Ally to adversary: mesenchymal stem cells and their transformation in leukaemia

The dynamic nature of MSC accounts for their role in diverse biological processes such as immunomodulation, tissue regeneration and repair, anti-fibrosis, angiogenesis and survival and maintenance of other cell populations in the niche [8]. One of the most noteworthy properties of MSC is their role in immune suppression and modulation which has been reported to aid in tumour development [8]. MSC inhibit: responses to alloreactive T lymphocytes [9], proliferation and cytotoxicity of natural killer cells [10], generation and maturation of dendritic cells and proliferation and maturation of B cells [11]. They have minimal expression of MHC class II molecules and lack the expression of costimulatory molecules CD40, CD40 ligand, CD80 and CD86 which prevents them from mounting an immune response to antigen presenting cells [12]. IFN-ɣ stimulation induces the expression of MHC II but not of costimulatory molecules [12]. MSC are also known to preserve neutrophil viability and function, induce macrophage M1/M2 phenotype transformation, promote generation and induction of CD4+CD25+ or CD8+ T regs both in vivo and in vitro [13]. M1 macrophages crosstalk with MSC via CD54 molecule in the form of ‘unconventional synapse’ leading to an increase in anti-inflammatory and immunomodulatory properties of MSC [14]. Their unique ability of immunosuppression also allows their use in the treatment of Graft vs Host Disease (GVDH) in bone marrow transplantations [15]. IFN-ɣ pre-treated MSC significantly suppress GVDH upon administration during bone marrow transplant [16]. Although the exact mechanism underlying the immunosuppressive effects of MSC is still ambiguous, most studies indicate that soluble factors like prostaglandin E2 (PGE2) [17, 18], indoleamine 2,3-dioxygenase (IDO) [19], hepatocyte growth factor (HGF) [20] and transforming growth factor (TGF)-β 1 are involved [21]. Bone marrow MSC mediate contact dependent suppression of proliferation and survival of T cells while simultaneously increasing the proportion of Tregs [22].

The plasticity of MSC to give rise to multiple lineages has implicated their role in tissue regeneration and repair. Single cell RNA sequencing of bone marrow MSC indicated simultaneous expression of markers for multiple lineages however, despite lineage priming MSC may change fate due to an environmental cue [23]. Single cell transcriptome also identified three distinct subpopulations of bone marrow MSC using FGFR2, PLAT, VCAM1, FGF2 and FGF5 [24]. It has been shown that addition of MSC to an active immune environment decreases the secretion of pro-inflammatory cytokines TNF-ɑ and IFN-γ, while concurrently increasing the production of anti-inflammatory cytokines IL-10 and IL-4 [25]. MSC paracrine signalling reduces inflammation, promotes angiogenesis and induces cell migration and proliferation by secreting various chemokines, cytokines and growth factors such as VEGF, PDGF, bFGF, EGF, FGF and TGF-β [26, 27]. These factors act as mediators for proliferation, migration and gene expression changes of epithelial/endothelial cells, fibroblasts and macrophages which further aid in the repair process [28].

One of the most intriguing roles of MSC is regulation of hematopoiesis, which they coordinate by secreting a plethora of soluble factors with diverse functions such as chemoattraction, migration, homing, induction of signalling, proliferation and maintenance. The maintenance of hematopoietic stem cells in their native state requires support from the underlying stromal cells through extracellular matrix molecules and soluble factors. A complex interplay of signalling networks between MSC and HSC is needed for maintenance of niche homeostasis. It is also important to note that the MSC regulate not just the normal hematopoiesis but also contribute immensely to the tumour growth by primarily regulating the microenvironment [2].

MSC can be transformed by acquiring genetic modifications which eventually manifests on the growth and maintenance of HSC. The origin and basis of these modifications are yet to be completely understood. The initial evidence on the role of genetically transformed stromal cells in the formation of myeloproliferative disease came from Rupec et al. who established that deletion of IkBɑ from the stromal cells of mice leads to MDS [51]. Loss of Retinoic acid receptor gamma (RARγ) gene from microenvironment exhibited an MPN-like phenotype wherein there was a reversion of the phenotype upon transplantation of bone marrow or spleen cells from RARγ⁻/⁻ mice to wild type mice, but not in RARγ⁻/⁻mice [52]. This exemplifies the role that a defective microenvironment may play in initiating neoplasms. In another study in mice, a link between deletion of Rb (Retinoblastoma) gene in both HSC as well as stromal cells established a myeloproliferative like disorder resulting in subsequent mobilization and loss of HSC. Interestingly, Rb deletion either in stromal cells or HSC alone did not manifest the disease, yet transplantation of Rb deleted HSC in Rb deficient mice lead to an induction of disease [53]. Similarly, Mib-1 (Mind bomb) deletion in MSC resulted in activation of Notch signalling in microenvironment leading to a myeloproliferative disorder. On transplantation into wild type mice, these cells did not show a similar phenotype which seems to be restricted only to Mib-1 null microenvironment [54]. Raaijmakers et al. suggested that alterations in progenitors of stromal cells can also result in hematological malignancies. They observed that deletion in Dicer1 along with reduced levels of sbds protein in mesenchymal osteo-progenitors (but not in mature osteoblasts) resulted in MDS and AML like phenotype [55]. Deletion of sbds gene in mesenchymal progenitor cells induced mitochondrial dysfunction, oxidative and genomic stress in HSPC and the damaged associated molecular pattern (DAMP) molecules s100a8 and s100a9, secreted by MSC mediated the effect [55]. Deletion of Sipa1 in mouse bone marrow stromal cells leads to the development of MPN/MDS phenotype. MSC and progenitors in bone marrow microenvironment are physically and functionally altered due to an enhanced inflammatory cytokine profile (TGF-β, TNF-α) and dysregulation of Dicer1, KITL, ANGPT1, Il-7, CXCL-12 and Thpo. Sipa1 gene is observed to be downregulated in leukaemia’s such as CML, CNL and CMML [56]. MSC from CML patients also cause immune suppression by reduced T-lymphocyte proliferation via granulocyte myeloid derived suppressor cell (G-MDSC). Up regulated immunomodulatory factors secreted by CML-MSC as compared to normal MSC include Cox2, TGF-β, Arg1 and IL-6, which modulate the function of the HSC [57]. Genetic, functional and immunologic characterization of bone marrow MSC from AML patients vs healthy donor bone marrow MSC showed an increase in anti-inflammatory and immunosuppressive mechanisms with increased IL-10 linked to disease outcome [58]. An altered microenvironment, due to aberrant signalling or production of pro-inflammatory cytokines can aid in the initiation and maintenance of hematological malignancies and favour the growth of leukemic cells (Fig. 1).

An upregulation in adhesion or homing molecules on both LSC and MSC can lead to prolonged survival and maintenance of these cells in vivo (Fig. 1). CXCL12 and CXCR4 serve as prognostic markers in various leukaemia’s with the expression of CXCR4 being highest in promyelocytic leukaemia, myelomonocytic AML and B-lineage ALL [59]. An upregulation of VCAM-1 on MSC and its ligand VLA-4 on leukemic cells had also been reported to be involved in the maintenance of residual disease in AML and resistance to therapy via activation of NF-κB [60‐62]. Knockdown of ITGA4 gene in CML murine cells led to decreased adhesion to VCAM-1, loss of CFU forming ability and cell survival [63]. Inhibition of ITGA4 using natalizumab sensitized primary ALL cells to therapy [64]. Zhang et al. observed activation of Wnt signalling pathway in CD34+ leukemic cells by N-cadherin from bone marrow MSC in CML. Human CML stem or progenitor cells co-cultured with bone marrow derived MSC showed an increased association of N-cadherin and cytoplasmic catenin. The concentration was found to be highest in imatinib treated cells thus speculating their role in chemoresistance in CML [65].

There is a bidirectional signalling in niche wherein LSC modulate the stromal cells and vice versa; ultimately transforming the niche into a self-reinforcing acclamatory leukemic niche (Fig. 1). LT-HSC (long term HSC) isolated from transgenic BCR-ABL⁺ mice and CML patients exhibited decreased bone marrow CXCL12 which debilitated their homing and engraftment in CML bone marrow. G-CSF secreted by leukemic cells down-regulated the levels of CXCL12 in CML bone marrow. This led to an anomalous proliferation and differentiation of LT-HSC subsequently conferring a selective advantage for their survival in the bone marrow [66]. High placental growth factor (PlGF) level in bone marrow plasma and peripheral blood correlated with BCR-ABL load in both CML mice and patients. CML cells induce bone marrow stromal cells to overproduce PlGF, which in return stimulates proliferation, migration, metabolism of CML cells along with bone marrow angiogenesis thus setting up a favourable vicious cycle. Since PlGF acts independent of BCR-ABL pathway it’s targeting overcame imatinib resistance and prolonged survival in imatinib resistant mice [67]. In transgenic mice leukemic cells remodelled the endosteal niche into a self-reinforcing stem cell niche. The leukemic myeloid cells were shown to interact with MSC and elevated expression of thrombopoietin (TPO) and CCL3 by these cells contributed to MSC expansion which in turn modulated osteoblasts in endosteal niche. Upregulated TGF-β and inflammatory signalling with elevated levels of CXCL12, SCF, Angiopoietin and Slit2 led to a complete transformation of the endosteal niche [68]. BMP1, PDPN, Nanog, MITF, FOXO3 and MET were upregulated in CML MSC as compared to healthy controls. Most of these except for FOXO3 were upregulated in MSC isolated from molecular response patients, suggesting their role in remodelling [69].

Communication between the LSC and bone marrow microenvironment is a hallmark of malignant transformation which is either guided by cell–cell contact or by the extracellular secretory molecules which could include circulating nucleic acids, protein, metabolites or lipids. There are several mechanisms which have been explored to understand how the LSC may regulate the MSC or vice versa (Fig. 1). The role of important signalling pathways, secretory factors, exosomes and miRNA have been explored by several groups to understand this relationship.

Dysregulated signalling is a premonition of cancer progression. Several critical pathways such as Notch, Wnt, Shh, BMP and TGF-β are involved in the crosstalk between HSC/LSC and MSC. Notch along with its ligand Jagged and Delta and regulator Hes1 guides differentiation, proliferation and cell fate decisions [70]. Human AML related MSC express higher levels of Notch1, Hes1 and Jagged1 as compared to normal MSC leading to increased Notch signalling in AML cells and conferring chemoresistance [71]. Abrogation of signalling by blocking Notch receptors along with chemotherapeutic agents lowered the supportive effect of MSC towards AML cells [71].

Bone marrow MSC have been observed to deliver functional mitochondria to AML leukaemia initiating CD34+ cells in co-culture system providing an edge towards chemotherapeutic resistance and survival [72].

The other important contributors are exosomes which are released by cells of both hematopoietic and non-hematopoietic origin. They accomplish bi-directional transfer of mRNA, miRNA and several proteins between cells. The role of exosomes in the context of hematological malignancies is well explored in last few years. AML LSC-derived exosomes were enriched with Flt3-ITD, NPM1, CXCR4, MMP-9 and IGF-1R mRNA within the microenvironment [73]. In CML, exosomes stimulate stromal cells to produce IL-8 which control many downstream signalling pathways under CXCR1 and CXCR2 receptor [74]. CLL derived exosomes delivered contents to endothelial cells and MSC which induced CXCL16 and CCL2 expression [75]. Patient derived CML exosomes activate epidermal growth factor receptor (EGFR) signalling in stromal cells and further increases the expression of SNAI and its targets, MMP9 and IL8 [74]. AML derived exosomes induce expression of Dkk1 while downregulating CXCL12, KITL and IGF1 in bone marrow stromal cells ablating normal hematopoiesis [76]. Microvesicles derived from leukemic cell lines have been reported to carry the fusion transcript of the parental cell line; which have been uptaken by normal MSC upon co-culture increasing their proliferation in vitro [77]. Extracellular vesicles released by murine MSC lead to loss of HSC quiescence and expansion of murine myeloid progenitors via Toll like receptor (TLR4) activation of MyD88 adaptor protein [78].

Interestingly MSC are possessed with both immunomodulating and immunosuppressive properties. The induction of either one of these depends on a complex balance of cues collected from many stimuli including the TLR. TLRs represent an arm of the innate immunity which is involved in recognizing Pattern and Damage Associated Molecular Patterns (PAMPs and DAMPs); thereby regulating processes of inflammation, anti-apoptosis and wound healing [79]. MSC express TLR1,2,3,4,5,6,7 and 9. TLRs triggered by stimuli can affect biological processes of MSC [79]. TLR3 and TLR9 deficiency were shown to result in acute lymphoblastic T cell leukaemia (post induction with endogenous retrovirus) in mice [80, 81].