Open access Original research J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. Bone marrow MSC from pediatric patients with B-­ALL highly immunosuppress T-­cell responses but do not compromise CD19-­CAR T-­ cell activity Samanta Romina Zanetti ‍ ‍ ,1 Paola Alejandra Romecin,1 Meritxell Vinyoles,1 Manel Juan,2 José Luis Fuster,3,4 Mireia Cámos,5,6,7 Sergi Querol,8 Mario Delgado,9 Pablo Menendez1,10,11 To cite: Zanetti SR, Romecin PA, ABSTRACT CD19+ B-­cell precursors.1 Of special interest Vinyoles M, et al. Bone marrow Background Although adoptive transfer of CD19-­directed is the B-­ALL harboring chromosomal rear- MSC from pediatric patients with chimeric antigen receptor (CAR) T-­cells (CD19-­CAR T-­cells) B-­ALL highly immunosuppress rangements of the mixed-­ lineage leukemia achieves high rates of complete response in patients with gene (MLL-­r), which presents unique clinical T-­cell responses but do not B-­cell acute lymphoblastic leukemia (B-­ALL), relapse is and biological features and has a dismal prog- compromise CD19-­CAR common. Bone marrow (BM) mesenchymal stem/stromal T-­cell activity. Journal for nosis compared with other B-­ALL.2 ImmunoTherapy of Cancer cells (BM-­MSC) are key components of the hematopoietic niche and are implicated in B-­ALL pathogenesis and B-­ ALL originates in the bone marrow 2020;8:e001419. doi:10.1136/ jitc-2020-001419 therapy resistance. MSC exert an immunosuppressive (BM),3 and the BM microenvironment effect on T-­cells; however, their impact on CD19-­CAR T-­ (BMM) is particularly important in the patho- SRZ and PAR contributed cell activity is understudied. genesis of the B-­ALL given the high depen- equally. Methods We performed a detailed characterization of dence of leukemic cells on external factors BM-­MSC from pediatric patients with B-­ALL (B-­ALL BM-­ for proliferation and survival.4 The BM is also Accepted 08 August 2020 MSC), evaluated their immunomodulatory properties and the primary site where residual leukemic cells their impact on CD19-­CAR T-­cell activity in vitro using survive during standard chemotherapy,5 6 microscopy, qRT-­PCR, ELISA, flow cytometry analysis and and is the most frequent location of B-­ALL in vivo using a preclinical model of severe colitis and a relapse.7 Mesenchymal stem/stromal cells B-­ALL xenograft model. Results While B-­ALL BM-­MSC were less proliferative (MSC) are major components of the BMM than those from age-­matched healthy donors (HD), the and are involved in the pathogenesis and morphology, immunophenotype, differentiation potential drug resistance of B-­ ALL cells during the and chemoprotection was very similar. Likewise, both BM-­ chemotherapy,8–14 underscoring the active MSC populations were equally immunosuppressive in vitro investigations aimed at unraveling the contri- and anti-­inflammatory in an in vivo model of severe colitis. bution of BM-­MSC to B-­ALL cell response to Interestingly, BM-­MSC failed to impair CD19-­CAR T-­cell treatment. cytotoxicity or cytokine production in vitro using B-­ALL Adoptive transfer of T-­ cells engineered cell lines and primary B-­ALL cells. Finally, the growth of to express an artificial chimeric antigen NALM6 cells was controlled in vivo by CD19-­CAR T-­cells receptor (CAR) targeting a tumor cell © Author(s) (or their irrespective of the absence/presence of BM-­MSC. Conclusions Collectively, our data demonstrate that surface-­specific antigen is a promising new employer(s)) 2020. Re-­use permitted under CC BY-­NC. No pediatric B-­ALL and HD BM-­MSC equally immunosuppress approach for cancer immunotherapy. In commercial re-­use. See rights T-­cell responses but do not compromise CD19-­CAR T-­cell the setting of B-­ALL, CAR T-­cells targeting and permissions. Published by activity. CD19 have generated unprecedented BMJ. results in multiple clinical trials, and ~50% For numbered affiliations see of treated patients are disease-­free after 12 end of article. BACKGROUND months.15 16 Despite these robust clinical Correspondence to Acute lymphoblastic leukemia (ALL) is the responses, however, leukemia relapse after Dr Samanta Romina Zanetti; most common malignancy in childhood, treatment is common and remains a major ​szanetti@​carrerasresearch.​org accounting for ~25% of all childhood cancers.1 challenge for CD19-­CAR T-­cell therapy.17 An Dr Pablo Menendez; B-­cell ALL (B-­ALL) represents ~80% of ALL intriguing result of CD19-­CAR T-­cell therapy ​pmenendez@​carrerasresearch.​ and is an aggressive hematologic malignancy in B-­ALL is that MLL-­r B-­ALL patients show org characterized by the clonal expansion of lower response rates and higher rates of Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 1 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. immune escape.18 The antitumor efficacy and persistence Anti-­CD3 (OKT3) and anti-­CD28 (CD28.2) monoclonal of CD19-­CAR T-­cells within the host depend on complex antibodies (mAbs), 7-­ amino-­ actinomycin D (7-­ AAD), in vivo processes and multicellular interactions and, in phycoerythrin (PE) annexin V apoptosis detection kit and recent years, it has been extensively demonstrated that fluorescein isothiocyanate, PE, peridin chlorophyll, allo- MSC regulate T-­cells by modifying their activation, prolif- phycocyanin (APC)-, PE/ cyanine7 (PE/Cy7)-, brilliant eration and effector functions.19 20 To the best of our violet 421 (BV421), BV510)-­conjugated mAbs specific for knowledge, the immunomodulatory properties of pedi- human CD3, CD19, CD45, CD13, CD73, CD105, CD90, atric B-­ALL-­derived BM-­MSC in vivo, and their impact on CD31, CD34 and isotype-­matched negative control mAbs CAR T-­cell efficacy/resistance in B-­ALL remains unex- were purchased from BD Bioscience, Franklin Lakes, plored. Thus, in the present study, we sought to charac- New Jersey, USA. APC-­conjugated anti-­Nestin mAb was terize in vitro and in vivo the immunomodulatory effects purchased from R&D System, Abingdon, UK. The B-­ALL of B-­ALL BM-­MSC on T-­cells and CD19-­CAR T-­cells, and cell lines SEM and NALM6 were obtained from DSMZ cell compared them with healthy donors (HD) BM-­MSC. A line bank, Braunschweig, Germany. Luciferase (Luc)/ better understanding of the interactions between T-­cells GFP-­expressing-­NALM6 cells were kindly provided by RJ and the B-­ALL BMM will help improving the efficiency of Brentjens (MSKCC, New York, USA). current immunotherapies in B-­ALL. Human samples Peripheral blood (PB) mononuclear cells (PBMC) MATERIALS AND METHODS were isolated from buffy coats of healthy volunteers by Reagents, drugs, antibodies and cell lines Ficoll-­Hypaque gradient centrifugation (GE Healthcare, Advanced DMEM, IMDM, RPMI-1640, L-­ glutamine Chicago, Illinois, USA).21 Buffy coats were obtained from and penicillin-­ streptomycin (P/S), insulin-­ transferrin-­ the Catalan Blood and Tissue Bank (BST) on Institu- selenium (ITS) were purchased from Gibco/Invitrogen, tional Review Board approval (HCB/2018/0030). BM Waltham, Massachusetts, USA. StemSpan SFEM was B-­ALL aspirates were obtained from pediatric patients purchased from STEMCELL Technologies StemCell with ETV6/RUNX1 or MLL-­r B-­ALL (table 1). BM aspi- Technologies, Vancouver, Canada. Phosphate buffered rates from age-­matched HD were obtained from the BST. saline was purchased from Merck Life Science SL, Madrid, Human samples were obtained after written informed Spain. MSC osteogenic and adipogenic differentiation consent in accordance with the Declaration of Helsinki. medium were purchased from Promocell, Heidelberg, Alemania. Human (h) Stem Cell Factor (SCF), hFMS-­like Isolation, expansion and characterization of BM-MSC tyrosine kinase 3 ligand (FLT3-­L), human interleukin-3 Mononuclear cells (MC) from BM samples were isolated (hIL-3), hIL-7 and hIL-15 were purchased from Miltenyi by centrifugation on Ficoll-­Paque Plus density gradient. Biotec, Bergisch Gladbach, Germany. Nitro blue tetra- MCs were seeded at 2×105 cells/cm2 in advanced DMEM zolium chloride (NBT), alizarin red, oil red-­O, dexameth- with 10% heat-­ inactivated FBS, L-­ glutamine and P/S asone, L-­phytohemagglutinin (PHA-­L) and fetal bovine (complete advanced DMEM medium) and maintained serum (FBS) were purchased from Sigma-­ Aldrich, St. in a humidified atmosphere with 5% CO2 at 37°C 2 Louis, Missouri, USA. L-­asparaginase was purchased from days, after which non-­adherent cells were collected and Jazz Pharmaceuticals Iberia SL, Barcelona, Spain. Cell- frozen and fresh medium was added. Adherent cells at Trace Violet Cell Proliferation Kit and CellTrace CFSE >85% confluence were trypsinized and replated at 5×103 Cell Proliferation Kit were purchased from Fisher Scien- cells/cm2.22 The resulting BM-­MSC cultures were main- tific SL, Madrid, Spain. Cell fixation and permeabiliza- tained in complete Advanced DMEM medium. BM-­MSC tion Kit was purchased from Nordic-­Mubio, Derio, Spain. cultures were assessed daily for changes in growth rate Table 1 Biological and molecular characteristics of HD and B-­ALL patients HD/patient Disease stage Cytogenetic Molecular Age (years) Gender Blasts (%) HD #1 N.A. N.A. N.A. 9 Female N.A. HD #2 N.A. N.A. N.A. 10 Female N.A. HD #3 N.A. N.A. N.A. 7 Male N.A. Pt #1 Diagnosis t(12;21) ETV6/RUNX1 5 Male 84 Pt #2 Diagnosis t(12;21) ETV6/RUNX1 2 Male 80 Pt #3 Diagnosis t(9;11) MLL/AF9 3 Male 18 Pt #4 Diagnosis t(1;11) MLL/EPS15 1 Female 80 Pt #5 Relapse t(4;11) MLL/AF4 4 Female 81 B-­ALL, B-­cell acute lymphoblastic leukemia; HD, healthy donors; N.A, not aplicable. 2 Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. and morphology. All experiments were performed with Quantification of the proinflammatory cytokines IL-2, cells harvested between passage (P)2 and P5. interferon-γ (IFN-γ) and tumor necrosis factor α (TNF-α) Growth kinetics of BM-­MSC were measured as cumula- was measured by ELISA using the OptEIA Human ELISA tive Population Doublings at each passage, as described.23 Kit (BD Biosciences) on supernatants from the afore- An equal number of BM-­MSC (5×103 cells/cm2) at P1 mentioned BM-­ MSC:PBMC cocultures. Unstimulated were initially plated for each patient/donor; cells were PBMC:BM-­ MSC cocultures were processed in parallel then harvested every 5 days, stained with trypan blue, as a control. ELISA determinations were performed in counted and re-­plated at 5×103 cells/cm2. The procedure triplicate. was repeated until P6. BM-­MSC characterization was performed following In vitro chemoresistance of B-ALL cell lines the guidelines proposed by the International Society for HD and B-­ALL BM-­MSC were seeded at 1×104 cells/well in Cellular Therapy (ISCT).24 The immunophenotype of flat-­bottom 96-­well plates containing complete Advanced cultured BM-­MSC was assessed by FACS. A total of 1×106 DMEM medium at 37°C for 24 hours. The next day, the cells were incubated for 30 min at 4°C in the dark. Data BM-­MSC culture medium was removed and SEM cells acquisition and analysis were performed using a FACS- (1×105 cells/well) were added to the adherent BM-­MSC Canto-­II flow cytometer equipped with FACSDiva soft- culture at a BM-­MSC:SEM ratio of 1:10 in IMDM/10% ware (BD Biosciences). Differentiation was evaluated heat-­inactivated FBS and P/S, in the absence or presence by plating BM-­MSC in specific differentiation induction of dexamethasone (10 UI) or L-­asparaginase (1 µM). Two medium for 3–4 weeks. For osteogenic differentiation, days after, non-­adherent cells were collected, washed and BM-­MSCs were stained to determine alkaline phospha- stained with an anti-­CD19 mAb, 7-­AAD and AnnexinV to tase (AP) activity using NBT and for calcium deposits determine cell apoptosis by FACS.27 with alizarin red.25 For adipogenic differentiation, BM-­MSCs were stained with oil red-­O to detect the lipid CD19-CAR vector, lentiviral production, T-cell transduction, droplets.22 Differentiation capacity was further quanti- activation and expansion fied by quantitative RT-­ PCR (qRT-­ PCR) and RNA was We used our clinically validated pCCL second-­generation isolated from differentiated BM-­MSC cultures using the lentiviral CD19-­CAR backbone, which contains a human Maxwell RSC simplyRNA Cells Kit (Promega, Madison, CD8 transmembrane domain and human 4-­ 1BB and Wisconsin, USA). RT-­PCR analysis was conducted using CD3ζ endodomains.28 We incorporated a GFP reporter the SYBR Green PCR Master Mix (Fisher Scientific SL) gene after a 2A ribosomal skip sequence at the C-­terminal with primers for specific adipogenic differentiation and CAR sequence to evaluate the transduction efficiency osteogenic differentiation transcription factors. Analysis and the tracking of CAR expression. CAR-­ expressing and mRNA quantification analysis were performed on viral particles pseudotyped with VSV-­G were generated the CFX384 Real-­ Time System (Bio-­ Rad Laboratorie, by transfection of HEK 293 T cells with pCCL, VSV-­ G Hercules, California, USA). Reactions were performed in and psPAX2 vectors using polyethylenimine (Poly- triplicate and gene expression values were normalized to sciences, Warrington, Pennsylvania, USA). Supernatants those of HPRT. Gene-­specific primer sequences for qRT-­ were collected at 48 and 72 hours after transfection and PCR are listed in online supplementary Table S1. concentrated by ultracentrifugation. T-­cells were activated by plate-­coating with anti-­CD3 and T-cell proliferation and cytokine secretion anti-­CD28 mAbs in complete RPMI medium for 2 days HD and B-­ALL BM-­MSC were seeded at 1×104 cells/well in and were then transduced with a CAR-­expressing lenti- flat-­bottom 96-­well plates containing complete Advanced virus at a multiplicity of infection of 10 in the presence of DMEM medium at 37°C for 24 hours. The next day, we hIL-7 and hIL-15 (10 ng/mL).29 T-­cells were expanded in evaluated T-­cell proliferation using the CellTrace CFSE complete RPMI medium plus hIL-7 and hIL-15 for up to Cell Proliferation Kit. To do this, PBMCs were labeled 6 days. CAR transduction efficiency in T-­cells was analyzed with 5 µM CellTrace CFSE for 20 min at 37°C and then by FACS. washed three times with RPMI 1640/1% heat-­inactivated FBS.26 Subsequently, the BM-­MSC culture medium was In vitro CD19-CAR T-cell, cytotoxicity assays and cytokine removed and CellTrace CFSE-­labeled cells were added to release determination the adherent BM-­MSC culture at a BM-­MSC:PBMC ratio For cytotoxicity assays, BM-­MSCs were seeded at 1×104 of 1:5 and 1:10 in RPMI-1640 with10% heat-­inactivated cells/well in flat-­ bottom 96-­ well plates containing FBS and P/S (complete RPMI medium) (or plated complete Advanced DMEM medium and incubated at alone), in the absence or presence of PHA-­L (1 µg/mL). 37°C for 24 hours. The next day, target cells (B-­ALL cell After 3 and 6 days of culture, non-­adherent cells were lines or primary B-­ALL cells, 1×105 target cells/well) were collected, washed and stained with an anti-­CD3 mAb and incubated with CD19-­CAR T-­cells at an Effector:Target 7-­AAD. Viable cells were gated as 7-­AAD-. Cell division (E:T) ratio of 1:1, or alone, for the indicated periods in the was measured using dye dilution on the CD3+ population absence/presence of BM-­MSC. Cell lines were cultured in by FACS and the number of proliferating cells was deter- RPMI complete medium and primary cells were cultured mined by gating on the CD3+CFSElow subset.26 in StemSpan supplemented with 20% heat-­ inactivated Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 3 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. FBS, P/S, ITS, hSCF (100 ng/mL), hFLT3-­ L (100 ng/ Maine, USA) were bred and housed under pathogen-­free mL), hIL-3 (10 ng/mL) and hIL-7 (10 ng/mL). After 2 conditions. NSG mice were intratibial (IT)-­injected with and 6 days of culture, non-­adherent cells were collected, 1×106 NALM6-­Luc+ cells alone or together with 3×105 washed and stained with anti-­CD3, anti-­CD19, anti-­CD10, BM-­MSC,27 followed 4 days later by an intravenous infu- anti-­CD22 mAbs and 7-­AAD. CAR T-­cell-­mediated cyto- sion of 4×106 CD19-­CAR T-­cells. Mice were followed up toxicity was determined by analyzing the residual living weekly by bioluminescence (BLI) using an in vivo imaging (CD10+CD22+7-­AAD-) target cells at each time point. BD system (Lumina III; Perkin-­Elmer, Wlatham, Massachu- TruCount absolute count tubes (BD Biosciences) were setts, USA). Mice were sacrificed at day 15 (controls) used for absolute cell counting. Quantification of the and day 35 (CAR-­treated), and PB and BM samples were pro-­inflammatory cytokines IL-2, IFN-γ and TNF-α was collected and analyzed by FACS to assess leukemic burden measured by ELISA using the OptEIA Human ELISA Kit and CAR T-­cell persistence. (BD Biosciences) on supernatants harvested after 2 days of coculture at a 1:1 E:T ratio. ELISA determinations were Statistical analysis performed in triplicate. Data were plotted as mean±SEM. One-­way analysis of vari- ance with Tukey’s post hoc test was used when three or Induction and treatment of experimental acute severe colitis more experimental groups were compared and Student’s A well-­established experimental mouse model for bowel paired t-­test was used when two experimental groups were inflammatory disease/acute colitis was employed.30 31 The compared. All analyzes were performed with Prism soft- in vivo procedure was approved by the Ethics Committee ware, V.8.0 (GraphPad Prism Software, San Diego, Cali- of the Spanish Council of Scientific Research. To fornia, USA). induce acute colitis, 2,4,6-­trinitrobenzene sulfonic acid (TNBS, 2 mg; Sigma) in 50% ethanol (100 µL) was administered intrarectally in male BALB/c mice aged RESULTS 6–8 weeks (n=10/group). Control mice received 50% Expansion and characterization of bona fide BM-MSC from ethanol alone. At 4 hours after TNBS infusion, animals pediatric patients with B-ALL were treated intraperitoneally with medium alone, with We isolated MSC from BM aspirates of 5 pediatric B-­ALL, 1×106 HD BM-­MSCs (n=3) or with 1×106 B-­ALL BM-­MSC (2 ETV6/RUNX1 and 3 MLL-­r), and we compared their (MLL-­r B-­ALL n=3), resuspended in 200 µl of medium. morphology, immunophenotype, growth kinetics and Animals were monitored daily for the appearance of diar- differentiation potential with that of BM-­MSC from three rhea, body weight loss, and survival. Colitis signs were pediatric HD counterparts using the standardized criteria scored based on stool consistency and rectal bleeding by outlined by the ISCT.24 Table 1 shows the main biolog- a blinded observer using the following scale: 0=normal ical and molecular/cytogenetic characteristics of B-­ALL stool appearance, 1=slight decrease in stool consistency, patients and HD. Adherent cell cultures were success- 2=moderate decrease in stool consistency, 3=moderate fully established from all samples using standard MSC decrease in stool consistency and presence of blood in culture conditions.24 BM-­MSC from both groups adhered stools, 4=severe watery diarrhea and moderate/severe to plastic and showed homogeneous spindle-­ shaped blooding in stools. At day 9 or immediately after the fibroblast-­like morphology (figure 1A, online supplemen- death of the animal, the colons were collected and eval- tary figure S1A). FACS analysis at P3 revealed that both uated for macroscopic damage (scale 0–8) based on the BM-­MSC groups display a bona fide MSC immunophe- grade of tissue adhesion, presence of ulceration and wall notype positive for CD105, CD90, CD73 and CD13 and thickness in a blinded fashion by a researcher: ulceration Nestin, while negative for hematopoietic (CD34 and (0=normal appearance, 1=focal hyperemia, no ulcers, CD45) and endothelial (CD31) cell markers (figure 1B, 2=ulceration without hyperemia or bowel wall thickening, online supplementary figure S1B).22 When we compared 3=ulceration with inflammation at 1 site, 4=two or more their expansion potential in vitro under identical culture sites of ulceration and inflammation, 5=major sites of conditions, HD BM-­ MSC showed a significantly supe- damage extending >1 cm along length of colon), adhe- rior proliferative potential than that of B-­ALL BM-­MSC sions (0=no adhesions, 1=minor adhesions, colon can be (figure 1C). We next evaluated the differentiation poten- easily separated from the other tissues, 2=major adhe- tial of P3 BM-­MSC cultures by exposing them to adipo- sions); and thickness (maximal bowel wall thickness, in genic and osteoblastic induction conditions. Both groups mm, measured with a caliper). PB was collected from the were similar with regard to adipogenic differentiation, tail vein (day 4) and by cardiac puncture (day 9) and the measured qualitatively by the formation of lipid droplets levels of proinflammatory cytokines were determined detected by oil red-­O staining; and to osteoblast differen- in sera by specific sandwich ELISA using capture/bioti- tiation, revealed by AP activity staining and by the pres- nylated detection antibodies from PreproTech. ence of calcium deposits detected by Alizarin Red staining (figure 1D). Adipogenic and osteoblastic differentia- In vivo CD19-CAR T cell-mediated cytotoxicity assay tion capacity was also confirmed quantitatively by qPCR, Ten-­week-­old non-­obese diabetic Cg-­Prkdcscid Il2rgtm1Wjl/ which showed a comparable and significant increase in SzJ (NSG) mice (The Jackson Laboratory, Bar Harbor, the expression of early and late master regulator genes 4 Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. Figure 1 Characterization of BM-­MSC from pediatric patients with B-­ALL and age-­matched HD. (A) Representative phase-­ contrast morphology of BM-­MSC. (B) Immunophenotype of expanded BM-­MSC at P3. Expression of CD105, CD90, CD73, CD13, nestin, CD45, CD34 and CD31 was analyzed by FACS. Shown are representative FACS histograms; gray histograms represent isotype-­matched negative control mAb and blue histograms represent mAb-­specific stained cells. (C) In vitro proliferation of BM-­MSC calculated as population doubling over six passages. (D, E) Osteogenic and adipogenic differentiation capacity of BM-­MSC. (D) Left panel, oil red-­O staining indicative of adipogenic differentiation capacity. Right panel, Alizarin red staining and detection of AP alkaline phosphatase activity with NTB, indicative of osteogenic differentiation capacity. (E) Relative quantification of mRNA expression of the adipogenic transcription factors C/EBP-α and PPAR-γ and the osteogenic transcription factors BMP2 and RUNX2 by qRT-­PCR. Data are shown as mean±SEM. *P<0.05, **p<0.01, ***p<0.001; one-­way ANOVA with Tukey’s post hoc test. HD BM-­MSC n=3 and B-­ALL BM-­MSC n=5. ANOVA, analysis of variance; B-­ALL, B-­cell acute lymphoblastic leukemia; BM-­MSC, bone marrow-­mesenchymal stem/stromal cells; HD, healthy donors; NTB, nitro blue tetrazolium. Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 5 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. Figure 2 In vitro immunomodulatory properties of BM-­MSC from pediatric patients with B-­ALL and age-­matched HD on T-­cells. (A) Left panel, percentage of proliferating T-­cells measured as percentage of CFSE+ T-­cells is shown. CellTrace CFSE-­ labeled PBMC (n=3 independent PBMC) was stimulated with PHA-­L in the absence/presence of BM-­MSC for 6 days at two different BM-­MSC:PBMC ratios (1:5 and 1:10). Right panel, representative FACS histograms of CellTrace CFSE-­labeled PBMC: R1 identifies non-­proliferating CFSE++ cells, and R2 identifies CFSElow proliferating cells. (B) Secretion of the proinflammatory cytokines IL-2, IFN-γ and TNF-α in cell-­culture supernatants after 6 days at a BM-­MSC:PBMC ratio of 1:10. Data are shown as mean±SEM. ***P<0.001, ****p<0.0001; one-­way ANOVA with Tukey’s post hoc test. HD BM-­MSC n=3 and B-­ALL BM-­MSC n=5. ANOVA, analysis of variance; B-­ALL, B-­cell acute lymphoblastic leukemia; BM-­MSC, bone marrow-­mesenchymal stem/stromal cells; HD, healthy donors; IFN-γ, interferon-γ; IL-2, interleukin-2; PBMC, peripheral blood mononuclear cell; TNF-α, tumor necrosis factor α. of adipogenic (CEBPα, PPAR-γ) and osteogenic (BMP2, B-­ALL BM-­MSC are equally immunosuppressive, as previ- RUNX2) differentiation (figure 1E). ously described.32 33 B-ALL BM-MSC immunosuppress T-cell response B-ALL BM-MSC exert anti-inflammatory effects in a MSCs are widely recognized for their immunomodu- preclinical model of severe acute colitis latory potential, including the inhibition of allogenic Having confirmed the immunosuppressive properties of cell proliferation and the production of proinflam- T-­ B-­ALL BM-­MSC in vitro, we wished to test their ability to matory cytokines.19 20 We, therefore, monitored PHA-­L-­ influence T-­cell functions in vivo. To do this, we used a well-­ stimulated T-­cell division in the absence or presence of established preclinical model of acute colitis (figure 3A) BM-­MSC in vitro. In line with published findings,32 33 we that shares clinical, histopathological and immunological found that HD BM-­MSC strongly inhibited T-­cell prolifer- features with Crohn’s disease.30 31 As expected, TNBS-­ ation in a dose-­dependent manner (figure 2A). Of note, treated mice developed severe, acute illness character- comparable inhibition of T-­cell proliferation was exerted ized by substantial (~20%) and sustained body weight by B-­ALL BM-­MSC (figure 2A). We next analyzed these loss (figure 3B), bloody diarrhea, rectal prolapse and supernatants to test whether BM-­MSC also regulate pro-­ pancolitis, accompanied by extensive wasting syndrome inflammatory cytokine secretion. The analysis of super- (figure 3C), which caused 25% mortality over a 9-­day natants showed that the levels of IL-2, IFN-γ and TNF-α period (figure 3D). Macroscopic examination of colons were comparably and significantly lower in HD and revealed profound signs of inflammation, hyperemia, B-­ALL BM-­MSC cocultures than in PBMC-­only controls ulceration and shortening (figure 3E). By contrast, mice (figure 2B). Overall, these results show that both HD and that were treated with either HD or B-­ ALL BM-­ MSC 6 Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. Figure 3 In vivo anti-­inflammatory properties of BM-­MSC from pediatric patients with B-­ALL and HD in a model of acute colitis. (A) Scheme of the experimental design. Acute colitis was induced in mice by intrarectal administration of TNBS. After 4 hours, mice (n=5–10/group) were injected intraperitoneally with medium or with HD or B-­ALL BM-­MSC. Body weight and survival were followed up for 9 days. (B) Daily change in body weight loss relative to day 0 (induction of colon damage). (C) Survival over a 9-­day follow-­up. (D) Colitis score determined at day 3 and 5. (E) Macroscopic damage score was evaluated in colons at sacrifice (day 9). (F) Serum levels of IL-6 and TNF-α at day 4 and 9. Data are shown as mean±SEM. ***P<0.001, ****p<0.0001; one-­way ANOVA test with Tukey’s post hoc test. HD BM-­MSC n=3 and B-­ALL BM-­MSC n=3. ANOVA, analysis of variance; B-­ALL, B-­cell acute lymphoblastic leukemia; BM-­MSC, bone marrow-­mesenchymal stem/stromal cells; HD, healthy donors; IL-6, interleukin-6; TNBS, 2,4,6-­trinitrobenzene sulfonic acid; TNF-α, tumor necrosis factor α. were largely protected from colitis, as evidenced by a that B-­ALL BM-­MSC can suppress inflammation in vivo to significant recovery of their body weight loss within 9 significantly protect mice against severe acute colitis. days (figure 3B), and by the improvement in the wasting syndrome and the signs of colon inflammation, which was BM-MSC do not compromise CD19-CAR T-cell activity in vitro reflected in the significantly lower mortality and colitis and in vivo score as compared with control animals (figure 3C–E). While it is well established the contribution of BM-­MSC In accord with their anti-­inflammatory effects, treatment in promoting survival, proliferation and chemotherapy with either HD or B-­ALL BM-­MSC decreased the levels of resistance in B-­ALL8–14 (online supplementary figure S2), the proinflammatory cytokines TNF-α and IL-6 in the sera the potential impact of B-­ALL BM-­MSC on CD19-­CAR of colitic mice (figure 3F). Collectively, our findings show T-­cell therapy has not been explored. Moreover, CAR Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 7 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. Figure 4 In vitro cytotoxicity of CD19-­CAR T-­cells against B-­ALL cell lines in the absence/presence of BM-­MSC. (A) Achematic of the experimental design for assessing the cytotoxicity of CD19-­CAR T-­cells against NALM6 or SEM B-­ALL cell lines in the absence or presence of BM-­MSC. (B, C) Absolute counts of alive target cells NALM6 (B) and SEM (C) measured by FACS after 2 and 6 days at 1:1 E:T ratio in the absence/presence of BM-­MSC. Data are shown as mean±SEM. ****P<0.0001; one-­way ANOVA test with Tukey’s post hoc test. HD BM-­MSC n=3 and B-­ALL BM-­MSC n=5. ANOVA, analysis of variance; B-­ALL, B-­ cell acute lymphoblastic leukemia; BM-­MSC, bone marrow-­mesenchymal stem/stromal cells; E:T, Effector:Target; HD, healthy donors. T-­cells are likely to experience the same loss of function/ (figure 4A). We found that the number of CD19-­CAR-­ persistence as T-­cells in an immunosuppressive B-­ALL resistant NALM6 or SEM cells cultured in the presence BMM. Consequently, we next explored the influence of either HD or B-­ALL BM-­MSC was very similar to that of BM-­MSC on CD19-­CAR T-­cell activity. We first evalu- of target cells exposed to CD19-­ CAR T-­ cells alone, a ated in vitro the impact of HD and B-­ALL BM-­MSC on non-­protective role of BM-­MSC on NALM6 (figure 4B) CD19-­CAR T-­cell performance. We generated CD19-­CAR and SEM (figure 4C) cell survival in CD19-­CAR T-­cell T-­cells following a standard protocol with anti-­CD3/anti-­ therapy, at least in vitro. We additionally studied the ex CD28 plus hIL-7 and hIL-15,29 and studied the effects of vivo impact of B-­ALL BM-­MSC on CD19 CAR T-­cell cyto- BM-­MSC on CD19-­CAR T-­cell cytotoxicity against NALM6 toxicity against autologous primary B-­ALL blasts, which and SEM cells cultured at a 1:1 E:T ratio for 2 and 6 days partially replicates the pathophysiological conditions of 8 Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. Figure 5 In vitro cytotoxicity of CD19-­CAR T-­cells against primary B-­ALL blasts in the presence/absence of patient-­matched BM-­MSC. (A) Schematic of the experimental design for assessing the cytotoxicity of CD19-­CAR T-­cells against primary B-­ ALL blasts in the absence or presence of patient-­matched (Pt-­m) BM-­MSC. (B) Absolute counts of alive primary B-­ALL blasts measured by FACS after 2 days cytotoxicity assay at a 1:1 E:T ratio in the absence/presence of Pt-­m BM-­MSC. (C) Secretion of IL-2, IFN-γ and TNF-α by CD19-­CAR T-­cells (n=3 independent PBMC) after 2 days exposure to primary B-­ALL cells at a 1:1 E:T ratio in the absence/presence of Pt-­m BM-­MSC. Data are shown as mean±SEM B-­ALL BM-­MSC n=5. B-­ALL, B-­cell acute lymphoblastic leukemia; BM-­MSC, bone marrow-­mesenchymal stem/stromal cells; CAR, chimeric antigen receptor; E:T, Effector:Target; IFN-γ, interferon-γ; IL-2, interleukin-2; PBMC, peripheral blood mononuclear cell; TNF-α, tumor necrosis factor α. patients. We performed cytotoxicity assays using B-­ALL Finally, we evaluated the impact of HD or B-­ALL BM-­MSC blasts and patient-­matched derived BM-­MSC from five on CD19 CAR T-­cell cytotoxicity in vivo, using a mouse pediatric B-­ALL patients (figure 5A). Irrespective of the xenograft model with NALM6 cells. To recapitulate the difference in survival of B-­ALL primary cells, after 2 days interactions between the BM-­MSC and leukemic cells, 1×106 of culture the number of alive/resistant primary blasts NALM6-­Luc+ cells were IT injected into NSG mice alone or from the five patients after CD19-­CAR T-­cell treatment together with 3×105 HD or B-­ALL BM-­MSC, followed 4 days was very similar in the absence or presence of autologous later by 5×106 CAR-­T cells (figure 6A). Mice were followed BM-­MSC (figure 5B), consistent with the results obtained up weekly by BLI. CD19-­CAR T-­cell-­untreated mice showed using B-­ALL cell lines (figure 4). In addition, we found massive leukemic burden and succumbed quickly to the no significant differences in the levels of IL-2, IFN-γ and disease (figure 6B and C). In contrast, mice treated with TNF-α in culture supernatants (figure 5C), confirming CD19-­CAR T-­cells, irrespective of whether they were trans- that BM-­MSC from pediatric B-­ALL do not promote resis- planted or not with BM-­ MSC, similarly controlled the tance to CD19-­CAR T-­cells in vitro. disease (figure 6B and C). FACS analysis of leukemic burden Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 9 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. Figure 6 In vivo cytotoxicity of CD19-­CAR T-­cells in an NALM6 mouse xenograft model in the absence/presence of BM-­MSC. (A) Schematic of the transplantation of B-­ALL cells and BM-­MSC followed by CD19-­CAR T-­cell infusion after 4 days to assess in vivo the influence of BM-­MSC on CD19-­CAR T-­cell activity. (B) Leukemic burden monitored weekly (until sacrifice of control mice) by BLI. (C) Total RADIANCE quantification (p/sec/cm2/sr) of mice shown in B. (D) Leukemic burden analyzed by FACS in PB and BM at the time control mice were sacrificed (day 15). (E) Representative FACS dot plots identifying target cells (red) and T-­cell persistence (blue) in BM at the time control mice were sacrificed. (F) Long-­term (35 days) BLI follow-­up of CD19-­CAR T-­ cell-­treated mice in the absence/presence of BM-­MSC. (G) Total RADIANCE quantification (p/sec/cm2/sr) of mice shown in F). Each dot represents a mouse. Data are shown as mean±SEM. *P<0.05, ***p<0.001, ****p<0.0001; a paired two-­tailed t-­test was used when two experimental groups were compared, and one-­way ANOVA test with Tukey’s post hoc test was used when three experimental groups were compared. ANOVA, analysis of variance; B-­ALL, B-­cell acute lymphoblastic leukemia; BM-­MSC, bone marrow-­mesenchymal stem/stromal cells; CAR, chimeric antigen receptor; E:T, Effector:Target; PB, peripheral blood. 10 Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. in PB and BM at day 15 confirmed the BLI data (figure 6D resistance of B-­ALL to CD19-­CAR T-­cell therapy. We found and E). All those mice treated with CD19-­CAR T-­cells with that CD19-­CAR T-­cells can efficiently eliminate both B-­ALL or without HD/B-­ALL BM-­MSCs in which the disease was cell lines and B-­ALL primary cells in vitro regardless the controlled at the time of sacrificing the untreated controls absence/presence of either HD or B-­ALL BM-­MSC. More- (day 15) were followed up to 35 days, and found that neither over, when we analyzed the proinflammatory cytokines HD nor B-­ALL BM-­MSC impaired long-­term CD19-­CAR secreted in the cocultures of B-­ ALL blasts with patient-­ T-­cell activity (figure 6F,G). Collectively, neither HD nor matched BM-­MSC derived from pediatric B-­ALL patients B-­ALL BM-­MSCs apparently contribute to CD19-­CAR T-­cell we found no significant differences as compared with those resistance. in the absence of BM-­MSC, suggesting that BM-­MSC from pediatric B-­ALL patients do not compromise CD19-­CAR T-­cell efficacy in vitro. These findings concur with a previous DISCUSSION study by Campana and colleagues.41 We also provide the There is increasing evidence that the BMM can influence first demonstration, to the best of our knowledge, that in leukemia progression.34–37 In this context, BM-­MSCs are vivo HD or B-­ALL BM-­MSC do not contribute to CD19-­CAR considered important players in the BM, promoting prolif- T-­cell resistance using a mouse xenograft model which reca- eration, survival, migration, vascularization and chemoresis- pitulates the interactions between the BM-­MSC and B-­ALL tance, and suppressing immune cells, all key processes for cells. leukemia progression and immune evasion.19 20 36 37 B-­ALL In conclusion, our data show that pediatric B-­ ALL is malignant proliferation of CD19+ B-­cell precursors,1 and BM-­ MSC and age-­ matched HD BM-­ MSC equally immu- BM-­MSCs have been shown to promote B-­ALL progres- nosuppressive T-­ cell response, but do not compromise sion and chemoresistance by several mechanisms.4 8–14 38 39 CD19-­CAR T-­cell activity either in vitro or in vivo, at least Here, we obtained BM-­MSC from pediatric patients with within the experimental constraints of this study. Although B-­ALL and compared their functional and immunological characteristics with those of BM-­MSC from age-­matched this study indicates that specific immunobiological proper- HD. We show that both HD and B-­ALL BM-­MSC display ties of BM-­MSCs do not seem to underlie the resistance of comparable morphology, immunophenotype and differen- B-­ALL cells to CD19-­CAR T-­cells, we must bear in mind that tiation potential in vitro; however, B-­ALL BM-­MSCs are less our study does not unequivocally assess the contribution proliferative. This is in line with defects in B-­ALL BM-­MSC of the intact BMM, composed of different cell types and expansion reported by others.32 33 In addition, we show that cellular interactions, to CD19-­CAR T-­cells mediated cyto- B-­ALL and HD BM-­MSC suppress mitogenic T-­cell prolifer- toxicity. Further in vivo studies using humanized animal ation by PHA-­L in a dose-­dependent manner, and block the models are necessary to elucidate the impact of the BMM secretion of pro-­inflammatory cytokines by T-­cells in cocul- on the response to adoptive cell therapy with CD19-­CAR ture. Overall, these data indicate that the typical in vitro T-­cells. biological properties of BM-­MSC, including their immuno- suppressive effects, are unchanged in patients with B-­ALL, Author affiliations 1 Josep Carreras Leukemia Research Institute, Barcelona, Spain which is in agreement with earlier reports.32 33 2 Servicio de Inmunología, Hospital Clínico de Barcelona, Hospital Clínico de The in vivo anti-­ inflammatory properties of BM-­ MSC Barcelona, Barcelona, Spain from patients with hematological malignancies have been 3 Sección de Oncohematología Pediátrica, Hospital Clínico Universitario Virgen de la scarcely investigated.40 In the present study, we describe, for Arrixaca, El Palmar, Murcia, Spain 4 the first time to our knowledge, the in vivo anti-­inflammatory Instituto Murciano de Investigación biosanitaria, Murcia, Spain 5 effects of BM-­MSC from pediatric patients with B-­ALL using Hematology Laboratory, Hospital Sant Joan de Déu, University of Barcelona, Barcelona, Spain a preclinical model of acute colitis. TNBS-­ treated mice 6 Institut de Recerca Hospital Sant Joan de Déu Barcelona, Barcelona, Spain develop a severe acute illness while mice treated with HD or 7 Centro de Investigación Biomédica en Red de Enfermedades Raras (CIBERER), B-­ALL BM-­MSC are protected against colitis. Also, the levels ISCIII, Barcelona, Spain of master pro-­inflammatory cytokines in the serum of colitis 8 Blood and Tissue Bank, Barcelona, Spain 9 mice treated with HD or B-­ALL BM-­MSC are diminished. Instituto de Parasitología y Biomedicina López-­Neyra-­CSIC, Barcelona, Spain 10 Collectively, our results show that HD and B-­ALL BM-­MSC Institució Catalana de Recerca i Estudis Avançats, Barcelona, Spain 11 Centro de Investigación Biomédica en Red de Cáncer (CIBERONC), ISCIII, have a similar capacity to dampen inflammation in a well-­ Barcelona, Spain characterized in vivo model. Given the accumulating evidence that BM-­MSCs protect Acknowledgements We are indebted to our laboratory colleagues and especially B-­ALL cells from chemotherapy in vitro,8–14 we confirmed to Ana Belén Carrillo (GENYO, Granada, Spain) y Rosa Maria Yañez (CIEMAT, Madrid, that coculturing B-­ALL BM-­MSC with B-­ALL cells reduces Spain) for their technical feedback and constructive discussions. We thank CERCA/ apoptosis in leukemic cells in response to dexametha- Generalitat de Catalunya and Fundació Josep Carreras-­Obra Social la Caixa for their institutional support. sone and L-­ asparaginase, with HD BM-­ MSC showing a similar capacity to protect leukemic cells from chemo- Contributors SRZ conceived the study, designed and performed experiments, analyzed data, interpreted data, performed figures and wrote the manuscript. therapy. Whether B-­ALL MSC could protect B-­ALL cells PAR designed and performed experiments, analyzed data, interpreted data, during CD19-­CAR T-­cell therapy was, however, not known. performed figures and reviewed the paper. MV performed experiments. JLF, MC, Thus, we evaluated whether BM-­MSC participated in the SQ and MJ provided human samples. MD performed experiments, analyzed data Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 11 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. and interpreted data. PM, conceived the study, designed experiments, wrote the leukemia cells and is a potential therapeutic target. Cancer Lett manuscript, and financially supported the study. 2013;333:9–17. 14 Yu K, Yin Y, Ma D, et al. Shp2 activation in bone marrow Funding Financial support for this work was obtained from the European microenvironment mediates the drug resistance of B-­cell acute Research Council (CoG-2014-646903, PoC-2018-811220), the Spanish Ministry lymphoblastic leukemia through enhancing the role of VCAM-1/VLA- of Economy and Competitiveness (MINECO, SAF2016-­80481R), the Fundación Uno 4. Int Immunopharmacol 2020;80:106008. entre Cienmil, the Obra Social La Caixa (LCF/PR/HR19/52160011), the Leo Messi 15 Rafei H, Kantarjian HM, Jabbour EJ. Recent advances in the Foundation, and the 'Heroes hasta la médula' initiative to PM. SRZ was supported treatment of acute lymphoblastic leukemia. Leuk Lymphoma by a Marie Sklodowska Curie Fellowship (GA 795833). MV is supported by a Juan 2019;60:2606–21. 16 Sermer D, Brentjens R. Car T-­cell therapy: full speed ahead. Hematol de la Cierva fellowship from the MINECO. PM is an investigator of the Spanish Cell Oncol 2019;37:95–100. Therapy cooperative network (TERCEL). 17 Ruella M, Maus MV. Catch me if you can: leukemia escape after Competing interests None declared. CD19-­Directed T cell immunotherapies. Comput Struct Biotechnol J 2016;14:357–62. Patient consent for publication Not required. 18 Gardner R, Wu D, Cherian S, et al. Acquisition of a CD19-­negative Ethics approval The study was approved by the institutional review board myeloid phenotype allows immune escape of MLL-­rearranged B-­ALL from CD19 CAR-­T-­cell therapy. Blood 2016;127:2406–10. of the Ethics Committee on Clinical Research of Clinic Hospital of Barcelona 19 Guerrouahen BS, Sidahmed H, Al Sulaiti A, et al. Enhancing (HCB/2017/0781). Animal studies of experimental acute severe colitis were mesenchymal stromal cell immunomodulation for treating conditions conducted with the approval of the Ethics Committee of the Spanish Council of influenced by the immune system. Stem Cells Int 2019;2019:1–11. Scientific Research and in vivo CD19-­CAR T cell-­mediated cytotoxicity assay was 20 Regmi S, Pathak S, Kim JO, et al. Mesenchymal stem cell therapy for approved by the Barcelona Biomedical Research Park ethics committee (HRH-17- the treatment of inflammatory diseases: challenges, opportunities, 0029-­P1). The in vivo procedure was approved by the local ethics committee (HRH- and future perspectives. Eur J Cell Biol 2019;98:151041. 17–0029 P1). 21 Zanetti SR, Ziblat A, Torres NI, et al. Expression and functional role of α7 nicotinic receptor in human cytokine-­stimulated natural killer (NK) Provenance and peer review Not commissioned; externally peer reviewed. cells. J Biol Chem 2016;291:16541–52. 22 Diaz de la Guardia R, Lopez-­Millan B, Lavoie JR, et al. Detailed Data availability statement Data are available on reasonable request. All data characterization of mesenchymal stem/stromal cells from a large relevant to the study are included in the article or uploaded as online supplementary cohort of AML patients demonstrates a definitive link to treatment information. Data are available on reasonable request. All data relevant to the study outcomes. Stem Cell Reports 2017;8:1573–86. are included in the article or uploaded as online supplementary information. 23 Lechanteur C, Briquet A, Giet O, et al. Clinical-­scale expansion of Open access This is an open access article distributed in accordance with the mesenchymal stromal cells: a large banking experience. J Transl Med 2016;14:145. Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which 24 Viswanathan S, Shi Y, Galipeau J, et al. Mesenchymal stem versus permits others to distribute, remix, adapt, build upon this work non-­commercially, stromal cells: International Society for Cell & Gene Therapy (ISCT®) and license their derivative works on different terms, provided the original work is Mesenchymal Stromal Cell committee position statement on properly cited, appropriate credit is given, any changes made indicated, and the use nomenclature. Cytotherapy 2019;21:1019–24. is non-­commercial. See http://c​ reativecommons.​org/​licenses/​by-​nc/​4.0​ /. 25 Xu X, Li R, Zhou Y, et al. Dysregulated systemic lymphocytes affect the balance of osteogenic/adipogenic differentiation of bone ORCID iD mesenchymal stem cells after local irradiation. Stem Cell Res Ther Samanta Romina Zanetti http://​orcid.​org/​0000-​0002-​9052-​8751 2017;8:71. 26 Borge M, Nannini PR, Morande PE, et al. Cxcl12 is a costimulator for CD4+ T cell activation and proliferation in chronic lymphocytic leukemia patients. Cancer Immunol Immunother 2013;62:113–24. REFERENCES 27 Lopez-­Millan B, Diaz de la Guardia R, Roca-­Ho H, et al. Imids 1 Katz AJ, Chia VM, Schoonen WM, et al. Acute lymphoblastic mobilize acute myeloid leukemia blasts to peripheral blood leukemia: an assessment of international incidence, survival, and through downregulation of CXCR4 but fail to potentiate AraC/ disease burden. Cancer Causes Control 2015;26:1627–42. Idarubicin activity in preclinical models of non del5q/5q- AML. 2 Winters AC, Bernt KM. Mll-­Rearranged Leukemias-­An update on Oncoimmunology 2018;7:e1477460. science and clinical approaches. Front Pediatr 2017;5:4. 28 Castella M, Boronat A, Martín-­Ibáñez R, et al. Development of 3 Pui CH, Crist WM. Biology and treatment of acute lymphoblastic a novel anti-­CD19 chimeric antigen receptor: a paradigm for an leukemia. J Pediatr 1994;124:491–503. affordable CAR T cell production at academic institutions. Mol Ther 4 Gibson LF. Survival of B lineage leukemic cells: signals from the bone Methods Clin Dev 2019;12:134–44. marrow microenvironment. Leuk Lymphoma 2002;43:19–27. 29 Baroni ML, Sanchez Martinez D, Gutierrez Aguera F, et al. 5 Bradstock KF, Gottlieb DJ. Interaction of acute leukemia cells with 41BB-­based and CD28-­based CD123-­redirected T-­cells ablate the bone marrow microenvironment: implications for control of human normal hematopoiesis in vivo. J Immunother Cancer minimal residual disease. Leuk Lymphoma 1995;18:1–16. 2020;8:e000845. 6 LeClerc JM, Billett AL, Gelber RD, et al. Treatment of childhood 30 Gonzalez-­Rey E, Anderson P, González MA, et al. Human adult stem acute lymphoblastic leukemia: results of Dana-­Farber all Consortium cells derived from adipose tissue protect against experimental colitis protocol 87-01. J Clin Oncol 2002;20:237–46. and sepsis. Gut 2009;58:929–39. 7 Hoogerbrugge PM, Gerritsen EJ, vd Does-­van den Berg A, et al. 31 Mizoguchi A, Mizoguchi E. Animal models of IBD: linkage to human Case-­Control analysis of allogeneic bone marrow transplantation disease. Curr Opin Pharmacol 2010;10:578–87. versus maintenance chemotherapy for relapsed all in children. Bone 32 Conforti A, Biagini S, Del Bufalo F, et al. Biological, functional and Marrow Transplant 1995;15:255–9. genetic characterization of bone marrow-­derived mesenchymal 8 Burt R, Dey A, Aref S, et al. Activated stromal cells transfer stromal cells from pediatric patients affected by acute lymphoblastic mitochondria to rescue acute lymphoblastic leukemia cells from leukemia. PLoS One 2013;8:e76989. oxidative stress. Blood 2019;134:1415–29. 33 Zhi-­Gang Z, Wei-­Ming L, Zhi-­Chao C, et al. Immunosuppressive 9 Moses BS, Slone WL, Thomas P, et al. Bone marrow properties of mesenchymal stem cells derived from bone marrow microenvironment modulation of acute lymphoblastic leukemia of patient with hematological malignant diseases. Leuk Lymphoma phenotype. Exp Hematol 2016;44:50–9. 2008;49:2187–95. 10 Mudry RE, Fortney JE, York T, et al. Stromal cells regulate 34 Ayala F, Dewar R, Kieran M, et al. Contribution of bone survival of B-­lineage leukemic cells during chemotherapy. Blood microenvironment to leukemogenesis and leukemia progression. 2000;96:1926–32. Leukemia 2009;23:2233–41. 11 Randhawa S, Cho BS, Ghosh D, et al. Effects of pharmacological 35 Méndez-­Ferrer S, Bonnet D, Steensma DP, et al. Bone marrow niches and genetic disruption of CXCR4 chemokine receptor function in B-­ in haematological malignancies. Nat Rev Cancer 2020;20:285–98. cell acute lymphoblastic leukaemia. Br J Haematol 2016;174:425–36. 36 Reagan MR, Rosen CJ. Navigating the bone marrow niche: 12 Tesfai Y, Ford J, Carter KW, et al. Interactions between acute translational insights and cancer-­driven dysfunction. Nat Rev lymphoblastic leukemia and bone marrow stromal cells influence Rheumatol 2016;12:154–68. response to therapy. Leuk Res 2012;36:299–306. 37 Schepers K, Campbell TB, Passegué E. Normal and leukemic stem 13 Yang Y, Mallampati S, Sun B, et al. Wnt pathway contributes to the cell niches: insights and therapeutic opportunities. Cell Stem Cell protection by bone marrow stromal cells of acute lymphoblastic 2015;16:254–67. 12 Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 Open access J Immunother Cancer: first published as 10.1136/jitc-2020-001419 on 31 August 2020. Downloaded from http://jitc.bmj.com/ on February 12, 2021 by guest. Protected by copyright. 38 Ma Z, Zhao X, Deng M, et al. Bone marrow mesenchymal stromal leukemia patients are anti-­inflammatory in in vivo preclinical models cell-­derived periostin promotes B-­ALL progression by modulating of hematopoietic reconstitution and severe colitis. Haematologica CCL2 in leukemia cells. Cell Rep 2019;26:1533–43. 2019;104:e54–8. 39 Verma D, Zanetti C, Godavarthy PS, et al. Bone marrow niche-­derived 41 Imai C, Mihara K, Andreansky M, et al. Chimeric receptors with extracellular matrix-­degrading enzymes influence the progression of 4-­1BB signaling capacity provoke potent cytotoxicity against acute B-­cell acute lymphoblastic leukemia. Leukemia 2020;34:1540–52. lymphoblastic leukemia. Leukemia 2004;18:676–84. 40 de la Guardia RD, Lopez-­Millan B, Roca-­Ho H, et al. Bone marrow mesenchymal stem/stromal cells from risk-­stratified acute myeloid Zanetti SR, et al. J Immunother Cancer 2020;8:e001419. doi:10.1136/jitc-2020-001419 13