Leukemia (2019) 33:2090–2125 https://doi.org/10.1038/s41375-019-0418-8 LETTER Immunotherapy CD133-directed CAR T-cells for MLL leukemia: on-target, off-tumor myeloablative toxicity Clara Bueno1 Talia Velasco-Hernandez1 Francisco Gutiérrez-Agüera1 Samanta Romina Zanetti1 ● ● ● ● Matteo L. Baroni1 Diego Sánchez-Martínez1 Oscar Molina1 Adria Closa2 Antonio Agraz-Doblás1,3 ● ● ● ● ● Pedro Marín4 Eduardo Eyras2,5 Ignacio Varela 3 Pablo Menéndez1,5,6 ● ● ● Received: 16 October 2018 / Revised: 28 November 2018 / Accepted: 12 December 2018 / Published online: 18 February 2019 © The Author(s) 2019. This article is published with open access To the Editor: cell ALL, an aggressive subtype of B-cell ALL (dismal in MLL-AF4+ infants) associated with lymphoid-to-myeloid Chimeric antigen receptors (CARs) have undoubtedly lineage switch [3, 5, 6]. We read with interest the work revolutionized immunotherapy, especially in the B-cell recently published in Leukemia by Li et al. reporting a 1234567890();,: 1234567890();,: acute lymphoblastic leukemia (ALL) arena where over 80% novel CAR targeting both CD19 and CD133 [7]. This study of complete remissions are observed in refractory/relapsed proposes to use a bi-specific CAR targeting both CD19 and (R/R) B-cell ALL patients treated with CD19-directed CAR CD133 antigens in a Boolean OR-gate approach for MLLr T-cells (CARTs) [1]. However, despite holding an unpre- B-cell ALL as a strategy to avoid and treat CD19- relapses. cedented promise, several issues still have to be resolved The authors reasoned that CD133, encoded by PROM1 before CARTs can be expanded to novel targets and/or gene, is a specific marker for MLLr leukemia because malignancies or even provided as first-line treatment in B- PROM1 is an MLL target, especially in MLL-AF4 B-cell cell ALL [2]. For instance, toxicities such as cytokine ALL [8–10]. They went on and performed in vitro assays release syndrome and immune escape mechanisms includ- showing than CD19/CD133 bi-specific CAR triggers robust ing loss of the antigen under CART-mediated pressure cytotoxicity against CD19 + CD133 + and CD19-CD133+ remain major concerns, urging further research on the B-cell lines [7], thus suggesting it may help in reducing mechanisms underlying CARTs cytotoxicity. subsequent lineage switch in MLLr B-cell ALL. In this sense, loss of CD19 antigen is frequently observed A major drawback for CD133 as target in immu- after CD19-directed CARTs therapy in B-cell ALL [3, 4], notherapy is its expression in hematopoietic stem and but is particularly common in MLL-rearranged (MLLr) B- progenitor cells (HSPCs), which would likely exert “on- target off-tumor” myeloablative, life-threatening toxicity * Clara Bueno [11, 12]. Because B-cell ALL is molecularly hetero- cbueno@carrerasresearch.org geneous and can be diagnosed during infancy, childhood * Pablo Menéndez pmenendez@carrerasresearch.org and adulthood, we have characterized PROM1/CD133 expression in a large cohort of cytogenetically distinct B- 1 cell ALL subgroups (n = 212 patients) as well as in dif- Josep Carreras Leukemia Research Institute and Department of Biomedicine, School of Medicine, University of Barcelona, ferent subpopulations of normal CD34+ HSPCs obtained Barcelona, Spain across hematopoietic ontogeny from 22-weeks old human 2 Pompeu Fabra University, Barcelona, Spain fetal liver (FL, prenatal), cord blood (CB, perinatal), and 3 adult G-CSF-mobilized peripheral blood/bone marrow Instituto de Biomedicina y Biotecnología de Cantabria (CSIC-UC- Sodercan), Departamento de Biología Molecular, Universidad de (PB/BM, postnatal). An initial analysis of publicly Cantabria, Santander, Spain available RNA-seq data [13] from 170 diagnostic B-cell 4 Hematology Department, Hospital Clínico de Barcelona, ALL patients confirmed that PROM1 is overexpressed in Barcelona, Spain patients with MLLr B-cell ALL, although its expression is 5 Instituciò Catalana de Recerca i Estudis Avançats (ICREA), not significantly higher than in other cytogenetic sub- Barcelona, Spain groups (Fig. 1a). We then analyzed PROM1 during HSPC 6 Centro de Investigación Biomédica en Red de Cáncer (CIBER- development and observed that PROM1 is highly ONC), ISCIII, Barcelona, Spain expressed in early normal hematopoietic stem cells (HSC) Letter 2091 A B PROM1 expression PROM1 expression *** *** n.s. n.s. n.s. 20000 Normalized Counts logCPM 10000 n.s. n.s. n.s. n.s. n.s. BCR-ABLL+ ETV6-RUNX1+ Hyperdiploid Hypodiploid Normal Karyotype TXCF3-PBX1 B-other 11q23/MLLr 0 HSC MPP LMPP CMP GMP MEP CLP 11q23/MLLr C 100 2000 MFI Total cell population 1500 %CD133+ Cells 50 1000 500 0 0 Prenatal CD34 Cells (FL/BM-derived cells) n=8 Perinatal CD34 Cells (CB-derived cells) n=7 Postnatal CD34 Cells (BM/PB-derived cells) n=7 MLL germline B-ALL n=5 (3 Relapses, 2 PDX) 11q23/MLLr B-ALL n=7 (3 Diagnostics, 4 PDX) Fig. 1 Characterization of CD133/PROM1 expression in B-cell ALL common lymphoid progenitors (CLP) [14]. Data shown as normalized and normal HSPCs. a Expression level of PROM1 in the indicated counts. The boxes define the first and third quartiles. The horizontal cytogenetic subgroups of B-cell ALL (n = 170 patients at diagnosis) line within the box represents the median. c Frequency (left) and mean determined by RNA-seq represented in log2(CPM) scale, with CPM fluorescence intensity (MFI, middle) of CD133+ BM blasts/cells in = counts per million [13]. b RNA-seq analysis comparing the MLLr (n = 7) and non-MLL B-cell ALLs (n = 5) primary diagnostic/ expression of PROM1 in 11q23/MLLr B-cell ALL (n = 29 patients) relapse samples or primografts (PDXs), and normal CD34+ HSPCs with that in distinct fractions of Lin-CD34 + CD38-CD19- non- derived from FL (n = 8), CB (n = 7) and adult PB/BM (n = 7). lymphoid normal HSPCs (HSC hematopoietic stem cells, MPP mul- Representative FACS dot plots for CD133 in normal CD34+ HSPCs tipotent progenitors, LMPP lymphoid-primed multipotent progenitors, (upper right) and BM samples from two independent MLLr B-cell CMP common myeloid progenitors, GMP granulocyte-monocyte ALL patients (bottom right) progenitor, MEP megakaryocyte-erythroid progenitors) and in and multipotent progenitors (MPP) with its expression (13/32) PROM1 levels that HSCs and MPPs, which raises decreasing from the lymphoid-primed multipotent pro- doubts about the suitability of PROM1 as a target for B- genitors (LMPP) onwards with its expression being mar- cell ALL immunotherapy [15]. ginal at later stages of myeloid differentiation FACS clinical immunophenotyping provides a priori a (megakaryocyte-erythroid progenitors, MEP) and com- more rapid and feasible clinically relevant diagnostic mon lymphoid progenitors (CLP) [14] (Fig. 1b). Impor- information than RNA-seq during the decision-making tantly, 70% (22/32) of 11q23/MLLr B-cell patients (both process. Thus, we next FACS-analyzed the expression of MLL-AF4 and MLL-AF9) express equal (9/32) or lower CD133 (PROM1 gene product) in the cell surface of BM- 2092 Letter derived primary blasts and primografts (PDXs) obtained financial support from the Obra Social La Caixa-Fundaciò Josep from 11q23/MLLr (n = 7) and non-MLL (n = 5) B-cell Carreras. SRZ and TV are supported by a Marie Curie fellowships. OM is supported by the Catalan Government through a Beatriu de ALL patients, and in comparison with healthy prenatal Pinos fellowship. MB is supported by MINECO through a PhD (22 weeks old FL), perinatal (CB) and adult (PB/BM) CD34 scholarship. PM is an investigator of the Spanish Cell Therapy + HSPCs (Fig. 1c). Consistent with the RNA-seq data, the cooperative network (TERCEL). expression of CD133 in 11q23/MLLr blasts is intermingled with that observed in CD34+ HSPCs across hematopoietic Compliance with ethical standards ontogeny (Fig. 1c). Our data demonstrates that PROM1/CD133 is similarly Conflict of interest The authors declare that they have no conflict of interest. expressed between MLLr B-cell ALL primary blasts and normal non-lymphoid HSPCs across ontogeny, thus indi- Publisher’s note: Springer Nature remains neutral with regard to cating that “on-target, off-tumor” toxic/myeloablative jurisdictional claims in published maps and institutional affiliations. effects are likely to occur if used in a bi-specific CAR approach where CD133 antigen will be constantly targeted Open Access This article is licensed under a Creative Commons regardless of the co-expression of CD19 in the same cell. Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as Our data therefore raises concerns about using CD133 as a long as you give appropriate credit to the original author(s) and the target for MLLr B-cell ALL immunotherapy. An alternative source, provide a link to the Creative Commons license, and indicate if to circumvent HSPC toxicity would be to engineer dual changes were made. The images or other third party material in this CAR T-cells with one CAR engaging an antigen (i.e., article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not CD19) mediating T-cell activation and another CAR included in the article’s Creative Commons license and your intended engaging a second antigen (i.e., CD133) mediating T-cell use is not permitted by statutory regulation or exceeds the permitted co-stimulation [16]. Unfortunately, although such a CD19/ use, you will need to obtain permission directly from the copyright CD133 dual CAR might be likely safe due to its cytotoxi- holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. city being restrained only to cells co-expressing CD19 and CD133, its specific cytotoxic performance will be poor since not the entire MLLr B-cell ALL blast population is CD19 + CD133+ (Fig. 1c). Another alternative approach to References prevent HSPC toxicity would be to have in place a potent 1. Maude, SL, Laetsch TW, Buechner J, Rives S, Boyer M, molecular switch (i.e., iCas9) to eliminate CAR133- Bittencourt H, et al. Tisagenlecleucel in Children and Young expressing T-cells as necessary [17]. Further long-term Adults with B-Cell Lymphoblastic Leukemia. N Engl J Med. in vivo studies using both primary B-cell ALL cells and 2018;378:439–48. normal HSCPs remain to be conducted to elucidate the 2. Ghorashian S, Amrolia P, Veys P. Open access? Widening access to chimeric antigen receptor (CAR) therapy for ALL. Exp efficacy versus the myeloablative toxicity of a CAR CD133 Hematol. 2018;66:5–16. [18, 19]. 3. Gardner R, Wu D, Cherian S, Fang M, Hanafi LA, Finney O, et al. Acquisition of a CD19-negative myeloid phenotype allows Data availability immune escape of MLL-rearranged B-ALL from CD19 CAR-T- cell therapy. Blood. 2016;127:2406–10. 4. Sotillo E, Barrett DM, Black KL, Bagashev A, Oldridge D, Wu G, All genomic data is already publicly available. A full data et al. Convergence of acquired mutations and alternative splicing availability will be provided. of CD19 enables resistance to CART-19 immunotherapy. Cancer Discov. 2015;5:1282–95. Acknowledgements We thank the Interfant treatment protocol and 5. Sanjuan-Pla A, Bueno C, Prieto C, Acha P, Stam RW, local physicians for contributing patient samples: Dr. Ronald W Stam Marschalek R, et al. Revisiting the biology of infant t(4;11)/MLL- (Princess Maxima Centre, Utrech), Dr. Mireia Camos and Dr. Jose AF4 + B-cell acute lymphoblastic leukemia. Blood. 2015; Luis Fuster (Spanish Society of Pediatric Hematoncology), Dr. Paola 126:2676–85. Ballerini (A. Trousseau Hospital, Paris). We also thank Prof. Paresh 6. Weiland J, Pal D, Case M, Irving J, Ponthan F, Koschmieder S, Vyas (Oxford Univeristy, UK) and Prof. Kajsa Paulsson (Lund Uni- et al. BCP-ALL blasts are not dependent on CD19 expression for versity, Sweden) for facilitating access to their RNA-seq database. leukaemic maintenance. Leukemia. 2016;30:1920–3. This work has been supported by the European Research Council 7. Li D, Hu Y, Jin Z, Zhai Y, Tan Y, Sun Y, et al. TanCAR T cells (CoG-2014-646903, PoC-2018-811220) to PM, the Spanish Ministry targeting CD19 and CD133 efficiently eliminate MLL leukemic of Economy and Competitiveness (MINECO, SAF-SAF2016-80481- cells. Leukemia. 2018;32:2012–6. R, BIO2017-85364-R) to PM and EE, the Generalitat de Catalunya 8. Guenther MG, Lawton LN, Rozovskaia T, Frampton GM, Levine (SGR330, SGR102 and PERIS) to PM and EE, the Spanish Asso- SS, Volkert TL, et al. Aberrant chromatin at genes encoding stem ciation against cancer (AECC-CI-2015) to CB, and the Health Institute cell regulators in human mixed-lineage leukemia. Genes Dev. Carlos III (ISCIII/FEDER, PI14-01191) to CB. PM also acknowledges 2008;22:3403–8. Letter 2093 9. Mak AB, Nixon AM, Moffat J. The mixed lineage leukemia 15. Agraz-Doblas A, Bueno C, Bashford-Rogers R, Anindita R, (MLL) fusion-associated gene AF4 promotes CD133 transcrip- Schneider P, Bardini M, et al. Unravelling the cellular origin and tion. Cancer Res. 2012;72:1929–34. clinical prognosis markers of infant B-cell acute lymphoblastic 10. Wuchter C, Ratei R, Spahn G, Schoch C, Harbott J, Schnittger leukemia using genome-wide analysis. Haematologica. 2018; in S, et al. Impact of CD133 (AC133) and CD90 expression press. analysis for acute leukemia immunophenotyping. Haematolo- 16. Aldoss I, Bargou RC, Nagorsen D, Friberg GR, Baeuerle PA, gica. 2001;86:154–61. Forman SJ. Redirecting T cells to eradicate B-cell acute lym- 11. Menendez P, Caballero MD, Prosper F, Del Canizo MC, Perez- phoblastic leukemia: bispecific T-cell engagers and chimeric Simon JA, Mateos MV, et al. The composition of leukapheresis antigen receptors. Leukemia. 2017;31:777–87. products impacts on the hematopoietic recovery after autologous 17. Diaconu I, Ballard B, Zhang M, Chen Y, West J, Dotti G, et al. transplantation independently of the mobilization regimen. Inducible Caspase-9 selectively modulates the toxicities of CD19- Transfusion. 2002;42:1159–72. Specific chimeric antigen receptor-modified T cells. Mol Ther. 12. Yin AH, Miraglia S, Zanjani ED, Almeida-Porada G, Ogawa M, 2017;25:580–92. Leary AG, et al. AC133, a novel marker for human hematopoietic 18. Pizzitola I, Anjos-Afonso F, Rouault-Pierre K, Lassailly F, Tet- stem and progenitor cells. Blood. 1997;90:5002–12. tamanti S, Spinelli O, et al. Chimeric antigen recptors against 13. Paulsson K, Lilljebjorn H, Biloglav A, Olsson L, Rissler M, Castor CD33/CD123 antigens efficiently target primary acute myeloid A, et al. The genomic landscape of high hyperdiploid childhood cells in vivo. Leukemia. 2014;28:1596–605. acute lymphoblastic leukemia. Nat Genet. 2015;47:672–6. 19. Kenderian SS, Ruella M, Shestova O, Klichincky M, Aikawa V, 14. Quek L, Otto GW, Garnett C, Lhermitte L, Karamitros D, Stoilova Morrissette JJD, et al. CD33-directed chimeric antigen receptor B, et al. Genetically distinct leukemic stem cells in human T cells exhibit potent preclinical activity against human acute CD34- acute myeloid leukemia are arrested at a hemopoietic myeloid leukemia. Leukemia. 2015;29:1637–47. precursor-like stage. J Exp Med. 2016;213:1513–35. Leukemia (2019) 33:2093–2097 https://doi.org/10.1038/s41375-019-0423-y Chronic lymphocytic leukemia Telomere fusions associate with coding sequence and copy number alterations in CLL 1 Laura Escudero ● Kez Cleal1 Kevin Ashelford1 Chris Fegan ● ● 1 ● Chris Pepper 2 ● Kate Liddiard1 ● 1 Duncan M. Baird Received: 6 September 2018 / Revised: 19 October 2018 / Accepted: 11 February 2019 / Published online: 22 February 2019 © The Author(s) 2019. This article is published with open access To the Editor: instability, driving disease progression. To investigate the impact of telomere dysfunction on the CLL genome, we Short-dysfunctional telomeres are detected prior to clinical performed a large-scale molecular characterisation of telo- progression in chronic lymphocytic leukaemia (CLL) and mere fusion events in CLL B-cells. A cohort of 276 CLL result in chromosomal fusions that propagate genome patient samples was selected for analysis based on short telomere length (TL) profiles, with the majority (97%, n = 269) having mean TL within the previously-defined fuso- genic range in CLL [1]. Patient samples were screened for These authors contributed equally: Kate Liddiard, Duncan M. Baird. the presence of telomere fusions using a single-molecule Supplementary information The online version of this article (https:// telomere fusion assay [2] modified to include the 5p telo- doi.org/10.1038/s41375-019-0423-y) contains supplementary mere (Supplementary Figure 1). Telomere fusions were material, which is available to authorized users. detected in 72% (198/276) of the samples, which were * Duncan M. Baird subsequently arbitrarily stratified by fusion frequency bairddm@cf.ac.uk (Supplementary Table 1). Fusions were detected for all telomeres assayed, including the 5p telomere, for which 1 Division of Cancer and Genetics, School of Medicine, Cardiff fusions were present in 23% (40/177) of patient samples University, Cardiff, UK (Supplementary Figure 2, Supplementary Table 2). 2 Brighton and Sussex Medical School, Sussex University, High-resolution characterisation of single-molecule Brighton, UK amplified telomere fusions from nine CLL patients with 2094 Letter Fig. 1 Signature of telomere fusions for 9 CLL patient samples. Stong et al. [12]. Colour code: telomere-telomere inter-chromosomal Circos plots showing the validated results obtained from the inter- (black), telomere-telomere intra-chromosomal for 5p, 17p and XpYp chromosomal and intra-chromosomal telomere fusion analysis from (blue), inter-chromosomal or intra-chromosomal for 16p and 21q nine CLL patient samples. Circos plot with each chromosome and families (light blue), and inter-chromosomal telomere-genomic its telomeres (1p telomere, Chr1, 1q telomere) around the circle (green), telomere-2q13 (orange) and telomere-ChrM (pink). Telo- orientated clockwise. Additional notches indicate linkages specifi- mere fusion events with unknown sub-telomeric sequence were not cally aligning with subtelomeric sequence references derived from included the highest fusion frequency was performed by Illumina These captured loci included; the ancestral telomere at HiSeq4000 paired-end sequencing. Following a customised Chr2q13-14 (n = 11), mitochondrial DNA (n = 4) and bioinformatics analysis pipeline [3] and manual curation other non-telomeric genomic loci (n = 78). Complex inter- (Supplementary Figure 3), 914 unique telomere fusions were chromosomal events involving multiple disparate loci were resolved (Supplementary Figure 4), of which 19% (172/914) also detected (n = 7) (Supplementary Figures 5–9). involved the 5p telomere (Supplementary Figure 5). Intra- Distinct signatures of telomere fusions across the gen- (sister-chromatid) or inter-chromosomal telomere fusion ome could be described for each CLL patient sample events were identified, as well as recombinations involving (Fig. 1). Two patients (CLL3 and CLL6) displayed sim- non-telomeric loci incorporated into telomere fusions ple signatures, defined by the presence solely of intra- (Fig. 1; Supplementary Figure 4, Supplementary Table 3). chromosomal and/or inter-chromosomal telomere-telomere Letter 2095 fusions. In contrast, the CLL8 sample telomere fusion CLL B-cells was revealed using GSEA Gene Set Enrichment profile revealed abundant genomic linkages, including with Analysis (GSEA, v5.2) Molecular Signatures Database the ancestral telomere at 2q13 and mitochondrial DNA. (MSigDB) [7]. This gene set included HTR7, KIF26B and Samples CLL1, CLL2, CLL4, CLL5, CLL7 and CLL9 were LPHN1 (p-value 1.5e−6; FDR q-value 2.7e−2) —genes pre- characterised by complex signatures with a combination of viously found to be upregulated in CD5+/CD19+/CD38+ most or all categories of telomere fusion events identified in CLL cells associated with worse patient prognosis, compared this study (Fig. 1; Supplementary Table 4). with patient-matched CD5−/CD19−/CD38− CLL cells in a Telomere dysfunction is associated with increased panel of six patient samples [8]. Strikingly, 36% (11/31) of genomic instability and disease progression in CLL [1, 4], all genes disrupted by a telomere fusion event for which the therefore a comprehensive analysis of all patient-derived junction could be validated were classified as expressed or telomere fusions with non-telomeric genomic loci was associated with B lymphocytes or CLL B-cells. These genes undertaken. Locations and junction sequences pertaining to included CD8A, RORA, TESPA1, DMD, NOX5, NTF3, EVI5 all 93 (10% total fusions) identified inter-chromosomal and FTO (Supplementary Table 7) with documented patho- fusions were investigated to determine commonality of logical relevance. A significant enrichment in genes posses- global or local sequence context as well as providing evi- sing binding motifs matching the B-cell-expressed homeobox dence for the engagement of specific DNA repair processes. transcription factor, HNF1α (TCF1) [9], within their pro- These inter-chromosomal genomic fusions were less abun- moters was also identified (DMD, RORA, NTF3 and HTR7; dant than pure telomeric inter-chromosomal fusions that p-value 2.51e−5; FDR q-value 1.31e−2; Supplementary represented 38% of all fusions characterised. Table 8). Furthermore, a noteworthy association of fusion- Inter-chromosomal fusions with non-telomeric genomic disrupted genes with gene sets over-expressed in other types loci were identified in all nine CLL patient samples. Indi- of cancer including breast and liver was also revealed by vidual events were validated by manual sequence analysis, these analyses. revealing 68% (63/93) had fusion junctions covered by We have previously shown that intra-chromosomal telo- junction-spanning sequence read pairs (mFJ) and 32% mere fusion is accompanied by extensive resection that results (30/93) had unmapped junctions (uFJ). Each fusion junction in asymmetric deletion of the participating sister-chromatids location was depicted on the ideogram in Fig. 2a. Notably, [3]. To assess whether this was true for CLL B-cells, the extent the loci disrupted by telomere fusions (summarised in of DNA end-processing at each sister-chromatid was exam- Supplementary Table 5) were not randomly distributed ined for intra-chromosomal fusions with mFJ. The distance throughout the genome since there was no simple correla- from the start of the telomere repeat sequences to the fusion tion with chromosome length (r2 = 0.44) or coding gene junction for each of the chromatids involved in the fusion density of the respective chromosomes (r2 = 0.32) (Fig. 2b, event was determined and the difference calculated to obtain a Supplementary Figure 10). However, loci with previously- measure of asymmetry (Fig. 2c, d; Supplementary Tables 9- reported copy number aberrations in CLL [5] were found to 10). The uneven distribution of fusion junctions across the 5p be incorporated into telomere fusions, including 2p15, sub-telomere (n = 14) is consistent with the location of a CpG 2p11.2 (2 events), 2q13 (11 events), 6q22.31, 11q22.2 and island and suggests that the GC-rich sequence may hamper the 18q21.32 (single events). In addition, a complex telomere detection of 5p fusion events (Supplementary Figure 11). fusion was detected involving four distinct loci including Thus, 5p telomere fusions may be under-represented in the 13q14.2 that is frequently deleted in CLL (Supplementary data and may have an even greater impact on CLL disease Figure 4B). than presently recognised. In contrast, telomere fusion junc- Inter-chromosomal telomere fusions occurred within cod- tions were effectively captured across the 17p (n = 30) and ing DNA more frequently than expected by chance. Over half XpYp (n = 20) telomeres (Fig. 2c). Asymmetry of sister- (57%) of mFJ were within introns and exons of protein- chromatids was observed for 5p, 17p and XpYp with a mean coding genes (Supplementary Table 5), significantly higher of 1408 bp, 1240 bp and 695 bp, respectively (Fig. 2d). The (Chi-squared analysis p = 0.0024) than the average 42% gene degree of asymmetry was significantly greater than the theo- content of the human genome (based on the hg19 RefGene). retical value 0 (one sample t-test, p < 0.001). This indicates We also observed 15% and 9% mFJ fused with Common that fusion occurs between sister-chromatids of different Fragile Sites (CFSs) and Alu elements, respectively; however, lengths in CLL B-cells, consistent with our observations in these were similar to the proportion of CFSs (15%) and Alu other models [2, 3, 10]. No significant differences were found sequences (11%) identified across the human genome [6]. in the extent of asymmetry between the 5p, 17p and XpYp All 31 protein-coding genes disrupted by telomere chromosome ends (Kruskal-Wallis, p = 0.1661). fusions with mFJ were further investigated for potential High-resolution analysis of each CLL mFJ was per- association with CLL pathogenesis (Supplementary Table 6). formed to investigate candidate DNA repair mechanisms An enrichment in genes overexpressed in CD38+ patient that may underlie distinct types of telomere fusion events. 2096 Letter Fig. 2 Characterisation of telomere fusions detected across the gen- on the 5p sub-telomere. Location of the fusion primer indicated, ome. a Validated inter-chromosomal telomere fusion events (n = 93) determines the limit of the assay from the telomere. d Level of on a karyotype map generated in Ensembl. Telomere fusions with asymmetry was determined by calculating the deletion difference genomic, ancestral telomere 2q13 and mitochondria DNA/Chr. Each between each chromatid of the same fusion event. e Microhomology colour represents a different patient sample. Continuous arrow-heads (bp) at the fusion junction was compared for the distinct type of indicate mapped fusion junctions (mFJ) and discontinuous arrow- events: TTAGGG-CCCTAA (00), Sub-telomere-TTAGGG (0), intra- heads represent unmapped fusion junctions (uFJ, location of the read chromosomal (1), intra-chromosoma or inter-chromosomal of 16p-16p represented). b Number of validated inter-chromosomal telomere- and 21q-21q families (1/2), inter-chromosomal telomeric fusion events genomic fusion junctions per Mb of DNA for each chromosome (2T), inter-chromosomal fusions with the ancestral telomere at 2q13 ordered by length (size obtained from Ensembl). c Sister-chromatid (2A) and inter-chromosomal fusions with genomic loci (2G). Mean, deletion and d asymmetry for the 5p, 17p and Xp chromosome ends of SD and SE are indicated below intra-chromosomal fusion events. Green box highlights the CpG island Insertions of templated, untemplated and/or potential telo- 1-21 nucleotides with a mean of 4.5 nucleotides. In contrast, mere variant repeat sequences were observed at 6% no insertions were identified at fusions with non-telomeric (50/796) of mFJ: 23/50 for Telomere-Sub-telomere, 4/50 loci. Statistically-significant differences in the extent of for intra-chromosomal, 2/50 for intra/inter, 19/50 for telo- microhomology usage at fusion junctions were determined meric inter-chromosomal, 1/50 for telomere-Chr2q13 and for the different types of telomere fusion events (Kruskal- 1/50 for telomere-ChrM fusions. Insertions ranged from Wallis p < 0.001 and Dunn’s Multiple Comparison Test) Letter 2097 (Fig. 2e; Supplementary Table 11). Inter-chromosomal Open Access This article is licensed under a Creative Commons fusions with non-telomeric loci (mean = 9.1 bp; n = 43), Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as together with intra-chromosomal sister-chromatid events long as you give appropriate credit to the original author(s) and the (mean = 4.1 bp; n = 32), displayed the greatest amounts of source, provide a link to the Creative Commons license, and indicate if junction microhomology. In contrast, very low or an changes were made. The images or other third party material in this absence of microhomology at the fusion point was observed article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not for inter-chromosomal telomeric fusions (mean = 1.5 bp; included in the article’s Creative Commons license and your intended n = 315), Telomere-Telomere (TTAGGG-CCCTAA; use is not permitted by statutory regulation or exceeds the permitted mean = 0.8 bp; n = 12) and Telomere-Sub-telomere use, you will need to obtain permission directly from the copyright (mean = 1.6 bp; n = 303) subgroups. Long tracts of holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/. microhomology of up to 23 bp, were observed at inter- chromosomal fusion junctions with non-telomeric loci (Fig. 2e). When the usage of microhomology was >10 bp, References the sequence was enriched for the repeat unit of (AC)n 1. Lin TT, Norris K, Heppel NH, Pratt G, Allan JM, Allsup DJ, et al. (Supplementary Figure 5); 40% (6/15) of events that con- Telomere dysfunction accurately predicts clinical outcome in tained at least (AC)5 (motif ACACACACAC), consistent chronic lymphocytic leukaemia, even in patients with early stage with repair utilizing single-stranded annealing [11]. disease. Br J Haematol. 2014;167:214–23. 2. Letsolo BT, Rowson J, Baird DM. Fusion of short telomeres in Taken together, our data reveal the impact of short- human cells is characterized by extensive deletion and micro- dysfunctional telomeres on the evolving CLL genome, homology, and can result in complex rearrangements. Nucleic generating tumour heterogeneity that may affect patient Acids Res. 2010;38:1841–52. prognosis. We have revealed that dysfunctional telomeres 3. Liddiard K, Ruis B, Takasugi T, Harvey A, Ashelford KE, Hendrickson EA, et al. Sister chromatid telomere fusions, predominantly fuse with protein-coding DNA including but not NHEJ-mediated inter-chromosomal telomere fusions, genes expressed in CLL B-cells and other tumours. We occur independently of DNA ligases 3 and 4. Genome Res. have also identified complex telomere fusions involving 2016;26:588–600. multiple non-telomeric loci across the CLL genome, 4. Lin TT, Letsolo BT, Jones RE, Rowson J, Pratt G, Hewamana S, et al. Telomere dysfunction and fusion during the progression of including those with known copy number aberrations in chronic lymphocytic leukemia: evidence for a telomere crisis. CLL. Our data implicate diverse DNA repair mechanisms at Blood. 2010;116:1899–907. play in CLL tumour initiation and progression, including 5. Salaverria I, Martin-Garcia D, Lopez C, Clot G, Garcia-Aragones C-NHEJ, A-NHEJ and SSA. These repair pathways provide M, Navarro A, et al. Detection of chromothripsis-like patterns with a custom array platform for chronic lymphocytic leukemia. potential therapeutic targets and combinations of ther- Genes Chromosomes Cancer. 2015;54:668–80. apeutic agents targeting these specific pathway components 6. Fungtammasan A, Walsh E, Chiaromonte F, Eckert KA, Makova may effectively sensitise CLL B-cell clones with ongoing KD. A genome-wide analysis of common fragile sites: what fea- telomere dysfunction to improve patient outcomes. tures determine chromosomal instability in the human genome? Genome Res. 2012;22:993–1005. 7. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Acknowledgements The authors would like to thank the CLL patients. Gillette MA, et al. Gene set enrichment analysis: a knowledge- The study was undertaken with the support of the UK CLL Clinical based approach for interpreting genome-wide expression profiles. Trials Sub-Group. This work was supported by Cancer Research Wales, Proc Natl Acad Sci USA. 2005;102:15545–50. Cardiff University and Cancer Research UK (C17199/A18246). Patient 8. Pepper C, Ward R, Lin TT, Brennan P, Starczynski J, Musson M, material was obtained from the UK CLL Trials Biobank, University of et al. Highly purified CD38 + and CD38- sub-clones derived from Liverpool, which is funded by Bloodwise. the same chronic lymphocytic leukemia patient have distinct gene expression signatures despite their monoclonal origin. Leukemia. Author contributions LE performed the experimental work, analysed 2007;21:687–96. the experimental and bioinformatics data and wrote the manuscript; 9. von Wnuck Lipinski K, Sattler K, Peters S, Weske S, Keul P, KC carried out the bioinformatics pipeline; CF provided clinical Klump H, et al. Hepatocyte nuclear factor 1A Is a cell-intrinsic samples and edited the manuscript; CP provided clinical input and transcription factor required for B cell differentiation and devel- edited the manuscript; KL jointly supervised the study and edited the opment in mice. J Immunol. 2016;196:1655–65. manuscript; DMB jointly supervised the study and edited the 10. Capper R, Britt-Compton B, Tankimanova M, Rowson J, manuscript. Letsolo B, Man S, et al. The nature of telomere fusion and a definition of the critical telomere length in human cells. Genes Compliance with ethical standards Dev. 2007;21:2495–508. 11. Bhargava R, Onyango DO, Stark JM. Regulation of single-strand annealing and its role in genome maintenance. Trends Genet. Conflict of interest The authors declare that they have no conflict of 2016;32:566–75. interest. 12. Stong N, Deng Z, Gupta R, Hu S, Paul S, Weiner AK, et al. Subtelomeric CTCF and cohesin binding site organization using Publisher’s note: Springer Nature remains neutral with regard to improved subtelomere assemblies and a novel annotation pipeline. jurisdictional claims in published maps and institutional affiliations. Genome Res. 2014;24:1039–50. 2098 Letter Leukemia (2019) 33:2098–2102 https://doi.org/10.1038/s41375-019-0421-0 Multiple myeloma gammopathies BCL2 blockade overcomes MCL1 resistance in multiple myeloma Ka Tat Siu1 Cherrie Huang2 Cristina Panaroni1 Kenta Mukaihara1 Keertik Fulzele1 Rosemary Soucy1 ● ● ● ● ● ● Cassandra Thorburn1 Justin Cidado3 Lisa Drew3 Shrikanta Chattopadhyay2 Noopur Raje1 ● ● ● ● Received: 24 December 2018 / Revised: 3 February 2019 / Accepted: 7 February 2019 / Published online: 28 February 2019 © Springer Nature Limited 2019 To the Editor: level of MCL1 expression correlates with disease progres- sion, and a higher MCL1 expression is associated with Disruption of the intrinsic apoptotic pathway by the aberrant shorter survival [5]. Since MM is heavily reliant on MCL1, expression of the BCL2 family members are frequent events in MM patients, particularly those at relapse, would benefit multiple myeloma (MM). In particular, the anti-apoptotic pro- from an MCL1-targeted therapy. However, there is no tein myeloid cell leukemia-1 (MCL1) is highly expressed in FDA-approved drug with the ability to selectively target MM and plays a crucial role in disease progression [1, 2]. MCL1. To address this unmet medical need, a few selective Using an unbiased approach to analyze cell death clustering, MCL1 inhibitors are currently being developed in pre- Gomez-Bougie and colleagues recently identified a group of clinical phase or clinical trials and have thus far shown MM patients insensitive to all the three classes of BH3 promising results as single agents or used in combination mimetics targeting MCL1, BCL2, and BCLxL. These BH3 with established therapies in various cancers, particularly in mimetic-resistant patients were mostly found at diagnosis, and hematologic malignancies [6–8]. To further explore MCL1 they often do not possess any recurrent chromosomal translo- biology in MM, we use a clinical-grade small-molecule cations. BCL2 dependency is mainly found in patients with t MCL1 inhibitor, AZD5991, to investigate the mechanistic (11;14) CCND1 translocation. BCLxL dependency is rare in underpinning of MCL1 inhibition in MM. MM as they are often co-dependent on either BCL2 or MCL1. AZD5991 is a potent and selective macrocyclic inhibitor of MCL1 dependency was strikingly predominant at relapse and MCL1 [9] that is currently in phase I clinical trial in patients in patients lacking common translocations and in the with relapsed or refractory MM and other hematologic CCND1 subgroup. These findings suggested a shift of cellular malignancies (ClinicalTrials.gov Identifier: NCT03218683). plasticity towards MCL1 dependence during disease progres- Using AZD5991 as a test compound, we aimed to determine sion as a result of prior treatments or clonal selection [3]. the survival dependency of human MM cells on the anti- A majority of well-established human MM cell lines and apoptotic protein MCL1. First, we evaluated the cytotoxicity low-passage patient-derived myeloma cell lines have been of AZD5991 on a panel of MM cell lines. MM cell lines shown to be MCL1 dependent using pharmacological showed a heterogeneous response to MCL1 inhibition. inhibitors or gene editing approaches that specifically target AZD5991 treatment resulted in dose-dependent cytotoxicity MCL1 [4]. Clinically, overexpression of MCL1 is observed with EC50 values (Table S1) ranging from 64 nM to 417 nM at in 52% of MM patients at diagnosis and 81% at relapse. The 24 h for AZD5991-sensitive cell lines (Fig. 1a). We next assessed the effect of AZD5991 in MM patient-derived CD138+ cells. AZD5991 treatment led to 40–82% decrease in Supplementary information The online version of this article (https:// viability of primary cells isolated from relapsed and refractory doi.org/10.1038/s41375-019-0421-0) contains supplementary MM patients at a dose of 300 nM at 24 h (Fig. 1b). AZD5991 material, which is available to authorized users. also induces potent anti-MM activity in vivo [9]. Together, * Noopur Raje these results indicate that AZD5991 has promising single- NRAJE@mgh.harvard.edu agent activity, but it would be prudent to study it in combi- nation with other anti-MM therapies. 1 Massachusetts General Hospital Cancer Center, Harvard Medical To understand the mechanism of cytotoxicity, we treated School, Boston, MA 02114, USA the AZD5991-sensitive MM.1S and H929 cells with 50 nM 2 Broad Institute, Cambridge, MA 02142, USA of AZD5991 for 24 h. The decrease in cell viability upon 3 Bioscience, Oncology, IMED Biotech Unit, AstraZeneca, MCL1 inhibition is due to an increase in apoptosis as shown Boston, MA, USA by an increase in Annexin V signals after MCL1 inhibition. Letter 2099 a AZD5991 Dose Response (24h) INA6 b Patient-derived CD138+ cells 140 MM.1R 120% OPM1 120 100% OPM2 AZD5991 % VIABILITY 100 LR5 80% 0nM % viability 80 KMS11 60% 50nM ANBL6VR 60 40% DOX40 300nM 40 H929 20% 20 LP-1 0% 0 MM.1S Paent 1 Paent 2 Paent 3 0 200 400 600 800 1000 RPMI-8226 (nM) c MCL1 inhibion in MM.1S (BMSC condioned media) e Cytokine enrichment upon MM-BMSC cell-cell contact 140% 150 BMSC+MM.1S 120% 0nM BMSC Normalized Signal Intensity 100% 25nM MM.1S % viability 80% 50nM 100 60% 100nM 200nM 40% 400nM 20% 50 800nM 0% 20% αMEM BMSC condioned 1600nM media 0 d MCL1 inhibion in MM.1S (BMSC coculture) IL-6 MCP-1 (CCL2) TIMP-2 IL-8 (CXCL8) IGFBP-2 TIMP-1 OSM G-CSF VEGF-A GRO a/ß/γ 800% 700% 0nM 600% 25nM % Growth 500% 50nM 400% 100nM 300% 200nM 200% 400nM 100% 800nM 0% BMSC BMSC + MM.1S 1600nM MM.1S Fig. 1 The BM microenvironment confers a protective effect on MM in the presence of (c) BMSC-conditioned media or (d) BMSCs. Cell cells against MCL1 inhibition. a A panel of human MM cell lines was viability in c was assessed by CellTiter-Glo® One Solution Assay. Cell treated with increasing doses of AZD5991 (0–1000 nM) for 24 h, and proliferation in d was evaluated by CyQUANT® NF Cell Proliferation cell viability was measured by CellTiter-Glo® One Solution Assay. b Assay. e Semi-quantitative detection of a panel of 80 human proteins Patient-derived CD138+ MM cells were treated with increasing doses in cell culture media of MM.1 S cells, BMSCs, and BMSC+MM.1S of AZD5991 for 24 h, and cell viability was measured as in a. All the co-culture. Shown here are the top ten most highly-expressed cyto- samples were derived from relapsed or refractory MM patients. c, d kines in the culture media. Signal intensity was normalized to uncul- MM.1S cells were cultured with increasing doses of AZD5991 for 24 h tured media On the other hand, the AZD5991-resistant DOX40 cells Since the bone marrow (BM) microenvironment enhan- showed no increase in Annexin V signals at the low dose ces tumor cell growth and survival in MM, we next assessed (Fig. S1). We next confirmed that the induction of apoptosis the effect of AZD5991 on the MCL1-sensitive MM.1S and in the AZD5991-sensitive cells is caspase-dependent and it H929 cells in the presence of BMSC-conditioned media and is induced primarily via the intrinsic apoptotic pathways as co-culture with bone marrow stromal cells (BMSCs). We shown by an increase in caspase-3 and caspase-9 signals found that soluble factors produced during the MM-BMSC after AZD5991 treatment. The caspase signals were com- interaction reduced the sensitivity of MM cells to pletely reversed with the addition of caspase inhibitors AZD5991, and direct MM-BMSC contact blunted the (Fig. S2). cytotoxic effect of AZD5991 (Fig. 1c, d and Fig. S3). The 2100 Letter Fig. 2 AZD5991 in combination with Venetoclax induces synergistic stacked bar graphs. c MM.1S cells were treated with increasing doses anti-MM activity and overcomes MCL1 resistance in MM. MM.1S of AZD5991 and Venetoclax for 24 h. Cell viability was assessed by cells were either cultured alone (a) or co-cultured with BMSCs (b) and CellTiter-Glo® One Solution Assay. The isobologram analysis con- treated with EC50 doses of AZD5991, Venetoclax (Table S1-S2), or in firms synergism. d The synergistic effect of the AZD5991/Venetoclax combination for 6 h. Protein lysates were prepared from MM.1S cells combination was examined on GFP-expressing MM.1S in the BMSC and subjected to co-immunoprecipitation with MCL1, BCL2, or co-culture setting over 72 h. Cell viability of MM.1S cells was BCLxL antibodies. The pull-down protein complexes were subjected assessed by quantitative fluorescence imaging. The isobologram ana- to Western blot analysis to examine Bim binding to MCL1, BCL2, and lysis confirms greater than additive effect. e The combinatorial effect BCLxL, respectively. Total protein lysates were subjected to Western of the AZD5991/Venetoclax/Dexamethasone regimen was examined blot analysis to determine the amount of protein input for each treat- on GFP-expressing MM.1 S cells in the BMSC co-culture setting over ment. Densitometric analysis of Bim binding was performed using the 72 h. Cell viability of MM.1S cells was assessed by quantitative ImageJ software, and the percentage of Bim bound to MCL1, BCL2, fluorescence imaging and BCLxL under different treatment conditions were presented in the protective effect of the BM microenvironment is mediated dependent. The fact that a higher concentration of via cytokines, as well as through MM-BMSC contact [10]. AZD5991 could overcome the soluble resistant factors in Importantly, MCL1 is the only anti-apoptotic protein within the BM milieu implied cytokines and growth factors only the BCL2 family members whose expression is controlled contribute partially to MCL1 resistance. Direct cell–cell by cytokine treatment of MM cells [11, 12]. Therefore, we contact in the BM microenvironment protects MM cells carried out a comprehensive analysis of 80 human proteins from AZD5991-induced cell death. The co-culture with in cell culture media. Our cytokine array analysis revealed BMSCs increased MCL1, BCLxL, and Bim expression in an enrichment of a panel of pro-survival cytokines and MM.1S cells (Fig. S5), further enhancing their co- growth factors, with the cytokine IL-6 being among the dependence on MCL1 and BCLxL for survival. most highly up-regulated proteins, upon cell–cell contact A shift in the balance of BCL2 family members is often the between MM.1S cells and BMSCs (Fig. 1e and Fig. S4). IL- primary reason for drug resistance [14, 15]. Although MM 6 is known to enhance MCL1 expression via STAT3 cells mostly depend on MCL1 for survival, we hypothesized signaling in MM [11, 13], making the cells more MCL1 that MM cells could switch their survival dependency to other Letter 2101 anti-apoptotic proteins upon stress. For example, when dosage of both AZD5991 and Venetoclax (Fig. 2e and AZD5991 displaces Bim from MCL1, the excess Bim may be Fig. S12), which is critical for clinical translation. sequestered by BCL2 or BCLxL, thereby allowing MM cells We discovered a MCL1 resistance mechanism in MM that to evade cell death. To test this hypothesis, we examined the is driven by Bim binding to other anti-apoptotic proteins upon impact of MCL1 and BCL2 inhibition on the binding pattern MCL1 inhibition. Our data demonstrated that the combined of Bim to anti-apoptotic proteins in MM. MM.1S cells and AZD5991/Venetoclax therapy overcomes MCL1 resistance in patient-derived CD138+ cells were either cultured alone or in MM. Concomitant suppression of both MCL1 and BCL2 co-culture with BMSCs and treated with AZD5991 or Vene- prevent MM cells from escaping apoptosis by releasing Bim toclax (a BH3 mimetic that selectively binds and inhibits from the anti-apoptotic proteins to activate the intrinsic BCL2 [16]) alone or in combination. The protein lysates apoptotic pathway. With the addition of Dex which enhances prepared from the MM cells were then co-immunoprecipitated BCL2 and Bim expression and promotes BCL2 dependence, with antibodies against MCL1, BCL2, and BCLxL to deter- we can achieve therapeutic dosage for both AZD5991 and mine the relative levels of Bim bound to each anti-apoptotic Venetoclax in MM treatment. As a proof of concept, our data protein under each drug treatment condition. We found that indicate combining therapeutics that selectively target the MCL1 inhibition by AZD5991 leads to release of Bim from anti-apoptotic proteins MCL1 and BCL2 could be an effective MCL1, but increased Bim binding to BCL2 and BCLxL. therapy for MM patients, particularly those who suffered from BCL2 inhibition by Venetoclax releases Bim from BCL2 but relapsed or refractory disease. results in increased Bim binding to MCL1. Cotreatment with AZD5991 and Venetoclax decreases the overall Bim bound to Acknowledgements This work is supported by the Multiple Myeloma the anti-apoptotic proteins (Fig. 2a, b and Fig. S6). These Research Fund at the Massachusetts General Hospital. AstraZeneca results suggest that MM cells switch their survival dependency provided the AZD5991 compound. The authors would like to thank to other anti-apoptotic proteins upon MCL1 inhibition and the Flow, Image, and Mass Cytometry Core in the Department of Pathology at the Massachusetts General Hospital for their support in simultaneous inhibition of both MCL1 and BCL2 could be an cell sorting and flow cytometry analysis. effective way to overcome MCL1 resistance in MM. Based on these observations, we combined AZD5991 with Author contributions K.T.S., S.C., and N.R. designed the research; K. Venetoclax for the treatment of MM. A significant decrease in T.S. and C.H. performed the research; K.T.S., C.H., C.P., K.M., K.F., cell viability was observed with the combined therapy com- R.S., C.T., S.C., and N.R. analyzed the data; J.C. and L.D. provided the AZD5991 compound; K.T.S. and N.R. wrote the manuscript; and pared with both drugs used alone (Fig. 2c and Fig. S7). Iso- all the authors reviewed the manuscript. bologram analysis confirmed greater than additive or synergistic effect upon co-treatment. The same in vitro Compliance with ethical standards synergism was observed in AZD5991-resistant DOX40 cells and Venetoclax-resistant ANBL6VR cells (Fig. S8). The Conflict of interest J.C. and L.D. are employees of AstraZeneca. N.R. is enhanced cytotoxic effect of the combined therapy was pre- on advisory boards of Amgen, Bristol-Myers Squibb, Celgene, Merck, served even when the MM cells are in co-culture with BMSCs Janssen and Takeda. N.R. has received research funding from AstraZe- neca. The remaining authors declare that they have no conflict of interest. (Fig. 2d and Fig. S9). No cytotoxic effect was observed when patient-derived BMSCs were exposed to this combined ther- Publisher’s note: Springer Nature remains neutral with regard to apy (Fig. S10). Cotreatment with AZD5991 and Venetoclax jurisdictional claims in published maps and institutional affiliations. also enhanced primary MM cell death in patient-derived bone marrow (Fig. S11), suggesting that this combination regimen References is effective in the BM milieu. Although AZD5991 in combination with Venetoclax is 1. Zhang B, Gojo I, Fenton RG. Myeloid cell factor-1 is a critical effective in inducing synergistic anti-MM activity, the survival factor for multiple myeloma. Blood. 2002;99: 1885–93. concentration of Venetoclax used in the initial testing was 2. Derenne S, Monia B, Dean NM, Taylor JK, Rapp MJ, Harousseau relatively high. A recent report showed that dexamethasone JL, et al. Antisense strategy shows that Mcl-1 rather than Bcl-2 or (Dex) enhances the expression of both BCL2 and Bim in Bcl-x(L) is an essential survival protein of human myeloma cells. MM, and consequently shifts Bim binding towards BCL2 Blood. 2002;100:194–9. 3. Gomez-Bougie P, Maiga S, Tessoulin B, Bourcier J, Bonnet A, and promotes BCL2 dependence in MM [17]. Thus, Dex Rodriguez MS, et al. BH3-mimetic toolkit guides the respective sensitizes MM cells to Venetoclax. To achieve therapeutic use of BCL2 and MCL1 BH3-mimetics in myeloma treatment. concentrations, we added low-dose Dex to the AZD5991/ Blood. 2018;132:2656–69. Venetoclax regimen and found that addition of Dex sig- 4. Gong JN, Khong T, Segal D, Yao Y, Riffkin CD, Garnier JM, et al. Hierarchy for targeting prosurvival BCL2 family proteins in multiple nificantly augments the effect of MCL1 and BCL2 blockade myeloma: pivotal role of MCL1. Blood. 2016;128:1834–44. in MM. We were able to achieve the same cytotoxic effect 5. Wuilleme-Toumi S, Robillard N, Gomez P, Moreau P, Le Gouill on MM cells in the BMSC coculture with a much lower S, Avet-Loiseau H, et al. Mcl-1 is overexpressed in multiple 2102 Letter myeloma and associated with relapse and shorter survival. Leu- 12. Jourdan M, De Vos J, Mechti N, Klein B. Regulation of Bcl-2- kemia. 2005;19:1248–52. family proteins in myeloma cells by three myeloma survival 6. Caenepeel S, Brown SP, Belmontes B, Moody G, Keegan KS, factors: interleukin-6, interferon-alpha and insulin-like growth Chui D, et al. AMG 176, a selective MCL1 inhibitor, is effective factor 1. Cell Death Differ. 2000;7:1244–52. in hematologic cancer models alone and in combination with 13. Puthier D, Derenne S, Barille S, Moreau P, Harousseau JL, established therapies. Cancer Discov. 2018;8:1582–97. Bataille R, et al. Mcl-1 and Bcl-xL are co-regulated by IL-6 in 7. Leverson JD, Zhang H, Chen J, Tahir SK, Phillips DC, Xue J, human myeloma cells. Br J Haematol. 1999;107:392–5. et al. Potent and selective small-molecule MCL-1 inhibitors 14. Luedtke DA, Niu X, Pan Y, Zhao J, Liu S, Edwards H, et al. demonstrate on-target cancer cell killing activity as single agents Inhibition of Mcl-1 enhances cell death induced by the Bcl-2- and in combination with ABT-263 (navitoclax). Cell Death Dis. selective inhibitor ABT-199 in acute myeloid leukemia cells. 2015;6:e1590. Signal Transduct Target Ther. 2017;2:17012. 8. Kotschy A, Szlavik Z, Murray J, Davidson J, Maragno AL, Le 15. Niu X, Zhao J, Ma J, Xie C, Edwards H, Wang G, et al. Binding Toumelin-Braizat G, et al. The MCL1 inhibitor S63845 is toler- of released Bim to Mcl-1 is a mechanism of intrinsic resistance to able and effective in diverse cancer models. Nature. 2016;538: ABT-199 which can be overcome by combination with daunor- 477–82. ubicin or cytarabine in AML cells. Clin Cancer Res. 9. Tron AE, Belmonte MA, Adam A, Aquila BM, Boise LH, 2016;22:4440–51. Chiarparin E, et al. Discovery of Mcl-1-specific inhibitor 16. Souers AJ, Leverson JD, Boghaert ER, Ackler SL, Catron ND, AZD5991 and preclinical activity in multiple myeloma and acute Chen J, et al. ABT-199, a potent and selective BCL-2 inhibitor, myeloid leukemia. Nat Commun. 2018;9:5341. achieves antitumor activity while sparing platelets. Nat Med. 10. Hideshima T, Bergsagel PL, Kuehl WM, Anderson KC. Advances 2013;19:202–8. in biology of multiple myeloma: clinical applications. Blood. 17. Matulis SM, Gupta VA, Nooka AK, Hollen HV, Kaufman JL, 2004;104:607–18. Lonial S, et al. Dexamethasone treatment promotes Bcl-2 depen- 11. Jourdan M, Veyrune JL, De Vos J, Redal N, Couderc G, Klein B. dence in multiple myeloma resulting in sensitivity to venetoclax. A major role for Mcl-1 antiapoptotic protein in the IL-6-induced Leukemia. 2016;30:1086–93. survival of human myeloma cells. Oncogene. 2003;22:2950–9. Leukemia (2019) 33:2102–2104 https://doi.org/10.1038/s41375-019-0437-5 Acute lymphoblastic leukemia Anti-CD19 CAR-T as a feasible and safe treatment against central nervous system leukemia after intrathecal chemotherapy in adults with relapsed or refractory B-ALL Xiaoyuan He1 Xia Xiao2 Qing Li2 Yanyu Jiang2 Yaqing Cao3 Rui Sun3 Xin Jin1 Ting Yuan2 Juanxia Meng2 ● ● ● ● ● ● ● ● ● Li Ma2 Wenyi Lu2 Cuicui Lyu2 Kaiqi Liu4 Mingfeng Zhao2,1 ● ● ● ● Received: 29 November 2018 / Revised: 14 February 2019 / Accepted: 15 February 2019 / Published online: 7 March 2019 © The Author(s) 2019. This article is published with open access To the Editor: Chimeric antigen receptor (CAR)-T cell treatment as an emerging tumor immunotherapy has produced exciting results These authors contributed equally: Mingfeng Zhao, Kaiqi Liu in relapsed or refractory B cell acute lymphoblastic leukemia (B-ALL) [1–4]. However, few data are available on the ther- Supplementary information The online version of this article (https:// apeutic effect of CAR-T against central nervous system leu- doi.org/10.1038/s41375-019-0437-5) contains supplementary material, which is available to authorized users. kemia (CNSL). Here, we evaluated the feasibility and safety of * Kaiqi Liu 2 Department of hematology, Tianjin First Central Hospital, No. 24 yanghua94583@sina.com Fu Kang Road, Tianjin, PR China * Mingfeng Zhao 3 Tianjin medical university, Tianjin, PR China mingfengzhao@sina.com 4 Leukemia Center, Institute of Hematology and Blood Diseases Hospital, Chinese Academy of Medical Sciences and Peking 1 Nankai University School of Medicine, Tianjin, PR China Union Medical College, 288 Nanjing Road, Tianjin, PR China Letter 2103 Fig. 1 Clinical response, expansion and persistence of anti- chemotherapy. d Malignant cells in the bone marrow (BM) of patient 2 CD19chimeric antigen receptor (CAR)-T cells. a–c Cell counts of and patient 3 pre and post-infusion. e The peak expansion levels of CAR-T cells and tumor cells in the cerebrospinal fluid (CSF) before CAR-T cells in the peripheral blood (PB), BM, and CSF and after CAR-T infusion. The black arrows indicate intrathecal anti-CD19 CAR-T against CNSL after intrathecal che- methodological information are described in Supplementary motherapy in three adults with relapsed or refractory B-ALL. Methods, Table S1, and Figure S1. Patient 1 with isolated CNSL, was refractory to high dose First, we assessed the clinical response of CNSL to CAR-T methotrexate plus vindesine and L-asparaginase, and intrathecal therapy in these patients. All patients with CNSL achieved CR chemotherapy, accompanied by bone marrow (BM) sustained approximately one to two weeks post CAR-T infusion remission with minimal residual disease (MRD) negative. (Fig. 1a–c), accompanied by BM remission with MRD Patient 2 initially experienced a CNS relapse, and underwent negative in patient 2 and patient 3 (Fig. 1d). One month after intrathecal chemotherapy, systemic chemotherapy and radio- CR, patient 2 received allogeneic hematopoietic stem cell therapy. However, her CNSL was not controlled, accompanied transplantation. Until the most recent follow-up, her leukemia by BM recurrence. Patient 3 received prophylactic intrathecal free survival has been over 2 months. Interestingly, patient 1 chemotherapy after his first complete remission (CR) but and patient 3 receiving no further therapy for CNSL after experienced a rapid recurrence in his BM and CNS. They were CAR-T infusion were in sustained remission for over enrolled in our anti-CD19 CAR-T clinical trial (ChiCTR-ONN- 5 months (Table S1). A phase 1 dose-escalation trial reported 16009862). Prior to CAR-T cell infusion, all the patients that two B-ALL patients with CNSL achieved CR after CAR- received conditioning chemotherapy with fludarabine and T therapy [3]. Another clinical trial showed that two CNSL cyclophosphamide, and intrathecal chemotherapy to reduce patients at the time of CAR-T infusion subsequently had no blasts in the cerebrospinal fluid (CSF). Detailed patient and blasts in the CSF [2]. Dai and colleagues also reported that two patients with active CNSL at the time of CAR-T infusion 2104 Letter became CNS negative [5]. Altogether, CAR-T can be a fea- (17JCZDJC35800, 15JCQNJC45500; to M. Zhao), and the Tianjin sible and effective treatment for CNSL. Key Science and Technology Program (2015K215, 15KG134, 16KG110; to M. Zhao), as well as Tianjin First Central Hospital. This We further evaluated the proliferation and persistence of work was also supported by the Ph.D. Candidate Research Innovation CAR-T in vivo. Two patients reached the peak expansion of Fund of Nankai University (X. He). CAR-T cells in the CSF on day 8 (Fig. 1b, c), which coincided with the disappearance of CSF blasts in the Author contributions MFZ and KQL designed the research; XYH, XX, QL, YYJ, TY, JXM and LM performed the research; XYH, YQC, responding patients. However, patient 1 exhibited peak RS, XJ, WYL and CCL analyzed the data; XYH wrote the manuscript; expansion on day 28, 2 weeks later than the disappearance XYH and MFZ revised the manuscript. All authors approved the final of the tumor cells in the CSF (Fig. 1a). The persistence time version of the manuscript. of CAR-T cells in the CSF of patient 2 and patient 3 was about 2–3 weeks (Fig. 1b, c), while in patient 1, 5.19% of Compliance with ethical standards CAR-T cells persisted in the CSF on day 90 (Fig. 1a). Ethical approval This study was conducted according to the principles Patient 1 who only had CNS relapse, showed significantly of the Declaration of Helsinki and with the approval of the Ethics higher peak proportions of CAR-T cells in the CSF than Committee of Tianjin First Central Hospital. those in the peripheral blood (PB) and BM. However, patient 2 and patient3 both with BM and CNS recurrence Informed consent All the enrolled patients or their families provided exhibited markedly lower peak levels of CAR-T cells in the written informed consent. CSF compared to patient 1 (Fig. 1e), but relatively high Conflict of interest The authors declare that they have no conflict of peak levels in the PB and BM. These different distributions interest. of CAR-T cells may be explained by chemotaxis and sti- mulated proliferation of effector cells at the tumor sites. Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. We next evaluated the adverse events associated with anti- CD19 CAR-T treatment. Patient 1 only showed grade 1 ane- Open Access This article is licensed under a Creative Commons mia and grade 4 lymphopenia on day 3 after infusion Attribution 4.0 International License, which permits use, sharing, (Table S2). She didn’t complain of any discomfort post CAR- adaptation, distribution and reproduction in any medium or format, as T therapy. It was reported that the severity of the cytokine long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if release syndrome (CRS) was correlated with tumor burdens changes were made. The images or other third party material in this and T cell proliferation [2, 3, 6]. However, the peak expansion article are included in the article’s Creative Commons license, unless time of patient 1 was two weeks later than the disappearance indicated otherwise in a credit line to the material. If material is not of the tumor cells in the CSF (Fig. 1a), which may explain the included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted low risk of CRS. Patient 2 and patient 3 experienced grade 2 use, you will need to obtain permission directly from the copyright fever, grade 3 febrile neutropenia, grade 2 CRS, and grade 1 holder. To view a copy of this license, visit http://creativecommons. reduced consciousness. Patient 2 also had grade 1 cognitive org/licenses/by/4.0/. impairment and grade 2 convulsion. Dexamethasone was administrated at 10 mg q8h to control her seizure on day 8, References and was de-escalated on day 9 and discontinued on day 10. These adverse events were well managed with supportive care 1. Maus MV, Grupp SA, Porter DL, June CH. Antibody-modified and dexamethasone. Other adverse events related to CAR-T T cells: CARs take the front seat for hematologic malignancies. Blood. 2014;123:2625–35. therapy are shown in Figure S2 and Table S2. 2. Maude SL, Frey N, Shaw PA, Aplenc R, Barrett DM, Bunin NJ, In summary, this study showed that anti-CD19 CAR-T et al. Chimeric antigen receptor T cells for sustained remissions in could effectively eliminated leukemia cells in the CNS with leukemia. N Engl J Med. 2014;371:1507–17. fully reversible toxicity. We also found that patient with 3. Lee DW, Kochenderfer JN, Stetler-Stevenson M, Cui YK, Delbrook C, Feldman SA. et al. T cells expressing CD19 chimeric antigen only CNS recurrence had higher levels of CAR-T in the receptors for acute lymphoblastic leukaemia in children and young CSF and relatively lower severity of toxic effects than adults: a phase 1 dose-escalation trial. Lancet. 2015;385:517–28. those with BM and CNS recurrence. This study shows that 4. Xiao X, He X, Li Q, Zhang H, Meng J, Jiang Y, et al. Plasma anti-CD19 CAR-T might be a feasible and safe treatment exchange can be an alternative therapeutic modality for severe cytokine release syndrome after chimeric antigen receptor-t cell for CNSL after intrathecal chemotherapy in adults with B- infusion: a case report. Clin Cancer Res. 2019;25:29–34. ALL, especially in isolated CNSL. More cases and further 5. Dai H, Zhang W, Li X, Han Q, Guo Y, Zhang Y, et al. Tolerance studies are needed to verify these findings. and efficacy of autologous or donor-derived T cells expressing CD19 chimeric antigen receptors in adult B-ALL with extra- Acknowledgements This work was supported by grants from the medullary leukemia. Oncoimmunology. 2015;4:e1027469. National Natural Sciences Foundation of China (81400092; to 6. Nellan A, Lee DW. Paving the road ahead for CD19 CAR T-cell M. Zhao), the Tianjin Key Natural Science Foundation therapy. Curr Opin Hematol. 2015;22:516–20. Letter 2105 Leukemia (2019) 33:2105–2110 https://doi.org/10.1038/s41375-019-0440-x Lymphoma Degradation of Bruton’s tyrosine kinase mutants by PROTACs for potential treatment of ibrutinib-resistant non-Hodgkin lymphomas Yonghui Sun1 Ning Ding2 Yuqin Song2 Zimo Yang1 Wanli Liu3 Jun Zhu2 Yu Rao1 ● ● ● ● ● ● Received: 24 December 2018 / Revised: 31 January 2019 / Accepted: 15 February 2019 / Published online: 11 March 2019 © Springer Nature Limited 2019 To the Editor: in MCL [4]. Unfortunately, there is no optimal therapy for MCL [5] and the 5-year survival rate for MCL patients in Diffuse large B-cell lymphoma (DLBCL) is the most the United States is <50% [6]. These statistics highlight an common type of non-Hodgkin lymphoma (NHL). In the urgent need for the development of more effective treat- United States and United Kingdom, the annual prevalence ments for patients with DLBCL or MCL. of DLBCL is as high as 0.08% [1]. The first-line clinical B-cell receptor (BCR) signaling is indispensable for the treatment for DLBCL is combined chemo-immunotherapy adhesion, survival, and growth of B cells. As an essential with a 5-year survival rate of around 58% for senior adults membrane proximal signal molecule in the BCR pathway, [2]. Mantle cell lymphoma (MCL) contributes to more than Bruton’s tyrosine kinase (BTK) plays a critical role in B- 6–8% of NHL worldwide [3]. Because of the t(11;14)(q13; cell activation and proliferation [7]. In 2013, the BTK q32) chromosomal translocation and upregulated expression covalent inhibitor ibrutinib was approved by the Food and of cyclin D1, malignant proliferation is commonly observed Drug Administration for the treatment of MCL. Addition- ally, activated B-cell-like (ABC)-DLBCL patients achieved remission after treatment with ibrutinib [8]. Unfortunately, These authors contributed equally: Yonghui Sun, Ning Ding, Yuqin drug-resistant tumor cells have been isolated from MCL Song patients during treatment with ibrutinib and the relapse- specific C481S missense BTK mutation contributes to this These authors jointly supervised this work: Wanli Liu, Jun Zhu, Yu Rao resistance [9]. In the growth inhibition of DLBCL, the BTK C481S mutant also resulted in resistance to ibrutinib [10]. Supplementary information The online version of this article (https:// Proteolysis-targeting chimera (PROTAC) is a novel doi.org/10.1038/s41375-019-0440-x) contains supplementary material, which is available to authorized users. strategy for selective knockdown of target proteins by small molecules [11]. PROTAC molecules are hetero- * Wanli Liu bifunctional compounds with three components: a target liuwanli@biomed.tsinghua.edu.cn protein-binding moiety, an E3 ligase ligand, and a linker * Jun Zhu connecting the two (Fig. 1a). The degradation machinery zhu-jun2017@outlook.com brings the target cellular protein to the corresponding E3 * Yu Rao yrao@tsinghua.edu.cn ligase, resulting in degradation by the ubiquitin- proteasome system. In recent years, this newly devel- 1 oped method has been widely used in antitumor studies MOE Key Laboratory of Protein Sciences, School of Pharmaceutical Sciences, MOE Key Laboratory of Bioorganic [12, 13]. Recently, we developed the first PROTAC- Phosphorus Chemistry & Chemical Biology, Tsinghua University, derived degrader for selective BTK degradation [10, 14, Beijing 100084, P.R. China 15]. Although preliminary data showed that our degrader 2 Key laboratory of Carcinogenesis and Translational Research effectively controlled BTK C481S mutant-induced ibru- (Ministry of Education), Department of Lymphoma, Peking tinib-resistant B-cell malignancies in vitro, the following University Cancer Hospital & Institute, No.52 Fucheng Road, critical questions remain unaddressed. (1) Is the BTK Haidian District, Beijing 100142, China 3 degrader also effective for DLBCL in vivo? Due to the MOE Key Laboratory of Protein Sciences, Center for Life poor aqueous solubility of the first generation of our Sciences, Institute for Immunology, School of Life Sciences, Beijing Key Lab for Immunological Research on Chronic degrader, a new generation of BTK degraders with Diseases, Tsinghua University, Beijing 100084, China improved solubility must be developed for in vivo 2106 Letter Letter 2107 Fig. 1 Schematic diagram and characterization of L18I, which pharmacological experiments in vitro and in vivo (Supple- degrades the ibrutinib-resistant BTK C481 mutant. a Mechanism of mentary Table 1). the newly developed BTK C481 mutant degrader. b Western blotting L18I efficiently degraded C481S BTK in HBL-1 cells, analysis for C481S BTK and β-actin proteins after 36 h of treatment of C481S BTK HBL-1 cells with indicated concentrations of L18I (left and the half maximal degradative concentration (DC50) was panel). Western blotting analysis for C481S BTK and β-actin proteins 29 nM (Fig. 1b and Supplementary Fig. S1). PROTAC- after treatment of C481S BTK HBL-1 cells with 100 nM L18I for the mediated degradation of C481S BTK occurred rapidly and indicated exposures (right panel). c Codons for amino acids in six specifically. The half-life of C481S BTK in DLBCL cells different C481 mutated BTK proteins. Catalytic function and degra- dation of C481 mutated BTK proteins were indicated. d Plasmids was <2 h in the presence of L18I (Fig. 1b and Supple- expressing full-length constructs of C481T (C/S) BTK were trans- mentary Figs. S2 and S3). As competitive inhibitors, ibru- fected into HeLa cells using Lipo2000. These C481 mutated BTK tinib and pomalidomide effectively blocked L18I-mediated proteins have stronger catalytic function. Western blotting analysis for degradation of C481S BTK. NEDD8-activating enzyme C481T (C/S) BTK and β-actin proteins after 4 days of treatment of HeLa cells with the indicated concentrations of L18I. e Plasmids inhibitor MLN4924 or proteasome inhibitor bortezomib expressing full-length constructs of C481G (W/A) BTK were trans- also effectively prevented the degradation (Fig. 1a and fected into HeLa cells using Lipo2000. Western blotting analysis for Supplementary Fig. S4). C481G (W/A) BTK and β-actin proteins after 4 days of treatment of Besides the C481S mutation, we investigated other HeLa cells with the indicated concentrations of L18I. f Schematic diagram of BTK degrader design. g Western blotting analysis for p- mutations at the 481 position of the BTK protein. C481T is PLCγ-2, PLCγ-2, p-ERK1/2, ERK1/2, p-p38, p38, and β-actin pro- catalytically active ibrutinib-resistant BTK mutant, and the teins after 36 h of treatment of C481S BTK HBL-1 cells with indicated C481G mutant has very weak catalytic activity [16], while concentrations of L18I the C481A point mutation results in no covalent binding to ibrutinib probe [17]. Therefore, new treatments to tackle evaluation. (2) Can our PROTAC molecules degrade these potential drug-resistant mutants are urgently needed. clinically relevant BTK mutants other than the C481S To confirm the degradation activity of L18I for different mutant? (3) Does our degrader work well for treatment of BTK single-point mutants at the 481 position, six plasmids MCL? (4) Can our degrader achieve synergistic effects expressing BTK C481 mutants were constructed. Except for with other inhibitors for ibrutinib-resistant B-cell alanine and tryptophan mutations, all BTK single-point malignancies? mutants retained catalytic function (Fig. 1c). L18I effi- Here we reported the development of a next-generation ciently degraded all BTK single-point mutants, and the half BTK degrader, L18I, based on the lenalidomide ligand maximal effective concentrations were around 30 nM for cereblon. L18I exhibited good solubility in (Fig. 1d, e). phosphate-buffered saline. Remarkably, L18I induced Next, to evaluate the antitumor effect of L18I in vitro, degradation of ibrutinib-resistant C481S BTK at a the C481S BTK HBL-1 cell line was tested. L18I effec- working concentration as low as 30 nM in a human ABC- tively inhibited the growth of mutant HBL-1 cells with a DLBCL cell line, HBL-1, with exogenous overexpression GI50 of 64 nM. By contrast, the GI50 of ibrutinib was of the BTK C481S mutant. Moreover, L18I achieved 2526 nM due to drug resistance (Fig. 2a). To study the much lower half maximal growth-inhibitory concentra- detailed effects of L18I in C481S BTK HBL-1 cells, we tion (GI50) values than ibrutinib in inhibiting the growth examined the activities of downstream BCR signaling of both DLBCL and MCL cells in vitro. As a further test, molecules upon BTK degradation. In remarkable contrast L18I effectively degraded different C481 mutated BTK to ibrutinib, L18I potently inhibited the phosphorylation proteins that were observed in clinically relevant B-cell of PLCγ-2, ERK1/2, and p38 (Fig. 1g and Supplementary tumors. More importantly, in vivo experiments showed Fig. S8). We also assessed the in vitro antitumor efficacy that L18I induced rapid tumor regression of C481S BTK of L18I in MCL cells (Mino cells and Z138 cells) with HBL-1 xenograft tumors with no signs of toxicity in exogenous expression of the BTK C481S mutant. As mice. Lastly, combined with SYK, PI3K, or lyn inhibitor, expected, L18I efficiently halted cell proliferation at L18I achieved even higher inhibitory activity against much lower concentrations than ibrutinib in C481S BTK C481S BTK HBL-1 cells. MCL cells. GI50s of L18I were all bellow 10 nM, almost To develop an ideal degrader with improved aqueous 300-fold lower than those of ibrutinib (Fig. 2b). These solubility for both in vitro and in vivo evaluations, we results illustrated that ibrutinib was unable to inhibit the optimized both ligand and linker (Fig. 1). After many C481S BTK mutant in vitro, while the newly developed attempts, a suitable degrader, L18I, with ibrutinib and degrader L18I showed excellent inhibitory effects in both lenalidomide ligands connected by a polyethylene gly- DLBCL and MCL BTK C481 mutant-expressing tumor col (PEG) linker was identified. The schematic diagram of cells. BTK degrader design was shown in Fig. 1f. This break- Based on these potent in vitro efficacies, L18I was fur- through provided an opportunity to carry out ther examined in a mouse xenograft model inoculated with 2108 Letter Letter 2109 Fig. 2 Pharmacodynamic and safety evaluation of L18I. a In vitro GS-9973, a SYK inhibitor, has been used in potential study of growth inhibition effect on C481S BTK HBL-1 cells by L18I overcoming ibrutinib-resistant clones [18]. PI3K upregu- and ibrutinib. In 96-well plates, 3000 cells were incubated in each well lation has been documented in MCL with ibrutinib at 37 °C for 72–96 h. The final GI50 was calculated using CCK-8. b In vitro study of growth inhibition effect on MCL (Mino or Z138) BTK resistance [9]. In addition, ibrutinib-resistant cells retain C481S mutant overexpression cells by L18I and ibrutinib. In 96-well sensitivity to lyn inhibitors, such as dasatinib [19]. Thus, plates, 3000 cells were incubated in each well at 37 °C for 72–96 h. it is of interest to examine the synergistic effect of L18I The final results were calculated using MTT. c L18I induced tumor with SYK, PI3K, or lyn inhibitors. We performed these regression in the xenograft model of human C481S BTK HBL-1 tumor cells with minimal body weight change compared with ibrutinib- experiments with C481S BTK HBL-1 cells. Remarkably, treated control. Tumor volume (means ± SEM) of vehicle-treated mice in combination with GS-9973, L18I was more effective in (n = 7), mice treated with ibrutinib (30 mg/kg, n = 7; 100 mg/kg, n = inhibiting C481S BTK HBL-1 cells, and the GI50 was as 7), or mice treated with L18I (30 mg/kg, n = 7; 100 mg/kg, n = 7). low as 11 nM. Compared with the treatment using L18I d Relative C481S BTK protein levels based on immunoblotting of tumor lysates from mice treated with L18I (30 mg/kg, n = 7; alone, a combination of L18I and a PI3K inhibitor, 100 mg/kg, n = 7) and vehicle-treated mice (n = 7). e Immunohis- copanlisib, dramatically improved the inhibitory effect on tochemistry results of tumor biopsies from mice treated with L18I C481S BTK HBL-1 cells, and the GI50 was lower than (30 and 100 mg/kg) and vehicle-treated mice. Black bars represent 5 nM. Additionally, a combination of L18I and dasatinib 100 μM. White bars represent 20 μM. f In vivo acute toxicity experi- ment results from mice treated with L18I (250 and 300 mg/kg). effectively inhibited the growth of ibrutinib-resistant g DLBCL drug combination therapies. In vitro study of growth inhi- HBL-1 cells with a GI50 under 5 nM (Fig. 2g and Sup- bition effect on C481S BTK HBL-1 cells by L18I combined with GS- plementary Fig. S9). 9973 or copanlisib. In 96-well plates, 3000 cells were incubated in In summary, we present a newly designed C481 BTK each well at 37 °C for 72–96 h. The final GI50 was calculated using CCK-8 mutant-specific degrader L18I. With lenalidomide as the cer- eblon E3 ubiquitin ligase ligand, the BTK degrader was considerably more soluble. A series of clinically relevant BTK mutants (C481S/T/G/W/A) were efficiently degraded by L18I. C481S BTK HBL-1 cells. Rapid tumor regression was Proliferation of BTK C481S mutant-expressing DLBCL and achieved with 30 or 100 mg/kg of L18I administered MCL cells was potently inhibited in vitro by compound L18I, intraperitoneally every day for 2 weeks. Ibrutinib (30 and whereas BTK covalent inhibitor ibrutinib exhibited very weak 100 mg/kg) was used as a negative control. Compared with efficacy. Meanwhile, downstream signaling activities were the vehicle group, tumor size of the 30 mg/kg L18I-treated strongly blocked by L18I, with effective concentration below group was reduced by 36% (p = 0.0126). In the 100 mg/kg 10 nM. More importantly, pharmacologic action of L18I L18I-treated group, tumor size was reduced by 63% (p = in vivo also resulted in an obvious antitumor effect in mouse 0.0004). By contrast, in the ibrutinib treatment groups, xenograft models inoculated with C481S BTK HBL-1 cells. almost no reduction in tumor size was observed at low or By contrast, ibrutinib had nearly no inhibitory activity in vivo, high doses (Fig. 2c). Relative C481S BTK protein levels in which was consistent with the results from the in vitro fresh ex vivo tumor cells were also tested by immunoblot- experiments. What’s more, compound L18I was found to be ting and immunohistochemistry. C481S BTK levels in safe and well-tolerated. Lastly, combined with GS-9973, isolated tumors from L18I-treated mice were significantly copanlisib, or dasatinib, L18I showed dramatically improved lower than those in the vehicle group (Fig. 2d, e). From the efficacy in inhibiting ibrutinib-resistant cells, indicating safety evaluation, it was clear that mice treated with ibru- potential for combinations of inhibitor and degrader in the tinib suffered from more severe body weight loss than mice treatment of drug-resistant cancers. treated with L18I. To further confirm the safety of L18I, acute toxicity experiments were conducted in the B6 mouse Acknowledgements This work was supported by National Natural Science Foundation of China (#81573277, 81622042, and 81773567), model. All mice in the L18I-treated groups (250 and 300 National Major Scientific and Technological Special Project for mg/kg) were in good health during the 3 weeks after treat- “Significant New Drugs Development” (#SQ2017ZX095003), and ment (Fig. 2f). Therefore, the newly developed BTK C481 Tsinghua University Initiative Scientific Research Program. This work mutant degrader L18I was well-tolerated and had a highly was also financially supported by NSFC (No. 81670187) and Beijing Natural Science Foundation (No. 7172047). efficacious antitumor effect in this mouse xenograft model. Drug combination strategies with small-molecule inhibitors are commonly used to treat refractory cancers. Compliance with ethical standards In the current study, we for the first time tested the pos- Conflict of interest The authors declare that they have no conflict of sibility of a drug combination therapy strategy with both interest. BTK degrader and BCR signaling molecule inhibitors. The inhibitory effect against ibrutinib-resistant cell lines Publisher’s note: Springer Nature remains neutral with regard to was tested using several different targets. For instance, jurisdictional claims in published maps and institutional affiliations. 2110 Letter References mutated BTK C481S induced ibrutinib-resistant B-cell malig- nancies. Cell Res. 2018;28:779–81. 1. Morton LM, Wang SS, Devesa SS, Hartge P, Weisenburger DD, 11. Sakamoto KM, Kim KB, Kumagai A, Mercurio F, Crews CM, Linet MS. Lymphoma incidence patterns by WHO subtype in the Deshaies RJ. Protacs: chimeric molecules that target proteins to United States, 1992-2001. Blood. 2006;107:265–76. the Skp1-Cullin-F box complex for ubiquitination and degrada- 2. Feugier P, van Hoof A, Sebban C, Solal-Celigny P, Bouabdallah tion. Proc Natl Acad Sci USA. 2001;98:8554–9. R, Fermé C, et al. Long-term results of the R-CHOP study in the 12. Toure M, Crews CM. Small-molecule PROTACS: new approa- treatment of elderly patients with diffuse large B-cell lymphoma: a ches to protein degradation. Angew Chem Int Ed. 2016;55: study by the groupe d'Etude des lymphomes de l’adulte. J Clin 1966–73. Oncol. 2005;23:4117–26. 13. Dobrovolsky D, Wang ES, Morrow S, Leahy C, Faust T, Nowak 3. Pérez-Galán P, Dreyling M, Wiestner A. Mantle cell lymphoma: RP, et al. Bruton’s tyrosine kinase degradation as a therapeutic biology, pathogenesis, and the molecular basis of treatment in the strategy for cancer. Blood. 2019;133:952–61. genomic era. Blood. 2011;117:26–38. 14. Buhimschi AD, Armstrong HA, Toure M, Jaime-Figueroa S, 4. Li JY, Gaillard F, Moreau A, Harousseau JL, Laboisse C, Milpied Chen TL, Lehman AM, et al. Targeting the C481S ibrutinib- N, et al. Detection of translocation t(11;14)(q13; q32) in mantle resistance mutation in Bruton’s tyrosine kinase using PROTAC- cell lymphoma by fluorescence in situ hybridization. Am J Pathol. mediated degradation. Biochemistry. 2018;57:3564–75. 1999;154:1449–52. 15. Zorba A, Nguyen C, Xu Y, Starr J, Borzilleri K, Smith J, et al. 5. Rajabi B, Sweetenham JW. Mantle cell lymphoma: observation to Delineating the role of cooperativity in the design of potent PRO- transplantation. Ther Adv Hematol. 2015;6:37–48. TACs for BTK. Proc Natl Acad Sci USA. 2018;115:E7285–92. 6. Herrmann A, Hoster E, Zwingers T, Brittinger G, Engelhard M, 16. Hamasy A, Wang Q, Blomberg KEM, Mohammad DK, Yu L, Meusers P, et al. Improvement of overall survival in advanced Vihinen M, et al. Substitution scanning identifies a novel, cata- stage mantle cell lymphoma. J Clin Oncol. 2009;27:511–8. lytically active ibrutinib-resistant BTK cysteine 481 to threonine 7. Wang ML, Rule S, Martin P, Goy A, Auer R, Kahl BS, et al. (C481T) variant. Leukemia. 2017;31:177–85. Targeting BTK with ibrutinib in relapsed or refractory mantle-cell 17. Furman RR, Cheng S, Lu P, Setty M, Perez AR, Guo A, et al. lymphoma. N Engl J Med. 2013;369:507–16. Ibrutinib resistance in chronic lymphocytic leukemia. N Engl J 8. Wilson WH, Young RM, Schmitz R, Yang Y, Pittaluga S, Wright Med. 2014;370:2352–4. G, et al. Targeting B cell receptor signaling with ibrutinib in 18. Liu TM, Woyach JA, Zhong Y, Lozanski A, Lozanski G, Dong S, diffuse large B cell lymphoma. Nat Med. 2015;21:922–6. et al. Hypermorphic mutation of phospholipase C, γ2 acquired in 9. Chiron D, Di Liberto M, Martin P, Huang X, Sharman J, Blecua ibrutinib-resistant CLL confers BTK independency upon B-cell P, et al. Cell-cycle reprogramming for PI3K inhibition overrides a receptor activation. Blood. 2015;126:61–8. relapse-specific C481S BTK mutation revealed by longitudinal 19. Cheng S, Guo A, Lu P, Ma J, Coleman M, Wang YL. Functional functional genomics in mantle cell lymphoma. Cancer Discov. characterization of BTKC481S mutation that confers ibrutinib 2014;4:1022–35. resistance: exploration of alternative kinase inhibitors. Leukemia. 10. Sun YH, Zhao XW, Ding N, Gao HY, Wu Y, Yang YQ, et al. 2015;29:895–900. PROTAC-induced BTK degradation as a novel therapy for Letter 2111 Leukemia (2019) 33:2111–2115 https://doi.org/10.1038/s41375-019-0444-6 Chronic lymphocytic leukemia KRAS, NRAS, and BRAF mutations are highly enriched in trisomy 12 chronic lymphocytic leukemia and are associated with shorter treatment-free survival Elena Vendramini1 Riccardo Bomben1 Federico Pozzo1 Dania Benedetti1 Tamara Bittolo1 ● ● ● ● ● Francesca Maria Rossi1 Michele Dal Bo1 Kari G. Rabe2 Gabriele Pozzato3 Francesco Zaja3 Annalisa Chiarenza4 ● ● ● ● ● ● Francesco Di Raimondo4 Esteban Braggio5 Sameer A. Parikh5 Neil E. Kay5 Tait D. Shanafelt6 ● ● ● ● ● Giovanni Del Poeta7 Valter Gattei1 Antonella Zucchetto1 ● ● Received: 4 March 2019 / Accepted: 6 March 2019 / Published online: 14 March 2019 © The Author(s) 2019. This article is published with open access To the Editor: members of the Ras–MAPK pathway in CLL cases with specific clinico-biological features [6, 7], including the KRAS mutations are among the most common oncogenic presence of trisomy 12, a cytogenetic aberration associated events in human carcinomas of endodermal origin, whose with a unique pathophysiology among CLL [8, 9], and/or an presence predicts for resistance to several target therapies unmutated (UM) configuration of IGHV genes, although a [1]. Conversely, little is known regarding the role and/or dedicated and comprehensive analysis of these aspects is clinical impact of KRAS mutations in the setting of the still missing. hematological malignancies, including chronic lymphocytic This study, approved by the IRB of the Aviano Centro di leukemia (CLL), and only in recent years extensive Riferimento Oncologico (Approvals n. IRB-05-2010 and n. sequencing data have highlighted the recurrent mutations of IRB-05-2015), included 534 primary CLL from treatment- genes affecting the Ras–MAPK pathway in CLL [2, 3]. naive patients. The cohort was purposely enriched in tris- These mutations, by leading to a constitutive activation of omy 12 CLL by including 110 cases from the Mayo Clinic, MAPK signaling pathway, have emerged as relevant in Rochester, MN [8] to better evaluate the incidence of driving impaired clinical responses to lenalidomide and mutations of the Ras–MAPK pathway in these subsets. chlorambucil, and acquired resistance to fludarabine as well Overall, out of 534 cases, trisomy 12 CLL accounted for as to PI3K and BCL2 inhibitors [4–6]. In this context, some 300 cases (190 with trisomy 12 as the sole abnormality studies pinpointed a higher frequency of mutations in [trisomy 12-only], and 110 with trisomy 12 plus another abnormality on FISH [trisomy 12-plus]), 332 cases had UM IGHV genes, and 214 cases had NOTCH1 aberrations (details in Table S1). CLL patients were diagnosed and These authors contributed equally: Valter Gattei, Antonella Zucchetto treated according to the current iwCLL 2018 guidelines Supplementary information The online version of this article (https:// [10], and all samples were collected at diagnosis from doi.org/10.1038/s41375-019-0444-6) contains supplementary treatment-naive patients. In 442/534 cases (clinical cohort), material, which is available to authorized users. * Valter Gattei 3 Department of Internal Medicine and Haematology, Maggiore vgattei@cro.it General Hospital, University of Trieste, Trieste, Italy * Antonella Zucchetto 4 Division of Hematology, Ferrarotto Hospital, Catania, Italy zucchetto.soecs@cro.it 5 Division of Hematology, Department of Medicine, Mayo Clinic, Rochester, MN, USA 1 Clinical and Experimental Onco-Hematology Unit, Centro di 6 Riferimento Oncologico di Aviano (CRO) IRCCS, Aviano (PN), Department of Hematology/Oncology, Stanford University, Italy Stanford, CA, USA 7 2 Division of Biomedical Statistics and Informatics, Department of Division of Hematology, S. Eugenio Hospital and University of Health Sciences Research, Mayo Clinic, Rochester, MN, USA Tor Vergata, Rome, Italy 2112 Letter treatment-free survival (TFS) data were available along with frequency was found in CLL patients with concomitant UM a comprehensive clinical and biological characterization IGHV and trisomy 12-only (38/133, 28.6%). This group (Table S1 and Supplemental Methods). This cohort showed was characterized by 25 KRAS (18.8%), 8 NRAS (6%) and the expected clinical behavior according to both the strati- 11 BRAF (8.3%) mutated cases. Of note, the UM IGHV/ fication of the established cytogenetic categories and to the trisomy 12-only group was characterized by a higher fre- canonical prognosticators by univariable and multivariable quency of KRAS/NRAS/BRAF mutations also when com- analyses (Supplemental Figure S1 and Table S2). Mutation pared to the UM IGHV/trisomy 12-plus group (8/49, testing for KRAS, NRAS, BRAF, TP53, NOTCH1, BIRC3, 16.3%). Finally, the lowest frequency of KRAS/NRAS/ and SF3B1 was performed on DNA from CD19+ enriched BRAF mutations was observed in the context of CLL CLL samples by Next Generation Sequencing (NGS) assays patients with M IGHV (8/186, 4.3%) and del13q as the sole with at least 1000 × coverage and 1% sensitivity (details in chromosomal aberration (2/94, 2.1% in the whole cohort, Supplemental Methods). Groups were compared by chi- and 2/53, 3.8% in the context of UM IGHV cases). square test; TFS was computed from diagnosis to treatment We then correlated the presence of KRAS/NRAS/BRAF and analyzed by log-rank test and Cox regression analysis mutations to other biological features (Table S4). When with a stepwise procedure using MedCalc Statistical Soft- considering the whole CLL cohort, the only variables ware version 16.8.4 (MedCalc Software bvba, Ostend, associated with a higher frequency of KRAS/NRAS/BRAF Belgium; https://www.medcalc.org; 2016). mutations were the absence of BIRC3 mutations (p = 0.02) The mutation analysis of the Ras–MAPK pathway was and the positive expression (≥30%) of CD49d (p = 0.04). focused on the KRAS, NRAS, and BRAF genes, previously On the other hand, if circumscribing the analysis to UM reported as the most frequently mutated genes among the IGHV/trisomy 12 CLL, CD49d positive expression lost its members of the pathway [2]. We found 91 missense point association with KRAS/NRAS/BRAF mutations, as expected mutations in 64 CLL cases, with a prevalence of KRAS due to the almost universal CD49d expression in trisomy (44 mutations in 38 [7.1%] patients), followed by BRAF 12 CLL patients [9]. Conversely, we observed a higher (32 mutations in 24 [4.5%] patients) and NRAS (15 muta- frequency of KRAS/NRAS/BRAF mutations in NOTCH1 tions in 13 [2.4%] patients). Nearly all mutations were wild type cases (29/92, 31.5%) and BIRC3 wild type cases previously associated with the gain-of-function phenotype (41/132, 31.1%) compared to their mutated counterparts and increased RAS/ERK downstream signaling (Fig. 1a and (NOTCH1 mutated: 17/90, 18.9%; BIRC3 mutated: 4/30, Table S3) [1]. In particular, among the most frequent KRAS/ 13.3%; p = 0.05 in both cases), pointing to a mutual NRAS mutations, almost half of the mutations (27/59, 45%), exclusivity of these mutations in the pathogenesis of the overall affecting 23/49 (47%) patients, involved the G12/ disease. No other significant associations with other known G13 codons, in keeping with what was observed in colon prognostic variables such as presence of TP53 mutations/ and lung cancers (Table S3) [1]. The co-occurrence of 2 disruption, SF3B1 mutations, ZAP-70 and CD38 expres- mutated genes was observed in 11 cases (KRAS and BRAF sion, Rai staging, age at diagnosis, and gender were in 8/11 cases, KRAS and NRAS in 2/11 cases, NRAS and observed either in the whole cohort or in the UM IGHV/ BRAF in 1/11 cases), whereas mutations affecting all three trisomy 12 cohort (Table S4). genes were not found in our cohort. The mutations were We finally evaluated the prognostic relevance of mainly subclonal (mean Variant Allele Fraction, VAF, KRAS, NRAS, and BRAF mutations as predictors of TFS. 12.3%, range 1.3–61.6%) with one-third of mutations (33/ In the context of the clinical cohort, the presence of either 91) above 10% VAF. The presence of multiple mutations KRAS or NRAS mutations or the concomitant presence of affecting the same gene occurred in 14 cases, including 5 KRAS/NRAS mutations were associated with shorter TFS cases that presented mutations in the same or adjacent (p = 0.07, p = 0.05, and p = 0.02, respectively) (Table 1, codons (i.e., one case with both K601N and K601E BRAF Fig. 1c). Conversely, BRAF mutations were not asso- mutations, one case with V600E and K601E BRAF muta- ciated with TFS, pointing to a secondary role of BRAF in tions, and three cases with two simultaneous KRAS muta- the Ras–MAPK pathway in CLL, in line with studies tions at the G12 and G13 codons) suggesting that multiple indicating the lack of therapeutic effects of BRAF inhi- genetic hits are positively selected in different subclones bition in CLL [11]. In a multivariable model that included within the same leukemia specimen. the main known CLL prognosticators, the presence of A strong association between KRAS/NRAS/BRAF muta- KRAS/NRAS mutations retained its independent prog- tions and the presence of an UM IGHV gene status and nostic power as predictor for shorter TFS (p = 0.03, trisomy 12 was observed (Fig. 1b and Table S4). Overall, Table 1). Moreover, circumscribing the analysis to the 87.3% of KRAS/NRAS/BRAF mutated cases had UM IGHV CLL subgroup with the highest incidence of these (p < 0.0001) and 79.7% were trisomy 12 CLL (p < 0.0001). mutations (i.e., UM IGHV/trisomy 12-only/NOTCH1- Concordantly, the highest KRAS/NRAS/BRAF mutation wt), both KRAS mutations alone (p = 0.005) and KRAS/ Letter 2113 a KRAS 11 G13D G12D/V/R # Mutations A146V/P/T Q22K/E L19F Q61H/L/R A18V T58I V14I K117R 0 Ras Sequenced 0 100 189 aa amplicons (aa) 7 24 57 68 110 128 141 150 NRAS 8 Q61R/K/L # Mutations G13D/V G12D/C A146T 0 Ras Sequenced 0 100 189 aa amplicons (aa) 9 24 56 69 115 131 145 150 BRAF 11 K601E/N # Mutations G469A/E/R/V D594G L597Q/R G466E V600E F468S 0 RBD C1_1 Pkinase_Tyr Sequenced 0 200 400 600 766 aa amplicons (aa) 451 478 590 608 b trisomy 12-only trisomy 12-plus trisomy 12 KRAS NRAS BRAF IGHV NOTCH1 -Trisomy 12-only -non-trisomy 12 -mutated NOTCH1 -unmutated NOTCH1 ND -unmutated IGHV -mutated IGHV c whole cohort d UM IGHV/trisomy 12-only/NOTCH1-wt 100 100 80 80 unmutated unmutated KRAS/NRAS KRAS 60 (n=400) 60 (n=45) TFS (%) TFS (%) mutated mutated 40 40 KRAS/NRAS KRAS (n=42) (n=16) 20 20 p=0.02 p=0.004 0 0 0 50 100 150 200 250 0 50 100 150 200 250 Time (months) Time (months) Fig. 1 Type, incidence and prognostic impact of KRAS, NRAS and tyrosine kinase domain. b Co-mutation plot of 534 CLL analyzed for BRAF mutations. a Lollipop plots of mutations found in KRAS, NRAS, KRAS, NRAS, BRAF mutations. Incidence of trisomy 12, KRAS, NRAS, and BRAF genes. Sites and frequency of missense point mutations, and and BRAF missense mutations, NOTCH1 aberrations and IGHV status schematic presentation of the protein structure and functional domains are shown. c Kaplan–Meier curves of treatment-free survival (TFS) of are shown (MutationMapper, cBioPortal Version 1.14.0, Gao et al. 442 CLL patients stratified by the presence of KRAS and/or NRAS Sci. Signal. 2013 and Cerami et al. Cancer Discov. 2012). Gray boxes mutations. d Kaplan–Meier curves of TFS of 61 CLL patients in the indicate amino acids (aa) regions corresponding to the sequenced IGHV unmutated/trisomy 12-only/NOTCH1-wt group stratified by the amplicons. RBD, Ras-binding domain; C1_1, phorbol esters/dia- presence of KRAS mutations cylglycerol binding domain (C1 domain); Pkinase_Tyr, protein 2114 Letter Table 1 Cox regression analysis Univariable Multivariable (n = 365) of treatment-free survival in the whole cohort N pts HR (95% CI) p-value HR (95% CI) p-value analyzed male gender 442 0.97 (0.76–1.22) 0.8 − − age ≥ 65 441 0.97 (0.77–1.23) 0.8 − − Rai stage II–III–IV 434 2.60 (2.03–3.32) <0.0001 2.59 (1.98–3.41) <0.0001 CD49d positive (≥30%) 440 1.48 (1.13–1.92) 0.004 1.65 (1.22–2.23) 0.001 CD38 positive (≥30%) 439 1.13 (0.89–1.43) 0.3 − − ZAP-70 positive (≥20%) 389 1.44 (1.11–1.87) 0.005 n.i. n.i. IGHV unmutated 428 2.16 (1.66–2.8) <0.0001 1.83 (1.37–2.45) <0.0001 TP53 disrupted (del17p and/or 442 1.70 (1.24–2.32) 0.0008 1.46 (1.05–2.03) 0.024 TP53 mutated) NOTCH1 mutated 442 1.37 (1.08–1.73) 0.009 n.i. n.i. SF3B1 mutated 327 1.46 (0.90–2.38) 0.1 − − BIRC3 mutated 343 0.93 (0.65–1.34) 0.7 − − KRAS mutated 442 1.49 (0.96–2.30) 0.073 − − NRAS mutated 442 1.86 (0.99–3.50) 0.055 − − BRAF mutated 442 1.34 (0.75–2.40) 0.3 − − KRAS/NRAS mutated 442 1.54 (1.05–2.25) 0.025 1.56 (1.04–2.36) 0.033 Factors with p-value <0.05 in univariable analysis were entered in the multivariable analysis HR hazard ratio, CI confidence interval, n.i. variables not included in the model after stepwise selection NRAS mutations (p = 0.05) were associated with shorter BCL2 axis, characterizes a CLL subset especially orien- TFS (Fig. 1d and Table S5), and the presence of KRAS ted toward the amplification of anti-apoptotic signals mutations retained its prognostic value in a multivariable [14]. On the other hand, in trisomy 12 CLL, the co- analysis that included all the variables with an impact in presence of KRAS/NRAS/BRAF mutations and/or univariable analysis (p = 0.01, Table S5). The subclonal NOTCH1 mutations and/or BIRC3 mutations along with a or clonal pattern of KRAS/NRAS mutations had similar UM IGHV gene status and over-expression of surface negative impact in our series (not shown), as previously receptors mediating microenvironment interactions (e.g., observed for other gene mutations in CLL [12], and in CD49d) more likely characterizes CLL with amplified keeping with the known capability of KRAS mutated pro-survival and proliferative signals [8, 9, 15]. This may tumor cells to enhance the overall tumor cell fitness by explain the clinical association between KRAS/NRAS influencing the non-mutated neoplastic component [13]. mutations and shorter TFS, as shown in the present In the present study, we demonstrated that KRAS, analysis. NRAS, and BRAF mutations were almost exclusively Given the reported high risk of poor response and found in UM IGHV/trisomy 12 CLL and were almost development of chemo-resistance characterizing CLL cases mutually exclusive with NOTCH1 and BIRC3 mutations. with KRAS/NRAS mutations [4–6], additional therapeutic The type of genomic structural variants, especially tris- strategies should be considered for the treatment of these omy 12 and del13q, strongly influenced KRAS/NRAS/ cases, including MEK/ERK inhibitors, employed alone or BRAF mutation incidence, that turned out to be at the in combination with conventional therapies. highest level in cases bearing trisomy 12 as the sole genomic aberration, intermediate in cases in which tris- Acknowledgements The study was supported by the Fondazione Umberto Veronesi, Post-doctoral Fellowships-year 2018 (to EV); omy 12 was associated with other genetic aberrations, Associazione Italiana Ricerca Cancro (AIRC), Investigator Grant IG- mainly del13q, and at the lowest level in cases bearing 21687; Progetto Giovani Ricercatori no. GR-2011-02346826, no. GR- del13q as the sole FISH detectable genetic aberration. 2011-02347441, and no.GR-2011-02351370, Ministero della Salute, This peculiar distribution of KRAS/NRAS/BRAF mutation Rome, Italy; Progetto Ricerca Finalizzata PE 2016-02362756, Minis- tero della Salute, Rome, Italy; Associazione Italiana contro le Leuce- incidence is in keeping with a CLL pathogenetic model in mie, linfomi e mielomi (AIL), Venezia Section, Pramaggiore Group, which the two main founder genetic lesions (i.e., trisomy Italy; Linfo-check - Bando ricerca - contributo art. 15, comma 2, lett b) 12 and del13q) identify CLL subgroups following dif- LR 17/2014; “5x1000 Intramural Program”, Centro di Riferimento ferent patho-biological pathways. In particular, the pre- Oncologico, Aviano, Italy; National Cancer Institute, CA197120 (to TDS and NEK). Authors wish to thank Gustavo Baldassarre (Division sence of del13q, given its link to the miR15/miR16- Letter 2115 of Molecular Oncology, Department of Translational Research, CRO mechanisms in chronic lymphocytic leukemia. Clin Cancer Res. Aviano, Italy) for helpful discussion. 2016;22:6217–27. 6. Takahashi K, Hu B, Wang F, Yan Y, Kim E, Vitale C, et al. Clinical implications of cancer gene mutations in patients with Compliance with ethical standards chronic lymphocytic leukemia treated with lenalidomide. Blood. 2018;131:1820–32. Conflict of interest The authors declare that they have no conflict of 7. Gimenez N, Martinez-Trillos A, Montraveta A, Lopez-Guerra interest. M, Rosich L, Nadeu F, et al. Mutations in RAS-BRAF-MAPK- ERK pathway define a specific subgroup of patients with Publisher’s note: Springer Nature remains neutral with regard to adverse clinical features and provide new therapeutic options in jurisdictional claims in published maps and institutional affiliations. chronic lymphocytic leukemia. Haematologica. 2018;104: 576–86. Open Access This article is licensed under a Creative Commons 8. Bulian P, Bomben R, Bo MD, Zucchetto A, Rossi FM, Degan M, Attribution 4.0 International License, which permits use, sharing, et al. Mutational status of IGHV is the most reliable prognostic adaptation, distribution and reproduction in any medium or format, as marker in trisomy 12 chronic lymphocytic leukemia. Haemato- long as you give appropriate credit to the original author(s) and the logica. 2017;102:e443–e446. source, provide a link to the Creative Commons license, and indicate if 9. Zucchetto A, Caldana C, Benedetti D, Tissino E, Rossi FM, changes were made. The images or other third party material in this Hutterer E, et al. CD49d is overexpressed by trisomy 12 chronic article are included in the article’s Creative Commons license, unless lymphocytic leukemia cells: evidence for a methylation-dependent indicated otherwise in a credit line to the material. If material is not regulation mechanism. Blood. 2013;122:3317–21. included in the article’s Creative Commons license and your intended 10. Hallek M, Cheson BD, Catovsky D, Caligaris-Cappio F, Dighiero use is not permitted by statutory regulation or exceeds the permitted G, Dohner H, et al. iwCLL guidelines for diagnosis, indications use, you will need to obtain permission directly from the copyright for treatment, response assessment, and supportive management holder. To view a copy of this license, visit http://creativecommons. of CLL. Blood. 2018;131:2745–60. org/licenses/by/4.0/. 11. Yaktapour N, Meiss F, Mastroianni J, Zenz T, Andrlova H, Mathew NR, et al. BRAF inhibitor-associated ERK activation drives development of chronic lymphocytic leukemia. J Clin References Invest. 2014;124:5074–84. 12. Nadeu F, Clot G, Delgado J, Martin-Garcia D, Baumann T, Sal- 1. Simanshu DK, Nissley DV, McCormick F. RAS proteins and their averria I, et al. Clinical impact of the subclonal architecture and regulators in human disease. Cell. 2017;170:17–33. mutational complexity in chronic lymphocytic leukemia. Leuke- 2. Landau DA, Tausch E, Taylor-Weiner AN, Stewart C, Reiter JG, mia. 2018;32:645–53. Bahlo J, et al. Mutations driving CLL and their evolution in 13. Grabocka E, Bar-Sagi D. Mutant KRAS enhances tumor cell fit- progression and relapse. Nature. 2015;526:525–30. ness by upregulating stress granules. Cell. 2016;167:1803–13. 3. Puente XS, Bea S, Valdes-Mas R, Villamor N, Gutierrez-Abril J, 14. Cimmino A, Calin GA, Fabbri M, Iorio MV, Ferracin M, Shimizu Martin-Subero JI, et al. Non-coding recurrent mutations in chronic M, et al. miR-15 and miR-16 induce apoptosis by targeting BCL2. lymphocytic leukaemia. Nature. 2015;526:519–24. Proc Natl Acad Sci USA. 2005;102:13944–9. 4. Herling CD, Abedpour N, Weiss J, Schmitt A, Jachimowicz RD, 15. Benedetti D, Tissino E, Pozzo F, Bittolo T, Caldana C, Perini C, Merkel O, et al. Clonal dynamics towards the development of et al. NOTCH1 mutations are associated with high CD49d venetoclax resistance in chronic lymphocytic leukemia. Nat expression in chronic lymphocytic leukemia: link between the Commun. 2018;9:727. NOTCH1 and the NF-kappaB pathways. Leukemia. 2018;32: 5. Pandzic T, Larsson J, He L, Kundu S, Ban K, Akhtar-Ali M, et al. 654–62. Transposon mutagenesis reveals fludarabine resistance 2116 Letter Leukemia (2019) 33:2116–2120 https://doi.org/10.1038/s41375-019-0443-7 Chronic myelogenous leukemia Skeletal muscle toxicity associated with tyrosine kinase inhibitor therapy in patients with chronic myeloid leukemia L. Janssen1,2 S. J. C. M. Frambach3,4 N. A. E. Allard1 M. T. E. Hopman1 T. J. J. Schirris3,4 N. C. Voermans5 ● ● ● ● ● ● R. J. Rodenburg4,6 N. M. A. Blijlevens2 S. Timmers1,7 ● ● Received: 3 December 2018 / Revised: 21 February 2019 / Accepted: 22 February 2019 / Published online: 14 March 2019 © The Author(s) 2019. This article is published with open access To the Editor: the Netherlands and registered at The Netherlands Trial Registry (NTR6373). Up to 80% of CML patients using tyrosine kinase inhibitors A total of twenty Ph+ CP-CML patients on TKI therapy (TKIs) reports muscle complaints [1]. These muscle com- aged ≥ 18 were recruited from the Department of Hematol- plaints are strongly related to the presence of fatigue and ogy at the Radboud University Medical Center (Nijmegen, contributes to both diminished disease control [2] and The Netherlands). CML patients were assigned to a group impaired quality of life [3]. Although the mechanism by with (CML + MC, N = 10) or without (CML - MC, N = 10) which TKIs cause muscle complaints is poorly understood, muscle complaints (MC) on the basis of presence, onset and mitochondrial dysfunction has been suggested to play a course of muscle cramps, pain, and/or weakness. This was pivotal role in TKI-induced cardiac muscle toxicity [4, 5]. quantified on a Likert scale from 1 (not at all) to 4 We assessed whether TKIs disturb skeletal muscle mito- (very much) resulting in significant different scores between chondrial density and function (cellular level), and if CML + MC and CML - MC (median 4.0 (interquartile this translates into alterations in muscle contractile function range (IQR) 3.0–4.0) and 2.5 (IQR 2.0–3.0), respectively; (muscle tissue level) and maximal exercise P = 0.002). The Brief Fatigue Inventory (BFI) [6] was used performance (whole-body level). To gain a better insight to compare the degree of fatigue in CML patients, showing into TKI-induced muscle complaints, these outcomes were higher fatigue levels in CML + MC when compared to compared between CML patients with and without muscle CML - MC (3.58 ± 2.19 and 0.95 ± 1.11, respectively; P = complaints. Written informed consent was obtained from 0.005). Ten control participants were matched on group the participants prior to study enrollment. This study was level for age, gender, BMI and physical activity level, approved by the Local Committee on Research Involving assessed by The Short Questionnaire to Assess Health- Human Subjects of the region Arnhem and Nijmegen, Enhancing Physical Activity (SQUASH) [7]. Subjects were ineligible if they had hereditary muscle defects, diabetes mellitus, hypo- or hyperthyroidism, severe electrolyte dis- turbances, or used co-medication known to cause muscle These authors contributed equally: S. J. C. M. Frambach, N. A. E. symptoms or have an effect on mitochondrial function Allard and N. M. A. Blijlevens, S. Timmers (e.g. statins, steroids, and metformin). Furthermore, subjects * N. M. A. Blijlevens 4 Center for Systems Biology and Bioenergetics, Radboud Center Nicole.Blijlevens@radboudumc.nl for Mitochondrial Medicine, Radboud University Medical Center, * S. Timmers Nijmegen, The Netherlands Silvie.Timmers@radboudumc.nl 5 Department of Neurology, Donders Institute for Brain, Cognition, and Behaviour, Radboud University Medical Center, 1 Radboud Institute for Health Sciences, Department of Physiology, Nijmegen, the Netherlands Radboud University Medical Center, Nijmegen, The Netherlands 6 Radboud Center for Mitochondrial Medicine, Translational 2 Radboud Institute for Health Sciences, Department of Metabolic Laboratory, Department of Pediatrics, Radboud Hematology, Radboud University Medical Center, Nijmegen, The University Medical Center, Nijmegen, The Netherlands Netherlands 7 Human and Animal Physiology, Wageningen University, 3 Department of Pharmacology and Toxicology, Radboud Institute Wageningen, The Netherlands for Molecular Life Sciences, Nijmegen, The Netherlands Letter 2117 with contra-indications for maximal exercise testing Furthermore, mitochondrial function, assessed by ATP pro- according to the ACC/AHA guidelines [8] and muscle duction capacity (Fig. 1a, c) and [1-14C]-pyruvate oxidation biopsy (e.g. anticoagulant therapy, bleeding disorders) were rates in the presence of malate or carnitine (Fig. 1b, d) was excluded. The demographic and hematological character- not different between groups. istics shown in Table 1 are not statistically significant dif- Maximal voluntary muscle strength of the dominant ferent between CML patients and controls and between quadriceps femoris muscle [11], did neither differ between CML + MC and CML - MC, except for a higher Charlson CML patients and controls (8.3 ± 2.0 N/kg and 7.9 ± 1.8 N/kg, Comorbidity Index [9] score in CML patients when com- respectively; P = 0.59), nor between CML + MC and pared to controls (median 2.0 (IQR 2.0–2.0) and 0.0 (IQR CML-MC (P = 0.97). Resistance to fatigue was assessed by 0.0–0.0), respectively; P < 0.001) caused by the presence electrically stimulating the quadriceps femoris muscle of CML. All participants completed the study protocol, repetitively at 40% of the MVC using 30 Hz bursts of one- i.e. a vastus lateralis muscle biopsy, electrical quadriceps second duration every other second for two minutes [11]. femoris stimulations, and an incremental cycling test. This fatigue protocol resulted in a significantly larger force Vastus lateralis muscle needle biopsies were performed decline in CML patients as compared to controls (31.8 ± under local anesthesia in overnight fasted state and processed 8.7% and 23.6 ± 7.7%, respectively; P = 0.010; Fig. 1e). for mitochondrial measurements according to standard lab Although a similar fatigability pattern was observed techniques as previously published [10]. Citrate synthase between CML + MC and CML - MC (force decline 29 ± activity, a marker for mitochondrial density, was not different 9% and 34 ± 9%, respectively; P = 0.24), the contractile between CML patients and controls (195 ± 80 mU/mg pro- properties of the quadriceps femoris muscle during tein and 171 ± 30 mU/mg protein, respectively, P = 0.24) repeated stimulation were explored in more detail. After and between CML + MC and CML - MC (P = 0.33). two minutes of stimulation CML + MC showed a Table 1 Subject and hematological characteristics Characteristics CML patients Controls P value CML + MC CML - MC P value Subject number, N 20 10 – 10 10 – Gender, male/female 14/6 7/3 1.00 7/3 7/3 1.00 Age, years 54 ± 8 58 ± 7 0.25 55 ± 9 54 ± 8 0.86 BMI, kg/m2 25.8 ± 4.1 27.5 ± 5.4 0.34 25.8 ± 3.6 25.8 ± 4.8 0.97 Physical activity, METmin/week; median (IQR) 2288 (1545–4982) 2850 (1965–4868) 0.53 2541 (1689–5488) 2070 (1403–4570) 0.63 Smoker, % 0 0 – 0 0 – Age at Dx, years 46 ± 8 N/A – 47 ± 7 46 ± 9 0.78 BCR-ABL level, N (%) N/A – 1.00 No MMR 1 (5) 1 (10) 0 (0) MMR or deeper 19 (95%) 9 (90) 10 (100) TKI, N (%) – 0.89 Imatinib 10 (50) N/A 6 (60) 4 (40) Dasatinib 5 (25) N/A 2 (20) 3 (30) Nilotinib 2 (10) N/A 1 (10) 1 (10) Bosutinib 2 (10) N/A 1 (10) 1 (10) Ponatinib 1 (5) N/A 0 (0) 1 (10) Duration of current TKI therapy, months; median (IQR) 42.5 (21.0–114.8) N/A – 66.5 (19.0–121.5) 38.0 (16.5–73.3) 0.63 Prior TKIs, N; median (IQR) 0.5 (0.0–1.0) N/A – 0.0 (0.0–1.0) 1.0 (0.0–1.0) 0.53 Charslon Comorbidity Index; median (IQR) 2.0 (2.0–2.0) 0.0 (0.0–0.0) <0.001 2.0 (2.0–2.0) 2.0 (2.0–2.0) 0.74 Potassium, mmol/l 3.9 ± 0.28 4.0 ± 0.20 0.50 3.9 ± 0.21 4.0 ± 0.34 0.76 Magnesium, mmol/l 0.82 ± 0.06 0.84 ± 0.06 0.49 0.82 ± 0.05 0.83 ± 0.06 0.63 Phosphate, mmol/l 0.82 ± 0.12 0.91 ± 0.17 0.11 0.83 ± 0.10 0.81 ± 0.15 0.75 Calcium, mmol/l 2.29 ± 0.09 2.30 ± 0.06 0.68 2.29 ± 0.09 2.29 ± 0.10 0.91 Albumin, g/l 38.7 ± 2.5 37.3 ± 2.0 0.13 39.1 ± 2.8 38.3 ± 2.2 0.48 TSH, mE/l; median (IQR) 2.1 (1.4–2.6) 2.0 (1.3–2.8) 0.90 2.0 (0.8–2.9) 2.1 (1.5–2.7) 0.61 CK, U/l; median (IQR) 137 (89–236) 120 (109–166) 0.61 169 (103–312) 127 (84–216) 0.90 Values are presented as mean ± SD unless indicated otherwise. There were no significant differences in subject and hematological characteristics between CML patients, except for a higher Charslon Comorbidity Index score in CML patients Also, there were no significant differences between CML patients with and without TKI induced muscle complaints MC muscle complaints, BMI body mass index, MET metabolic equivalent of task, IQR interquartile range, Dx diagnosis, MMR major molecular response, TKI tyrosine kinase inhibitor, TSH thyroid-stimulating hormone, CK creatine kinase 2118 Letter Fig. 1 Skeletal muscle mitochondrial function and muscle function responses are plotted every second during the complete (120 sec) parameters measured in CML patients and controls. a ATP production fatigue protocol. f Maximal force rise after two minutes of electrical rate normalized to muscle protein content and (b) [1-14C]-pyruvate quadriceps femoris stimulation was significantly lower in CML + MC oxidation rates in the presence of malate and carnitine normalized to when compared to CML - MC. g Quadriceps femoris muscle in muscle protein content measured in mitochondrial fractions from fresh CML + MC showed a tendency toward longer relaxation time vastus lateralis muscle biopsies were not affected by TKI use. There after two minute repetitive stimulation when compared to CML - MC. were also no differences in (c) ATP production rate and (d) [1-14C]- h Reported fatigue by CML patients (assessed by the Brief Fatigue pyruvate oxidation rates in the presence of malate and carnitine Inventory) correlates with half relaxation time in quadriceps femoris between CML + MC and CML - MC. e Force decline expressed as muscle after two minutes repetitive electrical stimulation (N = 15). percentage of the prefatigue value during two minutes repetitive Data are presented as means ± SEM for a–d; and means ± SD for e–g. electrical stimulation of the quadriceps femoris muscle shows a higher *P value < 0.05 is considered statistically significant level of muscle fatigue in CML patients compared to controls. Force Letter 2119 significantly lower maximal force rise (maximal slope MC showed delayed quadriceps femoris muscle force of force increment normalized for peak force) compared generation and a trend toward delayed relaxation in fatigued to CML - MC (0.54 ± 0.10%/ms and 0.67 ± 0.13%/ms, muscle compared to CML - MC. Since muscle fatigability, respectively; P = 0.038; Fig. 1f) and a tendency toward force generation, and relaxation are largely dependent on longer half relaxation time (time taken for force to decline Ca2+ regelulation by the SR, changes in SR functioning from 50 to 25% of the peak force; P = 0.07; Fig. 1g). The may underlie these findings [14]. In that respect, dis- half relaxation time at the end of the fatigue protocol turbances in Ca2+ homeostasis [15], and SR abnormalities strongly correlated with reported fatigue (rp = 0.72; P = (i.e., dilated SR with membrane whorls) [4] have been 0.002; Fig. 1h). Since muscle relaxation is dependent upon found upon imatinib treatment in myocytes, but have never the activity of sarco/endoplasmic reticulum Ca2+-ATPase been linked to muscle complaints. Although we found no (SERCA), an enzyme that mediates the re-uptake of cal- difference in SERCA activity between CML + MC and cium into the sarcoplasmic reticulum (SR) of skeletal CML - MC, muscle half relaxation time after 2-min stimu- muscle, SERCA activity was measured in whole-muscle lation correlated positively with the perception of fatigue in homogenates [11]. However, no significant differences in CML patients, and may therefore be an important key for SERCA activity were observed between CML + MC and understanding the mechanism underlying fatigue in CML. CML - MC (98.9 (IQR 79.3–110.7 mU/mg) and 101.5 (IQR To the best of our knowledge, maximal exercise capacity 77.0–109.8 mU/mg), respectively; P = 0.97). has not been assessed before in CML patients or other TKI- To measure whole-body fitness levels, all subjects per- users. Compared to controls, CML patients do not have formed an incremental cycle ergometer test (Lode Excali- diminished maximal exercise capacity as measured by bur; Groningen, the Netherlands) with continuous ECG VO2peak and have similar physical activity levels as controls. monitoring [10] to assess peak oxygen uptake (VO2peak). VO2peak was also not different between CML + MC VO2peak was 34.6 ± 8.4 ml/kg/min (100 ± 9% predicted) in and CML - MC, despite higher subjective fatigue levels in the CML patients, which was not different from the controls CML + MC. These findings fit with the unaltered mito- (VO2peak 35.6 ± 7.5 ml/kg/min, 100 ± 9% predicted; P = chondrial ATP production capacity, which is an important 0.75). In addition, no differences in VO2peak were observed determinant of VO22peak. between CML + MC and CML - MC (P = 0.38). There are several limitations to this study. Due to the Collectively, CML patients on TKI therapy show no signs exploratory character of the study a relatively large number of of skeletal muscle mitochondrial dysfunction. However, measurements were performed in a small sample size. quadriceps femoris muscle of TKI users fatigues to a larger Therefore, results should be cautiously interpreted. None- extent upon repetitive stimulation when compared to con- theless, this design made it possible to examine the influence trols. Changes in muscle contractile properties are associated of TKIs on multiple levels (i.e. cellular, muscle tissue and with TKI-induced muscle complaints, as CML + MC show whole body level) which offers broad insight into the effects a significant lower maximal force rise and a tendency toward of TKIs in CML patients. Secondly, participants were only a delayed muscle relaxation after two minutes of electrical included when they were able to perform all study measure- quadriceps femoris stimulations. CML patients did not have ments. Thus patients who were unable to perform exercise impaired maximal exercise performance. testing were excluded. Consequently, extreme cases of TKI- On a cellular level, no effects of TKI therapy on skeletal induced skeletal muscle complaints were not included in this muscle mitochondrial density and function were found. These study, which may have underestimated the results. results are in line with the only previous clinical case report in This study provides important information concerning which two CML patients, who had to interrupt or reduce the effects of TKIs on skeletal muscle function and whole therapy with nilotinib because of muscle pain, failed to show body fitness and lays foundation for further studies to elu- disturbances in mitochondrial oxidative enzyme reactions cidate the precise mechanism by which TKI therapy causes [12]. Intriguingly, in vitro studies in C2C12 myotubes muscle complaints and affects muscle function. showed no decline in ATP levels upon short-term imatinib incubation of 30 min [13], whereas long-term TKI-incubation Acknowledgements We would like to thank Berendien Stoltenborg- of 24 h showed decreased ATP levels overtime [5, 13]. Hogenkamp, Karina Horsting-Wethly and Theo van Lith of the Translational Metabolic Laboratory at the Radboudumc for their Disturbances in heart mitochondrial function are sug- assistance with the mitochondrial measurements. gested to occur secondary to activation of a stress response in the endoplasmic reticulum [4]. Perhaps, in skeletal Compliance with ethical standards muscle, changes in the function of other cellular organelles also precede mitochondrial disturbances. In support of this Conflict of interest The authors declare that they have no conflict of hypothesis, CML patients on TKI therapy showed sig- interest. nificantly more muscle fatigue than controls, and CML + 2120 Letter Publisher’s note: Springer Nature remains neutral with regard to mitochondrial function in isolated rat heart mitochondria and jurisdictional claims in published maps and institutional affiliations. H9c2 cells. Toxicol Sci. 2008;106:153–61. 6. Mendoza TR, Wang XS, Cleeland CS, Morrissey M, Johnson BA, Open Access This article is licensed under a Creative Commons Wendt JK, et al. The rapid assessment of fatigue severity in cancer Attribution 4.0 International License, which permits use, sharing, patients: use of the Brief Fatigue Inventory. Cancer. 1999;85: adaptation, distribution and reproduction in any medium or format, as 1186–96. long as you give appropriate credit to the original author(s) and the 7. Wendel-Vos GC, Schuit AJ, Saris WH, Kromhout D. Reprodu- source, provide a link to the Creative Commons license, and indicate if cibility and relative validity of the short questionnaire to assess changes were made. The images or other third party material in this health-enhancing physical activity. J Clin Epidemiol. 2003;56: article are included in the article’s Creative Commons license, unless 1163–9. indicated otherwise in a credit line to the material. If material is not 8. Gibbons RJ, Balady GJ, Bricker JT, Chaitman BR, Fletcher GF, included in the article’s Creative Commons license and your intended Froelicher VF, et al. ACC/AHA 2002 guideline update for exer- use is not permitted by statutory regulation or exceeds the permitted cise testing: summary article. A report of the American College of use, you will need to obtain permission directly from the copyright Cardiology/American Heart Association Task Force on Practice holder. To view a copy of this license, visit http://creativecommons. Guidelines (Committee to Update the 1997 Exercise Testing org/licenses/by/4.0/. Guidelines). J Am Coll Cardiol. 2002;40:1531–40. 9. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40:373–83. References 10. Allard NAE, Schirris TJJ, Verheggen RJ, Russel FGM, Rodenburg RJ, Smeitink JAM, et al. Statins affect skeletal muscle 1. Kekäle M, Peltoniemi M, Airaksinen M. Patient-reported performance: evidence for disturbances in energy metabolism. J adverse drug reactions and their influence on adherence and Clin Endocrinol Metab. 2018;103:75–84. quality of life of chronic myeloid leukemia patients on per oral 11. Gerrits HL, De Haan A, Hopman MT, van Der Woude LH, Jones tyrosine kinase inhibitor treatment. Patient Prefer Adherence. DA, Sargeant AJ. Contractile properties of the quadriceps muscle 2015;9:1733–40. in individuals with spinal cord injury. Muscle Nerve. 1999; 2. Marin D, Bazeos A, Mahon F-X, Eliasson L, Milojkovic D, Bua 22:1249–56. M, et al. Adherence is the critical factor for achieving molecular 12. Caocci G, Maioli MA, Atzeni S, Piras R, Carboni N, La Nasa G. responses in patients with chronic myeloid leukemia who achieve Absence of histological myopathy in chronic myeloid leukemia complete cytogenetic responses on imatinib. J Clin Oncol. patients complaining of muscle spasms and myalgia during 2010;28:2381–8. treatment with nilotinib. Leuk Res. 2012;36:e206–8. 3. Efficace F, Baccarani M, Breccia M, Cottone F, Alimena G, 13. Damaraju VL, Kuzma M, Cass CE, Putman CT, Sawyer MB. Deliliers GL, et al. Chronic fatigue is the most important Multitargeted kinase inhibitors imatinib, sorafenib and sunitinib factor limiting health-related quality of life of chronic myeloid perturb energy metabolism and cause cytotoxicity to cultured leukemia patients treated with imatinib. Leukemia. 2013;27: C2C12 skeletal muscle derived myotubes. Biochem Pharmacol. 1511–9. 2018;155:162–71. 4. Kerkela R, Grazette L, Yacobi R, Iliescu C, Patten R, Beahm C, 14. Westerblad H, Allen DG. Changes of myoplasmic calcium con- et al. Cardiotoxicity of the cancer therapeutic agent imatinib centration during fatigue in single mouse muscle fibers. J Gen mesylate. Nat Med. 2006;12:908–16. Physiol. 1991;98:615–35. 5. Will Y, Dykens JA, Nadanaciva S, Hirakawa B, Jamieson J, 15. Barr LA, Makarewich CA, Berretta RM, Gao H, Troupes CD, Marroquin LD, et al. Effect of the multitargeted tyrosine kinase Woitek F, et al. Imatinib activates pathological hypertrophy inhibitors imatinib, dasatinib, sunitinib, and sorafenib on by altering myocyte calcium regulation. Clin Transl Sci. 2014;7:360–7. Letter 2121 Leukemia (2019) 33:2121–2125 https://doi.org/10.1038/s41375-019-0445-5 Acute myeloid leukemia Insertional mutagenesis identifies cooperation between Setbp1 and Mllt3 in inducing myeloid leukemia development Bandana A. Vishwakarma1 Kristbjorn O. Gudmundsson1 Kevin Oakley1 Yufen Han1 Yang Du1 ● ● ● ● Received: 4 January 2019 / Revised: 20 February 2019 / Accepted: 8 March 2019 / Published online: 20 March 2019 © Springer Nature Limited 2019 To the Editor: 50% of the recipient mice in 10 months [5]. Such systems, as shown by us and others, also could be used as a powerful Multiple cooperating mutations are required to induce tool for identifying cooperating oncogenic mutations due to transformation during the development of acute myeloid insertional mutagenesis induced by retroviral vectors [7, 8]. leukemias (AMLs), and identification of these cooperating Therefore, to identify potential cooperating mutations for mutations is essential for developing better combinatorial SETBP1 overexpression in AML development, we cloned AML therapies. SETBP1, which encodes an AT-hook and sequenced a total of 26 retroviral integration sites from transcription factor [1, 2], likely plays a significant role in 16 Setbp1-induced mouse AMLs by splinkerette poly- driving human primary AML development as it was over- merase chain reaction (Supplementary Table 1). Integration expressed in up to 30% of primary AML patients [3, 4], and site sequences were subsequently matched against the its overexpression was capable of inducing AML develop- mouse genome sequence to identify their genomic loca- ment in mice [5]. Consistent with its association with poor tions. Interestingly, two leukemias carry independent patient prognosis [3, 4], SETBP1 overexpression also likely insertions at Mllt3 (Fig. 1a). Given that retroviruses inte- plays an important role in the maintenance of leukemia- grate in a largely random fashion into the genome, recurrent initiating cells (LICs) as it was identified as a top compo- integrations at Mllt3 strongly suggest that such integrations nent of a LIC gene expression signature for human AMLs may cooperate with Setbp1 in AML induction in these mice. [6]. SETBP1 may control the self-renewal of LICs through These two leukemias also expressed significantly higher direct transcription activation of oncogenes Hoxa9, levels of Mllt3 mRNA than leukemias without such inte- Hoxa10, and Myb, and repression of tumor suppressor gene grations (Fig. 1b), further indicating that Mllt3 over- Runx1 [1, 5]. SETBP1 may also promote AML develop- expression potentially induced by the viral integrations was ment through its interaction and stabilization of oncoprotein likely responsible for cooperating with Setbp1 in leukemia SET [3]. Despite the potentially significant contribution by induction. The lack of Mllt3 overexpression in the rest of SETBP1 overexpression to the initiation and maintenance of the mouse leukemias examined suggests a non-essential role primary human AMLs, its cooperating oncogenic partners for Mllt3 in Setbp1-induced leukemia development; how- in AML development remain unknown. ever, significantly higher levels of MLLT3 expression were Using a mouse bone marrow transduction and trans- detected in human AMLs with high-SETBP1 expression plantation system, we showed previously that over- than in samples with low-SETBP1 expression in an expression of Setbp1 induced AML development in over expression profiling study of human AMLs (Fig. 1c) [9], suggesting that this potential Setbp1/Mllt3 cooperation is likely conserved and prevalent in human AMLs with Supplementary information The online version of this article (https:// SETBP1 overexpression. doi.org/10.1038/s41375-019-0445-5) contains supplementary MLLT3, also known as AF9, first was identified as a material, which is available to authorized users. fusion partner for MLL1 in human myeloid leukemias [10]. * Yang Du Gene targeting in mice has revealed a critical role of Mllt3 yang.du@usuhs.edu in embryogenesis [11]. MLLT3 later was found to regulate transcriptional elongation as a component of the super 1 Department of Pediatrics, Uniformed Services University of the elongation complex (SEC) [12]. This function of MLLT3 is Health Sciences, Bethesda, MD, USA likely dependent on its ability to recognize H3K9 2122 Letter A D pMYs-Setbp1-IRES-GFP, Mllt3 (BL6) pMSCV-Mllt3-PGK-Puro, C57BL/6 or both Mllt3 (BL20) (5-FU) Scf Il-3 B Il-6 Scf, Il-3, Il-6 10 Mllt3 2 days 2 days Relative Expression 8 BM cells 6 E B6-Ly5.2 Relative colony formation 2.5 4 *** 2.0 **** S 2 M M+S 0 1.5 BM BL2 BL3 BL4 BL6 BL7 BL12 BL20 1.0 0.5 0 C 1st 2nd 3rd **** Relative Expression (log2) 6 F 4 100 Mllt3 Percent survival 2 (n=6) 0 50 Setbp1 -2 (n=7) -4 Mllt3+Setbp1 2° Trans (n=8) -6 (n=5) BM CD34+ SETBP1 SETBP1 0 low high 0 50 100 150 200 250 Days Fig. 1 Mllt3 cooperates with Setbp1 to induce AML development. (BM) progenitors transduced with pMYs-Setbp1-IRES-GFP (S), a Schematic diagrams of genomic locations of retroviral insertions (red pMSCV-Mllt3-PGK-Puro (M), or the combination (M + S) in primary triangles) at Mllt3 in two Setbp1-induced mouse leukemias BL6 and (1st), secondary (2nd), and tertiary (3rd) platings on methylcellulose BL20. b Real-time RT-PCR analysis of Mllt3 expression in indicated in the presence of murine Scf (100 ng/ml), Il-6 (10 ng/ml), and Il-3 Setbp1-induced leukemias relative to Rpl4 mRNA levels and control (6 ng/ml) (n = 3 for each transduction). The titer of Mllt3 and Setbp1 healthy bone marrow cells (BM). c Relative expression levels (log2) of virus used for co-transduction was reduced by half compared to the MLLT3 in human normal bone marrow (BM), CD34+ cells, and AML single-transduction groups to ensure same total viral titer being used cells with and without SETBP1 overexpression (SETBP1 high and for each transduction group. Depending on the virus used, transduced low). Normalized expression data from microarray dataset published cells were purified by selection with puromycin at 48 h post-trans- by Valk et al. [9] was downloaded from Oncomine website duction, by FACS based on GFP positivity at 72 h post-transduction, (http://www.oncomine.org/) and analyzed. d Schematic diagram or both before primary platings. Cells were re-plated every five days. depicting mouse bone marrow transduction and subsequent serial f Survival curves of irradiated B6-Ly5.2 mice receiving similarly replating and transplantation assays for testing Setbp1/Mllt3 coopera- transduced cells as in E and also secondary recipient mice receiving tion in inducing leukemia transformation. e Mean and SD of relative 1 × 106 spleen cells from primary Mllt3 + Setbp1 leukemic mice colony-forming potential of 5-fluorouracil (5-FU) treated bone marrow (2° Trans). ***P < 0.001; ****P < 0.0001 (two-tailed Student’s t test) acetylation [13]. A role for MLLT3 overexpression in cancer effects of co-transduction with pMSCV-Mllt3-PGK-Puro development has not been reported previously. plus pMYs-Setbp1-IRES-GFP viruses on 5-fluorouracil To confirm cooperativity between Setbp1 and Mllt3 in (5-FU)-treated bone marrow progenitors from C57BL/6 leukemic transformation, we cloned Mllt3 cDNA into the mice (Fig. 1d). We first compared the self-renewal potential pMSCV-PGK-Puro retroviral vector and examine the of co-transduced cells to cells singly transduced with Setbp1 Letter 2123 or Mllt3 virus by serial replating assay on methylcellulose. primary leukemic mice developed AML with a comparable Cells singly transduced with Mllt3 virus failed to form latency to recipients of Setbp1 leukemia cells (16 ± 2 days colonies after first plating, suggesting that Mllt3 over- vs. 17 ± 4 days, mean ± SD) (Supplementary Figure 2). expression is not capable of inducing their self-renewal Moreover, to confirm that co-expression of Setbp1 and in vitro (Fig. 1e). Both co-transduced cells and cells infected Mllt3 in same cell is required for their cooperativity, both with Setbp1 virus alone continued to form colonies of proviruses were detected in genomic DNA from all ran- myeloid progenitors on secondary and tertiary platings domly selected colonies generated by these leukemic cells (Fig. 1e), suggesting that the combined expression of (Supplementary Figure 3). Setbp1 and Mllt3 like Setbp1 expression alone is capable of Hoxa9 activation is essential for Setbp1-induced trans- immortalizing myeloid progenitors. Interestingly, con- formation [1]. It has been shown previously that Meis1 siderably increased colony formation by co-transduced cells overexpression can cooperate with Hoxa9 activation to compared to cells singly transduced by Setbp1 virus was induce AML development [14]. Therefore, we tested the observed at both secondary and tertiary plating (Fig. 1e), possibility that co-expression of Mllt3 with Setbp1 may further indicating that overexpression of Mllt3 may sig- increase Meis1 expression in hematopoietic stem and pro- nificantly enhance Setbp1-induced self-renewal of myeloid genitor cells by transducing mouse lin−Sca-1+c-kit+ (LSK) progenitors. cells and analyzing their Meis1 expression at 72 h after To test whether co-expression of Setbp1 and Mllt3 could transduction and also in their primary, secondary, and ter- lead to accelerated leukemia development in vivo, we also tiary colonies. Mllt3 overexpression alone is not sufficient transplanted unpurified transduced cells into lethally irra- to activate Meis1 as it failed to increase Meis1 mRNA levels diated B6-Ly5.2 mice. As expected from our previous stu- at 72 h post-transduction (Supplementary Figure 4). dies, three of seven mice receiving cells singly transduced Although both overexpression of Setbp1 alone and together by Setbp1 virus developed AML in 8 months (Fig. 1f). In with Mllt3 upregulated Meis1 mRNA to similar levels at contrast, none of the recipient mice in Mllt3 transduction 72 h (Supplementary Figure 4), significantly higher levels group developed leukemia in the same period (Fig. 1f), of Meis1 mRNA and protein were detected in all serial suggesting the overexpression of Mllt3 alone is not suffi- colonies overexpressing both Setbp1 and Mllt3 than in cient to induce leukemia development. Importantly, 100% colonies overexpressing Setbp1 only (Fig. 2a, b). Sig- of the recipient mice for the co-transduced cells developed nificantly higher levels of Meis1 expression also were AML at a much reduced latency [76 ± 13 days (mean ± SD); detected in Mllt3 + Setbp1 AMLs than in Setbp1 AMLs p < 0.0001, log-rank test] than mice of the Setbp1 group (Fig. 2c, d). Moreover, in supporting a critical role of (Fig. 1f). Resembling leukemic mice induced by Setbp1 increased Meis1 expression in transformation induced by alone, these mice developed similarly enlarged spleens and Setbp1/Mllt3 cooperation, Meis1 deletion by Cre/ERT2 in displayed significant infiltrations of leukemic cells into secondary colony cells produced from Meis1 conditional nonhematopoietic tissues including liver and lung (Sup- LSK cells after Mllt3 + Setbp1 transduction significantly plementary Table 2 and Supplementary Figure 1A). decreased their colony-forming capability (Fig. 2e). These Fluorescence-activated cell sorting analyses of the two results strongly suggest that a major cooperating mechanism leukemia types showed similar percentages of green fluor- for Setbp1 and Mllt3 in leukemia induction may be their escent protein-positive leukemia cells (83–95%) in the bone cooperation in inducing high levels of Meis1 expression. marrow and spleen with a majority (75–85%) of the leu- Interestingly, by performing chromatin immunoprecipita- kemia cells in the bone marrow being positive for Gr-1 and tion analysis on tertiary colony cells generated by LSK cells only a minority of the cells being weakly positive for CD19, co-expressing Setbp1 and Mllt3, we also detected significant CD4, or Ter119 (Supplementary Figure 1B). However, Mllt3 binding at the Meis1 locus (Fig. 2f), further sug- Mllt3 + Setbp1 AMLs have greatly increased number of c- gesting that Meis1 is a direct transcriptional target of Mllt3. kit-positive cells (59 ± 2% vs. 3 ± 1%, mean ± SD) and less This direct regulation of Meis1 transcription by Mllt3 is also Sca-1-positive cells (2 ± 1% vs. 10 ± 3%, mean ± SD) in conserved in human AML cells as a significant positive bone marrow than Setbp1 AMLs (Supplementary Fig- correlation between MLLT3 and MEIS1 expression was ure 1B). Consistent with the c-kit expression pattern, observed in human AMLs expressing high levels of cytospin preparations of leukemic bone marrow and spleens SETBP1 (Supplementary Figure 5). (Supplementary Figure 1C) revealed significantly higher Overexpression of SETBP1 in primary AMLs has been frequencies of myeloid blasts in Mllt3 + Setbp1 leukemic associated with poor disease prognosis, suggesting better mice than Setbp1 leukemic mice [70 ± 6% vs. 30 ± 8% therapies are critically needed to improve treatment out- (mean ± SD) of all nucleated cells in the bone marrow]. come. Our study identifies MLLT3 as a common cooperat- Mllt3 + Setbp1 leukemias are also transplantable as sec- ing partner for SETBP1 in inducing human AML ondary recipients receiving 1 × 106 spleen cells from development, indicating that inhibition of signaling 2124 Letter A B Meis1 1st 2nd 3rd 5 *** C+S M+S Relative Expression 4 ** Meis1 3 **** Mllt3 2 Setbp1 1 β-Actin 0 1st 2nd 3rd C Meis1 D 20 Relative Expression Meis1 15 β-actin 10 1 2 3 4 5 6 7 8 M+S S 5 0 WT 1 2 3 4 5 6 7 8 M+S S E F 1.5 0.20 *** **** 4-OHT IgG (C+S) Relative colony EtOH Mllt3 (C+S) formation 1.0 0.15 IgG (M+S) **** ** Mllt3 (M+S) % Input 0.5 0.10 0.05 0 fl wt 0.00 ko -1.2 kb 0 +2.3 kb Meis1fl/fl;Cre Meis1+/+;Cre/ /ERT2 ERT2 Fig. 2 Meis1 is a direct transcriptional target of Mllt3. a Real-time RT- d Western blotting analyses of whole cell extracts prepared from the PCR analysis of Meis1 expression in cells of primary (1st), secondary same leukemia cells as in C using indicated antibodies. e Upper panel, (2nd), and tertiary (3rd) colonies formed by LSK cells transduced by secondary colony cells generated by Mllt3 + Setbp1 co-transduced pMSCV-PGK-Puro + pMYs-Setbp1-IRES-GFP (C + S), or pMSCV- LSK cells of indicated genotypes were examined by colony-forming Mllt3-PGK-Puro + pMYs-Setbp1-IRES-GFP (M + S) viruses. Relative assay after treatment with 4-hydroxytamoxifen (4-OHT) or control expression levels (mean ± SD) were calculated by normalizing to Rpl4 ethanol (EtOH) and their relative colony-forming potentials are shown mRNA levels in the same sample and also to C + S cells (n = 3 for (mean ± SD, n = 3 for each transduction). Lower panel, different each transduction). GFP positive and puromycin resistant cells were Meis1 alleles including floxed (fl), knockout (ko), and wild-type (wt) collected at 72 h post-transduction for the serial replating assay. detected by PCR in the secondary colony cells after 4-OHT or EtOH b Western blotting analyses of whole cell extracts prepared from the treatment. f Quantitative ChIP analysis of indicated genomic regions same colony cells as in (a) using indicated antibodies. c Real-time RT- relative to Meis1 transcriptional start site using a Mllt3-specific anti- PCR analysis of Meis1 expression in spleen cells of leukemic mice body or control IgG in tertiary colony cells generated by mouse induced by Mllt3 + Setbp1 co-transduction (M + S) or Setbp1 LSK cells similarly transduced as in (a). **P < 0.01; ***P < 0.001; **** transduction alone (S) and control mouse bone marrow cells (WT). P < 0.0001 (two-tailed Student’s t test) Letter 2125 pathways downstream of MLLT3 may prove valuable for development genes by acting as an epigenetic hub. Nat Commun. treating such AMLs. MLLT3 activation in these AMLs is 2018;9:2192. 3. Cristobal I, Blanco FJ, Garcia-Orti L, Marcotegui N, Vicente C, likely independent of SETBP1 as no significant correlation Rifon J, et al. SETBP1 overexpression is a novel leukemogenic was detected between their expression levels and Setbp1 mechanism that predicts adverse outcome in elderly patients with overexpression also failed to increase Mllt3 expression in acute myeloid leukemia. Blood. 2010;115:615–25. mouse LSK cells (Supplementary Figures 6 and 7). Coop- 4. Lucas CM, Scott LJ, Carmell N, Holcroft AK, Hills RK, Burnett AK, et al. CIP2A- and SETBP1-mediated PP2A inhibition reveals erativity also may exist between MLLT3 and SETBP1 AKT S473 phosphorylation to be a new biomarker in AML. missense mutations in the development of other myeloid Blood Adv. 2018;2:964–8. neoplasms since SETBP1 missense mutants are known to 5. Vishwakarma BA, Nguyen N, Makishima H, Hosono N, Gud- activate similar targets as wild-type SETBP1 [15]. As a mundsson KO, Negi V, et al. Runx1 repression by histone dea- cetylation is critical for Setbp1-induced mouse myeloid leukemia component of SEC, MLLT3 was shown to regulate tran- development. Leukemia. 2016;30:200–8. scriptional elongation, but whether its overexpression could 6. Gentles AJ, Plevritis SK, Majeti R, Alizadeh AA. Association of a play any role in cancer development was unclear. Our study leukemic stem cell gene expression signature with clinical outcomes demonstrates for the first time that MLLT3 overexpression in acute myeloid leukemia. J Am Med Assoc. 2010;304:2706–15. 7. Du Y, Spence SE, Jenkins NA, Copeland NG. Cooperating can promote AML development by activating MEIS1 tran- cancer-gene identification through oncogenic-retrovirus-induced scription. Further studies using genomic and proteomic insertional mutagenesis. Blood. 2005;106:2498–505. approaches will be required to identify additional MLLT3 8. Jin G, Yamazaki Y, Takuwa M, Takahara T, Kaneko K, Kuwata transcriptional targets promoting AML development and T, et al. Trib1 and Evi1 cooperate with Hoxa and Meis1 in myeloid leukemogenesis. Blood. 2007;109:3998–4005. also other transcriptional cofactors critical for such 9. Valk PJ, Verhaak RG, Beijen MA, Erpelinck CA, Barjesteh van activation. Waalwijk van Doorn-Khosrovani S, Boer JM, et al. Prognostically useful gene-expression profiles in acute myeloid leukemia. New Acknowledgements This work was supported by National Institutes of Engl J Med. 2004;350:1617–28. Health (NIH) grant RO1CA143193 (Y.D.) and USUHS Pediatrics 10. Nakamura T, Alder H, Gu Y, Prasad R, Canaani O, Kamada N, Grant QP86GI (Y.D.). The views presented in this paper are those of et al. Genes on chromosomes 4, 9, and 19 involved in 11q23 the authors; no endorsement by the Uniformed Services University of abnormalities in acute leukemia share sequence homology and/or the Health Sciences or the Department of Defense has been given or common motifs. Proc Natl Acad Sci USA. 1993;90:4631–5. should be inferred. 11. Collins EC, Appert A, Ariza-McNaughton L, Pannell R, Yamada Y, Rabbitts TH. Mouse Af9 is a controller of embryo patterning, like Mll, whose human homologue fuses with Af9 after chromo- Compliance with ethical standards somal translocation in leukemia. Mol Cell Biol. 2002;22:7313–24. 12. Lin C, Smith ER, Takahashi H, Lai KC, Martin-Brown S, Florens Conflict of interest The authors declare that they have no conflict of L, et al. AFF4, a component of the ELL/P-TEFb elongation interest. complex and a shared subunit of MLL chimeras, can link tran- scription elongation to leukemia. Mol Cell. 2010;37:429–37. Publisher’s note: Springer Nature remains neutral with regard to 13. Li Y, Wen H, Xi Y, Tanaka K, Wang H, Peng D, et al. AF9 jurisdictional claims in published maps and institutional affiliations. YEATS domain links histone acetylation to DOT1L-mediated H3K79 methylation. Cell. 2014;159:558–71. 14. Thorsteinsdottir U, Kroon E, Jerome L, Blasi F, Sauvageau G. References Defining roles for HOX and MEIS1 genes in induction of acute myeloid leukemia. Mol Cell Biol. 2001;21:224–34. 1. Oakley K, Han Y, Vishwakarma BA, Chu S, Bhatia R, Gud- 15. Nguyen N, Vishwakarma BA, Oakley K, Han Y, Przychodzen B, mundsson KO, et al. Setbp1 promotes the self-renewal of murine Maciejewski JP, et al. Myb expression is critical for myeloid myeloid progenitors via activation of Hoxa9 and Hoxa10. Blood. leukemia development induced by Setbp1 activation. Oncotarget. 2012;119:6099–108. 2016;7:86300–12. 2. Piazza R, Magistroni V, Redaelli S, Mauri M, Massimino L, Sessa A, et al. SETBP1 induces transcription of a network of