| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Institute for Transplantation Diagnostics and Cell Therapeutics and
Institute of Pediatrics, Heinrich Heine University, Dusseldorf, Germany
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The KIR family consists of at least 13 different members with either two or three extracellular Ig domains (13). They can be further subdivided on functional and structural grounds into inhibitory and stimulatory receptors, the former containing inhibitory signal motifs in the cytoplasmic domain, which are deleted in the latter (14). Different inhibitory KIR are specific for HLA-A (KIR3DL2), HLA-B (KIR3DL1), and different subgroups of HLA-C (KIR2DL1–3) (15, 16, 17, 18, 19). The specificity of the six stimulatory KIR (KIR2DS1 through KIR2DS5 and KIR3DS1) is less well defined and might also include non-HLA class I ligands (12). Finally, the KIR2DL4 receptor, which binds to the nonclassical HLA class I molecule HLA-G, combines structural and functional features of inhibitory and stimulatory KIR and is also unique in its ubiquitous expression on all NK cells in most individuals (20, 21, 22).
The clonally distributed expression of KIR is a hallmark of human NK cells as is the clonal expression of Ly-49 on murine NK cells (23, 24). Individual NK cell clones express different sets of inhibitory and stimulatory KIR genes, ranging from one to eight different receptors (21). The distribution of KIR on NK cells appears to be largely stochastic: the frequency with which NK cells coexpress two KIR genes usually equals the product of the individual expression frequencies of both KIR genes. Combinatorial diversity of KIR leads to a broad range of functional NK cell specificities, which are thought to be critical for a rapid and sensitive detection of altered HLA class I levels on target cells. Nonetheless, certain restrictions appear to be operative, the most important being the requirement for expression of at least one self-specific inhibitory receptor, either a KIR or CD94:NKG2A (25). However, the educational process of tolerance induction in NK cells appears to have only subtle effects on the overall KIR repertoire (26). KIR-expressing T cells exhibit KIR repertoires very similar to NK cells from the same individual, yet T cells are thymically educated and are probably not subject to the same educational constraints as NK cells (27). A reasonable hypothesis emerging from these observations is that the KIR repertoire is largely determined by genetically defined factors with only minor contributions of education and selection of suitable NK cells. However, although KIR genes have diversified considerably in structure and function among the human population, the putative regulatory regions of KIR are remarkably conserved and provide few indications for genetically encoded regulatory mechanisms contributing to differential expression of KIR (13).
The observation that the putative promoter regions of KIR genes are highly conserved prompted us to speculate that epigenetic mechanisms rather than differences in KIR promoter sequences might be involved in establishing and maintaining differential expression patterns. Methylation of cytosines within CpG dinucleotides represents a frequent epigenetic modification, which in most cases is associated with repression of transcription (28). There is increasing evidence that epigenetic mechanisms are involved in modulation of immune responses (29). Most studies to date have focused on B and T lymphocytes, showing, for example, that developmental decisions, the control of B and TCR rearrangements, and the expression of cytokine genes are dependent on epigenetic modification of target genes (30, 31, 32, 33, 34). A characteristic feature of DNA methylation is that once established, DNA methylation patterns are maintained over many cell generations (35). Thus, DNA methylation has the potential of diversifying gene expression in individual cells of the same lineage as well as of transmitting the established expression status through clonal inheritance. Here, DNA methylation is shown to be critically involved in determination of clonal KIR expression patterns in NK cells.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The nomenclature used throughout the manuscript adheres to the HUGO Gene Nomenclature Committee (http://www.gene.ucl.ac.uk/nomenclature/genefamily/kir.html). KIR genes and transcripts are written in italics to distinguish them from their protein products. The HUGO Gene Nomenclature Committee nomenclature relates to the new CD nomenclature of MHC receptors (36) as follows: KIR2DL1(CD158a), KIR2DL2(CD158b1), KIR2DL3(CD158b2), KIR2DL4(CD158d), KIR2DL5(CD158f), KIR3DL1(CD158e1), KIR3DL2(CD158k), KIR2DS1(CD158 h), KIR2DS2(CD158j), KIR2DS3(no CD assigned), KIR2DS4(CD158i), KIR2DS5(CD158g), and KIR3DS1(CD158e2).
Preparation of DNA, RNA, and RT
DNA was extracted from human PBMC using the QIAamp DNA Blood Mini Kit (Qiagen, Hilden, Germany), and RNA was prepared using the RNeasy Mini Kit (Qiagen) according to the manufacturer’s instructions. First-strand cDNA synthesis was performed from 1.5 µg RNA by RT using oligo(dT) (PerkinElmer, Norwalk, CT) and Moloney murine leukemia virus reverse transcriptase (Life Technologies, Gaithersburg, MD) in a volume of 50 µl at 42°C for 1 h.
KIR typing
The PCR/single-strand polymorphism typing protocol originally developed by Uhrberg et al. (37) was modified to accommodate the recent discovery of additional KIR genes as well as the description of numerous novel alleles not included previously and was performed as previously described (38). In RNA-based KIR typing, GAPDH-specific primers were employed as a control for variation in RNA template amount.
Abs, cell separation, and flow cytometry
The following Abs, directly coupled to FITC, PE, or PC5, were
used in this study: anti-CD3, anti-CD16, anti-CD56, and the
KIR-specific reagents CD158a (anti-KIR2DL1/2DS1), KIR-NKAT2
(anti-KIR2DL2/2DL3/2DS2), and NKB1 (anti-KIR3DL1; all purchased
from BD PharMingen, San Diego, CA) as well as KARp50.3
(anti-KIR2DS4) and isotype-matched control mAbs for IgG1 and IgG2a
(obtained from Beckman Coulter, Krefeld, Germany). mAb DX31-PE
(anti-KIR3DL2) was provided by J. Phillips (DNAX, Palo Alto, CA).
Supernatants of hybridoma 33 (anti-KIR2DL4), provided by E. Long
(National Institute of Allergy and Infectious Disease, Bethesda, MD),
was stained with goat-anti-mouse IgG1-FITC (Beckman Coulter). PBMC
were separated by sedimentation over Ficoll-Paque (PAA Laboratories,
Linz, Austria). NK cells were enriched from PBMC using the CD56
MultiSort Kit, and residual T cells were subsequently depleted using
CD3 MicroBeads (Miltenyi Biotech, Bergisch-Gladbach, Germany) according
to the supplier’s instructions. The purity of NK cell preparations was
95%. Subsets of NK cells, expressing the indicated KIR, were
purified by five-parameter cell sorting to a purity of >99% using a
FACStar Plus (BD Biosciences, Mountain View, CA) equipped with a 2-W
argon ion laser (Innova 70, Coherent, Palo Alto, CA) operating at 488
nm and 140 mW.
Cell culture
Polyclonal NK cells were short term cultured in IMDM (Life Technologies) supplemented with 10% FCS, 2% human serum, and 200 U/ml IL-2. NK cell clones were generated from PBMC and maintained as previously described (21). The following NK cell lines were used: NKL (provided by M. J. Robertson, University of Indiana, Indianapolis, IN), NK-92 (provided by T. Tonn, Red Cross Blood Donor Service Hessia, Frankfurt, Germany), and NK3.3 (provided by J. Kornbluth, St. Louis University, MO). The T cell leukemia cell line Jurkat and the EBV-transformed B lymphoblastoid cell line RPMI 8866 were obtained from American Type Culture Collection (Manassas, VA). The prostate carcinoma cell lines DU145 and LNCaP were provided by W. A. Schulz (University Clinic of Dusseldorf, Dusseldorf, Germany). The compounds 5-aza-2'-deoxycytidine (5Aza-dC) and trichostatin A (both from Sigma-Aldrich, Munich, Germany) were used at final concentrations of 2 µM and 25 nM, respectively.
Genomic sequencing of bisulfite-converted DNA
Bisulfite conversion of DNA leads to conversion of all unmethylated cytosines into thymines, while methylated cytosines remain unchanged. Bisulfite conversion was performed using the CpGenome DNA Modification Kit (Intergen, NY). First-round PCR amplification of CpG islands of KIR2DL3 was performed using primers 2DL3-5'EcoRIb (5'-GGAATTCCGAAGAARAGTTTGAATTTTAGA-3') and 2DL3-3'BamHI (5'-CGGGATCCCGCCTATATCTCCAACTCTAAACC-3'), for KIR3DL2 using primers 2DL3-5'EcoRIb and 3DL2-3'BamHIB (5'-CGGGATCCCGCCTTCCAAAACTCACCAAC-3'), and for KIR3DL1 using primers 3DL1-5'EcoRI (5'-GGAATTCCGTTTTTTATGTTAGTATAGATTTTA-3') and 3DL1–3'BamHI (5'-CGGGATCCCGCCATATCTTTACCTCCAAATC-3'). Converted DNA (25 ng) was used as template in a 25-µl reaction volume, containing 25 pmol of each primer, 1 mM dNTPs, 3 mM MgCl2, and 0.5 U HotStarTaq DNA polymerase (Qiagen). After initial denaturation for 10 min at 95°C, 34 cycles were performed, each consisting of 90 s at 95°C, 55 s at 48°C, and 1 min at 72°C. One microliter of a 1/20 dilution of the first PCR was used for nested amplification of KIR2DL3, leading to a 319-bp fragment using primers 2DL3-5'EcoRI (5'-GGAATTCCGTGTATGAGAGGTTGGATTTG-3') and 2DL3-3'BamHIb (5'-CGGGATCCCGCYTTCCAAAACTCACCAAC-3') and to a 375-bp fragment using primers 2DL3-5'EcoRI and 2DL3-3'BamHI. Second-round PCR of KIR3DL2 was performed using primers 3DL2-5'EcoRI (5'-GGAATTCCATGTGTGAGAGGTTGGATTTG-3') and 3DL2-3'BamHIB, and of KIR3DL1 using primers 3DL1-5'EcoRIin (5'-GGAATTCCGTTAGTATAGATTTTAGGTATTT-3') and 3DL1-3'BamHI. The same amplification conditions were chosen as for first-round PCR. Following digestion with EcoRI and BamHI, amplification products were cloned into pBluescript SK+. Sequencing was performed using T7 primer (5'-TAATACGACTCACTATAGGG-3') and the DNA sequencing kit BigDye Terminator Cycle Sequencing Ready Reaction on a DNA analyzer (model 3700; PE Applied Biosystems, Foster City, CA).
In vitro methylation and NK cell transfection
A 225-bp fragment located directly upstream the initiation codon of the KIR2DL3 gene was amplified and subcloned into vector pGL3Basic (Promega, Mannheim, Germany) using XhoI/HindIII sites. A 812-bp BamHI/HindIII fragment from vector pRevTet-Off (Clontech, Heidelberg, Germany) containing the CMV promoter was subcloned into BglII/HindIII-precut pGL3Basic. Both constructs were methylated in vitro with SssI or HhaI methylase (both New England Biolabs, Frankfurt, Germany). Completion of methylation reactions was verified by control digestion with HhaI. NK3.3 cells were transfected with methylated and mock-methylated constructs using nucleofection technology (Amaxa, Cologne, Germany) according to the supplier’s instructions. Luciferase activity was measured by light emission measurement in a scintillation counter (LS6000 IC; Beckman, Munich, Germany). All samples were normalized for protein content, which was determined by absorption at 750 nm with the DC Protein Assay (Bio-Rad, Munich, Germany).
Computer sequence analysis
Genomic KIR sequences were obtained from GenBank (accession no. AC006293) and the Lawrence Livermore National Laboratories human Chr 19 sequence database (clone BC52946). CpG density blots of KIR genomic sequences were portrayed with CpG Island Grapher (http://tiamat.kaist.ac.kr/util/cgi-web/index.html). KIR sequences representing CpG islands were aligned using ClustalW at http://www2.ebi.ac.uk/clustalw/. The locations of repetitive elements belonging to Alu and LINE subfamilies were identified using RepeatMasker (http://ftp.genome.washington.edu/RM/RepeatMasker.html).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Analysis of the distribution of CpG dinucleotides in the
KIR locus revealed that CpG density is consistently
increased in areas surrounding the transcription initiation region of
KIR genes (data not shown). These CpG clusters largely
conform to the original definition of CpG islands, with a ratio of
expected-vs-observed CpG density of 0.6 (39). Although
comparatively small with a length of 350 bp, they are hereafter
referred to as CpG islands. The overall structure of the CpG islands is
similar in all expressed KIR, with complete conservation of
four CpG dinucleotides (-10, -9, -7, and -6) upstream of the
transcriptional start site with the exception of KIR2DL4,
which shows a highly divergent profile with lower CpG density (Fig. 1
). Furthermore, each KIR CpG
island, again with the exception of KIR2DL4, is consistently
found in close proximity to upstream repetitive elements of the Alu S
type, either represented by Alu Sx or Alu Sg elements, as shown for
KIR2DL3 (Fig. 1, upper part). Alu repeats have
been previously suggested to promote de novo methylation of neighboring
DNA regions (40, 41).
|
We then assessed whether methylation patterns of CpG islands
correlate with transcriptional activity of the respective
KIR genes. To this end, methylation patterns of
KIR2DL3 CpG islands were analyzed in
KIR2DL3-expressing NK3.3 and
KIR2DL3-nonexpressing NKL cell lines, employing the
bisulfite genomic sequencing technique. Cloning and sequencing of PCR
products from bisulfite-converted DNA enabled assessment of the
methylation status of each CpG site individually. As shown in Fig. 2A (left
panel), the CpG island of KIR2DL3 in NK3.3 cells
exhibited heterogeneous methylation patterns ranging from completely
unmethylated patterns (10 out of 57 obtained sequences) to complete
methylation represented by a single sequence. In contrast, CpG islands
of nonexpressed KIR2DL3 in NKL as well as nonexpressed
KIR3DL2 in NK3.3 showed overall dense methylation with only
sporadic occurrence of unmethylated CpG dinucleotides (Fig. 2A, middle and right panels).
Differences in methylation status are further illustrated by the fact
that CpG islands with <70% (10 of 14 analyzed CpG positions) of
methylated CpGs were exclusively found in the expressed
KIR2DL3 in NK3.3, whereas CpG islands of nonexpressed
KIR2DL3 in NKL and KIR3DL2 in NK3.3 were
consistently methylated at 70–100% of CpG dinucleotides (Fig. 2B). Analysis of the methylation status at individual CpG
sites demonstrated that differential methylation of expressed vs
nonexpressed KIR is not restricted to specific CpGs, but is
consistently found throughout the CpG island (Fig. 2C). CpG
positions -6 and especially -7 exhibited overall lower levels of
methylation than surrounding CpG sites.
|
). NK cells exhibiting surface
expression of KIR3DL1 were almost completely unmethylated at all CpG
sites in the 5'-untranslated region and the first exon of
KIR3DL1 (Fig. 3, A and C). A similarly
striking lack of CpG methylation was observed in the CpG island of
KIR2DL3 in primary NK cells purified from PBMC for surface
expression of KIR2DL3 (Fig. 3, B and D). A
considerable increase in methylated CpGs was observed further
downstream in the first intron at CpG site 8 in KIR3DL1 as
well as KIR2DL3-expressing NK cells (Fig. 3, C
and D). Nonexpressed KIR genes were generally
hypermethylated compared with expressed KIR genes, as shown
for the CpG island of KIR3DL1 in polyclonal NK cells lacking
surface expression of KIR3DL1, but exhibiting expression of other KIR
(Fig. 3, A and C). Similarly, polyclonal NK cells
lacking surface expression of KIR2DL3 were densely methylated (Fig. 3D).
|
We next investigated whether DNA methylation of CpG islands is required for the maintenance of a transcriptionally silent state of KIR genes. A panel of cell lines was exposed to the demethylating compound 5Aza-dC, which acts as a DNA methyltransferase inhibitor and leads to genome-wide demethylation in dividing cells (42). As a control for the effectiveness of 5Aza-dC treatment we used methylation-specific PCR, demonstrating that 5Aza-dC treatment consistently resulted in demethylation of KIR CpG islands (data not shown).
Three different NK cell lines were analyzed, exhibiting three different
KIR genotypes (Fig. 4, upper panel) and KIR expression patterns (Fig. 4, middle panel), as assessed by genomic as well as
RNA-based typing for KIR genes. Untreated NK cell lines
exhibited transcription of either no KIR (NKL), only
KIR2DL3 (NK3.3), or several different KIR
(NK-92). It should be mentioned that the NK-92 cell line was previously
described as expressing KIR2DL4 only (43). Following
incubation with 5Aza-dC, a generalized induction of transcription of
all genomically present KIR genes was seen in all three NK
cell lines within 48 h (Fig. 4, lower panel). In
contrast, lymphoid cell lines of non-NK cell type either exhibited
induction of some KIR genes only following 5Aza-dC exposure
in the case of the T cell line Jurkat or no induction at all in the
case of the B cell line RPMI 8866 (Fig. 4). Nonlymphoid cell lines such
as the prostate carcinoma cell lines DU145 and LNCaP did not show
induction of KIR expression either (data not shown).
|
). The NK
cell line NKL, for example, exhibited de novo surface expression of all
five mAb-reactive KIR groups following 5Aza-dC treatment. The overall
surface expression patterns, characterized by subpopulations of cells
positive for one or two different KIRs, remained stable when 5Aza-dC
was removed and NKL cells were cultured for another 2 wk (data not
shown). In NK-92, which shows the expression of certain KIR on small
subsets of cells, 5Aza-dC treatment led to induction of KIR expression
on >90% of cells. Furthermore, the flow cytometric analysis of NK-92
cells demonstrated that the levels of KIR expression on the cell
surface induced by 5Aza-dC treatment largely match the levels of KIRs
observed on the small KIR-expressing subpopulation of untreated cells.
Efficient induction of KIR protein expression was also seen in
5Aza-dC-treated NK cell clones as well as polyclonal NK cells (Fig. 5, left panel). In contrast, in the T cell line Jurkat a slight
induction of KIR3DL2 (corresponding to the induction of
KIR3DL2 transcription shown in Fig. 4
), but no other KIR,
was detectable following 5Aza-dC treatment (Fig. 5, right
panel). No KIR expression was detectable on the cell surface of
RPMI 8866 or any of the nonlymphoid cell lines following treatment with
5Aza-dC (data not shown).
|
Analysis of the initial phase of 5Aza-dC-mediated KIR
induction did not reveal a particular sequential order of
KIR expression (Fig. 6).
Instead, 5Aza-dC treatment caused the rapid and simultaneous expression
of multiple KIR genes regardless of location within the
KIR locus, functional distinction in stimulatory and
inhibitory KIR, or specificity. For most KIR,
transcription was detectable within the first 8 h of 5Aza-dC
treatment (Fig. 6A).
|
B). Similarly,
flow cytometric analysis did not exhibit any qualitative or
quantitative changes in KIR protein expression following TSA treatment
(data not shown). Additionally, combined treatment with 5Aza-dC and TSA
did not lead to a synergistic effect compared with 5Aza-dC alone (Fig. 6C). In contrast, surface expression levels of CD56 on NKL
were substantially up-regulated, providing a control for the
effectiveness of TSA treatment in these experiments (data not shown).
TSA treatment was similarly ineffective in freshly isolated NK cells
(data not shown). The expression of KIR2DL4 was not affected
by either demethylating or histone-acetylating treatment as assessed by
RT-PCR (Fig. 6) as well as flow cytometric analysis using a
KIR2DL4-specific mAb (data not shown). In vitro methylation of CpG islands leads to repression of transcriptional activity
To assess whether a causal relationship exists between methylation
of CpG islands of KIR and repression of KIR
transcription, a DNA fragment of KIR2DL3 encompassing the
CpG island from CpG position -16 up to the start codon (see Fig. 1 for
reference) was subcloned into a luciferase reporter gene vector and
subsequently methylated. Two different methylases were employed for in
vitro methylation experiments: SssI methylase, which leads
to methylation of all CpG sites, and HhaI methylase, which
targets only three CpG sites (CpG positions -7, -6, and -4).
Methylated constructs were transiently transfected into NK3.3 cells. In
vitro methylation of CpG islands using SssI methylase led to
repression of reporter gene activity down to baseline levels (Fig. 7A). Partial methylation of
selected CpG sites in the CpG island fragment using HhaI
methylase resulted in a less pronounced, but significant, decrease in
transcriptional activity. A CMV promoter, lacking HhaI
sites, was only marginally affected by HhaI methylase
treatment, demonstrating that the observed reduction in transcriptional
activity of KIR CpG islands is specific and not due to
methylation of the plasmid backbone (Fig. 7, A and
B).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Analysis of the methylation patterns of CpG islands surrounding the
transcriptional start site of each KIR gene revealed that
the methylation status of KIR genes consistently correlates
with their transcriptional activity in primary NK cells as well as NK
cell lines. In general, methylated CpG islands were associated with
transcriptionally silent KIR and unmethylated CpG islands
with expressed KIR genes. This correlation was most striking
in ex vivo-isolated NK cells, where expressed KIR genes were
strictly demethylated throughout the CpG island with few exceptions.
Remarkably, in ex vivo-isolated NK cells, CpG islands of expressed
KIR genes were unmethylated on both alleles as shown for an
individual with two alleles of KIR2DL3 (Fig. 3D).
It is thus possible that KIR genes are expressed in vivo in
a biallelic fashion, which is supported by a previous study showing
preferential biallelic expression of KIR3DL1 in NK cell
clones (46). The methylation patterns of the CpG island of
the expressed KIR2DL3 gene in NK3.3 were much more
heterogeneous, but it is unclear whether this exactly reflects the in
vivo situation, since cell type-specific genes are often progressively
methylated during long term culture (47). Our results thus
indicate that epigenetic regulation of KIR genes leads to a
preferential biallelic mode of expression in NK cells in vivo, which
contrasts with the functionally analogous murine Ly-49 genes, which are
expressed in a predominantly monoallelic fashion (48, 49).
Until now, attempts to induce KIR expression on NK cells have remained
unsuccessful. In this study the methyltransferase inhibitor 5Aza-dC for
the first time enabled experimental induction of KIR expression.
Exposure to 5Aza-dC effectively induced the whole range of KIR in NK
cells, but not in other cells of lymphoid or nonlymphoid origin. This
mirrors the physiological distribution of KIR, which is largely
restricted to NK cells and subsets of Ag-experienced T cells
(50). Notably, a weak induction of KIR3DL2 following
treatment with 5Aza-dC was observed in the T cell line Jurkat (Fig. 4).
These results suggest that in NK cells and T cells, but not other cell
types, repression of KIR gene transcription is dependent on
DNA methylation. Furthermore, it is reasonable to believe that the cell
type-specific effect of 5Aza-dC on KIR expression reflects
the presence of KIR-specific transcription factors in NK and
T cells, but not in other cell types. It is important to note that in
vitro methylation of the KIR2DL3 CpG island led to a
strongly repressed promotorial activity in reporter gene assays. These
results demonstrate that the induction of KIR transcription
is a direct cause of 5Aza-dC-mediated demethylation of KIR
CpG islands and not a secondary effect of 5Aza-dC on transcription
factor activity.
Recent studies suggest that CpG methylation frequently induces remodeling of chromatin and subsequent gene silencing (51). Transcriptionally competent chromatin is converted to a repressive state through specific binding of methyl-CpG-binding proteins, which are associated with histone deacetylases, leading to a condensed chromatin structure. In the present study altering the state of histone acetylation using the deacetylase inhibitor TSA did not change transcriptional activity of either expressed or silent KIR genes. KIR genes in NK cells might thus have already acquired a state of transcriptionally competent chromatin even in KIR genes, which are not transcribed. The dense clustering of KIR genes with short intergenic regions probably leads to an overall transcriptionally competent chromatin structure once KIR expression is initiated in NK cells. In general, methylation of KIR CpG islands appears to be the critical epigenetic modification required for silencing of specific KIR genes.
Differential methylation of common regulatory elements provides a suitable working model to explain clonal expression of KIR genes. Predictive analysis of transcription factor binding sites in the upstream regions of KIR genes suggested that most putative sites are shared by most KIR (13, 52). The high level of sequence conservation of putative promoter regions and transcription factor binding sites strongly argues in favor of a common mechanism of transcriptional activation. In a simple model of KIR gene regulation, clonal expression patterns would be determined by competition of DNA binding proteins for access to upstream CpG islands, some of which induce KIR expression by transcriptional activation and others that restrict KIR expression by mediating de novo methylation. As shown elsewhere, KIR CpG islands overlap with promoter regions in all clonally distributed KIR genes (our unpublished observations). This model of stochastic regulation would apply to all KIR genes except KIR2DL4, which is not clonally distributed but is expressed in all NK cells and might well be regulated in a methylation-independent way.
Once clonal patterns of KIR expression are determined, they are apparently stably maintained in NK cells over many cell generations (21, 53). Differential methylation patterns could provide the basis for stable propagation of the respective KIR expression patterns. Epigenetic stability of methylation patterns is attributable to the maintenance methyltransferase DNA methyltransferase I. Through conversion of hemimethylated to symmetrically methylated sites during replication, methylation patterns are faithfully transmitted to newly generated daughter cells (54). Whereas the maintenance of methylation patterns is a well-described feature of DNMT1, it is less clear how cell type-specific methylation patterns are established, in particular, how de novo methylation arises. Recently, additional methyltransferases with de novo methylation activity were described, but it is largely unknown how specific target genes are selected by these enzymes (55).
A common feature of KIR CpG islands is the proximity to Alu repetitive elements, a subclass of short interspersed nuclear elements. It has been proposed that Alu repeats and related short interspersed nuclear elements might provide targets for de novo methylases and promote spreading of methylation to neighboring regions (56, 57). In the KIR locus, different subclasses of Alu repeats were consistently found within several hundred base pairs upstream of CpG islands of KIR genes except the ubiquitously expressed KIR2DL4. The specific topology of upstream KIR regions with Alu elements neighboring CpG islands might play a role in directing de novo methylases to KIR genes.
There are several examples of clusters of homologous genes in the
mammalian genome, such as the hox genes and the
-globin genes, which are developmentally regulated in a
position-dependent way. In the case of the -globin genes,
a locus control region could be defined that mediates position- and
orientation-dependent expression (58). The present study
did not reveal evidence for position-dependent effects or a sequential
order of KIR gene expression following demethylating
treatment. The kinetics of 5Aza-dC-induced KIR expression
rather suggest that most if not all KIR genes are equally
likely to be targeted by the transcription machinery once
methylation-induced repression is released. This is also supported by
the conserved structure of the different KIR promoters as
well as the combinatorial frequencies of expressed KIR genes
in NK cell clones, which largely fit a model of random combination of
KIR (21).
It is currently unknown at which stage of NK cell development methylation patterns are imprinted on KIR genes. Although the majority of CpG dinucleotides in mammals are methylated, CpG islands usually remain unmethylated in all tissues including those not expressing the relevant genes (54). In that respect, methylation of KIR CpG islands would constitute a modification of the unmethylated default state. Analysis of hemopoietic stem cells as well as NK cell lineage-restricted progenitor populations will help to define whether determination of KIR expression patterns in developing NK cells is a consequence of CpG methylation of previously unmethylated CpG islands or vice versa if methylated CpG islands are subject to demethylation.
In summary, the present study suggests that in NK cells methylation of KIR CpG islands is necessary and sufficient to maintain clonal KIR expression patterns by differential repression of KIR transcription.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Markus Uhrberg, University Clinic of Dusseldorf, Institute for Transplantation Diagnostics and Cell Therapeutics, Building 14.80, Moorenstrasse 5, D-40225 Dusseldorf, Germany. E-mail address: uhrberg{at}itz.uni-duesseldorf.de
3 Abbreviations used in this paper: KIR, killer cell Ig-like receptors; 5Aza-dC, 5-aza-2'-deoxycytidine; TSA, trichostatin A.
Received for publication June 14, 2002. Accepted for publication August 14, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TCR rearrangement express highly diverse killer cell Ig-like receptor patterns. J. Immunol. 166:3923.
promoter demethylation in CD44highCD8+ T lymphocytes. J. Immunol. 162:5053.-fetoprotein control region. J. Biol. Chem. 269:1821.-globin locus regulation. Cell 100:499.[Medline]This article has been cited by other articles:
|
|
 |
|
  R. Farcas, E. Schneider, K. Frauenknecht, I. Kondova, R. Bontrop, J. Bohl, B. Navarro, M. Metzler, H. Zischler, U. Zechner, et al. Differences in DNA Methylation Patterns and Expression of the CCRK Gene in Human and Nonhuman Primate Cortices Mol. Biol. Evol., June 1, 2009; 26(6): 1379 - 1389. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Li, C. M. Weyand, and J. J. Goronzy Epigenetic mechanisms of age-dependent KIR2DL4 expression in T cells J. Leukoc. Biol., September 1, 2008; 84(3): 824 - 834. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Caligiuri Human natural killer cells Blood, August 1, 2008; 112(3): 461 - 469. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Santourlidis, N. Graffmann, J. Christ, and M. Uhrberg Lineage-Specific Transition of Histone Signatures in the Killer Cell Ig-Like Receptor Locus from Hematopoietic Progenitor to NK Cells J. Immunol., January 1, 2008; 180(1): 418 - 425. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Averdam, H. Kuhl, M. Sontag, T. Becker, A. L. Hughes, R. Reinhardt, and L. Walter Genomics and Diversity of the Common Marmoset Monkey NK Complex J. Immunol., June 1, 2007; 178(11): 7151 - 7161. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Romagnani, K. Juelke, M. Falco, B. Morandi, A. D'Agostino, R. Costa, G. Ratto, G. Forte, P. Carrega, G. Lui, et al. CD56brightCD16- Killer Ig-Like Receptor- NK Cells Display Longer Telomeres and Acquire Features of CD56dim NK Cells upon Activation J. Immunol., April 15, 2007; 178(8): 4947 - 4955. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Uhrberg, K. Schonberg, S. Weinhold, and H.-I. Trompeter Non-viral gene delivery into primary natural killer lymphocytes FASEB J, December 1, 2006; 20(14): 2660 - 2660. [Full Text] [PDF]
|
|
|
|
 |
|
 S. R. Presnell, L. Zhang, C. A. Ramilo, H.-W. Chan, and C. T. Lutz Functional redundancy of transcription factor-binding sites in the killer cell Ig-like receptor (KIR) gene promoter Int. Immunol., August 1, 2006; 18(8): 1221 - 1232. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. L. Rogers, A. Rouhi, F. Takei, and D. L. Mager A Role for DNA Hypomethylation and Histone Acetylation in Maintaining Allele-Specific Expression of Mouse NKG2A in Developing and Mature NK Cells J. Immunol., July 1, 2006; 177(1): 414 - 421. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Rouhi, L. Gagnier, F. Takei, and D. L. Mager Evidence for epigenetic maintenance of Ly49A monoallelic gene expression. J. Immunol., March 1, 2006; 176(5): 2991 - 2999. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. A. Moon, S. M. Veniamin, J. A. Parks-Dely, and K. E. Magor The MHC of the Duck (Anas platyrhynchos) Contains Five Differentially Expressed Class I Genes J. Immunol., November 15, 2005; 175(10): 6702 - 6712. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-W. Chan, J. S. Miller, M. B. Moore, and C. T. Lutz Epigenetic Control of Highly Homologous Killer Ig-Like Receptor Gene Alleles J. Immunol., November 1, 2005; 175(9): 5966 - 5974. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. S. Woll, C. H. Martin, J. S. Miller, and D. S. Kaufman Human Embryonic Stem Cell-Derived NK Cells Acquire Functional Receptors and Cytolytic Activity J. Immunol., October 15, 2005; 175(8): 5095 - 5103. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Xu, A. N. Vallejo, Y. Jiang, C. M. Weyand, and J. J. Goronzy Distinct Transcriptional Control Mechanisms of Killer Immunoglobulin-like Receptors in Natural Killer (NK) and in T Cells J. Biol. Chem., June 24, 2005; 280(25): 24277 - 24285. [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]
|
|
|
|
|
|
H.-I. Trompeter, N. Gomez-Lozano, S. Santourlidis, B. Eisermann, P. Wernet, C. Vilches, and M. Uhrberg Three Structurally and Functionally Divergent Kinds of Promoters Regulate Expression of Clonally Distributed Killer Cell Ig-Like Receptors (KIR), of KIR2DL4, and of KIR3DL3 J. Immunol., April 1, 2005; 174(7): 4135 - 4143. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. K. Epling-Burnette, J. S. Painter, P. Chaurasia, F. Bai, S. Wei, J. Y. Djeu, and T. P. Loughran Jr Dysregulated NK receptor expression in patients with lymphoproliferative disease of granular lymphocytes Blood, May 1, 2004; 103(9): 3431 - 3439. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M. Tanamachi, D. C. Moniot, D. Cado, S. D. Liu, J. K. Hsia, and D. H. Raulet Genomic Ly49A Transgenes: Basis of Variegated Ly49A Gene Expression and Identification of a Critical Regulatory Element J. Immunol., January 15, 2004; 172(2): 1074 - 1082. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Nakajima, A. Asai, A. Okada, L. Ping, F. Hamajima, T. Sata, and K. Isobe Transcriptional Regulation of ILT Family Receptors J. Immunol., December 15, 2003; 171(12): 6611 - 6620. [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]
|
|
|
|
|
|
B. T. Wilhelm, J.-R. Landry, F. Takei, and D. L. Mager Transcriptional Control of Murine CD94 Gene: Differential Usage of Dual Promoters by Lymphoid Cell Types J. Immunol., October 15, 2003; 171(8): 4219 - 4226. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Belkin, M. Torkar, C. Chang, R. Barten, M. Tolaini, A. Haude, R. Allen, M. J. Wilson, D. Kioussis, and J. Trowsdale Killer Cell Ig-Like Receptor and Leukocyte Ig-Like Receptor Transgenic Mice Exhibit Tissue- and Cell-Specific Transgene Expression J. Immunol., September 15, 2003; 171(6): 3056 - 3063. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Ioannidis, B. Kunz, D. M. Tanamachi, L. Scarpellino, and W. Held Initiation and Limitation of Ly-49A NK Cell Receptor Acquisition by T Cell Factor-1 J. Immunol., July 15, 2003; 171(2): 769 - 775. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Stewart, J. van Bergen, and J. Trowsdale Different and Divergent Regulation of the KIR2DL4 and KIR3DL1 Promoters J. Immunol., June 15, 2003; 170(12): 6073 - 6081. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H.-W. Chan, Z. B. Kurago, C. A. Stewart, M. J. Wilson, M. P. Martin, B. E. Mace, M. Carrington, J. Trowsdale, and C. T. Lutz DNA Methylation Maintains Allele-specific KIR Gene Expression in Human Natural Killer Cells J. Exp. Med., January 20, 2003; 197(2): 245 - 255. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and the location of CpG dinucleotides (small vertical
lines) indicated. The lower part of the figure
represents a schematic sequence alignment of CpG islands from 12
different KIR genes encompassing a genomic region from
position -271 bp to position +82 bp relative to the start codon of
KIR2DL3 (depicted by a vertical arrow). CpG
dinucleotides are depicted by vertical lines. The transcriptional start
site of KIR2DL3 is indicated by a horizontal arrow. The
row designated 
CpG is a cumulative portray of all CpG dinucleotides
found in any of the aligned KIR sequences except
KIR2DL4. Numbers below this row refer to consecutively
numbered CpG positions with plus/minus signs to indicate location
relative to the start codon. An insertion of five nucleotides in the
intron of KIR2DL4 is indicated by two slashes.
and • representing unmethylated and methylated cytosines,
respectively). The number of sequences with an identical methylation
pattern is shown at the right end of each row, and the
total number of analyzed sequences is given at the lower
right end of each panel. Positions of CpG dinucleotides are
shown below each column and are numbered as described in Fig. 1
), and
KIR3DL2 in NK3.3 (
). Data are obtained from the
sequence pools shown in A. C, The methylation frequency
at individual CpG positions is shown for CpG islands of
KIR2DL3 in NK3.3 (
