Downregulation of Death-Associated
Protein Kinase 1 (DAPK1) in
Chronic Lymphocytic Leukemia
Aparna Raval,1,10Stephan M. Tanner,1John C. Byrd,2Elizabeth B. Angerman,1James D. Perko,1
Shih-Shih Chen,1Bjo ¨rn Hackanson,1,8Michael R. Grever,2David M. Lucas,2Jennifer J. Matkovic,2
Thomas S. Lin,2Thomas J. Kipps,6Fiona Murray,7Dennis Weisenburger,4Warren Sanger,4Jane Lynch,4
Patrice Watson,4Mary Jansen,4Yuko Yoshinaga,3Richard Rosenquist,7Pieter J. de Jong,3Penny Coggill,5
Stephan Beck,5Henry Lynch,4Albert de la Chapelle,1,9,* and Christoph Plass1,9,*
1Department of Molecular Virology, Immunology, and Medical Genetics, Human Cancer Genetics Program,
The Comprehensive Cancer Center at The Ohio State University, Columbus, OH 43214, USA
2Department of Internal Medicine, Division of Hematology and Oncology, The Ohio State University, Columbus, OH 43214, USA
3Children’s Hospital Oakland Research Institute, Oakland, CA 94609, USA
4Department of Preventive Medicine and Public Health, Creighton University, Omaha, NB 68178, USA
5Wellcome Trust Sanger Institute, Genome Campus, Hinxton, Cambridge, UK
6Department of Internal Medicine, University of California at San Diego, San Diego, CA, 92093, USA
7Department of Genetics and Pathology, Rudbeck Laboratory, Uppsala University, SE-751 85 Uppsala, Sweden
8Department of Hematology/Oncology, University of Freiburg Medical Center, Freiburg, Germany
9Present address: The Ohio State University, Human Cancer Genetics Program, Tzagournis Medical Research Facility 464A,
420 West 12th Avenue, Columbus, OH 43210, USA.
10Present address: Department of Oncology, CCSR 2250, Stanford University, Stanford, CA 94305, USA.
*Correspondence: firstname.lastname@example.org (A.d.l.C.), email@example.com (C.P.)
The heritability of B cell chronic lymphocytic
leukemia (CLL) is relatively high; however, no
predisposing mutation has been convincingly
identified. We show that loss or reduced ex-
pression of death-associated protein kinase 1
(DAPK1) underlies cases of heritable predispo-
sition to CLL and the majority of sporadic CLL.
Epigenetic silencing of DAPK1 by promoter
methylation occurs in almost all sporadic CLL
cases. Furthermore, we defined a disease hap-
lotype, which segregates with the CLL pheno-
type in a large family. DAPK1 expression of
the CLL allele is downregulated by 75% in
germline cells due to increasedHOXB7 binding.
In the blood cells from affected family mem-
bers, promoter methylation results in additional
loss of DAPK1 expression. Thus, reduced ex-
pression of DAPK1 can result from germline
predisposition, as well as epigenetic or somatic
events causing or contributing to the CLL
Chronic lymphocytic leukemia (CLL) is one of the most
common types of adult leukemias and is currently incur-
able. There were 8190 new cases of CLL in the United
States in 2004 (Kasper and Harrison, 2005). Familial oc-
currence of CLL has been noted in up to 10% of cases
(Yuille et al., 2000); however, large pedigrees with many
affected individuals are exceedingly rare. Most familial
agglomerations consist of two or three affected first- or
second-degree relatives (Sellick et al., 2005). Large
case-control studies concluded that the risk ratio (RR),
a measurement of the disease frequency in first-degree
relatives of probands, was higher for CLL than for most
other cancers (Goldgar et al., 1994). While the average
RR for all cancers in a US study was approximately 2.1,
CLL showed a RR of 5.0, the fourth highest of all cancers
(Goldgar et al., 1994; Risch, 2001). For CLL a RR of 7.5
was recently calculated in Sweden (Goldin et al., 2004).
In a study of the entire Icelandic population, lymphoid
leukemia had the second highest RR of all malignancies
studied (Amundadottir et al., 2004).
This evidence of high heritability has prompted investi-
gators to conduct genome-wide searches for linkage in
samples from CLL families. The findings, in particular
from one US and one European-led consortium (Goldin
et al., 2003; Sellick et al., 2005), have been limited due
to weak evidence for linkage at several loci, most of which
were different between the two studies. Loci in chromo-
some bands 11p11 and 13q21, respectively, were backed
by statistically significant evidence, but no gene has been
implicated to date.
Studies uncovering epigenetic aberrations in CLL have
accelerated the search for CLL-related genes. A genome-
wide DNA methylation analysis of CLL samples identified
almost 200 novel epigenetically silenced genes in CLL.
Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc. 879
The study concluded that on average 4.8% of all CpG
islands in a CLL genome could be targeted for aberrant
DNA methylation and associated gene silencing (Rush
et al., 2004). The role of aberrant methylation in CLL
was highlighted by the finding of preferential promoter
methylation-directed gene silencing of ZAP-70 and
TWIST2 in subgroups of CLL defined by IgVHmutational
status (Corcoran et al., 2005; Raval et al., 2005). The
emerging concept from these findings is that DNA methyl-
changes that have been described for CLL (Rosenwald
et al., 2003).
tify a novel putative tumor suppressor in CLL. We present
evidence that epigenetic silencing and/or germline muta-
tions in death-associated protein kinase 1 (DAPK1), a
positive mediator of apoptosis, contribute to familial CLL
and that this gene is silenced in virtually all cases of
to global expression
Frequent Epigenetic Inactivation of DAPK1 in CLL
DAPK1 was initially isolated as a positive mediator of ap-
Previous studies, using methylation-specific PCR (MSP),
demonstrated low frequencies of aberrant DNA methyla-
tion of the DAPK1 promoter in CLL (Chim et al., 2006;
Katzenellenbogen et al., 1999). We expanded the analysis
of DNA methylation events in the DAPK1 CpG island and
performed quantitative high-throughput analysis of DNA
methylation in four amplicons using the MassARRAY sys-
A1–A4 was 17, 22, 20, and 29, respectively. The average
DNA methylation frequency in four samples of normal
peripheral blood mononuclear cells (PBMCs) isolated
from healthy volunteers and seven samples of normal
CD19+ B cells was 6.3% and 7.6% (range 5.1%–8.9%),
respectively. In contrast, we saw an average of 64%
Figure 1. DAPK1 Promoter Methylation
in CLL Samples
(A) Schematic representation of the DAPK1
gene showing the location of the CpG island
(black bar) and the four bisulfite reaction
amplicons (A1–A4). The arrow indicates the
predicted transcription start site. The lower
panel shows a graphical display of quantitative
DNA methylation data for the DAPK1 promoter
region. Each square represents asingle CpG or
a group of two or three CpGs analyzed, and
each row represents a sample. Methylation fre-
quencies extend from light yellow (0%) to dark
blue (100%). Gray indicates unavailable data.
Samples included seven CD19+ selected con-
trol B cells, four control normal PBMCs, seven
sporadic CLL samples, Raji and Jurkat cells.
Asterisks indicate CLL samples that showed
less than 11% methylation.
(B) Plot of average percentage methylation in
regions A1–A4, in controls (seven CD19+
selected B cells and four PBL samples; total =
11), CD19+ selected CLL cells (n = 7), unse-
lected CLL cells (n = 62), and Raji and Jurkat
880 Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc.
(range 55.9%–83.9%) DNA methylation in seven CD19+
selected CLL samples (Figure 1B). Next, we evaluated
the PBMCs from 62 sporadic CLL patient samples. These
CLL samples showed varying levels of DAPK1 promoter
methylation, ranging from 7.2% to 66.1%. Only 2 out of
62 CLL samples (marked as * in Figure 1A) showed meth-
ylation levels in the normal range (<11%), whereas all
other samples showed levels >11%. The distribution of
DNA methylation levels in CLL samples is significantly dif-
ferent from the one seen in normal cells (p < 0.0001). Raji
cells were methylated (83.9%), whereas Jurkat cells were
unmethylated in regions A1–A3 but methylated in A4. In
addition, we tested the DNA methylation status in several
leukemia/lymphoma cell lines—Ramos (Burkitt lym-
phoma), BJAB (atypical Burkitt lymphoma [EBV?]),
MEC-1 (CLL), MEC-2 (CLL), Wac3CD5 (CLL), Daudi (Bur-
kitt lymphoma), 697 (ALL), RS11846 (Non-Hodgkin lym-
phoma), and RS 4;11 (ALL) —using COBRA for the two
most densely methylated regions in the DAPK1 promoter.
With the exception of Wac3CD5, all cell lines demon-
strated aberrant DNA methylation as compared to
CD19+selected Bcells (Figures S1Aand S1B). Chromatin
immunoprecipitation (ChIP) assay, using antiacetylated
histone H3 and H4 antibodies, showed that in Raji cells
(methylated), the DAPK1 promoter was not associated
with either ac-H3 or ac-H4 histones, while in WaC3CD5
and Jurkat cell lines (unmethylated) acetylated histones
were found within this region (Figures S1C–S1F).
To study the relevance of promoter methylation for
structs to test DAPK1 promoter activity in Jurkat cells.
DAPK1 promoter construct no. 2 (c.1-1545–c.1-1151
bp), covering bisulfite region A1 showed a 30-fold in-
crease in the reporter activity relative to the vector control
(Figure 2A). Semiquantitative RT-PCR analysis of CD19+
CLL samples used in Figure 1A showed reduced DAPK1
expression in all seven samples (Figure 2B). Furthermore,
DAPK1 expression in 50 unselected CLL cells also
showed statistically significant (p < 0.01) reduction as
compared to normal CD19+ B cells (Figure 2C). The func-
tional relevance of DAPK1 promoter methylation was
tested in a reporter assay in which the promoter sequence
of construct no. 2 (Figure 2A) was ligated either following
SssI treatment or untreated. Luciferase expression was
Figure 2. Epigenetic Silencing of DAPK1
(A) pGL3 luciferase constructs ligated with dif-
ferent DAPK1 50upstream inserts (nos. 1–4)
were transfected into Jurkat cells, and reporter
activity was studied 48 hr after transfection.
tive control. Error bars indicate standard devia-
(B) Quantitative DAPK1 expression in normal B
cells and seven CD19+ selected CLL samples.
The expression in CLL samples is shown rela-
tive to the expression in B cells (defined as
1.0). Error bars indicate SD.
(C) Box plot of DAPK1 expression in 50 CLL
samples as measured by semiquantitative
RT-PCR and compared to its expression in
six normal B cell samples. The distribution is
significantly different (p < 0.01).
(D) Luciferase assays in 293T cells with either
methylated or unmethylated DAPK1 construct
no. 2. Error bars indicate SD.
(E) DAPK1 expression in Raji cells treated with
0.5 mM decitabine. RT-PCR was performed for
DAPK1 and GAPDH on untreated and treated
cell lines. Jurkat cells were used as a positive
(F) Bisulfite sequences in BS1 region (c.1-
1509–c.1-1262) for untreated, 6 and 12 days
decitabine-treated (0.5 mM) Raji cells. Each
row represents a clone. The open circles
indicate unmethylated CpG, and closed circles
indicate methylated CpG dinucleotides. The
overall methylation frequency is given in per-
centage. Error bars indicate SD.
Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc. 881
5-fold reduced in the methylated construct, further aug-
menting the finding that DAPK1 promoter methylation
results in gene silencing (Figure 2D).
9, and 12 days. RT-PCR for DAPK1 indicated that Decita-
bine treatment resulted in gradual upregulation of DAPK1
expression, while untreated Raji cells did not show any
detectable expression (Figure 2E). Bisulfite sequencing of
the BS1 region showed that 116 of the tested 210 CpG
dinucleotides (55%) were methylated in untreated Raji
cells. Treatment with Decitabine resulted in significantly
(p < 0.001) reduced DNA methylation with 29% (70/240)
on day 6 and 25% (68/270) on day 12 (Figure 2F).
DAPK1 Regulates Apoptosis in a Lymphoid Cell Line
We stably transfected Jurkat cells with DAPK1 siRNA that
resulted in its downregulation (Figure 3A). Cells trans-
fected with vector alone or DAPK1 siRNA were treated
with activating anti-Fas antibody, and apoptosis was
examined using annexin-V-FITC/Propidium Iodide (an-
nexin/PI) flow cytometry. Jurkat cells with inhibited
DAPK1 expression showed a statistically significant
increase in resistance to apoptosis compared to cells
transfected withvectoralone (p<0.001;Figure3B).These
data demonstrated that DAPK1 is involved in Fas-induced
extrinsic apoptosis in lymphoid cells. Next, we incubated
5 3 107cells from nine CLL patients with or without Fas-
activating antibody (50 or 100 ng/ml) for 24 hr and exam-
ined apoptosis by annexin/PI flow cytometry (Table S1).
No significant effect on the percentage of nonapoptotic
cells relative to the untreated control was seen (data not
shown). Immunoblot analysis to assess potential changes
in p53 expression following treatment showed no detect-
able differences (Figure 3C). In conclusion, these experi-
ments are in line with our current knowledge that DAPK1
is a mediator of Fas-induced apoptotic signaling and
(Cohen et al., 1999; Raveh et al., 2001).
A CLL Family with Linkage to Chromosome 9
extended family in which a father and four sons were
diagnosed with CLL (Lynch et al., 2002). We have since
identified additional family members, both affected and
unaffected (Figure 4A). Genome-wide linkage analysis
with a panel of 400 microsatellite markers was performed
using samples from individuals III-2, III-3, III-4, IV-3, IV-4,
and IV-5. This identified a region on chromosome 9 be-
tween markers D9S175–D9S1776 with the highest non-
parametric linkage (NPL) score of 0.96. The next highest
scores were less than 0.42. High-resolution genotyping
identified a common haplotype of 707 kb in all affected
family members for whom samples were available. The
segment of the presumed CLL haplotype includes three
known genes (DAPK1, CTSL, and CCRK) and 11 pre-
dicted genes. Based on the epigenetic data indicating fre-
quent loss of DAPK1 expression, we hypothesized that
DAPK1 might be the predisposing gene mutated in this
family. All exons and splice sites of DAPK1 were se-
quenced from DNA of the skin fibroblasts of four affected
(III-2, III-3, III-4, and IV-5) as well as two unaffected family
members (IV-3 and IV-4) and compared to the genomic
sequence (NM_004938). Although five DNA variants
were detected (one in exon 3, two in exon 16, and two in
exon 26), none was unique for the CLL haplotype. These
variants were useful in RT-PCR analysis of DAPK1 and
showed that gene expression was highly reduced in the
CLL allele of patients III-4 and III-3, whereas both wild-
type (WT) alleles were equally expressed in unaffected in-
dividuals IV-3 and IV-4 (Figure 4B).
Allelic Expression Imbalance in DAPK1 in Affected
As a next step we developed monochromosomal mouse-
human hybrid clones containing either the WT (one clone)
or the CLL (two clones) chromosome 9 from patient III-4.
Semiquantitative RT-PCR (Figure 4C) and immunoblotting
Figure 3. DAPK1 Regulates Apoptosis in
(A) DAPK1 expression in Jurkat cells stably
transfected with either vector alone, DAPK1
siRNA-A, or DAPK1 siRNA-C as measured by
western blot. Tubulin expression served as
(B) Percent live cells were measured in Jurkat
cells stably transfected with vector alone or
DAPK1 siRNA-C, treated with activating-Fas
antibody (100 ng/ml). After 16 hr, cells were
harvested and suspended in binding buffer
with annexin V-FITC and propidium iodide, fol-
lowed by flow cytometry to assess cell death.
Error bars indicate SD.
(C) p53 expression in cells treated with either
no, 50 ng/ml, or 100 ng/ml Fas-activating
882 Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc.
(Figure 4D) showed reduced DAPK1 expression from the
two clones containing the CLL chromosome (25% or
less) when compared to the clones containing the WT
chromosome (100%). Reduced DAPK1 expression from
the CLL allele was further confirmed in diploid cells. We
cloned RT-PCR products containing exon 16 comprising
the informative single nucleotide polymorphism (SNP)
c.1510A > G from an unaffected (IV-3) and an affected
(III-4) family member. Individual clones were genotyped
by PCR using allele-specific primers. In individual IV-3,
the ratio of A:G clones was close to the expected ratio
of 1:1 (Figure 4E), while in individual III-4, the number of
A-clones (WT allele) was approximately four times that of
the G clones (CLL allele; Figure 4F), and this difference
was statistically significant (p < 0.01).
Detection of a Rare Germline Mutation by DAPK1
In search of a germline mutation in DAPK1 regulatory se-
quences, we extended our genomic sequencing efforts.
BAC clones derived from both the affected and the
unaffected alleles of family member III-3 were generated,
and two overlapping BAC clones for the CLL (CR956620
and CR956432) and the WT allele (CT009543 and
imately 400 kbp were sequenced for each allele, extend-
ing from 45 kb upstream to 100 kb downstream of
DAPK1, covering the entire gene. No major rearrange-
ments were detected in the sequence. In total, 281 single
nucleotide differences were observed between the two
alleles, out of which 162 were reported SNPs and 87
were located within repeat elements. We further elimi-
nated 28 of the remaining 32 SNPs as candidate muta-
tions, since they occurred in additional controls tested.
One SNP seen in the CLL allele, c.1-6531A > G, was not
found among 383 control samples from the US (n = 281)
and from Northern Europe (n = 102). Screening of 263
CLL cases from the US (n = 129) and from Northern
Europe (n = 134) identified one additional CLL sample
from Scandinavia with SNP c.1-6531A > G. This Scandi-
navian sample (no. 102) showed the presence of this
SNP in purified CLL cells and in T cells.
We further investigated samples from 75 CLL patients
with one affected relative in order to determine if DAPK1
is a target for genetic mutations. We amplified and
sequenced all exonic sequences (including splice sites)
and the c.1-6531 suppressor region. As control samples
we included DNAs from Wac3CD5 and two healthy
Figure 4. DAPK1 Expression in CLL
and square represent unaffected female and
male, respectively, while closed circle and
square represent affected female and male.
(B) Sequencing of SNP c.114A > G from
genomic DNA and cDNA of two unaffected
(IV-3 and IV-4) and two affected (III-3 and
(C) RT-PCR on RNA isolated from monochro-
mosomal hybrid clones with either WT or the
CLL chromosome 9 from fibroblast cells of
individual III-4 is shown. RPL4 expression
was used as an internal control. Error bars indi-
(D) DAPK1 protein expression from WT and
CLL alleles in monochromosomal hybrids.
Jurkat and WaC3CD5 cells were used as posi-
tive controls, and NIH3T3 cells were used as
the negative control to show specificity to
(E and F) The RT-PCR product amplifying SNP
(III-4; F) fibroblast cell lines was cloned, and
individual clones were genotyped. Shown is
the percentage of A and G clones in IV-3 and
III-4, and n is the number of clones studied.
The difference in allelic expression of DAPK1
in III-4 and IV-3 was statistically significant
(p < 0.01).
Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc. 883
donors. We identified numerous SNPs in exons 1 (non-
coding), 3, 4, 16, 18, 19, and 26; some of them had
already been reported as common polymorphisms
(Table S2). We were unable to identify a missense muta-
tion, a splice-site mutation, or an additional case with
c.1-6531A > G.
HOXB7 Represses DAPK1 Transcription
To study the effect of c.1-6531A > G on DAPK1 transcrip-
tion, corresponding CLL and WT luciferase reporter con-
structs were designed (Figure 5A). A 357 bp fragment
including either c.1-6531A (DAP-A) or c.1-6531G (DAP-G)
was ligated upstream into luciferase construct no. 1 con-
taining the DAPK1 promoter. Transcription from the DAP-
A and DAP-G construct was suppressed by 39% and
70%, respectively, compared to control construct no. 1
(Figure 5B). This suggested that both sequences contain
a suppressor element; however, the CLL-derived se-
quence, (DAP-G), displayed stronger effects.
To identify the potential suppressor molecule that binds
was performed using oligonucleotides with either the WT
oligo (A) or the CLL oligo (G). Using nuclear extracts from
Jurkat cells, multiple bands were observed with both oli-
gos (I–V; Figure 5C). Competition assays with unlabeled
oligo confirmed specificity (Figure 5D). The specificity of
binding was further illustrated by mutating one base on
either side of the A or G within the respective oligos.
sulted in elimination of band V, the mutation in the CLL
oligo resulted in elimination of bands IV and V. This result
showed that protein binding at band IV is affected by
iments that the intensity of band IV with the CLL oligo was
stronger than with the WT oligo, hinting at a differential
binding strength (Figures 5C–5E). Transcription factor
binding-site prediction suggested that HOX family pro-
teins might have differential affinity to A or G at the c.1-
6531 SNP. In supershift assays using two HOX family
proteins (HOXB7 and MSX2), only the HOXB7 antibody
induced a clear shift, suggesting that HOXB7, or another
Figure 5. An A to G Change at c.1-6531
bp Regulates DAPK1 Expression in CLL
G was ligated upstream to DAPK1 promoter
(c.1-2215–c.1-1151) luciferase construct with
either A (DAP-A) or G (DAP-G) as SNP.
(B) The DAPK1 promoter alone (no. 1), DAP-A,
or DAP-G constructs were transfected into
Jurkat cells, and the luciferase activity was
measured. Error bars indicate SD.
lyzed by EMSA assay using WT or CLL oligo.
Five specific bands (I–V) are marked.
(D) Ten- or fifty-fold molar excess concentra-
tions of cold oligos were used for competition
assays. An unlabeled Oct-1 oligo was used as
a negative control.
mutant, CLL, and CLL mutant oligos. In mutant
oligos, the adjacent bases to the c.1-6531 SNP
(F) For the supershift assay, antibodies against
HOXB7, USF2, and MSX2 were added to the
using the CLL oligo.
884 Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc.
closely related HOX protein, interacts with this site
Next, we tested DAPK1 as well as HOXB7 expression in
B cells, T cells, and granulocytes of three healthy donors
was highest in normal B cells, and about one-fifth or one-
third of the expression is seen in T cells and in granulo-
cytes, respectively (Figure S2). As shown above, expres-
sion of DAPK1 was further reduced (or not detectable) in
selected CLL cells. HOXB7 was expressed equally in B
and T cells, however much reduced in granulocytes. Inter-
estingly, HOXB7 was variably expressed in selected
CD19+ CLL cells with several samples showing increased
expression (Figure S2).
To explore whether HOXB7 affects DAPK1 expression,
we transfected skin fibroblasts from an affected (III-4) and
an unaffected family member (IV-4) with HOXB7 siRNA.
Quantitative RT-PCR showed successful HOXB7 downre-
gulation in both samples on day 1 following transfection,
while scrambled siRNA did not show this downregulation
(Figures 6A and 6B). DAPK1 upregulation in HOXB7
siRNA-transfected fibroblasts was observed in both III-4
and IV-4, showing that HOXB7 is a repressor of DAPK1
expression (Figures 6Cand 6D). In addition, allele-specific
expression analysiswas studiedin skinfibroblast cellsbe-
fore and after siRNA HOXB7 transfection in family mem-
bers III-4 and IV-4. RT-PCR products (days 0 and 2) com-
prising informative SNP c.1510A > G were cloned, and
ber III-4, the increase in DAPK1 expression from the CLL
allele was significantly higher than in the WT allele (p <
0.01), while the relative ratio between the two DAPK1 al-
leles in unaffected family member IV-4 remained un-
changed (p = 0.4) (Figures 6E and F6). These results
show that DAPK1 expression is downregulated by
HOXB7 and suggest that c.1-6531A > G increases the af-
finity of HOXB7 binding, resulting in stronger repression of
DAPK1 from the CLL allele.
DAPK1 Promoter Methylation in CLL Cells
of Affected Family Members
Our results indicated that expression of the CLL allele was
reduced to 25% compared to the WT allele in affected
Expression by HOXB7
(A and B) Fibroblast cell lines from affected
(III-4) and unaffected (IV-4) family members
were transfected with 60 nM of HOXB7 siRNA
or scrambled siRNA, and HOXB7 expression
was studied at different time points by semi-
quantitative SYBR green RT-PCR. HOXB7 ex-
pression in untreated cells was set as 1.0. Error
bars indicate SD. (C and D) DAPK1 expression
was studied in the fibroblast cell lines trans-
fected with HOXB7 siRNA or scrambled siRNA
for different time points by quantitative SYBR
green RT-PCR.DAPK1 expression inuntreated
cells was set as 1.0. Error bars indicate SD.
an affected (E) and unaffected (F) family mem-
ber was studied in fibroblasts from III-4 and
IV-4, before and after transfection of HOXB7
Shown is the percentage of the number of
clones from the two alleles, and (n) indicates
number of clones studied.
Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc. 885
family members. Thus, total DAPK1 expression was re-
duced to approximately 62.5% of that of normal levels.
We hypothesized that additional DAPK1 repression
occurred due to promoter methylation. Indeed, DAPK1
expression in PBMCs of five affected family members
was remarkably downregulated when compared to three
normal CD19+ cell controls (Figure 7A). Selected CD19+
B cells from blood of unaffected family members were
not available as a control. Bisulfite sequencing of PBMC
DNA from affected (III-1 and III-4) and unaffected family
members (IV-1) for the BS1 and BS2 regions showed
that both regions were highly methylated in the two
affected family members (Figure 7B). Altogether, these
data suggestthatDAPK1 expressionin theaffected family
members is significantly reduced by a combination of
epigenetic and genetic aberrations.
We identified downregulaion of DAPK1 expression as
a mechanism predisposing to CLL in a large family with
seven individuals. A single nucleotide change enhances
the binding affinity of transcription factor HOXB7 and
results in downregulation of DAPK1 transcription. Impor-
tantly, DAPK1 is not only a target in familial CLL but is
also inactivated in the majority of sporadic cases of CLL
by epigenetic mechanisms. Interestingly, it has long
been known that some high-penetrance genes (e.g.,
MLH1, BRCA1, p16) identified in familial cancers are
also frequently silenced by epigenetic mechanisms in
sporadic cancers (Esteller, 2002).
ulin-dependent, serine/threonine kinase that promotes
Figure 7. DAPK1 Expression and Promoter Methylation in CLL Cells of Family No. 4532
(A) RT-PCR for DAPK1 with GAPDH as an internal control on RNA extracted from blood cells of no. 4532 family members III-1, III-2, III-3, III-4 and IV-5
and from selected CD19+ normal B-cells from 3 healthy volunteers. Error bars indicate SD.
(B) DNA methylation analysis in CLL cells from CD19+ selected control B-cells, total PBMCs from one unaffected family member (IV-1) and two
affected individuals (III-1 and III-4). Bisulfite treated DNA was amplified for the BS1 and BS2 regions. Each row represents a clone. The open circles
indicate unmethylated and closed circles indicate methylated CpGs.
886 Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc.
apoptosis in response to various stimuli, including Fas,
INF-g, and TNF-a (Bialik and Kimchi, 2006). Increased
ations and cell morphology changes (Cohen et al., 1997).
The DAPK1 promoter contains TGF-b response elements
as well as p53-binding sites (Martoriati et al., 2005).
regulate each other’s expression. DAPK1 suppresses
cMYC- and E2F-induced cell transformation by activating
p19ARF/p53-dependent apoptosis and also blocks tumor
metastasis in vivo (Inbal et al., 1997; Raveh et al., 2001).
DAPK1 inhibits extracellular signal-regulated kinase (ERK)
activity, counteracts its survival signal (Chen et al., 2005),
The observed frequent silencing of DAPK1 by promoter
methylation in CLL could be one of the events required by
the leukemic cells to escape cell death mediated by either
the intrinsic or extrinsic pathways of apoptosis. Downre-
gulation of DAPK1 by transfecting siRNA in Jurkat cells,
an acute T cell leukemia cell line, resulted in reduced
susceptibility to Fas-induced apoptosis. This suggests in-
volvement of DAPK1 in the extrinsic apoptosis pathway in
lymphocytes. Thus, it is reasonable to speculate that
DAPK1 silencing may be an additional mechanism con-
tributing to malignant progression by conferring resis-
tance to apoptosis.
Past efforts have identified numerous affected genes in
deletions, which are frequently observed in sporadic CLLs
(Stilgenbauer and Dohner, 2005). For example, deletion of
13q14 is commonly seen in up to 50% of sporadic CLL
(Dohneretal.,2000).Several candidate tumor-suppressor
genes from this region have been proposed, including
miR15 and miR16, two microRNAs targeting the onco-
gene BCL2 (Bullrich et al., 2001; Cimmino et al., 2005;
Mertens et al., 2002).
Mapping efforts in CLL families have been relatively
unsuccessful in identifying genes predisposing to CLL.
Several candidate genes and predisposing polymor-
phisms have been proposed, including ARLTS1, a gene
that resides on chromosome 13q14 (Calin et al., 2005)
and P2RX7, located on chromosome 12q24 (Wiley et al.,
2002). However, this predisposition was not supported
in follow-up studies (Dao-Ung et al., 2004; Sellick et al.,
2006; Thunberg et al., 2002).
How can these previous linkage results be reconciled
with our findings regarding the role of DAPK1? Heritable
germline mutations of DAPK1 itself may not underlie the
disease in many of the larger CLL families that contributed
to the previous linkage studies. The existence of multiple
linkage peaks, none or few of which are statistically signif-
icant, suggests considerable locus heterogeneity or the
existence of multiple mutated genes as recently proposed
in an association study using nonsynonymous SNPs in
candidate genes (Caporaso, 2006; Rudd et al., 2006).
Since our results implicate downregulation of DAPK1 not
only as the primary susceptibility factor in a large family,
but also as a major event in sporadic CLL, it is quite pos-
sible that trans-acting factors, such as RNA genes (Calin
and Croce, 2006; Esquela-Kerscher and Slack, 2006),
play roles in the deregulation of DAPK1, and other genes,
The occurrence of reduced DAPK1 expression in com-
bination with frequent promoter methylation in tumor cells
of familial CLL is intriguing. It is possible that reduced
expression itself becomes a trigger for DAPK1 promoter
methylation, which is initiated by transcriptional silencing
(or reduced expression), followed by chromatin conden-
sation and histone tail modifications, and finally by DNA
methylation. This sequence of gene silencing events has
been shown for target genes in the estrogen receptor sig-
naling pathway in breast cancer cells (Leu et al., 2004) and
for glutathione S-transferase in prostate cancer cells
(Stirzaker et al., 2004). To explain the frequent occurrence
of promoter methylation in sporadic CLL, one may postu-
late DAPK1 silencing due to modulation of upstream sig-
nals. One of these signals, HOXB7, was identified in this
work. HOXB7 is a homeobox-containing transcription
factor mediating a variety of developmental processes, in-
cluding hematopoietic differentiation and lymphoid devel-
opment (Lill et al., 1995; Shen et al., 1989). A scenario that
may explain frequent targeting of the DAPK1 promoter by
aberrant DNA methylation is the aberrant recruitment of
DNA methyltransferase activity to the DAPK1 promoter
by oncogenic proteins (Brenner et al., 2005; Di Croce
et al., 2002).
In conclusion, we have identified DAPK1 as a novel pu-
tative tumor suppressor gene in CLL. The combination of
downregulation of DAPK1 is common in CLL cells. The
frequent silencing in sporadic CLL suggests that aberrant
silencing of DAPK1 is an early and required event in
leukemogenesis. Screening of familial CLL cases for
of CLL, which is a late onset disease. In addition, as DNA
methylation events are reversible, our finding provides
hope for the development of novel treatment regimens in
CLL involving epigenetic therapies for gene reactivation.
Patient Selection, Sample Collection, and Cell Lines
Blood and fibroblast cell lines were obtained from patients with B cell
CLL through the CLL Research Consortium (CRC), from the Ohio State
University (OSU), Creighton University (CU), and Uppsala University
Hospital tissue banks. CLL patients and unaffected individuals from
OSU, CU, and CRC provided written informed consent using Institu-
tional Review Board-approved protocols. Written consent for Uppsala
samples was provided according to the declaration of Helsinki. Seven
CD19-selected sporadic CLL samples were obtained by positive
selection using magnetic beads conjugated to anti-CD19 (MACS,
Miltenyi Biotec, Auburn, CA). The resulting cells are at least 95%
CD19 positive. All patients had immunophenotypically defined CLL
as outlined by the modified NCI criteria (Cheson et al., 1996) and for
Uppsala samples, according to the WHO classification. Cell lines
Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc. 887
WaC3CD5, Jurkat, Raji, lymphoblastoid cell lines generated from III-4,
and skin fibroblasts from family 4532 were maintained in culture.
Quantitative DNA Methylation
Quantitative high-throughput DNA methylation analysis was done by
MassARRAY system as described elsewhere (Ehrich et al., 2005).
Four bisulfite reactions (A1–A4) were designed, which covered 17,
20, 22, and 29 CpGs, respectively, and extend from c.1-1573 to c.1-
239 bp. The primer sequences are available upon request. For fol-
low-up methylation studies on selected samples, bisulfite sequencing
was performed as described earlier (Rush et al., 2004). Two bisulfite
reactions covering regions c.1-1509–c.1-1262 (BS1) and c.1-1281–
c.1-903 (BS2) were amplified.
Semiquantitative Reverse Transcriptase PCR (RT-PCR)
RT-PCR was performed and analyzed as described previously (Raval
et al., 2005) using SUPERSCRIPT First-Strand Synthesis kit (Invitro-
gen). SYBRgreen PCR was done in triplicates with IQ SYBR Green
Supermix (Bio-Rad, Hercules, CA).
Development of Monochromosomal Mouse-Human Hybrid
Lymphoblastoid cells from samples III-4 was used to generate mono-
allelic clones as described previously (Papadopoulos et al., 1995).
Genotyping confirmed one clone with WT and two clones with CLL
Expression Vectors and Luciferase Reporter Assay
A modified pGL3 basic vector (Yu et al., 2004) was used to ligate the
PCR-amplified DAPK1 promoter constructs. The SV40 promoter re-
porter construct served as positive control. Monochromosomal
mouse-human hybrid cells were used to amplify SNP c.1-6531A > G
ina357bpfragmentwitheithertheA ortheGallele. ThePCRproducts
were ligated into the pGL3-promoter vector at the XhoI site, and the
SV40 promoter was replaced by DAPK1 promoter region (c.1-2215–
c.1-1151 bp) at the BglII and HindIII sites to create DAP-G or DAP-A
constructs. Constructs were confirmed by sequencing. For transfec-
tion into Jurkat cells, a density of 1.5 3 105cells/well was used in
a 24-well plate with no serum for 2 hr. After 2 hr 1 mg plasmid pGL3
vector and 20 ng of pRL-TK internal control vector (Promega) were co-
PA).After 4hr, RPMI medium with 10%FCS was added to each well to
make up the volume to 1 ml. The cells were further incubated for 48 hr,
and the luciferase assay was performed according to manufacturer’s
instructions (Promega). Luciferase activity was normalized using
pRL-TK activity. Each experiment was performed in triplicate. The
in vitro methylation assay was performed as recently described
(Yu et al., 2005).
For stable transfection of DAPK1 siRNA into Jurkat cells, the pRS vec-
tor with different DAPK1 siRNA inserts (siRNA A and C) from OriGene,
(Rockville, MD) were used. Ten micrograms of pRS-DAPK1 siRNA A,
siRNA C, or pRS vector alone were transfected into the amphotropic
Pheonix packaging cell line (60% confluent) using Superfect (Qiagen).
Virus-containing medium was collected from the Pheonix cells after 48
hr,and celldebris wasremovedbycentrifugation.Onemilliliterof fresh
medium plus 1 ml of infectious medium, containing either pRS-DAPK
siRNA A, siRNA C, or pRS vector alone were added to 2.5 3 105Jurkat
cells/well, and the 6-well plate was centrifuged at 2300 rpm for 90 min
at room temperature in a Beckman centrifuge GPH, Rotor 3.7. Follow-
ing thecentrifugationstep,2mloffreshmedium wasadded,and, 48hr
after incubation in 5% CO2, cells were selected for puromycin resis-
tance using medium supplemented with 2 mg/ml Puromycin (Sigma,
St Louis, MO). For transient transfection of HOXB7 siRNA (Ambion,
Austin, TX) into fibroblast cells, 2 3 105cells were plated, and 24 hr
later cells were transfected using polyethylenimine (Polysciences,
Warrington, PA), as described before.
Apoptosis and Flow Cytometric Studies
Jurkat cells stably transfected with pRS vector control or pRS-DAPK
siRNA C weretreated with100 ng/ml anti-Fas (human activating, clone
CH11)antibody (Upstate Biotechnology, LakePlacid,NY)for 16hrand
resuspended in binding buffer containing annexin V-fluorescein iso-
thiocyanate (FITC) and propidium iodide according to the supplier’s
instuctions (BD Biosciences, San Diego, CA), and assessed by flow
cytometry using a Beckman-Coulter model EPICS XL cytometer
(Beckman-Coulter, Miami, FL). Each sample was run in triplicate.
Immunoblotting was performed by transferring proteins to a nitrocellu-
lose membrane (Hybond-ECL, Amersham Biosciences, Germany).
Following the DAPK1 (Sigma-Aldrich, St. Louis, MO) and a-tubulin
antibody (Oncogene,Boston,MA)staining, theproteinsweredetected
with chemiluminescent substrate (SuperSignal, Pierce, Rockford, IL).
SSCP Analysis and Genotyping of Alleles by Colony PCR
PCR-amplified fragments were analyzed by single-strand conforma-
tion polymorphism (SSCP) as previously described (Liechti-Gallati
et al., 1999). Variant bands were reamplified and used for direct
sequencing on ABI Prism 3730 DNA analyzer (Applied Biosystems).
For genotyping of the alleles for c.1-1510A > G SNP, cDNA was ampli-
fied and cloned. Colony PCR was performed in each samples by using
primers specific to either ‘‘A’’ or the ‘‘G’’ SNP.
Electrophoretic Mobility Shift Assay (EMSA)
The oligonucleotides used for EMSA were WT oligo 50-cttgccttggtc-
gtgattacctacagatgcctgaat-30, WT oligo mutant 50-cttgccttggtcgtaac-
tacctacagatgcctgaat-30, CLL oligo 50-cttgccttggtcgtggttacctacagat-
gcctgaat-30, and CLL oligo mutant 50-cttgccttggtcgtagctacctacag
atgcctgaat-30. The double-stranded oligonucleotides were end-
labeled with [g-32P]ATP using T4 polynucleotide kinase enzyme
(NEB). The free probe was removed by purification in G50 Sephadex
spin columns. The binding reaction was conducted at room tempera-
radiolabeled oligonucleotide probe, in 1x Ficoll buffer (10 mM Tris (pH
7.5), 1 mM DTT, 1 mM EDTA, and 4% Ficoll), 250 ng of poly(deoxyino-
sinic-deoxycytidylic acid) in 75 mM KCl, and double-distilled H2O to
make the volume to 15 ml. For supershift assay, 1 mg each of
HOXB7, MSX2 (Cemines, Golden, CO), and USF2 (Santa Cruz Bio-
technologies, Santa Cruz, CA) was added, and the mixture was incu-
bated atroom temperaturefor anadditional30min. DNA-protein com-
plexes were fractionated by electrophoresis in 6% nondenaturating
polyacrylamide gel, and radioactivity in the gels was visualized by au-
toradiography analyzed using the STORM860 image analyzer (Amer-
sham Biosciences, NJ). Nuclear extracts from Jurkat cells were
made as described previously (Frissora et al., 2003).
Wilcoxon rank sum test was performed to compare the percent
DAPK1 promoter methylation between CLL samples and normal con-
trols and also to compare DAPK1 expression in 50 unselected CLL
samples with that of CD19+ normal B cells. To obtain proportions for
the number of ‘‘A’’ or ‘‘G’’ clones in unaffected and affected family
members and for the methylated and unmethylated CpGs within
DAPK1 promoter in Raji cells before and after decitabine treatment,
the values were compared using the Z test (http://faculty.vassar.edu/
with this article online at http://www.cell.com/cgi/content/full/129/5/
888 Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc.
The authors wish to thank all members of the Plass and Byrd labs for
critical discussions; Mattias Jansson from Uppsala University: and
the Sequencing Division at the Wellcome Trust Sanger Institute for ex-
cellent technical assistance: Ramana Davuluri, Sandya Liyanarachchi,
and Greg Singer for statistical support; Nickolas Papadopoulos for the
conversion to haploidy; and Paivi Lahermo of the Finnish Genome
Center for genotyping and linkage analyses. A.R. is supported by
a T32 CA106196 fellowship in Cancer Genetics. B.H. is supported by
a grant from the Dr. Mildred Scheel Foundation for Cancer Research,
Germany. This publication was supported by National Cancer Institute
grants CA110496 (J.C.B., C.P., and A.R.), CA101956 (C.P. and J.C.B.)
CA81534 to the CLL Research Consortium (J.C.B., M.G., and T.J.K.),
P30 CA16058 (A.d.l.C., C.P., and J.C.B.), The Leukemia and Lym-
phoma Society of America (J.C.B. and C.P.), and the D. Warren Brown
Foundation (J.C.B. and C.P.). Further supportcame from revenue from
Nebraska cigarette taxes awarded to Creighton University by the Ne-
braska Department of Health and Human Services (NDHHS). Its con-
tents are solely the responsibility of the authors and do not necessarily
represent the official views of the state of Nebraska or NDHHS. Sup-
port was also received by NIH grant 5U01 CA86389 and the Swedish
Cancer Society. C.P. is a Leukemia and Lymphoma Society scholar,
and J.C.B. is a Leukemia and Lymphoma Society clinical scholar.
P.C. and S.B. were supported by the Wellcome Trust.
Received: September 20, 2006
Revised: January 15, 2007
Accepted: March 12, 2007
Published: May 31, 2007
Amundadottir, L.T., Thorvaldsson, S., Gudbjartsson, D.F., Sulem, P.,
Kristjansson, K., Arnason, S., Gulcher, J.R., Bjornsson, J., Kong, A.,
Thorsteinsdottir, U., and Stefansson, K. (2004). Cancer as a complex
phenotype: pattern of cancer distribution within and beyond the nu-
clear family. PLoS Med. 1, e65.
Structure, Function, and Beyond. Annu. Rev. Biochem. 75, 189–200.
Brenner, C., Deplus, R., Didelot, C., Loriot, A., Vire, E., De Smet, C.,
Gutierrez, A., Danovi, D., Bernard, D., Boon, T., et al. (2005). Myc
represses transcription through recruitment of DNA methyltransferase
corepressor. EMBO J. 24, 336–346.
Bullrich, F., Fujii, H., Calin, G., Mabuchi, H., Negrini, M., Pekarsky, Y.,
Rassenti, L., Alder, H., Reed, J.C., Keating, M.J., et al. (2001). Charac-
terization of the 13q14 tumor suppressor locus in CLL: identification of
ALT1, an alternative splice variant of the LEU2 gene. Cancer Res. 61,
Calin, G.A., and Croce, C.M. (2006). Genomics of chronic lymphocytic
leukemia microRNAs as new players with clinical significance. Semin.
Oncol. 33, 167–173.
Calin, G.A., Trapasso, F., Shimizu, M., Dumitru, C.D., Yendamuri, S.,
Godwin, A.K., Ferracin, M., Bernardi, G., Chatterjee, D., Baldassarre,
G., et al. (2005). Familial cancer associated with a polymorphism in
ARLTS1. N. Engl. J. Med. 352, 1667–1676.
Caporaso, N. (2006). Chips, candidate genes, and CLL. Blood 108,
Chen, C.H., Wang, W.J., Kuo, J.C., Tsai, H.C., Lin, J.R., Chang, Z.F.,
and Chen, R.H. (2005). Bidirectional signals transduced by DAPK-
ERK interaction promote the apoptotic effect of DAPK. EMBO J. 24,
Cheson, B.D., Bennett, J.M., Grever, M., Kay, N., Keating, M.J.,
O’Brien, S., and Rai, K.R. (1996). National Cancer Institute-sponsored
Working Group guidelines for chronic lymphocytic leukemia: revised
guidelines for diagnosis and treatment. Blood 87, 4990–4997.
quent DAP kinase but not p14 or Apaf-1 hypermethylation in B-cell
chronic lymphocytic leukemia. J. Hum. Genet. 51, 832–838.
Cimmino, A., Calin, G.A., Fabbri, M., Iorio, M.V., Ferracin, M., Shimizu,
M., Wojcik, S.E., Aqeilan, R.I., Zupo, S., Dono, M., et al. (2005). miR-15
andmiR-16induceapoptosis bytargeting BCL2.Proc.Natl.Acad.Sci.
USA 102, 13944–13949.
Cohen, O., Feinstein, E., and Kimchi, A. (1997). DAP-kinase is a Ca2+/
calmodulin-dependent, cytoskeletal-associated protein kinase, with
cell death-inducing functions that depend on its catalytic activity.
EMBO J. 16, 998–1008.
Cohen, O., Inbal, B., Kissil, J.L., Raveh, T., Berissi, H., Spivak-
Kroizaman, T., Feinstein, E., and Kimchi, A. (1999). DAP-kinase partic-
ipates in TNF-alpha- and Fas-induced apoptosis and its function
requires the death domain. J. Cell Biol. 146, 141–148.
Corcoran, M., Parker, A., Orchard, J., Davis, Z., Wirtz, M., Schmitz,
O.J., and Oscier, D. (2005). ZAP-70 methylation status is associated
with ZAP-70 expression status in chronic lymphocytic leukemia. Hae-
matologica 90, 1078–1088.
Dao-Ung, L.P., Fuller, S.J., Sluyter, R., Skarratt, K.K., Thunberg, U.,
Tobin, G., Byth, K., Ban, M., Rosenquist, R., Stewart, G.J., and Wiley,
J.S. (2004). Association of the 1513C polymorphism in the P2X7 gene
with familial forms of chronic lymphocytic leukaemia. Br. J. Haematol.
Deiss, L.P., Feinstein, E., Berissi, H., Cohen, O., and Kimchi, A. (1995).
Identification of anovel serine/threonine kinase and anovel 15-kD pro-
tein as potential mediators of the gamma interferon-induced cell
death. Genes Dev. 9, 15–30.
Di Croce, L., Raker, V.A., Corsaro, M., Fazi, F., Fanelli, M., Faretta, M.,
Fuks, F., Lo Coco, F., Kouzarides, T., Nervi, C., et al. (2002). Methyl-
transferase recruitment and DNA hypermethylation of target pro-
moters by an oncogenic transcription factor. Science 295, 1079–1082.
Dohner, H., Stilgenbauer, S., Benner, A., Leupolt, E., Krober, A.,
Bullinger, L., Dohner, K., Bentz, M., and Lichter, P. (2000). Genomic
aberrations and survival in chronic lymphocytic leukemia. N. Engl. J.
Med. 343, 1910–1916.
Ehrich, M., Nelson, M.R., Stanssens, P., Zabeau, M., Liloglou, T.,
Xinarianos, G., Cantor, C.R., Field, J.K., and van den Boom, D.
(2005). Quantitative high-throughput analysis of DNA methylation pat-
terns by base-specific cleavage and mass spectrometry. Proc. Natl.
Acad. Sci. USA 102, 15785–15790.
Esquela-Kerscher, A., and Slack, F.J. (2006). Oncomirs - microRNAs
with a role in cancer. Nat. Rev. Cancer 6, 259–269.
Esteller, M. (2002). CpG island hypermethylation and tumor suppres-
sor genes: a booming present, a brighter future. Oncogene 21,
Frissora, F., Chen, H.C., Durbin, J., Bondada, S., and Muthusamy, N.
(2003). IFN-gamma-mediated inhibition of antigen receptor-induced
B cell proliferation and CREB-1 binding activity requires STAT-1 tran-
scription factor. Eur. J. Immunol. 33, 907–912.
Goldgar, D.E., Easton, D.F., Cannon-Albright, L.A., and Skolnick, M.H.
(1994). Systematic population-based assessment of cancer risk in
first-degree relatives of cancer probands. J. Natl. Cancer Inst. 86,
Goldin, L.R., Ishibe, N., Sgambati, M., Marti, G.E., Fontaine, L., Lee,
M.P., Kelley, J.M., Scherpbier, T., Buetow, K.H., and Caporaso, N.E.
(2003). A genome scan of 18 families with chronic lymphocytic leukae-
mia. Br. J. Haematol. 121, 866–873.
Goldin,L.R.,Pfeiffer, R.M.,Li, X.,andHemminki,K.(2004).Familialrisk
of lymphoproliferative tumors in families of patients with chronic
Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc. 889
lymphocytic leukemia: results from the Swedish Family-Cancer Data- Download full-text
base. Blood 104, 1850–1854.
Inbal, B., Cohen, O., Polak-Charcon, S., Kopolovic, J., Vadai, E.,
Eisenbach, L., and Kimchi, A. (1997). DAP kinase links the control of
apoptosis to metastasis. Nature 390, 180–184.
Jang, C.W., Chen, C.H., Chen, C.C., Chen, J.Y., Su, Y.H., and Chen,
R.H. (2002). TGF-beta induces apoptosis through Smad-mediated
expression of DAP-kinase. Nat. Cell Biol. 4, 51–58.
medicine, 16th edn (New York: McGraw-Hill, Medical Pub. Division).
Katzenellenbogen, R.A., Baylin, S.B., and Herman, J.G. (1999).
Hypermethylation of the DAP-kinase CpG island is a common alter-
ation in B-cell malignancies. Blood 93, 4347–4353.
Leu, Y.W., Yan, P.S., Fan, M., Jin, V.X., Liu, J.C., Curran, E.M.,
Welshons, W.V., Wei, S.H., Davuluri, R.V., Plass, C., et al. (2004).
Loss of estrogen receptor signaling triggers epigenetic silencing of
downstream targets in breast cancer. Cancer Res. 64, 8184–8192.
Liechti-Gallati, S., Schneider, V., Neeser, D., and Kraemer, R. (1999).
Two buffer PAGE system-based SSCP/HD analysis: a general proto-
col for rapid and sensitive mutation screening in cystic fibrosis and
any other human genetic disease. Eur. J. Hum. Genet. 7, 590–598.
Lill, M.C., Fuller, J.F., Herzig, R., Crooks, G.M., and Gasson, J.C.
(1995). The role of the homeobox gene, HOX B7, in human myelomo-
nocytic differentiation. Blood 85, 692–697.
Lynch, H.T., Weisenburger, D.D., Quinn-Laquer, B., Watson, P.,
Lynch, J.F., and Sanger, W.G. (2002). Hereditary chronic lymphocytic
leukemia: an extended family study and literature review. Am. J. Med.
Genet. 115, 113–117.
Martoriati, A., Doumont, G., Alcalay, M., Bellefroid, E., Pelicci, P.G.,
and Marine, J.C. (2005). dapk1, encoding an activator of a p19ARF-
p53-mediated apoptotic checkpoint, is a transcription target of p53.
Oncogene 24, 1461–1466.
Mertens, D., Wolf, S., Schroeter, P., Schaffner, C., Dohner, H.,
Stilgenbauer, S., and Lichter, P. (2002). Down-regulation of candidate
tumor suppressor genes within chromosome band 13q14.3 is inde-
pendent of the DNA methylation pattern in B-cell chronic lymphocytic
leukemia. Blood 99, 4116–4121.
Papadopoulos, N., Leach, F.S., Kinzler, K.W., and Vogelstein, B.
(1995). Monoallelic mutation analysis (MAMA) for identifying germline
mutations. Nat. Genet. 11, 99–102.
Young, D.C., Rassenti, L., Kipps, T.J., Grever, M.R., Byrd, J.C., and
Plass, C. (2005). TWIST2 Demonstrates Differential Methylation in
Immunoglobulin Variable Heavy Chain Mutated and Unmutated
Chronic Lymphocytic Leukemia. J. Clin. Oncol. 23, 3877–3885.
Raveh, T., Droguett, G., Horwitz, M.S., DePinho, R.A., and Kimchi, A.
(2001). DAP kinase activates a p19ARF/p53-mediated apoptotic
checkpoint to suppress oncogenic transformation. Nat. Cell Biol. 3,
Risch, N. (2001). The genetic epidemiology of cancer: interpreting
family and twin studies and their implications for molecular genetic
approaches. Cancer Epidemiol. Biomarkers Prev. 10, 733–741.
Rosenwald, A., Wright, G., Leroy, K., Yu, X., Gaulard, P., Gascoyne,
R.D., Chan, W.C., Zhao, T., Haioun, C., Greiner, T.C., et al. (2003).
Molecular diagnosis of primary mediastinal B cell lymphoma identifies
a clinically favorable subgroup of diffuse large B cell lymphoma related
to Hodgkin lymphoma. J. Exp. Med. 198, 851–862.
(2006). Variants in the ATM-BRCA2-CHEK2 axis predispose to chronic
lymphocytic leukemia. Blood 108, 638–644.
Rush, L.J., Raval, A., Funchain, P., Johnson, A.J., Smith, L., Lucas,
D.M., Bembea, M., Liu, T.H., Heerema, N.A., Rassenti, L., et al.
(2004). Epigenetic profiling in chronic lymphocytic leukemia reveals
novel methylation targets. Cancer Res. 64, 2424–2433.
Sellick, G.S., Webb, E.L., Allinson, R., Matutes, E., Dyer, M.J.,
Jonsson, V., Langerak, A.W., Mauro, F.R., Fuller, S., Wiley, J., et al.
(2005). A high-density SNP genomewide linkage scan for chronic
lymphocytic leukemia-susceptibility loci. Am. J. Hum. Genet. 77,
Sellick, G.S., Catovsky, D., and Houlston, R.S. (2006). Familial chronic
lymphocytic leukemia. Semin. Oncol. 33, 195–201.
Shen, W.F., Largman, C., Lowney, P., Hack, F.M., and Lawrence, H.J.
(1989). Expression of homeobox genes in human erythroleukemia
cells. Adv. Exp. Med. Biol. 271, 211–219.
Stilgenbauer, S., and Dohner, H. (2005). Genotypic prognostic
markers. Curr. Top. Microbiol. Immunol. 294, 147–164.
Stirzaker, C., Song, J.Z., Davidson, B., and Clark, S.J. (2004). Tran-
scriptional gene silencing promotes DNA hypermethylation through
asequential change in chromatin modifications incancer cells. Cancer
Res. 64, 3871–3877.
Thunberg, U., Tobin, G., Johnson, A., Soderberg, O., Padyukov, L.,
Hultdin, M., Klareskog, L., Enblad, G., Sundstrom, C., Roos, G., and
Rosenquist, R. (2002). Polymorphism in the P2X7 receptor gene and
survival in chronic lymphocytic leukaemia. Lancet 360, 1935–1939.
Wiley, J.S., Dao-Ung, L.P., Gu, B.J., Sluyter, R., Shemon, A.N., Li, C.,
Taper, J., Gallo, J., and Manoharan, A. (2002). A loss-of-function
polymorphic mutation in the cytolytic P2X7 receptor gene and chronic
lymphocytic leukaemia: a molecular study. Lancet 359, 1114–1119.
Yu, L., Liu, C., Bennett, K., Wu, Y.Z., Dai, Z., Vandeusen, J., Opavsky,
R., Raval, A., Trikha, P., Rodriguez, B., et al. (2004). A NotI-EcoRV pro-
moter library for studies of genetic and epigenetic alterations in mouse
models of human malignancies. Genomics 84, 647–660.
Yu, L., Liu, C., Vandeusen, J., Becknell, B., Dai, Z., Wu, Y.Z., Raval, A.,
Liu, T.H., Ding, W., Mao, C., et al. (2005). Global assessment of
promoter methylation in a mouse model of cancer identifies ID4 as
a putative tumor-suppressor gene in human leukemia. Nat. Genet.
Yuille, M.R., Matutes, E., Marossy, A., Hilditch, B., Catovsky, D., and
Houlston, R.S. (2000). Familial chronic lymphocytic leukaemia: a
890 Cell 129, 879–890, June 1, 2007 ª2007 Elsevier Inc.