Valeria BISIO1-3, Bérénice SCHELL1-3, Marion ESPELI1-3, Karl BALABANIAN1-3, Nicolas DULPHY1-4
1Institut Carnot OPALE, Hôpital Saint-Louis, F-75010 Paris, France
2Université Paris Cité, Institut de Recherche Saint Louis, INSERM UMR_1342, F-75010 Paris, France
3Leukemia Institute Paris Saint-Louis, F-75010 Paris, France
4Laboratoire d’Immunologie et d‘Histocompatibilité, Assistance Publique des Hôpitaux de Paris (AP-HP), Hôpital Saint-Louis, F-75010 Paris, France
The Acute Myeloid Leukemia microenvironment: a complex multicellular landscape.
Hematopoiesis is the process which allows the production of all blood cells, including red blood cells and immune cells. It mainly takes place in the bone marrow and is based on the differentiation of the hematopoietic stem cells (HSC) to a broad variety of progenitors and mature cells. However, HSC are not self-sufficient and require the support of the so-called bone marrow stroma, i.e. non hematopoietic cells which constitute the bone marrow landscape including the cells of the bone (osteoblasts and osteoclasts), fat tissues (adipocytes), blood vessels (endothelial cells) or nerves (neurons). Within this ecosystem of hematopoietic and non-hematopoietic cell subsets, HSC are localized in particular places, called hematopoietic niches, where they closely interact with the mesenchymal stromal cells (MSC). MSC not only support the hematopoietic differentiation but also give rise to adipocytes and osteoblasts among others.
Acute myeloid leukemia (AML) and related diseases, such as myelodysplastic neoplasms (MDN), also develop in the bone marrow and are characterized by the tumoral transformation of the HSC through successive mutational hits. AML leads to the accumulation of leukemic blasts within the bone marrow. The bone marrow stroma is absolutely required for AML initiation and progression. In return, the emergence of the tumoral cells will induce numerous perturbations into the bone marrow microenvironment: impaired differentiation of MSC to osteoblasts or adipocytes, reduced hematopoiesis, inhibition of immune cells including the ones (T lymphocytes, Natural Killer cells) involved in the detection and elimination of the tumor cells. Mechanisms leading to these perturbations can be associated to cell-cell contacts, notably during leukemic cell invasion, or soluble factors. They can be directly induced by the tumor cells, or indirectly by stromal cells affected by their pathological neighborhood. Studying the bone marrow microenvironment in the physiological or pathological contexts is the subject of tremendous efforts by many scientific teams worldwide. Investigating the diversity of the cell-cell crosstalk requires to model the pathological bone marrow microenvironment. Such models can first help to understand the leukemic process, how the disease emerges and how it evolves; and second to evaluate the pertinence of particular therapeutic strategies, including chemotherapy, targeted therapy or immune-based therapy.
How to model bone marrow complexity?
Among the various models developed over the past years to study the complexity of the tumor microenvironment, including hematologic malignancies, 3-dimensions (3D) in vitro cell culture models, systems so-called “organoids”, attracted the attention of the scientific community. Organoids are defined as small and simplified versions of an organ or tissue, grown in laboratory, with the use of stem, progenitor or mature cells. In function of their complexity, the word “organoid” can refer to a broad spectrum of systems, beginning with the relatively simple self-aggregating spheroids to the highly sophisticated “organ-on-chips” that use microfluidics devices to mimic tissue perfusion and/or blood circulation. Their goal is to recapitulate as close as possible the organization, three-dimensional structure, and physical and accessibility constraints of a particular tissue. The 3D in vitro organoids present the advantages to be reliable and versatile. Importantly, these systems can be developed as fully human models, avoiding the use cell lines from mouse or other animals. Therefore, they are compliant with the three Rs (Replacement, Reduction, Refinement) guiding principles for an ethical use of animals in medical research. If the organoids cannot totally replace animal models yet, they can strongly reduce their use without affecting the pertinence and reliability of the scientific observations.
To summarize, the challenge in modeling the bone marrow microenvironment during AML onset is to collect relevant cell subsets (healthy HSCs, leukemic cells, stromal cells, immune cells) from clinical bone marrow samples (aspirates or biopsies), and to organize them altogether in such a structure which could reproduce cell-cell direct or indirect interactions as well as nutrients and oxygen supply. In that context our group setup an original 3D self-organizing in vitro model to recapitulate the differentiation and production of innate immune cells (polynuclear neutrophils, monocytes, natural killer lymphocytes). This model opens new avenues to investigate the mechanisms of the hematopoietic and immune impairment during leukemia invasion, the role of the mesenchymal stroma in these processes, and to test new therapeutic strategies or drug combinations to restore hematopoiesis and repress leukemogenesis.
The Artificial Marrow Organoid – AMO model.
The AMO is an easy-to-access, reliable and reproducible self-organizing in vitro 3D model for studying progenitor and mature hematopoietic cells in the frame of their physical crosstalk with the human mesenchymal BM stroma.
The model is based on mesenchymal stromal cells from bone marrow aspirates harvested from individuals who were volunteers for such procedure. Herein, MSCs are used as a soft passive skeleton of the 3D structure as well as an active hematopoietic support. HSCs are added into the organoid to induce the hematopoietic differentiation of stem cells into progenitor and mature white blood cell subsets. The AMO is placed on the top of a permeable membrane at the liquid/air interphase, with a specific cell culture medium being localized under the membrane. Then soluble factors necessary for the cell development and differentiation are added into the culture medium to irrigate and diffuse into the organoid. Altogether, the MSCs and the additional soluble factors allow to initiate, orientate and support the hematopoietic differentiation to mature white blood cells, including anti-leukemic cytotoxic lymphocytes (Natural Killer cells) or anti-microbial myeloid cells (polynuclear and monocytes). Subsequent analysis can then be performed on either the whole 3D structure such as fluorescent microscopy, or, after the disruption of the organoid, on each of the constitutive cell subsets. Importantly, HSCs are still present at high frequencies after weeks of culture in the AMO. These cells keep some stemness capacities for all this time and can be transferred into a new AMO in an iterative in vitro culture. This procedure allows deeply analyzing the hematopoietic processes and the mechanisms by which the stroma support them.
The AMO model: A Platform for Understanding Leukemia Emergence and Progression, with Applications in Personalized Medicine
Based on this model, our group now develops new projects to study how preleukemic mutated HSC can give rise to a blood cancer. Thanks to patients’ implication, we can build the AMO with pathological cells (leukemic cells, MSC or immune cells from the leukemic BM, etc…) and decipher the interplay taking place between the different cell subsets. We developed experimental protocols to dissect the molecular signals delivered by each cell partner to the others. With this “bed-side to bench” projects, we aim to identify molecular pathways specific to some category of patients, and which could be targetable using drug repositioning or drugs association. The ultimate development of the AMO model will be to use this platform, associated with drug screening platforms, to design a personalized treatment for each patient , based on the precise immune and stromal characteristics of their disease.