Deletion 6Q Drives T-Cell Leukemia Progression by Ribosome Modulation

Published OnlineFirst September 28, 2018; DOI: 10.1158/2159-8290.CD-17-0831

RESEARCH ARTICLE

Deletion 6q Drives T-cell Leukemia

Progression by Ribosome Modulation

Stéphanie Gachet 1 ,2 , Tiama El-Chaar 1 ,2 , David Avran 1 ,2 ,3 , Eulalia Genesca 1 ,2 , Frédéric Catez 4 , Samuel Quentin 1 ,2 ,3 , Marc Delord 2 , Gabriel Thérizols 4 , Delphine Briot 1 ,2 ,3 , Godelieve Meunier 1 ,2 , Lucie Hernandez 1 ,2 , Marika Pla 2 ,5 , Willem K. Smits 6 , Jessica G. Buijs-Gladdines 6 , Wouter Van Loocke7 , Gerben Menschaert 7 , Isabelle André-Schmutz 8 , Tom Taghon 7 , Pieter Van Vlierberghe7 , Jules P. Meijerink 6 , André Baruchel 2 ,9 , Hervé Dombret 2 ,10 , Emmanuelle Clappier 1 ,2 ,3 , Jean-Jacques Diaz 4 , Claude Gazin 11 , Hugues de Thé1 ,2 , François Sigaux 1 ,2 ,3 , and Jean Soulier 1 ,2 ,3

ABSTRACT Deletion of chromosome 6q is a well-recognized abnormality found in poor- prognosis T-cell acute lymphoblastic leukemia (T-ALL). Using integrated genomic approaches, we identifi ed two candidate haploinsuffi cient genes contiguous at 6q14,SYNCRIP (encoding hnRNP-Q) and SNHG5 (that hosts snoRNAs), both involved in regulating RNA maturation and translation. Combined silencing of both genes, but not of either gene alone, accelerated leukemogeneis in a Tal1/Lmo1/Notch1 -driven mouse model, demonstrating the tumor-suppressive nature of the two-gene region. Proteomic and translational profi ling of cells in which we engineered a short 6q deletion by CRISPR/Cas9 genome editing indicated decreased ribosome and mitochondrial activities, suggesting that the resulting metabolic changes may regulate tumor progression. Indeed, xenograft experiments showed an increased leukemia-initiating cell activity of primary human leukemic cells upon coextinction of SYNCRIP and SNHG5. Our fi ndings not only elucidate the nature of 6q deletion but also highlight the role of ribosomes and mitochondria in T-ALL tumor progression.

SIGNIFICANCE: The oncogenic role of 6q deletion in T-ALL has remained elusive since this chromosomal abnormality was fi rst identifi ed more than 40 years ago. We combined genomic analysis and functional models to show that the codeletion of two contiguous genes at 6q14 enhances malignancy through deregulation of a ribosome–mitochondria axis, suggesting the potential for therapeutic intervention. Cancer Discov; 8(12); 1614–31. ©2018 AACR.

INTRODUCTION often associated with a worse prognosis. This is exemplifi ed by the considerable efforts that were developed over several The identifi cation of somatic genome aberrations in can- decades to understand and model oncogenesis for abnormali- cer cells has been a highly fruitful strategy to identify cancer ties such as monosomy 7 in myeloid malignancies or deletion genes, especially in hematopoietic malignancies. With the 6q in T-cell acute lymphoblastic leukemia (T-ALL). T-ALL is availability of massive sequencing tools, it is now relatively related to leukemic transformation of T-cell progenitors ( 2, easy to identify cancer genes from translocations, point muta- 3 ). In addition to a founding rearrangement that deregulates tions, or short insertion/deletions ( 1 ). However, the molecu- transcription factor genes and determines the oncogenic lar targets of large chromosomal losses or gains can remain subtype, such as TAL1/2, LMO1/2, TLX1/3, and HOXA, the diffi cult to elucidate, and these types of abnormalities are T-ALL genome harbors a myriad of additional mutations, deletions, and duplications that collectively lead to overt leukemia (3–10 ). These have a particular impact on the cell cycle (CDKN2A/p16/ARF deletion), the Notch pathway (NOTCH1 1INSERM UMR944 and CNRS UMR7212, Hôpital Saint-Louis, Paris, and FBXW7 mutations), and the JAK/STAT and PI3K/AKT France. 2 Institute of Hematology (IUH), Université Paris Diderot, pathways (IL7R and PTEN mutations). Deletion of the 3 Sorbonne Paris Cité, Paris, France. Hematology Laboratory APHP, Hôpital long arm of chromosome 6 (del6q) is a frequent karyotypic Saint-Louis, Paris, France. 4 Centre de Recherche en Cancérologie de Lyon, INSERM U1052, CNRS 5286, Centre Léon Bérard; Université abnormality in T-cell ALL and lymphoblastic lymphoma, Lyon 1, Lyon, France. 5 INSERM UMRS 940, Hôpital Saint-Louis, Paris, where it has been associated with an unfavorable prognosis France. 6 Department of Pediatric Oncology/Hematology, Princess ( 11–13 ). Although fi rst reported as a recurrent structural 7 Maxima Center for Pediatric Oncology, Utrecht, the Netherlands. Cancer abnormality in lymphoblastic leukemia in 1976, the under- Research Institute, Ghent University, Ghent, Belgium. 8U1163 INSERM, Université Paris Descartes-Sorbonne Paris Cité, Institut Imagine, Paris, lying molecular targets of this chromosomal event remain France. 9 Hematology Pediatry Department, Robert Debré Hospital, Paris, elusive (14 ). France. 10 Hematology Department, Hôpital Saint-Louis, Paris, France. In our study, we used a large range of integrated genomic 11Centre National de Recherche en Génomique Humaine (CNRGH), Insti- and functional analyses and identifi ed two genes simultane- tut de Biologie François Jacob, Direction de La Recherche Fondamentale, ously inactivated through del6q, the combined haploinsuf- CEA, Evry, France. fi ciency of which accelerates T-ALL progressionin vivo. By Note: Supplementary data for this article are available at Cancer Discovery Online (http://cancerdiscovery.aacrjournals.org/). using a genome-editing CRISPR/Cas9 approach, we precisely engineered a short del6q in human cells that enabled us S. Gachet, T. El-Chaar, D. Avran, and E. Genesca contributed equally to this article. to examine the proteomic and translatome profi le of the Corresponding Author: Jean Soulier, Hematology Department INSERM deleted cells. We found that the haploinsuffi ciency of these U944 and Hematology Laboratory, Hôpital Saint-Louis, 1, Av Claude velle- two genes, one encoding a ribonucleoprotein and the other faux, Paris, France. Phone: 33-1-53-72-40-41; Fax: 33-1-42-49-40-27; hosting noncoding small nucleolar RNA (snoRNA), deregu- E-mail: [email protected] lates cellular metabolism and ultimately affects the leukemia- doi: 10.1158/2159-8290.CD-17-0831 initiating cell (LIC) activity of human T-ALL cells through ©2018 American Association for Cancer Research. modulation of ribosomal functions.

DECEMBER 2018 CANCER DISCOVERY | 1615

RESEARCH ARTICLE Gachet et al.

AB * TAL HOX HOXA TLX1 TLX3 Immature Case TL08 −4 −2 −1 0 +1 +2 +4 −40−2 −1 0 +1 +2 +4 −4 −2 −1 +1 +2 +4 p25.2 p24.3 p24.1

p22.3 p22.1 p21.32 p21.2

p12.3 p12.1

q12

q14.1 del6q q14.3

q16.1 q16.3

q22.1

q22.31 q22.33 q23.2

q24.1 q24.3 q25.2

q26

Chr. 6 Diag Xeno Relapse

C 1032 1753 1815 1941 2120 2649 2775 9083 9160 9376 9027 TL05 TL06 TL04 TL11 TL14 TL02 TL03 TL08 TL10 TL17 TL25 TL27 335 1842 1950 2759 2787 8815 9963 10111 TL16 TL07 TL13 TL01 TL12 TL15 258 704 768 1632 1948 1949 2117 2322 2436 2486 2669 2720 2722 2760 2774 2788 2789 2794 2844 2846 9243 9323 9827 9938 10110 TL34 TL19 TL26 TL28 TL31 TL18 TL20 TL21 TL24 TL30 TL22 TL23 TL29 del6q NOTCH1 FBXW7 PTEN

Mutated 6q deleted Wild-type Nondeleted

Figure 1. Del6q is associated with the TAL1-related T-ALL subtype as a late chromosomal event. A, aCGH showing copy-number losses (green) and gains (red) for chromosome 6 in the first T-ALL cohort n( = 78 cases). Cases were ordered by oncogenic subtypes as described (4). The region of del6q is boxed. *, Association between del6q and TAL1-related subtype, P < 0.05 (Fisher test). B, Different-sized del6q were found in T-ALL samples from the same patient (TL08) at diagnosis and relapse and in a xenograft raised from the diagnosis sample; by contrast, a core of common events was found in this patient—i.e., SIL-TAL and CDKN2A deletion (not shown)—demonstrating late occurrence of del6q events. C, Co-occurring NOTCH1, FBXW7, and PTEN gene mutations in the TAL-R case cohorts (cases with available mutation data are shown).

RESULTS and analyzed a total of 107 TAL1-related (TAL-R) cases from Del6q Is a Late Chromosomal Event in TAL1 three patient cohorts (4, 5, 15), in which del6q was detected in 34 Oncogene–Related T-ALL cases (32%). Del6q could be subclonal, and distinct del6q could be observed in the same sample, or in the diagnosis and relapse To identify tumor suppressor genes in this deletion, we ini- leukemia samples from the same patient, or in primary and tially screened deletion 6q by array comparative genomic hybrid- patient-derived xenografts (PDX; Supplementary Fig. S2; Fig. ization (aCGH) in a cohort of 78 primary T-ALL cases that were 1B). Additional, co-occurring mutations in NOTCH1/FBXW7 or previously characterized by large-scale expression profiling (ref. PTEN genes were often found, as previously reported in TAL-R 4; see flow chart of the study in Supplementary Fig. S1). Del6q cases, suggesting multistage oncogenesis (Fig. 1C; refs. 3, 7). was associated with the T-ALL subtype that is characterized by Collectively, these data indicate that del6q is strongly associated aberrant expression of the TAL1 oncogene (P = 0.032), suggest- with the TAL-R subtype in which it occurs as a late-stage chro- ing oncogenic cooperation (Fig. 1A). We focused on this subtype mosomal event associated with leukemia progression.

1616 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

SYNCRIP

38 kb 52 kb

SNHG5 CDR del6q 4.9 Mb

Case TL97 Case ALEK21

MYB

–1 0 D Gene expression 4 C 2

Case 0

ALEK21 Score –2 TTTTGGAATTGC /ta/TCAACCCT TG A SYNCRIP/SNHG5 –4 Case TL97 4.9 Mb CDR del6q AGTCAGTG/ttagatcactccgatcacctctctcc/CCAGTGCTCT Nondel

1.5 Case ALEK21 (52.05 kb) 1.0 Case TL97 (38.26 kb) Centromeric BPs 0.5 SYNCRIP SNHG5 SYNCRIP SNHG5 0.0 SYNCRIP SNORD50A SYNCRIP SNORD50B –0.5 SYNCRIP SYNCRIP –1.0 SYNCRIP

Telomeric BPs values Centered expression –1.5 cases ALEK21 and TL97 DD2 ME1 IBTK NT5E TPBG PGM3 TBX18 SNX14 UBE3D MRAP2 SNHG5 RW PRSS35 SNAP91 CYB5R4 RIPPLY2 DOPEY1 SYNCRIP SNHG5 SNORD50A SNORD50B KIAA1009

U50B U50A * * * * *

Figure 2. Integrated genomic analysis of del6q in T-ALL and identification ofSYNCRIP–SNHG5 as candidate haploinsufficient tumor suppressor genes. A, Stacked aCGH 6q profiles of T-ALL cases from theTAL1 -related subgroup are shown along the chromosome. The log2 scale of copy-number ratio is indicated. Each vertical line corresponds to one deletion; thick lines indicate deletions with 0.75 to 1 ratio, and thin lines deletions with 0.1 to 0.75 ratios (subclonal). All 6q deletions were mapped, and a CDR of 4,891,544 bp containing 18 genes is shown in light red; a never-deleted region, containing the MYB gene, is shown in light blue. B, High-density custom genomic profiling detected a short deletion ofSYNCRIP–SNHG5 in two cases. No additional somatic point mutation or insertion/deletions involving these two genes was found in 30 additional TAL-R T-ALLs with or without del6q. C, Molecular characterization of the two SYNCRIP–SNHG5 short deletions. Genomic breakpoints of the deletions were amplified using the flanking primers Fw 5′-CCCCATCTCCAGAAAATCAA-3′ and Rev 5′-CCTGACACTTTTAACAGGTATGTG-3′ (case ALEK21), and Fw 5′-CACAGTGGAGCAGCTCTGAA-3′ and Rev 5′-TCACTGGCTACTCGTCCACA-3′ (case TL97). Arrows indicate the breakpoints, and the lower-case letters in the electropherograms indicate non– template-inserted nucleotides. All positions are given in hg19 assembly. The exon–intron organization is shown according to the UCSC genome browser (genome.ucsc.edu). D, Differential gene expression in TAL-R cases with or without del6q is shown with respect to chromosomal localization of the probe sets using the MACAT package. Red line, smoothed regularized t scores in del6q cases; yellow line, significant underexpression oft scores from del6q cases; gray lines, upper (97.5% quantile) and lower (2.5% quantile) significance borders of permutation scores in nondeleted cases. The red box indicates the CDR at 6q14 on chromosome 6; the arrow shows SYNCRIP and SNHG5 gene positions. Bottom, gene-expression box plots are shown corresponding to del6q status for each of the 18 genes of the CDR. Five genes with significant low expression (*,P < 0.05 using the Mann–Whitney test) are indicated, including SYNCRIP and SNHG5 (highlighted in red).

Two Contiguous Genes, SYNCRIP and SNHG5, Are 4.9 Mb at 6q14 that included 18 genes (Fig. 2A; Supplemen- Candidate Haploinsufficient Tumor Suppressors tary Table S1). Located just beside the telomeric border, the at 6q14 MYB oncogene was never included in the deletion, suggesting We then aimed to identify the molecular targets of chro- that loss of one MYB copy would limit T-ALL cell fitness (Fig. mosome 6q deletion. Del6q was interstitial in all cases, het- 2A; the alternative hypothesis of a MYB positional activation erogeneous in size, with a common deleted region (CDR) of driven by del6q was ruled out by allele expression analysis; see

DECEMBER 2018 CANCER DISCOVERY | 1617

RESEARCH ARTICLE Gachet et al.

Supplementary Fig. S3). We then focused on the CDR using a mouse model of tumor progression in the context of TAL1- custom 6q aCGH and high-throughput sequencing tools related leukemogenesis (Fig. 4A). To mimic human TAL-R (Supplementary Figs. S1 and S4). Although no deleterious primary cases (Fig. 1C), we collected bone marrow progenitor point mutation or short insertion/deletions were discovered, cells from young Tal1tgLmo1tg mice (known to develop an focal somatic deletions of 52 kb and 38 kb, respectively, were overt T-ALL by the age of 6 months; ref. 21), transduced them found in two T-ALL cases without large del6q (Fig. 2B; Sup- with a Notch1 construct (22, 23), and transplanted them into plementary Table S2). Critically, only two genes, SYNCRIP RAG−/−γc−/− mice. This resulted in a fully penetrant lethal and SNHG5, were involved in both short deletions (Fig. 2B T-ALL with a median latency of 60 days (Fig. 4B), in which we and C). tested whether additional silencing of murine Syncrip and/ In parallel, we analyzed gene-expression data in the T-ALL or Snhg5 by shRNAs might accelerate further the onset of cases (including n = 107 TAL-R with n = 34 cases harbor- overt leukemia. Notably, we chose an shRNA approach rather ing a del6q) and found an overall decreased expression of than a genetic inactivation of the Syncrip and Snhg5 mouse 6q13–q22 genes consistent with a gene dosage effect (Fig. locus, considering that the snoRNAs from Snhg5 are dupli- 2D). The downregulated transcripts in T-ALL cases with cated several times in the mouse genome. As shown in Fig. del6q included five genes from the CDR, andSYNCRIP and 4B, recipients of cells transduced with individual Syncrip or SNHG5 in particular (Fig. 2D; Supplementary Fig. S5A). We Snhg5 shRNA developed leukemia with a comparable median also analyzed healthy human thymic subset data and found latency to the control (59, 60, and 60 days, respectively). that both genes were differentially regulated during normal Strikingly, mice transplanted with progenitor cells silenced T-cell development, with a peak at the immature single posi- for both Syncrip and Snhg5 succumbed to T-ALL significantly tive (ISP) stage and downregulation after β-selection (Fig. earlier (median latency of 50 days; Global Cox likelihood 3A). In addition, chromatin immunoprecipitation sequenc- ratio test, P = 0.00013; Fig. 4B). Therefore, combined Syncrip ing data showed that TAL1 can bind regulatory regions and Snhg5 silencing, but not individual silencing, significantly of both the SYNCRIP and SNHG5 genes, suggesting tran- accelerates Tal1/Lmo1/Notch1-induced T-cell leukemogenesis scriptional regulation (Supplementary Fig. S5B). We sorted in mice. Collectively, these data demonstrate that Syncrip and double-negative (DN3 and DN4) thymocytes from young Snhg5 function as a tumor suppressor region, the silencing of pSil-TSCL (Tal1tg) and Lck-LMO1 (Lmo1tg) transgenic mice which drives tumor acceleration in vivo. (Tal1tgLmo1tg) and did indeed find a significant increase of Syncrip and Snhg5 expression compared with age-matched 6q14 Microdeletion Engineered by Genome wild-type littermates (Fig. 3B and C), suggesting that del6q Editing Affects Ribosomal Functions, Translation counteracts TAL1-related sustained expression of these Programs, and Mitochondrial Respiration in Human genes in the human TAL-R subtype, and thereby explaining T-ALL Cells why the deletion is found in that subtype. To investigate the mechanisms of T-ALL tumor progres- SYNCRIP (also known as hnRNP-Q or NSAP1) encodes sion induced by SYNCRIP and SNHG5 haploinsufficiency, the heterogeneous nuclear ribonucleoprotein hnRNP-Q, we used the CRISPR/Cas9 genome-editing approach to whereas SNHG5 is a non–protein-coding gene hosting two engineer a 6q14 microdeletion in a human T-ALL cell C/D-box small nucleolar RNAs in its introns (U50A and line harboring a prototypical TAL1 rearrangement (TAL1d, U50B snoRNAs, also known as SNORD50A and SNORD50B). i.e., SIL-TAL microdeletion) but no del6q, and then com- Both genes have known functions in the regulation of pared the resulting deleted and nondeleted clones. Briefly, mRNA processing (alternative splicing, editing, trans- we engineered the 38-kb deletion precisely, as seen in port, and degradation) and translation (16–19). Specifi- patient TL97, by using two sets of Cas9 target-flanking cally, U50 snoRNAs regulate ribosomal biogenesis through sequences simultaneously (Fig. 5A) and generated several site-specific recognition of preribosomal RNAs to mediate SYNCRIP–SNHG5 microdeleted CCRF-CEM clones (Fig. 2′-O-methylation (20). The expression of the two genes was 5B). Immunoblot showed reduced protein levels of hnRNP- correlated with pathways involved in mRNA processing and Q (SYNCRIP gene product) in the Del6q/ΔSYNCRIP/ ribosome activities in normal human thymic populations ΔSNHG5 clones (clones A6 and C4) compared with the (Fig. 3D and E). In both 6q short deletions from T-ALL nondeleted clone (C2) and isogenic cell line (Fig. 5C). Con- patients, the ATG and most exons of SYNCRIP were lost, sidering the known function of hnRNP-Q and snoRNAs as were the snoRNA sequences in the introns of SNHG5, in regulating mRNA processing and translation (16–20), suggesting inactivation of both genes (Fig. 2C). Collectively, we undertook a comprehensive proteomic approach to our integrated genomic and expression analysis of del6q in examine biological functions that might be affected. Mass T-ALL converged on two candidate tumor suppressor genes spectrometry–based label-free quantitative proteomic anal- differentially regulated during T-cell differentiation and ysis identified a list of 253 proteins that were expressed whose haploinsufficiency may play a role at a late stage of differentially between the deleted versus nondeleted cells TAL1-associated T-cell leukemia. out of a total of 2,352 significantly quantified proteins (Fig. 5D; Supplementary Table S3). As expected, the SYNCRIP/ Syncrip–Snhg5 Cosilencing Accelerates Tumor hnRNP-Q protein level was lower in the deleted clones Development in a Tal1/Lmo1/Notch1-Induced (Fig. 5D). Strikingly, pathway enrichment analysis found T-ALL Mouse Model a global downregulation of a number of pathways, includ- To functionally investigate the effects of single or com- ing protein metabolic process, translational initiation, bined SYNCRIP and SNHG5 haploinsufficiency, we developed protein/RNA complex assembly, and mRNA processing

1618 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

A Normal human T-cell differentiation

SYNCRIP_X1555427_s_at 2 SYNCRIP_X209024_s_at

SYNCRIP_X209025_s_at 1 SYNCRIP SYNCRIP_X217832_at SYNCRIP_X217833_at 0 SYNCRIP_X217834_s_at

SYNCRIP_X236146_at –1 SNHG5 SNHG5_X225155_at CD34 CD34 CD34 CD34 CD34 CD34 CD4 CD4 CD4 CD4 CD4 CD4 CD4 CD4 CD4 CD4 CD8 CD8 + + + + + + + + + + + + CD3 CD3 CD3 CD3 CD8 CD8 CD8 CD8 CD3 CD3 CD3 CD3 + + + + + + CD1 CD1 CD1 CD1 CD1 CD1 – – – – + + + + + + + + CD28 CD28 CD28 CD28 CD3 CD3 CD3 CD3 − − + + + + CD4 CD4 CD4 CD4 CD4 CD4 − − + + − − + + − − − − + +

BC P = 0.008 ns 0.15 P = 0.002 0.020 P = 0.03 0.015 0.10 to Gapdh 0.010 0.05 0.005 Snhg5 relati ve

relative to Gapdh Syncrip relative 0 0

WT DN3 WT DN4 WT DN3 WT DN4 Tal1/Lmo1 DN3 Tal1/Lmo1 DN4 Tal1/Lmo1 DN3 Tal1/Lmo1 DN4

D SYNCRIP E SNHG5

KEGG gene set # Genes P KEGG gene set# Genes P Ribosome 105 3.40E–14 Ribosome 115 4.90E–15 Spliceosome 85 8.20E–09 RNA transport 124 1.80E–08 Oxidative phosphorylation 86 1.00E–08 Oxidative phosphorylation 91 1.70E–07 Protein processing in endoplasmic reticulum 110 2.90E–07 Huntington disease 129 5.10E–06 RNA transport 109 6.60E–07 Spliceosome 86 5.70E–06 Nucleotide excision repair 37 8.80E–07 Parkinson disease 91 3.00E–05 Aminoacyl tRNA biosynthesis 35 6.60E–06 Epstein–Barr virus infection 134 3.70E–05 Huntington disease 116 1.00E–05 mRNA surveillance pathway 65 4.60E–05 Parkinson disease 84 1.10E–05 Cell cycle 84 5.20E–05 Ribosome biogenesis in eukaryotes 55 1.90E–05 Reticulum 113 9.20E–05 Epstein–Barr virus infection 121 4.00E–05 Proteasome 36 1.80E–04 DNA replication 29 9.20E–05 Cell cycle 75 1.50E–04 Ribosome biogenesis in eukaryotes 115 8.00E–04 Proteasome 33 2.10E–04

Figure 3. SYNCRIP and SNHG5 expression in normal and pathologic human T cells. A, Gene expression in normal human thymic subsets. Subsets were flow-sorted from two healthy thymus donors, and the resulting RNAs were processed on Affymetrix arrays as described (49, 50).SYNCRIP and SNHG5 were expressed in double negative (DN) cells, peaked in immature single positive (ISP) cells, and were downregulated after β-selection, at the double positive (DP) and single positive (SP) stages. Consistent data were found using two other data sets (ref. 4 and GSE33470; not shown). B and C, Syncrip (B) and Snhg5 (C) increased expression in flow-sorted DN3 and DN4 thymocytes from youngTal1 tgLmo1tg mice compared with control littermates. D and E, Biological pathways whose gene-expression correlates positively with SYNCRIP (D) and SNHG5 (E) expression, respectively, in normal thymic cells using the R2: Genomics Analysis and Visualization Platform (http://r2.amc.nl) and GSE33470 data set. KEGG, Kyoto Encyclopedia of Genes and Genomes.

(Fig. 5E and F). In the light of these results, we examined an unbiased manner the 2′-O-methylation occupancy on the ribosome population fractions in deleted and nonde- N = 106 sites using a highly sensitive RiboMethSeq leted cells by profiling them on sucrose gradient. Although approach (24). Strikingly, only the two known SNORD50A the overall profile was unchanged, indicating that transla- target sites (28S-Cm2848 and 28S-Cm2863) were signifi- tion remained active, we found that the 80S fraction was cantly hypomethylated in the deleted clones when com- higher in the deleted cells, indicating a change in the trans- pared with nondeleted and isogenic lines, whereas the 104 lation machinery activity (Fig. 5G). We also quantified in other sites remained unaffected (Fig. 5H), showing the

DECEMBER 2018 CANCER DISCOVERY | 1619

RESEARCH ARTICLE Gachet et al.

A shRNAs NotchICN-NGFR

Tal1tg Lmo1tg Leukemia 3–5 weeks latency Lin–BM cells RAG–/–γc–/–

B 100 shRNAs Ctrl (n = 21) Syncrip (n = 8) 80 Snhg5 (n = 8) Syncrip–Snhg5 (n = 20)

60 Ctrl vs.: Syncrip–Snhg5, P = 10 −5 Syncrip, P = 0.29 40

Percent survival Percent Snhg5, P = 0.21

Syncrip–Snhg5 vs.: 20 Syncrip, P = 0.007 Snhg5, P = 0.011

10 20 30 40 50 60 70 80 Latency (days)

Figure 4. In vivo cosilencing of Syncrip and Snhg5 accelerates T-ALL in mice. A, Schematic of the protocol for T-ALL acceleration experiments. Bone marrow (BM) Lin− cells from young Tal1tgLmo1tg mice, transduced with the shRNA vectors Syncrip, Snhg5, tandem Syncrip–Snhg5, or Ctrl and with the Notch-ICN vector (48), were injected into sublethally irradiated RAG−/−γc−/− recipients. The use of a Notch1 signal was necessary to get T-ALL development using Tal1tgLmo1tg cells because leukemia did not develop spontaneously within 18 months in the transplantation setting (n = 30 mice, data not shown). NotchICN likely replaces Notch1 mutations that spontaneously occur in expanded thymic progenitors in the Tal1tgLmo1tg mice, leading to preleukemic transition from the DN to DP stage (22, 23). NOTCH1/FBXW7 mutations can also be seen in human TAL-R T-ALL with del6q (Fig. 1C), and data inferred from clonal architecture showed that del6q occurred at late stage, i.e., after SIL-TAL1 and NOTCH1 mutations (not shown), as modeled by our experimental setting. B, Kaplan–Meier T-ALL survival curves. Numbers of mice for each shRNA are indicated. Indicated P values were calculated for paired analyses using likelihood ratio tests from the Cox model, the global significance beingP = 0.00013.

specificity of the methylation change uponSNHG5 haplo- specific gene transcripts, we analyzed the actively translated insufficiency. To evaluate whether altered ′2 -O-methylation mRNAs in deleted and nondeleted cells by sequencing the and ribosome assembly might affect the global translation polysome-bound and total cytoplasmic mRNA fractions rate, we evaluated global protein synthesis by quantify- (translatome). Using dedicated translatome profiling tools ing the L-azidohomoalanine incorporation into nascent (25, 26), we found a differential TE of transcripts from proteins (Click-iT AHA assay), but did not find consistent 475 genes, these being mainly linked to RNA binding (up) changes in basal culture condition (not shown). How- and mitochondrial function (down) using Gene Ontology ever, we reasoned that hypoxic condition, mimicking the (GO) analysis (Fig. 5K and L; Supplementary Table S4). hematopoietic niche, might stress the system and reveal Specifically, the differential TE of mitochondria transcripts a differential response. Indeed, the protein neosynthesis suggested a potential downstream mechanism of leukemia was significantly lower in the deleted clones in hypoxic cell regulation by the deletion (Fig. 5L). Notably, in normal condition (Fig. 5I). Consistent results were also observed in thymic cells, the expression of SYNCRIP and SNHG5 genes hypoxia in independent experiments of 35S-labeled methio- correlates positively with oxidative phosphorylation pathway nine incorporation into nascent peptides (Fig. 5J). To genes, in agreement with a functional link (Fig. 3D and E). We analyze whether ribosome changes can affect biological thus performed a direct analysis of mitochondrial respiration processes through differential translation efficiency (TE) of using real-time measurements of oxygen consumption rate

1620 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

(OCR) and did indeed find reduced basal and maximal mito- SYNCRIP–SNHG5 Cosilencing Increases the chondrial respiration in the deleted clones compared with In Vivo LIC Activity of Human T-ALL Cells along nondeleted clones (Fig. 5M), whereas glycolysis was increased with a Deregulated Ribosomal and Mitochondrial (Fig. 5N). Collectively, these results point to deregulated ribo- Signature some activity arising from the 6q microdeletion in human Low metabolism including differential oxidative phos- T-ALL cells that results in metabolic changes especially in phorylation is thought to regulate normal and leukemic mitochondrial respiration. stem cell fate and functions (27–33). Given the known

AB CCRF-CEM Nondel del6q Patient 38Kb TL97 Iso C2 A6 C4 SYNCRIP SNHG5 SNHG5 SNORD50A SYNCRIP SNORD50B

Cas9 nucleases 38 kb

SNHG5 SYNCRIP INTRON 4 EF 1RFP-T2A-Puro-PolyA SNHG5 INTRON 4 0.5 kb 0.6 kb

C D E CCRF-CEM 22 KEGG gene set #Genes P 144 up del6q Nondel Regulation of protein metabolic process 43 0.000 20 Regulation of cellular protein metabolic process 39 0.000 Translational initiation 25 0.000 A6 C4 Iso C2 18 Regulation of translational initiation 19 0.000 16 hnRNP-Q Regulation of cellular component organization and biogenesis 41 0.000 hn RNP-Q Enzyme inhibitor activity 24 0.000 14 Regulation of metabolic process 157 0.000

LFQ intensity del6q Active transmembrane transporter activity 18 0.000 2 12 Actin 109 down Mitochondrion 130 0.000

Log 10 Protein RNA complex assembly 45 0.003 10 12 14 16 18 20 22 Translation 87 0.003 Regulation of translation 26 0.005 Log2 LFQ intensity nondel Oxidoreductase activity 81 0.005

F G

Enrichment plot: 50 del6q (A6, C4) REGULATION_OF_PROTEIN_METABOLIC_PROCESS 80S

0.0 ) 40 −0.1 Nondel (C2, Iso) ISO

254 nm 30 C2 −0.2 60S 20 40S C4 −0.3 Polysomes A6 NES −2.14 10 −0.4 FDR q-val 0.007 Enrichment score (ES) −0.5 0

Absorbance (OD −10

−20 3456789 0.75 ‘na _pos’ (positively correlated) Time (min) 0.50 0.25 0.00 Zero cross at 1,295 −0.25 −0.50 ‘na_neg’ (negatively correlated) −0.75 0 250 500 750 1,000 1,250 1,500 1,750 2,000 2,250 Rank in ordered dataset Ranked list metric (Preranked) Ranked Enrichment profile Hits Ranking metric scores

Figure 5. 6q14 microdeletion engineered by genome editing induces ribosome and mitochondrial respiration changes in human T-ALL CEM cell lines. A, Schematic of the genome-editing approach using the CRISPR/Cas9 system to engineer precisely the SYNCRIP–SNHG5 microdeletion in the human T-ALL cell line CCRF-CEM (cf. also Methods section). B, aCGH profiles using 6q custom array confirmed the short 38-kb deletion, as seen in patient TL97, in two independent SYNCRIP–SNHG5 clones (A6 and C4) compared with a nondeleted clone (C2) and an isogenic CEM cell line. The initial screen of the deleted and nondeleted clones was performed using specific PCR systems as detailed in the Methods section. C, Immunoblot of hnRNP-Q (SYNCRIP gene product) showed reduced protein level in A6- and C4-deleted clones compared with the nondeleted clone C2 and isogenic cell line; actin was used as a loading control. D, Scatter plot showing protein changes in the deleted versus nondeleted clones; 144 proteins are upregulated (top left quadrant; ratio >1.5) and 109 proteins are downregulated (bottom right quadrant; ratio < 0.66); the red dot shows the SYNCRIP protein (hnRNP-Q) level. LFQ, label-free protein quantification. E, Pathway enrichment analysis using Gene Set Enrichment Analysis (GSEA; broadinstitute.org) tools in the SYNCRIP–SNHG5, CRISPR/Cas9-microdeleted clones (n = 2) versus nondeleted isogenic cells (n = 2). The number of genes in each gene set and P values are indicated. F, GSEA plot of the “Regulation of protein metabolic process” gene set in the microdeleted versus nondeleted cells. G, Profile of ribosome population of deleted and nondeleted clones. (continued on next page)

DECEMBER 2018 CANCER DISCOVERY | 1621

RESEARCH ARTICLE Gachet et al.

H I J 1 50,000 ISO Iso 1.5 Iso 40,000 C2 C2 C2 0.9 A6 A6 30,000 ** ** C4 1 * A6 ** ** C4 0.8 * * ***** C4 20,000 0.5 fluorescence

0.7 10,000 S incorporation Mean click iT-AHA Mean click 35 0 Methylation score Methylation

(a.u. normalized to ISO) (a.u. 0 0.6 28S-Um2824 28S-Cm2848 28S-Gm2863 28S-Cm3680

SNORD50A targets

K Translation up (290; 0) Translation down (185; 0) 2 2 Buffering down (0; 0) Buffering up (0; 0) mRNA abundance up (77; 0) mRNA abundance down (106; 0) 0 0 anslated mRNA log2FC anslated mRNA log2FC

Tr −2 Tr −2

−2 02 −2 02

Total mRNA log2 FC (contrast 1) Total mRNA log2 FC (contrast 1) L Go_gene set #Genes P RNA binding 56 0.000 Poly-A RNA binding 49 0.000 Mitochondrial envelope 30 0.000 Mitochondrion 40 0.000 Organelle inner membrane 27 0.000 Chromosome organization 32 0.000 Enzyme binding 40 0.000 Negative regulation of gene expression 36 0.000 Cell-cycle process 31 0.000 Nucleolus 45 0.000 Negative regulation of nitrogen compound metabolism 35 0.000 Oxydative phosphorylation 24 0.000 Ribonucleoprotein complex 18 0.000

Enrichment plot: GO_MITOCHONDRION Enrichment plot: GO_OXIDATIVE_PHOSPHORYLATION 0.0 0.0 −0.1 −0.1 −0.2 −0.3 −0.2 −0.4 −0.3 −0.5 NES 1.94 0.6 NES 2.19 −0.4 − − − FDR q-val 0.006 −0.7 FDR q-val 0.0 Enrichment score (ES) −0.5 Enrichment score (ES) −0.8 −0.9

‘na _pos’ (positively correlated) ‘na_pos’ (positively correlated) 20,000 20,000 15,000 15,000 10,000 Zero cross at 1,138 10,000 Zero cross at 1,138 5,000 5,000 0 0 ‘na_neg’ (negatively correlated) ‘na_neg’ (negatively correlated) 0 250 500 750 1,000 1,250 1,500 1,750 2,000 0 250 500 750 1,000 1,250 1,500 1,750 2,000

Rank in ordered dataset list metric (Preranked) Ranked Rank in ordered dataset Ranked list metric (Preranked) Ranked

Figure 5. (Continued) H, 2′-O-methylation levels measured by RiboMethSeq assay; methylation scores at the two SNORD50A target sites and at two representative neighbor target sites. N = 3; *, P < 0.05; **, P < 0.005, in accordance with the Mann–Whitney test. I and J, Global protein synthesis, measured by Click-iT AHA upon 3% hypoxia (I) and 35S-Methionine incorporation upon 2% hypoxia culture condition (J). N = 3; **, P < 0.005, in accordance with the Mann–Whitney test. K, Scatter plot showing gene transcripts with significant changes in TE between three pairs of deleted and nondeleted CCRF- CEM clones (left), with polysome and cytoplasmic fraction for each sample (total 12 files), resulting in a list of 475 significantly up- or downtranslated gene transcripts (Supplementary Table S4). The plot on the right of the panel shows the absence of significant TE change using an overall distribution that includes a permutation between a deleted and nondeleted label, in keeping with the high specificity of the Anota2Seq analysis algorithm using the experiment design. L, Main pathway enrichment by GO analysis using the list of differentially translated gene transcripts (top); GSEA plots of the “Mitochon- drion” and “Oxidative phosphorylation” gene sets in the microdeleted versus nondeleted cells (bottom; differentially translated transcript ratios were ranked based on the inverse of the P value). (continued on following page)

1622 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

M Figure 5. (Continued) M, Measurement of OCR. Oligomycin FCCP Antimycin A Left, real-time measurement of the OCR showing the + rotenone 400 500 key parameters of mitochondrial respiration for the A6 Nondel (Iso) **** **** microdeleted clone, compared with nondeleted cells; del6q (A6) 400 right, comparative analysis of the maximal respiration 300 in two individual microdeleted clones (A6 and C4), com- 300 pared with nondeleted cells. ****, P < 0.0001, in accord- 200 ance with the Mann–Whitney test. N, Measurement of 200 extracellular acidification rate (ECAR). Left, real-time 100 OCR (p mole/min) OCR (p mole/min) Basal Maximal 100 measurement of the ECAR showing the key parameters respiration respiration of glycolytic flux for the A6 microdeleted clone com- 0 0 pared with nondeleted cells; right, comparative analysis 020406080 Iso A6 Iso C4 of glycolysis in two individual microdeleted clones (A6 Time (minutes) and C4), compared with nondeleted cells. ***, P < 0.0005, N in accordance with the Mann–Whitney test. Glucose oligomycin 2-DG 200 200 Nondel (Iso) del6q (A6) *** *** 150 150

100 100 Glycolytic capacity 50 50 ECAR (mpH/min) ECAR (mpH/min) Glycolysis

0 0 020406080 Iso A6 Iso C4 Time (minutes)

functions of SYNCRIP and SNHG5 in mRNA processing SYNCRIP–SNHG5-silenced T-ALL cells also exhibited and translation (16–20), along with the consistent changes downregulation of many genes related to ribosome, RNA we observed when these genes were codeleted (Fig. 5E–N), degradation, and oxidative phosphorylation (Fig. 6F and we hypothesized that the del6q might increase malignancy G; Supplementary Table S5), i.e., pathways that are rel- by changing the metabolic state in the T-ALL cell popula- evant not only to the known functions of both genes, tion (Fig. 6A). We therefore aimed to directly test whether i.e., regulating mRNA processing and ribosome biogen- the SYNCRIP–SNHG5 deletion increases the LIC activity of esis (16–20), but also to the translatome and proteomic human T-ALL cells. We took advantage of our engraftment changes of ΔSYNCRIP/ΔSNHG5 clones (Fig. 5E–N). To model of primary T-ALL in immunodeficient NOD/SCID/ directly assess the effects of reduced oxidative phospho- IL2Rγ-null (NSG) mice (15) and used SYNCRIP–SNHG5 rylation in leukemia progression, we exposed PDX-derived shRNA transductions to cosilence both human genes (Fig. TAL1d, non-del6q, human T-ALL cells to tigecycline, an 6B). To provide a faithful model of the deletion as a late- antibiotic that inhibits mitochondrial protein translation stage chromosomal event in the context of TAL1-related (35), and subsequently performed ex vivo limiting dilution leukemogenesis, primary diagnosis T-ALL blast cells har- and immunophenotype experiments (Fig. 7A). We found boring a prototypical TAL1d rearrangement, but no del6q, that tigecycline conferred increased clonogenicity to T-ALL were lentivirally silenced, and the sorted cells were injected cells (Fig. 7B); in addition, tigecycline exposure partially into NSG mice as shown in Fig. 6B. Using this setting that reproduced PDX cell dedifferentiation that was seen after includes human SYNCRIP and SNHG5 shRNAs, we found SYNCRIP–SNHG5 cosilencing using shRNA (Fig. 7C). that silenced cells propagated leukemia significantly more Finally, to confirm the relevance of theSYNCRIP–SNHG5 - efficiently than control cells in both parallel and competitive focused codeletion with the usually larger 6q deletions seen in engraftment experiments, suggesting increased malignancy patients, we analyzed the expression of the “SYNCRIP/SNHG5- (Fig. 6C; Supplementary Fig. S6), consistent with the shorter silenced xenograft T-ALL” signature defined in silenced leukemia latency seen in our mouse model (Fig. 4B). More xenograft (Fig. 6F) in the cohort of TAL1-related patients. over, the accelerated engraftment of the SYNCRIP–SHNG5- Strikingly, the downregulated signature was also enriched in silenced leukemic cells persisted in dilution experiments, primary T-ALL cases with large del6q versus nondeleted cases indicating greater LIC activity (Fig. 6D). (Fig. 7D). We then characterized the human leukemic cells that expanded in the xenografts. We measured the 2′-O-methylation level of the 28S rRNA at the Cm2848 site, known to be a target DISCUSSION of SNORD50A (34), and found that it was lower in silenced The oncogenic meaning of del6q in T-ALL has been than in control T-ALL cells that emerged in NSG mice a long-standing unresolved question since the original (Fig. 6E). Comparing gene-expression data in these cells description of this structural chromosome rearrangement (Supplementary Table S5), relative SYNCRIP and SNHG5 in leukemia (11–14). In this study, we addressed del6q by levels were consistent with haploinsufficiency, suggesting investigating large cohorts of highly annotated primary in vivo selection of the optimal level of gene silencing and T-ALL samples, which allowed us to focus on the homogene- reinforcing the interest of these models. Moreover, the ous TAL1-related subtype, and by using original in vitro and

DECEMBER 2018 CANCER DISCOVERY | 1623

RESEARCH ARTICLE Gachet et al.

A B

6q/SYNCRIP-SNHG5 Primary shRNAs haplo insufficiency T-ALL cells SYNCRIP–SNHG5 or Ctrl

Decreased 2’-O-methylation and ribosome deregulation GFP+ sorting NSG Decreased mitochondrial respiration shRNAs: Ctrl SYNCRIP–SNHG5 Increased LIC activity 6 6 at late stage of tumor progression 5 5 Leukemia onset 4 4

3 3 shRNA Ctrl Tandem shRNA 2 2 1 1 hnRNP-Q 0 0 Expression value Actin SNHG5 SYNCRIP

C D 100 Sh SYNCRIP–SNHG5 ShSYNCRIP–SNHG5 Sh Ctrl P = 0.036 ShCtrl 80 5 × 104 cells 5 × 103 cells 5 × 102 cells 100 60 cells in blood + 10

40 cells in blood +

20 GFP 1 + % CD45-GFP 0 7 weeks 9 weeks 10 weeks 0.1 57911 13 15 57911 13 15 17 19 579111315 17 192123

% of CD45 Weeks after transplantation

E CH Site 2848 F 3 200 ** KEGG entry number Pathway description (KEGG) P (down) rRNA 150 hsa03010 Ribosome 1.73E–09 100 RT hsa00190 Oxidative phosphorylation 5.87E–06 Low [dNTP] High [dNTP] 50 hsa03018 RNA degradation 3.81E–04 0 hsa03050 Proteasome 1.07E–03 Methylation ratio Methylation ShRNA: Ctrl SYNCRIP–SNHG5 hsa04115 p53 signaling pathway 1.98E–03 60 Site 389 hsa04110 Cell cycle 1.98E–03 qPCR 50 ns hsa04120 Ubiquitin mediated proteolysis 2.34E–03 40 Methylation ratio 30 2(Ct low – Ct High) 20 10

Methylation ratio Methylation 0 ShRNA: Ctrl SYNCRIP–SNHG5

G Enrichment plot: MITOCHONDRION 0.00 Figure 6. Knockdown of both SYNCRIP and SNHG5 confers a selective advantage to human −0.05 −0.10 leukemic cells. A, Working model for del6q oncogenic function. B, Schematic of the protocol for −0.15 engraftment experiments of shRNA-mediated silenced primary T-ALL human cells. RT-qPCR analysis −0.20 −0.25 of SYNCRIP and SNHG5 gene expression and immunoblot of SYNCRIP products (SNHG5 is a noncod- −0.30 ing gene) show silencing efficiency. C, Engraftment kinetics of cells transduced with shRNA Ctrl and 0.35 NES −2.19 − + + 0.40 shRNA SYNCRIP–SNHG5 lentivectors, injected in parallel experiments, and measured by CD45 GFP Enrichment score (ES) − FDR q-val 0.002 −0.45 cell percentages in white blood cells. Each dot represents data from one mouse; horizontal bars represent medians. The indicated P value was obtained by the Mann–Whitney test. D, Experiments ed) similar to those in C were conducted with decreasing cell doses, as indicated. Each dot represents 0.50 ‘i’ (positively correlated) data from one mouse; lines connect the median values. E, Ribosomal RNA 2′-O-methylation analysis. 0.25

ic (Prerank 0.00 Left, schematic of the method used to measure methylation levels of specific RNA sites (see Meth- Zero cross at 1,138 −0.25 ods section). Right, methylation ratio of site 2848 of 28S rRNA, which is targeted by U50A snoRNA −0.50 (34), in human xenografted leukemic cells—SYNCRIP–SNHG5 silenced (n = 4) versus Ctrl (n = 4). As −0.75 ‘0’ (negatively correlated) 0 2,500 5,000 7,500 10,00012,50015,00017,500 20,000 a negative control, methylation levels were tested for site 389 of 28S rRNA, which is targeted by Rank in ordered dataset Ranked list metr Ranked another snoRNA (U26). **, P = 0.0041, in accordance with the Mann–Whitney test; ns, nonsignificant. F, Biological pathways (KEGG database) in human xenografted T-ALLs resulting from SYNCRIP– SNHG5 silenced (n = 4) versus Ctrl cells (n = 4). G, GSEA of the “Mitochondrion” gene set in human xenografted T-ALLs resulting from SYNCRIP–SNHG5 silenced (n = 4) versus Ctrl cells (n = 4).

1624 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

AB Tigecycline 3,000 DMSO (or DMSO) Limiting dilution experiment upon * Tigecycline hypoxia (2–4 weeks culture) 2,000 CD34 and CD8 immunophenotyping * Xenografted 1,000 human TAL1d 48 hours in vitro

T-ALL cells culture -ALL cells per well 4 T CD34 and CD8 immunophenotyping + 2 CD7 + shRNAs 0 SYNCRIP–SNHG5 (or Ctrl) 15,625 3,125 625 125 25 51 hCD45 hCD45+ CD7+ T-ALL cells seeded per well

C D Patient (TAL-RA) T-ALL cohorts Tigecycline exposure SYNCRIP–SNHG5 cosilencing SYNCRIP–SNHG5-silenced probeset + + + + CD34 CD8 CD34 CD8 0.0 ** ** −0.1 1.6 1.6 1.6 1.6 −0.2

−0.3 1.4 1.4 1.4 1.4 NES 1.46 chment score (ES) −0.4 −

1.2 1.2 1.2 1.2 En ri −0.5 FDR q-val 0.13

1.0 1.0 1.0 1.0

‘na_pos’ (positively correlated) 0.8 0.8 0.8 0.8 1.0 0.5

MFI tigecycline/DMSO Zero cross at 20,233 0.6 0.6 0.6 0.6 0.0 −0.5 MFI sh SYNCRIP–SNHG5 /shCtrl ‘na_der’ (negatively correlated) −1.0 0 10,000 20,000 30,000 40,000 50,000

Ranked list metric (signal2Noise) Ranked Rank in ordered dataset

Figure 7. Tigecycline exposure confers to primary T-ALL cell properties associated with gain of malignancy. A, Xenografted T-ALL diagnosis cells from a TAL1d, non-del6q patient were either cocultured in the presence of 2.5 μmol/L tigecycline (or DMSO) on MS5-DL1 feeder or transduced with SYNCRIP–SNHG5 tandem-GFP shRNA (or Ctrl-GFP) lentiviral vector and cultured on plates coated with DL4. Forty-eight hours later, hCD45+CD7+ PDX T-ALL cells were either FACS sorted and seeded on MS5-DL1 cells under limiting dilution conditions (from 15,625 to 1 cell per well, 24 wells per condition) and cultured upon hypoxia conditions or directly analyzed for CD34 and CD8 cell-surface expression. B, In the limiting dilution experiment, + + − FACS analysis of recovered human CD45 CD7 GFP cells was performed after 2 to 4 weeks of culture upon hypoxia condition (1.5% O2). *, P < 0.05, in accordance with the unpaired t test. C, Immunophenotype analysis of CD45+CD7+ PDX T-ALL cells shows a shift toward greater CD34 and lower CD8 expression upon 2.5 μmol/L tigecycline exposure (left) or SYNCRIP and SNHG5 cosilencing (right), consistent with an enrichment in LIC activity. Mean fluorescence intensity (MFI) of cell-surface expression for CD34 and CD8 (CD4 was negative in this T-ALL) was measured 48 hours after tigecycline exposure (left) or SYNCRIP–SNHG5 cosilencing (right) by FACS analysis. Ratios of cells expressing strongly CD34 and CD8 surface markers are shown for n = 4 experiments on cells originating from 4 distinct xenografted mice of the same patient. *, P < 0.05, in accordance with the paired t test. D, GSEA of the SYNCRIP–SNHG5-silenced signature in del6q versus nondeleted primary cases of the TAL-RA subtype.

in vivo models. Integrated global genomic and expression of T-ALL that can activate TAL1, and its silencing has been analyses identified two contiguous downregulated genes at shown to be deleterious for T-ALL cells (10, 38, 39). We 6q14, SNHG5 and SYNCRIP. Both genes were found induced then demonstrated that the codeletion was not a passenger upon TAL1 overexpression in thymocytes, which provided lesion but truly a driver of tumor progression using a mul- an explanation as to why del6q is found specifically in the tievent T-ALL in vivo model—i.e., acceleration of Tal1tgLmo1tg TAL-R subtype in the patients, i.e., to counteract abnor- Notch1IC mouse leukemia—in line with the human TAL-R mal (considering T-ALL differentiation stage) SNHG5- and subtype in which del6q occurs as a late-stage chromosomal SYNCRIP-sustained expression after β-selection (36, 37). event. Importantly, the acceleration was seen only when Regarding del6q, short or large deletions, but no single both genes were codeleted, demonstrating the tumor sup- point mutation of these two genes, were found, in line with pressor nature of the global SYNCRIP–SNHG5 region, but the idea that the simultaneous inactivation by a single chro- not of either gene alone. mosomal deletion is much more probable than two or more To mechanistically investigate the role of the combined concomitant mutations leading to combined haploinsuf- loss of the two genes, we made use of the CRISPR/Cas9 ficiency ofSNHG5 and SYNCRIP. Moreover, our mapping technology to engineer a precise 38 kb deletion in a T-ALL and expression results strongly suggest that the deletion cell line as seen in patient cells. Proteomic analyses of size must be sufficiently large to include the two-gene tumor del6q-engineered clones showed a decreased ribosomal sig- suppressor region, but interstitial rather than complete to nature, consistent with the known functions of these two prevent the loss of MYB. Indeed, MYB is a known oncogene genes, which both regulate mRNA processing and ribosome

DECEMBER 2018 CANCER DISCOVERY | 1625

RESEARCH ARTICLE Gachet et al. biogenesis. Such a downregulation of ribosomal functions METHODS may seem counterintuitive for the fitness of cancer cells. Patients and Collection of Biological Material However, on the basis of recent reports that ribosomal functions can regulate normal and leukemic stem cells Three annotated cohorts of patients with T-ALL were analyzed in (27–30), we hypothesized that these changes might induce this study. The CIT and ALEK cohort included children and adults, whereas the Rotterdam cohort included only children (4, 5, 15). metabolic changes and confer an oncogenic gain at late Informed written consent was obtained from the patients or relatives stage of T-ALL progression. Interestingly, our data showing in all cases, and the study was conducted in accordance with the differential translation of mitochondria gene transcripts, Declaration of Helsinki. The study was approved by the review board along with reduced mitochondrial respiration, suggested of the Institut Universitaire d’Hématologie, Hôpital Saint-Louis, direct oncogenic deregulation of the T-ALL LIC (31–33, 40). Paris, France. Because the nature of the LICs is not clearly identified in T-ALL, precluding consistent purification and characteriza- Genome-Wide DNA Array Copy-Number Analysis tion (41, 42), we performed a functional in vivo approach, A total of 107 T-ALL cases from the TAL1-related subtype were i.e., xenograft of leukemia cells in immunodeficient mice. tested using 105K, 180K, 244K, and/or custom 8 × 15K Human This enabled us to show that SYNCRIP–SNHG5 cosilenc- Genome CGH Microarrays arrays (Agilent Technologies), allowing ing did indeed confer greater LIC ability to primary TAL-R precise mapping of a CDR at 6q14. Hybridization was performed leukemia cells. In addition, ex vivo exposure to tigecycline, in accordance with the manufacturer’s recommendations, and an antibiotic that inhibits mitochondrial protein transla- copy numbers were analyzed using VAMP tools (Curie Institute) or Genomic Workbench software with the help of the ADM-2 algorithm tion, conferred to non-del6q primary human T-ALL cells (Agilent Technologies). All data were visually inspected and cured greater properties associated with LIC activity and gain of manually by at least two investigators (D. Avran and J. Soulier). malignancy, suggesting a direct driving role of the reduced mitochondrial respiration in T-cell oncogenesis. This differs Design of the 6q14 Custom Array from the inhibitory effect of tigecycline on myeloid leukemia The custom oligonucleotide-based microarray contained 15,689 (acute myeloid leukemia and chronic myeloid leukemia) 60-mer probes (including 1,391 manufacturer control probes) span- LICs, which relies on enhanced basal oxygen consumption ning both coding and noncoding genomic sequences of chromosome (43, 44). By contrast, T-ALL LICs would fit the classic figure 6 (Supplementary Fig. S4A). The multistep custom design included of malignant cells relying more on glycolysis than on oxi- 14,296 probes distributed among several groups with various probe dative phosphorylation. Differential mitochondrial activity densities. Group I comprised 2,574 probes spread along chromosome may also directly modulate T-ALL LIC self-renewal and 6p (∼22-kb average probe spacing); group II comprised 3,370 probes cell fate, as was reported in normal HSCs and thymocytes spread along chromosome 6q (∼32-kb average probe spacing). Several (33, 45). Notably, although our studies clearly establish intervals in chromosome 6q were made denser. the existence of a SYNCRIP/SNHG5 → ribosome → mitochondria → LIC axis in T-ALL, we cannot exclude that Next-Generation Sequencing at 6q14 SYNCRIP/SNHG5-dependent ribosome modifications also We designed a custom SureSelect Agilent capture in solution to impinge on the translation of other key regulators of cel- enrich genomic sequences from the CDR and flanking regions at lular self-renewal or proliferation. Importantly, we retrieved 6q14 (Supplementary Fig. S4B). The Agilent eArray (www.earray. reduced ribosome and oxidative phosphorylation signatures, chem.agilent.com/earray/) was used to design and assess coverage not only in the engineered deleted cells, but also in the SYN- across the target genomic regions of bait libraries. Different bait CRIP–SNHG5-silenced engrafted T-ALL cells, as well as in groups were established in accordance with various parameters (tiling, number of replicates, and orphan probes). Regions with del6q cases from our TAL-R patient cohorts. Collectively, the repeats were eliminated with the “repeat masker” option. The data also confirmed the general relevance of the shortSYN - final customized target capture contained 57,670 baits of 120 bp, CRIP–SNHG5 deletion to the larger 6q deletions seen in most corresponding to 5,676,317 bp. DNAs were sonicated to obtain patients. 150- to 200-bp fragments using a Covaris S220 Focused ultra- In conclusion, our study clarifies the role played by 6q dele- sonicator instrument (Covaris Inc.). 6q14 custom capture was tion in T-ALL oncogenesis and identifies SYNCRIP/hnRNP-Q performed using the SureSelect Target Enrichment System for and snoRNAs as tumor suppressors in this deletion. Together Illumina Paired-End Sequencing Library protocol (version 1.2, May with previous evidence of mutations in ribosomal genes in 2011) in accordance with the manufacturer’s recommendations. hematopoietic malignancies and other cancers (8, 46, 47), Captured samples were sequenced using a HiSeq 2000 (Illumina) our findings provide new insights for a critical driver role operated in the paired-end 2 × 100 bp mode. Somatic variants were identified by selecting the variants found in tumor reads that were of qualitative ribosome and mitochondria deregulation, in absent in the nontumor reads. All somatic variants were visually the context of multistage oncogenesis, consistent with the inspected and manually cured by at least two investigators (S. Quentin increasingly recognized regulatory functions of subtle meta- and J. Soulier). bolic changes on normal and leukemic stem cells (28–30, 40, 47, 48), and opening potential avenues for therapeutic target- Large-Scale Gene-Expression Analysis ing. For instance, therapies that change ribosomal activity or Microarray data files from the three T-ALL cohorts are avail- oxidative respiration may antagonize LICs in del6q T-ALL. able from ArrayExpress under accession number E-MEXP-313 (CIT In addition, ribosomal stress induced by del6q suggests that cohort), E-MTAB-604 (ALEK cohort), and from Gene Expression therapeutic intervention, such as synthetic lethality, may be Omnibus (GEO) accession number GSE26713 (Rotterdam cohort). used to target leukemic cells, thereby improving clinical out- Affymetrix GeneChip probe set summarization, background cor- comes in patients. rection, and quantile normalization were performed using the RMA

1626 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

routine, as implemented in the “R” Affymetrix package. Differential DNA and cloned respectively in the upstream and downstream gene expression arising from chromosomal gene localization was polylinkers of the HR110-PA1 homologous recombination vector analyzed using MACAT (microarray chromosomal analysis tool) (SBI). The linearized HR construct and the Cas9 vector set (2:1 and the associated R Bioconductor packages (www.bioconductor. ratio) were cotransfected by nucleofection in CCRF-CEM T-ALL org) using the following parameters: permute = labels, nperms = 750, cells with an Amaxa Nucleofector in accordance with the manufac- kernel = KNN, and step.with = 100,000. Biological gene-pathway turer’s recommendations. Bulk transfected cells were puromycin changes were scored using the Gene Set Enrichment Analysis algo- selected and cloned by RFP+ flow-cytometric cell sorting. Single- rithm (GSEA; broadinstitute.org). Human T-ALL samples from xen- cell–derived clones were expanded in RPMI medium (Life Technolo- ografts were hybridized using the GeneChip Human Gene 1.0 ST gies) supplemented with 10% heat-inactivated fetal bovine serum (Affymetrix); gene-expression profiling and splice analysis were (Life Technologies) under constant puromycin (Life Technologies) performed using the Easana and Fast DB tools (Genosplice Tech- selection at 2 μg/mL. The expanded individual RFP+ clones were nologies; www.genosplice.fr). screened for the microdeletion by PCR genotyping with two sets of Healthy T-cell subsets were prepared and analyzed as described in primers (system 1: CGGGGGAGGGACGTAATTAC, GCCTTTAC refs. 49 and 50; microarray data have been deposited in the GEO with TAAAATGGCGAAG; system 2: GCTCTCATTGAAACTGTAAGCA, accession number GSE62156. GTGGCGGCCGCTGTCTAGAT). The overall targeting efficiency was 8.3%, with 16 of 192 single-cell–derived clones harboring the Analysis of Syncrip and Snhg5 Expression Levels simultaneous recombination events at SYNCRIP and SNHG5 loci in Murine Thymocyte Subsets by Real-Time (data not shown). Individual SYNCRIP–SNHG5 microdeleted clones Quantitative PCR (RT-qPCR) were further validated by our custom 6q14 array (Fig. 5B; Sup- plementary Fig. S4). The level of hnRNP-Q protein was analyzed Murine thymocytes were extracted from thymi from 3- to by immunoblot using human SYNCRIP Ab (clone I8E4, sc-56703, tg tg 4-week-old Tal1 Lmo1 and age-matched wild-type littermates. Santa Cruz Biotechnology). Briefly, CD3-, CD4-, CD8-, and Ter119-positive cells were isolated by magnetic activated cell sorting. DN3 and DN4 thymocytes were purified by FACS-mediated cell sorting from the enriched positive Cell Lines cells using murine antibodies against lineage markers CD3, CD4, The CCRF-CEM T-ALL cell line had been purchased from DSMZ CD8, CD25, and CD44. Total RNA from DN3 and DN4 subsets in 2008, aliquoted, and authenticated by SIL-TAL1 testing and was extracted using the Direct-Zol RNA Mini Prep Plus kit (Zymo/ arrayCGH (not shown); for the current work, a fresh vial was Ozyme). The cDNA was synthesized using the SuperScript VILO defrozen, and the cell line was reauthenticated by arrayCGH. The Master Mix kit (Thermo Fisher). Syncrip and Snhg5 expression isogenic CCRF-CEM cell line and SYNCRIP–SNHG5 microdeleted levels were measured by RT-qPCR performed on the LightCycler clones were routinely tested for Mycoplasma using a luminescence- 480 (Roche) using the TaqMan Multiplex Master Mix (Thermo based array (MycoAlert Mycoplasma Detection Kit, Lonza), most Fisher). Each sample was analyzed in duplicate, and gene expres- recently in May 2018. Cells were passaged no more than 20 times sion was normalized relative to the expression of two housekeeping before thawing new low-passage batches that were also Mycoplasma genes, Hprt and Gapdh. TaqMan probes are listed in Supplementary tested before use. Methods. Label-Free Proteomic Quantitative Analysis Gene Silencing in Murine and Human Cells Two deleted and two nondeleted clones were analyzed, each of Short hairpin RNA sequences are listed in Supplementary Meth- them in duplicate (total 8 samples). Cells were lysed in 50 mmol/L Tris ods. They were cloned under the H1 promoter in the pTRIP/ SDS 2% pH8.5 and boiled 10 minutes at 95°C, and peptides were pre- ΔU3-MND-GFP lentivector. The vector that was used to construct pared using the filter-aided separation method. Proteins were digested tandem shRNAs was a kind gift from Bruno Verhasselt (Ghent Uni- for 14 hours at 37°C with 1 μg trypsin (Promega), and were fraction- versity Hospital, Belgium). pTRIP/ΔU3-MND-Ctrl-GFP and pTRIP/ ated by strong cationic exchange (SCX) StageTips. Mass spectrom- ΔU3-MND-Ctrl-Cherry for shRNA Luc were previously described etry analyses were performed on a U3000 RSLC nano-LC-system (15, 51). Silencing efficiency was checked by immunoblot (human coupled to an LTQ Orbitrap-Velos mass spectrometer (Thermo or murine SYNCRIP; Abcam clone I8E4, ab10687) and RT-qPCR Fisher Scientific). The data were analyzed using MaxQuant version (SYNCRIP, SNHG5, and U50A). RT-qPCR was performed on RNA 1.5.2.8 (53). The database used was a concatenation of human samples using TaqMan probes (Applied Biosystems); primers and sequences from the Uniprot–Swissprot database (Uniprot, release probes are listed in Supplementary Methods. Gene expression was 2015-02) and a list of contaminant sequences from MaxQuant. normalized relative to the expression of the reference genes GUS, The false discovery rate (FDR) was kept below 1% on both pep- Ywhaz, and/or Gapdh. tides and proteins. Label-free protein quantification (LFQ) was carried out using both unique and razor peptides. At least two Short del6q Engineering by CRISPR/Cas9 such peptides were required for LFQ. Data were imported into Perseus software (version 1.5.1.6). Protein copy numbers per cell Potential Cas9 target sites were searched within 2 kb of the two were calculated by standardization on total histone MS signal as germline regions encompassing the chromosomal breakpoint so described (54). as to minimize potential off-target effects (52). The corresponding oligonucleotides were cloned in a derivative of the GeckoV2 with the puromycin resistance gene inactivated (52). Two sets of Cas9 Polysome Fractionation target sequences bordering the inner edge of the intended deletion Polysome fractionation was performed as described previously were selected for genome engineering (set 1: GGCACACTACAAA (55). Briefly, a total of 30× 106 deleted and nondeleted CEM-CCRF GCACACC, ATCAGACCTTGTATATGTACC; set 2: GATTGTATTAC cells were incubated with 50 μg/mL Emetin (Sigma) and cytoplasmic TTTTTAACCT, GTATTATGTGTACATATACA). A 1.8-kb DNA frag- fractions were isolated by mechanical lysis of cells. One milligram of ment spanning the positions 86,349,297–86,347,424 of chromo- cytosolic proteins was separated on a 15% to 47% sucrose gradient some 6 (GRch37/hg19) and a 1.1-kb DNA fragment spanning by ultracentrifugation, and absorbance profiles were generated at the positions 86,387,432–86,388,539 were amplified from genomic 254 nm.

DECEMBER 2018 CANCER DISCOVERY | 1627

RESEARCH ARTICLE Gachet et al.

Translatome Analysis PBS supplemented with 3% BSA. The azide–alkyne cycloaddition mRNAs associated with polysomal ribosomes were analyzed and was performed using the Click-iT Cell Reaction Buffer Kit (Thermo compared with total cytoplasmic mRNAs in order to identify Scientific), with the alkyne conjugated to Alexa Fluor 488 (Thermo changes in TE between deleted and nondeleted CEM-CCRF cells. Scientific) at 1 μmol/L final concentration. Cells were incubated Briefly, polysomes were fractionated from 3 mg of cytosolic extract for 30 minutes for the reaction and washed twice with PBS before from each 6q deleted or control sample as described (55). Polysomal proceeding to flow cytometry. 35 fractions were pooled, and mRNAs were extracted from the polyso- As independent experiments, S-labeled methionine incorpora- mal and total cytoplasmic fractions using TriPure reagents (Roche) tion was measured in the deleted and nondeleted cells following and quality controlled using a BioAnalyzer (Agilent). Libraries 18-hour culture under hypoxia (2% O2). Cells were labeled 30 minutes 35 35 were prepared from 0.5 ng of polyA+ fraction per sample using with 75μCi of a [ S]-methionine and [ S]-cysteine mix as described the Scriptseq v2 (Illumina Epicentre), and RNA sequencing was in ref. 55. run on a HiSeq1000. Reads were mapped to the human reference genome GRCh38 using STAR v2.5 with the setting –quantMode Mitochondrial Respiration and Glycolytic gene counts to quantify the number of reads per gene. Differential Function Measurements expression analysis and identification of differences in TE were per- OCR was measured with the Seahorse XFp Cell Mito Stress Test formed using the Anota2Seq program (25, 26). Samples from three on the Seahorse XFp Analyzer (Agilent, Seahorse Bioscience) in independent cell extraction and polysome purification experiments accordance with the manufacturer’s recommendations. Briefly, cells were analyzed after cpm (counts per million) transformation, batch were seeded at 5 × 104 cells per well (optimal cell seeding density correction, quantile normalization, log2 (x + 10) transformation was determined in preliminary experiments) in an XFp-well mini- and selection of the expressed genes monitored by qqnorm fitting plate previously coated with Cell-Tak (Corning, Fisher Scientific) in resulting in a final paired comparison between three deleted and 180 μL of XF Base Medium, supplemented with fresh sodium pyru- three nondeleted samples (with polysome and cytoplasmic fraction vate (1 mmol/L), glutamine (2 mmol/L), and glucose (10 mmol/L; for each). Default Annota2seq parameters were used (anota2se- pH 7.4). Sequential compound injections of oligomycin A (1 μmol/L), qRun). Validation of the specificity of the TE changes was checked FCCP (1 μmol/L), rotenone (0.5 μmol/L), and antimycin A (0.5 using permutations between deleted and nondeleted labels (nine μmol/L) enabled measurements of basal respiration, ATP produc- aberrant combinations, the cytoplasmic or polysome origin being tion, proton leak, maximal respiration, spare respiratory capacity, kept for each sample). R version 3.4.2 (September 28, 2017) was and nonmitochondrial respiration. Extracellular acidification rate used. The resulting list of 475 genes was analyzed using the GSEA (ECAR) was measured with the Seahorse XFp Glycolysis Stress Test algorithm (broadinstitute.org). in accordance with the manufacturer’s recommendations. Cells (5 × 104 per well) were seeded in 180 μL of the glycolysis stress test RNA 2′-O-Methylation Profiling by High-Throughput medium [XF Base Medium supplemented with fresh glutamine Sequencing (RiboMethSeq) (2 mmol/L) only; pH adjusted to 7.4] on a previously coated XFp-well miniplate. Sequential compound injections of glucose A Site-specific rRNA methylation was determined by RiboMethSeq (10 mmol/L), oligomycin (1 μmol/L), and 2-deoxy-glucose (2-DG, as previously described (56). Briefly, 100 ng of total cellular RNA was 50 mmol/L) enabled measurements of glycolysis, glycolytic capac- subjected to alkaline hydrolysis in 50 mmol/L bicarbonate buffer ity, glycolytic reserve, and nonglycolytic acidification. Oxygen con- pH 9.2 for 10 to 12 minutes at 95°C. RNA fragments were purified sumption and glycolytic flux were recorded for 90 minutes after and converted to cDNA libraries using the NEBNext Small RNA cells were placed into the analyzer. OCR (pmol/minute) and ECAR Library kit (New England Biolabs). Libraries were subjected to high- (mpH/minute) were measured simultaneously in all wells, three throughput sequencing using an Illumina HiSeq 1000 instrument times at each step, by optical fluorescent O sensors, and a mini- with a 50-bp single-end read mode. Adapter sequence trimming was 2 mum of three replicates were analyzed per condition in any given conducted using Trimmomatic-0.32. Alignment to the reference experiment; all compounds and material were obtained from Agi- rRNA sequence was achieved by Bowtie2 (ver 2.2.4) in End-to-End lent Seahorse Bioscience. mode. 5′-end counting was carried out directly on *.sam file using dedicated Unix script. Final analysis was performed by calculation of T-ALL Mouse Model MethScore for quantification of ′2 -O-methylated residues. Animals were housed and handled in the IUH Département d’Expérimentation Animale in accordance with the guidelines of the Quantification of Global Protein Neosynthesis Animal Care and Use Committee (IUH, Hôpital Saint-Louis). The Total protein synthesis rates were measured with Click-iT AHA pSil-TSCL (Tal1tg) and Lck-LMO1 (Lmo1tg) transgenic mouse lines incorporation assay (Thermo Scientific) in accordance with the manu- were kind gifts from P. Aplan (NCI, Bethesda, MD; ref. 21). They facturer’s recommendations. Briefly, cells were seeded at 2.5× 105 cells/ were backcrossed onto the C57BL6/J background (The Jackson Lab- mL per well in 24-well plates in RPMI medium (Life Technologies) oratory) for more than 12 generations. Bone marrow was harvested supplemented with 10% heat-inactivated fetal bovine serum (Life from the femur of 3- to 4-week-old double-transgenic C57BL/6 Technologies) ± puromycin (Life Technologies) selection at 2 μg/mL mice, and Lin− cells were infected with lentiviral vectors and cultured and grown for 24 hours under basal (normoxia) or stress (3% O2 overnight. The cells were then infected with a retroviral vector for hypoxia) conditions. Before proceeding to Click-iT AHA incorpora- a Notch1-mutant (Notch-ICN, intracellular cleaved Notch1) that tion, cells were washed once with warm Ca2+and Mg2+-free phos- was coupled to a truncated nerve growth factor receptor (tNGFR), phate-buffered saline (PBS) and incubated for 60 minutes at 37°C kindly provided by J. Ghysdael (Institut Curie-Orsay, France) with in methionine-free medium to deplete methionine reserves. Click-iT the permission of W. Pear (ref. 57; University of Pennsylvania, AHA (L-azidohomoalanine; Thermo Scientific) was added to the cul- Philadelphia, PA). Next, 2 × 105 cells were i.v. injected into the tail of ture methionine-free medium to a final concentration of 100μ mol/L sublethally irradiated (3 Gy) RAG−/−γc−/− mice. Recipient mice were for 2.5 hours, and then cells were washed twice with PBS. Cells were monitored by FACS analysis of blood cells and clinical observation. fixed in 0.5 mL of 4% paraformaldehyde in PBS for 15 minutes on ice, Pathology analysis confirmed T-ALL diagnosis, with marrow and permeabilized in 0.5 mL 0.1% saponin (Sigma) in PBS supplemented spleen infiltration by lymphoblasts expressing the Thy-1, CD4, and/ with 1% BSA (Sigma) for 10 minutes on ice and washed once with or CD8 T-cell markers.

1628 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

Xenograft of Transduced Primary Leukemic of cDNAs was assessed by PCR using LightCycler FastStart DNA Cells from Patients Master SYBR Green I and a LightCycler 480 (Roche Applied Sci- ence). The primers were 5′-AGGTAAGGGAAGTCGGCAAG-3′ and This experimental procedure has been described in detail previ- 5′-CAGCCCTTAGAGCCAATCCT-3′ and 5′-CCGTAAGGGAAAGTT ously (15). Briefly, 3× 106 primary human leukemic cells were GAAAAG-3′ and 5′-CCCCACCCGTTTACCTCTTA-3′ for the 28S- transduced using a lentivirus containing a shRNA Ctrl-GFP or SYN- 2858 and 28S-389 sites, respectively. The methylation level was calcu- CRIP–SNHG5-GFP vector, cultured for 5 days, and sorted by FACS, lated as Log (Ctlo − Cthi), where the Ctlo and Cthi values corresponded and 5 × 103 cells were i.v. injected to sublethally irradiated (2.25 Gy) 2 to Ct obtained after RT performed at low and high dNTP concentra- NSG mice. For competitive experiments, cells were transduced with tions, respectively. a shRNA Ctrl-Cherry or SYNCRIP–SNHG5-GFP lentivirus, FACS- sorted, and mixed in equal number. The starting GFP+/Cherry+ ratio was rechecked by FACS, and 1 × 104 of these cells were i.v. injected Statistical Analyses into sublethally irradiated (2.25 Gy) NSG mice. Leukemia develop- Statistical analyses were performed using the Fisher, Mann–Whitney, ment was monitored by FACS analysis of blood cells. log-rank, unpaired t test and binomial tests, as indicated in the figure legends. In Vitro Limiting Dilution Experiment of Xenografted T-ALL Cells Exposed to Tigecycline Disclosure of Potential Conflicts of Interest PDX T-ALL cells from a TAL1d non-del6q primary case were H. de The is a consultant/advisory board member for Vector- exposed to 2.5 μmol/L tigecycline (Sigma) or an equivalent amount Lab. No potential conflicts of interest were disclosed by the other of DMSO and cocultured on MS5-DL1 feeder cells as previously authors. described (15) for 48 hours. CD45+CD7+GFP− human leukemic cells were then FACS sorted and seeded on irradiated (30 Gy) Authors’ Contributions MS5-DL1 cells under limiting dilution conditions from 15,625 to 1 cell per well (24 wells per condition). FACS analysis of recovered Conception and design: S. Gachet, T. El-Chaar, D. Avran, E. Genesca, human CD45+CD7+GFP− tigecycline or DMSO treated cells was S. Quentin, I. André-Schmutz, J.P. Meijerink, E. Clappier, C. Gazin, J. Soulier performed after 2 to 4 weeks of culture upon hypoxia (1.5% O2). Human PDX T-ALL cells were stained with APC anti-human CD45 Development of methodology: S. Gachet, T. El-Chaar, D. Avran, (clone 5B1; Miltenyi Biotec) and PE-Vio770 anti-human CD7 E. Genesca, S. Quentin, D. Briot, L. Hernandez, J.P. Meijerink, (clone 6B7; Miltenyi Biotec) antibodies. Flow cytometry analy- J.-J. Diaz, C. Gazin ses were performed on a FACS Canto II or a Fortessa Analyzer Acquisition of data (provided animals, acquired and managed (Beckton Dickinson). patients, provided facilities, etc.): S. Gachet, D. Avran, E. Genesca, F. Catez, S. Quentin, G. Thérizols, G. Meunier, L. Hernandez, M. Pla, W.K. Smits, J.G. Buijs-Gladdines, T. Taghon, P. Van Vlierberghe, Immunophenotype Analysis of Xenografted T-ALL J.P. Meijerink, A. Baruchel, H. Dombret, E. Clappier, C. Gazin Cells Exposed to Tigecycline or SYNCRIP–SNHG5 Analysis and interpretation of data (e.g., statistical analysis, shRNA-Mediated Silencing biostatistics, computational analysis): S. Gachet, T. El-Chaar, PDX T-ALL cells from a TAL1d non-del6q patient were exposed to D. Avran, E. Genesca, F. Catez, S. Quentin, M. Delord, W. Van Loocke, 2.5 μmol/L tigecycline (Sigma) as above or transduced with a lentivi- G. Menschaert, T. Taghon, P. Van Vlierberghe, E. Clappier, J.-J. Diaz, rus containing a shRNA Ctrl-GFP or SYNCRIP–SNHG5-GFP vector C. Gazin, H. de Thé, F. Sigaux, J. Soulier as indicated above for diagnosis cells, and cultured for 48 hours on Writing, review, and/or revision of the manuscript: S. Gachet, MS5-DL1 or on coated DL4, respectively. Mean fluorescence intensity T. El-Chaar, D. Avran, E. Genesca, S. Quentin, J.P. Meijerink, of cell-surface expression for CD34, CD8, and CD4 was measured by A. Baruchel, H. Dombret, J.-J. Diaz, H. de Thé, F. Sigaux, J. Soulier flow cytometry analysis and represented as ratios of cells expressing Administrative, technical, or material support (i.e., reporting or strongly each surface marker. Human PDX T-ALL were stained with organizing data, constructing databases): T. El-Chaar, D. Avran, Pacific Blue anti-human CD45 (Beckman Coulter), PE-Vio770 anti- S. Quentin, J.P. Meijerink, H. Dombret, E. Clappier human CD7 (clone 6B7; Miltenyi Biotec), APC anti-human CD34 Study supervision: S. Gachet, J. Soulier (clone 561; Sony), APC anti-human CD8 (Beckman Coulter), and PE anti-human CD4 (Beckman Coulter) antibodies. Immunophenotype Acknowledgments analyses were performed on a FACS Fortessa Analyzer (Beckton We would like to acknowledge contributions from François Dickinson). Guillonneau, Marjorie Leduc, and Patrick Mayeux (3P5 proteomic facility, Université Paris Descartes, Institut Cochin, Paris), Valerie Igel- Quantification of rRNA Methylation by RT-qPCR Bourguignon, Lilia Ayadi, Virginie Marchand, and Yuri Motorin (NGS This procedure, described by Belin and colleagues (34), was used on Core Facility, FR3209 CNRS-UL, Université de Lorraine, Vandoeuvre- RNAs extracted from xenografted human T-ALL samples. The first les-Nancy), Niclas Setterblad and Antonio Alberdi (Genomic Plat- step was reverse transcription (RT), performed at high (1 mmol/L) or form of Institut Universitaire d’Hématologie, University Paris low (10 μmol/L) dNTP concentration. RT is blocked by the presence Diderot, Paris, France), Carèle Fédronie (Institute of Hematology, of a methyl group at low dNTP concentration, but it is not affected Université Paris Diderot, Paris, France), Pierre De La Grange (Geno- when the reaction is performed at high dNTP concentration. The sec- splice, Paris, France), and Olivier Alibert and Amélie Rondot (CEA/ ond step was to quantify the respective cDNAs by qPCR with primers DSV/iRCM/LEFG Genopole, Evry, France). This work was sup- surrounding the methylation site. The methylation level was evalu- ported by grants to J. Soulier from the ERC St Grant Consolidator ated by the ratio of Ct obtained from the RT performed in the two #311660, the CIT program from the Ligue Contre le Cancer, the conditions. RT was performed using 50 ng total RNA in the presence Cancéropole IDF, and the ANR-10-IBHU-0002 Saint-Louis Insti- of 200 units of M-MLV reverse transcriptase (Invitrogen), 10 mmol/L tute program. Work from the J.J. Diaz team was supported by ANR DTT, 10 μmol/L or 1 mmol/L dNTPs, and 5 ng/μL of random prim- RiboMeth-13-BSV8-0012. T. El-Chaar, D. Avran, and G. Therizols ers (Promega). Reactions were incubated at 37°C for 50 minutes and were supported by fellowships from Fondation ARC, INCa, and the then stopped for 15 minutes at 70°C. Quantitative amplification Ligue Contre le Cancer.

DECEMBER 2018 CANCER DISCOVERY | 1629

RESEARCH ARTICLE Gachet et al.

The costs of publication of this article were defrayed in part by 19. Svitkin YV, Yanagiya A, Karetnikov AE, Alain T, Fabian MR, the payment of page charges. This article must therefore be hereby Khoutorsky A, et al. Control of translation and miRNA-dependent marked advertisement in accordance with 18 U.S.C. Section 1734 repression by a novel poly(A) binding protein, hnRNP-Q. PLoS Biol solely to indicate this fact. 2013;11:e1001564. 20. Kiss-Laszlo Z, Henry Y, Bachellerie JP, Caizergues-Ferrer M, Kiss T. Received July 30, 2017; revised February 12, 2018; accepted Site-specific ribose methylation of preribosomal RNA: a novel func- September 25, 2018; published first September 28, 2018. tion for small nucleolar RNAs. Cell 1996;85:1077–88. 21. Aplan PD, Jones CA, Chervinsky DS, Zhao X, Ellsworth M, Wu C, et al. An scl gene product lacking the transactivation domain induces bony abnormalities and cooperates with LMO1 to generate T-cell REFERENCES malignancies in transgenic mice. EMBO J 1997;16:2408–19. . 1 Martincorena I, Campbell PJ. Somatic mutation in cancer and nor- 22. O’Neil J, Calvo J, McKenna K, Krishnamoorthy V, Aster JC, Bassing mal cells. Science 2015;349:1483–9. CH, et al. Activating Notch1 mutations in mouse models of T-ALL. 2. Aifantis I, Raetz E, Buonamici S. Molecular pathogenesis of T-cell Blood 2006;107:781–5. leukaemia and lymphoma. Nat Rev Immunol 2008;8:380–90. 23. Tremblay M, Tremblay CS, Herblot S, Aplan PD, Hebert J, Perreault 3. Van Vlierberghe P, Ferrando A. The molecular basis of T cell acute C, et al. Modeling T-cell acute lymphoblastic leukemia induced by the lymphoblastic leukemia. J Clin Invest 2012;122:3398–406. SCL and LMO1 oncogenes. Genes Dev 2010;24:1093–105. 4. Soulier J, Clappier E, Cayuela JM, Regnault A, Garcia-Peydro M, 24. Krogh N, Jansson MD, Hafner SJ, Tehler D, Birkedal U, Christensen- Dombret H, et al. HOXA genes are included in genetic and bio- Dalsgaard M, et al. Profiling of ′2 -O-Me in human rRNA reveals a logic networks defining human acute T-cell leukemia (T-ALL). Blood subset of fractionally modified positions and provides evidence for 2005;106:274–86. ribosome heterogeneity. Nucleic Acids Res 2016;44:7884–95. 5. Homminga I, Pieters R, Langerak AW, de Rooi JJ, Stubbs A, Verstegen 25. Larsson O, Sonenberg N, Nadon R. Identification of differential M, et al. Integrated transcript and genome analyses reveal NKX2-1 translation in genome wide studies. Proc Natl Acad Sci U S A and MEF2C as potential oncogenes in T cell acute lymphoblastic 2010;107:21487–92. leukemia. Cancer Cell 2011;19:484–97. 26. Oertlin C, Lorent J, Gandin V, Murie C, Masvidal L, Cargnello M, et 6. Zenatti PP, Ribeiro D, Li W, Zuurbier L, Silva MC, Paganin M, al. Genome-wide analysis of differential translation and differential et al. Oncogenic IL7R gain-of-function mutations in childhood T-cell translational buffering using anota2seq. bioRxiv 2017. acute lymphoblastic leukemia. Nat Genet 2011;43:932–9. 27. Signer RA, Magee JA, Salic A, Morrison SJ. Haematopoietic stem cells 7. Liu Y, Easton J, Shao Y, Maciaszek J, Wang Z, Wilkinson MR, et al. require a highly regulated protein synthesis rate. Nature 2014;509: The genomic landscape of pediatric and young adult T-lineage acute 49–54. lymphoblastic leukemia. Nat Genet 2017;49:1211–8. 28. Suda T, Takubo K, Semenza GL. Metabolic regulation of hematopoi- 8. De Keersmaecker K, Atak ZK, Li N, Vicente C, Patchett S, Girardi T, etic stem cells in the hypoxic niche. Cell Stem Cell 2011;9:298–310. et al. Exome sequencing identifies mutation in CNOT3 and ribosomal 29. Blanco S, Bandiera R, Popis M, Hussain S, Lombard P, Aleksic J, genes RPL5 and RPL10 in T-cell acute lymphoblastic leukemia. Nat et al. Stem cell function and stress response are controlled by protein Genet 2013;45:186–90. synthesis. Nature 2016;534:335–40. 9. Weng AP, Ferrando AA, Lee W, Morris JPt, Silverman LB, Sanchez- 30. Cai X, Gao L, Teng L, Ge J, Oo ZM, Kumar AR, et al. Runx1 deficiency Irizarry C, et al. Activating mutations of NOTCH1 in human T cell decreases ribosome biogenesis and confers stress resistance to hemat- acute lymphoblastic leukemia. Science 2004;306:269–71. opoietic stem and progenitor cells. Cell Stem Cell 2015;17:165–77. 10. Mansour MR, Abraham BJ, Anders L, Berezovskaya A, Gutierrez A, 31. Burgess RJ, Agathocleous M, Morrison SJ. Metabolic regulation of Durbin AD, et al. Oncogene regulation. An oncogenic super-enhancer stem cell function. J Intern Med 2014;276:12–24. formed through somatic mutation of a noncoding intergenic ele- 32. Khacho M, Slack RS. Mitochondrial activity in the regulation of stem ment. Science 2014;346:1373–7. cell self-renewal and differentiation. Curr Opin Cell Biol 2017;49:1–8. 11. Schneider NR, Carroll AJ, Shuster JJ, Pullen DJ, Link MP, Borowitz 33. Giambra V, Jenkins CR, Wang H, Lam SH, Shevchuk OO, Nemirovsky MJ, et al. New recurring cytogenetic abnormalities and association O, et al. NOTCH1 promotes T cell leukemia-initiating activity by of blast cell karyotypes with prognosis in childhood T-cell acute RUNX-mediated regulation of PKC-theta and reactive oxygen species. lymphoblastic leukemia: a pediatric oncology group report of 343 Nat Med 2012;18:1693–8. cases. Blood 2000;96:2543–9. 34. Belin S, Beghin A, Solano-Gonzalez E, Bezin L, Brunet-Manquat S, 12. Remke M, Pfister S, Kox C, Toedt G, Becker N, Benner A, et al. High- Textoris J, et al. Dysregulation of ribosome biogenesis and transla- resolution genomic profiling of childhood T-ALL reveals frequent tional capacity is associated with tumor progression of human breast copy-number alterations affecting the TGF-beta and PI3K-AKT path- cancer cells. PLoS One 2009;4:e7147. ways and deletions at 6q15-16.1 as a genomic marker for unfavorable 35. Norberg E, Lako A, Chen PH, Stanley IA, Zhou F, Ficarro SB, et al. early treatment response. Blood 2009;114:1053–62. Differential contribution of the mitochondrial translation pathway 13. Bonn BR, Rohde M, Zimmermann M, Krieger D, Oschlies I, Niggli to the survival of diffuse large B-cell lymphoma subsets. Cell Death F, et al. Incidence and prognostic relevance of genetic variations in Differ 2017;24:251–62. T-cell lymphoblastic lymphoma in childhood and adolescence. Blood 36. Asnafi ,V Beldjord K, Libura M, Villarese P, Millien C, Ballerini P, 2013;121:3153–60. et al. Age-related phenotypic and oncogenic differences in T-cell 14. Oshimura M, Sandberg AA. Chromosomal 6q-anomaly in acute acute lymphoblastic leukemias may reflect thymic atrophy. Blood lymphoblastic leukaemia. Lancet 1976;2:1405–6. 2004;104:4173–80. 15. Clappier E, Gerby B, Sigaux F, Delord M, Touzri F, Hernandez L, 37. Ferrando AA, Neuberg DS, Staunton J, Loh ML, Huard C, Raimondi et al. Clonal selection in xenografted human T cell acute lymphoblastic SC, et al. Gene expression signatures define novel oncogenic pathways leukemia recapitulates gain of malignancy at relapse. J Exp Med 2011; in T cell acute lymphoblastic leukemia. Cancer Cell 2002;1:75–87. 208:653–61. 38. Lahortiga I, De Keersmaecker K, Van Vlierberghe P, Graux C, Cau- 16. Kiss T. Small nucleolar RNAs: an abundant group of noncoding welier B, Lambert F, et al. Duplication of the MYB oncogene in T cell RNAs with diverse cellular functions. Cell 2002;109:145–8. acute lymphoblastic leukemia. Nat Genet 2007;39:593–5. 17. Williams GT, Farzaneh F. Are snoRNAs and snoRNA host genes new 39. Clappier E, Cuccuini W, Kalota A, Crinquette A, Cayuela JM, Dik WA, players in cancer? Nat Rev Cancer 2012;12:84–8. et al. The C-MYB locus is involved in chromosomal translocation and 18. Mourelatos Z, Abel L, Yong J, Kataoka N, Dreyfuss G. SMN interacts genomic duplications in human T-cell acute leukemia (T-ALL), the with a novel family of hnRNP and spliceosomal proteins. EMBO J translocation defining a new T-ALL subtype in very young children. 2001;20:5443–52. Blood 2007;110:1251–61.

1630 | CANCER DISCOVERY DECEMBER 2018 www.aacrjournals.org

Deletion 6q Drives Tumor Evolution in T-ALL RESEARCH ARTICLE

40. Vannini N, Girotra M, Naveiras O, Nikitin G, Campos V, Giger S, therapeutic strategy in T-cell acute lymphoblastic leukemia. Blood et al. Specification of haematopoietic stem cell fate via modulation of 2014;124:3738–47. mitochondrial activity. Nat Commun 2016;7:13125. 50. Taghon T, Waegemans E, Van de Walle I. Notch signaling during 41. Tan SH, Bertulfo FC, Sanda T. Leukemia-initiating cells in T-cell human T cell development. Curr Top Microbiol Immunol 2012; acute lymphoblastic leukemia. Front Oncol 2017;7:218. 360:75–97. 42. Gerby B, Clappier E, Armstrong F, Deswarte C, Calvo J, Poglio S, et al. 51. Ceccaldi R, Parmar K, Mouly E, Delord M, Kim JM, Regairaz M, et al. Expression of CD34 and CD7 on human T-cell acute lymphoblastic Bone marrow failure in Fanconi anemia is triggered by an exacerbated leukemia discriminates functionally heterogeneous cell populations. p53/p21 DNA damage response that impairs hematopoietic stem Leukemia 2011;25:1249–58. and progenitor cells. Cell Stem Cell 2012;11:36–49. 43. Skrtic M, Sriskanthadevan S, Jhas B, Gebbia M, Wang X, Wang Z, 52. Shalem O, Sanjana NE, Hartenian E, Shi X, Scott DA, Mikkelsen TS, et al. Inhibition of mitochondrial translation as a therapeutic strategy et al. Genome-scale CRISPR-Cas9 knockout screening in human cells. for human acute myeloid leukemia. Cancer Cell 2011;20:674–88. Science 2014;343:84–7. 44. Kuntz EM, Baquero P, Michie AM, Dunn K, Tardito S, Holyoake TL, 53. Cox J, Hein MY, Luber CA, Paron I, Nagaraj N, Mann M. Accurate et al. Targeting mitochondrial oxidative phosphorylation eradicates proteome-wide label-free quantification by delayed normalization therapy-resistant chronic myeloid leukemia stem cells. Nat Med and maximal peptide ratio extraction, termed MaxLFQ. Mol Cell 2017;23:1234–40. Proteomics 2014;13:2513–26. 45. Yang K, Blanco DB, Chen X, Dash P, Neale G, Rosencrance C, et al. 54. Wisniewski JR, Hein MY, Cox J, Mann M. A "proteomic ruler" for Metabolic signaling directs the reciprocal lineage decisions of alpha- protein copy number and concentration estimation without spike-in beta and gammadelta T cells. Sci Immunol 2018;3. pii: eaas9818. doi: standards. Mol Cell Proteomics 2014;13:3497–506. 10.1126/sciimmunol.aas9818. 55. Bash-Imam Z, Therizols G, Vincent A, Laforets F, Polay Espinoza M, 46. Narla A, Ebert BL. Ribosomopathies: human disorders of ribosome Pion N, et al. Translational reprogramming of colorectal cancer cells dysfunction. Blood 2010;115:3196–205. induced by 5-fluorouracil through a miRNA-dependent mechanism. 47. Pelletier J, Thomas G, Volarevic S. Ribosome biogenesis in cancer: new Oncotarget 2017;8:46219–33. players and therapeutic avenues. Nat Rev Cancer 2018;18:51–63. 56. Marchand V, Blanloeil-Oillo F, Helm M, Motorin Y. Illumina-based 48. Vu LP, Prieto C, Amin EM, Chhangawala S, Krivtsov A, Calvo-Vidal RiboMethSeq approach for mapping of 2′-O-Me residues in RNA. MN, et al. Functional screen of MSI2 interactors identifies an essen- Nucleic Acids Res 2016;44:e135. tial role for SYNCRIP in myeloid leukemia stem cells. Nat Genet 57. Pear WS, Aster JC, Scott ML, Hasserjian RP, Soffer B, Sklar J, et al. 2017;49:866–75. Exclusive development of T cell neoplasms in mice transplanted 49. Peirs S, Matthijssens F, Goossens S, Van de Walle I, Ruggero K, de with bone marrow expressing activated Notch alleles. J Exp Med Bock CE, et al. ABT-199 mediated inhibition of BCL-2 as a novel 1996;183:2283–91.

DECEMBER 2018 CANCER DISCOVERY | 1631

Deletion 6q Drives T-cell Leukemia Progression by Ribosome Modulation

Stéphanie Gachet, Tiama El-Chaar, David Avran, et al.

Cancer Discov 2018;8:1614-1631. Published OnlineFirst September 28, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/2159-8290.CD-17-0831

Supplementary Access the most recent supplemental material at: Material http://cancerdiscovery.aacrjournals.org/content/suppl/2018/09/28/2159-8290.CD-17-0831.DC1

Cited articles This article cites 55 articles, 21 of which you can access for free at: http://cancerdiscovery.aacrjournals.org/content/8/12/1614.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://cancerdiscovery.aacrjournals.org/content/8/12/1614.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department Subscriptions at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerdiscovery.aacrjournals.org/content/8/12/1614. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.