3 – Bone marrow microenvironment and cytokine network
The bone marrow microenvironment (BMMe) is a complex network made up of a cellular compartment, an extracellular matrix (collagens, fibronectin, laminin, proteoglycans, and glycosaminoglycans), and soluble factors in a liquid milieu (cytokines, growth factors, and chemokines) . The cellular compartment is composed of hematopoietic cells including immune cells (macrophages and lymphocytes) and non-hematopoietic cells including MSCs (which can be defined as multipotent stromal cells or mesenchymal stem cells) and their progeny (osteoblasts and adipocytes), endothelial cells, and osteoclasts (which are derived from macrophages). In physiologic conditions, BMMe is essential to hematopoiesis: stromal cells create ‘niches’ that maintain hematopoietic cells and supply necessary factors for their development .
Bone marrow localization is a fundamental characteristic of MM; the dissemination of MM cells in other tissues and organs generally occurs at a late stage of the disease, suggesting that these cells are highly dependent on the BMMe. Hence in MM, the BMMe is defined as a ‘tumoral microenvironment’ because it plays a pivotal role in tumor growth, survival, migration, evasion of the immune system, and drug resistance of MM cells [22–24]. The involved mechanisms are complex, multiple, and overlapping. They include direct interactions between MM cells and BMMe cells (via adhesion molecules) or extracellular matrix components, and indirect interactions through soluble factors of cellular origin. These interactions are reciprocal as they occur from the BMMe toward MM cells, but also in the opposite direction creating a ’vicious cycle’ (Figure 1.1).
Cells from the BMMe secrete the chemokine stromal cell-derived factor 1α (SDF-1α), which allows the homing of MM cells through their CXCR4 receptor. The adhesion of MM cells to both the extracellular matrix and BMMe cells is mediated by several adhesion molecules including: CD44, syndecan-1 (CD138), very late antigen 4 (VLA4), very late antigen 5 (VLA5), leukocyte function-associated antigen 1 (LFA1), neuronal adhesion molecule (NCAM), intercellular adhesion molecule 1 (ICAM1), P-selectin glycoprotein ligand 1 (PSGL1), and signaling lymphocyte adhesion molecule family 7 (SLAMF7). For example, VLA4 expressed by MM cells allows adhesion to fibronectin but also to MSCs through vascular cell adhesion molecule 1 (VCAM1). The interaction between the co-stimulatory molecule CD40 expressed by MM cells and its ligand CD40L on MSCs induces an upregulation of adhesion molecules. Adhesion of MM cells to MSCs induces an upregulation of cytokine production by both cell types, promoting paracrine and autocrine loops. Of note, adhesion-mediated activation of the NF-κB pathway upregulates adhesion molecules on both MSCs and MM cells, which is a good illustration of the vicious cycle referred to previously. Such mechanisms result in so-called cell-adhesion mediated drug resistance (CAMDR) [22–24]. Some of the adhesion molecules, such as syndecan-1, are also able to sequester growth factors in their site of secretion and thus favor autocrine and paracrine bioactivity.
The BMMe cellular compartment produces a large variety of cytokines secreted in the liquid milieu [25,26]. These cytokines activate major signaling pathways involved in growth, survival, anti-apoptosis, drug resistance, and migration of MM cells: the phosphoinositide 3-kinase (PI3K)/Akt/mammalian target of rapamycin (mTOR)/p70S6K cascade, the IkappaB kinase alpha (IKK-α)/NFκB pathway, Ras/Raf/mitogen-activated protein kinase (MAPK), and Janus-activated kinase (JAK)/signal transducer and activator of transcription 3 (STAT3). Factors of interest include: interleukin 6 (IL-6), insulin-like growth factor 1 (IGF1), vascular endothelial growth factor (VEGF), B-cell activating factor (BAFF), a proliferation-inducing ligand (APRIL), stromal cell-derived factor 1α (SDF1α), hepatocyte growth factor (HGF), interferon α (INFα), tumor necrosis factor α (TNFα), fibroblast growth factor 2 (FGF2), transforming growth factor β (TGFβ), macrophage inflammatory protein 1α (MIP1α), interleukin 1β (IL-1β), interleukin 3 (IL-3), interleukin 7(IL-7), and growth differentiation factor 15 (GDF15).
Figure 1.1 Interaction of multiple myeloma cells in their bone marrow milieu
Adhesion of multiple myeloma cells to BMSCs triggers cytokine-mediated tumor cell growth, survival, drug resistance, and migration. MM cell binding to BMSCs upregulates cytokine secretion from both BMSCs and tumor cells. These cytokines activate major signaling pathways: ERK; JAK2–STAT 3; PI3K–Akt; and/or NFκB. Their downstream targets include: cytokines, such as IL-6, IGF1, and VEGF; anti-apoptotic proteins, such as BCL-XL, IAPs, MCL1; and cell-cycle modulators (cyclin D). Adhesion-mediated activation of NFκB upregulates adhesion molecules such as ICAM1 and VCAM1 on both MM cells and BMSCs, thereby further increasing the binding of MM cells to BMSCs (the blue boxes in the BMSC nucleus represent NFκB binding sequences in the promoter region of a target gene). Secretion of angiogenic factors, such as VEGF, bFGF, and HGF, from MM cells and BMSCs stimulates neo-angiogenesis. RANKL produced by BMSCs, and MIP1α produced by MM cells, stimulate osteoclastogenesis. By contrast, OPG secreted from osteoblasts and BMSCs inhibits osteoclastogenesis. Osteoblastogenesis is inhibited by MM cells through the secretion of IL-3 and DKK1 from MM cells and HGF from BMSCs. Stimulation of osteoclastogenesis and inhibition of osteoblastogenesis promote osteolysis. APRIL, a proliferation-inducing ligand; BAD, BCL-XL associated death promoter; BAFF, B-cell activating factor; bFGF, basic fibroblast growth factor; BMECs, bone marrow endothelial cells; BMSCs, bone marrow stromal cells; CAMDR, cell adhesion-mediated drug resistance; DKK1, Dickkopf 1; ERK, extracellular signal-regulated kinase; FKHR, forkhead in rhabdomyosarcoma; GSK3β, glycogen synthase kinase 3β; HGF, hepatocyte growth factor; IAPs, inhibitor of apoptosis proteins; ICAM1, intercellular adhesion molecule 1; IL-3/6, interleukin 3/6; IGF1, insulin-like growth factor 1; JAK2, Janus kinase 2; MAPK, mitogen-activated protein kinase; MCL1, myeloid cell leukemia sequence 1; MEK, MAPK/ERK kinase; MIP1 α, macrophage inflammatory protein-1α; MM, multiple myeloma; mTOR, mammalian target of rapamycin; NFκB, nuclear factorκB; OPG, osteoprotegerin; PIK3, phosphatidylinositol 3-kinase; RANKL, receptor activator of NFκB ligand; SDF-1α, stromal cell-derived factor 1α; STAT3, signal transducer and activator of transcription 3; TGFβ, transforming growth factor β; TNFα, tumor necrosis factor α; VCAM1, vascular cell adhesion molecule 1; VEGF, vascular endothelial growth factor. Reproduced with permission from © Nature Publishing Group, 2007. All rights reserved. Hideshim et al .
Several groups have reported that IGF1 (produced by MSCs and osteoblasts) is the major growth factor for MM cells, probably because the bioactivity of some others, such as IL-6 and HGF, is dependent on the activation of IGF1R by IGF1 . This leads to activation of MEK/MAPK and PI3K/Akt signaling pathways and consequently to activation of anti-apoptotic proteins Bcl-XL and Bcl-2 whereas the proapoptotic protein Bim is downregulated. IL-6 is also a key factor in MM . In contrast to IGF1, IL-6 has a critical role in the development of normal plasma cells, and can be produced in an autocrine manner by MM cells. IL-6 secretion from MSCs is upregulated by many cytokines, including IL-1β (also aberrantly produced by MM cells), VEGF, and TNFα. After binding to its receptor, IL-6 triggers activation of MEK/MAPK, JAK/STAT3, and to a lower extent PI3/Akt signaling pathways. This leads notably to the activation of anti-apoptotic proteins Mcl-1, Bcl-XL, and c-Myc.
The cytokine network complexity is due to the cross-talk between MM and BMMe cells and the inherent redundancy of cytokine signaling. For example, the levels of VEGF and IL-6 secretion measured in the supernatants of MM cells and MSC co-cultures, are higher than the sum of the levels of the same cytokines from individual MM and MSC cultures .
Separate from the effects on the tumor itself, the dialogue between BMMe and MM cells promotes two essential features of MM: angiogenesis and osteolysis, which both occur during the progression of the disease.
The adhesion of MM cells to bone marrow endothelial cells upregulates the production of cytokines with angiogenic activity by both cell types (such as VEGF, FGF2, HGF, transforming growth factor-β, IL-8, and platelet-derived endothelial growth factor), which in turn promotes tumor growth by increasing delivery of nutrients and oxygen . Consequently, the level of bone marrow angiogenesis is increased in patients with MM and has been described as an adverse prognostic factor .
In physiologic conditions, osteoclastogenesis is regulated by MSCs and osteoblasts, which produce the receptor activator of NFκB ligand (RANKL) and its decoy receptor osteoprotegerin (OPG); the binding of RANKL to RANK on osteoclasts stimulates their differentiation and activity, whereas this is prevented by the binding of OPG. In MM, the OPG/RANKL ratio is unbalanced by adhesion of MM cells to MSCs, contributing to an excessive osteoclastogenesis . MSCs and MM cells produce many other osteoclastic activating factors such as IL-6, MIP1α, IL-1β, TNFα, VEGF, and HGF. In turn, osteoclasts themselves are important producers of IL-6. Decreased osteoblastogenesis also contributes to osteolytic lesions. MM cells induce an inhibition of MSC differentiation into functional osteoblasts either directly by adhesion (VLA-4 on MM cells to VCAM1 on osteoblastic progenitors) or indirectly by producing soluble inhibitors including DKK1, IL-3, IL-7, and Frizzle related-protein 2 (FRP2). The common mechanism is a downregulation of runt-related transcription factor 2 (RUNX2) .
Compelling evidence revealed that MSCs from patients with MM are abnormal and contribute to disease progression . Importantly, these abnormalities have been described as outside of MM cells influence and persist after several weeks of culture, suggesting MSCs are, or have become, abnormal ‘by themselves’. MSCs from patients show differences in gene and protein expression profiles when compared with an age-matched healthy MSC donor, and functional abnormalities such as an altered ability to differentiate into osteoblastic lineage . MM MSC-derived exosomes have been shown to be key regulators of direct interactions with MM cells (Figure 1.2) .
The obvious causes of the extraordinary biological heterogeneity of MM are notably cytogenetic lesions and clonal heterogeneity of MM cells, but BMMe may also be considered. Clinical studies have led to a better understanding into the mechanism by which the abnormal BMMe affects the pathophysiology of myeloma. Some factors specific to BMMe have been related to the prognosis of the patients and potentially to the response to treatment [36–38]. In addition, emerging data suggest that alterations of the BMMe may not only be supportive of tumor growth but also required for tumorigenesis induction . Finally, as in the tumor, BMMe is probably an entity that evolves during the course of the disease, maybe with different functions (anti/protumor) according to the stage.
Hence, BMMe has become a therapeutic target that cannot be ignored in MM. This has led to the development of drugs such as immunomodulatory drugs and proteasome inhibitors, which target not only myeloma cells but also the BMMe, and have clearly contributed to the clear improvement of patient survival. However, the dissection of the hierarchy into this tumoral BMMe and the identification of specific targets are urgently needed for the development of next generation therapies.
Figure 1.2 Phenotypic differences: MM-MSCs versus healthy MSCs.
Diagram of the phenotypic differences currently described between MM-MSCs and healthy, non-diseased MSCs. Compared with healthy MSCs, MM-MSCs have the following traits: increased expression of EphB4 receptor, ICAM, VCAM, IL-3, IL-6, IL-10, IL-1b, VEGF, SCF, TNF-a, TGF-b1, BAFF, HGF, RANKL, DKK1, GDF15, HoxB, MMP1, and MMP2; decreased expression of MMP3, TGF-b2, TGF-b3, FasL, and hyaluronan synthase 1 (Has1); increased production of fibronectin, osteopontin, and hyaluronan; and reduced immunosuppressive properties due to a loss in the ability to inhibit T cells. Downstream effects of MM-MSCs on myeloma cells include increased chemotherapeutic resistance, Bcl-2 signaling, NF-kB signaling, and cancer stem cell population concentrations. BAFF, B-cell activating
factor; DKK1, Dickkopf 1; FasL, Apo-1/CD95 ligand; GDF15, growth differentiation factor 15; Has1, hyaluronan synthase 1; HGF, hepatocyte growth factor; HoxB, homeobox protein B; ICAM, intercellular adhesion molecule; IL-1b/3/6/10, interleukin 1b/3/6/10; MM-MSCs, multiple myeloma mesenchymal stem cells; MMP1/2/3, matrix metalloproteinase 1/2/3; NFkB, nuclear factorκB; RANKL, receptor activator of NFκB ligand; SCF, stem cell factor; TGFβ1/2/3, transforming growth factor β1/2/3; TNFα, tumor necrosis factor α; VCAM, vascular cell adhesion molecule; VEGF, vascular endothelial growth factor. Reproduced with permission from © American Association for Cancer Research, 2012. All rights reserved. Reagan and Ghobrial .