psQ

MTa

b

c

d

e

f

g

h

a

ARRAA

KpMPMP

1

dbattpbDith

US

0d

Leukemia Research 34 (2010) 718–722

Contents lists available at ScienceDirect

Leukemia Research

journa l homepage: www.e lsev ier .com/ locate / leukres

15INK4b methylation correlates with thrombocytopenia, blast percentage, andurvival in myelodysplastic syndromes in a dose dependent manner:uantitation using pyrosequencing study

iyoung Kima, Bora Ohb,c, Song-yee Kima, Hyun-Kyung Parkd, Sang Mee Hwanga,ae Young Kimb,c, Cha Ja Shea, Inchul Yange, Sung Soo Yoonf, Jong Hyun Yoong, Dong Soon Leea,b,c,h,∗

Department of Laboratory Medicine, Seoul National University College of Medicine, 28 Youngon-Dong, Chongno-Gu, Seoul 110-744, South KoreaDepartment of Molecular and Clinical Oncology, Seoul National University College of Medicine, Seoul, South KoreaCancer Research Institute, Seoul National University College of Medicine, Seoul, South KoreaSeoul Clinical Laboratory, Seoul Medical Science Institute, Seoul, South KoreaOrganic and Bio-analysis Group, Korea Research Institute of Standards and Science, Daegeon, South KoreaDepartment of Internal Medicine, Seoul National University Hospital, Seoul, South KoreaDepartment of Laboratory Medicine, Seoul National University Boramae Hospital, Seoul, South KoreaDepartment of Tumor Biology, Seoul National University College of Medicine, Seoul, South Korea

r t i c l e i n f o

rticle history:eceived 29 June 2009eceived in revised form 2 September 2009ccepted 6 September 2009vailable online 25 September 2009

a b s t r a c t

We investigated how the quantity of p15INK4b methylation related to International Prognosic ScoringSystem variables and survival in 74 patients with de novo myelodysplastic syndrome (MDS). Pyrose-quencing of 11 consecutive CpG sites of the p15INK4b promotor region was performed, with the extent ofCpG cytosine methylation assessed in terms of methylation level (MtL). Patients with >5% bone marrow

eywords:15INK4bethylation

yrosequencingyelodysplastic syndromes

blasts had higher MtL than patients with <5% blasts (10.1% vs. 6.1%, p = 0.030, respectively). Methylationwas not associated with chromosomal aberrations. The MtL of patients with thrombocytopenia werehigher than patients without thrombocytopenia (11.2% vs. 6.2%, p = 0.036, respectively); they were higherin patients with cytopenias in ≥2 lineages than in patients with either unilineage or no cytopenia (9.8%vs. 4.1%, p = 0.036, respectively). The survival of patients with >7% MtL was worse than patients with <7%MtL (p = 0.031). Heavy p15INK4b methylation in MDS is associated with IPSS predictors of poor prognosis

rognosis and adverse survival.

. Introduction

The myelodysplastic syndromes (MDSs) are a group of disor-ers that differ in the degree of dysplasia, the extent of blasticone marrow involvement, and the rate of progression towardscute leukemia [1–3]. The inactivation by promotor hypermethyla-ion of tumor suppressor genes, particularly p15INK4b, is believedo contribute to the initiation and progression of MDS [2–5]. Therotein encoded by p15INK4b negatively regulates the cell cycley inhibiting cyclin-dependent kinases (CDKs) 4 and 6, and cyclin

-CDK4/6 complexes [2–4]. Transforming growth factor � (TGF�)

ncreases p15INK4b expression and downregulates CDK 4 synthesis,hereby inhibiting progression from G1 phase to S phase in normalematopoietic progenitors [2–4].

∗ Corresponding author at: Department of Laboratory Medicine, Seoul Nationalniversity College of Medicine, 28 Youngon-Dong, Chongno-Gu, Seoul 110-744,outh Korea. Tel.: +82 2 2072 3986; fax: +82 2 747 0359.

E-mail address: soonlee@plaza.snu.ac.kr (D.S. Lee).

145-2126/$ – see front matter. Crown Copyright © 2009 Published by Elsevier Ltd. All rioi:10.1016/j.leukres.2009.09.007

Crown Copyright © 2009 Published by Elsevier Ltd. All rights reserved.

Many reports have suggested that the extent to which thep15INK4b promotor is methylated helps to account for the hetero-geneity among MDSs. p15INK4b methylation was found to be morefrequent in high-risk MDS (RAEB) at the time of diagnosis [2,3],and to be associated with progression to AML [3,4]. Methylation ofp15INK4b also appears to be associated with chromosomal aber-rations, including monosomy and deletion 7q, in therapy-relatedMDS [5]. However, most of these studies used qualitative meth-ods such as the methylation-specific polymerase chain reaction(MSPCR) to analyze DNA methylation.

It is believed that methylation density correlates with transcrip-tional repression [6,7]. Indeed, the extent of methylation appearsto be a critical determinant of the degree to which p15INK4b issilenced. In the present study, we hypothesized that differencesin the extent of p15INK4b methylation contribute to the hetero-

geneity among MDSs. This was investigated in 74 patients withde novo MDS, and the methylation of 11 consecutive CpG sitesin the p15INK4b promotor region quantified using real-time DNAsequencing PyrosequencingTM technology [8–12]. We determinedthe extent to which methylation was related to clinical variables of

ghts reserved.

Resea

ts

2

2

opamt0Cm

2

2

cfoae

2

(biTp(rCbawfipU5s

2

Agowmo

r

2

h8cret[

2

2

sIcm17n2

3.1.2. KaryotypeOnly 56 patients had the data of cytogenetic study. The MtL data

for those 56 patients with or without known prognostic cytoge-

Table 1The quantity of p15INK4b methylation across bone marrow blast percentage groups.

Percent blasts No. ofcases

MtL (%)

Mean ± SD p

A: blasts in bone marrow < 5% 25 6.1 ± 6.4 A and B: 0.113B: 5% < blasts in bone marrow < 10% 26 10.4 ± 11.9 A and C: 0.029

M. Kim et al. / Leukemia

he International Prognostic Scoring System (IPSS) [13] and patienturvival.

. Materials and methods

.1. Patients

Data for the study came from 74 de novo MDS patients (53 males and 21 females)f Seoul National University Hospital between January 2000 and August 2008. Theatients were 20–86 years old (mean = 56.7 years). Bone marrow aspirates obtainedt the time of initial diagnosis were used, with the aspirates of 20 healthy bonearrow transplant donors serving as normal controls. The study was approved by

he Institutional Research Board of Seoul National University Hospital (IRB No. 0907-23-285). The KG1a (American Type Culture Collection [ATCC], Rockville, MD, No.CL246.1) and SNU484 (Korea Cell Line Bank, Seoul) cell lines were used as p15INK4bethylated and unmethylated controls, respectively.

.2. Pyrosequencing

.2.1. DNA extraction and sodium bisulfite modificationGenomic DNA was extracted from bone marrow mononuclear cells in control

ell lines with QIAamp DNA Blood Mini Kit (QIAGEN, CA, USA), as per the manu-acturer’s instructions. Bisulfite modification was undertaken with 1-�g samplesf DNA using DNA Methylation KitTM (Zymo Research, HiSS Diagnostics, CA, USA),s per the manufacturer’s instructions. The samples were then eluted in 15 �L oflution buffer and stored at −20 ◦C.

.2.2. Generation of PCR products and pyrosequencingWe pyrosequenced the first 63 base pairs of p15INK4b, from c.−318 to c.−256

NC 000009.10, NM.078487.2). This sequence was selected on the basis of the studyy Herman et al. [14], and included 11 CpGs used in that study. PCR and pyrosequenc-

ng primers were designed using Assay Design Software (Biotage, Uppsala, Sweden).he PCR volume was 50 �L, and incorporated 400 nmol/L of forward and reverserimers, 200 mmol/L of each dNTP, 2.5 mmol/L MgCl2, 1.25 units PlatinumTaqTM

Invitrogen, CA, USA), and 50–100 ng of bisulfite-modified DNA. The forward andeverse primer sequences were 5′-GGTTGGTTTTTTATTTTGTTAGA-3′ and 5′-biotin-CTAAATTACTTCTAAAAAAAAAC-3′ , respectively. Hot start PCR was performed. Thisegan at 95 ◦C for 15 min, followed by 45 cycles of denaturation at 95 ◦C for 20 s,nnealing at 51 ◦C for 20 s, and elongation at 72 ◦C for 20 s. The procedure finishedith the final elongation period extended to 5 min. PCR product quality was con-rmed on 1% agarose gels with ethidium bromide staining. Pyrosequencing waserformed with PyroGold SQATM Reagent Kit and PSQ96MA SystemTM (Biotage,ppsala, Sweden), as per the manufacturer’s instructions. The primer sequence was′-GGGGTAGTGAGGATTT-3′ . CpG methylation was quantified with PyroQ-CpGTM

oftware (Biotage).

.2.3. Analysis of pyrosequencing dataPyrosequencing data (pyrograms) were successfully generated with all samples.

cross the 11 CpGs, the proportion of cytosines methylated was less than 1.8% andreater than 26.6% in the SNU484 and KG1a cell lines, respectively. The proportionf methylation was less than 5% in normal controls. Low-level methylation (0–5%)as interpreted as background noise [10,12]. Methylation was quantified in terms ofethylation level (MtL). As per the equations below, MtL was the average percentage

f cytosines methylated per CpG: MtL (%) = (∑

%methylated cytosines)/11.For the KG1a cell line and SNU464 cell line the MtL value was 57.1% and 0.0%,

espectively. For MDS patients the mean (±SD) MtL was 8.7 ± 9.3%.

.3. Bone marrow blast percentages

The patients of the present study had a heterogeneous range of disorders: 11ad refractory anemia (RA), 5 had refractory anemia with ringed sideroblast (RARS),had refractory cytopenia with multilineage dysplasia (RCMD), 1 had refractory

ytopenia with multilineage dysplasia and ringed sideroblast (RCMD-RS), 26 hadefractory anemia with excess blast-1 (RAEB-1), and 23 had refractory anemia withxcess blast-2 (RAEB-2) [13]. Each of these groups was assigned a score based onheir average bone marrow blast percentage: <5% = 0, 5–10% = 0.5, and 11–19% = 1.513].

.4. Karyotype

.4.1. Probes for interphase FISH (fluorescence in situ hybridization)We used the interphase FISH technique to detect any abnormalities of chromo-

omes 5q, 7q, 8, and 1. All probes were obtained from Vysis Inc. (Downers Grove,L, USA): LSI (locus-specific identifier) EGR1/D5S23, D5S721 Dual Color Probe for

hromosome 5q; LSI D7S522/CEP 7 Probe for chromosome 7q; CEP (centromere enu-eration probe) 8 DNA Probe for chromosome 8; LSI p58 (1p36)/Telvysion 1p/LSI

q25 Probe for chromosome 1. The normal cut-off values of deletion 5q, deletionq, trisomy 8, and trisomy 1 abnormalities were based on the mean (±3 SD) of 40egative controls, and equal to 2.0%, 2.0%, 2.0%, and 1.5%, respectively. No less than00 nuclei of a sample were scored for the deletion and the rearrangement in each

rch 34 (2010) 718–722 719

sample. We analyzed interphase cells, as per the manufacturer’s instructions andISCN (2005) criteria [15].

2.4.2. Conventional G-bandingA standard protocol was used to perform G-banding with heparinized whole

bone marrow samples [16]. No less than 20 metaphase cells were karyotyped inaccordance with ISCN (2005) criteria [15].

2.4.3. Cytogenetic findings and IPSS scoresEither one of the three risk categories was assigned on the basis of cytoge-

netic findings. These were good risk: normal cytogenetics, isolated del(5q), isolateddel(20q), and −Y; poor risk: complex abnormalities or abnormalities of chromo-some 7; intermediate risk: all other abnormalities. Each risk category was assigneda score: good =−0, intermediate = 1.5, and poor = 1.0 [13].

2.5. Cytopenia

The presence of cytopenia was determined from the complete blood cell countof the bone marrow aspirate obtained at the time of initial diagnosis. The crite-ria for anemia, neutropenia, and thrombocytopenia were Hgb <10 g/dL, neutrophils<1.8 × 109/L, and platelets <100 × 109/L, respectively. Cytopenia in 0 or 1 lineagewas assigned a score of 0, and cytopenia in 2 or 3 lineages assigned a score of 0.5[13].

2.6. IPSS risk groups

Four risk groups are assigned based on the summed scores of variables describedabove: low, 0; INT (intermediate)-1, 0.5–1.0; INT-2, 1.5–2.0; high, ≥2.5 [13].

2.7. Statistical analysis

All statistical analyses were performed using SPSS 12.0 (SPSS Inc., Chicago, IL,USA). The values for IPSS variables were based on data obtained at the time of initialdiagnosis [13]. Differences in MtI and MtL across the various levels of each IPSSvariable were investigated with Mann–Whitney U-tests. Data concerning patientdeaths was obtained in October 2008. Survival curves were estimated using theKaplan–Meier method. Log-rank tests were used to conduct univariate comparisonsof overall survival among subgroups. Multivariate analyses adjusted for significantprognostic factors were performed using Cox’s hazard regression model. Prognosticsignificance was evaluated in terms of 95% confidence intervals.

3. Results

3.1. Relationships between the quantity of p15INK4b promotormethylation and IPSS scores

3.1.1. Bone marrow blast percentageAs shown in Table 1, patients with 5–20% blasts had MtL val-

ues significantly higher than patients with <5%. Patients were alsocategorized into three groups in accordance with IPSS criteria:blasts <5%, 5% ≤ blasts < 10%, and 10% ≤ blasts < 20%. MtI increasedacross these groups in line with increasing percentage of blasts (seeTable 1). However, the mean MtL was highest in the 5–10% blastgroup and lowest in the <5% blast group, but these values were notsignificantly different.

C: 10% < blasts in bone marrow < 20% 23 9.8 ± 8.5 B and C: 0.582Blasts in bone marrow < 5% 25 6.1 ± 6.4

0.0305% < blasts in bone marrow < 20% 49 10.1 ± 10.3

Total 74 8.7 ± 9.3

Mann–Whitney U-test.

720 M. Kim et al. / Leukemia Research 34 (2010) 718–722

Table 2The quantity of p15INK4b methylation across karyotype groups.

Karyotype No. of cases MtL (%): mean ± SD

(+) (−) (+) (−) p

GoodNormal 31 25 10.1 ± 11.8 7.2 ± 6.6 0.824−Y 1 55 29.8 ± 0.0 8.4 ± 9.5 0.071del(5q) 2 54 9.2 ± 1.4 10.0 ± 8.8 0.273del(20q) 2 54 6.7 ± 3.1 8.8 ± 10.0 0.657

IntermediateOthers 24 32 6.6 ± 6.2 10.4 ± 11.7 0.275Trisomy 8 9 47 5.6 ± 5.1 9.4 ± 10.4 0.415Trisomy 1 12 44 7.7 ± 7.5 9.1 ± 10.4 0.780

PoorComplex 10 46 7.5 ± 5.3 9.1 ± 10.6 0.814Chromosome 7 abnormalities 9 47 6.4 ± 3.7 9.2 ± 10.6 0.920

Good 25 31 7.2 ± 6.6 10.1 ± 11.8 0.824Intermediate 24 32 6.6 ± 6.2 10.4 ± 11.7 0.275Poor 12 44 7.0 ± 4.9 9.2 ± 10.8 0.811

Total 56 8.8 ± 9.8

Mann–Whitney U-test.

Table 3The quantity of p15INK4b methylation across cytopenia groups.

Cytopenias Number of cases MtL (%): mean ± SD

Total (+) (−) (+) (−) p

Anemia (Hb < 10 g/dL) 69 56 13 9.0 ± 10.0 7.8 ± 7.4 0.586Neutropenia (neutrophils < 1.8 × 109/L) 69 53 16 9.1 ± 10.0 7.8 ± 7.9 0.665Thrombocytopenia (platelets < 100 × 109/L) 69 36 33 11.2 ± 10.9 6.2 ± 7.0 0.003

0–1 cytopenia 69 12 4.1 ± 2.5 0.036

M

ns

3

Ttepcp

3

sn

TT

2–3 cytopenias 69

Total 69

ann–Whitney U-test.

etic factors is summarized in Table 2. There were no statisticallyignificant differences between any of the subgroups.

.1.3. CytopeniaOnly 69 patients had the data of complete blood count. As seen in

able 3, MtLs were higher in patients with anemia, neutropenia, orhrombocytopenia than in patients without these conditions. How-ver, the only statistically significant differences were betweenatients with thrombocytopenia and patients without thrombo-ytopenia. Patients with bilineage cytopenia had higher MtL thanatients with unilineage or no cytopenia (see Table 3).

.1.4. IPSS risk groupThe MtL data for 56 patients grouped according to IPSS risk

cores are summarized in Table 4. There were no statistically sig-ificant differences between any of the risk groups.

able 4he quantity of p15INK4b methylation across IPSS risk groups.

Risk group IPSS No. of cases MtL (%): mean ± SD

Low 0.0 3 5.2 ± 3.2

Intermediate-10.5 12 6.3 ± 7.8

10.3 ± 12.11.0 18 13.1 ± 13.9

Intermediate-21.5 6 3.5 ± 1.2

7.6 ± 7.32.0 9 10.4 ± 8.4

High2.5 3 3.6 ± 1.1

63 ± 363.0 5 8.0 ± 3.7

Total 56 8.7 ± 9.8

57 9.8 ± 10.1

12.2 ± 8.8

3.2. The quantity of p15INK4b promotor methylation andsurvival prognosis in MDS patients

Patient survival Of the 74 patients investigated, 20 had died(from any cause) and 54 had either survived or were cen-sored as of October 2008. Patients were divided into heavyand light methylation groups on the basis of MtL cut-off val-ues of 7%. Patients with MtL >7% survived for a lesser meanduration than patients with MtL <7% (738 days vs. 1804days, p = 0.031) (see Fig. 1). However, these differences didnot remain statistically significant with multivariate analyses inwhich blast percentage was the only significant prognostic fac-tor (representing the relationship between methylation and blastpercentage).

4. Discussion

In vitro studies using tumor cell lines have revealed a correlationbetween the extent to which tumor suppressor genes are expressedand promotor methylation density [6,7]. However, studies whichhave considered p15INK4b promotor methylation in relation toMDS were performed using MSPCR, a semi-quantitative method.MSPCR generates frequent false positives and does not provideinformation on the methylation status of individual CpG sites[10,17]. Pyrosequencing overcomes these limitations [10,18,19]

and produces results that correlate highly with those of othermethods, including COBRA, SnaPmeth, and mass spectrometry[20–22]. In the present study we pyrosequenced 11 consecutiveCpGs in the 5′ upstream region of the p15INK4b promotor usingPyrosequencingTM. It was our hypothesis that the extent to which

M. Kim et al. / Leukemia Resea

Ff>

pp

wissmfRttp

rCi[wnc

wcsdttl

wsmddam

[

[

ig. 1. Heavy methylation of the p15INK4b promotor region is a prognostic factoror poor survival in MDS patients. Kaplan–Meier curve of MDS patients with MtL7% and patients with MtL <7% (738 days vs. 1804 days, respectively).

15INK4b is methylated correlates with factors relating to MDSatient survival.

There was a greater degree of p15INK4b methylation in patientsith >5% bone marrow blasts than in patients with <5% blasts. Hav-

ng >5% bone marrow blasts is an important indicator of shorterurvival time and higher incidence of leukemic transformation. Atudy using MSPCR was reported as finding p15INK4b methylationore frequently in patients with RAEB or RAEB-t (RAEB in trans-

ormation, MDS with 20–30% blasts) than in patients with RA orARS [2]. In the present study, we showed with pyrosequencinghat it is not only the incidence of p15INK4b methylation, but alsohe extent of this methylation that is associated with marrow blastercentage.

Alterations in G1/S phase-regulating proteins like theetinoblastoma protein, cyclins D and E, CDKs 4 and 6, andDK inhibitors p16INK4A and p15INK4b, cause karyotypic instabil-

ty by uncontrolled centrosome replication, leading to neoplasia23]. The presence of p15INK4b methylation had relationshipith certain chromosomal changes [5,24]. However, we foundo association between the extent of p15INK4b methylation andhromosomal aberrations addressed by the IPSS.

There was a greater degree of p15INK4b methylation in patientsith thrombocytopenia than in patients with a normal platelet

ount. This should be of particular interest to morphologists, con-idering that dysmegakaryopoiesis is rather pathognomonic thanyserythropoiesis or dysgranulopoiesis in MDS [25]. The results ofhe present study also indicate that there is a relationship betweenhe extent of p15INK4b methylation and the number of cytopenicineages.

We did find that the initial extent of p15INK4b methylationas related to survival in MDS patients. With univariate analy-

es, patients with a higher degree of p15INK4b methylation (heavyethylation) showed worse survival than patients with a lower

egree of p15INK4b methylation (light methylation). However, thisifference did not remain statistically significant with multivari-te analysis. This can be interpreted as the quantity of p15INK4bethylation contributing to the prognosis of MDS patients in com-

[

[

rch 34 (2010) 718–722 721

bination with other factors to which it is related, such as marrowblast percentage or cytopenia.

The management of MDS is influenced by an assessment of therisks associated with the disease [26]. Immunosuppressive therapy,like cyclosporine or anti-thymocyte globulin, is used for low-riskMDS patients. With high-risk MDS, hypomethylating agents can bean appropriate option. We found in the present study that p15INK4bmethylation was more extensive in high-risk MDS patients and thatit was related to poor prognostic factors (blast % and cytopenia)and reduced survival. This supports the use of hypomethylatingagents in treating high-risk MDS patients. Further, the quantityof p15INK4b promotor methylation could serve as a biomarker bywhich to track treatment effects.

Using the new method of pyrosequencing, our study has shownthat the extent of p15INK4b promotor methylation in MDS is asso-ciated with known prognostic factors, including bone marrow blastpercentage, thrombocytopenia, and number of cytopenic lineages.Heavy methylation of the p15INK4b promotor region is a prognos-tic factor for poor survival in MDS patients. Assessing the extent ofp15INK4b methylation could be helpful in guiding the managementof MDS patients.

Acknowledgements

This work was supported in part by (1) the National R&D Pro-gram for Cancer Control, Ministry of Health, Welfare and FamilyAffairs, Republic of Korea (0720440), (2) the Korea Science andEngineering Foundation (KOSEF) funded by the Ministry of Edu-cation, Science, and Technology (M1064152000108N415200110).Tae Young Kim and Bora Oh are grateful for being awarded a BK21fellowship. The authors are very appreciative of the technical sup-port of BMS in performing pyrosequencing.

References

[1] Nimer SD. Myelodysplastic syndromes. Blood 2008;111:4841–51.[2] Uchida T, Kinoshita T, Nagai H, Nakahara Y, Saito H, Hotta T, et al. Hyper-

methylation of the p15INK4B Gene in myelodysplastic syndromes. Blood1997;90:1403–9.

[3] Quesnel B, Guillerm G, Vereecque R, Wattel E, Preudhomme C, Bauters F, et al.Methylation of the p15(INK4b) gene in myelodysplastic syndromes is frequentand acquired during disease progression. Blood 1998;91:2985–90.

[4] Tien HF, Tang JH, Tsay W, Liu MC, Lee FY, Wang CH, et al. Methylation of thep15(INK4B) gene in myelodysplastic syndrome: it can be detected early at diag-nosis or during disease progression and is highly associated with leukaemictransformation. Br J Haematol 2001;112:148–54.

[5] Au WY, Fung A, Man C, Ma SK, Wan TS, Liang R, et al. Aberrant p15gene promoter methylation in therapy-related myelodysplastic syndrome andacute myeloid leukaemia: clinicopathological and karyotypic associations. Br JHaematol 2003;120:1062–5.

[6] Cameron EE, Baylin SB, Herman JG. p15(INK4B) CpG island methylation in pri-mary acute leukemia is heterogeneous and suggests density as a critical factorfor transcriptional silencing. Blood 1999;94:2445–51.

[7] Yang R, Li WW, Hoang BH, Kim H, Banerjee D, Kheradpour A, et al. Quantitativecorrelation between promoter methylation and messenger RNA levels of thereduced folate carrier. BMC Cancer 2008;8:124.

[8] Lee ES, Issa JP, Roberts DB, Williams MD, Weber RS, Kies MS, et al. Quantitativepromoter hypermethylation analysis of cancer-related genes in salivary glandcarcinomas: comparison with methylation-specific PCR technique and clinicalsignificance. Clin Cancer Res 2008;14:2664–72.

[9] Tost J, Gut IG. DNA methylation analysis by pyrosequencing. Nat Protoc2007;2:2265–75.

10] Brakensiek K, Wingen LU, Länger F, Kreipe H, Lehmann U. Quantitative high-resolution CpG island mapping with pyrosequencing reveals disease-specificmethylation patterns of the CDKN2B gene in myelodysplastic syndrome andmyeloid leukemia. Clin Chem 2007;53:17–23.

11] Gao W, Kondo Y, Shen L, Shimizu Y, Sano T, Yamao K, et al. Variable DNAmethylation patterns associated with progression of disease in hepatocellular

carcinomas. Carcinogenesis 2008;29:1901–10.

12] Shaw RJ, Liloglou T, Rogers SN, Brown JS, Vaughan ED, Lowe D, et al. Promotermethylation of P16, RARb, E-cadherin, cyclin A1 and cytoglobin in oral cancer:quantitative evaluation using pyrosequencing. Br J Cancer 2006;94:561–8.

13] List AF, Vardiman J, Issa JP, DeWitte TM. Myelodysplastic syndromes. HematolAm Soc Hematol Educ Program 2004:297–317.

7 Resea

[

[

[

[

[

[

[

[

[

[

[pathway in acute leukemias. Ann Hematol 2003;82:738–42.

[25] Thiele J, Quitmann H, Wagner S, Fischer R. Dysmegakaryopoiesis in myelodys-

22 M. Kim et al. / Leukemia

14] Herman JG, Graff JR, Myöhänen S, Nelkin BD, Baylin SB. Methylation-specificPCR: a novel PCR assay for methylation status of CpG islands. Proc Natl AcadSci USA 1996;93:9821–6.

15] Shaffer LG, Tommerup N, editors. ISCN 2005: an international system for humancytogenetic nomenclature (2005). Basel: S. Karger; 2005.

16] Czepulkowski BH, Bhatt B, Rooney DE. Basic techniques for the preparationand analysis of chromosomes from bone marrow and leukaemic blood. In:Human cytogenetics: malignancy and acquired abnormalities. Oxford: OxfordUniversity Press; 2001. pp. 1–26.

17] Cottrell SE. Molecular diagnostic applications of DNA methylation technology.Clin Biochem 2004;37:595–604.

18] Ronaghi M, Uhlen M, Nyren P. A sequencing method based on real-time

pyrophosphate. Science 1998;281:363–5.

19] Marsh S, King CR, Garsa AA, McLeod HL. Pyrosequencing of clinically relevantpolymorphisms. Methods Mol Biol 2005;311:97–114.

20] Uhlmann K, Brinckmann A, Toliat MR, Ritter H, Nurnberg P. Evaluation of apotential epigenetic biomarker by quantitative methyl-single nucleotide poly-morphism analysis. Electrophoresis 2002;23:4072–9.

[

rch 34 (2010) 718–722

21] Colella S, Shen L, Baggerly KA, Issa JP, Krahe R. Sensitive and quantitativeuniversal pyrosequencing methylation analysis of CpG sites. Biotechniques2003;35:146–50.

22] Dupont JM, Tost J, Jammes H, Gut IG. De novo quantitative bisulfitesequencing using the pyrosequencing technology. Anal Biochem 2004;333:119–27.

23] McDermott KM, Zhang J, Holst CR, Kozakiewicz BK, Singla V, Tlsty TD. p16INK4aprevents centrosome dysfunction and genomic instability in primary cells. PLoSBiol 2006;4:e51.

24] Chim CS, Wong AS, Kwong YL. Epigenetic inactivation of INK4/CDK/RB cell cycle

plastic syndromes (MDS): an immunomorphometric study of bone marrowtrephine biopsy specimens. J Clin Pathol 1991;44:300–5.

26] Mufti GJ, Chen TL. Changing the treatment paradigm in myelodysplastic syn-dromes. Cancer Control 2008;15:S14–28.