![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
|
Vol. 10, Issue 10, 1579-1586, October 2000
LETTER
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
RESULTS AND DISCUSSION
METHODS
REFERENCES
|
|---|
The human major histocompatibility complex (MHC) is characterized by polymorphic multicopy gene families, such as HLA and MIC (PERB11); duplications; insertions and deletions (indels); and uneven rates of recombination. Polymorphisms at the antigen recognition sites of the HLA class I and II genes and at associated neutral sites have been attributed to balancing selection and a hitchhiking effect, respectively. We, and others, have previously shown that nucleotide diversity between MHC haplotypes at non-HLA sites is unusually high (>10%) and up to several times greater than elsewhere in the genome (0.08%-0.2%). We report here the most extensive analysis of nucleotide diversity within a continuous sequence in the genome. We constructed a single nucleotide polymorphism (SNP) profile that reveals a pattern of extreme but interrupted levels of nucleotide diversity by comparing a continuous sequence within haplotypes in three genomic subregions of the MHC. A comparison of several haplotypes within one of the genomic subregions containing the HLA-B and -C loci suggests that positive selection is operating over the whole subgenomic region, including HLA and non-HLA genes.
[The sequence data for the multiple haplotype comparisons within the class I region have been submitted to DDBJ/EMBL/GenBank under accession nos. AF029061, AF029062, and AB031005-AB031010. Additional sequence data have been submitted to the DDBJ data library under accession nos. AB031005-AB03101 and AF029061-AF029062.]
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS AND DISCUSSION
METHODS
REFERENCES
|
|---|
Nucleotide diversity within the human genome has
been estimated to be between 0.08% and 0.2% (Li and Saddler 1991
;
Rowen et al. 1996; Horton et al. 1998; Lai et al. 1998; Satta et al.
1998). However, average pairwise comparisons between the HLA class I genes in the major histocompatibility complex (MHC) on chromosome 6 are
much higher (up to 8.6%) (Satta et al. 1998), and genomic differences
remote from the HLA class I genes may be >10% when two haplotypes
are compared (Guillaudeux et al. 1998; Horton et al. 1998; Gaudieri et
al. 1999). The elevated level of nucleotide diversity within the
antigen-presenting HLA class I and II genes has been attributed to
balancing selection acting on the antigen recognition sites (Hughes and
Nei 1988, 1989), with differences outside of the HLA coding region
associated with a hitchhiking effect (Grimsley et al. 1998, Guillaudeux
et al. 1998; Horton et al. 1998). In Drosophila, it has been
shown that the hitchhiking effect of balancing selection on neutral
sites is affected by mutation and recombination rates (Kreitman and
Hudson 1991; Aquadro 1992).
We have analyzed genomic subregions within the MHC described as
polymorphic frozen blocks (PFB) (Marshall et al. 1993; Dawkins et al.
1999). These PFBs can be up to several hundred kilobases in length, and
in cis combinations are observed in a population as MHC
haplotypes (Degli-Esposti et al. 1992). PFBs contain polymorphic genes
and have been shown to possess extensive genomic nucleotide diversity
that suppresses recombination within the blocks but not between the
blocks (Dawkins et al. 1999).
In this study, we constructed a single nucleotide polymorphism (SNP)
profile of a continuous sequence from three separate genomic subregions
of the MHC, including the region containing HLA-B and -C termed the
block and the region spanning HLA-A, -G, and -F termed the 
block. In this paper, SNP will refer only to nucleotide substitutions
and not to indels. Given the very low meiotic recombination rate
(Dawkins et al. 1999) within the blocks and the balancing selection
occurring at the HLA class I loci (HLA-A, -B, and -C), the SNP profile
is expected to show peaks at these loci with decreasing levels of
nucleotide diversity at distant neutral sites (Kreitman and Hudson
1991; Aquadro 1992; Satta et al. 1998). However, our results clearly
show the SNP profiles are extreme and interrupted with numerous peaks
and troughs within the MHC, suggesting that selection is occurring at
HLA and non-HLA class I loci.
| |
RESULTS AND DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS AND DISCUSSION
METHODS
REFERENCES
|
|---|
Extreme and Interrupted Nucleotide Diversity Profile Within the MHC
Our own continuous sequence within the MHC has been enhanced by
three sequencing groups (Mizuki et al. 1997; Guillaudeux et al. 1998;
Shiina et al. 1998; including sequence submissions by A. Hampe from
Centre National de la Recherche Scientifique, Rennes, France), allowing
an extension of earlier analyses of the nucleotide diversity between
two haplotypes at sites distant from the HLA class I loci (Fig. 1)
(Abraham et al. 1993). The SNP profiles within the
MHC are much more extensive and complex than those within another
region on chromosome 6 (6p23) that contain the polymorphic SCA1 gene
(Horton et al. 1998) and other regions of the genome (Fig. 1; Table
1). The SNP profiles we obtained within the genomic
subregions of the MHC are extreme and interrupted with several peaks
(Fig.1). With the addition of retroelement indels (such as
Alus) and other smaller indels, the level of nucleotide diversity within the MHC is even greater (Table 1).
|
|
Multiple Haplotype Comparisons Reveal a Similar Nucleotide Diversity Profile Within the MHC
The variation in nucleotide diversity within the class II region
appears to be related to the different haplotype comparisons (Fig. 1).
In contrast, each haplotype comparison in the class I region contains
regions of low nucleotide diversity (<1%) and peaks (>10%)
(Table 1). The SNP profiles in Figure 1 only compare two haplotypes at
any one site within the MHC. We predict that when multiple haplotypes
are compared the shape of the SNP profile will be similar, but the
level of nucleotide diversity between any two MHC haplotypes will
reflect the age of their last common ancestor. To determine whether the
level of nucleotide diversity in Figure 1 is consistent between
haplotypes, we compared five regions of low, medium, and high
nucleotide diversity within the block of different MHC haplotypes
(Table 2). The only exception was the comparison of
44.1 and 57.1 haplotypes in region ii (Table 2). As expected, the
comparison between the recently diverged 7.1 and 8.1 haplotypes shows a
low mean nucleotide diversity (Table 2). Overall, these results
indicate that the level of nucleotide diversity between different
haplotype comparisons will reflect the SNP profile observed in Figure
1.
|
To test for nucleotide diversity heterogeneity within the five regions
described in Table 2, we used the goodness-of-fit statistic described
by Kreitman and Hudson (1991). There was heterogeneity within the five
regions at the P = 0.001 level of significance.
Evolutionary History of the MHC Plays a Role in Shaping the Nucleotide Diversity Profile
To investigate the factors influencing the shape of the SNP
profiles, we examined the duplications and indels characteristic of the
MHC (Gaudieri et al. 1997a,b; Kulski et al. 1999b). In the block,
HLA-B and -C, MICB (PERB11.2), and MICA (PERB11.1) genes are contained
within two sets of duplicated segments that each share approximately 30 kb of sequence (Fig. 1) (Gaudieri et al. 1997a). The segments contain
all the major peaks within this region except for the TA-rich expansion
within the LTR region of human endogenous retroviral (HERV)-L (Fig. 1)
(Kulski et al. 1999a). Each duplicated segment contains at least one
major peak in nucleotide diversity (Fig. 1A), with the level of
nucleotide difference between them probably caused by the earlier
duplication of the HLA-B and -C segments (Gaudieri et al. 1997a; Kulski
et al. 1999b). Some of the troughs within and between the duplicated segments can be explained by recent insertion events. For example, the
HERV-K9I sequence telomeric of HLA-C inserted into the HLA-C duplication segment shows a low level of nucleotide diversity (Fig. 1).
This HERV has still retained large open reading frames (Kulski et al.
1999a), suggesting it is a recent insertion event. Furthermore, a 10-kb
region between the HLA-B and -C duplication segments is duplicated in a
telomeric region between HLA-30 and MICC (PERB11.3), which may also be
the result of a recent translocation because it shows a low level of
nucleotide diversity (Fig. 1). Thus, several troughs within the SNP
profile of the block can be accounted for by recent insertions
and translocations. However, even after excluding all indels from the
duplication segments within the block, the SNP profile remains
extreme and interrupted with peaks at non-HLA class I loci.
Within the block, the SNP profile shows three broad but distinct
peaks in the level of nucleotide diversity (Fig. 1). This block is
subject to flawed multisegmental duplications that have been separated
into three tripartite segmental regions: I, II, and III (Kulski et al.
1999b). Kulski et al. (1999b) show that the segments (duplicons)
containing HLA-A, -G, and -F duplicated during different times, with
the segment containing HLA-F diverging first, then HLA-G and -A,
respectively. The greater nucleotide diversity around HLA-A compared
with HLA-G and -F is opposite to that expected from the evolutionary
history of the segmental regions (Fig. 1) (Kulski et al. 1999b). This
suggests that other forces besides neutral accumulation of nucleotide
differences are occurring within this region.
Low Nucleotide Diversity Coincides with the Predicted End Points of
the Block
Two regions within the block centromeric of MICB (PERB11.2)
and telomeric of HLA-C show very low levels of nucleotide diversity (0% to ~2%) (Fig. 1). These two regions are rich in Alu
sequences (Fig. 1C). The Alus within these regions belong to
different subtypes, ranging from Alu J sequences that have
been inserted in early primates to more recent Alu Y inserts
in apes (Kapitonov and Jurka 1996). Alu sequences have been
associated with microsatellites and polymorphism (Epstein et al. 1990),
with a likely positive correlation with time of insertion. In addition,
the Alu-rich regions are also rich in hypermutatable CpG
dinucleotides (Fig. 1B) (Holliday and Grigg 1993). Thus, the low level
of nucleotide diversity observed within the Alu-rich regions
suggests that there is a suppression of nucleotide diversity. These
regions of low nucleotide diversity coincide with the predicted end
points of the block (Marshall et al. 1993; Dawkins et al. 1999).
In addition, two regions of low nucleotide diversity (0%-2%) within
the block centromeric of MICB (PERB11.2) and telomeric of HLA-C
coincide with the proposed centromeric and telomeric boundaries of the PFB (Marshall et al. 1993; Dawkins et al. 1999).
A decrease in nucleotide diversity is expected at the ends of the PFBs where recombination may occur, and this is reflected in the SNP pattern observed in Figure 1. Similarly, hitchhiking from balancing selection acting on the HLA loci would result in a decrease in nucleotide diversity flanking the loci when the recombination rate increases. Thus, the hitchhiking effect from the HLA class I genes is expected to contribute to only a single peak at the loci, which is clearly not the case in the HLA class I duplicated region of the MHC (Fig. 1).
Selection Pressure on Non-HLA class I Sequences in the MHC
Figure 1 shows that peaks in nucleotide diversity correspond to HLA
and non-HLA class I genes and certain retroelements. Two peaks in
nucleotide diversity at non-HLA class I regions are greater than the
HLA-B and -C peaks in the block. The two peaks correspond to the
HERV-I sequence and its flanking L1 sequences and to a CpG and
G+C-rich region telomeric of HERV-I containing a mixture of
Alu and L1 sequences with a large open reading frame
corresponding to the reverse transcriptase domain in the L1 sequence
(Fig. 1). Within the SNP profile of the block, the highest peak
in nucleotide diversity occurs centromeric of HLA-A in a region
containing a copy of HERV-16 (Fig. 1). Other non-HLA class I peaks in
the SNP profile within the and blocks include regions
telomeric of the transcribed genes MICB (PERB11.2) and MICA (PERB11.1).
As discussed above, these peaks are within the more recently duplicated MIC (PERB11) segments. Therefore, the SNP profiles within the MHC do
reflect the expected profile of selection occurring not only at the
antigen presenting HLA class I genes (Hughes and Nei 1988; Satta et al.
1998), but also at other loci, such as MIC (PERB11) genes, some HERV
and L1 sequences, and, potentially, the whole genomic subregion.
Other Non-HLA Genes Within the MHC that Are Transcribed and Polymorphic
Non-HLA class I polymorphic sequences that are transcribed in the
block include polymorphic MIC (PERB11) genes (Gaudieri et al.
1997c) and HERVs. The MIC (PERB11) genes have been shown to be
involved in the activation of NK and T cells (Bauer et al. 1999) and
are associated with susceptibility to several diseases (Dawkins et
al. 1999). However, the type of selection acting on the MIC (PERB11)
genes is so far unknown. The level of nucleotide diversity
within HERV-I and flanking L1 sequences is higher or at least
equivalent to that observed at HLA-B and -C (Fig. 1A) (Guillaudeux et
al. 1998; Gaudieri et al. 1999). Thus, although the role of HERV-I and
L1 sequences within the block is unknown, it seems likely they
are under selection. The duplicated HERV-16 sequences within the block differ in their level of nucleotide diversity (Fig. 1). One of
the copies of HERV-16, named P5-1, is transcribed in lymphoid cells
and tissues in an antisense direction to its internal RTase
sequence, and it has been suggested that this transcript may have an
antiviral role (Kulski and Dawkins 1999)
In addition, we could not find an overall correlation between CpG
frequency and the level of nucleotide diversity in the MHC genomic
subregions we had examined (Fig. 1B). The correlation between CpG
frequency and nucleotide diversity is expected when mutation pressure
is stronger than selection, given the hypermutatable change from
methylated cytosine in CpG to TpG (Holliday and Grigg 1993). Moreover,
it has recently been shown that the level of variation in synonymous
substitutions within genes correlates to the frequency of CpG
dinucleotide sequences (K. Tsunoyama, pers. comm.). This result is
consistent with our proposal that selection occurs over the whole
genomic subregion and not only at the HLA class I loci under balancing selection.
We constructed SNP profiles within genomic subregions of the MHC under the expectation that balancing selection was occurring at the antigen-presenting HLA class I loci (HLA-A, -B, and -C). However, our results clearly show that the SNP profiles within the genomic subregions are extreme and interrupted with several peaks and troughs. Although duplications and indels have contributed to the SNP profiles constructed within the MHC, we propose that selection has also acted to shape the SNP profiles not only at HLA class I genes but at other sites. The SNP profiles suggest that selection may be occurring at sites outside of the HLA class I genes and over the whole genomic subregion because there are peaks within the profile at non-HLA class I loci and highly polymorphic non-HLA class I genes are transcribed within the region.
Our hypothesis of selection occurring at multiple sites within the genomic subregions assumes a constant mutation rate. We cannot eliminate the possibility that there is variation in the mutation rate; however, one indicator of mutation rate, CpG%, does not correlate with nucleotide diversity.
We conclude that hitchhiking and other factors influence the nucleotide diversity profile within the MHC and that selection operates on non-HLA class I sequences and potentially over the entire genomic subregion. The nucleotide diversity seen in Figure 1, and usually attributed to hitchhiking and balancing selection at the HLA genes, is probably further confounded by the segmental duplications and retroelement indel events occurring at different times in primate history.
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS AND DISCUSSION
METHODS
REFERENCES
|
|---|
Sequences
The sequences used in the SCA1 and class II region have been
previously described (Horton et al. 1998). The SNP profile spanning IkBL to telomeric of HLA-C in the block is broken into three different haplotype comparisons. From IkBL to MICA (PERB11.1), cosmids
from the Mann cell line (HLA-A29; -B44; -Cw4; -DR7) (AC004181, AC006046, AC004183, AC004184, AC004215, AC004214) (Guillaudeux et al.
1998) were compared with the Boleth cell line (HLA-A2; -B62; -Cw10;
-DR4) (AB000882) (Shiina et al. 1998). From MICA (PERB11.1) to HERV-I,
the Mann cell line (AC004180 and AC004182) was compared with the
heterozygous CGMI cell line (in this comparison, HLA-A3; -B8; -Cw;
-DR3) (D84394) (Mizuki et al. 1997). The region from HERV-I to
telomeric of HLA-C was compared with that between the two haplotypes in
CGMI (HLA-A3,29; -B8,14; -Cw-; -DR3,7) (AC004205, AC004204, AC006048, AC004185, and AC006047 were compared with D84394) (Guillaudeux et al.
1998; Shiina et al. 1998).
To determine the level of sequence error within the block, we
compared a sequence from the same haplotype from two different sequencing groups. In this case, cosmid Y5C028 (AC004210) was compared
with D84394, with a resultant substitution and indel error rate of less
than 0.05%. To determine the degree of nucleotide diversity within the
block, cosmids from the CGMI cell line (AC004178, AC004199,
AC005404, AC004200, AC004203, AC004194, AC004193, AC004172, AC004192,
AC004173, AC004170, and AC004213) (Guillaudeux et al. 1998) were
compared with the DDBJ/EMBL/GenBank accession numbers U51588 and
AF055066 (submitted by A. Hampe from Centre National de la Recherche
Scientifique, Rennes, France). The probing, mapping, and sequencing of
the clones for the 57.1, 8.1, 7.1, and 18.2 haplotypes within the
regions i-v in Figure 1 have been previously described (Leelayuwat et al. 1992; Gaudieri et al. 1997b). The following DDBJ/EMBL/GenBank accession numbers for the regions i-v were used: AF029062 (8.1) and
AF029061 (57.1) for region i (Gaudieri et al. 1997b); AB031005 (57.1)
and AB031008 (18.2) for region ii (Leelayuwat et al. 1992; Gaudieri et
al. 1997b); AB031007 (7.1) for region iii (Gaudieri et al. 1997b);
AB031010 (57.1) for region iv (Gaudieri et al. 1997b); and AB031006
(57.1) and AB031009 (7.1) for region v (Leelayuwat et al. 1992;
Gaudieri et al. 1997b). For the calculation of nucleotide diversity in
Table 2, only sequences with twofold coverage or greater were used.
Sequence analysis
All sequence alignments were produced using the program ClustalW (http://www.ddbj.nig.ac.jp/E-mail/clustalw-e.html), and the resultant outputs were used in the program CLTOSS (http://193.50.234.246/~beaudoin/anrs/cgi-bin/Pre_align_process2.cgi). CLTOSS removed all gaps from the alignments to normalize the number of nucleotides examined in each window. The nucleotide diversity comparisons, G+C%, and CpG changes were calculated using an in-house program called Window6.pl. RepeatMasker2 (http://ftp.genome.washington.edu/cgi-bin/RepeatMasker) was used to identify retroelement sequences, and its output was illustrated using an in-house program called DrawRep.pl.
The correlation between CpG% and nucleotide diversity was calculated using Pearson's correlation coefficient (Microsoft Excel version 5.0) after the removal of the CpG islands of reported genes and the TA-rich region in HERV-L.
To test whether nucleotide diversity levels were statistically
different in regions of the block profile, we used the method described by Kreitman and Hudson (1991; Hartl and Clark 1997). To test
for heterogeneity, a goodness-of-fit statistic was used as described by
Kreitman and Hudson (1991):
|
| |
ACKNOWLEDGMENTS |
|---|
We acknowledge the efforts of Dr. Chanvit Leelayuwat, David Sayer, Dr. Maria Pia Degli-Esposti, and Linda Smith in the preparation and sequencing of the 57.1, 18.2, 8.1, and 7.1 haplotype sequences described in this study. We thank Dr. Katsuho Ikeo and Professor Joergen Epplen for helpful suggestions with the manuscript. We also thank two anonymous reviewers for their helpful suggestions and comments. S.G. is supported by a Japanese Society for the Promotion of Science (JSPS) fellowship. T.G. is supported by the Ministry of Education, Science, Sports and Culture of Japan. J.K.K. and R.L.D. are grateful for support from the National Health and Medical Research Council, Australia.
The publication costs of this article were defrayed in part by payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 USC section 1734 solely to indicate this fact.
| |
FOOTNOTES |
|---|
3 Corresponding author.
E-MAIL tgojobor{at}genes.nig.ac.jp; FAX 81-559-81-6848.
Article and publication are at www.genome.org/cgi/doi/10.1101/gr.127200.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
RESULTS AND DISCUSSION
METHODS
REFERENCES
|
|---|
T cell receptor locus.
Science
272:
1755-1762[Abstract].Received November 29, 1999; accepted in revised form July 20, 2000.
CiteULike 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. Takuno, R. Fujimoto, T. Sugimura, K. Sato, S. Okamoto, S.-L. Zhang, and T. Nishio Effects of Recombination on Hitchhiking Diversity in the Brassica Self-incompatibility Locus Complex Genetics, October 1, 2007; 177(2): 949 - 958. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Shiina, M. Ota, S. Shimizu, Y. Katsuyama, N. Hashimoto, M. Takasu, T. Anzai, J. K. Kulski, E. Kikkawa, T. Naruse, et al. Rapid Evolution of Major Histocompatibility Complex Class I Genes in Primates Generates New Disease Alleles in Humans via Hitchhiking Diversity Genetics, July 1, 2006; 173(3): 1555 - 1570. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. K. Raymond, A. Kas, M. Paddock, R. Qiu, Y. Zhou, S. Subramanian, J. Chang, A. Palmieri, E. Haugen, R. Kaul, et al. Ancient haplotypes of the HLA Class II region Genome Res., September 1, 2005; 15(9): 1250 - 1257. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. K Kulski, T. Anzai, T. Shiina, and H. Inoko Rhesus Macaque Class I Duplicon Structures, Organization, and Evolution Within the Alpha Block of the Major Histocompatibility Complex Mol. Biol. Evol., November 1, 2004; 21(11): 2079 - 2091. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Anzai, T. Shiina, N. Kimura, K. Yanagiya, S. Kohara, A. Shigenari, T. Yamagata, J. K. Kulski, T. K. Naruse, Y. Fujimori, et al. Comparative sequencing of human and chimpanzee MHC class I regions unveils insertions/deletions as the major path to genomic divergence PNAS, June 24, 2003; 100(13): 7708 - 7713. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Innan, B. Padhukasahasram, and M. Nordborg The Pattern of Polymorphism on Human Chromosome 21 Genome Res., June 1, 2003; 13(6): 1158 - 1168. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Toomajian and M. Kreitman Sequence Variation and Haplotype Structure at the Human HFE Locus Genetics, August 1, 2002; 161(4): 1609 - 1623. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Navarro and N. H. Barton The Effects of Multilocus Balancing Selection on Neutral Variability Genetics, June 1, 2002; 161(2): 849 - 863. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
