Letters to the Editor 246 7 Life Sciences Division, Department of Statistics, University of 3 Michaud J, Simpson KM, Escher R, Buchet-Poyau K, Beissbarth T, California, Lawrence Berkeley National Laboratory, Carmichael C et al. Integrative analysis of RUNX1 downstream Berkeley, CA, USA and pathways and target . BMC Genomics 2008; 31: 363. 8Department of Pediatric Hemato-oncology, Radboud 4 Mikhail FM, Sinha KK, Saunthararajah Y, Nucifora G. Normal and University Nijmegen Medical Center, Nijmegen, transforming functions of RUNX1: a perspective. J Cell Physiol The Netherlands 2006; 207: 582–593. E-mails: [email protected] or 5 Renneville A, Roumier C, Biggio V, Nibourel O, Boissel N, [email protected] Fenaux P et al. Cooperating mutations in acute myeloid 9These authors contributed equally to this work. leukemia: a review of the literature. Leukemia 2008; 22: 915–931. 6 Ripperger T, Steinemann D, Go¨hring G, Finke J, Niemeyer CM, Strahm B et al. A novel pedigree with heterozygous germline RUNX1 mutation causing familial MDS-related AML: can these families serve as a multistep model for leukemic transformation? References Leukemia 2009; 23: 1364–1366. 7 Michaud J, Wu F, Osato M, Cottles GM, Yanagida M, Asou N et al. 1 Owen CJ, Toze CL, Koochin A, Forrest DL, Smith CA, Stevens JM In vitro analyses of known and novel RUNX1/AML1 mutations in et al. Five new pedigrees with inherited RUNX1 mutations causing dominant familial platelet disorder with predisposition to acute familial platelet disorder with propensity to myeloid malignancy. myelogenous leukemia: implications for mechanisms of pathogen- Blood 2008; 112: 4639–4645. esis. Blood 2002; 99: 1364–1372. 2 Song WJ, Sullivan MG, Legare RD, Hutchings S, Tan X, Kufrin D 8 Buijs A, Poddighe P, van WR, van SW, Borst E, Verdonck L et al. et al. Haploinsufficiency of CBFA2 causes familial thrombocyto- A novel CBFA2 single-nucleotide mutation in familial platelet penia with propensity to develop acute myelogenous leukaemia. disorder with propensity to develop myeloid malignancies. Blood Nat Genet 1999; 23: 166–175. 2001; 98: 2856–2858.

Aberrant structure is characteristic of resistant chronic lymphocytic leukaemia cells

Leukemia (2010) 24, 246–251; doi:10.1038/leu.2009.213; and activity are further prognostic markers as published online 22 October 2009 unfavourable disease was associated with short telomeres4 (reviewed in Van Bockstaele et al.3). This, together with altered expression of several telomeric components,5 indicates that are the capping structures of ends and telomeres are profoundly rearranged in CLL cells. are composed of repeated DNA sequences (B10 kb in adult A subgroup of CLL cells is resistant to DNA damage-induced somatic cells) and a specific set of associated proteins. The in vitro (reviewed in Bouley et al.6) and patients regulation of telomere length results from the action of telomere- harbouring these cells inevitably develop the aggressive form of lengthening mechanisms, such as the telomerase complex disease, suggesting that this is a clinically relevant factor. Here, (hTERT, hTR and dyskerin), and of telomere-shortening mechan- we investigated whether specific telomere alterations can isms, such as replication and recombination. Telomerase distinguish CLL cells sensitive (CLL-S) or resistant (CLL-R) to activity is regulated in cis by the hexa–protein complex DNA damage-induced apoptosis in vitro. A total of 35 patients (TRF1, TRF2, hRAP1, POT1, TIN2 and TPP1) and epigenetic were included in this study (16 CLL-R and 19 CLL-S, clinical and factors.1,2 Proteins involved in DNA replication and repair are biological behaviours were presented in Table 1). Southern blot also associated in telomeric structure and function.2 The analysis showed a significant difference in telomere length shortening of telomere sequences upon cell divisions in most between the two subgroups of patients (Figure 1a). The mean somatic cells results in irreversible cell growth arrest called telomere length was twofold shorter (U-test, Po0.001) in senescence or in apoptosis. A change in telomere function is CLL-R compared with the CLL-S (3552±1151 bp for CLL-R one of the main mechanisms underlying the evolution and and 8010±1668 bp for CLL-S) (Figure 1b). Furthermore, we maintenance of cancer cells. Telomere erosion may be critical observed a positive correlation (R2 ¼ 0.675) between telomere in tumour suppression as it impairs cell proliferation. In restriction fragment length and the in vitro sensitivity to many cancer cells, this is circumvented by reactivating the apoptosis (Figure 1c) that appeared independent of previous expression and/or activity of telomerase or by homologous clinical treatment of patients (Figures 1c and d). The use of the recombination.1,2 Spearman’s rank correlation showed that this association was Chronic lymphocytic leukaemia (CLL) is a B-cell malignancy significant (Po0.0001). These results clearly show that resistant with increased incidence among the elderly. The disease is due CLL cells have shorter telomeres than sensitive CLL cells. to an imbalance between cell death and proliferation resulting The mRNA level of hTERT was assessed by quantitative PCR analysis in an accumulation of malignant cells in the bone marrow and and revealed no significant difference (U-test, P40.05) between peripheral blood (reviewed in Van Bockstaele et al.3). Clinically, the two subgroups of patients (Figure 2a). We also analyzed the the disease varies from indolent, which may progress to/or mRNA level of the negative regulator of the telomerase, PinX1, appear at diagnostic as an aggressive incurable form. The most by quantitative PCR analysis in samples already included in the commonly used biological markers for the disease outcome are analysis for the expression of hTERT. No significant difference the type of chromosomal aberration and the mutational status of (U-test, P40.05) was observed between the two subgroups of immunoglobulin heavy-chain variable region genes (IgVH), cells (Figure 2b). Telomere maintenance and protection mainly often associated with the expression status of Zap70 tyrosine depends on the presence of the six shelterin components and kinase (reviewed in Van Bockstaele et al.3). The telomere length their stoichiometry at chromosome ends is crucial for telomere

Leukemia Table 1 Clinical characteristics and outcome of CLL patients

Patient’s ID Age Sex Binet’s Treatment Date of treatment Date of sampling IgVH mutation Matutes Chromosomal aberrations Apoptosis stage status score in vitro

LLC-S 1 77 m A No January 2007 M 4 del 13q14 monoallelic Sensitive LLC-S 2 77 m A Yes 1994 April 2007 M 4 del 13q14 monoallelic Sensitive LLC-S 3 53 m A Yes 2001 April 2007 M 4 trisomy 12, t(14;17) Sensitive LLC-S 4 89 m A Yes March 2006 December 2006 UM 4 11q- Sensitive LLC-S 5 81 m A No March 2007 M 4 ND Sensitive LLC-S 6 78 m A No December 2006 ND 4 del 13q monoallelic Sensitive LLC-S 7 58 m B No December 2006 M 4 del 13q monoallelic Sensitive LLC-S 8 69 m A No November 2006 M 4 del 13q monoallelic Sensitive LLC-S 9 79 m A No November 2006 ND 4 ND Sensitive LLC-S 10 81 m A No January 2006 ND 5 del 13q monoallelic Sensitive LLC-S 11 63 m A No 2006 M 5 ND Sensitive LLC-S 12 81 f A No May 2007 ND 4 ND Sensitive LLC-S 13 73 f A No November 2006 M 5 ND Sensitive LLC-S 14 63 f A No April 2008 M 5 del 13q14 monoallelic, del 17p monoallelic Sensitive LLC-S 15 78 f A No May 2007 ND 5 ND Sensitive LLC-S 16 85 f A No March 2007 ND 5 ND Sensitive LLC-S 17 74 f A No January 2007 UM 5 trisomy 12, t(14;17 ) Sensitive LLC-S 18 73 f A No December 2006 M 4 ND Sensitive LLC-S 19 63 f A No May 2007 ND 4 ND Sensitive LLC-R 1 78 m A Yes 2002 March 2007 M 5 del 13q14 biallelic Resistant LLC-R 2 56 m C Yes January 2006 October 2006 ND 5 del 13q14 biallelic Resistant Editor the to Letters LLC-R 3 57 m A Yes 1997 January 2006 UM 4 del 13q14 monoallelic Resistant LLC-R 4 72 m A No July 2007 ND 5 Complex karyotype Resistant LLC-R 5 64 m A Yes 2003 January 2007 M 4 del 13q biallelic, del 17p Resistant LLC-R 6 70 m B Yes October 2006 April 2007 ND 4 del 13q14 biallelic, del 17p monoallelic Resistant LLC-R 7 89 m A No July 2007 UM 5 ND Resistant LLC-R 8 75 m ND No October 2007 ND 4 del 13q, del 11q Resistant LLC-R 9 76 m ND No October 2007 ND 5 del 13q, t(14;18) Resistant LLC-R 10 82 m A Yes January 2003 June 2004 M 5 del 13q14 biallelic, del 17p monoallelic Resistant LLC-R 11 68 f A Yes 2005 January 2007 M 5 del 13q14 biallelic, del 17p monoallelic Resistant LLC-R 12 76 f A/B Yes September 2006 April 2007 UM 5 del 13q14monoallelic, del 17p monoallelic Resistant LLC-R 13 60 f A No November 2006 UM 5 del 13q14 biallelic, del 17p monoallelic Resistant LLC-R 14 76 f A Yes January 2006 December 2005 UM 5 ND Resistant LLC-R 15 77 f C Yes 2000 2006 ND 4 ND Resistant LLC-R 16 75 f A Yes September 2007 March 2006 M 5 del 13q14 monoallelic, del 17p monoallelic Resistant

Abbreviations: CLL, chronic lymphocytic leukaemia; del, deletion; f, female; IgVH, immunoglobulin heavy-chain variable genes; m, male; M, mutated; ND, not determined; t, translocation; UM, unmutated. Sex was defined as male and female, and age is expressed in years. Staging was according to the Binet’s staging system. Treatments (if any) were received at least 3 months before cell sampling for 6 DNA extractions, chromatin immunoprecipitation assays and apoptosis in vitro determination. Apoptosis score was established as described. The somatic mutational status of IgVH has been established according to the 98% homology cutoff value to the closest germline gene; that is, o98% homology was considered as mutated (M), whereas X98% homology was considered as unmutated. Chromosomal aberrations were established by fluorescent in situ hybridization using LSI ATM/LSI p53 and LSI D13S319/LSI 13q34/CEP12 probe sets (Abbott, Rueil Malmaison, France), which are complementary to 11q22.3, 17p13.1, 13q14.3, 13q34 and 12p11.1-q11 genome regions, respectively. Presented data take into account the percentage of the interphase nuclei with the same chromosomal anomaly (that is, at least 5% of nuclei should contain the same aberration). Leukemia 247 Letters to the Editor 248 Size MW CLL-S CLL-R (bp) 12000

10000 10 000 8 000 6 000 5 000 8000 4 000

3 000 6000 ∗∗∗

4000 Mean TRF Length (bp)

2000

0 CLL-R CLL-S

5000 CLL-R 100 r2= 0.675 CLL-S 4500 90 P < 0.0001 4000 80 3500 70 3000 60 50 2500 40 2000

Apoptotic cells (%) 30 1500 Mean TRF Length (bp) 20 1000 CLL-R 10 500 0 0 2000 4000 6000 8000 10000 12000 0 Mean TRF Length (bp) Treated Untreated

Figure 1 Short mean telomere length is characteristic of chronic lymphocytic leukaemia (CLL) cells resistant (CLL-R) to DNA damage-induced apoptosis in vitro. Blood from leukaemic patients was collected in heparin-coated tubes as previously described (reviewed in Bouley et al.6 and the references within). The cells samples were separated into two groups according to their sensitivity to apoptosis.6 TRF indicates telomere restriction fragment (TRF) lengths and are expressed as base pairs (bps). (a) Representative example of TRF length measurement using Southern blot. MW indicates molecular weight ladder. For Southern blot analysis, Hinf1- and Rsa1-restricted genomic DNA was hybridized with a telomeric 32 (CCCTAA)10 probe labelled with g P (dATP). Telomere length was calculated after the intensity signal quantification on the Typhoon 9400 (Amersham-Pharmacia Biotech, Orsay, France) and using the ImageQuant software (Amersham-Pharmacia Biotech). (b) Comparison of TRF length between CLL-R and CLL-S (sensitive) cells. Values represent the mean TRF lengths (bp)±s.d. (CLL-R: n ¼ 16 samples; CLL-S: n ¼ 19 samples). A nnn Mann–Whitney U-test was applied for statistical analysis. Po0.001. (c) Correlation between TRF length and sensitivity to apoptosis. Apoptosis represents the difference between percentage of cell death before and after irradiation. A significant correlation was found (R2 ¼ 0.675, Po0.0001), giving the mathematical relation Y ¼À29.781 þ 0.11X. CLL-S are presented by trapezoids; CLL-R derived from patients treated before cell sampling are presented by rectangles and CLL-R from untreated patients by triangles. (d) Mean telomere length in CLL-R derived from chemotherapy-treated or -untreated patients.

stability.2 We determined the mRNA level of five telomeric was established by chromatin immunoprecipitation (ChlP) proteins (TRF1, TRF2, TIN2, hRAP1 and POT1) using quantita- assays (Figure 2e). tive PCR analysis in CLL-R and CLL-S samples. No significant Telomeres have epigenetic marks characteristic of hetero- difference between the two subgroups (U-test, P40.05) was chromatin, including trimethylation of histone H3 lysine 9 and observed (Figure 2c). TRF2 is involved in telomeres structure histone H4 lysine 20 and hypoacetylation of histone H3. In and protection, whereas TRF1 is implicated in telomere length particular, decreased levels of H3K9 trimethylation at telomeres control through negative regulation of the telomerase activity. are associated with aberrant telomere elongation.1 To determine We therefore analyzed TRF2 protein levels by western blot. No whether telomeric chromatin may be discriminating between significant difference (U-test, P40.05) was observed between resistant and sensitive CLL cells, we have analyzed histone CLL-R and CLL-S (Figure 2d). Similarly, no significant difference H3K9 methylation (H3K9me3) and acetylation (H3K9ac) at (U-test, P40.05) in the levels of theses two proteins at telomeres telomeres of B-CLL cells by chromatin immunoprecipitation

Leukemia Letters to the Editor 249 14

0.30 12 10 0.25 10 8 0.20 8 6 0.15 6 4 0.10 4

0.05 2 Expression Level (ratio) 2

0 hTERT expression Level (ratio) 0.00 Pinx1 expression Level (ratio) 0 CLL-R CLL-S CLL-R CLL-S TRF1 TRF2 TIN2 hRAP1 POT1

1.2

1 TRF2 No Ab Input 10 0.8 CLL-R 9 8 0.6 CLL-S 7 (CCCTAA) Level (ratio) 0.4 Telomeric probe 6 5

TRF2 protein expression 0.2 4 TRF1 No Ab Input 0 3

CLL-R CLL-S 10 CLL-R 2 1 CLL-R CLL-S

Telomeric DNA, percentage of input 0

(CCCTAA) CLL-S TRF2 TRF2 TRF1 Telomeric probe β-Actin

Figure 2 (a) The hTERT expression level in chronic lymphocytic leukaemia (CLL) cells (Q-PCR). Reverse transcription and quantitative PCR (RT–Q-PCR) were carried out as previously described.5 Values represent the average of hTERT expression level relative to b-actin±s.d. (CLL-R (resistant) : n ¼ 11 samples; CLL-S (sensitive): n ¼ 14 samples). (b) PinX1 expression level in CLL cells. The expression of PinX1 was established by Q-PCR analysis. b-actin was used as the reference gene. Values represent the average of PinX1 expression level±s.d. (CLL-R: n ¼ 11 samples; CLL-S: n ¼ 14 samples). (c) Telomere maintenance by the shelterin complex in CLL cells. Dark columns represent resistant samples and clear columns represent the sensitive samples. Comparison of the shelterin components expression level between CLL-R and CLL-S. Values represent the average of expression level (Q-PCR) relative to b-actin±s.d. (CLL-R: n ¼ 11 samples; CLL-S: n ¼ 14 samples). (d) Comparison of the amount of TRF2 protein between CLL-R (dark columns) and CLL-S (light columns, western blot). Values represent the average amount of protein relative to b-actin±s.d. (CLL-R: n ¼ 8; CLL-S: n ¼ 7). (e) Comparison of average TRF1 and TRF2 protein level at the telomeric chromatin between CLL-S (light columns) and CLL-R (dark columns). Chromatin immunoprecipitations (ChIPs) were carried out as previously described8 using polyclonal TRF1 and TRF2 antibodies (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Immunoprecipitated DNA was extracted with phenol–chloroform–isoamylic alcohol and dot-spotted on Hybond N þ membrane (Amersham-Pharmacia Biotech, Orsay, France). Hybridization 32 was in CHURCH buffer with a telomeric probe, g- P radiolabelled (TTAGGG)10 oligonucleotide complementary to the double-strand part of telomeres. Membranes were exposed to Phosphoimager screen and signals were quantified by ImageQuant software (Amersham-Pharmacia Biotech). In each experiment, equal amounts of DNA (Quant-iT PicoGreen dsDNA Reagent from Molecular Probes, Invitrogen, Cergy-Pontoise, France) and identical immunoprecipitation conditions were used. DNA load was also controlled by hybridizing membranes with an Alu probe (50-AAGTCGCGGCCGCTTGCAGTGAGCCGAGAT-30). The relative enrichment in telomeric DNA for each ChIP assays was calculated as follows: (telomeric signal of immunoprecipitated DNAÀtelomeric signal of no antibody control)/total telomeric signal (input). In all cases, the value obtained represents the percentage of the total input telomeric DNA (thus correcting for differences in the number of telomeric repeats) adjusted with respect to the amount of DNA used for each assay. To exclude unspecific binding between beads and crosslinked chromatin, a ChIP was performed in the absence of any antibody (No Ab). Input represents total DNA used in each ChIP assay. Quantification of the telomeric signals is shown in the right panel and is expressed as the mean percentage±S.D. of the total telomeric DNA. A Mann-Whitney U-test was applied for statistical analysis and was considered as significant if P-value is o0.05.

experiments. A total of 20 CLL samples (10 CLL-R and 10 CLL-S) CLL-R cells are associated with an increase in the heterochro- were analyzed for H3K9me3 and H3K9ac. As shown in Figure 3, matic H3K9me3 mark, indicating a more compact telomeric although no significant difference (U-test, P40.5) was observed chromatin configuration. for the acetylated form of histone H3, we noted a significant In conclusion, we report here that short telomeres occur in the increase (U-test, Po0.001) in H3K9me3 at the telomeres of restricted subset of CLL cells derived exclusively from patients CLL-R samples (50±18% in CLL-R vs 14±6% in CLL-S). As an refractory to front-line treatment with purine analogues and/or internal control, we measured the density of H3 bound to alkylating agents. This subset of CLL cells is characterized by telomeres of CLL cells, which was similar in sensitive and their inability to activate apoptotic death following DNA resistant cells. These results therefore show that telomeres of damage in vitro. This feature is independent of IgVH mutational

Leukemia Letters to the Editor 250 Total H3 ab0.8 80 70 *** 0.7 60 0.6 H3K9 3meH3K9 ac No Ab Input 50 0.5

10 0.4 CLL-R 40 of input of input 30 0.3 0.2 CLL-S 20 (CCCTAA)

Telomeric probe 10 0.1 Telomeric DNA, percentage Telomeric DNA, percentage 0 0 H3K9 3me H3K9 ac CLL-R CLL-S

Figure 3 Histone modifications at telomeric chromatin in chronic lymphocytic leukaemia (CLL) cells. (a) Chromatin immunoprecipitations (ChIPs) and hybridization conditions were carried out as described in Figure 2 using polyclonal anti-trimethylated H3K9 (H3K9me3) and anti- acetylated histone H3 (H3K9ac) antibodies (Upstate, Nottingham, UK). A no antibody (No Ab) control and anti-histone H3 immunoprecipitation (rabbit polyclonal anti-histone H3, Abcam, Cambridge, UK) were carried out as controls. Input represents total DNA used in each ChIP assay. Quantification of the telomeric signals in the right panel is expressed as a percentage of total telomeric DNA. Bars represent the mean percentage±s.d. A Mann–Whitney U-test was applied for statistical analysis and was considered as significant (Po0.001). Dark columns represent resistant samples and clear columns represent the sensitive samples (b). Density of H3 bound to telomeres of CLL cells was measured by ChIP experiments using anti-histone H3 antibodies as above and used as control for total nucleosome input.

status, but seems to be associated with multiple chromosomal telomeres found in resistant cells are the cause or the abnormality, as well as with resistance to treatment (Table 1). consequence of the resistance mechanisms operating in these Moreover, we provide evidence of how these short telomeres cells. Nevertheless, our results suggest that heterochromatiza- are organized in chromatin context of resistant CLL cells. tion of telomeres through increased H3K9me3 prevents telo- Mammalian telomeres are associated with both shelterin mere elongation, thus causing increased chromosomal proteins and histones carrying heterochromatic modifications. instability, which was already observed in CLL cells showing Proper assembly of the six shelterin subunits and association short telomeres.4 Ultimately, this might favour the entry of CLL with telomere sequences is crucial for telomere maintenance cells into a short telomere-initiated senescence-like state, which and control of telomerase activity.2 Changes in the expression of could contribute to their maintenance. CLL cells may therefore shelterin genes have recently been established in CLL by be a paradigm for the role of telomeres in cancer, as short comparison to normal B cells, suggesting a global telomere telomeres and high telomerase levels are often associated with dysfunction in B-CLL.5 However, despite enhanced telomere poor clinical outcome in many types of cancer. As these cells shortening, no evidence of altered shelterin protein expression exhibit a deregulated non homologous end joining (NHEJ) DNA or binding of TRF1 and TRF2 at telomeres was observed repair (reviewed in Bouley et al.6), this is reminiscent of the between sensitive and resistant CLL cells. These results indicate recent findings showing that mice with dysfunctional telomeres that if shelterin dysfunction occurs in CLL, it cannot alone survive better with deficient DNA mismatch repair.7 explain the differences in telomere length between indolent and refractory forms of the disease. However, we found that telomeres of resistant cells were associated with elevated levels Conflict of interest of histone lysine 9 trimethylation as compared with sensitive CLL cells. This histone modification mediates chromatin The authors declare no conflict of interest. condensation. The elegant studies of Blasco1 have shown that loss of heterochromatic histone modifications leads to abnormal Acknowledgements telomere elongation, indicating that these histone modifications are negative regulators of telomere elongation. Our finding of We are grateful to all leukaemic donors who have signed elevated levels H3K9me3 at telomeres of resistant CLL implies informed consent form. This work was supported by ARC (ARECA aberrant epigenetic regulation in these malignant cells, leading project). Approval for this study was obtained from the institu- to a more condensed telomeric structure. This may impair tional review boards of the CEA and the Pitie´-Salpeˆtrie`re Hospital. telomere elongation, perhaps by inhibiting the accessibility of telomerase or elongation factors in resistant cells, hence T Brugat1, N Gault1, I Baccelli1, J Mae¨s2,3, A Roborel de explaining the shorter telomeres despite similar telomerase Climens4, F Nguyen-Khac5, F Davi5, H Merle-Be´ral5, expression in resistant and sensitive cells. We also investigate E Gilson4, M Goodhardt2,3 and J Delic1 1 whether p53 deficiency or IgVH mutational status (Table 1) may Laboratoire d’Onco-He´matologie, iRCM, DSV, be linked to short telomeres in resistant cells; an attractive Commissariat a` l’Energie Atomique (CEA), hypothesis that would also explain the defect in apoptotic death Fontenay aux Roses, France; 2Institut Universitaire d’He´matologie, signalling in these cells. However, CLL cells resistant to DNA Universite´ Paris 7, Paris, France; damage-induced apoptosis have a similar proportion of p53wt 3 6 INSERM U662, Paris, France; and p53mut (reviewed in Bouley et al. ) and unmutated and 4Laboratoire de Biologie Mole´culaire mutated IgVH genes, suggesting that multiple chromosomal et de Cellule, CNRS UMR5239, aberrations, rather than p53 deficiency or lack of IgVH Faculte´ de Me´decine Lyon Sud, mutations, are associated with short telomeres in CLL-R. It Universite´ Lyon 1, Lyon, France and 5 remains to be determined whether the shorter heterochromatic Service d’He´matologie Biologique,

Leukemia Letters to the Editor 251 Hoˆpital Pitie´-Salpeˆtrie`re and Universite´ genomic aberrations, and short survival in chronic lymphocytic Paris 6, Paris, France leukemia. Blood 2008; 111: 2246–2252. E-mail: [email protected] 5 Poncet D, Belleville A, t’kint de Roodenbeke C, Roborel de Climens A, Ben Simon E, Merle-Beral H et al. Changes in the expression of telomere maintenance genes suggest global telomere dysfunction in References B-chronic lymphocytic leukemia. Blood 2008; 111: 2388–2391. 6 Bouley J, Deriano L, Delic J, Merle-Be´ral H. New molecular markers 1 Blasco MA. The epigenetic regulation of mammalian telomeres. Nat in resistant B-CLL. Leuk Lymphoma 2006; 47: 791–801. Rev Genet 2007; 8: 299–309. 7 Siegl-Cachedenier I, Mun˜oz P, Flores JM, Klatt P, Blasco MA. 2 Palm W, De Lange T. How shelterin protects mammalian telomeres. Deficient mismatch repair improves organismal fitness and Annu Rev Genet 2008; 42: 301–334. survival of mice with dysfunctional telomeres. Genes Dev 2007; 3 Van Bockstaele F, Verhasselt B, Philippe´ J. Pronostic markers in 21: 2234–2247. chronic lymphocytic leukemia: a comprehensive review. Blood Rev 8 Maes J, Maleszewska M, Guillemin C, Pflumio F, Six E, Andre´- 2009; 23: 25–47. Schmutz I et al. Lymphoid-affiliated genes are associated with active 4 Roos G, Kro¨ber A, Grabowski P, Kienle D, Bu¨hler A, Do¨hner H et al. histone modifications in human hematopoietic stem cells. Blood Short telomeres are associated with genetic complexity, high-risk 2008; 112: 2722–2729.

Leukemia