| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Departments of Structural Biology and Microbiology and Immunology, Stanford University School of Medicine, Stanford, CA 94305;
Department of Hematology and Bone Marrow Transplantation, City of Hope National Medical Center, Duarte, CA 91010; and
Department of Pediatrics, Cedars-Sinai Medical Center, Los Angeles, CA 90048
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During their development, NK cells are programmed to express different numbers and combinations of KIR in a stable manner. Thus, human peripheral blood NK cells have clonal diversity in receptors due to this mode of KIR gene expression. Contributing to this diversity, but to lesser extent, is expression of the HLA-E-specific receptors CD94:NKG2A and CD94:NKG2C (7). In a similar manner, KIR expression in T cells can produce diversity within populations of cells expressing the same TCR (8, 9). In NK cell responses, KIR are proposed to be receptors that contribute to recognition and response to cells damaged by infection or malignant transformation (1). In T cell responses, the expression of KIR after activation by Ag may modify their response on subsequent encounter with Ag (10).
Diversity is an important characteristic of the human KIR gene family and its counterpart in apes and Old World monkeys (11, 12, 13, 14). Sequence analysis of KIR cDNA populations obtained from individual donors has demonstrated that several of the KIR genes are polymorphic, and in this regard KIR3DL1 and KIR3DL2 have been particularly illustrative (15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26). Humans also differ in the KIR genes they inherit, as shown by typing genomic DNA for the presence and absence of each KIR gene. In combination with RFLP analysis, these data suggested two types of KIR haplotype, segregating at roughly equal frequency in the largely Caucasoid population studied. Group A haplotypes contain seven KIR genes and have KIR2DS4 as the only activating receptor. Group B haplotypes are more diverse in the KIR genes they contain, have more activating receptors, and are characterized by the 2DL2, 2DS1, 2DS2, 2DS3, and 2DS5 genes (11). These results have subsequently been confirmed in studies of other populations (27, 28, 29, 30).
Comparison of gene organization for two A haplotypes and one B haplotype showed that human KIR haplotypes are organized around three conserved framework genes: KIR3DL2 and -3DL3 at either end; and KIR2DL4 in the middle. Between the framework genes are two regions of variable gene content. In these regions, gene duplications, deletions, and hybridization by asymmetrical recombination are likely facilitated by the close proximity of the genes and the sequence similarity of the intergenic sequences (4, 5).
Although it is now clear that KIR haplotypes vary in gene content and that KIR genes are polymorphic, what has yet to be understood is how these two mechanisms work together in producing KIR diversity in human populations. In other words, how does allelic polymorphism further diversify KIR haplotypes defined by gene content, and to what extent? The investigation described here addresses these questions by combining analysis of families with KIR typing directed at the level of both genes and alleles. The results show that allelic polymorphism further differentiates KIR haplotypes having the same gene content. Together these two mechanisms of diversification individualize the human KIR genotype.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
KIR2DL1, -2DL3, -3DL1, and -3DL2 alleles
were named according to guidelines used in naming HLA alleles (Table I
). Briefly, an asterisk separates the
accepted gene designations (31) from three digits that
distinguish alleles that differ by nonsynonymous substitutions; fourth
and fifth digits were assigned to alleles that differ only by
synonymous substitutions. Numerical order was assigned based on GenBank
release dates. Partial sequences and splice variants were excluded, as
were sequences of single PCR-derived clones.
|
Donor population and DNA isolation
Samples were obtained from a total of 143 donors (predominantly Caucasian) from 34 families. PBMC from 64 donors representing 15 families were prepared at the Cedars-Sinai Medical Center (Los Angeles, CA); PBMC from 24 individuals representing 4 families were prepared at the City of Hope Histocompatibility Laboratory (Duarte, CA). Peripheral blood from 50 individuals representing 14 families was collected at the Stanford Medical Center Histocompatibility Laboratory (Stanford, CA), PBMC being isolated by Ficoll-Hypaque gradient separation. Genomic DNA was prepared from 2 x 106–1 x 107 PBMC using a QIAamp Blood Kit (Qiagen, Chatsworth, CA). One family described by Gardiner et al. (23) was analyzed for KIR at higher resolution in this study.
KIR genotyping
Generic KIR typing of genomic DNA was performed by
PCR amplification with primers based on conserved regions specific to
each KIR locus, as described by Uhrberg et al.
(11), with modification. These primers detect five genes
encoding inhibitory KIR (2DL1-3 and
3DL1-2) and five genes encoding stimulatory KIR
(2DS1–4 and 3DS1). Detection of
KIR2DS5 was as modified by Vilches et al. (26).
KIR2DL2v1 (17) was assayed using the forward
primer (5'-CAG CAC TTC CTT CTG CAC AC-3') and the 2DL2
locus-specific reverse primer (5'-GCC CTG CAG AGA ACC TAC A-3')
(11). KIR2DL4 was amplified as described by
Norman et al. (29), with substituted internal control
primers (listed below). KIR2DL5 was assayed with typing
primers described by Vilches et al. (33), using the same
PCR conditions as for other generic KIR typing reactions. As
an internal control, primers specific for a 650-bp fragment of the
-globin gene (5'-GCT GTC ATC ACT TAG ACC TCA CC-3' and 5'-GAA AGA
AAA CAT CAA GCG TCC C-3'), were included in the typing reactions for
2DL2v1, 2DL4, and 2DL5.
For KIR subtyping, primers designed to discriminate
allele-specific polymorphisms were paired with KIR2DL1, -2DL3,
-3DL1, or -3DL2 locus-specific primers.
KIR3DL1 and -3DL2 subtyping were performed as
described by Gardiner et al. (23), with the addition of
one primer pair (3DL1-607T(F): 5'-GGT TCT GTT ACT CAC ACC T-3' and
3DL1-882T(R): 5'-AGA GTG ACG GAA GCC A-3') which amplifies
3DL1*004 and 3DL1*005 alleles. Fig. 1 shows the alleles amplified by
2DL1- and 2DL3-specific primer combinations. PCR
amplification conditions are listed as H (high), N (normal), or L (low)
reflecting the annealing temperature used. Under condition N, initial
denaturation was 95°C for 5 min; five cycles of 97°C for 20 s,
62°C for 45 s, and 72°C for 90s were followed by 26–30 cycles
of 95°C for 20 s, 60°C for 45 s, and 72°C for 90
s; finally, a 7-min extension at 72°C was performed. For condition H,
annealing temperatures were 68°C during the initial five cycles and
64°C for the remaining cycles; for condition L, the annealing
temperatures were 60°C and 58°C. All other parameters of the
amplification remained constant. Amplification of genomic DNA was
performed in 25-µl reactions using 100–300 ng DNA, 0.625 U AmpliTaq
polymerase, 2.5 µl 10x buffer (PerkinElmer, Norwalk, CT), and 0.2 mM
dNTPs (Promega, Madison, WI). KIR primers were used at 0.5
mM. Internal control forward (5'-CTT CGA GCA AGA GAT GGC CAC-3') and
reverse (5'-TTG CTG ATC CAC ATC TGC TGG AAG-3') primers, specific for a
350-bp fragment of -actin and -actin pseudogenes, were included
in each subtyping PCR at a concentration of 0.033 mM.
|
Total cellular RNA was prepared from 2 x
106–1 x 107 PBMC
using RNAzol (Tel-Test, Friendswood, TX). First-strand cDNA was
synthesized from 
5.0 µg RNA using oligo(dT) (Applied Biosystems,
Foster City, CA) and avian myeloblastosis virus reverse transcriptase
(Promega) at 42°C for 90 min.
Sequencing and subtyping of novel KIR alleles
When KIR subtyping results could not be explained in terms of known alleles, full length cDNA sequences were determined. KIR3D sequences were PCR amplified from PMBC-derived cDNA using oligonucleotide primers based on conserved KIR sequences as described by Valiante et al. (7). The resulting products were subcloned using a TOPO-TA Cloning Kit, according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA) and then partially sequenced using a standard T7 or T3 primer and dye terminator automated sequencing (Applied Biosystems). Based on partial sequences, three to four representatives of each allele were selected and sequenced completely on both strands to obtain a consensus sequence. Alternatively, 7–10 clones representing each new allele were selected and pooled; the pool template was sequenced completely on both strands to obtain a consensus sequence.
In KIR3DL2 subtyping, 3DL2*010 and 3DL2*011 amplify with the 3DL2 subtyping primer pairs: 474G(F)/756C(R); 322G(F)/393G(R); 497G(F)/756C(R); 563T(F)/755T(R); 1019A(F)/1190T(R); and 122G(F)/394C(R). 3DL2*010 is additionally amplified by 377C(F)/393G(R), whereas 3DL2*011 is recognized by 377G(F)/393G(R). 3DL2*012 is positive with primer pairs 474G(F)/756C(R), 322G(F)/393G(R), 377C(F)/393G(R), 497A(F)/756C(R), 563T(F)/755T(R), and 1019A(F)/1190C(R) (23). These new reactivities result in ambiguous subtyping patterns for several combinations of 3DL2 alleles: 3DL2*010 plus *006 is indistinguishable from *001/9 and *007; as are 3DL2*010 and *003 from *001/9 and *011; *010 and *005 from *011 and *012; *011 and *006 from *003 and *007; and *012 plus *003 from *001/9 and *005. However, we were able to resolve the potential ambiguities in the data obtained here through segregation analysis.
KIR3DL1/2v is recognized by 3DL1 primer pairs 202C(F)/607C(R) and 607T(F)/882T(R), and 3LD2 primers 122G(F)/394C(R) and 1019A(F)/1190C(R) (23). Donors typing positive in these reactions were further tested with the 3DL1 locus-specific forward primer (5'-CCA TCG GTC CCA TGA TGC T-3') (11) and reverse primer KIR-921A(R) (5'-AAC AAG CAG TGG GTC ACT T-3'), specific for a 1.9-kb band of 3DL1/2v. No additional 3DL1/2v-positive donors were found outside of the family in which it was first identified.
Determination of KIR haplotypes
KIR haplotypes were determined by segregation analysis in families. Assumptions made to facilitate assignments of certain KIR alleles and loci included: 1) all haplotypes were assumed to include KIR2DL4 and 3DL2; 2) individuals having only one allele for 2DL1 and 2DL3 were assumed to be homozygous for both 2DL1 and 2DL3, unless segregation analysis discriminated between homo- and hemizygosity at these loci; 3) haplotypes were assumed to include either 2DL1/2DL3 or 2DL2, but not both; 4) KIR3DS1 was assumed to segregate as an allele of 3DL1; 5) KIR2DS4 was assumed to be present on haplotypes having no other KIR2DS genes. These assumptions are supported by the KIR haplotypes studied at high resolution (4, 5) and KIR linkage disequilibria (LDs) reported to date (11, 27, 28, 29, 30).
Statistical analysis
LD between pairs of KIR alleles and/or loci was
calculated from parental haplotypes using the computer package Arlequin
version 2.0 (34). For this purpose, KIR2DL3*006
(haplotype 22; Fig. 3A) was considered equivalent to
2DL3*002/6, and 3DL1/2v (haplotype 37; Fig. 3B) was scored as an allele of both 3DL1 and
3DL2. To calculate LD between subtyped KIR
alleles and other, nonsubtyped KIR loci, the latter were
scored as present or absent; ambiguous KIR2DS and
2DL5 loci were considered as being present. The significance
of LD values was assessed by 
2 analysis.
|
HLA-A and HLA-B Ags were determined serologically by the laboratories supplying the samples (35). HLA-C type was determined serologically or by PCR-sequence-specific priming (SSP) analysis of genomic DNA using C locus SSP Unitray test kits (Pel-Freez, Brown Deer, WI).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous studies of KIR diversity have focused on either low resolution genotyping that determines the presence or absence of KIR genes (11, 27, 28, 29, 30) or cDNA sequencing to assess their allelic polymorphism (15, 16, 17, 20, 21, 22, 23, 24, 25, 26). The goal of this study was to determine how these two levels of genetic diversity are related and, more specifically, to examine the extent of allelic polymorphism underlying KIR haplotypes sharing similar or identical gene content. To this end, we assembled a panel of genomic DNAs from 143 donors (predominantly Caucasoid) belonging to 34 families; families were selected for having three or more members which could include siblings, parents, and grandparents. Initial low resolution PCR-SSP typing was performed, as in prior studies, to determine the presence or absence of 13 KIR genes (2DL1–5, 2DS1–5, 3DL1–2, and 3DS1). This yielded 22 distinct KIR genotypes, of which 19 have been previously reported (data not shown).
High resolution typing was targeted to the four KIR loci for
which most variants had been reported at the time this study began:
KIR2DL1; KIR2DL3; KIR3DL1; and
KIR3DL2. For each of these genes, locus-specific primers
were paired with allele-specific primers designed to survey polymorphic
sites spanning the coding sequence. Given that individual
KIR alleles are identified by the combination of
polymorphisms present, this scheme had the potential for discovering
KIR alleles with novel configurations of known polymorphic
sites. Indeed, four new KIR alleles were discovered.
KIR3DL2*010, -3DL2*011, and -3DL2*012 differ from
other 3DL2 sequences in that they have different
combinations of previously described 3DL2 polymorphisms
(Fig. 2A). The fourth allele,
provisionally called KIR3DL1/2v, is a chimeric sequence
encoding a molecule that combines 3DL1 extracellular domains with a
3DL2 cytoplasmic tail. This arrangement is similar to that of the
common chimpanzee KIR, Pt-KIR3DL1/2 (12). The sequence of
3DL1/2v can be derived from 3DL1*005 and
3DL2*003 (Fig. 2B), however, and shares an
average of only 96.6% nucleotide sequence identity with the three
alleles of Pt-KIR3DL1/2. Taken together, these data indicate
that 3DL1/2v and Pt-KIR3DL1/2 probably arose
independently, albeit through analogous recombination events. The
haplotype containing 3DL1/2v lacks conventional
3DL1 or 3DL2 genes, indicating that
3DL1/2v was generated by unequal crossing over between these
two genes, which gave rise to a haplotype in which the intervening
region is absent.
|
C). Given that 3DL2*010
differs from 3DL2*001 and 3DL2*007 by single
nucleotides, it could have arisen from recombination or gene conversion
between these two alleles or point mutation of either one. Similarly,
3DL2*011 could have originated by recombination or gene
conversion between 3DL2*003 and 3DL2*007 or by
point mutation of 3DL2*003 (Fig. 2C). Patterns of LD in 22 A haplotypes
Within each family, the segregation of KIR alleles was
determined and used to define KIR haplotypes. Of the 34
families, 26 were informative in this regard, and 98 of the 110
parental haplotypes within the informative families were defined. The
most common combination of KIR genes is also the simplest
and corresponds to that of the A haplotype group. These
haplotypes consist of KIR2DL1, -2DL3, -2DL4, -3DL1, -3DL2,
-3DL3, and -2DS4, the last being the only
KIR2DS gene. Of the 98 parental haplotypes, 48 (49%) have
this combination of KIR genes. Within this group, 22
KIR haplotypes were distinguished by their combination of
2DL1, 2DL3, 3DL1, and 3DL2 alleles (Fig. 3A).
Comparison of the A haplotypes revealed striking patterns of
LD, most notably between alleles of KIR2DL1 and
KIR2DL3 (Table II). The
2DL1*002 and 2DL3*002/6 alleles were partnered
exclusively with each other, as were 2DL1*003 and
2DL3*001. These 2DL1/2DL3 pairs appeared on
haplotypes in 12 and 16 families, respectively. Certain pairs of
3DL1 and 3DL2 alleles are also in LD, including
3DL1*001 and 3DL2*001/9, 3DL1*005 and
3DL2*001/9, 3DL1*004 and 3DL2*003, and
3DL1* 002/3/6/7/8 and 3DL2*002. However, there is
more evidence of recombination between 3DL1 and
3DL2 than between 2DL1 and 2DL3;
KIR3DL1*004, for example, was associated with
3DL2*005, -*011, and -*012, as well as
with -*003, whereas 3DL2*001/9
appeared with 3DL1*002/3/6/7/8,
-*001, and -*005. KIR3DL1*002/3/6/7/8, a
collection of similar alleles, was associated with 3DL2*001/9,
-*002, -*006, -*008, and -*010. The strong LD between
2DL1 and 2DL3 does not extend to 3DL1,
the predominant 2DL1/2DL3 pairs being found in linkage with
all four of the 3DL1 alleles distinguished by our
subtyping.
|
Haplotypes having more complicated combinations of KIR
genes than the A haplotypes, and which are characterized by
KIR2DS genes not present in A haplotypes, are
called collectively B haplotypes (11). These
differ primarily in the number and type of KIR2DS genes
present but can also vary by the presence or absence of 2DL1,
2DL2, 2DL3, and 3DL1. From the family analysis, 24
B haplotypes were unambiguously defined. They represent 15
distinct KIR haplotypes, involving 9 different gene
combinations (Fig. 3B). An additional 26 parental haplotypes
were partially defined, but ambiguities in assignment of
KIR2DS loci prevented their complete definition.
Different combinations of KIR alleles were seen to underlie
three of the B haplotype gene combinations (Fig. 3B). In the first group, three KIR2DL1/2DL3 pairs
are combined with 3DL2*007 and 3DS1; these
haplotypes also include 2DS1, 2DS5, and 2DL5
genes. In the second and third groups, it is the combinations of
3DL1 and 3DL2 alleles that differentiate the
haplotypes. Four distinct 3DL1/3DL2 pairs were associated
with 2DL2 and 2DS2; these haplotypes, unlike most
of the B haplotypes, do not include 2DL5. Two additional
3DL1/3DL2 pairs are combined with 2DS2, 2DS3, 2DL5,
2DL2, and 2DL1*004.
The KIR2DL1/2DL3 allele combinations seen on these B haplotypes include those pairs that were prominent on the A haplotypes: KIR2DL1*002 and KIR2DL3*002/6; KIR2DL1*003 and KIR2DL3*001; and KIR2DL1*001 and KIR2DL3*004/5. 2DL1*004, a chimeric 2DL1/2DS1 sequence (18) not seen on any A haplotype, was present only on B haplotypes that lacked 2DL3, including two haplotypes with 2DL2. KIR2DL2 also appeared without 2DL1 or 2DL3 on several haplotypes involving various combinations of KIR2DS loci and 3DL1 and 3DL2 alleles.
Several KIR3DL1/3DL2 allele combinations were present in
both A and B haplotypes. Additionally,
3DS1, which segregates as a 3DL1 allele, was in
strong LD with 3DL2*007, an allele that is common on
B haplotypes but absent from A haplotypes. The
combination of 3DS1 and 3DL2*007 appeared on
haplotypes that differed in both gene and allele content, including
haplotypes with different combinations of KIR2DS loci (Fig. 3B).
The gene combinations seen in the B haplotypes listed on
Fig. 3B encompass all but two of the partial, low resolution
B haplotypes reported by Crum (27) et al. from
analysis of 13 Irish families. Of these, the partial haplotypes
2DL1-2DL3-3DS1-2DS5-2DS1,
and 3DS1-2DS5-2DS1 correspond to
haplotypes 23–25, as does the partial haplotype
3DS1-2DS5 (also reported by Witt et al.
(28)). Haplotypes 26–31, and 34 have the
3DL1-2DS2 combination, and the
3DL1-2DS2-2DS4 combination is part of
haplotype 28. The combination of
3DS1-2DS5-2DS1-2DS2 is seen
in haplotype 33, as are
3DS1-2DS5-2DS1, and
3DS1-2DS5. Finally, haplotypes 30–31 have the
combination 2DL2-2DS3-2DS2, although
in the Irish family from which this partial haplotype was defined,
these genes are probably linked to 3DS1 instead of
3DL1. All but one of these partial haplotypes form part of
KIR gene combinations for which we have identified more than
one distinct haplotype. Of the two partial haplotypes of Crum et al.
that do not match any of the B haplotypes defined here, the
2DL2-3DL1-2DS3-2DS1-2DS2
combination is suggested in several of the haplotypes that could not be
determined fully (data not shown). By contrast, there was no suggestion
of the 2DS3-2DS5-2DS1 combination in
our data. Conversely, four of the B haplotypes identified
here (haplotypes 32 and 35–37) did not include any of the gene
combinations reported as partial haplotypes.
Strong LD within but not between the two halves of the KIR gene complex
Calculations of LD were performed using all defined parental
haplotypes. For the four subtyped genes (KIR2DL1,
2DL2, 3DL1, 3DL2), allele pairs
in significant LD are shown in Table II with their significance levels.
The results quantitatively establish the strong LD between
KIR2DL1 and 2DL3 alleles first apparent from
inspection of A haplotypes (Fig. 3A). In
addition, KIR2DL1*004 and 2DL1blank are in
significant positive LD with 2DL3blank, as is evident on
inspection of B haplotypes (Fig. 3B). There is a
corresponding significant negative LD between 2DL1*002 and
2DL3*001, 2DL1*003 and 2DL3*002/6, and
2DL1*004 and 2DL3*001, as well as between
2DL1*002 or 2DL1*003 and 2DL3blank and
between 2DL3*001 or 2DL3*002/6 and
2DL1blank. LD calculations also demonstrate significant
positive LD between pairs of 3DL1 and 3DL2
alleles, including that between 3DL2*007 and
3DS1. In addition, there are several 3DL1/3DL2
pairs in negative LD, including 3DL1*002/3/6/7/8 and
3DL2*005 or 3DL2*007, 3DL1*004 and
3DL2*002 or 3DL2*007, and 3DS1 and
3DL2*001/9.
The stronger LD between KIR2DL1 and 2DL3 than
between 3DL1 and 3DL2 is consistent with the
physical separation of these loci. KIR2DL1 and
2DL3 are closer together, being separated only by the
pseudogene KIR2DP1 (also called KIRZ), whereas
the distance between the 3DL1/3DS1 and 3DL2 loci
varies with haplotype and can include several KIR2DS genes.
In contrast, the LD between either 2DL1 or 2DL3
and 3DL1 or 3DL2 is much weaker. The four pairs
of alleles in strongest positive LD were KIR2DL1*002 and
KIR2DL3*002/6 with 3DL1*004 and
3DL2*005; this combination of alleles is represented in
haplotype 8 (Fig. 3A), an A haplotype defined
independently in four families.
LD between alleles of the four subtyped genes and the presence or
absence of other KIR genes is shown in Table III. These data show that for genes that
are present in both A and B haplotypes, certain
alleles appear restricted to B haplotypes.
KIR2DL1*004 is in strong LD with the B haplotype
genes KIR2DL2, -2DL5, -2DS2, and
-2DS3 and appears to be exclusively found on B
haplotypes. Similarly, 3DL2*007 is also strongly associated
with the B haplotype genes 2DL5, 2DS1,
and 2DS5 which make up a different subset of B
haplotype genes than that for 2DL1*004. There are
corresponding negative LDs between 2DL1*004 and
3DL2*007 and the absence of B group loci (or
B group blanks).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In this study, family analysis and high resolution genotyping of the KIR2DL1, -2DL3, -3DL1, and -3DL2 genes were combined with low resolution typing of 13 genes to define 37 KIR haplotypes involving 10 different gene combinations. Demonstrated by these data is the extent to which allelic polymorphism diversifies KIR haplotypes having the same set of genes. This effect was most apparent for the A haplotypes, which all have the same arrangement of KIR genes. A total of 48 independent A haplotypes were defined, and within this group 22 different haplotypes were represented. Because allele level typing was focused on genes found on both A and B haplotypes, and not those that distinguish B haplotypes, the number of B haplotypes unambiguously defined, 24, was less than for A haplotypes. Moreover, because B haplotypes vary in their gene content, the number of defined B haplotypes having any given arrangement of KIR genes was further reduced. Fifteen haplotypes involving 9 different gene combinations were represented in the 24 independently defined haplotypes. Thus, there is considerable diversification of B haplotypes through the combination of different alleles within the same arrangement of genes. Our assessments of A and B haplotype diversity are likely to be an underestimate because allele level typing was not done for all KIR genes; for those genes thus analyzed, not all the known alleles could be discriminated. Despite these limitations, our approach was robust enough to identify four previously undiscovered KIR variants, one of which (KIR3DL2*010) is present on both A and B haplotypes.
Analysis of the high resolution typing obtained for the KIR2DL1, -2DL3, -3DL1, and -3DL2 genes revealed striking patterns of LD. Between 2DL1 and 2DL3 alleles, there is very high LD, and it is also high between 3DL1 and 3DL2 alleles. In contrast, between these two pairs of genes, which mark the two halves of the human KIR gene complex, LD was negligible. High LD is consistent with the compact nature of the KIR gene family (3, 4, 5), and its absence in the central interval between 2DL1 and 3DL1 suggest that this region is a focus for recombination. This region contains 2DL4, the pseudogene KIR3DP1 (also called KIRX) and a unique 14-kb stretch which includes L1 repeat sequences not present in the other KIR genes (5).
In general, the pressure from diverse and rapidly evolving pathogens selects for diversity and novelty in immune system genes. In this regard, the class I and II genes of the major histocompatibility complex have provided the precedent for genes so polymorphic that unrelated human individuals rarely have identical HLA genotype (36, 37, 38). Although HLA diversity has been appreciated for some decades, KIR genes and their diversity are recently discovered phenomena (15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26). Whereas HLA diversity almost entirely derives from allelic polymorphism, with differences in gene content playing a minor role, we demonstrate here that KIR diversity involves major contributions from both gene content and allelic polymorphism. Moreover, the data obtained permit assessment of the extent to which KIR genes individualize human genomes. A total of 9761 pairwise comparisons between the unrelated individuals in our panel revealed only 0.24% of them to be KIR identical. This number provides an estimate of the frequency with which unrelated individuals have identical KIR type. However, it is likely to be an overestimate because polymorphisms for only four KIR genes were typed at the allele level, and not all alleles at those loci were distinguished. Similar comparisons of the HLA class I types of the panel members showed that 0.01% of unrelated pairs had identical HLA class I genotype. From these data, it is seen that diversity of the KIR gene system ensures that the large majority of individuals within a population have different KIR genotype, a situation analogous to that seen for HLA. As for HLA, it seems likely that such diversity in human KIR genotype is the result of natural selection by pathogens.
The approach we describe here of targeting allele level typing to the KIR2DL1, -2DL3, -3DL1, and -3DL2 genes works well in defining A haplotypes because all four targeted genes are invariant components of these haplotypes. A haplotypes comprise an estimated 47–59% of KIR haplotypes in Caucasoid populations, including the one studied here (11, 27, 28, 29, 30). Of the four KIR genes typed at high resolution, only 3DL1 and 3DL2 are common components of B haplotypes. Consequently, the approach we took has been less effective in tracking the segregation of genes associated with B haplotypes and thus in defining B haplotypes. This limitation could be addressed by extending high resolution to the other KIR genes.
That the KIR gene system individualizes human genomes to an
extent approaching that of the HLA system means that bone
marrow transplantation between unrelated HLA-matched donors
and recipients will involve, with few exceptions, KIR
incompatibility. Also, 75% of the transplants involving
HLA-matched sibling donors will involve KIR
incompatibility. The detriment or benefit of KIR
incompatibilities in transplantation is less easy to predict but is
amenable to study in patients and donors who are typed at high
resolution for both HLA and KIR.
|
Acknowledgments |
|---|
-globin
primer sequences. |
Footnotes |
|---|
2 Current address: Department of Medicine and Therapeutics, University College Dublin, Dublin 7, Ireland.
3 Address correspondence and reprint requests to Dr. Peter Parham, Departments of Structural Biology and Microbiology and Immunology, Sherman Fairchild Building, Stanford University School of Medicine, Stanford, CA 94305. E-mail address: peropa{at}leland.stanford.edu
4 Abbreviations used in this paper: KIR, killer cell Ig-like receptors; LD, linkage disequilibrium; SSP, sequence-specific priming.
Received for publication October 4, 2001. Accepted for publication December 19, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. A. Guethlein, A. M. Older Aguilar, L. Abi-Rached, and P. Parham Evolution of Killer Cell Ig-Like Receptor (KIR) Genes: Definition of an Orangutan KIR Haplotype Reveals Expansion of Lineage III KIR Associated with the Emergence of MHC-C J. Immunol., July 1, 2007; 179(1): 491 - 504. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Ruggeri, A. Mancusi, M. Capanni, E. Urbani, A. Carotti, T. Aloisi, M. Stern, D. Pende, K. Perruccio, E. Burchielli, et al. Donor natural killer cell allorecognition of missing self in haploidentical hematopoietic transplantation for acute myeloid leukemia: challenging its predictive value. Blood, July 1, 2007; 110(1): 433 - 440. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-J. Liaw, W.-R. Chen, Y.-C. Huang, C.-W. Tsai, K.-C. Chang, and C.-L. Kuo Genomic Organization of the Chicken CD8 Locus Reveals a Novel Family of Immunoreceptor Genes J. Immunol., March 1, 2007; 178(5): 3023 - 3030. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Thananchai, G. Gillespie, M. P. Martin, A. Bashirova, N. Yawata, M. Yawata, P. Easterbrook, D. W. McVicar, K. Maenaka, P. Parham, et al. Cutting Edge: Allele-Specific and Peptide-Dependent Interactions between KIR3DL1 and HLA-A and HLA-B J. Immunol., January 1, 2007; 178(1): 33 - 37. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Yawata, N. Yawata, M. Draghi, A.-M. Little, F. Partheniou, and P. Parham Roles for HLA and KIR polymorphisms in natural killer cell repertoire selection and modulation of effector function J. Exp. Med., March 20, 2006; 203(3): 633 - 645. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Wills, O. Ashiru, M. B. Reeves, G. Okecha, J. Trowsdale, P. Tomasec, G. W. G. Wilkinson, J. Sinclair, and J. G. P. Sissons Human Cytomegalovirus Encodes an MHC Class I-Like Molecule (UL142) That Functions to Inhibit NK Cell Lysis J. Immunol., December 1, 2005; 175(11): 7457 - 7465. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Kuroki, N. Tsuchiya, M. Shiroishi, L. Rasubala, Y. Yamashita, K. Matsuta, T. Fukazawa, M. Kusaoi, Y. Murakami, M. Takiguchi, et al. Extensive polymorphisms of LILRB1 (ILT2, LIR1) and their association with HLA-DRB1 shared epitope negative rheumatoid arthritis Hum. Mol. Genet., August 15, 2005; 14(16): 2469 - 2480. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. P. Williams, A. R. Bateman, and S. I. Khakoo HANGING IN THE BALANCE: KIR and Their Role in Disease Mol. Interv., August 1, 2005; 5(4): 226 - 240. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Leung, R. Iyengar, B. Triplett, V. Turner, F. G. Behm, M. S. Holladay, J. Houston, and R. Handgretinger Comparison of Killer Ig-Like Receptor Genotyping and Phenotyping for Selection of Allogeneic Blood Stem Cell Donors J. Immunol., May 15, 2005; 174(10): 6540 - 6545. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Draghi, N. Yawata, M. Gleimer, M. Yawata, N. M. Valiante, and P. Parham Single-cell analysis of the human NK cell response to missing self and its inhibition by HLA class I Blood, March 1, 2005; 105(5): 2028 - 2035. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Rajalingam, P. Parham, and L. Abi-Rached Domain Shuffling Has Been the Main Mechanism Forming New Hominoid Killer Cell Ig-Like Receptors J. Immunol., January 1, 2004; 172(1): 356 - 369. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Pando, C. M. Gardiner, M. Gleimer, K. L. McQueen, and P. Parham The Protein Made from a Common Allele of KIR3DL1 (3DL1*004) Is Poorly Expressed at Cell Surfaces due to Substitution at Positions 86 in Ig Domain 0 and 182 in Ig Domain 1 J. Immunol., December 15, 2003; 171(12): 6640 - 6649. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Artavanis-Tsakonas, K. Eleme, K. L. McQueen, N. W. Cheng, P. Parham, D. M. Davis, and E. M. Riley Activation of a Subset of Human NK Cells upon Contact with Plasmodium falciparum-Infected Erythrocytes J. Immunol., November 15, 2003; 171(10): 5396 - 5405. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. R. van der Slik, B. P.C. Koeleman, W. Verduijn, G. J. Bruining, B. O. Roep, and M. J. Giphart KIR in Type 1 Diabetes: Disparate Distribution of Activating and Inhibitory Natural Killer Cell Receptors in Patients Versus HLA-Matched Control Subjects Diabetes, October 1, 2003; 52(10): 2639 - 2642. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. P. Martin, A. Bashirova, J. Traherne, J. Trowsdale, and M. Carrington Cutting Edge: Expansion of the KIR Locus by Unequal Crossing Over J. Immunol., September 1, 2003; 171(5): 2192 - 2195. [Abstract] [Full Text] [PDF]
|
|
|
|
, Alleles amplified by each primer pair. Approximate size of the
amplified DNA fragment is shown to the right. H or N
indicates high or normal annealing temperatures, respectively (detailed
in Materials and Methods). C, Locus- and
allele-specific primer sequences; priming sites correspond to the
nucleotide position, relative to the start codon, of the 3' end of each
primer. KIR2DS1v1 (
