MIF Involvement in AML

The study aims at studying the involvement of MIF in adult AML.

1. DESCRIPTION OF THE ELEMENT OR ELEMENTS UNDER INVESTIGATION 1.1. Measurement of MIF concentrations Plasma and BM supernatant from patients and controls (100 age-matched controls for blood and BM plasma, available at Tours Hospital, ClinicalTrials.gov #NCT02789839 ) will be analyzed by ELISA. BM supernatants will be obtained after a 2-step centrifugation procedure: 1 mL of fresh BM from AML patients or control BM will be centrifuged at 540 g for 5 min to recover viable cells in the pellet (for cytometric and expression studies); the supernatant will be centrifuged at 1000 g/15 min/4°C to prepare the plasma. MIF concentrations in BM and blood plasma will be systematically compared. As MIF concentrations may be affected by different factors, including inflammation, leukocytosis, and monocytosis, blood cell counts and plasma C-reactive protein levels will be measured and taken into account in the interpretation of the data.

1.2. Measurement of MIF, CD74, and CXCR4 expression Hematopoietic stem cells, progenitor cells and blasts from AML patients will be sorted according to the expression of CD34, CD38, CD90, CD45-RA, IL-3R antigens, taking into account the initial immunophenotype of the disease. Hematopoietic stem cells and progenitor cells from control BM samples will be sorted similarly. MIF, CD74 and CXCR4 RNA expression will be studied by quantitative RT-PCR in sorted cell subpopulations. The surface expression of CXCR4 and anti-CD74 as well as intracellular CD74 will be studied by multiparametric flow with cytometry.

1.3. Analysis of MIF, CD74, CXCR4 expression, and MIF concentrations at time of complete remission.

The analyses performed in a) and b) will be repeated in complete remission samples in 30 to 50 selected patients. Patients will be selected according to their response to therapy (assessment of complete remission and normal blood counts), and to the genotype of their disease (15-25 patients with either TET2 or DNMT3A mutations at diagnosis vs 15-25 patients without these mutations). Samples will be collected after two chemotherapy courses at time of remission evaluation. Whenever possible, bone marrow controls of patients in long term remission (> 1 year post diagnosis) will be collected and studied. In the meantime, NGS, ddPCR, FISH or RT-PCR techniques will be used to assess the persistence or absence of the initial AML lesions.

It is expected that complete material will be obtained for 80% of included patients making the objective of 240 patients achievable in less than three years. Cell purification may be an issue in the context of AML with aberrant immunophenotype: this will be taken into account by the use of appropriate FACS protocols. Regarding the focus on MIF pathway, it cannot rule out that other chemokines/cytokines, or others molecules may be involved in the communication of AML cells and their micro-environment. Thus, aliquots of BM and blood plasma will be cryopreserved to allow a refinement of our strategy (for example to perform cytokine arrays, ELISA, or MS analyses which are available at Saint-Antoine Hospital).

1.4. Xenotransplantation model of primary AML in NSG mice Xenotransplantation models of primary AML cells has been developed in a cohort of 70 AMLs in which MIF expression will be analysed, the cytogenetic and molecular genotype, and correlate these parameters to different engrafting phenotypes.

1.5. MIF inhibition in NSG mice transplanted with AML cells and in NSG mice transplanted with TET2-depleted or DNMT3A-depleted cord blood HSPCs The role of MIF on leukemia development in NSG mice xenotransplanted with primary cells will be accessed. In these experiments, primary AML patient cells from 6 patients with TET2 mutations of our cohort that engraft will be used to generate leukemia. When leukemic development will be seen in blood (between 8 to 10 weeks), MIF expression will be measured in the serum of mice. Mice will be randomized into three groups with 7 mice per group: One group will be treated with a commercially available MIF inhibitor (ISO-1) and a second group with an anti MIF antibody and one group will be treated by the vehicle for three weeks (one injection a week). Mice will then be monitored to detect pre-leukemic engraftment and AML development using weekly blood sampling and anti-human CD45 staining.

A model of xenotransplantation by cord blood CD34+ cells transduced with lentivectors expressing TET2 shRNAs have been developed. These engineered CD34 positive cells showed increased MIF expression and secretion and demonstrate enhanced multilineage NSG engraftment. Il will be tested whether inhibition of MIF using MIF inhibitor or anti MIF antibody leads to a correction of the enhanced multilineage NSG engraftment by TET2-depleted cells when compared to control cells. Similar experiments will be performed with a DNMT3A-depleted model using DNMT3A shRNAs that are currently available in the laboratory. In addition, the effects of MIF inhibition (inhibitor and antibody) and recombinant human MIF will be analyzed in NSG repopulation assays of unprocessed normal cord blood CD34+ cells.

1.6. MIF inhibition in NSG mice transplanted with cells from patients in complete remission with persistent clonal hematopoiesis (DNMT3A / TET2 mutant) Complete remission samples from six patients with persistent clonal hematopoiesis in complete remission will be selected and treated as in a) to test whether MIF inhibition is also able to counteract NSG repopulation by residual pre-leukemic stem cells.

It is expected that MIF inhibition will impair AML development in mice. However, stability and side effects of anti MIF antibody or inhibitors may be an issue in treated animals. In this case, shRNAs will be used to knock-down MIF in AML cells before injection to NSG mice. Another issue may be the need for co-injection of MSCs at time of xenotransplantation to increase the relevance of the model. If needed, normal MSCs and AML MSCs will be collected, prepared, and co-injected with AML cells.

1.7. Full characterization of the up-regulation of MIF in TET2 depleted cells How TET2 might impact gene regulation remains still not completely understood. Recently, Cao X's team has shown that loss of Tet2 resulted in the upregulation of several inflammatory mediators. In this model, Tet2 recruited Hdac2 and repressed transcription of Il6 via histone deacetylation. It has been found that TET2 binds the proximal promoter of MIF gene which is enriched in CpG islands in Kasumi cells. Loss of TET2 results in MIF accumulation in several leukemic models through EGR1. i) MIF promoter methylation will be analyzed by sequencing after bisulfite treatment in normal CD34+ cells (20 samples) and cell populations from AMLs sorted as in c)(150 samples) ii) EGR1 mRNA expression will be analyzed by qRT-PCR in normal and AML pre-leukemic stem cells to see if there is a correlation with MIF expression as observed in other models iii) ChIP-PCR analyses focusing on MIF promoter will be performed for EGR1, H3K4me3, H3K27me3 and HDAC2 in control and AML patient samples. If EGR1 is not involved in this system, MIF promoter luciferase assay will be used to identify transcription factor sites that are essential for its up-regulation in AML samples and validate the targets by ChIP-PCR. All these technologies are available.

1.8. bidirectional communication between MSCs and AML cells or normal HSPCs The impact of MIF overexpression in AML cells will be tested on the physiology of normal MSCs (differentiation, sustain of hematopoiesis, production of cytokines / chemokines, interactome). For this aim, normal primary MSCs from adult bone marrow (already available) will be isolated and expanded as described (Reinisch A et al, Blood 2015, 125:249-260). The Cells will be amplified up to passage 2 (P2). MSCs will be treated with different concentrations of recombinant MIF for 24 and 48 hours and the proliferation, production of cytokines/ chemokines will be evaluated. As it has been reported that MIF stimulates glycolysis, the effects of MIF will also decipher will be also deciphered in the energetic metabolism of MSCs by measuring mitochondrial respiration and glycolysis of MSCs (Seahorse XF extracellular flux analyzer) and by determining their MIF-induced metabolic signature using high throughput multiplexed fingerprinting (Omnilog Biolog). To model the effects of MIF overexpression by AML cells, will perform co-culture experiments between human primary AML cells (5 samples) that express or not (shRNA) of MIF. After 24 and 48 hours, MSCs will be sorted and their proliferation, cell cycle and differentiation potential into osteoblasts, adipocytes and chondrocytes will be evaluated. Previous studies indicate that MIF-CXCR4 is a novel axis for MSC recruitment to tumors in vivo.

Thus it will be studied whether recombinant MIF may recruit MSCs using in vitro transwell assays. Effect of MIF production by AML cells on MSCs recruitment will be tested using AML conditioned medium. AML cells engineered with MIF shRNA will be used in this context. As it cannot rule out that MSCs may also contribute to MIF production, mRNA expression will be tested by primary normal MSC cocultured with AML cells.

It is expected that recombinant MIF will have an impact on either proliferation, cell cycle, differentiation potential or migration of MSCs. However, MIF produced in BM microenvironment could also modify the biology of other populations of the stem cell niche including osteoblasts, endothelial cells or macrophages. To investigate alternative cells on which MIF may have an impact, it will be test its effects on these populations. MSCs will be induced to differentiate into osteoblasts, adipocytes or chondrocytes. MIF will then be added to investigate proliferation, cell cycle and migration of populations of interest. Effect on osteoblasts and endothelial cells will be modelized using MG63 and Huvec cell lines, respectively.

1.9. Study population 1.9.1. Participant recruitment 300 adult patients (>18) newly diagnosed with AML will be included in the study. 85 patients diagnosed since novembe 2015 at Saint-Antoine Hospital will be retrospectively included in the study. 215 patients newly diagnosed with AML at Saint-Antoine and Tours University hospitals will be prospectively included [>100 AML patients per year are referred to Saint-Antoine hospital (>70 patients per year) and Tours University Hospital for AML (>30 patients per year)].

1.9.2. Participant follow-up

Patient sample collection :

Only standard care blood and bone marrow samples will be used in the study :

• Blood samples for biochemistry, blood cell counts, immunophenotyping, cytogenetics, molecular biology analyses, and tumor bank cryopreservation
• Bone marrow samples for immunophenotyping, cytogenetics, molecular biology analyses, and tumor bank cryopreservation

Time points:

Only standard care time point samples will be collected:

• Diagnosis time point (day 0, Dx sample)
• After the first course of chemotherapy (C1 time point)
• After the second course of chemotherapy (C2 time point)
• After the third course of chemotherapy (C3 time point)
• In case of relapse or refractoriness,
• at time of diagnosis of first relapse or refractoriness (RR1)
• at time of diagnosis of second relapse or refractoriness (RR2)
• Long term remission control samples at 1, 2, 3 years after diagnosis (LTR1, 2, 3)