Effective control of acute myeloid leukaemia and acute lymphoblastic leukaemia progression by telomerase specific adoptive T-cell therapy

Telomerase (TERT) is a ribonucleoprotein enzyme that preserves the molecular organization at the ends of eukaryotic chromosomes. Since TERT deregulation is a common step in leukaemia, treatments targeting telomerase might be useful for the therapy of hematologic malignancies. Despite a large spectrum of potential drugs, their bench-to-bedside translation is quite limited, with only a therapeutic vaccine in the clinic and a telomerase inhibitor at late stage of preclinical validation. We recently demonstrated that the adoptive transfer of T cell transduced with an HLA-A2-restricted T-cell receptor (TCR), which recognize human TERT with high avidity, controls human B-cell chronic lymphocytic leukaemia (B-CLL) progression without severe side-effects in humanized mice. In the present report, we show the ability of our approach to limit the progression of more aggressive leukemic pathologies, such as acute myeloid leukaemia (AML) and B-cell acute lymphoblastic leukaemia (B-ALL). Together, our findings demonstrate that TERT-based adoptive cell therapy is a concrete platform of T cell-mediated immunotherapy for leukaemia treatment.


INTRODUCTION
Acute myeloid leukaemia (AML) and B-cell acute lymphoblastic leukaemia (B-ALL) are still incurable diseases for a number of patients and a leading cause of death in children and adults, respectively [1][2][3]. Currently, the worldwide standard treatments, including induction chemotherapy followed by either consolidation Priority Research Paper chemotherapy, autologous or allogeneic stem cell transplantation, generated remission in a high percentage of treated patients: approximately, 90% of B-ALL patients diagnosed between 1 and 18 years of age are expected to be long-term and event-free survivors [4][5][6], and 70-80% of AML patients younger than 60 years can achieve complete remission [7,8]. However, the outcome remains insufficient for patients who develop adverse disease-related events or treatment-related toxicity. In fact, the 40-50% of AML patients that undergo relapse showed a 5-year overall survival (O.S.) [9][10][11]. On the contrary, the 5-year O.S. of relapsed ALL patients was 10% in adults [12,13] and 10-20% in children [9,14]. In the last years, the immunotherapy of cancer has emerged as realistic treatment option for B-ALL and AML patients [15][16][17]. The minimal residual disease negative status of treated patients, which is around 79%, suggests that immunotherapy promotes long-lasting efficacy compared to standard procedures [18]. Also the use of Blinatumomab, bispecific T-cell engager (BiTE) directed towards CD19 and CD3 antigens, provided impressive clinical effects in B-ALL patients with either relapsed or refractory disease [19][20][21]. These exciting results in B-ALL have generated several adaptations to treat AML patients that are now under clinical evaluation. Since CD33 and CD123 markers are expressed on the majority of AML cells, they have been used to generate innovative tools such as CD33/CD3 (AMG330) [22,23] or CD3/WT1-HLA-A * 0201 [24] BiTEs or chimeric antigen receptors (CAR) targeting CD123 epitope [25][26][27]. BiTE demonstrated specific ability to contrast tumour progression in AML mouse models and a selective in vitro killing ability against human AML blasts, but currently no studies about toxicity has been reported yet. On the contrary, preclinical experiments using a selective hCD123 CAR able to eliminate human AML cells caused complete eradication of normal bone marrow (BM) cells in mice engrafted with human CD34 + stem cells [27]. These data also emphasize the deleterious effects on normal myeloid cells caused by the use of potent immunebased therapies specific for a widespread antigen and highlight the relevance of selecting the correct target for the development of anti-cancer immunotherapy.
Human Telomerase (TERT) has been identified as a common hallmark of cancer, since it plays a critical role in aberrant cell proliferation and immortalization in the majority of tumours [28]. Variable levels of telomerase have been detected in up to 85% of all AML [29][30][31] and, normally, relapsed AML patients showed highest telomerase activity [30]. Among all the subtypes of acute leukemia, B-ALL cells showed the greatest level of telomerase activity and the shortest telomeres, conditions generally associated with reduced response to therapy, faster leukemic progression and poor prognosis [32][33][34]. All these findings designate TERT as ideal tumourassociated antigen (TAA) that could be exploited to design a selective cancer immunotherapy for the treatment of leukemias [35]. Indeed, TERT generates immunogenic epitopes for both major histocompatibility complex (MHC) class I and II pathways, able to trigger an adaptive cytotoxic T lymphocytes (CTL) response against tumour cells [36,37]. A spontaneous immune response against TERT was reported in different tumour settings and anti-TERT specific CD8 + T cells were detected, with a higher frequency, in the blood of patients affected by chronic lymphocytic leukemia (B-CLL), as well as breast, lung and colorectal cancers, compared to healthy donors (HD) [38][39][40][41][42]. However, the endogenous anti-TERT T cells isolated from B-CLL normally display a very low affinity in their TCR, thus limiting their use in adoptive cell therapy (ACT) [38]. To overcome this limitation, we recently demonstrated the ability of hTERT 865-873 -specific, TCR-engineered T-cells both to efficiently recognize different solid human tumour cells and restrict human B-CLL tumour progression in vivo without inducing dramatic toxicity. In fact, the hTERT-specific ACT did not induce myeloid precursor depletion in tumour-bearing humanized mice, affecting only BM resident mature granulocytes and preserving the ability of hCD34 + cells to generate mature leukocytes [38]. We describe here the exploitation of the therapeutic efficacy of our anti-TERTbased ACT approach in more aggressive haematological cancer settings, such as AML and B-ALL, to validate its versatility as a widespread anti-tumour immunotherapy for leukemic diseases.

RESULTS hTERT 865-873 -specific, TCR-engineered T-cells reduce AML progression in vivo
In our recent work, we characterized the anti-cancer efficacy of hTERT 865-873 -specific, TCRengineered T-cells transferred to human tumour-bearing immunocompromised (NOG) mice. In particular, we emphasized the ability of TERT-based ACT to promote a strong control of tumor growth in several cancer settings, while ignoring hematopoietic progenitors and activated T lymphocytes [38]. Conversely, hTERT 865-873 -specific, TCRengineered T-cells specifically induced the contraction of the mature granulocytic cell subset resident in the BM of human immune reconstituted (HIR) mice [38]. Starting from these data, we explored the killing ability of our hTERT 865-873 -specific TCR-engineered T-cells on AML cells. We initially verified TERT expression and activity in both PBMCs isolated from HLA-A2 + AML patients or HD and THP1 human myeloid leukaemia-immortalized cells ( Figure 1A). All leukemic cells showed a heightened TERT activity compared to normal leukocytes. Moreover, hTERT 865-873 -specific TCR-engineered T-lymphocytes were able to discriminate normal from tumour cells: in fact, in vitro, they efficiently recognized AML PBMCs and THP1 cells, but not HD PBMCs ( Figure 1B). Indeed, hTERT 865-873 -specific TCR-engineered T-cells secreted interferon (IFN)-γ in response to patients' samples, at similar levels to those reached using THP1 cells as target ( Figure 1B, left panel). Cell recognition was confirmed by a cytofluorimetric cytotoxic assay in which leukemic cells were selectively lysed by hTERT 865-873 -specific TCR-engineered T cells ( Figure 1B, right panel). The hTERT 865-873 TCR-engineered T-cells are specific for the target antigen, since they recognized PBMCs pulsed with hTERT 865-873 peptide but not PBMCs pulsed with control hHCV 1406-1415 peptide ( Figure 1B). These data highlight that AML blasts and tumour cells were able to process TERT protein to generate the HLA-A2 class I/hTERT 865-873 peptide complex able to activate anti-TERT CTLs.
To test the therapeutic effect of hTERT 865-873 -specific TCR-engineered T-cells on controlling AML progression, we subcutaneously (s.c.) challenged immunodeficient NOG mice with THP1 cells. Our immunotherapeutic treatment based on anti-TERT CTLs infusion significantly controlled tumour growth inducing a survival benefit on treated mice compared to mice treated with hHCV 1406-1415 -specific TCR-engineered T-cells (Figure 2A-2B).
Subsequently, we generated firefly luciferase-expressing THP1 cells (THP1-Luc) to track the in vivo spreading of AML cells after intravenous (i.v.) injection delivery, to mimic the disseminated disease in patients. THP1-Luc cells were recognized in vitro by the hTERT 865-873 -specific TCR-engineered T-lymphocytes at levels comparable with wild type (WT) THP1 cells (data not shown). Immunocompromised NOG mice were injected i.v. with 3x10 5 THP1-Luc cells and treated with three weekly infusions of either hTERT 865-873 -specific TCRengineered or hHCV 1406-1415 -specific TCR-engineered T-cells seven days after tumour challenge. Tumour progression was evaluated through bioluminescence imaging. Figure 2C demonstrates the ability of TERTbased ACT to significantly limit leukemic progression. We monitored tumour dissemination until day twentyseven from tumour challenge when control group reached the endpoint threshold (3x10 6 p/s/cm 2 /sr). Lymphopoietic organs (spleen and BM) isolated from control HCV-based ACT treated mice presented a more severe leukemic cell infiltration compared to TERT-based ACT treated mice ( Figure 2D). Finally, to confirm the therapeutic effectiveness of TERT-based ACT, we analyzed THP1 accumulation, identified as human (h) CD45 + cells by flow cytometry analysis ( Figure 2E). These data were confirmed also by immunohistochemistry (IHC), showing a significant contraction of leukemic cells in TERT-based ACT treated mice compared to controls ( Figure 2F). All these findings suggest that TERT antigen can be pursued as a target for T-cell therapy of AML.

hTERT 865-873 -specific TCR-engineered T-cells reduce ALL progression in vivo
We previously demonstrated that hTERT 865-873specific TCR-engineered T-cells were able to significantly control B-CLL progression [38]. Once the efficacy of the adoptive transfer of TERT-specific T cells against human AML was confirmed, we investigated their antitumour activity towards human B-ALL. B-lymphocytes isolated from HLA-A2 + B-ALL patients and the HLA-A2 + ALL-CM immortalized cell line (both co-expressing hCD19 and hCD20 markers) were characterized by higher levels of telomerase expression and activity, when compared to B cells isolated from HD PBMCs ( Figure 3A). Moreover, hTERT 865-873 -specific TCR-engineered T-cells efficiently recognized in vitro leukemic cells compared to normal B cells ( Figure 3B). To confirm B-ALL targeting capability, NOG mice were s.c. challenged with ALL-CM cells and then treated with either hTERT 865-873 -or hHCV 1406-1415 specific TCR-engineered T-cell infusion. The anti-TERT based immunotherapy significantly controlled tumour progression ( Figure 4A) improving the survival of leukaemia-bearing mice ( Figure 4B). To evaluate the therapeutic impact of TERT-targeting immunotherapy in a more relevant setting, NOG mice were i.v. injected with Assessment of hTERT-specific ACT capability to ameliorate mouse survival upon THP1-Luc subcutaneous challenge. Either hTERT 865-873or HCV 1406-1415 -TCR-engineered T cell were transferred twice, followed by the i.p. administration of IL-2 to test the ability to control tumor growth. B. Kaplan-Meier survival analysis (hTERT n = 8; hHCV n = 6) of a representative experiment. C. NOG mice were then intravenously injected with THP1-Luc cells and treated with three, weekly ACTs of hTERT 865-873 -specific or control HCV 1406-1415 -specific T-cells (representative data of 1 out of 2 independent experiments of total n = 12 mice per group). Tumor growth was weekly evaluated by bioluminescence imaging. D. At 27 days from tumor-challenge, leukemic cells spread to spleen and BM was tested by bioluminescence imaging. E. The percentage of infiltrating malignant hCD45 + cells was also evaluated by Flow cytometric analysis and F. by IHC (20X magnification). Statistical analysis was performed with Student's t test. www.impactjournals.com/oncotarget 2.5x10 6 ALL-CM cells and tumour progression was weekly assessed by enumerating the frequency of circulating hCD19 + cells by flow cytometry, as previously described [43,44]. We set up a threshold of ~60% of human B cells in the peripheral blood as disease endpoint. Three weekly administration of hTERT 865-873 -specific TCR-engineered T-cells significantly controlled ALL progression compared to control treatment (HCV-based ACT, Figure 4C). This therapeutic effect resulted in a significant reduction of tissue-infiltrating hCD20 + cells compared to controls ( Figure 4D). For a translational perspective, we next evaluated the ability of TERT-based ACT to limit the in vivo progression of HLA-A2 + hCD19 + cells isolated from two different cryopreserved B-ALL patient's BM aspirates (referred as B-ALL#1 and B-ALL#2 respectively). We engrafted immunocompromised NOG mice with 1.5x10 6 hCD19 + cells and then assessed the leukemic progression by quantifying the frequency of circulating hCD19 + lymphocytes. Mice were treated with three consecutive ACT, starting one week after tumour challenge. The TERT-based immunotherapeutic approach significantly controlled leukemic expansion compared to the control therapy ( Figure 5A). To validate the therapeutic effect of TERT-based ACT, we sacrificed a cohort of leukemic mice when the control group showed around 60% of circulating hCD19 + lymphocytes. The IHC analysis of hCD20 + leukemic cells revealed a remarkable reduction in tumour accumulation in different tissues (spleen, BM, liver and kidney) in TERT-treated mice compared to control mice, therefore suggesting an effective control of leukemic spread ( Figure 5B). Indeed, TERT-based ACT promoted a significant survival benefit ( Figure 5C): in particular, in the case of mice engrafted with B-ALL#1 blasts, only the 16% of mice treated with hTERT-specific ACT developed, after 60 days from tumour challenge, an expansion of malignant human B cells that reached the threshold for sacrifice. The hTERT 865-873 -specific TCR-engineered T-cells ( Figure 6A) exhibited a very modest in vivo persistence; in fact their presence in peripheral blood of treated mice decreased as soon as 7 days after injection ( Figure 6B). These findings could in part explain the limited therapeutic impact on controlling tumour with more aggressive features in this experimental setting.

DISCUSSION
The results of our study demonstrated that: a) AML and ALL cells not only overexpress TERT protein but also show an higher TERT activity compared to normal cells; b) AML and ALL cells process TERT protein, generate the immunodominant hTERT 865-873 peptide and expose the immunogenic hTERT-HLA-A2 complex on the cell surface, promoting a specific T-cell recognition; c) hTERT 865-873 -specific TCR-engineered T-cells show specificity in recognizing both primary and immortalized AML and ALL cells but not target cells pulsed with unrelated antigen, normal PBMCs or B-cells; d) TERTbased ACT controls in vivo the systemic dissemination of AML and ALL cells; e) TERT-based ACT significantly prolongs the survival of ALL affected hosts. All these findings depict TERT-targeting immunotherapy as a promising strategy for treating leukemic patients. Although the outstanding clinical success achieved by CAR-mediated immunotherapy in contrasting B-ALL progression has comprehensibly attracted the interest of scientific community for the development and improvement of this type of engineered T-cell therapy, the idea of redirecting T-lymphocytes with high affinity TCR selective for universal tumour-associated antigens, such as TERT, may open new avenues on the clinical procedure. Indeed, TERT is shared and actively expressed in about 85% of human tumours with various histological type [45]. Transformed tumour cells initially express TERT to acquire both an "immortalized" phenotype preventing cell apoptosis induced by the telomeres reduction and inducing self-renewal ability [46][47][48]. In the late tumorigenesis  events, cancer cells develop mobilization and extravasation skills that promote the metastatic dissemination via TERT overexpression [49]. These findings demonstrate a key role for telomerase in different stages of tumour differentiation, therefore indicating a broad potential use of TERT-based ACT therapy. In agreement with this idea, our previous data showed both the ability of anti-TERT T lymphocytes to selectively destroy the pool of cancer stem cells (CSC), which usually survive conventional therapies favouring metastasis dissemination, and the therapeutic impact on controlling melanoma lung metastases [50]. We additionally proved the effectiveness of hTERT 865-873specific TCR-engineered T-cells ACT to control different human solid tumour growth [38]. Finally, we recently demonstrated the ability to further improve the therapeutic efficacy of TERT-based ACT by manipulating the cell composition in the tumour-microenvironment through the infusion of inducible nitric oxide synthase iNOS-and tumour necrosis factor (TNF) α-producing dendritic cells [51], as well as the use of either low dose of chemotherapy [52,53] or drugs able to contain the generation of nitrogen reactive species [54], such as peroxynitrites [55], within the tumour microenvironment to defeat the immunosuppressive cancer barriers, favouring the antitumour function and persistence of the infused T cells.
The limited therapeutic impact of our immunotherapy is probably due to the low in vivo persistence of hTERT 865-873 -specific TCR-engineered T-cells rather than their exhaustion. In fact, before their in vivo administration, TCR-transduced T-cells had an effector/memory phenotype (CD45RA -CD62L -CCR7cells) characterized by the low-expression of exhaustion markers (LAG3/PD-1/Tim-3). Other immuneevasion mechanisms, such as antigen loss or antigen presentation deficiency, seem not to play a critical role in our experimental setting since residual leukemia cells in TERT-treated mice maintained telomerase expression and HLA-A2 molecules (data not shown). As recently reported, the identification and characterization of T memory stem cells (T SCM ) generated an increased interest for their therapeutic use for cancer treatment. T SCM cells, that are the earliest developmental stage of memory T cells, exhibit stem cell-like properties and display a gene profile between naïve and central memory T cells  [56]. Moreover, T SCM cells show in preclinical models an intense self-renewing capacity following antigen loss, a long lifespan since they are refractory to apoptosis and a robust proliferative potential [57] but overall, as described by Biasco and colleagues, transferred genetically modified T SCM cells in human showed an extended longevity persisting in the host for up to 12 years post infusion [58]. T SCM cells in vivo persistence not required cytokine supply [58]; therefore, ACT based on engineered T SCM cells might completely eliminate all the adverse effects produced by the administration of high dose of IL-2 [59]. We plan to generate, in the next future, hTERT 865-873 -specific TCR-engineered T SCM -cells, taking advantage of recent protocols for T SCM cells in vitro isolation and expansion, to provide their increased benefit on tumor control [56,60], in association with checkpoint inhibitors, in order to design more appealing immunotherapeutic strategies. Obviously, a second limitation to envision the real impact of TERT-based therapy is linked to the selected xenograft mouse models, which do not recapitulate the integrated biology of an intact human immune system and the cross-talk between human transferred T cells with human stromal and endothelial cells, as well as with other human leukocyte subsets such as monocytes, granulocytes, macrophages, NK cells and T regulatory cells [61,62]. A partial mitigation of this limitations could come from the creation of novel leukemia-affected, humanized mice, which permit in vivo analysis of anti-leukemic cell-based therapy in the presence of a human immune system [63].
The most serious concern about the impact of TERT-based ACT in clinic is related to its safety. In fact, even if tumour tissues showed a higher TERT expression and telomere length compared to normal tissues, it is impossible to predict whether and to what extent normal cells would be recognized by TERT-specific T cells. Prudence is needed when designing a selective TERTtargeting immunotherapy, as pointed out by our previous data about the treatment of prostate tumor-bearing mice with three cycles of ACT with mouse TERT-specific CD8 + T cells that caused a transient B-cell lymphopenia. This toxicity might be related to the schedule and type of immunotherapeutic approach. In fact, the side effects were observed when repeated infusions of TERT-specific CTLs were associated with the injection of an adenovirus encoding mouse TERT protein and the administration of high dose of IL-2, whereas DNA vaccination did not induce any detectable adverse event [50,64]. However, to address issues about safety, it will be mandatory to assess the in vitro activation of hTERT 865-873 -specific TCR-engineered T-cells in contact with vital tissue slides of specimens with different histology. Moreover, to limit an uncontrolled activation of our transduced T cell that may produce severe immunopathologies, we plan to incorporate a suicide gene in the retroviral vector encoding TERT-specific TCR sequences. Indeed the use of an inducible caspase 9 (iCas9) "safety switch" has already produced excellent results on controlling graft-versus-host disease (GVHD) development in patients of haploidentical stem-cell transplants [65].
Collectively, our data proved and reinforced the potential clinical translation of TERT-based immunotherapy on treating different leukemic pathologies.

Primary and immortalized cell lines
PBMCs isolated from AML patients and HDs and BM (BM) cells from B-ALL were collected at the Hematology Unit, Azienda Ospedaliera Universitaria Integrata (AOUI) in Verona (Italy). All participating people provided written informed consent for the collection and use of their samples for research purposes, in compliance with the Declaration of Helsinki. The study was approved by the local Ethics Committee (Comitato Etico per la Sperimentazione, AOUI of Verona, n. 1496). THP1 acute monocytic leukemia cell line were obtained from American Type Culture Collection (ATCC), and were maintained in RPMI 1640 medium (Lonza). ALL-CM acute lymphocytic leukemia cell line were provided by Dr. A. Bondanza under restricted condition with Leiden University Medical Center (Leiden, The Neatherland) [43] and cultured in X-VIVO medium (Lonza) containing 2% human serum (Gibco). Both media were also supplemented with 2 mM L-Glutamine, 10 mM HEPES, 100 U/ml Penicillin, and 100 U/ml Streptomycin (all reagents were purchased from Lonza). All cell lines were maintained at 37°C in a 5% CO 2 atmosphere.

Telomerase expression and activity
Telomerase expression was assessed by western blotting as previously described [38]. Telomerase activity was measured using TRAP assay (TRAPeze ® XL telomerase detection kit; Millipore), following manufacturer's instructions.

Isolation of hCD19 + cells from BM of ALL patients
Crio-preserved BM aspirates from ALL patients were thawed and the total cells were sorted with hCD19-Microbeads (Miltenyi) according with the standard procedures. Purified cells were analyzed by flow citometry and they reached the ~95% of CD19 + cells.

Functional assays
IFN-γ secretion was quantified by a sandwich enzyme-linked immunosorbent assay (ELISA; eBioscience) according to manufacturer's instructions. Briefly, 3x10 5 engineered CTLs were co-cultured at 3:1 ratio in presence of with human malignant cells (PBMCs isolated from AML patients, purified hCD19 + cells isolated from BM of B-ALL patients, THP1 or ALL-CM cell lines) or control cell targets (PBMCs or purified hCD19 + cells isolated from HDs). Negative and positive controls were represented by HD PBMCs pulsed with either hHCV 1406-1415 or hTERT 865-873 peptides, respectively. After 24 hours (h) of co-culture, the supernatants were harvested to measure the released IFN-γ. Moreover, functional response was also evaluated by flow cytometric cytotoxic activity assay as previously described [38]. In brief, target cells were firstly stained with 0.2 µM or 2 µM CFSE or CellTrace (both from Invitrogen) allowing the evaluation of four different cell targets simultaneously. Mixed targets were incubated at 2:1 ratio with effectors lymphocytes for 18 h and then analyzed by flow cytometry. The formula: 1-[(% pre negative control / % pre sample) /(% post negative control / % post sample)]x100; was used to determine specific lysis.

Systemic treatment of subcutaneous tumours
For in vivo testing of the therapeutic activity of hTERT 865-873 -specific T cells, 5x10 6 THP-1 or 5x10 6 ALL-CM (in Matrigel -Corning, USA) cells were inoculated s.c. in the left flank of NOG mice. Tumour volume was calculated according to the following equation: V (mm3) = (d2 x D)/2, where d (mm) and D (mm) are the smallest and largest perpendicular tumour diameters, respectively, as assessed by caliper measurement. Treatments started when tumour volume reached 100 mm 3 . hTERT 865-873 -or control HCV 1406-1415 -specific T cells were injected i.v. in the tail vein once a week, for 2 consecutive weeks. Every adoptive transfer was followed by the administration of recombinant human IL-2 (30,000 IU/mouse) intraperitoneally every 12h, for a total of 6 doses.

ACT on human B-ALL setting
Eight-week-old NOG mice were intravenously (i.v.) injected with ALL-CM cells (2.5x10 6 ) or with hCD19 + cells isolated from a HLA-A2 + B-ALL patient (1.5x10 6 ). Mice were then treated with three, weekly ACT of hTERT 865-873 -specific or control HCV 1406-1415 -specific T-cells starting one week after tumour challenge, followed by IL-2 administration, as previously indicated. Leukemic spread was monitored as percentage of circulating hCD19 + cells in peripheral blood and we set up a survival threshold of hCD19 + circulating leukemic cells in the amount of either 60% in the case of ALL-CM tumour setting or 80% in the case of patient derived ALL tumour setting .

ACT on human AML setting
NOG mice (8-10 weeks of age) were firstly injected i.v. with 3x10 5 THP1-LUC cells obtained by infection of lentivirus built up with the pLENTI.EGFP.Luc vector, kindly provided by Dr. D. Melisi (University of Verona). THP1-Luc cells were sorted to obtain a 95% of EGFP + cell population that was used for the study. A total of 3 infusions of 2.5x10 6 hTERT 865-873 -or hHCV 1406-1415specific CTLs were retro-orbitally injected every 7 days, starting 1 week after tumour injection, followed by IL-2 administration, as previously indicated. Leukaemia burden was monitored by BLI (photons/second/cm 2 /sr) using the IVIS Spectrum Imaging System (PerkinElmer). Images were acquired prior to treatment and then weekly, 7 days after the last treatment. Animals were anesthetized with isoflurane/oxygen. For each animal, dorsal and ventral images were obtained pre and 10 minutes after i.p. administration with 150 mg/kg of D-luciferin (PerkinElmer) in PBS in order to discriminate specific light signal from basal emission. Five mice were imaged simultaneously with the subsequent parameters: exposure time = 5 minutes, field of view = 19×19 cm, binning B = 8 and f/stop = 1. Images were quantified tracing the region of interest (ROI) on the entire animal body. Living Image Software 4.4 (PerkinElmer) was used to acquire and quantify the bioluminescence.

Histological and immunohistochemical analysis
Spleen and BM samples were fixed in 10% neutral buffered formalin; after fixation, BM samples were decalcified with 10% EDTA (pH 7.4) for 3 weeks and embedded into paraffin; 5 µm slides were cut and stained with Hematoxylin (BioOptica) and Eosin (BioOptica) for histological examination. For immunohistochemistry, sections were deparaffinized, serially rehydrated and after the appropriate antigen retrieval procedure, incubated with the following primary antibodies: monoclonal Mouse Anti-Human CD45 (M0701, Dako Corporation) or monoclonal Mouse Anti-Human CD20 (M0755, Dako Corporation), followed by the secondary antibody Dako EnVision System-HRP Labelled Polymer Anti-mouse (K4001, Dako Corporation). After chromogen incubation, slides were counterstained in Hematoxylin and images were acquired by Leica DMRD optical microscope (Leica). The absence of cross-reactions between human and mouse antigens was verified testing anti-human antibodies on a normal mouse spleen. The percentage of positive cells was evaluated on the digital images of 5-6 samples per group (6-10 X 200 microscopic fields per sample) by 2 pathologists, independently and in a blind fashion.

Statistical analysis
Data were indicated as the mean ± SD. Student's t test was used to determine statistically significant differences between two treatment groups, while ANOVA test was used in case of multiple comparisons. Growth curves were analyzed with Repeated Measures (RM) ANOVA. Survival analysis was performed using the Kaplan-Meier survival analysis (Log-Rank) method. P values less than 0.05 were considered statistically significant.