Leukemic Spleen Research Articles

Murine Regulatory CD4+ T Cells Are Increased in Leukemic Spleens and Show High Co-Expression of Co-Regulatory Molecules CD39, CD73, PD1, and TIGIT

Murine Regulatory CD4+ T Cells Are Increased in Leukemic Spleens and Show High Co-Expression of Co-Regulatory Molecules CD39, CD73, PD1, and TIGIT

In Vivo Ablation of NFκB Cascade Effectors Alleviates Disease Burden in Myeloproliferative Neoplasms

ABAL and TK equal contributors Multilevel regulation of NFκB activation exists in response to microenvironmental cues. Aberrant hyperactivation of the NFκB cascade propagates oncogenic signaling and inflammation, which together augment disease burden in myeloproliferative neoplasms (MPNs). Here, we systematically ablated NFκB signaling effectors to identify core dependencies using a series of primary samples and syngeneic and patient-derived xenograft (PDX) mouse models. We first examined the impact of Rela abrogation in the Jak2 V617F knock-in mouse model that exemplifies polycythemia vera hallmarks and generated Jak2V617FRela flox/flox Ubc-Cre-ERT2 ( Jak2- Rela fl) mice. Evident suppression of leukocytosis was observed within a week of Rela excision and sustained until endpoint. Elevated hematocrit, monocytes, and CD45.2 myeloid cells were suppressed in both the peripheral blood (PB) and bone marrow (BM) and Jak2- Rela flmice had significant reduction of splenomegaly. We then adopted a second MPN model, where MPL W515L retroviral expression produces a phenotype resembling myelofibrosis. Consistent reduction of pathologic WBC counts, monocytes, and PB and BM CD45.2 myeloid cells were observed in MPL- Rela fl mice. Rela ablation also reduced hematocrit, platelet counts, and CD45.2 GFP+ cells. Importantly, we observed a significant attenuation of disease hallmarks with a decrease of spleen weight and dysmegakaryopoiesis, enhanced survival, and reduction of BM reticulin deposition. We next established PDX models where CD34+ cells from MF patients and healthy BM donors (NBM) were transduced with an sgRNA targeting RELA and engrafted into NSGS mice. Across two MPN PDXs, there was significant suppression of leukemic cells and improvement of disease burden, reduced spleen weights and prolonged survival, and subsequent BM histology revealing improved sclerosis, fibrosis, and dysmegakaryopoiesis. In contrast, NSGS mice engrafted with NBM CD34+ cells did not yield disease symptoms, nor did RELA targeting in healthy cells affect hematopoiesis or measured disease hallmarks. RNA-seq on CD34+ cells from four MPN patients following RELA abrogation revealed augmented and altered expression of genes that are 1) regulators of survival and proliferative pathways: MDM2, TP53INP1, and FAS 2) major ferroptosis mediator GPX4, and 3) cell cycle effectors: CDKN1A (p21), E2F4, and CHEK2. In further investigation of NFκB pathway effectors, we genetically targeted upstream effector Myd88 and transduced Kit+ cells from Myd88-null mice with MPL W515L. Consistent with Rela-knockout, Myd88-deficiency also diminished pathogenic WBC and monocyte count upsurgence, inhibited the accumulation of BM CD45.2 myeloid cells, and reduced spleen weights. The rapid lethality governed by mutant MPL was also delayed upon Myd88 loss, consistent with reduced pathology in spleen and BM sections. While Myd88 and Rela activate NFκB, miR-146a is repressive. We performed further knock-out of MIR146A using PDX models and transplanted MIR146A-KO CD34+ cells into NSGS mice and observed enhanced leukemic engraftment, increased spleen weights, earlier lethality, and evidence of BM sclerosis or fibrosis across both MF PDXs. Importantly, these disease parameters remained unchanged in a NBM PDX following MIR146A-KO. Lastly, in our WashU Cohort of RNA-sequencing performed on 90 CD34+ samples encompassing the spectrum of MPN and secondary acute myeloid leukemia (AML) with NBM controls, we observed significant upregulation of IRAK4 and IL1B in MF and sAML compared to NBM, and their significant expression correlation in MPN. We then pharmacologically inhibited IRAK4 using a small molecule, CA-4948, which is currently under evaluation in phase 1/2a trials of relapsed/refractory AML/myelodysplastic syndrome. In PDX experiments, CA-4948 treatment effectively reduced leukemic engraftment and spleen weights of two MF models without inducing toxicity in a NBM PDX. Through qRT-PCR and mass cytometry, we demonstrate that CA-4948 treatment dampened inflammatory cytokine production induced by IL-1β in primary MPN CD14+ monocytes. Overall, we demonstrate that targeting mediators including Rela, Myd88, and IRAK4 specifically alleviated MPN disease burden without toxicity to healthy tissue and further establish CA-4948 as a promising therapeutic avenue for the treatment of MPN.

Importance of FDG-PET (Fluorodeoxyglucose Positron Emission Tomography) in Staging and Response Assessment of Hairy Cell Leukemia (HCL)

Alessandro Mancini and Luca De Carolis are Co-first authors. Nicodemo Baffa, Brunangelo Falini and Enrico Tiacci are Co-last authors. INTRODUCTION HCL is an indolent BRAF-mutated leukemia usually presenting with cytopenias, splenomegaly, little or no lymphadenopathy, heavy bone marrow (BM) infiltration and few circulating leukemic cells. Imaging exams for staging include abdominal echography and computed tomography (CT), while the role of PET is largely unknown. METHODS 84 patients (pts) with relapsed/refractory HCL participating to our clinical trials of BRAF inhibitor-based therapies prospectively underwent PET/CT pre- and post-treatment, in addition to usual examinations for response assessment. Since FDG uptake is normally higher in liver than spleen, BM and lymph nodes, uptake in the latter tissues was considered abnormal if equal to or higher than the liver. Splenomegaly was defined by a longest spleen diameter >13 cm on PET/CT. The standard definition of complete remission (CR) in HCL just requires no palpable splenomegaly on physical examination, in addition to resolution of cytopenias as well as no hairy cells visible in the blood smear and the BM biopsy by non-immunological stains. RESULTS Pre-therapy, 58/77 (75%) non-splenectomized pts showed abnormal splenic FDG uptake (maximal standardized uptake value/SUVmax: median 4.8, vs 3.9 for the liver), which was always diffuse without focal lesions. Conversely, focal splenic uptake is not infrequent in splenic marginal zone lymphoma (18%, 7/39 pts - Abdom Radiol 2018;43:2721; vs of 0/58 HCL pts, p=0.0003), which may aid in the differential diagnosis with this HCL-mimicker. Moreover, 14/58 (24%) HCL pts. with a PET+ spleen did not have splenomegaly, and splenic uptake returned to normal in 10/11 evaluable cases (91%) achieving at least a partial response post-therapy, which suggests leukemic spleen involvement even without splenomegaly. Abnormal diffuse BM uptake was observed in 41/82 evaluable pts (50%; median SUVmax 5, vs 3.4 for the liver). In these cases, leukemic involvement in the BM biopsy was greater (median 80%) than in the other 41 pts (median 65%; p-value <0.01), pointing to HCL infiltration as the cause of abnormal BM uptake rather than to reactive hyperplasia of non-involved BM. Post-therapy, pathologic spleen uptake normalized in 46/51 evaluable pts (90%), and their CR rate was higher (32/46 pts, 70%) than in the other 5 pts remaining PET+ (1/5, 20%; p=0.047). Among the 46 PET- cases, 31 (67%) were not splenomegalic while 15 (33%) had residual splenomegaly (up to 17 cm; Fig. 1A); interestingly, within PET- cases relapse-free survival (RFS) was similar not only in conventionally defined CR cases with (n=8) vs without (n=24) residual non-palpable splenomegaly (p=0.84 after a median follow-up of 19 months), but also when the 6 cases meeting all conventional CR criteria except for still palpable splenomegaly were added to the 8 CR pts with residual non-palpable splenomegaly (p=0.7 after a median follow-up of 20.5 months - Fig. 1B). These findings suggest that metabolic status reflects splenic HCL involvement more reliably than spleen size or palpability, a concept that could improve the current definition of CR in HCL. Abnormal BM uptake resolved after treatment in 34/37 evaluable pts (92%), and their CR rate was again higher (23/34, 68%) than in the other 3 pts remaining PET+ (0/3; p=0.047). Leukemic infiltration of the BM biopsy was lower in the 34 PET- vs the 3 PET+ pts (median 5% vs 70%-80%, respectively; p=0.0004), indicating that BM abnormal uptake reflects HCL infiltration load rather than reactive BM hyperplasia post-therapy. Finally, abnormal lymph node uptake was observed pre-therapy in 9/84 (11%) pts (median SUVmax 7.1, vs 3.4 for the liver) and normalized after treatment in 7/8 evaluable pts (88%), including one case with residual lymph node enlargement (2.2 cm); all these 7 pts had achieved a CR while the remaining pt had no response to therapy. CONCLUSIONS PET is potentially useful in the differential diagnosis with splenic marginal zone lymphoma and may aid in the clinical management of HCL pts by detecting metabolic involvement of the main disease sites even in the absence of organomegaly. Importantly, PET signal tracks with response to therapy even in case of persistent organomegaly, which may lead to refine the definition of CR in HCL.

Interleukin-27 potentiates CD8+ T-cell-mediated anti-tumor immunity in chronic lymphocytic leukemia.

Chronic Lymphocytic Leukemia (CLL) cells are highly dependent on interactions with the immunosuppressive tumor microenvironment (TME) for survival and proliferation. In the search for novel treatments, pro-inflammatory cytokines have emerged as candidates to reactivate the immune system. Among those, IL-27 has recently gained attention, but its effects differ among malignancies. Here, we utilized the Eμ-TCL1 and EBI3 knock-out mouse models as well as clinical samples from patients to investigate the role of IL-27 in CLL. Characterization of murine leukemic spleens revealed that the absence of IL-27 leads to enhanced CLL development and a more immunosuppressive TME in transgenic mice. Gene-profiling of T-cell subsets from EBI3 knock-out highlighted transcriptional changes in the CD8+ T-cell population associated with T-cell activation, proliferation, and cytotoxicity. We also observed an increased anti-tumor activity of CD8+ T cells in presence of IL-27 ex vivo with murine and clinical samples. Notably, IL-27 treatment led to the reactivation of autologous T cells from CLL patients. Finally, we detected a decrease in IL-27 serum levels during CLL development in both pre-clinical and patient samples. Altogether, we demonstrated that IL-27 has a strong antitumorigenic role in CLL and postulate this cytokine as a promising treatment or adjuvant for this malignancy.

A Predictive Endothelial-Leukemia Pre-Clinical Platform to Uncover Drug Vulnerabilities for Personalized Treatments

Background. T-cell acute lymphoblastic leukemia (T-ALL) is an aggressive disease with few innovative treatment options. This is also contributed by the lack of models capable of capturing the complexity of the tumor and its microenvironment.Aims. To identify patient-specific vulnerabilities and novel therapeutic strategies in T-ALL and interrogate the mechanisms of the crosstalk between leukemic and stromal elements.Methods. We established a drug-testing platform using patient-derived-tumor-xenografts (PDTX) and a mixed-culture approach using E4ORF1-transduced endothelial cells (ECs) (Seandel M et al, PNAS 2008) to overcome host-mediated chemoresistance. We performed functional experiments using total and single-cell RNA sequencing.Results. First, we established a battery of 22 T-ALL PDTX models that matched both phenotypically (immune-histochemistry, flow cytometry) and genotypically (TCR rearrangement, transcriptome) with the primary patients' samples. We then challenged these models (n=14 samples belonging to different PDTX and serial passages within each model) with a library of compounds (n=433) targeting redundant proteins (n=634). Unsupervised clustering and Principal Component Analysis (PCA) demonstrated two clusters of T-ALL samples based on differential drug susceptibility. We could at least partially correlate these differences to specific transcriptomic signatures predictive of drug response (Figure 1A).We then defined a group of pan-active compounds across all models (n=40), which we validated using an independent screening with/without ECs (Figure 1B). We found that ECs counteracted the activity of selected compounds (i.e. TSA, THZ1 and MLN2238). By PCA, we observed distinct response profiles based on different T-ALL models. We vectorized the EC-rescue and found that the direction was the same across all 3 models tested, indicating that it relied on similar mechanisms regardless of model identity. Based on the known role of IGF1-IGFR1 as a supportive EC-rescue axis (Medyouf H et al, J Exp Med 2011), we performed the same screening with/without recombinant IGFBP-7 (500 ng/mL), a decoy IGF1 molecule. Remarkably, IGFBP-7 completely or partially abrogated the EC-mediated rescue of selected compounds [enzastaurin (PKC-β inhibitor), SC144 (GP130 inhibitor), CHIR124 (Chk1 inhibitor) and YM155 (Survivin inhibitor)] (Figure 1B). Drugs not rescued by ECs (n=30) were considered positive hits and 5 of them (ruxolitinib, tofacitinib, panobinostat, bortezomib, irinotecan) ultimately proved to be effective in vivo in randomized pre-clinical trials either alone or in combination (Figure 1C). Our stepwise endothelial-leukemia platform led to the discovery of “public” and “private” vulnerabilities and the proof-of-principle of prediction-guided in vivo pre-clinical trials. We propose a list of compounds that could be readily translated into T-ALL clinical trials (Figure 1D). We finally proved the validity of our platform using other disease models (i.e. B and T-lymphoma PDTXs).Mechanistically, at single-cell resolution, in vitro interacting T-ALL cells and ECs underwent reciprocal transcriptome changes, with T-ALL shifting towards stemness/undifferentiation and ECs towards tumor-ECs (TECs) phenotypes. Furthermore, in vitro EC-educated T-ALL cells mimicked distinct T-ALL subsets of the leukemic spleen of corresponding PDTX mice (Figure 1E).Conclusions. These data demonstrate that our EC-T-ALL culture system simulates in vivo conditions, offering a robust platform to study drug response, leukemia-host interactions and cell plasticity. This approach will improve the pre-clinical predictability of novel drugs/combinations for T-ALL, as well as for other hematologic malignancies, and propel the development of patient-tailored treatments. [Display omitted] DisclosuresMelnick: Janssen Pharmaceuticals: Research Funding; Sanofi: Research Funding; Daiichi Sankyo: Research Funding; Epizyme: Consultancy; Constellation: Consultancy; KDAC Pharma: Membership on an entity's Board of Directors or advisory committees. Elemento: AstraZeneca: Research Funding; Freenome: Consultancy, Other: Current equity holder in a privately-held company; Volastra Therapeutics: Consultancy, Other: Current equity holder, Research Funding; Champions Oncology: Consultancy; Owkin: Consultancy, Other: Current equity holder; One Three Biotech: Consultancy, Other: Current equity holder; Eli Lilly: Research Funding; Johnson and Johnson: Research Funding; Janssen: Research Funding. Chiaretti: amgen: Consultancy; pfizer: Consultancy; novartis: Consultancy; Incyte: Consultancy. Cerchietti: Celgene: Research Funding; Bristol Myers Squibb: Research Funding.

Association of aberrant ASNS imprinting with asparaginase sensitivity and chromosomal abnormality in childhood BCP-ALL

AbstractKaryotype is an important prognostic factor in childhood B-cell precursor acute lymphoblastic leukemia (BCP-ALL), but the underlying pharmacogenomics remain unknown. Asparaginase is an integral component in current chemotherapy for childhood BCP-ALL. Asparaginase therapy depletes serum asparagine. Normal hematopoietic cells can produce asparagine by asparagine synthetase (ASNS) activity, but ALL cells are unable to synthesize adequate amounts of asparagine. The ASNS gene has a typical CpG island in its promoter. Thus, methylation of the ASNS CpG island could be one of the epigenetic mechanisms for ASNS gene silencing in BCP-ALL. To gain deep insights into the pharmacogenomics of asparaginase therapy, we investigated the association of ASNS methylation status with asparaginase sensitivity. The ASNS CpG island is largely unmethylated in normal hematopoietic cells, but it is allele-specifically methylated in BCP-ALL cells. The ASNS gene is located at 7q21, an evolutionally conserved imprinted gene cluster. ASNS methylation in childhood BCP-ALL is associated with an aberrant methylation of the imprinted gene cluster at 7q21. Aberrant methylation of mouse Asns and a syntenic imprinted gene cluster is also confirmed in leukemic spleen samples from ETV6-RUNX1 knockin mice. In 3 childhood BCP-ALL cohorts, ASNS is highly methylated in BCP-ALL patients with favorable karyotypes but is mostly unmethylated in BCP-ALL patients with poor prognostic karyotypes. Higher ASNS methylation is associated with higher L-asparaginase sensitivity in BCP-ALL through lower ASNS gene and protein expression levels. These observations demonstrate that silencing of the ASNS gene as a result of aberrant imprinting is a pharmacogenetic mechanism for the leukemia-specific activity of asparaginase therapy in BCP-ALL.

3021 – A NOVEL HISTONE CHAPERONE PATHWAY DRIVES IL-1β-MEDIATED AML PROGRESSION

Genetic heterogeneity remains an ongoing challenge in treating acute myeloid leukemia (AML). Identifying microenvironment-driven pathways that are active across AML genetic subtypes should allow the development of broadly effective therapies. We have shown that AML microenvironment is rich in proinflammatory cytokine interleukin-1β (IL-1β) and promotes AML progression. The transcriptome analysis revealed that ASF1B (anti-silencing function1B), a histone chaperone, is one of the most differentially expressed genes in AML progenitors as compared to normal cells upon IL-1β stimulation. The functional roles of ASF1B in hematopoiesis and leukemia remain elusive. We identified that IL-1β upregulates ASF1B in various genetic subtypes of AML at both mRNA and protein levels. Upon shRNA-mediated ASF1B depletion, AML cells exhibited reduced cell viability, inhibited growth, cell cycle arrest, and impaired colony formation ability. In vivo xenograft model showed 80% reduction in leukemic burden and spleen size following ASF1B silencing. Using a MLL-ENL murine bone marrow transplantation model, we found that ASF1B overexpression alone shortened survival (p = 0.02) of MLL-ENL leukemic mice to the same extent as IL-1β-treated recipients versus vehicle-treated group (p = 0.03), suggesting that high ASF1B levels mimic IL-1β-mediated growth effects in AML. In a complementary study, both heterozygous and complete deletion of ASF1B extended survival of mice (p < 0.01). Notably, ASF1B deficiency attenuated IL-1β-mediated leukemia progression (p = 0.05) as compared to wild type leukemic cells. Mechanistically, activation of the ASF1B pathway confers higher DNA damage and replication stress tolerance in AML cells. Additionally, normal hematopoiesis is not disrupted in ASF1B-deficient mice, indicating targeting of the ASF1B pathway may spare normal cells. In summary, we demonstrate that elevated ASF1B expression promotes disease progression in AML. We also provide in vitro and in vivo evidence that ASF1B plays a critical role in potentiating IL-1β-dependent AML growth. Therefore, ASF1B can serve as a novel route of therapeutic strategy to suppress microenvironmental-driven growth in various AML genetic subtypes.

Antiproliferative and Immunoregulatory Effects of Azelaic Acid Against Acute Myeloid Leukemia via the Activation of Notch Signaling Pathway.

Acute myeloid leukemia (AML) is a common type of hematological malignancy that can progress rapidly. AML has a poor prognosis and a high incidence of relapse due to therapeutic resistance. Azelaic acid (AZA), a small molecular compound is known to exhibit antitumor effect on various tumor cells. This study aimed to evaluate the antiproliferative and immunoregulatory effects of AZA against AMLviathe activation of the notch signaling pathway. We found that AZA can inhibit the proliferation of AML cells. In addition, laser confocal microscopy showed AZA-treated AML cells began to swelling and undergo cytoplasmic vacuolization. Importantly, AZA promoted the proliferation of NK and T cells and increased the secretion of TNF-αand IFN-γ. AZA also increased the expression levels of CD107a and TRAIL in NK cells, and CD25 and CD69 in T cells to influence their activation and cytotoxic ability. AZA-treated NK cells can kill AML cells more efficiently at the single-cell level as observed under the microfluidic chips. Further mechanistic analysis using protein mass spectrometry analysis and Notch signaling reporter assay demonstrated that Notch1and Notch2 were up-regulated and the Notch signaling pathway was activated. Moreover, combining AZA with the Notch inhibitor, RO4929097, decreased the expression of Notch1and Notch2, and downstream HES1 and HEY1, which rendered AML cells insensitive to AZA-induced apoptosis and alleviated AZA-mediated cytotoxicity in AML. In vivo, AZA relieved the leukemic spleen infiltration and extended the survival. The percentage of CD3-CD56+NK cells and CD4+CD8+T cells as well as the secretion of cytotoxic cytokines was increased after the treatment of AZA. The overall findings reveal that AZA is a potential Notch agonist against AML in activating the Notch signaling pathway.

Hepatic leukemia-associated macrophages exhibit a pro-inflammatory phenotype in Notch1-induced acute T cell leukemia

Tumor-associated macrophages (TAMs) are well accepted and the pathological role of macrophages in hematopoietic malignancies have been proposed. Hepatomegaly is frequently observed in T cell acute lymphoblastic leukemia (T-ALL) patients with poor prognosis. However, the role of leukemia-associated macrophages (LAMs) in hepatic microenvironment remains unclear. Here, the characteristics of hepatic LAMs (H-LAMs) were studied in Notch1 induced T-ALL model. Increase in proportion and absolute counts of H-LAMs was detected with infiltration of inflammatory cells. Furthermore, H-LAMs exhibited a more M1-like phenotype distinct from that of TAMs in hepatocellular carcinoma and LAMs from BM or spleen in leukemia. Moreover, H-LAMs expressed increased level of cytokines in charge of recruiting inflammatory cells, which contributed to pro-inflammatory hepatic microenvironment.

Morinda Citrifolia Leaf Extract Ameliorated Leukemia in Mice Model

Leukemia affects over a million people worldwide and disturbs the bone marrow, the spleen and the lymph system. In finding dietary therapy against leukemia the vegetable Morinda citrifolia leaves were investigated and compared with ATRA (alltrans-retinoic acid) in mice. The extract standardised to epicatechin and scopoletin was anti-proliferative against human JURKAT and murine WEHI-3B leukemia cells (IC50 of 14.5 and 17μg /ml respectively) in vitro yet was non-toxic to normal peripheral blood mononuclear cells. Leukemia was induced in the BALB/c mice by inoculating WEHI-3B cells subcutaneously. Three days after leukemia development the extract (100 and 200mg /Kg BW) or ATRA (5mg/kg BW) were fed daily to the mice for four weeks. The extract decreased the leukemic mice peripheral blood immature leukocytes, anemia, total white blood cells; bone marrow blasts cells content and blast cells invasions to the spleen and liver. Electron microscopy and histopathology observations confirmed the extract induced leukemia cells apoptosis similar to ATRA. The extract also mitigated the changes in T-cells and B-cells levels in the leukemic spleen tissues in vivo. The extract dose dependently induced leukemia cells apoptosis both in vitro and in vivo restored all pathological blood, liver and spleen changes in vivo and reinstated the endogenous immune responses to normal healthy levels in vivo.

A screen for Fli-1 transcriptional modulators identifies PKC agonists that induce erythroid to megakaryocytic differentiation and suppress leukemogenesis.

The ETS-related transcription factor Fli-1 affects many developmental programs including erythroid and megakaryocytic differentiation, and is frequently de-regulated in cancer. Fli-1 was initially isolated following retrovirus insertional mutagenesis screens for leukemic initiator genes, and accordingly, inhibition of this transcription factor can suppress leukemia through induction of erythroid differentiation. To search for modulators of Fli-1, we hereby performed repurposing drug screens with compounds isolated from Chinese medicinal plants. We identified agents that can transcriptionally activate or inhibit a Fli-1 reporter. Remarkably, agents that increased Fli-1 transcriptional activity conferred a strong anti-cancer activity upon Fli-1-expressing leukemic cells in culture. As opposed to drugs that suppress Fli1 activity and lead to erythroid differentiation, growth suppression by these new Fli-1 transactivating compounds involved erythroid to megakaryocytic conversion (EMC). The identified compounds are structurally related to diterpene family of small molecules, which are known agonists of protein kinase C (PKC). In accordance, these PKC agonists (PKCAs) induced PKC phosphorylation leading to activation of the mitogen-activated protein kinase (MAPK) pathway, increased cell attachment and EMC, whereas pharmacological inhibition of PKC or MAPK diminished the effect of our PKCAs. Moreover, in a mouse model of leukemia initiated by Fli-1 activation, the PKCA compounds exhibited strong anti-cancer activity, which was accompanied by increased presence of CD41/CD61 positive megakaryocytic cells in leukemic spleens. Thus, PKC agonists offer a novel approach to combat Fli-1-induced leukemia, and possibly other cancers,by inducing EMC in part through over-activation of the PKC-MAPK-Fli-1 pathway.

Leukemia-Associated Macrophages Educated By Organ-Specific Leukemic Microenvironment Acquire Diverse Functions and Gene-Expression Signatures

Macrophages are important components of microenvironments and participate diverse physiological and pathological processes. Though CD68+macrophages have been demonstrated a potentially clinical marker to predict adverse prognosis in lymphoma, the role of macrophages in leukemia have not been fully elucidated. Most recently, we reported leukemia-associated macrophages (LAMs), which showed leukemia-promoting effects, in Notch1-induced mouse T-ALL leukemic microenvironment. In the present study, we studied the characteristics of LAMs from BM and spleen in MLL-AF9 induced mouse AML model, and compared the gene expression diversities among LAMs from BM and SP of AML and T-ALL models.LAMs were sorted from leukemic BM and spleen in the middle stage of AML, control macrophages were sorted from normal mouse. Then LAMs were cocultured with AML cells for 24, 48 and 72hrs. Cell count, BrdU incorporation assay and Annexin V assay were used to detect the proliferation and apoptosis of leukemia cells, respectively. Significant increase of BrdU positive and decrease of apoptotic AML cells were detected in leukemia cells cocultured with LAMs than control Mϕ. Latex-beads uptaken assay and transwell assay were used to detect the phagocytosis and migration of LAMs. Significant decrease of phagocytosis and increase of migration ability were detected in LAMs than control Mϕ. Furthermore, T cell coculture assay revealed that spleen LAMs suppress the CD25 and CD69 expression in both CD4+ and CD8+T cells but increase induction of Treg cells than BM LAMs. Moreover SP LAMs had a significantly higher proliferation promoting effect on leukemia cells than BM LAMs. To explore the diverse effect among LAMs from BM and spleen, we compared the transcriptional data of LAMs from BM and spleen from AML and T-ALL. The ordered list algorithm revealed the similarity score in LAM samples with same microenvironment (991.8) is higher than those with same type of leukemia (709.1), also the expression pattern analysis revealed that LAMs from leukemic BM and spleen are much more different, the phenotypic gene validated by qRT-PCR revealed that the BM LAMs are more M1-like while the spleen LAMs are more M2-like.Our results suggested that the diverse response of LAMs to leukemia are largely determined by organ-specific leukemic microenvironment.The work was supported by the Grants 81370634, 81570153, 81300376 from the National Natural Science Foundation of China (NSFC). DisclosuresNo relevant conflicts of interest to declare.

Impact of Genomic Alterations on Outcomes in Myelofibrosis Patients Undergoing JAK1/2 Inhibitor Therapy

Introduction:The advent of next generation sequencing (NGS) has brought intense interest to the complex genetic landscape of myeloproliferative neoplasms (MPN). However, data regarding clinical outcomes in the context of novel MPN therapies such as JAK inhibitors are scarce. Limited data indicate that high molecular risk signature (HMR, presence of at least one mutation in ASXL1, EZH2, IDH1/2, SRSF2) or multiple mutations may be associated with decreased spleen response and a shorter time to discontinuation of Ruxolitinib in myelofibrosis (Patel et al, Blood 2015).Methods:All myelofibrosis patients seen in the MPN program at Princess Margaret Hospital between November 2009 and May 2016 and treated with JAK1/2 inhibitor therapy were identified. NGS molecular profiling of 54 genes (39 hotspot region; 15 complete coding region coverage) was performed on peripheral blood or bone marrow samples using the TruSight Myeloid Sequencing Panel. Reporting was restricted to well-covered, exonic nonsynonymous, intronic splice site, and known pathogenic synonymous variants. Variants with global mean allele frequency >1% were identified using multiple population databases (1000 genomes, ESP, ExAC) and excluded. The primary endpoint was time from start of JAK1/2 inhibitor therapy to treatment failure (TTF) defined as treatment discontinuation, progression to accelerated phase or leukemic transformation, spleen progression or death. Secondary endpoints included best spleen, anemia and IWG response achieved by 48 weeks of treatment and overall survival. Response was assessed according to the 2013 revised IWG-MRT criteria. Transfusion dependency was assessed as any transfusion in 12 weeks prior to treatment or being identified as transfusion dependent in medical history.Results:Of 159 patients treated with JAK1/2 inhibitors at our institution, 102 met the inclusion criteria (see Table 1). Patients were excluded if; no sample was available for analysis (19), short use of JAK inhibitor prior to transplant (9), active clinical trial (5), in accelerated phase/acute leukemia (4) and others (20). First JAK inhibitor used was ruxolitinib in 77 patients and momelotinib in 25. At least one mutation was identified in every patient. Twenty (20%) patients had one mutation, 32 (31%) had 2 mutations and 50 (49%) patients had ≥ 3 mutations. Eighty (82%) patients had the JAK2V617F mutation, 15 (15%) had mutations in CALR, 4 (4%) had MPL mutations and one patient was triple negative. One patient had mutations in both CALR and JAK2 while another had mutations in MPL and CALR. Forty-eight (47%) patients had mutations consistent with HMR profile. Mutation profile is summarized in Table 2.With median follow-up of 2.5 years, 51 (50%) patients experienced treatment failure. On univariate analysis, TTF was associated with DIPSS, pre-treatment transfusion status and Hb <100 prior to initiating JAK inhibitor therapy. However, TTF was not associated with specific driver mutations, the number of mutations or HMR profile. Exploratory analysis of genes mutated in ≥5% of the population showed EZH2 (p=0.004) and CBL (p=0.005) mutated patients had shorter TTF. Multi-variable analysis employing anemia <100 and DIPSS with either of the number of mutations or HMR profile did not show any association with TTF. There was a trend towards TTF in patients with Hb<100 (HR 2.21, 95%CI: 0.99-4.95, p=0.05). Of 102 patients, 91 were evaluable for spleen response with 43 (47%) achieving a spleen response by 48 weeks of treatment. Twelve (23%) of the 53 patients evaluated met criteria for anemia response. Neither spleen nor anemia response was associated with the use of a particular JAK inhibitor. Of the 95 patients with at least 48-week follow-up, best response while on treatment was clinical improvement and spleen response in 29 (31%) and 19 (20%) patients respectively.Conclusions:In this study of myelofibrosis patients treated with JAK inhibitors, EZH2 and CBL mutated patients had shorter TTF. We did not find any association between TTF and number of mutations or other high risk mutations such as ASXL1/SRSF2. Anemia was the only significant independent predictor of shorter TTF. Our findings highlight the need for multicenter collaborative studies on a large number of patients and cautious use of mutation profiling results in routine clinical decision making with current treatment approaches.Spiegel and McNamara are co-primary authors. [Display omitted] [Display omitted] DisclosuresPanzarella:Cellgene: Consultancy. Schimmer:Novartis: Honoraria. Schuh:Amgen: Membership on an entity's Board of Directors or advisory committees. Yee:Novartis Canada: Membership on an entity's Board of Directors or advisory committees, Research Funding. Kamel-Reid:BMS: Research Funding. Gupta:Novartis: Consultancy, Honoraria, Research Funding; Incyte Corporation: Consultancy, Research Funding.

The role of CXCR3 in the microenvironment of chronic lymphocytic leukemia

Abstract Chronic lymphocytic leukemia (CLL) is a B-cell malignancy that is stringently associated with a tumor-supportive microenvironment and defective anti-tumor immunity. T cells from CLL patients show features of exhaustion, including expression of PD-1, and are highly impaired in their activity. Our previous work showed that immune checkpoint blockade using anti-PD-L1 effectively prevents disease and restores T-cell activity in the Eμ-TCL1 mouse model of CLL. Development of disease in these mice was shown to be associated with a relative loss of naïve T cells and an enrichment of effector T cells displaying an exhausted phenotype. Our work showed that T cells in the Eμ-TCL1 mice express high levels of the chemokine receptor CXCR3 accompanied with elevated serum levels of the corresponding chemokines CXCL9 and CXCL10, suggesting a role for CXCR3 in T-cell skewing in CLL. Intracellular flow cytometric analysis showed an enrichment of CXCL9+ CD45+ hematopoietic cells in leukemic spleens which were identified as Ly6C+ monocytes. Along this line, depletion of phagocytic cells by Clodronate Liposomes resulted in significantly decreased serum levels of CXCL9, a reversal of T-cell skewing and slowed down development of CLL in mice. To evaluate the role of this receptor in CLL in vivo, adoptive transfer of murine CLL cells in CXCR3 knockout mice is currently ongoing. We further perform an in-depth phenotypical analysis of T-cell subsets in blood samples of CLL patients from different prognostic subgroups to characterize T-cell aberrations along with disease progression. Taken together, our data suggest a pathomechanistic role for CXCL9, CXCL10 and their receptor CXCR3 in CLL. Their potential as novel drug targets needs to be explored in the future.

Microenvironmental interleukin-6 suppresses toll-like receptor signaling in human leukemia cells through miR-17/19A

AbstractThe regulation of toll-like receptor (TLR) signaling in a tumor microenvironment is poorly understood despite its importance in cancer biology. To address this problem, TLR7-responses of chronic lymphocytic leukemia (CLL) cells were studied in the presence and absence of a human stromal cell-line derived from a leukemic spleen. CLL cells alone produced high levels of tumor necrosis factor (TNF)-α and proliferated in response to TLR7-agonists. A signal transducer and activator of transcription 3 -activating stromal factor, identified as interleukin (IL)-6, was found to upregulate microRNA (miR)-17 and miR-19a, target TLR7 and TNFA messenger RNA, and induce a state of tolerance to TLR7-agonists in CLL cells. Overexpression of the miR-17-92 cluster tolerized CLL cells directly and miR-17 and miR-19a antagomiRs restored TLR7-signaling. Inhibition of IL-6 signaling with antibodies or small-molecule Janus kinase inhibitors reversed tolerization and increased TLR7-stimulated CLL cell numbers in vitro and in NOD-SCIDγcnull mice. These results suggest IL-6 can act as tumor suppressor in CLL by inhibiting TLR-signaling.

The Zinc Finger Transcription Factor Mds1‐Evi1 Forms a Novel Protein Complex in MLL leukemia

The dual transcription factor‐methyltransferase Mds1‐Evi1 is frequently overexpressed in high‐risk leukemias. Mds1‐Evi1 contributes to leukemias induced by the mixed lineage leukemia (MLL) fusion protein MLL‐AF9, although the mechanism is unknown. MLL contains chromatin‐targeting and histone lysine methyltransferase domains, and fusion to AF9 can cause aberrant protein recruitment to target genes. Although Mds1‐Evi1 contains a methyltransferase domain, it does not appear to act on histones. Therefore, we are interested in how protein‐protein interactions involving Mds1‐Evi1 contribute to leukemogenesis. Many known interactors of MLL and Mds1‐Evi1 include chromatin modifying enzymes, suggesting that coordination between epigenetic modulators is key in leukemia pathogenesis.Here, we report the biochemical purification of a novel Mds1‐Evi1‐containing protein complex using a unique murine model of leukemia. To isolate the complex, we induced acute MLL‐AF9 leukemia in mice containing a tandem affinity purification (TAP) tag knocked into the endogenous Mds1‐Evi1 gene. Leukemic spleen tissue from these mice was used to purify the TAP‐tagged protein complex. Proteomic analysis revealed several members of the SWI/SNF chromatin remodeling complex, which is involved in epigenetic control of a variety of gene programs. Association between Mds1‐Evi1 and SWI/SNF may allow the activation of cell‐cycle and differentiation programs, contributing to transformation. SWI/SNF also coordinates with MLL fusion proteins to activate the Hox family of leukemia oncogenes. Our results provide a link between Mds1‐Evi1 and the chromatin remodeling machinery, where abnormal coordination of Mds1‐Evi1, MLL‐AF9 and SWI/SNF may lead to gene activation and the development of leukemia.

The Dual PI3K/mTOR Inhibitor PF-04691502 Induces Substantial Apoptosis in Chronic Lymphocytic Leukemia Cells in Vitro and Prolongs Survival in the Eµ-TCL1 Mouse Model

Pharmacological inhibition of specific B cell receptor signalling pathways within chronic lymphocytic leukemia (CLL) cells offers the potential for improved therapeutic options with reduced off target toxicity. Idelalisib, the PI3Kδ selective inhibitor, has been approved for CLL and significantly improved overall survival among patients with relapsed CLL in combination with rituximab compared to rituximab alone. In addition to PI3Kδ however, there are three other Class I PI3K isoforms, PI3Kα, PI3Kβ and PI3Kγ, with PI3Kα known to have a role in CLL survival and chemotaxis. In neutrophils, functional redundancy between PI3K isoforms is evident, with inhibition of at least three PI3K isoforms required for maximal apoptosis. Inhibition of mTOR is known to induce cell cycle arrest and apoptosis in CLL cells, however prolonged inhibition of mTOR results in activation of a positive feedback loop resulting in PI3K\\Akt reactivation. To overcome these caveats, pan PI3K inhibition alongside mTOR inhibition may achieve superior cytotoxicity against CLL cells compared to PI3Kδ or mTOR inhibition alone. We therefore sought to investigate the effect of a dual pan PI3K and mTOR inhibitor, PF-04691502, on primary CLL cells and in the Eµ-TCL1 mouse model of CLL.Twenty five primary CLL samples were treated with PF-04691502 and downstream signalling and subsequent apoptosis assessed by immunoblotting and Annexin V/PI staining respectively. In primary CLL cells, PF-04691502 induced 80% apoptosis after 24 hours at 10µM concentration. PF-04691502 had an IC50 value of 1µM as assessed by Annexin V/PI staining, with minimal toxicity to normal human B or T cells and a trend towards more specific killing in unmutated CLL (p=0.09) compared to mutated CLL. PF-04691502 inhibited both soluble and immobilised anti-IgM induced signalling and overcame anti-IgM induced survival signals. PF-04691502 abrogated CXCL12 induced signalling and impaired subsequent CLL migration towards the chemokine in a transwell chemotaxis assay. Finally, PF-04691502 was able to overcome protection from co-culture with stroma inducing significant apoptosis of CLL cells when added continuously and in wash out experiments.To assess the effect of PF-04691502 in vivo, twenty Eµ-TCL1 mice were inoculated with tumour cells and after day 21 dosed once daily with vehicle, 5mg/kg or 10mg/kg PF-04691502 for a further 14 days. Mice were assessed for leukemic cell counts, spleen size and white blood cell count throughout the duration of the experiment. The effect of PF-04691502 on B cell receptor and chemokine receptor induced signalling in ex vivo Eµ-TCL1 cells was assessed by immunoblotting.Near identical results to the CLL data above were observed. Using Eµ-TCL1 cells isolated from the spleen, PF-04691502 was shown to profoundly inhibit anti-IgM and CXCL12 induced signalling and chemotaxis as well as induce substantial apoptosis as measured by Annexin V/PI staining. Therefore, we subsequently assessed the effect of PF-04691502 in vivo using this model. Mice treated with PF-04691502 displayed a transient increase in leukemic cell numbers in the blood persisting for 1-4 days, followed by a reduction to levels significantly below that of the vehicle control. The whole white blood cell count remained stable in the PF-04691502 treated mice during the course of treatment, whilst the vehicle treated mice showed significant increases in tumour load up to 40×106 cells/ml 33 days post inoculation. Spleen sizes increased steadily over time in the control mice, whilst PF-04691502 treatment prevented this. After 33 days, PF-04691502 treated mice had spleen sizes comparable to non tumor-recipient control animals. In addition, CD5+ B220+ Eµ-TCL1 cells were significantly reduced in the bone marrow, spleen and lymph nodes following PF-04691502 treatment compared to control mice (p=0.0198; p< 0.0001; p=0.0151 respectively). These results demonstrate that PF-04691502 induces substantial apoptosis of primary CLL cells in vitro and significantly prolongs survival in an in vivo murine model of CLL. Such data indicate that dual pan PI3K/mTOR inhibitors may prove efficacious in the treatment of CLL patients. DisclosuresNo relevant conflicts of interest to declare.

Global Inhibition of Bruton's Tyrosine Kinase (BTK) Delays the Development and Expansion of Chronic Lymphocytic Leukemia (CLL) in the TCL1 Mouse Model of Disease

AbstractAbstract 183CLL is a common leukemia which demonstrates variable reactivity of the B cell receptor (BCR) to antigen ligation, but constitutive activation of this pathway. Bruton’s Tyrosine Kinase (BTK) is a member of this pathway which shows transcriptional upregulation and constitutive activity in CLL. Ongoing trials of ibrutinib, an orally bioavailable inhibitor of BTK, have shown outstanding preliminary activity in CLL. However results with other known reversible and more selective irreversible BTK inhibitors have had variable results. As ibrutinib potentially has other targets in addition to BTK, this raises the question of whether BTK is an important target in CLL. To answer this question and determine the role of BTK in the development and expansion of CLL, we have used Eμ-TCL1 (TCL1) transgenic mouse model of CLL, where the TCL1 oncogene is under the control of the VH promoter-IgH-Eμ enhancer and adult transgenic mice spontaneously develop CLL.To determine whether BTK is an important target in CLL as opposed to an alternative target of ibrutinib, we crossed the B6/TCL1 mouse with the XID mouse, which harbors a mutation in the PH domain of BTK, rendering it kinase-inactive. In B-cells derived from these mice we observed continued TCL1 expression and diminished BCR signaling with BCR ligation. XID/TCL1 mice (n=65) had a significantly prolonged overall survival (OS) compared to those with wild-type (WT) BTK (n=79) (Median not reached versus (vs) 13.4 months respectively, p<0.0001), and at similar timepoints, XID/TCL1 mice were less likely to be leukemic as measured by CD5/CD19+ B-cells in blood than a WT TCL1 cohort. This study confirms both that kinase functional Btk is critical for leukemia development in this model, and that BTK is an important target of ibrutinib. Next, we wanted to determine whether ibrutinib was effective in this mouse model, and whether ibrutinib could inhibit CLL development similar to genetic manipulation of BTK. To evaluate the efficacy of ibrutinib, we transplanted leukemic TCL1 spleen lymphocytes into CB17/SCID mice, and at the time of leukemia diagnosis by flow cytometry, treated mice with vehicle or ibrutinib administered continuously via drinking water giving an approximate daily dose of 25 mg/kg. Mice treated with ibrutinib had significantly prolonged OS from the time of leukemia diagnosis compared to vehicle treated mice (46 days versus 24 days respectively, p=0.003), confirming that ibrutinib inhibits the expansion phase of CLL in this model. To evaluate the effects of ibrutinib on the development of CLL, we treated B6/TCL1 (n=24) and C3H/B6/TCL1 (n=24) mice from the time of weaning with continuous vehicle or ibrutinib administered via drinking water. We found that mice treated with ibrutinib had a significantly prolonged time to leukemia development compared to vehicle (p<0.0001 for B6/TCL1 mice and p=0.001 for C3H/B6/TCL1 mice), and that OS as well favors ibrutinib-treated mice (p=0.016). Because ibrutinib inhibits Btk in the tumor microenvironment as well as the tumor cells, we set to determine whether global BTK inhibition was necessary to inhibit leukemia development, or whether inhibition in the tumor cells or the microenvironment alone was sufficient. As ibrutinib is an irreversible inhibitor of BTK, to evaluate BTK inhibition in tumor cells alone, we treated TCL1 spleen lymphocytes with vehicle, 1uM ibrutinib, or 1uM fludarabine for 1 hour in vitro prior to engraftment into CB17/SCID mice (n=124 total). The three conditions were not significantly different in terms of leukemia free or overall survival (p=0.26, p=0.98 respectively), suggesting that inhibition of BTK in the CLL cell alone is not sufficient to prevent tumor engraftment following transplant. To determine whether inhibition of BTK in the microenvironment alone was sufficient to inhibit leukemia development, we transplanted leukemic B6/TCL1 lymphocytes into B6 or XID mice (n=46 per group). There was no difference in the time to development of leukemia between B6 mice and XID mice, demonstrating that BTK inhibition in the tumor microenvironment alone is not sufficient to inhibit leukemia development.In conclusion, our data show that active BTK is critical for the development and expansion of CLL in the TCL1 mouse model of CLL. Continuous global inhibition of this kinase through mutation or pharmacologic inhibition delays the development of CLL, confirming that BTK is an important target for drug development. Disclosures:Chang:Pharmacyclics, Inc.: Employment. Buggy:Pharmacyclics: Employment, Equity Ownership. Byrd:Pharmacyclics: Research Funding.

The Activation of STAT3 and 5 in t(6;9)(DEK/CAN)-Positive Acute Myeloid Leukemia Models Is Inhibited by Exposure to Arsenic Trioxide

AbstractAbstract 1384▪▪This icon denotes a clinically relevant abstractIn acute myeloid leukemia (AML) chimeric genes derived from specific chromosomal translocations, such as t(15;17), t(8;21) or t(6;9) encode aberrant fusion proteins (FPs). These Fps, i.e. PML/RARα (PR), AML-1/ETO (AE) or DEK/CAN (DC), recapitulate the leukemic phenotype in the mouse. In about 50% af AMLs constitutive activation of STAT3 and/or 5 is observed and it is postulated that the aberrant stem cell capacity in leukemia is correlated to the activation of STATs. Furthermore it has been shown in models of the t(15;17)(PR)-positive AML that there is a relationship between STAT-activation and response to arsenic trioxide (ATO). t(6;9)(DC)-positive AML is classified as a separate entity, because of its young age of onset and poor prognosis. The mechanisms by which DC induces leukemia are nearly completely unknown.In order to investigate whether the STAT-activation plays a role for the pathogenesis of t(6;9)-positive AML we investigated the activation of STAT3 and 5 in models of DEK/CAN positive leukemia.As models of DC-induced AML we used either primary Sca1+/lin- murine hematopoietic stem and progenitor cells (mHSPC) retrovirally transduced with DC or as controls with PR. Samples of DC-positive pre-leukemia as well as primary leukemia and 2° leukemia from established DC-induced AML were used for the in vivo studies. Expression and activation of STAT3 and 5 was assessed by either intracellular FACS or Western blotting using antibodies against the phosphorylated (activated) and total STAT3 or 5. The effect of constitutively activated STAT3 (STAT3*) on the stem cell capacity was studied by the retroviral expression of STAT3* in mHSPC which then were subjected to serial replating and colony-forming unit spleen day 12 (CFU-S12) assays. ATO treatment was performed on sublethally irradiated recipients (8 mice/group), from day 5 after inoculation of DC-positive primary leukemic spleen cells for the induction of a 2° leukemia for 14 days. Leukemia development and response to ATO was assessed 30 days post-transplantation by spleen size and determination of the proportion of leukemic cells in the peripheral blood (PB), spleen and bone marrow (BM) in comparison to solvent treated controls.Here we show that i.) both DC and PR strongly activated both STAT3 and 5 in primary blasts from mouse leukemia as as well as in retrovirally transduced primary HSCPs; ii) the expression of constitutively activated STAT3 (STAT*) in HSC led to an increased replating efficiency in the presence of mIL-3, mIL-6 and SCF; iii.) STAT* increased CFU-S12 as a read out for short-term stem cell and early progenitor potential to a similar extent as PR. iv.) the pre-leukemic state of DC-positive AML was characterized by the only activation of STAT3 but not of STAT5; v.) STAT5 activation appeared only in the full blown leukemia; v.) exposure to ATO inhibited phosphorylation of STAT3 and STAT5 in vivo in the PB of the leukemic mice; vi.) the treatment with ATO significantly decreased spleen size and the proportion of leukemic cells in the BM and spleen as compared to solvent treated controls.In summary, our results strongly suggest a direct relationship between the expression of the leukemogenic oncogenes DC or PR and the activation of STATs, which seems to play an important role for the maintenance of the leukemic stem cell. Furthermore ATO seems to represent a novel therapeutic option for the t(6;9)-positive AML. Disclosures:No relevant conflicts of interest to declare.

Leukemic spleen cells are more potent than bone marrow-derived cells in a transgenic mouse model of CML

Spleen size ranks among the most important risk factors in chronic myeloid leukemia (CML), but the pathogenic mechanisms of splenic hematopoiesis in CML remain poorly defined. Here, we studied the biology of Bcr-Abl positive leukemia-initiating cells in the spleen, using an inducible transgenic mouse model of CML. Disease kinetics showed greater increases of immature leukemic cells in spleen vs bone marrow (BM). To assess how Bcr-Abl alters the behavior of spleen-derived CML cells, we transplanted these cells either before ('pre-uninduced') or 44 days after ('pre-induced') expression of the oncogene. Mice transplanted with pre-induced spleen cells showed significantly increased neutrophilia and splenomegaly compared with mice receiving pre-uninduced spleen cells, suggesting that Bcr-Abl expression in the donors had increased splenic tumor burden. However, pre-induction also altered the biology of these cells, as shown by a striking increase in erythropoietic potential. These results differ from those of BM-derived CML stem cells where pre-induction of Bcr-Abl had previously been shown to decrease disease transplantability. Moreover, splenic cells were less sensitive to imatinib than BM cells. In conclusion, Bcr-Abl alters the biology of splenic leukemic stem cells by a cell-autonomous mechanism, but the disease phenotype is also influenced by the microenvironment of these cells.