Abstract

Acute lymphoblastic leukemia (ALL) accounts for ∼25% of pediatric malignancies. Of interest, the incidence of ALL is observed ∼20% higher in males relative to females. The mechanism behind the phenomenon of sex-specific differences is presently not understood. Employing genome-wide genetic aberration screening in 19 ALL samples, one of the most recurrent lesions identified was monoallelic deletion of the 5′ region of SLX4IP. We characterized this deletion by conventional molecular genetic techniques and analyzed its interrelationships with biological and clinical characteristics using specimens and data from 993 pediatric patients enrolled into trial AIEOP-BFM ALL 2000. Deletion of SLX4IP was detected in ∼30% of patients. Breakpoints within SLX4IP were defined to recurrent positions and revealed junctions with typical characteristics of illegitimate V(D)J-mediated recombination. In initial and validation analyses, SLX4IP deletions were significantly associated with male gender and ETV6/RUNX1-rearranged ALL (both overall P < 0.0001). For mechanistic validation, a second recurrent deletion affecting TAL1 and caused by the same molecular mechanism was analyzed in 1149 T-cell ALL patients. Validating a differential role by sex of illegitimate V(D)J-mediated recombination at the TAL1 locus, 128 out of 1149 T-cell ALL samples bore a deletion and males were significantly more often affected (P = 0.002). The repeatedly detected association of SLX4IP deletion with male sex and the extension of the sex bias to deletion of the TAL1 locus suggest that differential illegitimate V(D)J-mediated recombination events at specific loci may contribute to the consistent observation of higher incidence rates of childhood ALL in boys compared with girls.

INTRODUCTION

Acute lymphoblastic leukemia (ALL) is the most common childhood malignancy (1–3). Despite intensive research efforts, the causes of childhood ALL remain largely unknown. Sex bias—characterized by a 20% higher incidence in males relative to females—is consistently observed in childhood ALL (2–4). In addition, male sex is associated with worse treatment response and outcome (5,6). Therefore, a better understanding of this phenomenon could directly enhance our understanding of disease etiology and improve therapeutic approaches to ALL.

From a genetic perspective, ALL is characterized by recurrent numeric and/or structural somatic aberrations (7). Hyperdiploidy or the cryptic chromosomal translocation t(12;21)—leading to an ETV6/RUNX1 gene fusion—are detectable in 20–25% of cases each, making them the most common genetic subtypes of childhood ALL (7,8). Important for pathomechanistic insights, the molecular features associated with the breakpoints of structural genetic aberrations in ALL are multifold and include, for example, an open chromatin context during gene transcription, Alu and other repeat sequences, or illegitimate V(D)J-mediated recombination (7).

V(D)J recombination is a physiological process by which segments (variable (V), diversity (D), joining (J)) of immunoglobulin (Ig) or T-cell receptor (TCR) genes are rearranged and lead to great diversity of the Ig/TCR repertoire. This process is mediated by lymphocyte-specific endonucleases (RAG1, RAG2) which cut the regional V(D)J genes at flanking recombination signal sequences (RSS) consisting of specific highly conserved heptamer and nonamer sequences with an unconserved spacer (12 or 23 nucleotides) in-between (9,10). Subsequently, the coding segments are joined using the classical non-homologous end-joining (NHEJ) pathway. Paradoxically, RAG proteins can recognize a variety of slightly modified RSS sequences increasing the chance of erroneously targeting RSS-like sequences (‘cryptic RSS’) elsewhere in the genome. Several chromosomal translocations between Ig/TCR loci and proto-oncogenes as well as deletions of non-antigen receptor loci in lymphoid malignancies are thought to be generated by such illegitimate V(D)J-mediated recombination events (11,12). Recently, interstitial deletion of the B-cell translocation gene 1 (BTG1) in 9% of precursor B-cell ALL has been attributed to this mechanism (13). Also, subtypes of CDKN2A and IKZF1 deletions as well as all TAL1 deletions have been associated with off-target action of the RAG complex (12,14–16).

In the present study, we describe a highly frequent deletion in SLX4IP—a gene encoding a currently uncharacterized DNA repair-related protein—by illegitimate V(D)J-mediated recombination (17). SLX4IP deletion has previously been reported in childhood ALL, but at a far lower frequency (18,19). We correlate this finding to demographic and clinical characteristics in a large cohort of children with ALL and describe its specific association with ETV6/RUNX1-rearranged ALL and male sex.

RESULTS

In our initial screen for recurrent genetic aberrations in childhood ALL employing comparative genomic hybridization (CGH) analysis, 5 out of 19 samples—making it one of the most common observations—harbored a monoallelic deletion encompassing the first two exons of the SLX4IP gene with tight breakpoint clustering to a defined position (HG18, chromosome 20: ∼10 363 654–10 404 199 bp; Fig. 1A). Sequencing analysis in the five deletion-positive samples revealed patient-specific breakpoints within only a few nucleotides of each other—confirming site specificity of the recombination event—and demonstrated near perfect matches to 5′ and 3′ heptamer sequences (consensus 5′-CACAGTG), non-templated ‘N’ nucleotides and ‘nibbling away’ of nucleotides at the junction points as hallmarks of illegitimate V(D)J-mediated recombination (Fig. 1B–D) (11,12,14,16,20). No convincing AT-rich nonamer sequences (consensus 5′-ACAAAAACC) were found at 12 or 23 nucleotides distance of the heptamers. As all five patients displaying the SLX4IP deletion at initial diagnosis relapsed, we next analyzed ALL samples obtained at disease recurrence for maintenance of aberrations during leukemic evolution and detected the same patient-specific deletion breakpoints in all samples (Fig. 1D).

Figure 1.

Deletion of the first two exons of SLX4IP on chromosome 20 in five ETV6/RUNX1-negative ALL samples. (A) Custom made NimbleGen fine-tiling CGH-array illustrating a uniform deletion at diagnosis. (B) PCR product of the same samples showing the SLX4IP deletion at diagnosis and relapse. No PCR products regarding SLX4IP deletion are discovered in remission (germline). Patient samples are depicted in the following order: KY-3: lines 1, 6 and 11; KY-2: lines 2, 7 and 12; KY-5: lines 3, 8 and 13; KY-1: lines 4, 9 and 14; KY-4: lines 5, 10 and 15. bp = 100 base pair DNA ladder. L = negative control. HPRT exons 7 and 8 were used as positive control. (C) The SLX4IP deletion junction sequence is displayed and remains unchanged from diagnosis to relapse. Dotted lines illustrate breakpoints at the 5′ and 3′ end of the deletion and N-nucleotides are allocated in-between. (D) The SLX4IP deletion junction sequence at diagnosis and relapse is shown in alignment with wild-type sequence. Signs of illegitimate V(D)J-mediated recombination are displayed. The pseudo-heptamer as part of the pseudo-RSS is depicted as black box. The potential pseudo-nonamer sequence is underlined, separated either by a 12-RSS or 23-RSS.

Figure 1.

Deletion of the first two exons of SLX4IP on chromosome 20 in five ETV6/RUNX1-negative ALL samples. (A) Custom made NimbleGen fine-tiling CGH-array illustrating a uniform deletion at diagnosis. (B) PCR product of the same samples showing the SLX4IP deletion at diagnosis and relapse. No PCR products regarding SLX4IP deletion are discovered in remission (germline). Patient samples are depicted in the following order: KY-3: lines 1, 6 and 11; KY-2: lines 2, 7 and 12; KY-5: lines 3, 8 and 13; KY-1: lines 4, 9 and 14; KY-4: lines 5, 10 and 15. bp = 100 base pair DNA ladder. L = negative control. HPRT exons 7 and 8 were used as positive control. (C) The SLX4IP deletion junction sequence is displayed and remains unchanged from diagnosis to relapse. Dotted lines illustrate breakpoints at the 5′ and 3′ end of the deletion and N-nucleotides are allocated in-between. (D) The SLX4IP deletion junction sequence at diagnosis and relapse is shown in alignment with wild-type sequence. Signs of illegitimate V(D)J-mediated recombination are displayed. The pseudo-heptamer as part of the pseudo-RSS is depicted as black box. The potential pseudo-nonamer sequence is underlined, separated either by a 12-RSS or 23-RSS.

Figure 1.

Continued

Figure 1.

Continued

As the above analyses were performed on selected patients, we next aimed at a reliable assessment of the SLX4IP deletion frequency in a cohort of 512 patients enrolled into AIEOP-BFM ALL 2000 (Cohort 1, Table 1, Supplementary Material, Fig. S1) (5,6). Employing a PCR assay, the deletion could be detected in 164 children (32.0%). When analyzing the association of SLX4IP deletion with clinical characteristics, the most significant positive interrelationships were observed for male sex (P < 0.001) and ETV6/RUNX1-rearranged ALL (P < 0.001) (Table 1). In an analysis stratified by ETV6/RUNX1 translocation status, the male sex bias stayed significant in translocation positive (P < 0.001) as well as negative patients (P < 0.001) (Tables 2 and 3). No association of SLX4IP deletion with treatment outcome was detected (Supplementary Material, Fig. S2).

Table 1.

Clinical characteristics of 512 patients with ALL from trial AIEOP-BFM ALL 2000 (Cohort 1) by SLX4IP deletion status

 Patients positive for SLX4IP deletion (n = 164) n (%) Patients negative for SLX4IP deletion (n = 348) n (%) P-valued 
Gender 
 Male 124 (75.6) 181 (52.0)  
 Female 40 (24.4) 167 (48.0) <0.001 
Age at diagnosis (years) 
 <1 0 (0.0) 0 (0.0)  
 1–<5 79 (48.2) 161 (46.3)  
 5–<10 49 (29.9) 90 (25.9)  
 10–<15 23 (14.0) 73 (21.0)  
 ≥15 13 (7.9) 24 (6.9) 0.286 
Initial WBCa (µl−1
 <10 000 80 (48.8) 156 (44.8)  
 10 000–<50 000 56 (34.1) 111 (31.9)  
 50 000 16 (9.8) 46 (13.2)  
 ≥100 000 12 (7.3) 35 (10.1) 0.463 
Immunophenotype 
 Pre-B 142 (86.6) 260 (74.7)  
 Pro-B 5 (3.0) 14 (4.0)  
 T-cell 17 (10.4) 72 (20.7) 0.011 
 Other/unknown 0 (0.0) 2 (0.6)  
DNA indexb 
<
1.16 
117 (71.3) 208 (59.8)  
1.16 
8 (4.9) 64 (18.4) <0.001 
 Unknown 39 (23.8) 76 (21.8)  
ETV6/RUNX1 
 Positive 71 (43.3) 39 (11.2)  
 Negative 82 (50.0) 283 (81.3) <0.001 
 Unknown 11 (6.7) 26 (7.5)  
BCR/ABL 
 Positive 5 (3.0) 8 (2.3)  
 Negative 158 (96.4) 332 (95.4) 0.765 
 Unknown 1 (0.6) 8 (2.3)  
MLL/AF4 
 Positive 0 (0.0) 0 (0.0)  
 Negative 152 (92.7) 320 (92.0) NAe 
 Unknown 12 (7.3) 28 (8.0)  
Prednisone responsec 
 Good 151 (92.1) 294 (84.5)  
 Poor 12 (7.3) 52 (14.9) 0.015 
 Unknown 1 (0.6) 2 (0.6)  
Risk group 
 Standard 61 (37.2) 112 (32.2)  
 Intermediate 83 (50.6) 169 (48.6)  
 High 20 (12.2) 67 (19.3) 0.123 
 Patients positive for SLX4IP deletion (n = 164) n (%) Patients negative for SLX4IP deletion (n = 348) n (%) P-valued 
Gender 
 Male 124 (75.6) 181 (52.0)  
 Female 40 (24.4) 167 (48.0) <0.001 
Age at diagnosis (years) 
 <1 0 (0.0) 0 (0.0)  
 1–<5 79 (48.2) 161 (46.3)  
 5–<10 49 (29.9) 90 (25.9)  
 10–<15 23 (14.0) 73 (21.0)  
 ≥15 13 (7.9) 24 (6.9) 0.286 
Initial WBCa (µl−1
 <10 000 80 (48.8) 156 (44.8)  
 10 000–<50 000 56 (34.1) 111 (31.9)  
 50 000 16 (9.8) 46 (13.2)  
 ≥100 000 12 (7.3) 35 (10.1) 0.463 
Immunophenotype 
 Pre-B 142 (86.6) 260 (74.7)  
 Pro-B 5 (3.0) 14 (4.0)  
 T-cell 17 (10.4) 72 (20.7) 0.011 
 Other/unknown 0 (0.0) 2 (0.6)  
DNA indexb 
<
1.16 
117 (71.3) 208 (59.8)  
1.16 
8 (4.9) 64 (18.4) <0.001 
 Unknown 39 (23.8) 76 (21.8)  
ETV6/RUNX1 
 Positive 71 (43.3) 39 (11.2)  
 Negative 82 (50.0) 283 (81.3) <0.001 
 Unknown 11 (6.7) 26 (7.5)  
BCR/ABL 
 Positive 5 (3.0) 8 (2.3)  
 Negative 158 (96.4) 332 (95.4) 0.765 
 Unknown 1 (0.6) 8 (2.3)  
MLL/AF4 
 Positive 0 (0.0) 0 (0.0)  
 Negative 152 (92.7) 320 (92.0) NAe 
 Unknown 12 (7.3) 28 (8.0)  
Prednisone responsec 
 Good 151 (92.1) 294 (84.5)  
 Poor 12 (7.3) 52 (14.9) 0.015 
 Unknown 1 (0.6) 2 (0.6)  
Risk group 
 Standard 61 (37.2) 112 (32.2)  
 Intermediate 83 (50.6) 169 (48.6)  
 High 20 (12.2) 67 (19.3) 0.123 

aWBC, white blood cell count at diagnosis.

bRatio of DNA content of leukemic G0/G1 cells to normal diploid lymphocytes.

cGood: <1000 leukemic blood blasts/µl on treatment Day 8; poor: ≥1000 µl−1.

dχ2- or Fisher's exact test.

eNA, not applicable.

Table 2.

Clinical characteristics of Cohort 1 according to SLX4IP deletion status in ETV6/RUNX1-positive and -negative ALL patients treated on AIEOP-BFM ALL 2000

  ETV6/RUNX1-positive patients (n = 110)
 
P-valued ETV6/RUNX1-negative patients (n = 365)
 
P-valued 
Positive for SLX4IP deletion (n = 71) n (%) Negative for SLX4IP deletion (n = 39) n (%) Positive for SLX4IP deletion (n = 82) n (%) Negative for SLX4IP deletion (n = 283) n (%) 
Gender 
 Male 47 (66.2) 12 (30.8)  68 (82.9) 154 (54.4)  
 Female 24 (33.8) 27 (69.2) <0.001 14 (17.1) 129 (45.6) <0.001 
Age at diagnosis (years) 
 <1 0 (1.4) 0 (0.0)  0 (0.0) 0 (0.0)  
 1–<5 42 (59.2) 22 (56.4)  30 (36.6) 126 (44.5)  
 5–<10 23 (32.4) 14 (35.9)  24 (29.3) 70 (24.7)  
 10–<15 4 (5.6) 3 (7.7)  19 (23.2) 64 (22.6)  
 ≥15 2 (2.8) 0 (0.0) 0.708 9 (11.0) 23 (8.1) 0.564 
Initial WBCa (µl−1
 <10 000 40 (56.3) 23(59.0)  36 (43.9) 118 (41.7)  
 10 000–<50 000 21 (29.6) 9 (23.1)  32 (39.0) 95 (33.6)  
 50 000 7 (9.9) 6 (15.4)  5 (6.1) 39 (13.8)  
 ≥100 000 3 (4.2) 1 (2.6) 0.738 9 (11.0) 31 (11.0) 0.292 
Immunophenotype 
 Pre-B 70 (98.6) 38 (97.4)  63 (76.8) 202 (71.4)  
 Pro-B 1 (1.4) 1 (2.6)  4 (4.9) 11 (3.9)  
 T-cell 0 (0.0) 0 (0.0) NAe 15 (18.3) 68 (24.0) 0.516 
 Other/unknown 0 (0.0) 0 (0.0)  0 (0.0) 2 (0.7)  
DNA indexb 
<
1.16 
49 (69.0) 30 (76.9)  60 (73.2) 167 (59.0)  
1.16 
4 (5.6) 1 (2.6) 0.647 3 (3.6) 57 (20.1) <0.001 
 Unknown 18 (25.4) 8 (20.5)  19 (23.2) 59 (20.9)  
Prednisone responsec 
 Good 69 (97.2) 39 (100.0)  73 (89.0) 230 (81.3)  
 Poor 2 (2.8) 0 (0.0) 0.538 8 (9.6) 52 (18.4) 0.067 
 Unknown 0 (0.0) 0 (0.0)  1 (1.2) 1 (0.4)  
Risk group 
 Standard 35 (49.3) 28 (71.8)  21 (25.6) 78 (27.6)  
 Intermediate 34 (47.9) 11 (28.2)  46 (56.1) 142 (50.2)  
 High 2 (2.8) 0 (0.0) 0.058 15 (18.3) 63 (22.3) 0.609 
  ETV6/RUNX1-positive patients (n = 110)
 
P-valued ETV6/RUNX1-negative patients (n = 365)
 
P-valued 
Positive for SLX4IP deletion (n = 71) n (%) Negative for SLX4IP deletion (n = 39) n (%) Positive for SLX4IP deletion (n = 82) n (%) Negative for SLX4IP deletion (n = 283) n (%) 
Gender 
 Male 47 (66.2) 12 (30.8)  68 (82.9) 154 (54.4)  
 Female 24 (33.8) 27 (69.2) <0.001 14 (17.1) 129 (45.6) <0.001 
Age at diagnosis (years) 
 <1 0 (1.4) 0 (0.0)  0 (0.0) 0 (0.0)  
 1–<5 42 (59.2) 22 (56.4)  30 (36.6) 126 (44.5)  
 5–<10 23 (32.4) 14 (35.9)  24 (29.3) 70 (24.7)  
 10–<15 4 (5.6) 3 (7.7)  19 (23.2) 64 (22.6)  
 ≥15 2 (2.8) 0 (0.0) 0.708 9 (11.0) 23 (8.1) 0.564 
Initial WBCa (µl−1
 <10 000 40 (56.3) 23(59.0)  36 (43.9) 118 (41.7)  
 10 000–<50 000 21 (29.6) 9 (23.1)  32 (39.0) 95 (33.6)  
 50 000 7 (9.9) 6 (15.4)  5 (6.1) 39 (13.8)  
 ≥100 000 3 (4.2) 1 (2.6) 0.738 9 (11.0) 31 (11.0) 0.292 
Immunophenotype 
 Pre-B 70 (98.6) 38 (97.4)  63 (76.8) 202 (71.4)  
 Pro-B 1 (1.4) 1 (2.6)  4 (4.9) 11 (3.9)  
 T-cell 0 (0.0) 0 (0.0) NAe 15 (18.3) 68 (24.0) 0.516 
 Other/unknown 0 (0.0) 0 (0.0)  0 (0.0) 2 (0.7)  
DNA indexb 
<
1.16 
49 (69.0) 30 (76.9)  60 (73.2) 167 (59.0)  
1.16 
4 (5.6) 1 (2.6) 0.647 3 (3.6) 57 (20.1) <0.001 
 Unknown 18 (25.4) 8 (20.5)  19 (23.2) 59 (20.9)  
Prednisone responsec 
 Good 69 (97.2) 39 (100.0)  73 (89.0) 230 (81.3)  
 Poor 2 (2.8) 0 (0.0) 0.538 8 (9.6) 52 (18.4) 0.067 
 Unknown 0 (0.0) 0 (0.0)  1 (1.2) 1 (0.4)  
Risk group 
 Standard 35 (49.3) 28 (71.8)  21 (25.6) 78 (27.6)  
 Intermediate 34 (47.9) 11 (28.2)  46 (56.1) 142 (50.2)  
 High 2 (2.8) 0 (0.0) 0.058 15 (18.3) 63 (22.3) 0.609 

aWBC, white blood cell count at diagnosis.

bRatio of DNA content of leukemic G0/G1 cells to normal diploid lymphocytes.

cGood: <1000 leukemic blood blasts/µl on treatment Day 8; poor: ≥1000 µl−1.

dχ2- or Fisher's exact test.

eNA, not applicable.

Table 3.

Gender distribution by SLX4IP or TAL1 deletion status in four childhood ALL cohorts from trials AIEOP-BFM ALL 2000 and ALL IC-BFM 2002

 Cohort 1a (n = 512)
 
 All patients
 
ETV6/RUNX1-positive (n = 110) ETV6/RUNX1-negative (n = 365) 
 SLX4IP deletion-positive (n = 164) n (%) SLX4IP deletion-negative (n = 348) n (%) P-valueb SLX4IP deletion-positive (n = 71) n (%) SLX4IP deletion-negative (n = 39) n (%) P-valueb SLX4IP deletion-positive (n = 82) n (%) SLX4IP deletion-negative (n = 283) n (%) P-valueb 
Gender 
 mc 124 (75.6) 181 (52.0)  47 (66.2) 12 (30.8)  68 (82.9) 154 (54.4)  
 fc  40 (24.4) 167 (48.0) <0.001 24 (33.8) 27 (69.2) <0.001 14 (17.1) 129 (45.6) <0.001 
  
 Cohort 2d (n = 232) Cohort 3e (n = 249) Cohort 4f (n = 1149) 
 SLX4IP deletion-positive (n = 145) n (%) SLX4IP deletion-negative (n = 87) n (%) P-valueb SLX4IP deletion-positive (n = 25) n (%) SLX4IP deletion-negative (n = 224) n (%) P-valueb TAL1 deletion-positive T-cell ALL (n = 128) n (%) TAL1 deletion-negative T-cell ALL (n = 1021) n (%) P-valueb 
Gender 
 mc 81 (55.9) 37 (42.5)  19 (76.0) 119 (53.1)  110 (85.9) 750 (73.5)  
 fc 64 (44.1) 50 (57.5) 0.049  6 (24.0) 105 (46.9) 0.029  18 (14.1) 271 (26.5) 0.002 
 Cohort 1a (n = 512)
 
 All patients
 
ETV6/RUNX1-positive (n = 110) ETV6/RUNX1-negative (n = 365) 
 SLX4IP deletion-positive (n = 164) n (%) SLX4IP deletion-negative (n = 348) n (%) P-valueb SLX4IP deletion-positive (n = 71) n (%) SLX4IP deletion-negative (n = 39) n (%) P-valueb SLX4IP deletion-positive (n = 82) n (%) SLX4IP deletion-negative (n = 283) n (%) P-valueb 
Gender 
 mc 124 (75.6) 181 (52.0)  47 (66.2) 12 (30.8)  68 (82.9) 154 (54.4)  
 fc  40 (24.4) 167 (48.0) <0.001 24 (33.8) 27 (69.2) <0.001 14 (17.1) 129 (45.6) <0.001 
  
 Cohort 2d (n = 232) Cohort 3e (n = 249) Cohort 4f (n = 1149) 
 SLX4IP deletion-positive (n = 145) n (%) SLX4IP deletion-negative (n = 87) n (%) P-valueb SLX4IP deletion-positive (n = 25) n (%) SLX4IP deletion-negative (n = 224) n (%) P-valueb TAL1 deletion-positive T-cell ALL (n = 128) n (%) TAL1 deletion-negative T-cell ALL (n = 1021) n (%) P-valueb 
Gender 
 mc 81 (55.9) 37 (42.5)  19 (76.0) 119 (53.1)  110 (85.9) 750 (73.5)  
 fc 64 (44.1) 50 (57.5) 0.049  6 (24.0) 105 (46.9) 0.029  18 (14.1) 271 (26.5) 0.002 

aAnalysis in Cohort 1 included stratification regarding ETV6/RUNX1 translocation and DNA index.

bχ2- or Fisher's exact test.

cm, male; f, female.

dCohort 2 comprised only ETV6/RUNX1 translocation-positive patients.

eCohort 3 comprised only ETV6/RUNX1 translocation-negative patients.

fCohort 4 consisted of 1149 children with T-cell ALL who have been investigated for TAL1 deletions and were recruited from the German ALL-BFM 2000 and additional national trial groups from Austria, Italy, Czech Republic and Israel.

To validate the observed interrelationships, we next screened an independent cohort of 232 patients with ETV6/RUNX1-rearranged ALL from AIEOP-BFM ALL 2000 (Cohort 2) and identified 145 (62.5%) deletion-positive samples. In these analyses, both the association of SLX4IP deletion with male sex (P = 0.049) and its high incidence in ETV6/RUNX1-rearranged ALL could be confirmed (Table 3, Supplementary Material, Table S1). In a second validation step, we screened 249 leukemic ETV6/RUNX1-negative samples from AIEOP-BFM ALL 2000 patients enrolled in the years 2001 and 2002 (Cohort 3) and, again, found a significant positive association of SLX4IP deletion status with male sex (P = 0.029) (Table 3, Supplementary Material, Table S2).

To consolidate our findings on SLX4IP deletion characteristics at the nucleotide level which were based on five samples only, we continued by sequencing 38 additional leukemic patient DNA samples and detected site-specific SLX4IP breakpoint junctions with typical characteristics of illegitimate V(D)J-mediated recombination in all specimens (Supplementary Material, Table S3). This suggests that—similar to T-cell ALL with its high frequency of illegitimate V(D)J-mediated recombination events—precursor B-cell ALL bears site-specific recombination events with non-dispersed breakpoints in association with cryptic RSS at a previously underestimated high frequency (>25%). Of importance, ∼50% of samples exhibited signs of oligoclonality, indicating that different SLX4IP-deleted subclones existed in parallel.

Recently, it was demonstrated that subclones in ALL have variegated genetics associated with a dynamic clonal architecture in the lead-up to a diagnosis and in relapse (21). To characterize the behavior of the SLX4IP deletion during leukemic evolution, we extended our above described preliminary analyses of initial diagnosis and relapse samples to a total of 22 pairs (Fig. 1; Supplementary Material, Table S4). The deletion was confirmed at relapse in 21 paired samples by PCR. Sequence analysis of the respective breakpoints showed the same monoclonal sequences in nine patients indicating stability of SLX4IP deletions during disease progression. The remaining samples either developed heterogeneity of SLX4IP breakpoints at relapse subsequent to monoclonal sequences (n = 1) or already showed an oligoclonal pattern at initial diagnosis (n = 11) (Supplementary Material, Fig. S3). Backtracking of SLX4IP deletions to birth in seven ETV6/RUNX1-positive patients using material derived from Guthrie cards did not yield any positive results suggesting that deletion of SLX4IP is a secondary event occurring after birth (Supplementary Material, Fig. S4). Although backtracking of SLX4IP deletions may have been hampered by the sensitivity of the detection method, a secondary nature of the deletions is supported by the observation that in monozygotic twins with ETV6/RUNX1-positive ALL and identical translocation breakpoints, the SLX4IP deletion was detectable in one diagnostic leukemic sample only while the second twin was negative (Supplementary Material, Fig. S5). None of 134 available matching bone marrow remission samples from Cohort 1 exhibited the SLX4IP deletion as well as none of 145 peripheral blood samples derived from healthy blood donors.

To gain information on functional consequences of SLX4IP deletion, we next compared its expression in 60 diagnostic ALL specimens carrying the deletion with 60 deletion-negative samples (Supplementary Material, Table S5). We did not find significant differences by SLX4IP deletion status (P = 0.793), but ETV6/RUNX1-positive samples demonstrated higher SLX4IP expression compared with ETV6/RUNX1 negatives (Fig. 2, P < 0.001). Next, we investigated a potential effect of SLX4IP deletion on the expression of neighboring genes (MKKS and JAG1, Supplementary Material, Table S5 and S6) (22,23). While no differential expression of the MKKS gene was seen (P = 0.826), patients carrying a SLX4IP deletion demonstrated significantly higher expression of JAG1 (P < 0.001). Also here, we performed stratified analysis by ETV6/RUNX1 status and observed significantly higher expression of JAG1 in SLX4IP-deleted samples in the ETV6/RUNX1-negative subgroup (P = 0.011), whereas positive patients demonstrated uniformly high JAG1 expression with no discernable differences according to SLX4IP deletion status (P = 0.999) (Fig. 2).

Figure 2.

Expression of SLX4IP, MKKS and JAG1 by SLX4IP deletion and ETV6/RUNX1 rearrangement status. SLX4IP, MKKS and JAG1 expression in leukemic bone marrow cells at diagnosis was measured by real-time quantitative polymerase chain reaction after reverse transcription. The horizontal line in each box indicates the median. The top and bottom of each box indicate the first and third quartiles, respectively, and the tails of the boxes extend to the most extreme values not considered to be outliers. P-values of the Mann–Whitney U-test are depicted. (A) No difference in SLX4IP expression is observed in SLX4IP deletion-positive and -negative patients. Patients with ETV6/RUNX1 rearrangement have a slightly higher SLX4IP expression than ETV6/RUNX1-negative patients. (B) MKKS expression is the same in SLX4IP deletion positive and negative as well as in ETV6/RUNX1 rearrangement-positive and -negative patients. (C) JAG1 expression is depicted in all examined patients (N = 319). SLX4IP deletion-positive patients as well as ETV6/RUNX1 rearrangement-positive patients have significantly higher JAG1 expression in comparison to deletion- or rearrangement-negative patients. (D) JAG1 expression after stratification according to ETV6/RUNX1 rearrangement: SLX4IP deletion-positive ALL samples do express significantly higher JAG1 levels compared with ETV6/RUNX1 rearrangement-negative samples, whereas ETV6/RUNX1 rearrangement-positives all demonstrate high JAG1 expression independent of SLX4IP deletion status. AU, arbitrary units.

Figure 2.

Expression of SLX4IP, MKKS and JAG1 by SLX4IP deletion and ETV6/RUNX1 rearrangement status. SLX4IP, MKKS and JAG1 expression in leukemic bone marrow cells at diagnosis was measured by real-time quantitative polymerase chain reaction after reverse transcription. The horizontal line in each box indicates the median. The top and bottom of each box indicate the first and third quartiles, respectively, and the tails of the boxes extend to the most extreme values not considered to be outliers. P-values of the Mann–Whitney U-test are depicted. (A) No difference in SLX4IP expression is observed in SLX4IP deletion-positive and -negative patients. Patients with ETV6/RUNX1 rearrangement have a slightly higher SLX4IP expression than ETV6/RUNX1-negative patients. (B) MKKS expression is the same in SLX4IP deletion positive and negative as well as in ETV6/RUNX1 rearrangement-positive and -negative patients. (C) JAG1 expression is depicted in all examined patients (N = 319). SLX4IP deletion-positive patients as well as ETV6/RUNX1 rearrangement-positive patients have significantly higher JAG1 expression in comparison to deletion- or rearrangement-negative patients. (D) JAG1 expression after stratification according to ETV6/RUNX1 rearrangement: SLX4IP deletion-positive ALL samples do express significantly higher JAG1 levels compared with ETV6/RUNX1 rearrangement-negative samples, whereas ETV6/RUNX1 rearrangement-positives all demonstrate high JAG1 expression independent of SLX4IP deletion status. AU, arbitrary units.

Figure 2.

Continued

Figure 2.

Continued

To find out whether an increased male susceptibility to illegitimate V(D)J-mediated recombination was restricted to the SLX4IP locus or could be extended to other loci, we finally analyzed illegitimate V(D)J-mediated deletions of TAL1 in a cohort 1149 T-cell ALL patients from trials AIEOP-BFM ALL 2000 and ALL IC-BFM 2002 (Cohort 4). We were able to identify 128 (11.1%) TAL1-deleted patients with 110 (85.9%) out of them being male. Similar to SLX4IP, this distribution resulted in a significant positive association of TAL1 deletion status with male sex (P = 0.002) (Table 3). Thus, an increased male susceptibility to the event of illegitimate V(D)J-mediated recombination could also be observed at the TAL1 locus. In additional analyses employing multiplex ligation-dependent probe amplification (MPLA) data on IKZF1 and BTG1 deletions which were previously described as being V(D)J-mediated we could not detect a sex-biased distribution of deletions (Supplementary Material, Table S7) (13,24,25).

DISCUSSION

Here, we describe a site-specific deletion within the 5′ region of the SLX4IP locus in ∼30% of childhood ALL in general and >60% of ETV6/RUNX1-rearranged ALL, making the deletion of SLX4IP one of the most common aberrations in childhood ALL described so far. As SLX4IP deletions were previously only described at far lower frequencies in array-based genome-wide analyses of representative large cohorts (18,19), our observation of SLX4IP deletion at a much higher percentage by PCR analysis implies that the subclonal architecture of ALL in association with methodological sensitivity thresholds introduces bias in frequency estimates of genetic aberrations by conventional SNP array or CGH analyses. Most likely, a larger proportion of SLX4IP deletions detected in our study occurred at later stages of leukemic evolution secondary to ETV6/RUNX1 rearrangements and were, therefore, only present in a fraction of leukemic cells. These subclonal levels would have been missed by array techniques requiring presence of a lesion in a majority of analyzed cells for reliable detection. The particular sensitivity of the SLX4IP locus to secondary recombination events is further exemplified by the fact that breakpoint sequencing revealed oligoclonality in nearly half of the SLX4IP deletion-positive samples analyzed at initial diagnosis of ALL.

The described deletion in the 5′ region of SLX4IP differs from other recurrently detectable genetic aberrations of comparable frequencies in childhood ALL in the way that it occurs in a site-specific fashion and consistently demonstrates features of a single causal mechanism—illegitimate V(D)J-mediated recombination. The repeatedly detected association of SLX4IP deletion with male sex in our study and the extension of the sex bias to deletion of the TAL1 locus imply that differential susceptibility by sex to illegitimate V(D)J-mediated recombination at specific loci may contribute to the consistent observation of higher incidence rates of childhood ALL in boys compared with girls. We were not able to extend our findings to two additional loci recurrently displaying features of illegitimate V(D)J-mediated recombination at their breakpoints—IKZF1 and BTG1 (13,14). This suggests that our thoroughly validated observation of sex-specific acquisition of SLX4IP and TAL1 deletions by illegitimate V(D)J-mediated recombination cannot be generalized to all loci demonstrating susceptibility to illegitimate V(D)J-mediated recombination. The mechanism underlying these locus-specific findings remains to be clarified. Potential explanations could include locus-specific differences, variation in sex-specific susceptibility to illegitimate V(D)J-mediated recombination for biological subgroups of ALL, or SLX4IP and TAL1 deletions occurring with equal frequencies in males and females, but underlying sex-specific differences in selection processes during leukemia development.

Sex differences exist in a variety of diseases including, besides others, autoimmune disorders and infectious diseases, and are largely believed to represent differential effects of sex-specific hormone action even in prepubertal children (26,27). For the hematopoietic system, it was demonstrated that 17β-estradiol treatment can modify the differentiation, proliferation and survival of early B-cell precursors with a direct influence on Ig gene rearrangements (28). How these observations could mechanistically relate to sex-specific differences in illegitimate V(D)J-mediated recombination in childhood ALL is currently only subject to speculation, but it may well be that differences in endogenous sex hormone exposure could play a modifying role here. Although sex hormone levels between boys and girls already differ during infancy, it is of interest that sex bias in childhood ALL becomes most obvious in children during puberty when hormonal differences become more profound (27). That nuclear receptors themselves are implicated in the mechanism of genomic aberrations was recently demonstrated by an intriguing observation demonstrating that binding of the androgen receptor to intronic regions introduces inter- and intrachromosomal interactions, leading to double-strand breaks and ligation by the NHEJ pathway (29). Future mechanistic studies taking into account this available information may shed some light on the complex interplay of factors with potential involvement in sex-biased genetic aberrations in childhood ALL.

If the high frequency of the deletion indicates a pathomechanistic role for SLX4IP in ALL leukemogenesis and confers a selective advantage to the affected leukemic clone or simply reflects an increased locus-specific susceptibility to recombination events without ‘leukemogenic’ consequences remains to be evaluated in further detail. Nevertheless, we found that SLX4IP expression did not differ between leukemic samples carrying the deletion and those not, indicating that loss at SLX4IP seems to be predominantly monoallelic as demonstrated by array CGH and does not lead to haploinsufficiency at the transcriptional level. In contrast, we could demonstrate higher expression of the SLX4IP neighboring gene JAG1 in SLX4IP-deleted patients. Overexpression of JAG1—a WNT-dependent Notch signaling activator with important functions in stem and progenitor cell homeostasis—has been previously reported in acute myeloid leukemia as well as ETV6/RUNX1-positive ALL (22,23). Since ETV6—through a consensus sequence on the promoter—may function as a transcriptional repressor on JAG1, the decrease of ETV6 activity as a consequence of the ETV6/RUNX1 translocation was suggested to be responsible for the increase of JAG1 transcription in ETV6/RUNX1-positive ALL (23). However, although we detected differential JAG1 expression by SLX4IP deletion status in the overall patient group and in the ETV6/RUNX1-negative subgroup, this effect could not be observed in ETV6/RUNX1-positive patients. This observation contradicts a selective advantage of a leukemic cell acquiring a SLX4IP deletion by dysregulation of JAG1 and supports a model where higher regional transcriptional activity at SLX4IP and the neighboring JAG1 is common to ETV6/RUNX1-rearranged and a subgroup of alike other ALLs which predisposes to SLX4IP deletion through an open chromatin context. Thus, SLX4IP deletion may act as a surrogate of a local transcriptional profile common to ETV6/RUNX1-positive ALL and a group of ETV6/RUNX1-like ALL with a similarly good prognosis.

In conclusion, the here reported increased susceptibility of males to illegitimate V(D)J-mediated recombination at SLX4IP and TAL1 may serve as an important etiological hint to the currently only poorly understood sex bias in childhood ALL and, at the same time, may serve the hypothesis of sex-specific differences in recombination frequencies being associated with higher male incidences in a majority of hematological neoplasms and their partly less favorable treatment outcome.

MATERIALS AND METHODS

Study individuals

Patients from Austria, Germany, Italy and Switzerland were enrolled in multicenter trial AIEOP-BFM ALL 2000 on treatment of childhood ALL. Diagnosis, characterization and treatment of ALL were performed as previously described (see also Supplementary Material) (5,6,30–35).

German patients of AIEOP-BFM ALL 2000 were included for SLX4IP deletion screening and sequencing, TAL1 deletion analysis as well as gene expression analysis, Austrian and Italian patients were only included in TAL1 deletion analysis. Patients from Czech Republic and Israel were treated in multicenter trial ALL IC-BFM 2002. Czech patients were included in backtracking experiments and subjected to TAL1 deletion analysis, Israeli patients only employed for TAL1 analysis. German patients with recurrent ALL were enrolled in relapse trials ALL-Rez BFM 95/96 and ALL-Rez BFM 2002 and included in SLX4IP sequencing analyses. Outcome analysis was only performed for German patients treated on AIEOP-BFM ALL 2000. All trials are described in detail elsewhere and were approved by the respective institutional review boards (Hannover Medical School, Hannover; 2nd Faculty of Medicine, Charles University Prague; Charité University Medicine, Berlin) (5,6,36–38). Informed consent for the use of specimen for research was obtained from all study individuals, parents or legal guardians.

Samples, cell and DNA isolation

Bone marrow samples were obtained at initial diagnosis, treatment Days 15, 33, 52, 78 and at relapse. Mononuclear cells were isolated by Ficoll-Paque gradient centrifugation (Pharmacia, Freiburg, Germany) from bone marrow samples followed by extraction of high-molecular-weight DNA according to standardized protocols using either the Puregene DNA isolation system (Gentra Systems, Minneapolis, MN, USA) or Qiagen DNA Blood Kits (Qiagen, Hilden, Germany). Quality and quantity of genomic DNA was determined by spectrophotometry. DNA yielded from bone marrow aspirates collected in remission served for germline SLX4IP deletion screening. An anonymized cohort of blood donors served as healthy control group.

Comparative genomic hybridization analysis

Nineteen diagnostic DNA samples (blast content >80%) from German AIEOP-BFM ALL 2000 patients (Supplementary Material, Table S8) who later relapsed and matching remission bone marrows were hybridized to high-resolution custom-made fine-tiling CGH arrays covering 62 loci comprising 27 210 197 bp (Supplementary Material, Table S9). Loci selection was based on own preliminary experiments employing 100K Affymetrix arrays (Affymetrix, Santa Clara, CA, USA) and publications reporting genomic aberrations in ALL (18).

UCSC Human genome build March 2006 (HG18) was used as a reference. The designed oligonucleotide array comprised 385 000 probes (probe length 50–75mer) and was manufactured, hybridized and scanned by imaGenes GmbH (NimbleGen service: ROCHE NimbleGen, Reykjavik, Iceland). Repeat regions were excluded and oligonucleotide spacing was <71 bp on average (Supplementary Material, Table S9). Data were analyzed using SignalMap Version 1.9.0.03 (NimbleGen Systems, Inc., Madison, WI, USA).

SLX4IP and TAL1 deletion screening and genomic sequencing, MLPA

SLX4IP deletion breakpoints were amplified by genomic PCR using primers (5′ → 3′) forward: gataattcacccggcatttccccatc and reverse: atgcccccgaggcctctctacaaact (39). Product length was ∼705 bp when SLX4IP deletion was present; sequencing primers (5′ → 3′) were forward: tgttcaaacatggctatttttatt and reverse: cctgtctatgagactgccaaa. For TAL1 deletion analysis, diagnostic leukemic DNA samples were screened using the BIOMED-1 primer set (40).

Amplification of SLX4IP deletion breakpoints was performed by genomic PCR (forward primer (5′–3′): gataattcacccggcatttccccatc and reverse primer (5′–3′): atgcccccgaggcctctctacaaact) (39). Fifty nanograms of leukemic blast DNA or human control DNA were amplified using the Qiagen Mutation Detect PCR Kit (Qiagen). As a positive control, a fragment encompassing HPRT exons 7 and 8 was amplified using the following primers: forward (5′–3′): gtaatattttgtaattaacagcttgctgg and reverse (5′–3′): tcagtctggtcaaatgacgaggtgc (product length 423 bp). PCR conditions were: 95°C for 5 min followed by 35 cycles of 95°C for 30 s and 61°C for 1 min and 72°C for 3 min. PCR products were analyzed on a multicapillary electrophoresis system (QIAxcel, Qiagen). Before direct sequencing, PCR products were run on a 1.0% agarose gel and purified using the Qiagen gel extraction Kit (Qiagen). Product length was ∼705 bp when SLX4IP deletion was present. The nucleotide sequences of both strands of the PCR products were directly determined using an automated fluorescent sequencer (ABI Prism 310 Genetic Analyzer, Perkin Elmer Corporation, Wellesley, MA, USA). Primer sequences used for sequencing were the following: forward (5′–3′): tgttcaaacatggctatttttatt and reverse (5′–3′): cctgtctatgagactgccaaa. When distinct breakpoints with non-ambiguous sequencing results continuing behind the breakpoint junction in forward and reverse sequencing reactions were observed, monoclonality of SLX4IP deletion was assumed (Fig. 1C). In the case of non-ambiguous sequencing results before the breakpoint junction (in forward and reverse reactions) with sudden indistinct sequencing signature or with one major-sequence as well as another upcoming minor sequence signature behind the typical breakpoint junction, oligoclonality was hypothesized (Supplementary Material, Tables S3 and S4 and Fig. S3). In these cases, true breakpoints might deviate by 1–2 nucleotides from those depicted in Supplementary Material, Tables S5 and S6, due to ambiguous breakpoint junction assignment.

Primer design in the BIOMED-1 approach for the TAL1 deletion was restricted to the breakpoint in the SIL gene and the two most frequent breakpoints in the TAL1 gene, type 1 and type 2 (taldb1 and taldb2), covering at least 95% of all known TAL1 deletions. PCR products obtained were further examined by heteroduplex or gene scanning analyses to discriminate between amplifications derived from monoclonal or polyclonal lymphoblastic cell populations and junctional regions of clonal PCR products were sequenced.

MLPA was performed as described previously (25).

Deletion backtracking in Guthrie cards and ETV6/RUNX1 genomic breakpoint cloning

For backtracking of SLX4IP deletions to the time of birth, genomic DNA from Guthrie cards was isolated using the InstaGene Dry Blood Kit (Bio-Rad, Hercules, CA, USA). Three separate sections (1/8) of one Guthrie card were used in PCR reactions. For SLX4IP deletion screening, specific primers were designed. The amplified products were analyzed by automated chip-based microcapillary electrophoresis on an Agilent 2100 Bioanalyzer instrument (Agilent Technologies, Santa Clara, CA, USA). Assay specificity was determined using serial dilutions of individual diagnostic DNA in healthy donor DNA. Guthrie card PCRs were performed in a separate room in a separate PCR box. For SLX4IP deletion screening in Guthrie cards specific primers were designed: forward (5′–3′): agaaaacacaccacagaaaagc and reverse (5′–3′): gggaggtgaggagccactat. PCR conditions were: initial denaturation at 95°C for 10 min; 35 cycles at 94°C for 30 s, 58.5°C for 30 s, 72°C for 30 s, final extension 72°C for 7 min. ETV6/RUNX1 genomic breakpoints in monozygotic twins were cloned as described previously (41).

RNA isolation, reverse transcription and gene expression analysis by real-time quantitative PCR

Total RNA was isolated with Triazol reagent (Invitrogen, Paisley, UK) and subsequently passed over a Qiagen RNeasy column (Qiagen) for removal of small fragments. Total RNA was quantified and validated for integrity using the Bioanalyzer 2100 (Agilent, Palo Alto, CA, USA). Real-time quantitative polymerase chain reaction (RQ-PCR) analysis was performed after random hexamer priming and MuLV reverse transcription (Fermentas, Hanover, MD, USA) to generate cDNA. PCR was carried out on an Applied Biosystems Model 7900 HT Sequence Detector (Darmstadt, Germany) using the QuantiTect SYBR Green PCR kit (Qiagen) as described in the manufacturer's instructions. The expression level of the SDHA and ABL1 were used to normalize for differences in input cDNA. QuantiTect Primer Assays were used to measure mRNA abundances of the SLX4IP, MKKS and JAG1 genes (Qiagen). Melting curve analyses were performed to verify the amplification specificity. Each sample was tested in duplicate. The expression ratio was calculated as 2n, where n was the threshold cycle (CT) value difference normalized by the CT difference of a calibrator sample.

For SLX4IP and MKKS expression analysis 120 German patients from trial AIEOP-BFM ALL 2000 with known SLX4IP deletion status and RNA availability were selected. Selection criteria were as following: half of the 120 patient samples were either SLX4IP deletion-negative (n = 60) or positive (n = 60). Within each of the two subgroups, half of the patients were ETV6/RUNX1 translocation-negative (n = 30) or positive (n = 30). Exclusion criteria comprised: BCR/ABL1 or MLL rearrangement, T-ALL and hyperdiploidy (Supplementary Material, Table S5). For JAG1 expression analysis, 319 German patients from trial AIEOP-BFM ALL 2000 with known SLX4IP deletion status and RNA availability were selected (Supplementary Material, Table S6).

Statistical analysis

Differences in the distribution of categorical variables among patient subsets were analyzed using Fisher's exact or χ2-test. Comparisons of gene expression levels between groups were performed by Mann–Whitney U-test. Event-free survival was defined as the time from diagnosis to the date of last follow-up in complete remission or to the first event. Events were resistance to therapy (non-response), relapse, secondary neoplasm or death from any cause. Failure to achieve remission due to early death or non-response was considered as event at time zero. Patients lost to follow-up were censored at the time of their withdrawal. The Kaplan–Meier method was used to estimate survival rates, differences were compared with the two-sided log-rank test (42,43). Cumulative incidence functions for competing events were constructed by the method of Kalbfleisch and Prentice, and were compared employing the Gray's test (44,45). Computations were performed using SPSS Version 17.0 (SPSS, Armonk, NY, USA) and SAS Version 9.1 (SAS, Cary, NC, USA) statistical programs.

SUPPLEMENTARY MATERIAL

Supplementary Material is available at HMG online.

FUNDING

This work was supported by the Deutsche José Carreras Leukämie-Stiftung, the Madeleine Schickedanz Kinderkrebs-Stiftung, the Deutsche Krebshilfe, the Israel Cancer Association, Fondazione Tettamanti (Monza), Fondazione Citta della Speranza, Fondazione Cariparo (Padova) and St. Anna Kinderkrebsforschung Austria.

ACKNOWLEDGEMENTS

We are indebted to all participants and personnel involved in the clinical trials associated with this study.

Conflict of Interest statement. None declared.

REFERENCES

1
Pui
C.H.
Evans
W.E.
Treatment of acute lymphoblastic leukemia
N. Engl. J. Med.
 , 
2006
, vol. 
354
 (pg. 
166
-
178
)
2
Redaelli
A.
Laskin
B.L.
Stephens
J.M.
Botteman
M.F.
Pashos
C.L.
A systematic literature review of the clinical and epidemiological burden of acute lymphoblastic leukaemia (ALL)
Eur. J. Cancer Care (Engl)
 , 
2006
, vol. 
14
 (pg. 
53
-
62
)
3
Tower
R.L.
Spector
L.G.
The epidemiology of childhood leukemia with a focus on birth weight and diet
Crit. Rev. Clin. Lab. Sci.
 , 
2007
, vol. 
44
 (pg. 
203
-
242
)
4
Linabery
A.M.
Ross
J.A.
Trends in childhood cancer incidence in the U.S. (1992–2004)
Cancer
 , 
2008
, vol. 
112
 (pg. 
416
-
432
)
5
Schrappe
M.
Valsecchi
M.G.
Bartram
C.R.
Schrauder
A.
Panzer-Grümayer
R.
Möricke
A.
Parasole
R.
Zimmermann
M.
Dworzak
M.
Buldini
B.
, et al.  . 
Late MRD response determines relapse risk overall and in subsets of childhood T-cell ALL: results of the AIEOP-BFM-ALL 2000 study
Blood
 , 
2011
, vol. 
118
 (pg. 
2077
-
2084
)
6
Conter
V.
Bartram
C.R.
Valsecchi
M.G.
Schrauder
A.
Panzer-Grümayer
R.
Möricke
A.
Aricò
M.
Zimmermann
M.
Mann
G.
De Rossi
G.
, et al.  . 
Molecular response to treatment redefines all prognostic factors in children and adolescents with B-cell precursor acute lymphoblastic leukemia: results in 3184 patients of the AIEOP-BFM ALL 2000 study
Blood
 , 
2010
, vol. 
115
 (pg. 
3206
-
3214
)
7
Greaves
M.F.
Wiemels
J.
Origins of chromosome translocations in childhood leukaemia
Nat. Rev. Cancer
 , 
2003
, vol. 
3
 (pg. 
639
-
649
)
8
Zelent
A.
Greaves
M.
Enver
T.
Role of the TEL-AML1 fusion gene in the molecular pathogenesis of childhood acute lymphoblastic leukaemia
Oncogene
 , 
2004
, vol. 
23
 (pg. 
4275
-
4283
)
9
Gellert
M.
V(D)J recombination: RAG proteins, repair factors, and regulation
Annu. Rev. Biochem.
 , 
2002
, vol. 
71
 (pg. 
101
-
132
)
10
Schatz
D.G.
V(D)J recombination
Immunol. Rev.
 , 
2004
, vol. 
200
 (pg. 
5
-
11
)
11
Raghavan
S.C.
Kirsch
I.R.
Lieber
M.R.
Analysis of the V(D)J recombination efficiency at lymphoid chromosomal translocation breakpoints
J. Biol. Chem.
 , 
2001
, vol. 
276
 (pg. 
29126
-
29133
)
12
Marculescu
R.
Vanura
K.
Montpellier
B.
Roulland
S.
Le
T.
Navarro
J.M.
Jäger
U.
McBlane
F.
Nadel
B.
Recombinase, chromosomal translocations and lymphoid neoplasia: targeting mistakes and repair failures
DNA Repair
 , 
2006
, vol. 
5
 (pg. 
1246
-
1258
)
13
Waanders
E.
Scheijen
B.
van der Meer
L.T.
van Reijmersdal
S.V.
van Emst
L.
Kroeze
Y.
Sonneveld
E.
Hoogerbrugge
P.M.
van Kessel
A.G.
van Leeuwen
F.N.
, et al.  . 
The origin and nature of tightly clustered BTG1 deletions in precursor B-cell acute lymphoblastic leukemia support a model of multiclonal evolution
PLoS Genet.
 , 
2012
, vol. 
8
 pg. 
e1002533
 
14
Onozawa
M.
Aplan
P.D.
Illegitimate V(D)J recombination involving nonantigen receptor loci in lymphoid malignancy
Genes Chromosomes Cancer
 , 
2012
, vol. 
51
 (pg. 
525
-
535
)
15
van der Velden
V.H.
Hochhaus
A.
Cazzaniga
G.
Szczepanski
T.
Gabert
J.
van Dongen
J.J.
Detection of minimal residual disease in hematologic malignancies by real-time quantitative PCR: principles, approaches, and laboratory aspects
Leukemia
 , 
2003
, vol. 
17
 (pg. 
1013
-
1034
)
16
Aplan
P.D.
Lombardi
D.P.
Ginsberg
A.M.
Cossman
J.
Bertness
V.L.
Kirsch
I.R.
Disruption of the human SCL locus by ‘illegitimate’ V(D)J recombinase activity
Science
 , 
1990
, vol. 
250
 (pg. 
1426
-
1429
)
17
Svendsen
J.M.
Smogorzewska
A.
Sowa
M.E.
O'Connell
B.C.
Gygi
S.P.
Elledge
S.J.
Harper
J.W.
Mammalian BTBD12/SLX4 assembles a holliday junction resolvase and is required for DNA repair
Cell
 , 
2009
, vol. 
138
 (pg. 
63
-
77
)
18
Mullighan
C.G.
Goorha
S.
Radtke
I.
Miller
C.B.
Coustan-Smith
E.
Dalton
J.D.
Girtman
K.
Mathew
S.
Ma
J.
Pounds
S.B.
, et al.  . 
Genome-wide analysis of genetic alterations in acute lymphoblastic leukemia
Nature
 , 
2007
, vol. 
446
 (pg. 
758
-
764
)
19
Lilljebjörn
H.
Soneson
C.
Andersson
A.
Heldrup
J.
Behrendtz
M.
Kawamata
N.
Ogawa
S.
Koeffler
H.P.
Mitelman
F.
Johansson
B.
, et al.  . 
The correlation pattern of acquired copy number changes in 164 ETV6/RUNX1-positive childhood acute lymphoblastic leukemias
Hum. Mol. Genet.
 , 
2010
, vol. 
19
 (pg. 
3150
-
3158
)
20
Tsai
A.G.
Lieber
M.R.
Mechanisms of chromosomal rearrangement in the human genome
BMC Genomics
 , 
2010
, vol. 
11
 
Suppl. 1
pg. 
S1
 
21
Anderson
K.
Lutz
C.
van Delft
F.W.
Bateman
C.M.
Guo
Y.
Colman
S.M.
Kempski
H.
Moorman
A.V.
Titley
I.
Swansbury
J.
, et al.  . 
Genetic variegation of clonal architecture and propagating cells in leukemia
Nature
 , 
2011
, vol. 
469
 (pg. 
356
-
361
)
22
Stirewalt
D.L.
Meshinchi
S.
Kopecky
K.J.
Fan
W.
Pogosova-Agadjanyan
E.L.
Engel
J.H.
Cronk
M.R.
Dorcy
K.S.
McQuary
A.R.
Hockenbery
D.
, et al.  . 
Identification of genes with abnormal expression changes in acute myeloid leukemia
Genes Chromosomes Cancer
 , 
2008
, vol. 
47
 (pg. 
8
-
20
)
23
Bungaro
S.
Dell'Orto
M.C.
Zangrando
A.
Basso
D.
Gorletta
T.
Lo Nigro
L.
Leszl
A.
Young
B.D.
Basso
G.
Bicciato
S.
, et al.  . 
Integration of genomic and gene expression data of childhood ALL without known aberrations identifies subgroups with specific genetic hallmarks
Genes Chromosomes Cancer
 , 
2009
, vol. 
48
 (pg. 
22
-
38
)
24
Mullighan
C.G.
Miller
C.B.
Radtke
I.
Phillips
L.A.
Dalton
J.
Ma
J.
White
D.
Hughes
T.P.
Le Beau
M.M.
Pui
C.H.
, et al.  . 
BCR-ABL1 lymphoblastic leukaemia is characterized by the deletion of Ikaros
Nature
 , 
2008
, vol. 
453
 (pg. 
110
-
114
)
25
Dörge
P.
Meissner
B.
Zimmermann
M.
Möricke
A.
Schrauder
A.
Bouquin
J.P.
Schewe
D.
Harbott
J.
Teigler-Schlegel
A.
Ratei
R.
, et al.  . 
IKZF1 Deletion is an independent predictor of outcome in pediatric acute lymphoblastic leukemia treated according to the ALL-BFM 2000 protocol
Haematologica
 , 
2013
, vol. 
98
 (pg. 
428
-
432
)
26
Fish
E.N.
The X-files in immunity: sex-based differences predispose immune responses
Nat. Rev. Immunol.
 , 
2008
, vol. 
8
 (pg. 
737
-
744
)
27
Ober
C.
Loisel
D.A.
Gilad
Y.
Sex-specific genetic architecture of human disease
Nat. Rev. Genet.
 , 
2008
, vol. 
9
 (pg. 
911
-
922
)
28
Medina
K.l.
Strasser
A.
Kincade
P.W.
Estrogen influences the differentiation, proliferation, and survival of early B-lineage precursors
Blood
 , 
2000
, vol. 
95
 (pg. 
2059
-
2067
)
29
Lin
C.
Yang
L.
Tanasa
B.
Hutt
K.
Ju
B.G.
Ohgi
K.
Zhang
J.
Rose
D.W.
Fu
X.D.
Glass
C.K.
, et al.  . 
Nuclear receptor-induced chromosomal proximity and DNA breaks underlie specific translocations in cancer
Cell
 , 
2009
, vol. 
139
 (pg. 
1069
-
1083
)
30
Ludwig
W.D.
Rieder
H.
Bartram
C.R.
Heinze
B.
Schwartz
S.
Gassmann
W.
Löffler
H.
Hossfeld
D.
Heil
G.
Handt
S.
, et al.  . 
Immunophenotypic and genotypic features, clinical characteristics, and treatment outcome of adult pro-B acute lymphoblastic leukemia: results of the German multicenter trials GMALL 03/87 and 04/89
Blood
 , 
1998
, vol. 
92
 (pg. 
1898
-
1909
)
31
Hiddemann
W.
Wörmann
B.
Ritter
J.
Thiel
E.
Göhde
W.
Lahme
B.
Henze
G.
Schellong
G.
Riehm
H.
Büchner
T.
Frequency and clinical significance of DNA aneuploidy in acute leukemia
Ann. N. Y. Acad. Sci.
 , 
1986
, vol. 
468
 (pg. 
227
-
240
)
32
Harbott
J.
Ritterbach
J.
Ludwig
W.D.
Bartram
C.R.
Reiter
A.
Lampert
F.
Clinical significance of cytogenetic studies in childhood acute lymphoblastic leukemia: experience of the BFM trials
Recent Results Cancer Res.
 , 
1993
, vol. 
131
 (pg. 
123
-
132
)
33
Schlieben
S.
Borkhardt
A.
Reinisch
I.
Ritterbach
J.
Janssen
J.W.
Ratei
R.
Schrappe
M.
Repp
R.
Zimmermann
M.
Kabisch
H.
, et al.  . 
Incidence and clinical outcome of children with BCR/ABL-positive acute lymphoblastic leukemia (ALL): a prospective RT-PCR study based on 673 patients enrolled in the German pediatric multicenter therapy trials ALL-BFM-90 and CoALL-05-92
Leukemia
 , 
1996
, vol. 
10
 (pg. 
957
-
963
)
34
Viehmann
S.
Borkhardt
A.
Lampert
F.
Harbott
J.
Multiplex PCR: a rapid screening method for detection of gene rearrangements in childhood acute lymphoblastic leukemia
Ann. Hematol.
 , 
1999
, vol. 
78
 (pg. 
157
-
162
)
35
Borkhardt
A.
Cazzaniga
G.
Viehmann
S.
Valsecchi
M.G.
Ludwig
W.D.
Burci
L.
Mangioni
S.
Schrappe
M.
Riehm
H.
Lampert
F.
, et al.  . 
Incidence and clinical relevance of TEL/AML1 fusion genes in children with acute lymphoblastic leukemia enrolled in the German and Italian multicenter therapy trials: Associazione Italiana Ematologia Oncologia Pediatrica and the Berlin-Frankfurt-Munster Study Group
Blood
 , 
1997
, vol. 
90
 (pg. 
571
-
577
)
36
Volejnikova
J.
Mejstrikova
E.
Valova
T.
Reznickova
L.
Hodonska
L.
Mihal
V.
Sterba
J.
Jabali
Y.
Prochazkova
D.
Blazek
B.
, et al.  . 
Minimal residual disease in peripheral blood at day 15 identifies subgroup of childhood B-cell precursor acute lymphoblastic leukemia with superior prognosis
Haematologica
 , 
2011
, vol. 
96
 (pg. 
1815
-
1821
)
37
Bader
P.
Kreyenberg
H.
Henze
G.H.
Eckert
C.
Reising
M.
Willasch
A.
Barth
A.
Borkhardt
A.
Peters
C.
Handgretinger
R.
, et al.  . 
Prognostic value of minimal residual disease quantification before allogeneic stem-cell transplantation in relapsed childhood acute lymphoblastic leukemia: the ALL-REZ BFM Study Group
J. Clin. Oncol.
 , 
2009
, vol. 
27
 (pg. 
377
-
384
)
38
Eckert
C.
Flohr
T.
Koehler
R.
Hagedorn
N.
Moericke
A.
Stanulla
M.
Kirschner-Schwabe
R.
Cario
G.
Stackelberg
A.
Bartram
C.R.
, et al.  . 
Very early/early relapses of acute lymphoblastic leukemia show unexpected changes of clonal markers and high heterogeneity in response to initial and relapse treatment
Leukemia
 , 
2011
, vol. 
25
 (pg. 
1305
-
1313
)
39
Mullighan
C.G.
Phillips
L.A.
Su
X.
Ma
J.
Miller
C.B.
Shurtleff
S.A.
Downing
J.R.
Genomic analysis of the clonal origins of relapsed acute lymphoblastic leukemia
Science
 , 
2008
, vol. 
322
 (pg. 
1377
-
1380
)
40
Pongers-Willemse
M.J.
Seriu
T.
Stolz
F.
d'Aniello
E.
Gameiro
P.
Pisa
P.
Gonzalez
M.
Bartram
C.R.
Panzer-Grümayer
E.R.
Biondi
A.
, et al.  . 
Primers and protocols for standardized detection of minimal residual disease in acute lymphoblastic leukemia using immunoglobulin and T cell receptor gene rearrangements and TAL1 deletions as PCR targets: report of the BIOMED-1 CONCERTED ACTION: investigation of minimal residual disease in acute leukemia
Leukemia
 , 
1999
, vol. 
13
 (pg. 
110
-
118
)
41
von Goessel
H.
Jacobs
U.
Semper
S.
Krumbholz
M.
Langer
T.
Keller
T.
Schrauder
A.
van der Velden
V.H.
van Dongen
J.J.
Harbott
J.
, et al.  . 
Cluster analysis of genomic ETV6-RUNX1 (TEL-AML1) fusion sites in childhood acute lymphoblastic leukemia
Leuk. Res.
 , 
2009
, vol. 
33
 (pg. 
1082
-
1088
)
42
Kaplan
E.L.
Meier
P.
Nonparametric estimation from incomplete observations
J. Am. Stat. Assoc.
 , 
1958
, vol. 
53
 (pg. 
457
-
481
)
43
Mantel
N.
Evaluation of survival data and two new rank order statistics arising in its consideration
Cancer Chemother. Rep.
 , 
1966
, vol. 
50
 (pg. 
163
-
170
)
44
Kalbfleisch
J.D.
Prentice
R.L.
The Statistical Analysis of Failure Time Data
 , 
1980
New York
John Wiley and Sons
(pg. 
163
-
188
)
45
Gray
R.J.
A class of K-sample tests for comparing the cumulative incidence of a competing risk
Ann. Stat.
 , 
1988
, vol. 
16
 (pg. 
1141
-
1154
)

Supplementary data