Genomic Analysis Reveals Few Genetic Alterations in Pediatric Acute Myeloid Leukemia

Genomic analysis reveals few genetic alterations in pediatric acute myeloid leukemia

Ina Radtkea, Charles G. Mullighana, Masami Ishiia, Xiaoping Sua, Jinjun Chenga, Jing Mab, Ramapriya Gantia, Zhongling Caia, Salil Goorhaa, Stanley B. Poundsc, Xueyuan Caoc, Caroline Obertb, Jianling Armstrongb, Jinghui Zhangd, Guangchun Songa, Raul C. Ribeiroe, Jeffrey E. Rubnitze, Susana C. Raimondia, Sheila A. Shurtleffa, and James R. Downinga,1

Departments of aPathology, cBiostatistics, and eOncology, and the bHartwell Center for Bioinformatics and Biotechnology, St. Jude Children’s Research Hospital, 262 Danny Thomas Place, Memphis, TN 38105; and dCenter for Biomedical Informatics and Information Technology, National Cancer Institute, National Institutes of Health, 2115 E. Jefferson Street, Rockville, MD 20892

Edited by Janet D. Rowley, University of Chicago Medical Center, Chicago, IL, and approved June 11, 2009 (received for review March 20, 2009) Pediatric de novo acute myeloid leukemia (AML) is an aggressive disease with an inferior treatment outcome compared to ALL. malignancy with current therapy resulting in cure rates of only 60%. Despite the introduction of new drugs and allogeneic bone marrow To better understand the cause of the marked heterogeneity in transplantation, overall cure rates in most contemporary treatment therapeutic response and to identify new prognostic markers and protocols remain below 60% (10–12). therapeutic targets a comprehensive list of the genetic mutations that Like pediatric ALL, de novo AML is a heterogeneous disease underlie the pathogenesis of AML is needed. To approach this goal, composed of different genetic subtypes with distinct clinical fea- we examined diagnostic leukemic samples from a cohort of 111 tures and responses to contemporary therapies. The best charac- children with de novo AML using single-nucleotide-polymorphism terized subtypes include the core-binding factor leukemias microarrays and candidate gene resequencing. Our data demonstrate (t(8;21)[RUNX1(AML1)-RUNX1T1(ETO)] and inv(16)/ that, in contrast to pediatric acute lymphoblastic leukemia (ALL), de t(16;16)[CBF␤-MYH11]), cases with rearrangements of the MLL novo AML is characterized by a very low burden of genomic alter- gene on chromosome 11q23, cases with distinct morphology in- ations, with a mean of only 2.38 somatic copy-number alterations per cluding acute promyeloctic leukemia with t(15;17)[PML-RARA] leukemia, and less than 1 nonsynonymous point mutation per leu- and acute megakaryoblastic leukemia (FAB-M7), and cases with kemia in the 25 genes analyzed. Even more surprising was the normal cytogenetics. Although some cooperating lesions have been observation that 34% of the leukemias lacked any identifiable copy- identified in AMLs, including point mutations or CNAs of NRAS, number alterations, and 28% of the leukemias with recurrent trans- KRAS, FLT3, KIT, PTPN11, RUNX1, MLL, NPM1, CEBPA, and locations lacked any identifiable sequence or numerical abnormali- TP53 (13–17), the full complement of cooperating lesions remains ties. The only exception to the presence of few mutations was acute to be defined. The identification of the complete complement of megakaryocytic leukemias, with the majority of these leukemias genetic lesions within AML will not only improve our understand- being characterized by a high number of copy-number alterations but ing of the molecular pathology of acute leukemia, but should also rare point mutations. Despite the low overall number of lesions across directly impact diagnosis and risk stratification, and may lead to the the patient cohort, novel recurring regions of genetic alteration were identification of new targets against which novel therapies can be identified that harbor known, and potential new cancer genes. These developed. data reflect a remarkably low burden of genomic alterations within We report the results of a study of genome-wide DNA CNAs, pediatric de novo AML, which is in stark contrast to most other human LOH, and targeted gene resequencing analyses on primary leuke- malignancies. mic blasts from 111 pediatric AML patients. Our data demonstrate that, in contrast to pediatric ALL, de novo AML is characterized copy number alterations ͉ single-nucleotide-polymorphism (SNP) ͉ by a very low burden of genomic alterations. Despite the low microarray ͉ candidate gene resequencing ͉ loss-of-heterozygosity (LOH) number of lesions, however, unique recurring regions of genetic alteration were identified that harbor known, and potential new eukemia results from multiple genetic and epigenetic alterations cancer genes. Moreover, the spectrum of CNAs and sequence Lwithin hematopoietic stem cells (HSCs) or progenitors that alter mutations was found to vary significantly across the different their normal self-renewal, proliferation, differentiation, and apop- genetic subtypes of AML. totic pathways (1–3). These alterations include point mutations, Results gene rearrangements, deletions, amplifications, and a diverse array of epigenetic changes that influence gene expression. For most AML Leukemic Cells Contain Few Copy-Number Alterations. As an initial leukemias the full complement of oncogenic lesions remains to be approach to define the total complement of genetic lesions in defined. pediatric de novo AML, we performed high resolution genome- To define the lesions in acute leukemia, we recently used wide analysis on leukemic blasts from diagnostic bone marrow single-nucleotide-polymorphism (SNP) microarrays to perform aspirates from 111 patients using both Affymetrix 100K and 500K genome-wide DNA copy-number and loss-of-heterozygosity SNP microarrays (combined resolution of 615K). The leukemias (LOH) analyses on primary leukemic blasts from pediatric patients with acute lymphoblastic leukemia (ALL) (4, 5). These studies Author contributions: I.R., C.G.M., S.A.S., and J.R.D. designed research; I.R., C.G.M., M.I., identified a high frequency of genetic alterations of key regulators X.S., J.C., R.G., Z.C., S.G., J.A., and S.C.R. performed research; X.S., S.B.P., and J.Z. contributed of B lymphoid development and cell cycle in B-progenitor ALL. new reagents/analytic tools; I.R., C.G.M., M.I., X.S., J.C., J.M., S.G., S.B.P., X.C., C.O., J.Z., G.S., More recently, similar approaches have been used to explore the R.C.R., J.E.R., S.C.R., and S.A.S. analyzed data; and I.R. and J.R.D. wrote the paper. type of copy-number alterations (CNAs) in adult myeloid malig- The authors declare no conflict of interest. nancies (6–9), although these studies have used relatively low This article is a PNAS Direct Submission. resolution platforms. Freely available online through the PNAS open access option. We have now extended these analyses to pediatric de novo acute 1To whom correspondence should be addressed. E-mail: [email protected]. myeloid leukemia (AML). AML comprises 15–20% of the acute This article contains supporting information online at www.pnas.org/cgi/content/full/ leukemias diagnosed in this age group and remains a challenging 0903142106/DCSupplemental.

12944–12949 ͉ PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903142106 Downloaded by guest on September 30, 2021 A t(8;21) inv(16) MLL t(15;17) M7 Miscellaneous Normal 1

4 5

6 7 8 9 10 11 SNPs by chromosome 12 13 14 15-16 17-19 20-22 X patients -2 0 +2 log2ratio B Amplifications C Deletions

Deletions

3 genes

4 genes

22 genes

RUNX1T1

CCDC26 12 genes (*CDKN2A) TUSC1

13 genes (*XPA)

7 genes

4 genes (*MLL)

31 genes ABCC4

5 genes (*MYH11)

88 genes 17 genes (*CBFB)

81 genes No gene (*ERG, *TMPRSS2) 410 genes -0.85 -1.1 -1.6 -2.6 -4.7 -8.8 10 10 10 10 10 10 10-0.810-110-1.4 10-2.2 10-3.8 10-7

0.25 False Discovery Rate 0.25 False Discovery Rate

Fig. 1. DNA copy-number abnormalities in pediatric de novo AML. (A) Summary of CNA (log2 ratio) from a combined 100K and 500K Affymetrix SNP array analysis of diagnostic leukemia cells from 111 pediatric de novo AML patients. Each column represents a case and the 615K SNPs are arranged in rows according to chromosomal location. Cases are arranged by subgroup. Diploid regions are white. Blue represents deletion, red amplification (see color scale). Gross changes can be observed for example in chromosome 8 (10 cases with trisomy 8). (B) GISTIC (18) analysis of copy-number gains. (C) GISTIC analysis of copy-number losses. False discovery rate q values are plotted along the x axis with chromosomal position along the y axis. Altered regions with significance levels exceeding 0.25 (marked by vertical green line) are deemed significant. Five significant regions of amplification and 13 significant regions of deletion were identified. Chromosomal position and relevant genes are shown for each significant region on the right side of the plots. Genes indicated in blue are associated with known translocations, genes marked with * are cancer census genes (19).

included a representation of the different genetic subtypes of the Gene Targets of Recurrent Copy-Number Alterations. Recurrent CNAs pediatric de novo AML (in SI Appendix, Tables S1 and S2). Germ within a patient cohort can be used to identify alterations of line DNA was available for 65 of the patients allowing a definitive potential biological significance (driver versus passenger muta- identification of somatically acquired CNAs. Two-hundred seven tions). Surprisingly, when large regions (whole chromosomes or CNAs were detected across the cohort with a mean number of chromosome arms) of gains or losses were excluded the majority of CNAs/patient of 2.38 (range 0–45), with no significant difference the remaining lesions were nonrecurrent, being identified in only a in the average number of gains (1.32, range 0–41) and losses (1.06, single patient (SI Appendix, Table S2). Using the genomic identi- range 0–12) (Fig. 1 and SI Appendix, Table S2). The frequency of fication of significant targets in cancer (GISTIC) algorithm (18), CNAs was similar across the various AML genetic subtypes with the only 5 significant regions of gains and 13 regions of deletion were exception of FAB-M7, which had an average of 9.33 CNAs/patient, altered more often than would be predicted by chance (Fig. 1 B and with the majority consisting of gains (SI Appendix, Table S2, P ϭ C). These included 1 broad lesion (gain of chromosome 22), 4 focal 0.013). Excluding FAB-M7 leukemias the average number of lesions that were contained within broader lesions [8q24.21 (0.091 CNAs/de novo AML patient was 1.76. Notably, 34% of the patients Mb), 7p21.3 (3.523 Mb), 7q36.1 (1.372 Mb), and 9q22.3 (0.943 lacked any identifiable CNAs (SI Appendix, Table S2). Importantly, Mb)], and 13 focal lesions containing Ͼ25 genes (3 loci), 6–25 genes no association was detected between clinical outcome and the (3 loci), 2–5 genes (3 loci), a single gene (3 loci), or lacking an number of gains, losses, total CNAs, or the amount of the genome annotated gene (1 locus) (Fig. 1, Fig. 2, and SI Appendix, Table S3). altered in either univariate or multivariate analysis. In addition, we identified 24 other recurrent lesions where the MEDICAL SCIENCES

Radtke et al. PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 ͉ 12945 Downloaded by guest on September 30, 2021 tel cent

8q24.21 50 kb ORF e1e1a e2 e3 e4 CCDC26

lincRNA

AML-AE-#20 AML-N-#15 AML-misc-#5 AML-AE-#8 n=11 NB-4 HL60

Fig. 2. Amplification of CCDC26 in pediatric AML. The genomic organization of CCDC26 is illustrated relative to the telomere (tel) and centromere (cent) of chromosome 8, with exons labeled by lowercase e, and an alternative transcript initiating from exon e1a. The vertical blue lines show the location of the SNPs on combined 100K and 500K Affymetrix arrays. The putative ORF encoded by exons e3 and e4 is shown by a dotted line with the arrowhead illustrating the direction for transcription. The vertical red arrow marks the integration site found in a retroviral integration screen of retinoic acid resistant myeloid cell lines (20). The green box represents the 10-kb lincRNA, identified in ref. 22. The extent of amplification across this genomic locus for each case is illustrated by a horizontal red line with the case ID number next to the line or the number of cases that contained large amplifications that included the entire locus. Two AML cell lines, HL-60 and NB-4, were also found to have focal amplification of this locus.

minimal altered regions (MAR) was less than 20 Mb, but fell below the AML-associated t(6;11) translocations (27), and PRDM5 (28), statistical significance by the GISTIC algorithm (SI Appendix, Table a putative tumor suppressor. The CNAs of RUNX1T1 were ob- S4). Together, these 41 recurrent focal lesions contained a total of served in three t(8;21) leukemias, whereas the MLLT4 alteration 1,158 genes (290 from GISTIC peaks, 868 from other MARs), of was detected in 2 patients, 1 with normal cytogenetics and the other which only 30 (2.6%) were contained within the Cancer Gene with a complex karyotype that did not include a detectable struc- Census listing (http://www.sanger.ac.uk/genetics/CGP/Census/) tural alteration of 6q (SI Appendix,Fig.S1). (19), suggesting that new AML cancer genes are likely to exist In addition to RUNX1T1 and MLLT4, 7 other genes that are the within the identified MAR. Twenty-one of the 30 cancer consensus targets of AML-associated chromosomal translocations were af- genes have been previously implicated in AML as targets of fected by recurrent CNAs, including MYH11, CBF␤, MLL, NSD1, translocations or sequence mutation including CBFB, CDKN2A, MLF1, ERG, and MYST4 (SI Appendix, Tables S3 and S4). Prior ELL, ERG, ETV6, FLI1, GMPS, JAK3, LYL1, MLF1, MLL, studies have demonstrated that focal micro deletions and amplifi- MLLT1, MLLT4, MYH11, MYST4, NPM1, NSD1, PBX1, RUNX1, cations can occur near the breakpoints of chromosomal transloca- SH3GL1, and WT1. A number of other cancer consensus genes tions (4, 29). Consistent with these observations, the CNAs of were contained within nonrecurrent CNAs suggesting that a subset MYH11 and CBF␤ were observed in cases with inv(16)/t(16;16), and of the nonrecurrent lesions may provide a selective advantage to the the CNAs of MLL were seen in a subset of cases with MLL leukemic cell and thus constitute driver mutations (SI Appendix, translocations (SI Appendix, Fig. S1 and S2). On the basis of these Tables S5 and S6). observations, we examined whether any other recurrent CNAs of When broad and focal CNAs were combined the most common genes were the result of cryptic translocations. Our analysis iden- affected region was on chromosome 8, band q24, which showed a tified 4 leukemias that expressed the t(5;11)-encoded NUP98-NSD1 copy-number gain in 14% of patients (Fig. 2 and SI Appendix, Table chimeric transcript, with 2 having CNAs adjacent to 1 or both genes S6). Focal gains were identified in 4 patients, and in an additional (SI Appendix,Fig.S2) and 2 cases that expressed the t(6;11)- 11 patients an increase in copy number was observed either as a encoded MLL-MLLT4(AF6) chimeric transcript, with associated result of a larger region of chromosomal gains (1 case), or as the CNAs involving both genes (SI Appendix,Fig.S2). These data result of trisomy 8. No significant difference in the frequency of this indicated that at least 14% of the cytogenetically normal or alteration was detected across the different AML genetic subtypes. miscellaneous karyotype subgroups within this cohort contained a Interestingly, the MAR targets a locus that was originally identified cryptic translocation. Whether the CNAs involving MLF1, ERG, in a forward genetic screen to identify genes that are required for and MYST4 are also the result of cryptic translocations remains to retinoic acid (RA)-induced myeloid cell differentiation (20, 21). be determined. Retroviral integration into this locus disrupted RA-induced differentiation. The region contains a putative gene referred to as Copy Neutral Loss of Heterozygosity. The genotypes generated by the CCDC26, and a highly conserved large intervening noncoding Affymetrix SNP array platform enable the detection of regions of RNA of unknown function (22). Which of these transcripts nor- somatic copy neutral-loss of heterozygosity (CN-LOH), which may mally functions in myeloid differentiation, and whether the iden- identify reduplication of mutated or aberrantly methylated genes tified copy-number changes alter their expression and function that contribute to tumorigenesis. Prior publications have identified remain to be determined. CN-LOH of 11p (involving WT1), 13q (FLT3), and 19q (CEPBA) The MARs in 8 other recurrent CNAs were each limited to a in AML (30, 31), and have suggested that CN-LOH is the most single gene in at least 1 patient, thus identifying the gene as a target frequent lesion in myeloid malignancies, occurring in up to 48% of of the lesion (SI Appendix, Table S6). These included monoallelic cases (6, 7). However, these studies have often failed to compare the deletions involving RUNX1T1 (ETO), a known target of the AML- SNP genotypes of the tumor to the patient’s own constitutional associated t(8;21) translocations, MLL involved in 11q23 translo- DNA and consequently have been unable to definitively distinguish cations, FAM20C, which is expressed in hematopoietic cells and is somatic CN-LOH from inherited homozygosity. We initially per- the target of mutations in Raine’s syndrome (23) a lethal osteo- formed paired CN-LOH analysis for 60 patients with matched sclerotic bone disease, and the putative tumor suppressors TUSC1 constitutional and leukemia cell DNA and identified only 6 leuke- (24) and BCOR (25), and amplifications of ABCC4, encoding a mias that contained somatic CN-LOH that were defined by 3 or multidrug resistance membrane protein (26), MLLT4, a target of more contiguous SNPs (SI Appendix, Fig. S3 and Table S7). These

12946 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903142106 Radtke et al. Downloaded by guest on September 30, 2021 Fig. 3. Integrated analysis of copy-number alterations, CN-LOH, and point mutations. Each column illustrates the results of the integrated mutational analysis for a single patient, with the patients grouped into the 7 genetic AML subtypes, which are color coded as illustrated. The rows depict the presence of mutations from sequence analysis (Top 13 rows) or copy-number alterations (Bottom 2 rows). The presence of a mutation in a gene is shown as a box, with the type of mutation indicated by the color of the box as shown. The presence of copy-number alterations is shown in the Bottom 2 rows with ampliﬁcations in red, and deletions in green, and the intensity of the color corresponding to the number of copy-number alterations according to the scale shown at bottom right.

leukemias included 5 cases with large regions of CN-LOH (chro- of genes within the RAS-signaling pathway or upstream receptor mosome 21, n ϭ 1; 11p, n ϭ 2; 9p, n ϭ 1; and 15q, n ϭ 1) along with tyrosine kinases (NRAS, n ϭ 26; KRAS, n ϭ 2; PTPN11, n ϭ 3; a single case with a focal region of CN-LOH on chromosome 8q11 FLT3-ITD, n ϭ 15; and KIT, n ϭ 5). Consistent with previously that contained no genes. published data, mutations of NRAS were frequent in the core We next performed unpaired CN-LOH analysis for the 51 binding factor (44%) and MLL rearranged AMLs (24%) (33), but patients that lacked constitutional DNA by using the 60 constitu- were uncommon in other AML subtypes (12%) (P ϭ 0.015) (Fig. tional DNA samples from our AML cohort as a reference pool to 3 and SI Appendix, Table S2). By contrast, mutations in FLT3 were eliminate common regions of inherited homozygosity (32). Because only found in acute promyelocytic leukemia and in patients with a this approach can fail to exclude private regions of inherited LOH, normal or miscellaneous karyotype (Fig. 3, P ϭ 0.0015). The we decided to limit our calls to regions of CN-LOH that were Ͼ20 mutations in these genes were heterozygous with the exception of Mb, thus significantly improving the accuracy of identifying true 4 leukemias that contained homozygous FLT3 mutations associ- lesions (SI Appendix,Fig.S3andS7). When the unpaired and paired ated with chromosome 13 CN-LOH. Although mutations in these analyses were combined, we identified somatic CN-LOH in only genes are usually mutually exclusive (33), two t(8;21) containing 13% of the entire cohort, including chromosome 13 (n ϭ 4), 11p AMLs had mutation in both NRAS and KIT (Fig. 3). (n ϭ 3), 9p (n ϭ 2), and 6p, 8q, 15q, 17q, and chromosome 21 (1 Mutations in CEBPA, RUNX1, ETV6, and MLL partial tandem patient each). The 4 leukemias with chromosome 13 CN-LOH had duplications (MLL-PTD) were also common (Fig. 3 and SI Appen- homozygous FLT3-ITD mutations (see below), and the two leuke- dix, Table S2) and were more frequent in AMLs with either a mias with 9p CN-LOH had homozygous deletions of CDKN2A/B normal or miscellaneous karyotype. For patients with matched that were included in the regions of LOH. Thus, somatically constitutional DNA, we were able to show that each mutation was acquired CN-LOH occurs in 13% of de novo AML, with many of somatic except for CEBPA, which was germ line in 3 of 6 patients the identified lesions targeting genes previously shown to be (SI Appendix, Table S2). Germ line mutations in CEBPA have been involved in the pathogenesis of AML. previously identified in rare pedigrees of familial AML (34). Point mutations in the other analyzed AML-associated cancer genes were Integration of Sequence Mutation and Copy-Number Alteration Data. To rare in this patient cohort (Fig. 3 and SI Appendix, Table S2). gain further insights into the complement of oncogenic lesions in Importantly, 42% of patients had no point mutations in the 25 AML, we sequenced genes previously found to be mutated in AML genes analyzed, 38% had only 1 gene mutated, 13% had 2 genes (AML-associated cancer genes, including NRAS, KRAS, PTPN11, mutated, and only 7% had 3 genes with point mutations. Moreover, FLT3, KIT, RUNX1, CEBPA, ETV6, NPM1, TP53, CDKN2A/B, the frequency of mutations varied across the AML genetic subtypes GATA1, and MLL), along with a subset of the genes targeted by with a higher number of lesions observed in the cases with normal CNAs in this cohort (CCDC26, FAM20C, TUSC1, ERG, IKZF1, or miscellaneous karyotypes as compared to the other subtypes PAXIP1, PTEN, FBXW7, BTG1, XRCC2, BRAF, and LYL1). (1.36 sequence mutations/patients in normal or miscellaneous Surprisingly, nonsynonymous somatic sequence mutations were karyotypes versus 0.54/patient in other subtypes, P Ͻ 0.0001). only identified in the AML-associated cancer genes (Fig. 3 and see Integration of the CNA/CN-LOH analysis and candidate gene SI Appendix, SI Text). resequencing revealed several important findings (Fig. 3 and SI Forty-seven leukemias contained somatic activating mutations Appendix, Table S2). First, 28% of leukemias with recurrent trans- MEDICAL SCIENCES

Radtke et al. PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 ͉ 12947 Downloaded by guest on September 30, 2021 A Acute Myeloid Leukemia Cytogenetics Cytogenetics Cytogenetics + 615K SNP analysis + 615K SNP array + sequence mutation analysis

B Acute Lymphoblastic Leukemia Cytogenetics Cytogenetics + 615K SNP analysis

Fig. 4. Spectrum of number of lesions per case in pediatric AML and ALL. Percentage of cases containing zero (normal), 1, 2, 3, 4, 5, or Ͼ5 alterations based on cytogenetics, cytogenetics plus CNAs using 615K SNP arrays, or cytogenetics, 615K SNP arrays, and targeted gene resequencing of 25 candidate genesinA AML (n ϭ 111 cases) and B ALL (n ϭ 212, cases from refs. 4, 5).

locations lacked any identifiable sequence or numerical abnormal- It is generally believed that the large number of mutations within ities. Second, although the FAB-M7 leukemias had the highest cancer cells arise either from inherent genomic instability or as the number of CNAs per case, they rarely contained point mutations, result of a single mutational crisis. Although we have performed with only a single patient having a GATA1 mutation. Third, the genomewide analysis for only a single type of mutation (CNAs), and leukemic cells from AML patients with normal or miscellaneous have sequenced only a limited number of potential cancer genes, karyotypes each contained 1 or more alterations, with over 40% of our data suggest that AML may arise in the absence of an increased these leukemias containing both CNAs and sequence mutations. mutational rate. Moreover, our data raise the possibility that the These data demonstrate that genomewide copy-number alteration development of AML may require fewer genetic alterations than and target gene resequencing complement routine cytogenetic other cancers and that a very limited number of biological processes analysis and identify new genetic lesions in more than half of may need to be altered in hematopoietic stem cells, multipotential pediatric de novo AMLs (Fig. 3). However, although new lesions progenitors, or committed myeloid progenitors to convert them are identified, the total burden of genomic alterations is low, with from a normal cell into an acute myeloid leukemic cell. An only 21% of the AML having Ͼ5 lesions (Fig. 4 and SI Appendix, Table alternative but not mutually exclusive possibility is that epigenetic S2). This is in stark contrast to results from pediatric ALL where 77% changes may play a more predominant role in AML, working in of cases have Ͼ5 lesions excluding point mutations (4, 5). concert with genetic alterations to alter a wider range of biological processes to induce overt leukemia. Detailed genomewide epige- Discussion netic analysis and whole genome resequencing will ultimately be Our analyses of CNAs, CN-LOH, and sequence mutations in required to determine the range of mutations and epigenetic pediatric de novo AML identified remarkably few somatic changes required for the development of AML. However, the genetic alterations within the leukemia cells. We identified a recent sequence of all coding exons in the DNA of an AML mean of only 2.38 somatic CNAs per leukemia and less than 1 patient’s leukemia cells revealed only 10 somatically acquired nonsynonymous point mutation per leukemia in the 25 genes nonsynonymous mutations in the coding region of annotated genes analyzed. Moreover, somatic CN-LOH was observed in only (41), suggesting that even with single base pair resolution few 13% of patients. Even more surprising was the observation that mutations may be the rule in AML. 34% of the leukemias lacked any identifiable CNAs, and 28% of Although the majority of the focal CNAs occurred in only a single the leukemias with recurrent translocation lacked any identifi- case, the recurrent lesions targeted 30 genes known to be involved able sequence or numerical abnormalities. The only exception to in cancer, including 21 that have been previously implicated in the presence of few mutations was acute megakaryocytic leu- AML. In addition, many of the recurrent lesions lacked any cancer kemias, with the majority of these leukemias being characterized consensus genes, suggesting that new AML cancer genes exist by a high number of CNAs but rare point mutations. within the identified regions. A few of the latter lesions target single These data reflect a very low burden of genomic alterations in genes, thereby directly implicating the target genes in AML patho- pediatric de novo AML, which is in stark contrast to most other genesis. The top-ranked recurrent lesion in this category targeted cancers. Recent studies using similar approaches in pediatric ALL, a MAR that contained a putative gene CCDC26, along with a and adult cancers [lung (35), pancreatic (36), glioblastoma multi- recently described Ϸ10-kb highly conserved noncoding RNA (lin- forme (37, 38), breast and colon (39, 40)], have demonstrated a cRNA-CCDC26) that resides within an intron of CCDC26 (22). much higher number of CNAs and point mutations, with the Interestingly, the lincRNA-CCDC26 resides within 1 kb of the majority of these cancers containing a very large number of retroviral integration site that disrupted myeloid differentiation mutations. Although the possibility exists that AMLs may contain (20). Moreover, multiple transcripts are encoded within this locus small regions of CNAs that are below the resolution of detection and some are normally downregulated during myeloid cell differ- using the combined 100K and 500K SNP platforms, this appears entiation (20, 21). Exactly how the AML-associated CNAs affect unlikely on the basis of the absence of smaller lesions in pediatric expression of these transcriptional units and whether their alter- and adult AMLs analyzed using the higher resolution Affymetrix ations contribute to the development of AML remains to be 6.0 microarrays. determined. Nevertheless, the locus is a good candidate to contain

12948 ͉ www.pnas.org͞cgi͞doi͞10.1073͞pnas.0903142106 Radtke et al. Downloaded by guest on September 30, 2021 gene(s) whose alterations may contribute to the development of Methods AML. Other genes implicated include the known or putative tumor Patients and Samples. Leukemic samples from 111 pediatric AML patients treated suppressors TUSC1, BCOR, and PRDM5, ABCC4, which encodes at St. Jude Children’s Research Hospital (SJCRH) were studied. Written informed a multidrug resistance membrane protein, and FAM20C, a gene consent and institutional-review-board approval was obtained for each patient. expressed in hematopoietic cells that is involved in the pathogenesis No commercial entity was involved in the conduct of the study, the analysis or of osteosclerotic bone disease. Determining how genetic alterations storage of the data, or the preparation of the manuscript. The authors vouch for of these genes contribute to the AML pathogenesis will require the completeness and accuracy of the data and analysis. direct functional studies on the biological role of the encoded gene products in normal and leukemic hematopoiesis. Genomic Analyses. DNA extracted from leukemic cells obtained at diagnosis and The limited number of recurrent lesions and the low frequency from samples obtained during remission was genotyped with the use of 50K Hind of individual recurrent lesions across the cohort are notable and 240, 50K Xba 240, 250K Sty, and 250K Nsp SNP arrays (Affymetrix) for each sample. likely reflect the marked heterogeneity among the AML cases that Data from all 4 arrays were combined before analysis for the presence of CNAs, were analyzed. Within the patient cohort, 7 distinct genetic subtypes resulting in an average intermarker distance of less than 5 kb. SNP array analysis of copy-number alterations and LOH was performed as previously reported (4) of de novo AML were included. Each subtype might arise from a and is described in SI. A subset of the identified copy-number alterations was different combination of genetic lesions, with the complement of validated using fluorescence in situ hybridization (FISH) or quantitative genomic lesions being heavily influenced by the initiating event (a chromo- PCR. The primary SNP data are available upon request. Identifiers for each case somal translocation in many pediatric de novo AMLs). Our inte- are listed in SI Appendix, Table S1. grated analysis of cytogenetics, CNAs, and sequence mutations is consistent with this interpretation, with marked difference in the Genomic Resequencing of Candidate AML Cancer Genes. Genomic sequencing of spectrum of changes observed across the AML genetic subtypes. exons and adjacent splice sites was performed on the genes listed in the text. A Extending the detailed genomic characterization of AML to a large detailed description of the sequencing methods are in SI. number of leukemias for each individual known genetic subtype of de novo AML should yield valuable information on the spectrum ACKNOWLEDGMENTS. This study was supported by a Cancer Center Core Grant of mutations in this disease. It will also be of interest to see whether 21765 from the National Cancer Institute, a Leukemia and Lymphoma Society the limited number of CNAs and point mutations found in de novo Specialized Center of Research Grant (LLS7015, to J.R.D.), a grant from the National Health and Medical Research Council of Australia (C.G.M.), and the pediatric AML, will also be observed in other subtypes of AML American Lebanese Syrian Associated Charities of St. Jude Children’s Research including myelodysplasia-related AML, and secondary AML that Hospital. We thank Claire Boltz, Letha Phillips, and James Dalton for technical results from prior chemotherapy and/or radiation therapy. assistance.

1. Kelly LM, Gilliland DG (2002) Genetics of myeloid leukemias. Annu Rev Genomics Hum 22. Guttman M, et al. (2009) Chromatin signature reveals over a thousand highly conserved Genet 3:179–198. large non-coding RNAs in mammals. Nature 458:223–227. 2. Reya T, Morrison SJ, Clarke MF, Weissman IL (2001) Stem cells, cancer, and cancer stem 23. Simpson MA, et al. (2007) Mutations in FAM20C are associated with lethal osteoscle- cells. Nature 414:105–111. rotic bone dysplasia (Raine syndrome), highlighting a crucial molecule in bone devel- 3. Schimmer AD (2006) Induction of apoptosis in lymphoid and myeloid leukemia. Curr opment. Am J Hum Genet 81:906–912. Oncol Rep 8:430–436. 24. Shan Z, Parker T, Wiest JS (2004) Identifying novel homozygous deletions by micro- 4. Mullighan CG, et al. (2007) Genome-wide analysis of genetic alterations in acute satellite analysis and characterization of tumor suppressor candidate 1 gene, TUSC1, lymphoblastic leukaemia. Nature 446:758–764. on chromosome 9p in human lung cancer. Oncogene 23:6612–6620. 5. Mullighan CG, et al. (2008) BCR-ABL1 lymphoblastic leukaemia is characterized by the 25. Huynh KD, Fischle W, Verdin E, Bardwell VJ (2000) BCoR, a novel corepressor involved deletion of Ikaros. Nature 453:110–114. in BCL-6 repression. Genes Dev 14:1810–1823. 6. Dunbar AJ, et al. (2008) 250K single nucleotide polymorphism array karyotyping 26. Rius M, Hummel-Eisenbeiss J, Keppler D (2008) ATP-dependent transport of leukotri- identifies acquired uniparental disomy and homozygous mutations, including novel enes B4 and C4 by the multidrug resistance protein ABCC4 (MRP4). J Pharmacol Exp missense substitutions of c-Cbl, in myeloid malignancies. Cancer Res 68:10349–10357. Ther 324:86–94. 7. Gondek LP, et al. (2008) Chromosomal lesions and uniparental disomy detected by SNP 27. Taki T, et al. (1996) Fusion of the MLL gene with two different genes, AF-6 and arrays in MDS, MDS/MPD, and MDS-derived AML. Blood 111:1534–1542. AF-5alpha, by a complex translocation involving chromosomes 5, 6, 8 and 11 in infant 8. Raghavan M, et al. (2005) Genome-wide single nucleotide polymorphism analysis leukemia. Oncogene 13:2121–2130. reveals frequent partial uniparental disomy due to somatic recombination in acute 28. Deng Q, Huang S (2004) PRDM5 is silenced in human cancers and has growth suppres- myeloid leukemias. Cancer Res 65:375–378. sive activities. Oncogene 23:4903–4910. 9. Rucker FG, et al. (2006) Disclosure of candidate genes in acute myeloid leukemia with 29. Zhang Y, Rowley JD (2006) Chromatin structural elements and chromosomal translo- complex karyotypes using microarray-based molecular characterization. J Clin Oncol cations in leukemia. DNA Repair (Amst) 5:1282–1297. 24:3887–3894. 30. Gupta M, et al. (2008) Novel regions of acquired uniparental disomy discovered in 10. Ravindranath Y, et al. (2005) Pediatric Oncology Group (POG) studies of acute myeloid acute myeloid leukemia. Genes Chromosomes Cancer 47:729–739. leukemia (AML): A review of four consecutive childhood AML trials conducted be- 31. Serrano E, et al. (2008) Uniparental disomy may be associated with microsatellite tween 1981 and 2000. Leukemia 19:2101–2116. instability in acute myeloid leukemia (AML) with a normal karyotype. Leuk Lymphoma 11. Ribeiro RC, et al. (2005) Successive clinical trials for childhood acute myeloid leukemia 49:1178–1183. at St Jude Children’s Research Hospital, from 1980 to 2000. Leukemia 19:2125–2129. 32. Beroukhim R, et al. (2006) Inferring loss-of-heterozygosity from unpaired tumors using 12. Rubnitz JE, et al. (2007) Prognostic factors and outcome of recurrence in childhood high-density oligonucleotide SNP arrays. PLoS Comput Biol 2:e41. acute myeloid leukemia. Cancer 109:157–163. 33. Nanri T, et al. (2005) Mutations in the receptor tyrosine kinase pathway are associated 13. Bacher U, Haferlach T, Schoch C, Kern W, Schnittger S (2006) Implications of NRAS with clinical outcome in patients with acute myeloblastic leukemia harboring mutations in AML: A study of 2502 patients. Blood 107:3847–3853. 14. Frohling S, et al. (2004) CEBPA mutations in younger adults with acute myeloid t(8;21)(q22;q22). Leukemia 19:1361–1366. leukemia and normal cytogenetics: Prognostic relevance and analysis of cooperating 34. Owen C, Barnett M, Fitzgibbon J (2008) Familial myelodysplasia and acute myeloid mutations. J Clin Oncol 22:624–633. leukaemia: A review. Br J Haematol 140:123–132. 15. Loh ML, et al. (2004) PTPN11 mutations in pediatric patients with acute myeloid 35. Weir BA, et al. (2007) Characterizing the cancer genome in lung adenocarcinoma. leukemia: Results from the Children’s Cancer Group. Leukemia 18:1831–1834. Nature 450:893–898. 16. Renneville A, et al. (2008) Cooperating gene mutations in acute myeloid leukemia: A 36. Jones S, et al. (2008) Core signaling pathways in human pancreatic cancers revealed by review of the literature. Leukemia 22:915–931. global genomic analyses. Science 321:1801–1806. 17. Zwaan CM, et al. (2003) FLT3 internal tandem duplication in 234 children with acute 37. Anonymous (2008) Comprehensive genomic characterization defines human glioblas- myeloid leukemia: Prognostic significance and relation to cellular drug resistance. toma genes and core pathways. Nature 455:1061–1068. Blood 102:2387–2394. 38. Parsons DW, et al. (2008) An integrated genomic analysis of human glioblastoma 18. Beroukhim R, et al. (2007) Assessing the significance of chromosomal aberrations in cancer: multiforme. Science 321:1807–1812. Methodology and application to glioma. Proc Natl Acad Sci USA 104:20007–20012. 39. Leary RJ, et al. (2008) Integrated analysis of homozygous deletions, focal amplifica- 19. Futreal PA, et al. (2004) A census of human cancer genes. Nat Rev Cancer 4:177–183. tions, and sequence alterations in breast and colorectal cancers. Proc Natl Acad Sci USA 20. Yin W, Rossin A, Clifford JL, Gronemeyer H (2006) Co-resistance to retinoic acid and 105:16224–16229. TRAIL by insertion mutagenesis into RAM. Oncogene 25:3735–3744. 40. Sjoblom T, et al. (2006) The consensus coding sequences of human breast and colorectal 21. Hirano T, et al. (2008) Genes encoded within 8q24 on the amplicon of a large cancers. Science 314:268–274. extrachromosomal element are selectively repressed during the terminal differentia- 41. Ley TJ, et al. (2008) DNA sequencing of a cytogenetically normal acute myeloid MEDICAL SCIENCES tion of HL-60 cells. Mutat Res 640:97–106. leukaemia genome. Nature 456:66–72.

Radtke et al. PNAS ͉ August 4, 2009 ͉ vol. 106 ͉ no. 31 ͉ 12949 Downloaded by guest on September 30, 2021