Supplementary Figure 1 TCF3 breakpoints of TCF3-PBX1 (patients 1a 5a) and TCF3-HLF (patients 6a 9a and11a) translocations. The CpG motifs closest to the breakpoints are highlighted in red boxes and the closest cac motifs in blue.
Supplementary Figure 2 The TCF3-HLF translocation is only detected in the lymphoid progenitor gate. Experiments, combining sorting of specific cell populations using flow cytometry (a) and subsequent detection of the fusion gene in sorted cells employing FISH probes or real-time PCR (b), were performed using bone marrow from the TCF3-HLF-positive patients 15,
16, and 17 of the validation cohort. Bone marrow cells were sorted into HSC (CD34 + CD38 - CD90 + CD45RA - ) and MLP (CD34 + CD38 - CD90 neg-low CD45RA + ) populations as previously described (Doulatov et al. Nature Immunology 11, 585-593 (2010)). Leukemic cells (CD34 - CD38 + CD19 + CD10 + ) were sorted according to standard diagnostic markers. Sorted cell populations were analyzed by FISH and real-time PCR as described in the Online Methods section. All sorted leukemia cells carried the translocation (FISH signals 2R, 1G, 1B, 2B/G fusion) while none of the scored HSC or MLP (2R, 2G, 2B) populations were positive. The TCF3-HLF fusion was detected only in blast populations and not in myeloid cells indicating that the TCF3-HLF-positive leukemia originated in committed lymphoid cells. For patient 15, the results were confirmed using a TCF3-HLF-fusion-specific quantitative real-time PCR on a fraction of sorted cells as described in the supplementary online methods section. Leukemic cells were 100% positive for the TCF3-HLF-fusion, whereas the MLP and the HSC fractions were negative. As a positive control the reference gene -globin was employed. The TCF3-HLF fusion was detected by real-time PCR only in the blast population, but not in myeloid cells, indicating that the TCF3-HLF-positive leukemia may have originated from committed lymphoid cells.
Supplementary Figure 3 BTG1 deletions. (a) Whole genome sequencing (WGS) read ratio plot showing a BTG1 deletion identified in the diagnostic sample of patient 8. This BTG1 deletion was present in a subclonal population of ~14% of leukemia cells. (b) WGS read ratio plot showing a BTG1 deletion identified in the diagnostic sample of patient 9. (c) Exome sequencing read ratio plot for a xenograft (7b_X) derived from the remission sample of patient 7 showing a BTG1 deletion. (d) The BTG1 deletion in xenograft 7b_X ia also visible in the SNP allele frequency data from exome sequencing, inspecting the SNP rs709222 in BTG1 exon 2.
a chromosome 1p35.1 223 kb deletion in 6a chr1: 32.50 32.55 32.60 32.65 32.70 Mbp [hg19] KHDRBS1 TMEM39B KPNA6 DCDC2B LCK TXLNA EIF3I CCDC28B FAM167B IQCC b Fusion border RNA-seq read coverage (RPM) 1.2 0 KHDRBS1 c 0.8 0 LCK isoform 2 FAM167B LCK isoform 1 d chr1:32479954-32479978(+)...cacgccatgcagctgctgacggcag chr1:32739926-32739950(+) GGACCATGGGCTGTGGCTGCAGCTC... fused exons e GCTCGACCCGTCCTTCACTCACGCCATGCAGCTGCTGACGGCAG GTCCTTCACTCACGCCATGCAGCTGCTGACGGCAG CTTCACTCACGCCATGCAGCTGCTGACGGCAG ACTCACGCCATGCAGCTGCTGACGGCAG GGACCAT (1x) GGACCATGGGCTGTGG (1x) GGACCATGGGCTGTGGCTG (1x) GGACCATGGGCTGTGGCTGCAGC (1x) RNA-seq reads on fusion TGCAGCTGCTGACGGCAG CTGCTGACGGCAG GCTGACGGCAG GGACCATGGGCTGTGGCTGCAGCTCACACCCGG (5x) GGACCATGGGCTGTGGCTGCAGCTCACACCCGGAAGAT (3x) GGACCATGGGCTGTGGCTGCAGCTCACACCCGGAAGATGA (1x) CTGACGGCAG GGACCATGGGCTGTGGCTGCAGCTCACACCCGGAAGATGAC (1x) KHDRBS1 exon 1... D P S F T H A M Q L L T A LCK exon 2 (in-frame) fusion protein G T M G C G C S S H P E D D... Supplementary Figure 4 KHDRBS1-LCK fusion gene in patient 6a. (a) A heterozygous deletion encompassing 223 kb of genomic DNA on chromosome 1 fused intron 1 of KHDRBS1 to the promoter region of the LCK proto-oncogene. (b) Transcriptome sequencing showed a decrease of KHDRBS1 expression after exon 1. (c) LCK expression was induced starting from exon 2. (d) Analysis of chimeric fusion reads showed the fusion of KHDRBS1 exon 1 with LCK exon 2, confirmed by 14 chimeric reads spanning the fusion border. (e) Amino acid sequence of the in-frame fusion protein encoded by the KHDRBS1-LCK fusion transcript.
Supplementary Figure 5 Recurrent fusion of the genes KHDRBS1 and LCK in ALL. RNA was isolated from samples of a cohort of 74 unselected pediatric ALL and transcribed into cdna. Nested PCR was performed using specific primers for the KHDRBS1-LCK fusion transcript detected in patient 6a. (a) The gel electrophoresis presents the expected specific PCR product of 226 bp in the sample 6a and three patient samples of the validation cohort as well as controls. The negative control is representative of the 71 patient samples of this validation cohort tested negative. (b) The chromatograms display the sequencing result for patient 6a (upper panel) and a patient of the validation cohort (patient D3, lower panel) as an example. The fusion of KHDRBS1 exon 1 to LCK exon 2 was confirmed by Sanger sequencing in all positive cases (3/74).
Supplementary Figure 6 Pattern of recurrent genomic lesions and model of leukemogenesis in TCF3-HLF positive leukemia. Left panel: Recurrent genetic lesions identified in TCF3-HLF-positive ALL affect transcription factors regulating lymphoid differentiation, the Ras pathway and cell cycle regulation. Besides the 12 cases analyzed in the present study, a single case reported in a recent study is included (Ma et al. Nature Communications 6, 6604 (2015), DOI: 10.1038/ncomms7604) (x indicates a KRAS mutation that was detected only in a patient derived xenograft sample, not in the primary tissue. Note that the cases 16 and 17 have not been analyzed by next generation sequencing and besides PAX5 deletion, NRAS and KRAS hotspot mutations other genetic alterations possibly present are not yet analyzed). Right panel: The TCF3-HLF gene fusion probably occurs in an early B cell progenitor. B cell differentiation is blocked by mutations in transcription factors playing key roles in lymphoid development. Secondary mutations involving the Ras pathway and cell cycle regulators lead to leukemia development.
Supplementary Figure 7 Engraftment of TCF3-PBX1 and TCF3-HLF positive ALL samples. ALL cells were recovered from cryopreserved samples and transplanted into NSG mice. The percentage of mcd45 - hcd19 + hcd45 + leukemic cells was monitored by flow cytometry every 4 weeks. Upper panels: Primary and secondary transplantation of diagnostic TCF3-PBX1-positive samples. Middle panels: Primary and secondary transplantation of diagnostic TCF3-HLF-positive samples. Lower panels: Primary and secondary transplantation of MRD and relapse TCF3-HLF-positive samples. MRD samples were successfully amplified only for the TCF3-HLF-positive subtype. Patient samples were transplanted intrafemorally (primary transplantation). Xenograft-samples were transplanted intravenously (secondary transplantation). Mean and SD of 2-3 mice is shown when available. The number of living cells transplanted is indicated in the legend.
Supplementary Figure 8 Expression of PAX5 in primary and patient derived xenografts (PDX) of TCF3-PBX1 and TCF3-HLF positive ALL samples. The significant differential expression of PAX5 between TCF3-PBX1-positive and TCF3-HLF-positive primary ALL is preserved in patient derived xenograft samples.
Supplementary Figure 9 Cell survival and cell cycle analysis of TCF3-PBX1 and TCF3-HLF positive ALL cells in coculture on hmscs. Upper panel: ALL cell viability comparing co-cultures to ALL cell suspension cultures after 72 h by flow cytometry using 7-AAD staining. The percentage of living cells normalized to the input cells seeded on day 0 is shown (mean with SD of duplicates). Lower panel: Percentage of cells of the population in the G 0 G 1, S and G 2 M phases of the cell cycle. Flow cytometric analysis was performed at day 1 of co-culture with hmsc after 20 h of EdU incubation. Mean and SD of at least 2 independent experiments performed in triplicates are shown.
Supplementary Figure 10 Drug profiling of TCF3-PBX1 and TCF3-HLF ALL xenografts derived from diagnostic, MRD and relapse samples. Unsupervised clustering of xenograft samples according to their response (log IC 50 ) to 98 compounds in co-culture on hmsc after 72 h. Xenografts derived from diagnostic a, MRD b and relapse c samples are included. Fitted values are provided in the supplementary section (absolute IC 50 ). Compounds are arranged based on the family of the targets. Numbers identify the compounds shown in Figure 5c.
Supplementary Figure 11 In vitro dose response curves to ABT-199 (venetoclax) comparing xenografts derived from TCF3-HLF-positive ALL patients at the time point of diagnosis 'a', MRD 'b' and relapse 'c'. ALL co-cultures were treated with ABT-199 for 72 hours, the responses were normalized against DMSO-treated controls..
Supplementary Figure 12 Drug activity profiling of two additional TCF3-HLF translocated leukemia cases of the validation cohort (15, 17) reveals similar patterns of drug sensitivities and high responsiveness to ABT-199 (venetoclax) compared to the TCF3-HLF positive patients of the discovery cohort. Selection of drugs based on differences in sensitivity or resistance in TCF3-PBX1-positive compared to TCF3-HLF-positive patients (Figure 5c). Similar patterns of drug sensitivity were observed testing patient derived primary (patient 15, 17) and xenograft cells (other TCF3-PBX1- and TCF3-HLF-positive cases shown). Drugs active across both patient cohorts include doxorubicin, idarubicin, mitoxantrone, bortezomib, and ABT-199 (venetoclax).
Supplementary Figure 13 In vitro dose response co-titration curves to ABT-199/dexamethasone (top) and ABT-199/vincristine (bottom) for patient derived xenografts from TCF3-HLF positive (6a 11a) ALL patients. Xenograft cells were seeded on a layer of stroma cell and treated with the indicated combinations and concentrations of ABT-199, vincristine and dexamethasone, respectively, for 72 h. Drug concentrations used in these assays are provided in Supplementary Table 26. The response was normalized against DMSO-treated controls. The co-titration curves indicate that ABT-199 combined with either vincristine or dexamethasone likely exhibits synergistic activity in the patients marked with (*); combination indices (see Table) were calculated based on the Chou-Talaly method (Chou and Talaly JBC 252: 6438-6442, (1977)).
Supplementary Table 1. Characteristics of patients with TCF3-PBX1 and TCF3-HLF fusion-positive pediatric acute lymphoblastic leukemia included in initial sequencing analyses. TCF3-PBX1 TCF3-HLF Patient 1 2 3 4 5 6 7 8 9 11 Sex male female male female female male female male male male Age at diagnosis (years) 15.5 14.5 10.8 1.8 15.5 16.5 14.5 14.1 13.8 15.5 WBC a /!l at diagnosis 26,800 90,600 10,370 102,000 15,520 14,300 2,200 6,300 5,800 20,300 CNS status b CNS1 CNS1 CNS1 CNS2 CNS1 CNS1 CNS1 CNS1 CNS1 CNS1 Prednisone response c good poor poor good poor good good poor good good Morphological remission in bone marrow after induction treatment d M1 M1 M1! M1! M1! M1! M1! M1! M1 M3 MRD after induction negative 10-4 treatment e Treatment outcome f CCR CCR after HSCT negative negative negative 10-3 10-3 10-1 10-3 10-2 CCR CCR CCR TRM TRM relapse and death after HSCT relapse and death relapse and death after HSCT Leukemic specimen g used for sequencing (% blasts) BM (96) PB (92) BM (98) BM (90) BM (94) BM (98) BM (81) BM (61) BM (72) BM (88) Normal reference tissue BM BM BM BM BM BM BM BM BM BM Treatment protocol h BFM BFM BFM! BFM! BFM! BFM BFM BFM FRALLE BFM a white blood cell count; b central nervous system (CNS) status: CNS1 (lumbar puncture nontraumatic, no leukemic blasts in the cerebrospinal fluid (CSF) after cytocentrifugation); CNS2 (puncture nontraumatic, "5 leukocytes/!l CSF with identifiable blasts; the prognostic impact of CNS2 status is debated); CNS3 (puncture nontraumatic, >5 leukocytes/!l CSF with identifiable blasts); TLP+ (traumatic lumbar puncture with identifiable leukemic blasts); a TLP with no identifiable blasts is not an adverse factor. For cytomorphological examination, CSF samples should be analyzed after cytospin preparation, a method through which cellular components within the CSF are concentrated by centrifugation; c assessed after 7 days induction with daily prednisone and a single intrathecal dose of methotrexate on treatment day 1, good: < 1000/!l blood blasts, poor: # 1000/!l blood blasts; d < 5% blasts (M1), # 5% blasts (M2 or M3); e assessed by DNA-PCR-based monitoring of leukemic immunoglobuline and/or T-cell receptor rearrangements, markers required to have a sensitivity of at least 10-4 ; f CCR: continuous complete remission, HSCT: hematopoietic stem cell transplantation, TRM: treatment-related mortality; g BM: bone marrow, PB: peripheral blood; h BFM: Berlin-Frankfurt-Münster study group.!!!
Supplementary Table 2. Characteristics of TCF3-HLF-positive pediatric acute lymphoblastic leukemia patients included in validation analysis. Patient 10 12 13 14 15 16 17 Sex female female male male male female female Age at diagnosis (years) 12.7 18.2 14.7 3.4 13.6 14.3 14 WBC a /µl at diagnosis 5,000 5,200 unknown 4,240 3,470 6,150 4,450 CNS status b CNS1 CNS1 CNS1 CNS1 CNS1 CNS1 CNS1 Prednisone response c not applicable not applicable good good good good not applicable Morphological remission in bone marrow after induction M1 M1 M1 M3 M1 M2 no treatment d 6x10 MRD after induction treatment e 10-1 unknown negative 3.5x10-1 1x10-4 10-1 unknown Treatment outcome f relapse and death after SCT relapse and death after SCT relapse and death relapse and death after SCT -1 remission after SCT relapse after HSCT primary refractory Leukemic specimen g used for sequencing (% blasts) BM (53) BM (unknown) PB (79) BM (80) BM (93.5) not applies not applies Normal reference tissue BM BM BM not available not available not applies not applies Treatment protocol h UKALL UKALL FRALLE BFM BFM BFM AALL1131 PAX5 deletion/mutation no no yes yes yes unknown unknown BTG1 deletion/mutation uncertain yes no no no unknown unknown VPREB1 deletion no yes yes no no unknown unknown TCF3 mutation - - S467G - - unknown unknown RAS pathway mutation i - SPHK1 P16H - PTPN11 T73I PTPN11 Q86R unknown unknown
a white blood cell count; b central nervous system (CNS) status: CNS1 (lumbar puncture nontraumatic, no leukemic blasts in the cerebrospinal fluid (CSF) after cytocentrifugation); CNS2 (puncture nontraumatic, 5 leukocytes/µl CSF with identifiable blasts; the prognostic impact of CNS2 status is debated); CNS3 (puncture nontraumatic, >5 leukocytes/µl CSF with identifiable blasts); TLP+ (traumatic lumbar puncture with identifiable leukemic blasts); a TLP with no identifiable blasts is not an adverse factor. For cytomorphological examination, CSF samples should be analyzed after cytospin preparation, a method through which cellular components within the CSF are concentrated by centrifugation; c assessed after 7 days induction with daily prednisone and a single intrathecal dose of methotrexate on treatment day 1, good: < 1000/µl blood blasts, poor: 1000/µl blood blasts; d < 5% blasts (M1), 5% blasts (M2 or M3); e assessed by DNA-PCR-based monitoring of leukemic immunoglobuline and/or T-cell receptor rearrangements, markers required to have a sensitivity of at least 10-4 ; f CCR: continuous complete remission, HSCT: hematopoietic stem cell transplantation, TRM: treatment-related mortality; g BM: bone marrow, PB: peripheral blood; h BFM: Berlin-Frankfurt-Münster study group.
Supplementary Table 3. Characteristics of patients with TCF3-PBX1 fusion-positive pediatric acute lymphoblastic leukemia included in validation analyses a. TCF3-PBX1 (n = 24) n (%) Sex male 11 (45.8) female 13 (54.2) Age at diagnosis (years) < 1-1 - 9 14 (58.3) # 10 10 (41.7) WBC b /!l at diagnosis < 10,000 3 (12.5) 10,000 49,999 15 (62.5) 50,000 99,999 3 (12.5) #100,000 3 (12.5) Prednisone response c good 20 (83.3) poor 4 (16.7) MRD d TP 1 + 2 negative 9 (37.5) other 10 (41.7) TP 2 # 10-3 4 (16.7) unknown 1 (4.2) EBF1 deletion e negative 16 (66.7) positive 3 (12.5) unknown 5 (20.8) CDKN2A/B deletion e negative 16 (66.7) positive 3 (12.5) unknown 5 (20.8) PAX5 deletion e negative 18 (75.0) positive 1 (4.2) unknown 5 (20.8) ETV6 deletion e negative 18 (75.0) positive 1 (4.2) unknown 5 (20.8) RAS pathway mutation f negative 23 (95.8) positive 1 (4.2) Relapse no 22 (91.7) yes 2 (8.3)
a 24 patients included in validation of RAS pathway mutation screening, 19 patients included in validation of deletions associated with B-cell differentiation (e.g. PAX5) by MLPA; b white blood cell count; c assessed after 7 days induction with daily prednisone and a single intrathecal dose of methotrexate on treatment day 1, good: < 1000/!l blood blasts, poor: # 1000/!l blood blasts; d MRD: minimal residual disease as assessed by DNA-PCR-based monitoring of leukemic immunoglobuline and/or T-cell receptor rearrangements, markers required to have a sensitivity of at least 10-4 ; TP: time point, TP1: MRD after induction at treatment day 33, TP2: MRD at treatment day 78, MRD #10-3 qualifies for the high-risk group; e five out of 19 patients affected, one patient harbored several deletions (CDKN2A and B, PAX5, ETV6, BTG1), one patient demonstrated co-occurrence of EBF1 and CDKN2A and B, the remaining three patients had single occurrences of EBF1 (n=2) or CDKN2A and B; f RAS pathway screening by Sanger sequencing included: KRAS exon 1, NRAS exon 1 and exon 2, FLT3 exon 14 and 20, PTPN11 exons 3 and 13, one patient found positive for an NRAS exon 2, codon 61 (CAA>CGA, Q61R) activating mutation.!!!
Supplementary Table 4. Genomic breakpoint coordinates, fused exons and nucleotides at the junctions of TCF3-PBX1 and TCF3-HLF gene fusions detected in pediatric ALL patient samples. Coordinates of breakpoints were obtained from Sanger sequencing of PCR products from genomic DNA or cdna. Ins=Insertion; Mhom=microhomology; TpIns=Templated insertion [*:Chr17; $:ChrX].
Supplementary Table 8. Characteristics of pediatric acute lymphoblastic leukemia patients from trial AIEOP-BFM ALL 2000 included in mutation screening of TCF3 exon 19. AIEOP-BFM ALL 2000 cohort n = 1033 n (%) Sex male 593 (57.4) female 440 (42.6) Age at diagnosis (years) < 1 1 (0.1) 1-9 749 (72.5)! 10 283 (27.4) WBC a /"l at diagnosis < 10,000 405 (39.2) 10,000 49,999 365 (35.3) 50,000 99,999 128 (12.4)!100,000 135 (13.1) Prednisone response b good 899 (87.0) poor 126 (12.2) unknown 8 (0.8) MRD c TP 1 + 2 negative 359 (34.8) other 433 (41.9) TP 2! 10-3 65 (6.3) unknown 176 (17.0) Immunophenotype B-cell 840 (81.3) T-cell 181 (17.5) unknown/other 12 (1.2) BCR/ABL1 negative 1014 (98.2) positive 19 (1.8). MLL/AF4 negative 1029 (99.6) positive 4 (0.4) ETV6/RUNX1 negative 754 (73.0) positive 208 (20.1) unknown 71 (6.9) a white blood cell count; b assessed after 7 days induction with daily prednisone and a single intrathecal dose of methotrexate on treatment day 1, good: < 1000/"l blood blasts, poor:! 1000/"l blood blasts; c MRD: minimal residual disease as assessed by DNA-PCR-based monitoring of leukemic immunoglobuline and/or T-cell receptor rearrangements, markers required to have a sensitivity of at least 10-4 ; TP: time point, TP1: MRD after induction at treatment day 33, TP2: MRD at treatment day 78, MRD!10-3 qualifies for the high-risk group.!!