Leukemia (2020) 34:1158–1162 https://doi.org/10.1038/s41375-019-0643-1 LETTER Acute lymphoblastic leukemia B-cell leukemia transdifferentiation to macrophage involves reconfiguration of DNA methylation for long-range regulation Alberto Bueno-Costa1 David Piñeyro 1 Marta Soler1 Biola M. Javierre1 Helena Raurell-Vila2 ● ● ● ● ● Marc Subirana-Granés2 Lorenzo Pasquali 1,2,3 Jose A. Martinez-Climent4,5 Manel Esteller 1,5,6,7 ● ● ● Received: 4 July 2019 / Revised: 18 October 2019 / Accepted: 3 November 2019 / Published online: 12 November 2019 © The Author(s) 2019. This article is published with open access To the Editor: lineage priming that generate distinct blood cell types according to the physiological or pathological environment Hematopoiesis is a highly regulated process that, starting [2]. In this context, there are increasing data of hemato- from hematopoietic stem cells (HSCs) with self-renewal poietic plasticity and cell lineage conversion, particularly in capacity in the adult human bone marrow, is able to gen- leukemogenesis. Examples of transdifferentiation include 1234567890();,: 1234567890();,: erate all different types of mature blood cells. The classical B-cell lymphomas that can transform to histiocytic/dendritic view of hematopoiesis defines binary branching points from cell sarcoma, erythroid/megakaryocytic lineages changing these HSCs that segregate lineages and direct differentiation to granulomonocytic-like lineage upon use of a histone to terminally differentiated functional cell types [1]. How- demethylase LSD1 inhibitor or B-ALL (acute lympho- ever, the described hierarchical model can be complemented blastic leukemia) patients that evaded CD19-directed anti- with the emerging data that suggest the existence of body therapy (blinatumomab) by undergoing myeloid- hematopoietic stem and progenitor cells with a continuum lineage switch. Related to the latter scenario, lineage of transitory differentiation stages, including cells with early switching has also been reported as a cause of antigen loss in chimeric antigen receptor T-cell therapies, where B-ALL patients transdifferentiate in their relapse as acute myelo- blastic leukemia in response to the initial CD19-directed Supplementary information The online version of this article (https:// doi.org/10.1038/s41375-019-0643-1) contains supplementary immunotherapy [3]. Due to the central role of epigenetics, material, which is available to authorized users. particularly DNA methylation, in the successful generation of differentiated blood cell types and its plasticity during * Manel Esteller mesteller@carrerasresearch.org lineage specification [4], we wondered about its function in hematopoietic transdifferentiation, a largely unexplored 1 Josep Carreras Leukaemia Research Institute (IJC), Badalona, field. Barcelona, Catalonia, Spain Our studied model of transdifferentiation is a well- 2 Program of Predictive and Personalized Medicine of Cancer defined experimental system that converts B cells into (PMPPC), Endocrine Regulatory Genomics Laboratory, macrophages. Following initial work that demonstrated that Department of Endocrinology and Nutrition, Germans Trias i normal murine B-cell precursors as well as mature Pujol University Hospital and Research Institute (IGTP), Badalona, Catalonia, Spain antibody-producing B cells can be induced by C/EBPα to 3 transdifferentiate into functional macrophages [5], a murine CIBER de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM), Barcelona, Spain cellular model was established of pre-B cells containing a 4 fusion of C/EBPα with the estrogen receptor hormone Division of Hematological Oncology, Center for Applied Medical Research (CIMA), University of Navarra, Pamplona, Spain binding domain (C/EBPαER) that converts them to 5 macrophage-like cells upon 17β-estradiol exposure [6]. We Centro de Investigacion Biomedica en Red Cancer (CIBERONC), 28029 Madrid, Spain have recently translated this model to human B-lymphoma 6 and leukemia cell lines that can be induced by C/EBPα to Institucio Catalana de Recerca i Estudis Avançats (ICREA), Barcelona, Catalonia, Spain transdifferentiate into functional macrophages [7]. Impor- 7 tantly, primary human BCR-ABL1(+) B-ALL cells could Physiological Sciences Department, School of Medicine and Health Sciences, University of Barcelona (UB), also be induced to reprogram into macrophage-like cells by Barcelona, Catalonia, Spain transient expression of C/EBPα [8]. To explore the changes B-cell leukemia transdifferentiation to macrophage involves reconfiguration of DNA methylation for. . . 1159 that the DNA methylome undergoes upon transdifferentia- associated programs (such as FOXC1, GATA2, and tion, we have herein applied this experimental system. SPI1). We proceeded further to validate two candidate Thus, we have analyzed the human precursor B-ALL cell genes detected as undergoing CpG demethylation- line RCV-ACH transfected with the transgene C/EBPαER, associated reactivation for our multiomic approaches in thereafter termed BLaER1, upon 17β-estradiol-mediated the transdifferentiation model: the interleukin-1 receptor transdifferentiation at seven timepoints (0, 3, 12, 24, 48, 72, antagonist (IL1RN) and integrin alpha X (ITGAX), both and 168 h) using a comprehensive DNA methylation genes that are almost exclusively expressed in macrophages microarray that interrogates more than 850,000 CpG sites [10, 11]. The methylation changes of the identified CpG (Supplementary Fig. 1a and Supplementary Methods). sites associated with IL1RN and ITGAX were further DNA methylation data are available on the GEO repository confirmed by bisulfite genomic sequencing of multiple under accession number GSE132845. We have observed a clones and bisulfite pyrosequencing (Supplementary Fig. 4 significant change in the methylation status of 251 CpG and Supplementary Methods). Quantitative real-time PCR sites during the transdifferentiation process (p-value < 0.05 (Supplementary Methods) demonstrated lack of expression and CpG B-value change ≥0.66) (Supplementary Table 1 for both transcripts in the initial B-ALL cells in the absence and Supplementary Methods). Most strikingly, all except of 17b-estradiol treatment and a strong gain of expression at one (250 of 251, 99.6%) were hypomethylation changes the end of the transdifferentiation process in the (Fig. 1a and Supplementary Fig. 1a). In this regard, these macrophage-like cells (Fig. 1c). The assessment of protein hypomethylation events occurred in the context of down- levels for IL1RN (western blot) and ITGAX (cell cyto- regulation of the DNA methyltransferases DNMT1 and metry) (Supplementary Methods) obtained the same DNMT3B, but not DNMT3A, in our transdifferentiation expression patterns (Fig. 1c). The use of the demethylating model (Supplementary Fig. 2). The DNA methylation pat- agent 5-aza-2′-deoxycytidine in BLaER1 cells induced tern of the endpoint of transdifferentiation (BlaER1 at 168 IL1RN and ITGAX expression (Supplementary Fig. 5), h) for these sites mimicked the CpG methylation status of supporting the role of CpG methylation in gene silencing. naive macrophages (Fig. 1a and Supplementary Table 1). Most importantly, 17β-estradiol treatment of the original According to genomic distribution of the identified CpG naive pre-B-ALL leukemia cells RCH-ACV, which were sites, 141 CpGs (56.2%) had an associated gene, whereas not transfected with the transgene C/EBPαER and thus do 110 CpGs (43.8%) were in regions of the genome without no transdifferentiate, maintained the gene silencing of any annotated gene (Fig. 1b). IL1RN and ITGAX (Fig. 1c). These data support the role of Due to our interest in epigenetic modifications that can a wave of DNA demethylation that, acting locally in the actively contribute to the transdifferentiation phenomenon CpG-associated genes, confers macrophage identity to the and the classical view of the impact of DNA methylation on original pre-B-ALL cells. transcription, we first studied the 141 CpG sites that are Strikingly, our analysis workflow showed, as described annotated as having an associated gene in relation with the above (Fig. 1b), that among the differentially methylated available expression levels of the corresponding gene [9]. CpGs in our transdifferentiation model, 110 CpGs (43.8%) We identified 41 CpG sites (29.1%), corresponding to 39 were located in genome contexts without any genes in their genes (Supplementary Table 2), for which the methylation vicinity. Thus, in these cases, it is possible that a candidate status significantly correlated with the expression of the regulatory event of those CpGs occurs in the 3D organi- associated gene: in most cases the CpG hypomethylation zation of the human genome. In this regard, regulation at event was linked to gene expression (32 of 39, 82%), long-range distance can occur due to the formation of whereas only in a minority of cases was demethylation “loops” in the DNA that, for example, collocate enhancer or associated with gene repression (7 of 39, 18%) (Supple- silencer sequences and minimal promoters. DNA methyla- mentary Table 2). Using data mining (Supplementary tion patterns at enhancers are relevant to the determination Methods), we observed that these 41 CpG sites were in of cell identity [12] and their aberrant CpG methylation binding sites for 80 transcription factors (Supplementary profile has been observed in human cancer [13], including Fig. 3a). Gene ontology analysis using a hypergeometric hematological malignancies. Thus, we investigated how test to find biological processes overrepresented in our set many of these nongene associated CpGs were located in of transcription factors (Supplementary Methods) demon- distant regulatory regions, taking advantage of our available strated, in addition to global regulatory networks, an Promoter Capture Hi-C (PCHi-C) data in macrophages [14]. enrichment in the “immune system development,” “cell fate We also included in our analyses the 141 CpG sites with commitment,” and “leukocyte differentiation” categories associated genes due to the possibility that, in addition to (Supplementary Fig. 3b), including homeobox proteins regulating genes in lineal proximity, these CpG sites could related to leukemia biology (such as MEIS1, HOXA2, and also act as long-range interactor sequences. We found that HOXC11) and lymphoid and myeloid-differentiation- 72 of the overall 251 CpGs (28.7%) were located in 1160 A. Bueno-Costa et al. RCH- Distribution of the 251 CpGs BLaER1 ACV 9 6 A Hours after induction (3.6%) (2.4%) B 3 TSS1500 20 (8.0%) (1.2%) TSS200 5’UTR Gene-Associated 110 CpG sites (43.8%) 1st Exon 99 Body (39.4%) 3’UTR Non-Gene Associated 4 CpG sites (1.6%) C IL1RN RNA Expression (AU) ** BLaER1 RCH-ACV 80000 0h 72h 168h 0h 72h 168h 60000 IL1RN 40000 20000 Lamin B1 0 0h 168h BLaER1 BLaER1 0h BLaER1 168h RCH-ACV 168h ITGAX 250 0.1% 250 80.1% 250 RNA Expression (AU) 6 0.2% 6×10 *** 99.9% 19.9% 99.8% SSC-A 6 150 150 150 4×10 6 2×10 50 50 50 0 0 0 0 100% 0% 0h 168h 2 3 0 10 10 10 10 4 5 2 3 0 10 10 10 10 4 5 2 3 0 10 10 10 10 4 5 Methylation (Beta value) BLaER1 ITGAX (APC) ITGAX (APC) ITGAX (APC) Fig. 1 DNA methylation analysis at different timepoints of B-ALL-to- controls, respectively. b Pie chart showing the genomic distribution of Macrophage transdifferentiation. a Heatmap showing the methylation the 251 significant CpGs. c qRT-PCR (***p < 0.001, T-Test) western state of the 251 significant hyper/hypomethylated CpGs during 7 days blot and flow cytometry analysis at different timepoints of transdif- of transdifferentiation. RCH-ACV (treated 7 days with 17β-estradiol, ferentiation of IL1RN and ITGAX, whose associated promoter CpGs IL-3, and M-CSF) and macrophage are taken as negative and positive are significantly demethylated during transdifferentiation candidate distant regulatory regions defined by PCHi-C in located in nonpromoter regions of the gene (11 of 16, 69%) macrophages (Supplementary Fig. 1b and Supplementary (Supplementary Table 4). Using data mining (Supplemen- Table 3) [14]. To address the functionality of the observed tary Methods), we observed that four of the PCHi-C-derived DNA methylation changes, we then interrogated whether CpG dinucleotides contained a binding motif for the the methylation status of these 72 CpG sites had any impact CCCTC-binding factor (CTCF), the most widely recog- on the expression of the genes whose promoters are targeted nized protein controlling three-dimensional structures of by these long-range regulatory sequences [14]. We found DNA. Related to the 52 identified target genes, for 31 genes that the methylation status of 34 of the 72 CpG sites (60%) the CpG hypomethylation event acted as an enhancer (47.2%) correlated with the expression of 52 genes that event associated with the expression of the corresponding interacted with these regions (Supplementary Fig. 1b and gene, whereas for 21 genes (40%) the demethylation event Supplementary Table 4). Importantly, all except one (33 of was associated with gene inactivation and can be classified 34, 97%) were CpG hypomethylation changes (Supple- as a candidate silencer sequence in our pre-B-ALL to mentary Table 4). Most of the 34 PCHi-C-derived CpG macrophage transdifferentiation model (Supplementary dinucleotides represented unique interactions between one Fig. 1b and Supplementary Table 4). methylation site and one gene (21 of 34, 62%) (Fig. 2a), We experimentally validated the available expression followed by dual interactions (one CpG interacting with two microarray data [7] using quantitative real-time PCR for genes, 7 of 34, 21%), but more complex interactions were five candidate targets: three genes with CpG hypo- also observed (Fig. 2a). Interestingly, although many of methylation associated expression upon transdifferentia- these CpG sites were in nonassociated gene regions (18 of tion (ras homolog family member G, RHOG; C-X-C motif 34, 53%), we also frequently observed CpGs within asso- chemokine ligand 8, CXCL8; and C-C motif chemokine ciated genes (16 of 34, 47%) (Fig. 2b). It is important to receptor 1, CCR1) and two genes with CpG highlight that the CpGs in the latter subset were mainly demethylation-associated expression reduction upon cell B-cell leukemia transdifferentiation to macrophage involves reconfiguration of DNA methylation for. . . 1161 Single Interaction 21 Distribution of the 34 significant CpGs 2 Double Interaction (5.9%) 3 7 (8.8%) TSS1500 Triple Interaction 2 5’UTR Gene-Associated 18 Body CpG sites 10 Quadruple Interaction (52.9%) (29.4%) 3’UTR 2 Non Gene-Associated Quintuple Interaction CpG sites 2 1 CpG Position Gene (2.9%) # Normalized UMIs RHOG CCR1 CXCL8 3000 BLaER1 CCR1 RNA Expression (AU) 0h Contacts/ 5 *** 4000 *** 8000 *** 2000 Maximum 1000 0% 100% 3 2000 4000 0 1 0 0 0 0h 168h 0h 168h 0h 168h BLaER1 BLaER1 BLaER1 45.8 46.0 46.2 46.2 46.4 Coordinates Chr3 (Mb) RNA Expression (AU) CHML DBF4 # Normalized UMIs 1.5 1.5 3000 BLaER1 CCR1 ** *** 168h Contacts/ 1.0 1.0 2000 Maximum 0.5 0.5 1000 0% 100% 0 0 0 0h 168h 0h 168h BLaER1 BLaER1 45.8 46.0 46.2 46.2 46.4 Coordinates Chr3 (Mb) Fig. 2 Analysis of the loop formation between differentially methy- smoothed trend line of the contact profiles of the bait at time 0 (top) lated distant regions and gene promoters. a Illustrative representation and at the end of BlaER1 transdifferentiation (168 h) (below). A of CpG interaction with promoter regions that significantly modulated bottom heatmap (domainogram) reports the differential mean contacts expression after transdifferentiation. b Distribution of the 34 sig- per a series of restriction enzyme fragments. The UMI-4C analysis nificant CpGs on the human genome. c RT-qPCR (*p < 0.05, ***p < shows that upon BlaER1 transdifferentiation the hypomethylated CpG- 0.001, T-Test) of three enhancer (top) and two candidate silencer- containing sequence engages a strong distal chromatin contact (black controlled genes (below). d View of the 3D chromatin contacts, arrow) with the promoter of CCR1. Of note the chromatin contact obtained by UMI-4C, for a bait located at the CpG-containing can- intensity for the same region was very low for the hypermethylated didate enhancer sequence for the CCR1 promoter in the BlaER1 regulatory sequence at time 0 h of transdifferentiation (Chi-square transdifferentiation model. The top graph of the panel shows the adjusted p < 0.0001) conversion (CHM like Rab escort protein, CHML; and contact quantifications only between DNA sequences DBF4 zinc finger, DBF4) (Fig. 2c). The methylation located within a ~0.5 Kb and ~1 Mb interval [15]: among changes of the distant CpG sites associated with the target the five studied genes, only the CCR1 promoter and its genes was further confirmed by bisulfite genomic putative PCHi-C obtained CpG-containing regulatory sequencing of multiple clones and bisulfite pyrosequen- region fulfilled these parameters. UMI-4C data are avail- cing (Supplementary Fig. 6 and Supplementary Methods). able at the SRA repository under accession number The use of the demethylating agent 5-aza-2′-deox- PRJNA548887. Most importantly, we found that the CpG ycytidine in BLaER1 cells induced RHOG, CXCL8, and demethylation event occurring during transdifferentiation CCR1 expression (Supplementary Fig. 5), further sup- for the PCHi-C derived candidate long-distance regulatory porting a role of the identified CpGs in the candidate sequence of CCR1 was not only associated with the enhancer sequences. Finally, using targeted chromosome expression of the gene (Fig. 2c), but also with the for- capture with unique molecular identifiers (UMI-4C), a mation of a new loop between the CpG-containing distant recently developed method to quantitatively measure 3D regulatory region and the CCR1 proximal promoter (Chi- interaction intensities between distant regulatory regions square test p < 0.0001) (Fig. 2d), supporting a role as an and their corresponding promoters [15], we decided to enhancer. These data highlight the relevance of DNA confirm the loop dynamics of the singled-out targets. One methylation events at distant regulatory regions to confer technical limitation of UMI-4C is that it provides reliable cell identity for both B-ALL and macrophage cells. 1162 A. Bueno-Costa et al. In conclusion, we report that transdifferentiation events indicated otherwise in a credit line to the material. If material is not in the context of hematopoietic lineage plasticity, such as included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted the pre-B-ALL lineage conversion to macrophage studied use, you will need to obtain permission directly from the copyright herein, involve DNA methylation shifts that not only affect holder. To view a copy of this license, visit http://creativecommons. CpG sites in lineal proximity to genes, but also incur epi- org/licenses/by/4.0/. genetic changes in long-range interactor sequences derived from the 3D genome architecture of the living cell. These References results may help to improve our knowledge of the critical determinant for cell type specification and to understand 1. Chao MP, Seita J, Weissman IL. Establishment of a normal hematopoietic and leukemia stem cell hierarchy. Cold Spring Harb what goes awry in hematological malignancies that, in Symp Quant Biol. 2008;73:439–49. response to pharmacological or cellular therapies, undergo 2. Velten L, Haas SF, Raffel S, Blaszkiewicz S, Islam S, Hennig B, lineage switching to develop resistance to the applied et al. Human haematopoietic stem cell lineage commitment is a treatment. continuous process. Nat Cell Biol. 2017;19:271–81. 3. Majzner RG, Mackall CL. Tumor antigen escape from CAR T-cell therapy. Cancer Discov. 2018;8:1219–26. Acknowledgements We thank CERCA Programme/Generalitat de 4. Langstein J, Milsom MD, Lipka DB. Impact of DNA methylation Catalunya for institutional support. ABC is a fellow of the Spanish programming on normal and pre-leukemic hematopoiesis. Semin Ministry of Education and Vocational Training, under FPU contract Cancer Biol. 2018;51:89–100. no. FPU17/02423. ME is an ICREA Research Professor. BMJ is a 5. Xie H, Ye M, Feng R, Graf T. Stepwise reprogramming of B cells Ramon y Cajal fellow (RYC-2016-19655). This work was supported into macrophages. Cell. 2004;117:663–76. by the Health Department PERIS-project no. SLT/002/16/00374 and 6. Bussmann LH, Schubert A, Vu Manh TP, De Andres L, Des- AGAUR-projects no. 2017SGR1080, 2014SGR633, 2017SGR1033, bordes SC, Parra M, et al. A robust and highly efficient immune and 2009SGR1315 of the Catalan Government (Generalitat de Cata- cell reprogramming system. Cell Stem Cell. 2009;5:554–66. lunya); the Spanish Institute of Health Carlos III (ISCIII) project no. 7. Rapino F, Robles EF, Richter-Larrea JA, Kallin EM, Martinez- DTS16/00153, and CIBERONC CB16/12/00312 and CB16/12/00489; Climent JA, Graf T. C/EBPα induces highly efficient macrophage and Ministerio de Economía y Competitividad (MINECO) project transdifferentiation of B lymphoma and leukemia cell lines and nos. SAF2014-55000-R, BFU2014-57466-P, SAF2017-89673-R, impairs their tumorigenicity. Cell Rep. 2017;19:1281. SAF2017-86242-R, and SAF2015-70270-REDT cofinanced by the 8. McClellan JS, Dove C, Gentles AJ, Ryan CE, Majeti R. Repro- European Development Regional Fund, ‘A way to achieve Europe’ gramming of primary human Philadelphia chromosome-positive B ERDF; the Cellex Foundation; and “la Caixa” Bank Foundation (LCF/ cell acute lymphoblastic leukemia cells into nonleukemic macro- PR/GN18/51140001). We thank Dr Thomas Graf for providing the phages. Proc Natl Acad Sci USA. 2015;112:4074–9. transdifferentiation model. 9. Svensson PA, Hägg DA, Jernås M, Englund MC, Hulten LM, Ohlsson BG, et al. Identification of genes predominantly Author contribution Contribution: ABC and ME conceived and expressed in human macrophages. Atherosclerosis. 2004; designed the study; ABC performed all experiments; ABC, DP, and 177:287–90. BMJ analyzed multiomics data; MS performed western blots; HRV, 10. Scott CL, T’Jonck W, Martens L, Todorov H, Sichien D, Soen B, MSG, and LP performed UMI-4C analysis; JMC provided cellular et al. The transcription factor ZEB2 Is required to maintain the models and primary samples; and ABC and ME wrote the manuscript tissue-specific identities of macrophages. Immunity. 2018;49: with contributions and approval from all authors. 312–25. 11. Heinz S, Romanoski CE, Benner C, Glass CK. The selection and Compliance with ethical standards function of cell type-specific enhancers. Nat Rev Mol Cell Biol. 2015;16:144–54. 12. Heyn H, Vidal E, Ferreira HJ, Vizoso M, Sayols S, Gomez A, Conflict of interest ME is a consultant of Ferrer International and et al. Epigenomic analysis detects aberrant super-enhancer DNA Quimatryx. The authors declare that they have no conflict of interest. methylation in human cancer. Genome Biol. 2016;17:11. 13. Benetatos L, Vartholomatos G. Enhancer DNA methylation in Publisher’s note Springer Nature remains neutral with regard to acute myeloid leukemia and myelodysplastic syndromes. Cell Mol jurisdictional claims in published maps and institutional affiliations. Life Sci. 2018;75:1999–2009. 14. Javierre BM, Burren OS, Wilder SP, Kreuzhuber R, Hill SM, Open Access This article is licensed under a Creative Commons Sewitz S, et al. Lineage-specific genome architecture links Attribution 4.0 International License, which permits use, sharing, enhancers and non-coding disease variants to target gene pro- adaptation, distribution and reproduction in any medium or format, as moters. Cell. 2016;167:1369–84. long as you give appropriate credit to the original author(s) and the 15. Schwartzman O, Mukamel Z, Oded-Elkayam N, Olivares-Chauvet source, provide a link to the Creative Commons license, and indicate if P, Lubling Y, Landan G, et al. UMI-4C for quantitative and tar- changes were made. The images or other third party material in this geted chromosomal contact profiling. Nat Methods. 2016;13: article are included in the article’s Creative Commons license, unless 685–91.