Med Oncol (2012) 29:2083–2087 DOI 10.1007/s12032-011-0108-5
SHORT COMMUNICATION
Additional chromosomal abnormalities detected by array comparative genomic hybridization in AML Ana Rosa S. Costa • Sintia I. Belangero Maria Isabel Melaragno • Maria de Lourdes Chauffaille
•
Received: 11 July 2011 / Accepted: 1 November 2011 / Published online: 25 November 2011 Ó Springer Science+Business Media, LLC 2011
Abstract Improved outcome of acute myeloid leukemia (AML) depends on the better differentiation of subtypes to predict treatment response and the identification of new target for treatment. In this study, array comparative genomic hybridization (aCGH) was used to distinguish eight cases of AML cases. Validation was performed by FISH and quantitative genomic PCR. The aCGH revealed new large and small recurrent genomic imbalances, such as gains of 1p36, 10q26, 11p15, 20q13, 22q23, harboring many proto-oncogenes. These results better define genetically the studied cases and could be used to understand the multiple phenomena involved in leukemogenesis. Keywords Acute myeloid leukemia Chromosome aberration Array CGH
All authors contributed equally to this article. A. R. S. Costa M. L. Chauffaille (&) Division of Hematology and Hemotherapy, UNIFESP/Escola Paulista de Medicina, Rua Botucatu, 740, 38 andar, Sa˜o Paulo, SP CEP: 04023-900, Brazil e-mail:
[email protected] S. I. Belangero M. I. Melaragno Division of Genetics, UNIFESP/Escola Paulista de Medicina, Sa˜o Paulo, SP, Brazil
Introduction Chromosomal abnormalities detected by G-banding karyotype provide important information to classify and to treat acute myeloid leukemia (AML) [1, 2]. In fact, due to cytogenetics studies, this group of disorders is properly categorized, and patients are stratified into three different prognosis subgroups: favorable, intermediate and adverse, according to overall survival [1, 2]. However, despite the established clinical prognostic value of karyotype, wide differences are found among patients with similar cytogenetic results. This clinical divergence could be explained, partially, by G-banding resolution, which would not allow the detection of aberrations smaller than approximately 5 Mb. Additionally, a number of genes have been shown to be mutated in AML, such as FLT3 (internal tandem duplication or tyrosine kinase domain), NMP1, c-KIT and CEPBA, among others, all with prognosis importance [2] especially in normal karyotypes, but even in abnormal ones. Notwithstanding this, AML is characterized by chromosomal alterations that result in deletions or gains of genome regions. In this context, genomic technologies are useful tools to identify subgroups of the disease with distinct abnormalities undetected by the resolution of G-banding karyotype. The array comparative genomic hybridization (aCGH), a molecular cytogenetic methodology, has been used to evaluate congenital malformation syndromes and has become an essential and routine diagnostic tool in this area [3]. aCGH allows the detection of unbalanced alterations with high resolution (1 Mb or less) by co-hybridizing two differentially labeled DNAs to a known DNA segment spotted onto array, providing a genome-wide screen for imbalances. This methodology holds considerable promise also for AML [4]. This study was conducted in order to increase chromosomal
123
2084
abnormalities detection by using aCGH in addition to karyotyping in AML cases.
Patient and methods Eight cases of AML with known recurrent chromosomal alterations and dismal outcome were retrospectively selected: two cases with trisomy 8, five with -7/7q- and one with both alterations (Table 1). The median blast count in bone marrow (BM) was 80% (range, 67–97%). The study was approved by the UNIFESP/EPM Ethics Committee. Conventional G-banding on bone marrow metaphases was performed after short-term culturing (24 h), using standard procedures, and the karyotypes were described according to ISCN (2009) [5]. The AML genomic DNAs (gDNA) were extracted from BM cultured cells stored in Carnoy’s fixative, at -20°C. These archived bone marrow samples were washed in PBS, and the gDNA was isolated with the QIAamp DNA Blood Mini Kit according to manufacturer’s protocol (Qiagen, Valencia, CA). The DNA quality was analyzed by agarose gel electrophoresis and all samples contained high-molecular-weight DNA. Sex-mismatch reference gDNA (Promega, Madison, WI, USA) was used as normal reference for seven cases. In case 8, DNA from a sample in remission was used as reference to reduce the germline alterations effect. A commercially available 1 Mb resolution aCGH TM Spectral Chip 2600 (PerkinElmer, Boston, MA) was used, consisting of 2,632 Bacterial Artificial Chromosomes clones (BAC), 1 Mb spaced across the entire human genome. The procedure was carried out according to the manufacturer’s protocols. Each sample was run in duplicate using a dye-swap design. The arrays were analyzed by SpectralWare Software (PerkinElmer). The normalization method was local LOESS. The threshold for abnormalities was 2 standard deviations from the mean [6]. Whenever large-scale alterations (larger than 5 Mb, G-banding karyotype resolution) were found on aCGH analysis, they were validated by interphase FISH (iFISH), using the same fixed chromosome preparations, as employed for gDNA extraction. The following commercial probes were performed as manufacturer’s protocol: RUNX1/RUNX1T1, IGH/CCND1, MLL and BCR/ABL1 genes rearrangements, chromosome 8 centromere, 5q31 and 7q31 regions (Abbott Vysis, Downers Grove, IL). The cutoff levels for trisomy 8 (2.7%), del(7q31) (1,6%) and del(5q31) (3%) were established according to the iFISH patterns observed in a group of 4 normal control BM samples studied with the same probes.
123
Med Oncol (2012) 29:2083–2087
To validate small-scale aberrations (smaller than 5 Mb) found on aCGH analyses, quantitative PCR (qPCR) was performed using the Taqman Copy Number Assay (Applied Biosystems, Foster City, CA). Five regions were chosen between those overlapping areas with at least two adjacent BAC clones considered to be deviating in more than one case (Table 2): a 600 Kb duplication of 1pter1p36.23; a 720 Kb duplication of 10q26; a 1,5 Mb duplication of 11p15; 1,7 Mb duplication of 20q13; 700 Kb duplication of 22q13. The qPCR was carried out as manufacturer’s protocol, using primers to genes: DVL1 (1p36.23 region), TUBGCP2 (10q26 region), BRSK2 (11p15 region), HRH3 (20q13 region), TTLL8 (22q13 region). Briefly, the comparative Ct method was used. The housekeeping gene was RNase P, and the reference DNA was the same used to aCGH experiment. Every test and reference samples were run in four replicates. The experiments were run on 7500 Fast equipment, and the results were analyzed by 7500 v.2.0.3 and by CopyCaller software (Applied Biosystems).
Results Delineation of breakpoints by aCGH Table 1 shows the abnormalities detected in each case by G-banding and aCHG. The aCGH analysis confirmed most of the large-scale aberrations identified by G-banding karyotype. Some abnormalities were not revealed by aCGH: ?8 and del(7)(q31) in case 6 and ?22 in case 8, likely due to a small subclone population. The 7q deletions were confirmed by aCGH but in a different position: two interstitial deletions, del(7)(q21.1q21.3) and del(7)(q31.1q35) in case 1; an interstitial deletion del(7)(q21.2q22) in case 2. Also in case 2, the breakpoint in 5q deletion proved to be in a different position, more proximal (Table 1). Two cases presented large-scale abnormalities detected only by aCGH: del(4)(q21q24), del(4)(q28.2q32), del(5) (q35), del(6)(q12q22.1), del(14)(q11.2q24.3) and del(21) (q21.3q22.12), in case 2 and chromosomes 8 and 14 trisomies, in addition to duplication of 11q23 and loss of 4p15.2 in case 3 (Table 1). iFISH validations Large scales abnormalities of cases 2, 3, 6 and 8 were validated by iFISH, whose aCGH and G-banding results showed divergences. In case 2, probes to regions 5q31 and RUNX1 validated aCGH result of deletion in both regions not evident by G-banding. In case 3, centromeric probe for chromosome 8, probes to 14q32 (IGH) and 11q23 (MLL)
78
47
53
59
57
89
29
2
3
4
5
6
7
8
c
b
F
M
AML-MDS Ra
t-AMLb
AML with recurrent genetic abnormalities
AML without maturation
AML-MDS R
AML with recurrent genetic abnormalities
F
F
F
M
M
M
AML-MDS Ra
AML with minimal differentiation
Sex
WHO Diagnosis
47,XX, ? 8,inv(16)(p13.1q22) [11]/47,XX,inv(16) (p13.1q22), ? 22[4]/ 46,XX,inv16[1]/46,XX[2]
46,XX, ? 8[20]
47,XX,del(7)(q31), ? 8[3]/ 46,XX[13]
45,XY,t(3;3)(q21;q26),-7[15]
45,XY,-7[6]
47,XY,del(7)(q32),add(11)(p15), ? i(?)[cp9]/46,XY[1]
44 * 46,XY,del(5)(q32), del(7)(q32),del(11)(q22)[cp5]
46,XY,del(7)(q32)[11]
G-banding analysis
t-AML Therapy-related myeloid neoplasms IC nuclei interphase cells
AML MDS R AML with myelodysplasia-related changes
74
1
a
Age years
Case
Table 1 Clinical aspects of AML cases, large-scale aCGH result and FISH validation
arr 8p23.3q24.3(1,240,000–146,040,000)93
arr 8p23.3q24.3(1,720,000–146,000,000)93
No large-scale aberration
arr 7p22.3-q36.3(170,000–158,380,000)91
arr 7p21.3-q31.1(13,230,000–114,220,000)91
arr 4p15.2(130,000–26,000,000)91,8p23q24.3 (660,000–146,110,000)93, 11q22(104,400,000–134,600,000)93,14q11.2 (18,600,000–106,180,000)93
arr 1p32.2p21.3(50,460,000 –92,650,000)91,4q21q24(84,800,000 –107,740,000)91,4q28q32(130,000,000 –161,500,000)91,4q35(184,380,000 –190,940,000)91,5q11.2(54,380,000 –169,200,000)91,6q12q22.1(69,440,000 –114,060,000)91,7q21.2q22(88,390,000 –106,300,000)91,11p15.1(6,720,000 –15,800,000)91,11q14.1(78,290,000 –134,600,000)91,14q11.2q24.3(18,600,000 –76,600,000)91,21q21.3q22.12(29,800,000 –35,300,000)91
arr 7q21.1q21.3(84,680,000– 92,000,000)91, 7q31.1q35(108,000,000– 142,000,000)91
aCGH large-scale alteration
BCR/ABL1 probe- three copies of BCR gene (22q) in 17% ICc
No need further evaluation
CEP 8 probe- three copies of Chr 8 in 3% ICc (cutoff 2,7%) FISH 7q31del(7)(q31) in 2,5% ICc (cutoff 1,6%)
No need further evaluation
No need further evaluation
CEP 8 probe- three copies of Chr 8 in 88% ICc; IGH/CCND1 probethree copies of gene on Chr 14 in 60% ICc; MLL probe – three copies of MLL in 92% ICc
RUNX1/RUNX1T1 probe- RUNX1 deletion (21q) in 82,5% IC#
5q31 probe- del(5)(q31) in 72,5% ICc
No need further evaluation
FISH validation
Med Oncol (2012) 29:2083–2087 2085
123
2086
Med Oncol (2012) 29:2083–2087
Table 2 Small aberrations found in two or more samples Chromosome band
Position (hg 18) (bp)
Size (Kb)
qPCR primers to genes
Validation
Known gene(s) of interest to cancera
1p36.23
1,020,000–1,610,000
600
DVL1
Gain. Cases: 2, 4, 6
DVL1, AURKAIP1,CDCL2 and others
10q26
134,500,00–135,220,000
1,200
TUBGCP2
Gain. Cases: 2, 4, 5, 6, 7
ADAM8
11p15
190,000–1,630,000/ 6,720,000
1,500–5,000
BRSK2
Gain. Cases: 6, 7
HRAS, IGF2, ILK, NUP98 and others
20q13
61,320,000–63,100,000
2,000
HRH3
Gain. Cases: 1, 2, 4, 5, 6, 7
TNFRSF6B, RTEL1, EEF1A2, PTK6, BIRC7, NTSR1, CDH4, DIDO1 and others
22q13
48,800,000–49,500,000
700
TTLL8
Gain. Cases: 2, 6
TYMP, HDAC10, PIM3, RABL2B, and others
bp base pair, Kb kilobase a
According to the UCSC Genome Browser, http://genome.ucsc.edu
regions were revealed and three copies of chromosome 8, three copies of regions 14q32 and 11q23 were not revealed by G-banding. Finally, iFISH was performed to investigate small clones in cases 6 and 8. In case 6, iFISH revealed ?8 in 3% of the nuclei examined and deletion of 7q31 in 2.5% of the nuclei examined. In case 8, three copies of BCR gene were found in 17% of the nuclei examined, revealing ?22 not evident by aCGH analysis. Detection of hidden genomic imbalances All cases presented many small-scale alterations, but only the recurrent regions chosen to be validated as described above. All of them were found as a gain. The qPCR were able to confirm most of the aberrations: 22 out of 35 reactions, including more than one copy gain in 1p36 region in two cases: 2 and 6 (Table 2). In case 5, aCGH analysis revealed that t(3;3)(q21;q26.2) is unbalanced. An almost 3 Mb deletion in region 3q26 (175,860,000–178,710,000), near by EVI1 gene, was revealed by aCGH.
Discussion Nowadays, chromosomal abnormalities revealed by G-banding karyotype are the most important AML prognostic factor [1, 2]. However, metaphase cytogenetics solely cannot reveal and discriminate all biological heterogeneity observed in AML cases. PCR and FISH analysis have their place in AML diagnostic algorithm and are also important tools. Nevertheless, they are targeted methods that require prior knowledge of chromosomal region(s) of interest or probable rearrangements or mutations to be searched. The aCGH appears to be a valuable highthroughput tool to study additional cryptic genomic imbalances [3] as here shown.
123
In this study, a BAC-based aCGH confirmed abnormalities already detected by G-banding karyotype and allowed to better define them. The aCGH revealed new breakpoints to 7q and 5q deletions, as well as added new large- and small-scale aberrations to these cases. Regarding to large-scale abnormalities, aCGH revealed additional aberrations in cases 2 and 3, some of them in regions of known AML-related genes: deletion of RUNX1 in case 2 and duplication of MLL, in addition to ?8 and ?14 in case 3. MLL amplifications is a recurrent abnormality in therapy-related AML, as in case 3 and have been associated with karyotype complexity, extremely poor survival, and mutations in TP53 [7]. Case 4 illustrates an advantage of this methodology as a karyotyping method. aCGH analysis revealed monosomy 7, but metaphases were not found in the first cytogenetic analysis performed, a problem that would be overcome by the aCGH test if it had been available at that moment. It is well known that FISH could easily show monosomy 7 if a panel of probes for the most frequent abnormalities had been used initially, but aCGH, as a whole genome screening tool, showed advantage in this case. Concerning to small clones or subclones, represented here by cases 6 and 8, aCGH revealed only large clonal expansions irrespective of cell growth. As a common sense, this methodology is not able to detect abnormalities in less than 20–35% of cells. This study analyzed cases with at least 67% of leukemic cells in BM in order to bypass this pitfall. However, regardless the large amount of blast cells analyzed, in cases 6 and 8, the small clones identified by karyotype were not detected by aCGH, indicating that this method was not able to reveal subclones not significantly expanded, information proved by FISH analysis in these cases (Table 1). Improved outcome of AML depends on the better differentiation of subtypes to predict treatment response and the identification of new target for treatment. In unbalanced
Med Oncol (2012) 29:2083–2087
karyotype AML, whose pathogenesis is unclear, aCGH seems to be very useful not only to detect large aberrations but also to elucidate many cryptic alterations, revealing new regions to serve as novel targets for molecular treatment or as prognostic factors. According to literature, many small aberrations would be expected, as we found in all of these cases. Here we focused on the small aberrations in the recurrent regions, all of them harboring some amount of genes with interest in cancer research. Chromosome 1p36 abnormality is a common event in many types of cancer. This region seems harboring more than one cancer-related gene, turning it into an interesting region for further studies. The DVL1 gene located in this region acts as a central role to propagate Wnt signaling by canonical and non-canonical pathways [8]. This signaling pathway plays a critical role in many biological process and its deregulation is associated with cancer and leukemia stem cells development as well [9]. Garnis et al. [10] studied some candidate genes in this pathway on lung tumors samples and found concordance between amplification by aCGH and significant overexpression of DVL1. Other relevant regions found recurrently altered in this study were as follows: 10q26.3, harboring ADAM8 gene, implicated in tumor progression and invasiveness whose overexpression was associated with myelodysplastic syndromes advanced stages [11]; 20q13, a known region deleted in many hematological neoplasias, and regions 11p15 and 22q13, harboring many known proto-oncogenes. There are still limitations to the use of aCGH methodology on clinical settings: it is not able to identify balanced rearrangements, which are an important AML abnormalities; it does not allow to detect a small clones or subclones as shown, and finally, aCGH detects copy number variation (CNV), whose meaning is still of unclear significance [12]. This later one could be surpassed by the use of patient germline tissue as the reference DNA [13], as used here in case 8, although some genes included in this CNV region could be de novo alteration found only in cancer cells [13], acting as a disease progression contributors or as an initiation event. In conclusion, this study calls attention to the importance of aCGH method as an AML diagnostic tool to better distinguish its genomic alterations and could be used to understand the multiple phenomena involved in leukemogenesis. Other studies should be conducted to establish which of these aCGH alterations found here have a real
2087
prognostic value and could be used as a prognostic tool or even as new target regions for therapy. Acknowledgments This work was supported by Fundac¸a˜o de Amparo a Pesquisa do Estado de Sa˜o Paulo (FAPESP 07/55665-6), Sa˜o Paulo, Brazil. Conflict of interest report.
All authors have no conflict of interest to
References 1. Grimwade D, Walker H, Oliver F, et al. The importance of diagnostic cytogenetics on outcome in AML: analysis of 1, 612 patients entered into the MRC AML 10 trial. The Medical Research Council Adult and Children’s Leukaemia Working Parties. Blood. 1998;92:2322–33. 2. Mrozek K, Bloomfield CD. Chromosome aberrations, gene mutations and expression changes, and prognosis in adult acute myeloid leukemia. Hematol Am Soc Hematol Educ Program. 2006;169–177. 3. Shinawi M, Cheung SW. The array CGH and its clinical applications. Drug Discov Today. 2008;13:760–70. 4. Maciejewski JP, Mufti GJ. Whole genome scanning as a cytogenetic tool in hematologic malignancies. Blood. 2008;112: 965–74. 5. Shaffer LG, Slovak ML, Campbell LJ, editors. ISCN 2009: an international system for human cytogenetic nomenclature 2009. Basel: S. Karger; 2009. 6. Iafrate AJ, Feuk L, Rivera MN, Listewnik ML, Donahoe PK, Qi Y, Scherer SW, Lee C. Detection of large-scale variation in the human genome. Nat Genet. 2004;36:949–51. 7. Andersen MK, Christiansen DH, Kirchhoff M, et al. Duplication or amplification of chromosome band 11q23, including the unrearranged MLL gene, is a recurrent abnormality in therapyrelated MDS and AML, and is closely related to mutation of the TP53 gene and to previous therapy with alkylating agents. Genes Chromosomes Cancer. 2001;31:33–41. 8. Gao C, Chen YG. Dishevelled: the hub of Wnt signaling. Cell Signal. 2010;22:717–27. 9. Reya T, Clevers H. Wnt signaling in stem cells and cancer. Nature. 2005;434:843–50. 10. Garnis C, Campbell J, Davies JJ, Macaulay C, Lam S, Lam WL. Involvement of multiple developmental genes on chromosome 1p in lung tumorigenesis. Hum Mol Genet. 2005;14:475–82. 11. Vasikova A, Budinska E, Belickova M, Cermak J, Bruchova H. Differential gene expression of bone marrow CD34? cells in early and advanced myelodysplastic syndrome. Neoplasma. 2009;56:335–42. 12. Feuk L, Carson AR, Scherer SW. Structural variation in the human genome. Nat Rev Genet. 2006;7:85–97. 13. Starczynowski DT, Vercauteren S, Sung S, et al. Copy number alterations at polymorphic loci may be acquired somatically in patients with myelodysplastic syndromes. Leuk Res. 2010;35: 444–7.
123
