with altered
function or no function at all (see above). This hypothesis
is supported by the evidences of positive or balancing
selection observed in this study and by others.41,42
MICA was found to be in LD with HLA-B in many
populations. Most of the MICA–HLA-B haplotypes
observed in this study were also found in other
populations, with the exception of MICA*010–HLA-B51
and MICA*020–HLA-B35.4
To conclude, this is the first report of populations with
high frequency of MICA*020. We present evidence that
MICA*020 probably originated de novo in South America
in a haplotype MICA*00201–HLA-B*3505–HLA-C*0401.
Its high frequency in the Kaingang population may be
because of natural selection favoring MICA or HLA-B or
other MHC alleles in LD in this haplotype. Another
possibility is that the frequency of MICA*020 allele rose
because of genetic drift or other demographic events.
Othercommon alleles
MICA*00201, MICA*010 and MICA*027. The first two
have a worldwide distribution and high frequencies in
inAmerindians were
MICA polymorphisms in Amerindians
LA Oliveira et al
702
Genes and Immunity

Page 7

almost all populations analyzed, whereas MICA*027
occurs in a subset of the Amerindian, and possibly also
of the Eastern Asian populations. A complete deletion of
the gene was also found with high frequency in the
Kaingang population. MICA*00801 is very common
worldwide, but in Guarani and Kaingang, it seems to
occur only because of gene flow with non-Amerindian
populations, as do MICA*001, -*004, -*00701, -*011,
-*01201, -*015, -*01801 and -*019. Moreover, all popula-
tions have a significant frequency of alleles that seem to
have an altered function or no function at all, indicating
that these alleles may not be prejudicial and may even
have been subjected to positive selection and that much
remains to be learned about MICA function and
evolution. Natural selection seems to have an important
function in MICA evolution, and positive selection was
observed for exon 3. Selection may account for the high
frequencies of MICA*00801 in almost all non-Amerindian
populations and the high frequency of MICA deletion in
Asians and Amerindians. The comparison between
populations is hampered by the fact that, in several
studies, certainalleleswere
MICA*020, MICA*027 and MICA*049), and this may
result in false assignments. We emphasize the impor-
tance of typing the whole gene, including exons 5 (with
the STR) and 6. Many MICA alleles still do not have their
sequences determined for these exons and new alleles
may be recognized when the whole gene is analyzed in
more populations. So, MICA diversity may be even
higher than what we consider nowadays. The correct
description of MICA polymorphism is of great impor-
tance in view of the increasing awareness of the
functional and medical implications of its variability.
nottested (namely
Materials and methods
Population samples
The Kaingang population sample (n¼245) is from Ivaı ´
(241300S, 511400W) and Rio das Cobras (251180S, 521320W),
both in Parana ´ state, Brazil. The Guarani M’bya ´ popula-
tion (n¼100) is also from Rio das Cobras. This reserva-
tion area is shared by Guarani M’bya ´ and Kaingang
populations. The Guarani N˜andeva population (n¼89) is
from Amambaı ´ (231060S; 551120W) and Porto Lindo
(231480S; 541300W), in the state of Mato Grosso do Sul,
Brazil. The Guarani Kaiowa ´ population (n¼174) is from
Lima ˜o Verde (231120S; 551060W) and also from Amambaı ´
(Figure 2).
Collection of the blood samples and DNA extraction
were as described before.25,26,43
MICA typing
The STR typing consisted of PCR amplification of the
region containing the microsatellite with the following
primers (50-30sequences): STRMICAF: cctttttttcagggaaa
gtgc and STRMICAR: ccttaccatctccagaaactgc,44followed
by electrophoresis in an automated DNA sequencer,
MegaBACE 1000 (GE Healthcare, Pollards Wood, UK).
STRMICAF was labeled with VIC (a fluorescent dye).
PCR had 0.25mM dNTP, 1.5mM MgCl2, 0.25U of Taq
DNA Polymerase Platinum (Invitrogen, Carlsbad, CA,
USA), 1? buffer (20mM Tris HCl (pH¼8), 40mM NaCl,
2mM Na2SO4, 0.1mM dithiothreitol, glycerol 50%), 0.05mg
of DNA and 0.2mM of each primer. The reaction was
performed in a thermocycler (Eppendorf Mastercycler
epgradient S) with the first step at 941C for 3min,
followed by 30 cycles at 941C for 30s, 54.51C for 30s and
721C for 1min, and a final step at 721C for 40min. After
PCR, the products were diluted in ultrapure water (2ml
of the product in 8ml of water). Then, 2ml of the diluted
product were mixed with 0.3ml of the size standard
ET400-R (GE Healthcare) and 6.7ml of 0.02? Tween. The
products were denaturated for 3min at 941C. The
analyses were performed using Fragment Profiler ver-
sion 1.2 (GE Healthcare).
On the basis of STR_MICA–HLA-B–HLA-C haplotypes
(see below), samples were selected for MICA typing,
which consisted of PCR-RSSOP (PCR-sequence-specific
oligonucleotide probes reverse) using LABType RSSO
MICA kit (One Lambda Inc., Canoga Park, CA, USA)
and Luminex technology. All procedures followed the
manufacturer’s instructions. Assignment of each indivi-
dual’s genotype was achieved through the software HLA
Visual 2.2 (One Lambda Inc.). Ambiguous results were
solved by PCR-SSP (PCR-sequence-specific primers) as
described by Collins et al.45and Rees et al.31The
ambiguitybetweenthe
MICA*049 was solved by PCR-RFLP (PCR-restriction
fragment length polymorphism), amplifying the region
containing the SNP that differs between the two alleles
with the primers MICAF: 50-agctcgtgagcctgca-30and
MICAR: 50-agtggagccagtggacccaag-30and then digesting
the products with BseGI restriction enzyme.
The HLA-B and -C genes were typed earlier26,36
(Tsuneto et al., manuscript in preparation).
alleles
MICA*00901
and
Statistical analysis
First, haplotypic and allelic frequencies were obtained
for STR_MICA, HLA-B and HLA-C using the ELB and
Figure 2
reservation areas where the Guarani and Kaingang populations
from this study live. (1) Ivaı ´ (KIV—Kaingang from Ivaı ´); (2) Rio das
Cobras (KRC and GRC—Kaingang and Guarani from Rio das
Cobras); (3) Amambaı ´ and Lima ˜o Verde; and (4) Porto Lindo
(GND—Guarani N˜andeva and GKW—Guarani Kaiowa ´).
Map of South America showing the location of the
MICA polymorphisms in Amerindians
LA Oliveira et al
703
Genes and Immunity

Page 8

EM algorithms included in the software package
Arlequin v. 3.01,46attaining similar results. When the
observed frequency of a haplotype is different from the
expected one, the loci are said to be in LD. The
parameters to measure LD (D, D0and P) between alleles
from STR_MICA and HLA-B were also estimated using
Arlequin v. 3.01.46
The haplotypic frequencies calculated for the STR in
exon 5 of MICA and the HLA genes B and C were used to
select a subsample for MICA allelic typing with LABType
RSSOP MICA kit (One Lambda Inc., Canoga Park, CA,
USA). This subsample was carefully chosen to allow an
accurate inference of MICA alleles in the remaining
samples and contained at least two and up to 20
individuals with each of the haplotypes, with the
exception of rare haplotypes for which just one indivi-
dual was available. To allow this selection of individuals
to the subsample and the posterior inference of MICA
alleles, it was necessary to know each individual
haplotype. The haplotypes for each individual were
manually checked based on the list of common haplo-
types (frequencies X1%) estimated by ELB and EM
algorithms (see above). When only one haplotype from
this list could be assigned for the individual, the other
haplotype was deduced from the remaining alleles. In
the great majority of cases (558 of 608), the two
haplotypes were successfully inferred. For 48 indivi-
duals, some information was missing for at least one
locus, making the deduction of both haplotypes im-
possible. For the other two individuals, there was an
ambiguity in the typing result for one of the loci. When
the haplotype was not inferred, if MICA was not typed
for that individual sample, the allele was also not
inferred. MICA was then typed by PCR-RSSOP in a
subsample that included 107 Guarani and 67 Kaingang
individuals. PCR-SSP typing was performed in some of
those individuals to solve ambiguities and in an
additional sample including 52 Guarani and 37 Kaingang
individuals to enlarge the sample for some of the
common haplotypes. The MICA alleles found were
assumed to occur in the other individuals carrying the
same STR_MICA–HLA-B–HLA-C haplotypes as the in-
dividuals who had their MICA alleles determined. This
was possible because of the extreme LD seen among the
STR_MICA, MICA, HLA-B and HLA-C alleles in all
instances (see results). In addition, PCR-RFLP (as
described above) was performed for all 39 individuals
assumed to have allele MICA*00901 to distinguish it
from allele MICA*049.
Cavalli-Sforza’s chord measure of genetic distances
was estimated47and used to construct a dendrogram by
the Fitch-Margoliash method.48Both analyses were made
with the software package Phylip.49The dendrogram
was drawn using TreeView 1.6.6.50
AMOVA was carried out using the software package
Arlequin v. 3.01.46For this analysis, the three Guarani
populations defined a group, as did the two Kaingang
populations.
The software DNA Sequence Polymorphism 451was
used to estimate the number of synonymous and
nonsynonymous sites and the number of synonymous
and nonsynonymous substitutions. To test the hypoth-
esis of selective neutrality, the codon-based test of
neutrality averaged over all pairs of the 64 MICA alleles
was used, with the help of MEGAversion 4.52These tests
were done for the whole sequence and also for each exon
separately.
Acknowledgements
We thank the individuals of the populations analyzed in
this research for their collaboration. The support and
friendship of our laboratory colleagues are highly
valued. We are also grateful for the support of Conselho
Nacional de Desenvolvimento Cientı ´fico e Tecnolo ´gico
(CNPq), Institutos do Mile ˆnio, Fundac ¸a ˜o Arauca ´ria de
Apoio ao Desenvolvimento Cientı ´fico e Tecnolo ´gico do
Parana ´, Fundac ¸a ˜o da Universidade Federal do Parana ´
(FUNPAR) and Coordenac ¸a ˜o de Aperfeic ¸oamento de
Pessoal de Nı ´vel Superior (CAPES).
References
1 The MHC sequencing consortium. Complete sequence and
gene map of a human major histocompatibility complex.
Nature 1999; 401: 921–923.
2 Bahram S, Spies T. The MIC gene family. Res Immunol 1996;
147: 328–333.
3 Bahram S, Bresnahan M, Geraghty DE, Spies T. A second
lineage of mammalian major histocompatibility complex class
I genes. Proc Natl Acad Sci USA 1994; 91: 6259–6263.
4 Collins RW. Human MHC class I chain related (MIC) genes:
their biological function and relevance to disease and
transplantation. Eur J Immunogenet 2004; 31: 105–114.
5 Schrambach S, Ardizzone M, Leymarie V, Sibilia J, Bahram S.
In vivo expression pattern of MICA and MICB and its
relevance to auto-immunity and cancer. PLoS ONE 2007; 2:
e518.
6 Groh V, Bahram S, Bauer S, Herman A, Beauchamp M, Spies T.
Cell stress-regulated human major histocompatibility complex
class I gene expressed in gastrointestinal epithelium. Proc Natl
Acad Sci USA 1996; 93: 12445–12450.
7 Groh V, Rhinehart R, Secrist H, Bauer S, Grabstein KH, Spies T.
Broad tumor-associated expression and recognition by tumor-
derived gamma delta T cells of MICA and MICB. Proc Natl
Acad Sci USA 1999; 96: 6879–6884.
8 Zwirner NW, Fuertes MB, Girart MV, Domaica CI, Rossi LE.
Cytokine-driven regulation of NK cell functions in tumor
immunity: role of the MICA-NKG2D system. Cytokine Growth
Factor Rev 2007; 18: 159–170.
9 Bauer S, Groh V, Wu J, Steinle A, Phillips JH, Lanier LL et al.
Activation of NK cells and T cells by NKG2D, a receptor for
stress-inducible MICA. Science 1999; 285: 727–729.
10 Groh V, Rhinehart R, Randolph-Habecker J, Topp MS,
Riddell SR, Spies T. Costimulation of CD8alphabeta T cells
by NKG2D via engagement by MIC induced on virus-infected
cells. Nat Immunol 2001; 2: 255–260.
11 Steinle A, Groh V, Spies T. Diversification, expression, and
gamma delta T cell recognition of evolutionarily distant
members of the MIC family of major histocompatibility
complex class I-related molecules. Proc Natl Acad Sci USA
1998; 95: 12510–12515.
12 Steinle A, Li P, Morris DL, Groh V, Lanier LL, Strong RK et al.
Interactions of human NKG2D with its ligands MICA, MICB,
and homologs of the mouse RAE-1 protein family. Immuno-
genetics 2001; 53: 279–287.
13 Li Z, Groh V, Strong RK, Spies T. A single amino acid
substitution causes loss of expression of a MICA allele.
Immunogenetics 2000; 51: 246–248.
14 Mizuki N, Ota M, Kimura M, Ohno S, Ando H, Katsuyama Y
et al. Triplet repeat polymorphism in the transmembrane
region of the MICA gene: a strong association of six GCT
MICA polymorphisms in Amerindians
LA Oliveira et al
704
Genes and Immunity

Page 9

repetitions with Behc ¸et disease. Proc Natl Acad Sci USA 1997;
94: 1298–1303.
15 Suemizu H, Radosavljevic M, Kimura M, Sadahiro S,
Yoshimura S, Bahram S et al. A basolateral sorting motif in
the MICA cytoplasmic tail. Proc Natl Acad Sci USA 2002; 99:
2971–2976.
16 Rueda B, Pascual M, Lopez-Nevot MA, Gonzalez E, Martin J.
A new allele within the transmembrane region of the human
MICA gene with seven GCT repeats. Tissue Antigens 2002; 60:
526–528.
17 Pe ´rez-Rodrı ´guez M, Corell A, Argu ¨ello JR, Cox ST, McWhin-
nie A, Marsh SGE et al. A new MICA allele with ten alanine
residues in the exon 5 microsatellite. Tissue Antigens 2000; 55:
162–165.
18 Tian W, Zeng XM, Li LX, Jin HK, Luo QZ, Wang F et al.
Gender-specific associations between MICA-STR and naso-
pharyngeal carcinoma in a southern Chinese Han population.
Immunogenetics 2006; 58: 113–121.
19 Obuchi N, Takahashi M, Nouchi T, Satoh M, Arimura T, Ueda K
et al. Identification of MICA alleles with a long Leu-repeat in
the transmembrane region and no cytoplasmic tail due to a
frameshift-deletion in exon 4. Tissue Antigens 2001; 57: 520–535.
20 Komatsu-Wakui M, Tokunaga K, Ishikawa Y, Kashiwase K,
Moriyama S, Tsuchiya N et al. MIC-A polymorphism in
Japanese and a MIC-A-MIC-B null haplotype. Immunogenetics
1999; 49: 620–628.
21 Komatsu-Wakui M, Tokunaga K, Ishikawa Y, Leelayuwat C,
Kashiwase K, Tanaka H et al. Wide distribution of the MICA-
MICB null haplotype in East Asians. Tissue Antigens 2001; 57:
1–8.
22 Aida K, Russomando G, Kikuchi M, Candia N, Franco L,
Almiron M et al. High frequency of MIC null haplotype (HLA-
B48-MICA-del-MICB*0107 N) in the Angaite Amerindian
community in Paraguay. Immunogenetics 2002; 54: 439–441.
23 Frigoul A, Lefranc MP. MICA: standardized IMGT allele
nomenclature, polymorphisms and diseases. Recent Res Dev
Hum Genet 2005; 3: 95–145.
24 Belich MP, Madrigal JA, Hildebrand WH, Zemmour J,
Williams RC, Luz R et al. Unusual HLA-B alleles in two tribes
of Brazilian Indians. Nature 1992; 357: 326–329.
25 Tsuneto LT, Probst CM, Hutz MH, Salzano FM, Rodriguez-
Delfin LA, Zago MA et al. HLA class II diversity in seven
Amerindian populations. Clues about the origins of the Ache ´.
Tissue Antigens 2003; 62: 512–526.
26 Petzl-Erler ML, Luz R, Sotomaior VS. The HLA polymorphism
of two distinctive South-American Indian tribes: the Kaingang
and the Guarani. Tissue Antigens 1993; 41: 227–237.
27 Zhang Y, Lazaro AM, Zou Y, Lavignia B, Moraes EM, Moraes
RJ et al. MICA polymorphism in South American Indians.
Immunogenetics 2002; 53: 900–906.
28 Katsuyama Y, Ota M, Ando H, Saito S, Mizuki N, Kera J et al.
Sequencing based typing for genetic polymorphisms in
exons, 2, 3 and 4 of the MICA gene. Tissue Antigens 1999; 54:
178–184.
29 Pyo CW, Hur SS, Kim YK, Choi HB, Kim TY, Kim TG.
Distribution of MICA alleles and haplotypes associated with
HLA in the Korean population. Hum Immunol 2003; 64:
378–384.
30 Romphruk AV, Naruse TK, Romphruk A, Kawata H,
Puapairoj C, Kulski JK et al. Diversity of MICA (PERB11.1)
and HLA haplotypes in Northeastern Thais. Tissue Antigens
2001; 58: 83–89.
31 Rees MT, Downing J, Darke C. A typing system for the major
histocompatibility complex class I chain related genes A and B
using polymerase chain reaction with sequence-specific
primers. Genet Test 2005; 9: 93–110.
32 Gao X, Single RM, Karacki P, Marti D, O0Brien SJ, Carrington
M. Diversity of MICA and linkage disequilibrium with HLA-B
in two North American populations. Hum Immunol 2006; 67:
152–158.
33 Petersdorf EW, Shuler KB, Longton GM, Spies T, Hansen JA.
Population study of allelic diversity in the human MHC class
I-related MIC-A gene. Immunogenetics 1999; 49: 605–612.
34 Zhang Y, Han M, Vorhaben R, Giang C, Lavingia B, Stastny P.
Study of MICA alleles in 201 African Americans by multi-
plexed single nucleotide extension (MSNE) typing. Hum
Immunol 2003; 64: 130–136.
35 Piancatelli D, Del Beato T, Oumhani K, El Aouad R, Adorno D.
MICA polymorphism in a population from north Morocco,
Metalsa Berbers, using sequence-based typing. Hum Immunol
2005; 66: 931–936.
36 Parham P, Arnett KL, Adams EJ, Little AM, Tees K, Barber LD
et al. Episodic evolution and turnover of HLA-B in the
indigenous human populations of the Americas. Tissue
Antigens 1997; 50: 219–232.
37 Ellegren H. Microsatellites: simple sequences with complex
evolution. Nat Rev Genet 2004; 5: 435–445.
38 Chalupny NJ, Rein-Weston A, Dosch S, Cosman D. Down-
regulation of the NKG2D ligand MICA by the human
cytomegalovirus glycoprotein UL142. Biochem Biophys Res
Commun 2006; 346: 175–181.
39 Choy MK, Phipps ME. Possible polyphyletic origin of major
histocompatibility complex class I chain-related gene A
(MICA) alleles. J Mol Evol 2003; 57: 38–43.
40 Martinez-Naves E, Barber LD, Madrigal JA, Vullo CM,
Clayberger C, Lyu SC et al. Interactions of HLA-B*4801 with
peptide and CD8. Tissue Antigens 1997; 50: 258–264.
41 Elsner HA, Schroeder M, Blasczyc R. The nucleotide diversity
of MICA and MICB suggests the effect of overdominant
selection. Tissue Antigens 2001; 58: 419–421.
42 Ohashi J, Tokunaga T. Computer simulation analysis suggests
weak balancing selection operative at the MICA locus.
Hereditas 2000; 133: 25–28.
43 Petzl-Erler ML, McDevitt HO. Molecular analysis of the HLA-
DRB genes in two tribes of Brazilian Indians. Hum Immunol
1994; 41: 180–184.
44 Ota M, Katsuyama Y, Mizuki N, Ando H, Furihata K,
Ono S et al. Trinucleotide repeat polymorphism within
exon 5 of the MICA gene (MHC class I chain-related gene
A): allele frequency data in the nine population groups
Japanese, Northern Han, Hui, Uygur, Kazakhstan, Iranian,
Saudi, Arabian, Greek and Italian. Tissue Antigens 1997; 49:
448–454.
45 Collins RW, Stephens HA, Clare MA, Vaughan RW. High
resolution molecular phototyping of MICA and MICB
alleles using sequence specific primers. Hum Immunol 2002; 63:
783–794.
46 Excoffier L, Laval G, Schneider S. Arlequin ver. 3.0: An
integrated software package for population genetics data
analysis. Evol Bioinform Online 2005; 1: 47–50.
47 Cavalli-Sforza LL, Edwards AWF. Phylogenetic analysis:
models and estimation procedures. Evolution 1967; 32: 550–570.
48 Fitch WM, Margoliash E. Construction of phylogenetic trees.
Science 1967; 155: 279–284.
49 Felsenstein J. PHYLIP (Phylogeny Inference Package). version
3.6. Distributed by the author Department of Genome
Sciences, University of Washington: Seattle, 2004.
50 Page RDM. TREEVIEW: An application to display phyloge-
netic trees on personal computers. Comput Appl Biosci 1996; 12:
357–358.
51 Rozas J, Sa ´nchez-DelBarrio JC, Messeguer X, Rozas R. DnaSP,
DNA polymorphism analyses by the coalescent and other
methods. Bioinformatics 2003; 19: 2496–2497.
52 Tamura K, Dudley J, Nei M, Kumar S. MEGA4: Molecular
Evolutionary Genetics Analysis (MEGA) software version 4.0.
Mol Biol Evol 2007; 24: 1596–1599.
MICA polymorphisms in Amerindians
LA Oliveira et al
705
Genes and Immunity