Exosomes are small (30–150 nm) membranous vesicles of endocytic origin produced by all cells under physiological and pathological conditions. They have recently emerged as vehicles for intercellular transfer of molecular and genetic contents from parent to recipient cells. Exosome-mediated transfer of proteins or genes (RNA, miRNA, DNA) results in reprogramming of recipient cell functions. Exosomes carry and deliver information that is essential for health, and they participate in pathological events, including malignant transformation. Within the hematopoietic system, exosomes maintain crosstalk between cells located in the bone marrow compartment and at distant tissue sites. In hematological malignancies, tumor-derived exosomes (TEX) reprogram the bone marrow environment, suppress anti-leukemia immunity, mediate drug resistance and interfere with immunotherapies. TEX are also viewed as promising biomarkers of malignant progression and as potential therapeutic targets. The involvement of TEX in nearly all aspects of malignant transformation has generated much interest in their biology, mechanisms responsible for information transfer and the role they play in cancer escape from the host immune system.
It has been well established that cells in multicellular organisms have the ability to communicate with one another via direct contacts established between cells or via cytokine/chemokine-mediated networks.1 Recently, a new system of intercellular communication has been recognized that is evolutionarily conserved and universally utilized by single cell as well as multicellular organisms.2 This system involves information transfer between cells by means of extracellular vesicles (EVs).3, 4 EVs mediate short- or long-distance delivery of cellular contents and provide a protective mechanism via the vesicular membrane that prevents in-route destruction of sensitive messages. EVs deliver the messages upon their direct contact with recipient cells. This communication system operates in all body parts, including the neural, endocrine, hematopoietic and immune systems.
EVs were first recognized in the late 1960s, when electron microscopy (EM) showed that 20–50 nm vesicles carrying a clotting factor, later identified as the tissue factor (Tf), were present in human platelet-free plasma.5 In the early 1980s, Johnstone and colleagues6 reported that maturing reticulocytes released 50 nm microvesicles (MVs) following fusion of intracellular storage vesicles and multivesicular bodies (MVB) with the plasma membrane. The presence of MVs was also observed in tumor cell cultures, and they were described as ‘exfoliations’ of the cell membrane.7 At the time, MVs were considered to be ‘cellular waste,’ used to eliminate unnecessary cell contents. Today, MVs re-named ‘EVs’ are viewed as vehicles for horizontal transfer of molecular signals as well as genes between cells.3
Mammalian cells spontaneously release EVs of various sizes (30–5000 nm) into all body fluids.8 Unfortunately, no consensus exists as to the EV classification. The current EV nomenclature is partly based on size. It recognizes small EVs (sEVs, exosomes, 30–150 nm); intermediate-sized EVs, also called MVs (200–1000 nm); and the largest apoptotic bodies (1000–5000 nm). Exosomes differ from the larger EVs not only by size, but also by their unique origin and cargo. Exosomes are formed within MVBs in the endosomal compartment of parent cells and carry endosome-associated proteins such as TSG101 or ALIX.9, 10 In contrast to exosomes, which are released when MVBs fuse with the cell membrane, larger EVs are shed as ‘blebs,’ exfoliating directly from the plasma membrane. The exosome content in part reflects that of the parent cell surface, but exosomes also carry components of the cytosol as well as various RNA species and DNA.11, 12, 13 For this reason, exosomes have been of special interest among various EVs, and their characteristics and biological activities are being intensely investigated.
Emerging evidence suggests that EVs play a key role in normal physiology, including tissue differentiation and repair, hematopoietic stem cell development, coagulation, pregnancy or immune surveillance.14, 15, 16 Exosomes also participate in various pathological processes, including tumor growth and metastasis.17 The roster of human diseases in which exosomes mediate pathology extends to HIV-1,18 autoimmunity,19 Parkinson’s disease,20 Alzheimer’s disease,20 numerous inflammatory conditions21 and others. Although it is still unclear that all EVs mediate these pathologies, attention has focused on exosomes as potential non-invasive biomarkers and perhaps also as future therapeutic targets. However, the lack of specific markers that could reliably distinguish exosomes from larger EVs is a significant barrier to the development of exosomes as future disease biomarkers.
Leukemic blasts, like other tumors, produce and utilize exosomes to promote their own progression. Tumor- or blast-derived exosomes, which we denote as tumor-derived exosomes (TEX), are implicated in reprogramming of the bone marrow (BM) microenvironment,22 promoting drug resistance,23 inhibiting anti-leukemia immunity potentially contributing to leukemic relapse.24 TEX carry oncogenic proteins promoting tumor growth.25 Many aspects of the exosome biology are yet to be discovered. Nevertheless, considerable insights into their properties and interactions with diverse cells, including cells of the hematopoietic system, exist and are reviewed here. Overwhelmingly, emerging evidence supports the role of TEX in the development, progression and therapy of human hematological malignancies.
The biogenesis of exosomes begins with endocytosis of the plasma membrane and the formation of an early endosome. Next, inward budding of the endosomal membrane generates intraluminal vesicles coated with clathrin and other plasma membrane components and enclosing cytosolic components.26 Late endosomes or MVBs contain multiple intraluminal vesicles. When MVBs fuse with the plasma membrane, these vesicles are released into extracellular space as virus-size exosomes with topology resembling the parental cell membrane. The vesicular lumen contains nucleic acids, nucleoproteins, enzymes, soluble factors and a variety of molecules derived from the parental cell cytosol.26 Released exosomes travel to recipient cells situated locally or at a distance from the parent cell, delivering molecular and/or genetic information and inducing phenotypic and/or functional changes in the recipient cells.27 It is suspected that exosome secretion is a regulated process calibrated to deliver information to the specific cellular address.
Exosome formation in late endosomes is accomplished via the Endosomal Sorting Complex Required for Transport, which requires participation of various accessory proteins such as ALIX and tumor-susceptibility gene 101 (TSG101) as recently reviewed.10, 24 Exosomes can also utilize the Endosomal Sorting Complex Required for Transport-independent pathway which is less well understood, but it is known to involve lipids, such as sphingosine-1-phosphate and ceramide, tetraspanin-enriched microdomains and sphingomyelinase.26, 28 The exosome biogenesis pathway and its various molecular components are not yet firmly established. Leukemic cells apparently utilize the same biogenesis pathway for exosome production and release as do all other cells.
Isolation of exosomes from body fluids
Studies of TEX have been largely confined to vesicles isolated from supernatants of cultured tumor cell lines, where tumor cells are the only source of exosomes. In contrast, body fluids of cancer patients contain TEX mixed with non-tumor-derived exosomes; thus, the ratios of TEX/non-TEX might vary broadly. Therefore, to study TEX in body fluids, it is first necessary to isolate the total exosome fraction from plasma and then separate TEX from other exosomes. While seemingly simple and rational, strategies for TEX isolation from body fluids have been difficult to implement. TEX isolation is a two-step process. First, isolation of total exosome fractions from plasma is performed using one of several methodologies available today, including microfluidics, precipitation or affinity capture as recently reviewed.29 Numerous commercially developed exosome isolation procedures are available. Many of these methods are designed for the isolation of nucleic acids, especially microRNAs, from fractionated exosomes without consideration for other cargo components. Many of these methods might not distinguish exosomes from larger EVs30 or might compromise exosome integrity and functions.31 The author’s laboratory isolates exosomes from body fluids or cell supernatants by mini-size exclusion chromatography as recently reported.32 Initial differential centrifugations and the ultrafiltration steps are used to pre-clear body fluids, and mini-size exclusion chromatography facilitates the removal of non-relevant ‘contaminating’ plasma proteins or nucleoprotein aggregates. The method rapidly processes small plasma volumes (0.5–1.0 ml), minimizes exosome losses and ensures recovery of morphologically intact, functionally active exosomes that can be used for biochemical, genetic and immunological characterization, EM or flow cytometry studies.32 Exosome concentrations and size are determined using special instruments designed to measure size and concentrations of nanoparticles in suspensions. Because exosomes are virus size, special methods are required for their imaging, including scanning or transmission EM, atomic force microscopy or fluorescent imaging. Isolation of TEX from total exosomes requires methods for a selective capture of TEX and their quantitative recovery. This approach is based on the principle that TEX carry membrane-embedded molecules that are either uniquely expressed or highly enriched in parental tumor cells. We used this approach for capture of CD34+ TEX from plasma of patients with CD34+ acute myeloid leukemia (AML).24 More recently, Melo et al.33 used immunocapture to isolate glypican 1+ exosomes from plasma of patients with pancreatic cancer. Others have used immunocapture to isolate exosomes carrying prostate-specific membrane antigen from the blood of patients with prostate cancer.34 Methods for TEX immunocapture with Abs specific for leukemia-associated antigens (LAAs) such as CD123 or CCL-1(ref. 35) using exosomes isolated by mini-size exclusion chromatography from plasma of AML patients are in development. The two-step method enables the recovery of TEX in parallel with non-TEX exosomes in order to compare their cargos and effects mediated in recipient cells.
Morphological and molecular features of TEX
When examined by EM, TEX look no different than other exosomes. They are membrane-bound vesicles heterogeneous in size (often <50 nm in diameter). Sections of Epon-embedded TEX examined by EM confirm that a bona fide membrane surrounds each vesicle.36 Immuno-EM has been used to illustrate the presence of FasL,36 glypican-1(ref. 33) or other biologically significant molecules37 embedded in the TEX membrane.
Information about the TEX molecular content largely comes from studies of TEX produced by tumor cell lines. The molecular content of TEX membrane is partly derived from the parent tumor cell surface and partly from the endosome.38 The MVB-related proteins, ALIX or TSG101, and tetraspanins, CD9, CD63, CD81, CD82, have been widely used as exosome markers.26 Figure 1 shows that the TEX molecular content includes a broad variety of proteins, lipids and glycans as well as LAAs. TEX produced by different tumor cell lines carry distinct molecular signatures.32, 39 The intra-vesicular content of TEX, rich in proteins and RNA species, only partly resembles that of the parent cell cytosol, suggesting that loading of the exosome cargo may not be a random process but might involve selective packaging by mechanisms not yet clearly understood.
To interrogate TEX molecular signatures, western blots, immune arrays and mass spectrometry can be used. Antibody-based techniques allow for the amplification of specific signals and have been most useful for analysis of plasma-derived exosomes. The amplification is necessary, because these exosomes are coated with an excess of plasma proteins camouflaging the true exosome components. For this reason, mass spectrometry of plasma-derived exosomes is challenging, although it has been successfully used to chart contents of TEX isolated from supernatants of tumor cell lines40 or from urine.41 Proteins and RNA species that have been identified in TEX so far are deposited in the ExoCarta database accessible on line.42 The proteins most commonly identified in exosomes are CD63, heat shock proteins, actin, tubulin and components of such cellular signaling pathways as β-catenin, WNT and/or Notch.43, 44 TEX cargo contains several RNAs such as mRNA, miRNA and ncRNA as well as gDNA.45, 46 The molecular and genetic contents of TEX are being intensively interrogated because they are perceived as potentially useful future cancer biomarkers.33, 47
Functions mediated by TEX
Exosomes modulate a variety of physiological and pathological activities. They do so by the delivery of information to neighboring or distantly located recipient cells and reprogramming their functions. TEX also transfer information from one tumor cell to another and can deliver autocrine signals.48 TEX are well equipped for the task of information transfer: their surface is decorated with signaling and adhesion molecules derived from the parent cell surface membrane. Tumor cells, for example, breast cancer cells, which are normally under oxidative stress, secrete excessive numbers of TEX.49 TEX distribute freely throughout all body fluids, establishing a communication network between the tumor and host cells. Upon their isolation from plasma of cancer patients, TEX exert biological activities in vitro and in vivo. Importantly, TEX deliver to recipient cells the membrane-protected content in the form specified by parent cells by one of several mechanisms as indicated in Table 1. TEX are internalized by recipient cells via fusion, phagocytosis or endocytosis. They are either directed to the lysosomes for degradation/clearance or are incorporated into the cellular machinery to initiate recipient cell reprogramming.27
Since TEX are products of tumor cells, the communication network they establish and reprogramming they initiate are entirely tumor-driven. TEX-induced changes are designed to promote tumor progression and metastasis.50 Utilizing autocrine and juxtacrine signaling, TEX facilitate tumor progression by delivery of depleted growth receptors, ectoenzymes or factors necessary for sustaining tumor growth.50 TEX also modify functions of stromal cells to favor tumor growth,51 promote angiogenesis52 and mediate immune suppression targeting anti-tumor immune cells.53 TEX play a critical role in oncogenic transformation of BM progenitor cells.22 TEX have also been shown to interfere with functions of mature hematopoietic cells in the TME.4 Numerous signaling molecules present in the TEX cargo have been implicated in inducing alterations of the responder cell phenotype to the malignant or metastatic phenotype.54
TEX carry and deliver immunosuppressive molecules to innate and adaptive anti-tumor immune cells, inhibiting their functions.55 Interestingly, T lymphocytes do not readily internalize TEX. Instead, TEX interacting with surface molecules on T cells deliver signals, which initiate a Ca2+ flux and activate downstream signaling, resulting in alterations of the recipient cell transcriptome and re-programing of T-cell functions.56 TEX also carry factors promoting differentiation and proliferation of regulatory T cells, myeloid-derived suppressor cells and regulatory B cells.53 Importantly, TEX carry and deliver TAAs to dendritic cells, drive antigen processing and presentation to T cells and thus promote anti-tumor immunity.57 The dual role of TEX in modulating anti-tumor immunity is not yet understood. It is, however, clear that interactions of TEX with the host immune system have a profound impact on cancer development, progression and metastasis.
TEX in hematological malignancies
The role of TEX in patients with hematological malignancies is rapidly emerging as a determinant of outcome. Recent data suggest that blast-derived oncosomes influence every aspect of the development, differentiation, maturation, functional integrity and progression of hematological malignancies.
High levels of TEX in patients’ plasma
Plasma of patients with hematological malignancies is enriched in exosomes compared to plasma of normal controls (NCs). Newly diagnosed AML patients have significantly elevated plasma levels of blast-derived CD34+, CD117+, CD33+ exosomes.58 Exosome plasma levels decrease after induction chemotherapy, concomitant with blast reduction in the BM, but then increase during consolidation therapy, approaching levels measured at diagnosis. Further, high exosome levels persist in many, but not all patients in CR, at the time when leukemic blasts are undetectable in the BM by hematopathological methods.58 This suggests that exosome plasma levels might serve as a predictor of leukemic relapse. However, it is not yet known whether these exosomes derive from minimal residual leukemia. In those AML patients who achieve long-term remission following consolidation therapy, exosome plasma levels normalize to NC’s plasma levels, implying prognostic significance.58 In addition, levels of individual exosomal proteins, for example, TGF-β1 and its isoforms, especially, the mature, biologically active form of TGF-β1, were elevated in AML exosomes. Further, AML exosomes isolated before, during and after chemotherapy differed in levels of active TGF-β1.58 In aggregate, these correlative ex vivo data suggest that TEX plasma levels and the TEX molecular content could serve as potential indicators of AML response to therapy and disease outcome.
Caivano et al.59 isolated serum EVs and determined their numbers, size and phenotypic profiles in patients with chronic lymphocytic leukemia (CLL), non-Hodgkin’s lymphoma, Waldenstrom’s macroglobulinemia, Hodgkin’s lymphoma (HL), multiple myeloma (MM), AML, myeloproliferative neoplasms, myelodysplastic syndromes and NC.59 Compared to NC, EV levels were found to be significantly elevated in Waldenstrom’s macroglobulinemia, HL, MM, AML, CLL, non-Hodgkin’s lymphoma and some myeloproliferative neoplasm patients. EVs isolated from patients expressed TAAs, for example, CD19 in B-cell neoplasms, CD38 in MM, CD33 in myeloid tumors and CD30 in HL. The counts of total and antigen-specific EVs significantly correlated with clinical features defined by the established staging criteria in CLL, Waldenstrom’s macroglobulinemia, MM and HL. These data underscore the potential role of plasma EVs, and especially of TEX, as potentially useful biomarkers of the disease presence and progression.
TEX carry and deliver unique cargos to BM cells
Similar to TEX produced by solid tumors, leukemia-derived oncosomes carry a unique cargo of genes and molecules that are distinct from those present on exosomes produced by normal cells.32 TEX isolated from supernatants of AML cell lines were enriched in CD34, CD44, CD117 and CD200.58 In AML, plasma-derived exosomes carry myeloid-blast markers (CD34, CD117 and CD33) and LAA, such as CD123, CCL-1, CD-96 and CD44.58 Also, AML exosomes are enriched in immunosuppressive molecules, including membrane-associated TGF-β1, FasL, PD-1/PDL-1, MICA/MICB, CD39/CD73 and others.32, 60 These molecules are biologically active and able to mediate immune suppression.32, 60 In B-CLL, plasma-derived exosomes were highly enriched in CD19, an indication they originated from leukemic B cells.61 In MM, a proteomic content of cell lines was evaluated by LC-MS/MS and compared to that of whole-cell lysates.62 Among a number of common exosomal proteins, such as the MHC molecules, adhesion proteins, membrane transporters or cytoskeletal components, several proteins unique to vesicles derived from MM cells were identified. Also, an overlap in the protein content of the tumor cell lysates and TEX was observed, suggesting that the exosome molecular profile resembles that of parent cancer cells.
EVs in the circulation of patients with leukemia have been long known to carry the clotting factor.5 Recent data confirm that Tf+ TEX induce cancer-associated life-threatening hypercoagulation in patients with advanced malignancies.63 Cells undergoing malignant transformation increase the production of Tf-rich exosomes. Upon internalization by endothelilal cells, Tf+ TEX induce a rapid conversion of endothelilal cells to the pro-coagulant phenotype. The presence of high levels of pro-coagulant TEX in patients with cancer predicts poor outcome.63 It is suspected that dissemination and delivery of Tf by oncosomes contribute to cancer progression by promoting blood coagulation, which impedes access of immune effector cells or their products to sites of tumorigenesis.
The TEX lumen contains proteins originating in the cytosol of parent cells and nucleic acids (mRNA, miRNA and DNA). The ratio of mRNA/miRNA in the EVs cargo is variable, and the proportion of this cargo reflecting the parent cell transcriptome is unknown. In AML, exosomes produced by primary blasts or cell lines were enriched in coding and non-coding RNAs implicated in AML pathogenesis, including transcripts relevant to AML prognosis (NPM1, FLT3-ITD), response to therapy (CXCR4, IGF-IR) and the leukemic niche formation (IGF-IR, CXCR4, MMP9).64 Huan et al.64, 65 reported that the total RNA content of TEX varied considerably among AML cell lines, primary AML cells and primary CD34+ BM cells. An abundance of small RNAs (0–150 nt) and miRNAs (40–80 nt) was present in TEX, with 5- to 13-fold enrichment in miRNAs relative to parental cells. Also, different AML cell lines produced TEX carrying distinct miRNA species: for example, the release of the miR-17-92 cluster, especially miR-92a, was a distinct feature of K562 cells.66 The enrichment in various mRNA and miRNA species in leukemia-derived exosomes is under intense scrutiny. Transfer of these materials to recipient cells is directly responsible for the induction of a broad variety of pro-tumorigenic effects. The critical impact of exosomal miRNAs on tumorigenesis has been broadly discussed in the literature.67
TEX also contain >10-kb fragments of double-stranded genomic (g)DNA fragments. TEX subpopulations differ from each other in the DNA content and harbor different mutations in for example, KRAS and p53 genes. This suggests that DNA might be selectively packaged into vesicles during their production in parent cells,68 and that TEX could be used for assessments of gDNA mutations in the parent tumor. However, DNA origin, presence and content in exosomes is yet to be validated.
TEX reprogram the BM microenvironment
The hematopoietic stem cell (HSC) niche is composed of a complex mix of stromal cells and blood elements necessary for hematopoiesis. Leukemic cells reorganize the HSC niche to promote their own survival and growth using exosomes.15, 51, 69 The ‘proof of principle’ for the role of TEX in reprogramming of the bone marrow environment (BME) was provided by David Lyden and colleagues, who reported that transfer of TEX derived from murine metastatic melanoma to the BME reprogrammed it to a pro-metastatic niche, which now supported the development of highly invasive melanoma cells.22 Emerging evidence suggests that oncosomes migrating into the BME are equipped for interactions with stromal elements, blood vessels and immune cells.15 They carry a cargo of genetic and molecular factors that impede the maturation of normal HSC and interfere with functions of stromal, endothelial and immune cells (Figure 2). The mechanisms responsible for this reprogramming may include the oncogene transfer, transcriptional changes in cellular mRNAs and/or post-translational alterations in the protein repertoire and functions of recipient cells. TEX-mediated oncogenic transformation25, 68, 70 via horizontal gene transfer indicates that EVs can be dangerous by contributing to malignant progression.
Mechanisms TEX employ to reprogram the BME
In addition to transfer of oncogenes or oncogenic signals from cancer to normal cells, TEX transfer other genetic elements to the BME, as indicated in Table 2. The integration of these elements into the genome of recipient cells in the BM leads to reprogramming of cellular pathways, especially those responsible for growth factor production15, 69 and to the emergence of aggressive pro-leukemia phenotypes. Not all interactions between exosomes and recipient cells are mediated by transfer of genetic information. TEX also directly modify the proteome of recipient cells by inducing alterations in major signaling pathways and by mediating enzymatic modifications of proteins (Table 2). For example, exosomes produced by CML cells stimulate BM stromal cells to produce IL-8 that, in turn, is able to modulate leukemia phenotype in vitro and in vivo.51 Another potential mechanism of the BME reprogramming may be a switch from oxidative phosphorylation to aerobic glycolysis reported to be induced by leukemia-derived EVs.71 This metabolic alteration and hypoxia, known to be prominent in the leukemic BME, further promote exosome release from blasts72 and/or mesenchymal stromal cells (MSC), stabilize HIF-1α, up-regulate TGF-β1 expression and activate the CXCR4/SDF-1 axis in AML cells. The latter drives immune cell suppression and, concomitantly, increases expression levels of E-selectin, which regulates cell adhesion and migration out of the BME.73 In AML, these TEX-driven effects protect AML blasts and ensure their survival.72 Table 2 lists selected examples of changes in the proteome and metabolome of recipient cells known to be induced by leukemia TEX, and Table 3 provides a list of cells in the BME that may be targeted by leukemic TEX, leading to major functional alterations in these cells.
In aggregate, studies of oncosomes in various hematologic malignancies support their role in transfer of nucleic acids, proteins and soluble factors from the blasts to other cells in the BME. Importantly, emerging data suggest that leukemic TEX can simultaneously utilize different mechanisms (Table 2) to alter functions of various recipient cells, and that genetic, molecular or metabolic changes that are induced in these various cell types all favor leukemia progression. It also appears that the leukemia/stroma interactions are bi-directional: TEX reprogram the stroma, which in turn produces exosomes favoring proliferation of leukemic cells and suppressing normal hematopoiesis (Figure 2). Thus, MSC-derived exosomes carrying growth factors and IL-6 were shown to promote proliferation of MM cells in vitro and in vivo.74 TEX can also directly interfere with HSC differentiation and migration. In the BME, pericellular accumulations of TEX depend on their production rate by parent cells and the level of uptake by recipient cells. In this context, the TEX/non-TEX ratio in the BME might be a critical factor in balancing normal versus leukemic hematopoiesis.
While available evidence suggests that TEX have the ability to reprogram the BME and convert it into one supporting leukemia progression by several distinct mechanisms, it is not clear whether all blast-derived oncosomes can mediate all of the above described alterations in recipient cells (Tables 2, 3, 4). Possibly, a division of labor exists among TEX and distinct subsets of TEX specialize in mediating individual effects, depending on the program imparted by the parent leukemic cell. Most likely, the selection and regulation of mechanisms involved in the BME reprogramming rest with leukemic blasts and are mediated via TEX. An alternative possibility endows recipient cells with the capability to accept or reject TEX, depending upon conditions prevailing in the BME.
TEX interfere with anti-leukemia immunity
As discussed, TEX can exert suppressive or stimulatory effects on immune cells. However, the overall effect of TEX-immune cell interactions in the TME is immunosuppression (Table 4). The TEX cargo contains genes, molecules and factors, which can induce immune cell dysfunction by several distinct mechanisms, as recently reviewed.53 TEX can deliver tolerogenic signals, inhibit cell differentiation and/or anti-leukemia functions, induce apoptosis in activated CD8+ T cells or promote expansion of regulatory T cells and myeloid-derived suppressor cells.53 Thus, TEX-mediated effects can be direct (that is, delivery of immunoinhibitory signals to recipient cells) or indirect by promoting differentiation of regulatory immune cells.55 TEX have free access to lymph nodes, spleen, gut lymphoid tissues and other mucosal surfaces rich in immune cells. In vivo studies using mouse tumor models show that functions of immune cells in the periphery and in situ are altered following intravenous injections of TEX.17 Using fluorescently labeled TEX, it has been possible to visualize TEX redistribution to various lymphoid and non-lymphoid tissues.27 In subjects with cancer, TEX interacting with T or B lymphocytes, NK cells, monocytes or granulocytes create an immunoinhibitory milieu favoring escape of leukemia from the immune system. The immunoinhibitory effects of leukemic oncosomes have been extensively examined in AML. Patients with AML have a low NK-cell frequency and significantly depressed NK-cell activity in the peripheral blood with concomitant decreases in expression levels of activating natural cytotoxic receptors, NKp30, NKp44 and NKp46, and C-type lectin receptors, NKG2D and NKG2C.60 They also have an increased frequency of regulatory T cells mediating suppression in the peripheral circulation.75 TEX isolated at the time of AML diagnosis carry TGF-β1 and, upon co-incubation with normal human NK cells, downregulate NKG2D expression levels and NK-cell lytic activity. Antibody-mediated neutralization of TGF-β1 abrogated suppressive activity of AML exosomes, confirming that TEX-associated TGF-β1-induced NK-cell dysfunction. Similarly, co-incubation of exosomes isolated from plasma of AML patients in CR with normal NK cells downregulated NKG2D expression levels and NK-cell cytotoxicity. These data suggested that persistently elevated levels of biologically active exosomes in AML patients’ plasma impair anti-leukemia immune responses and thus contribute to leukemia relapse.58 TGF-β1 is but one of multiple suppressive factors delivered by TEX to immune cells. Our data linking the presence on TEX of biologically active CD39/CD73, Fas/FasL and PD-1/PDL-1, as discussed above, links immunoinhibitory activity of AML exosomes to at least four different suppressive pathways signaling in NK cells interacting with TEX.
TEX can also exert immunostimulatory effects
In addition to inducing immunoinhibitory signaling, leukemia oncosomes can also mediate anti-tumor immunity. The presence of LAAs, the MHC class I and II molecules and chaperones, such as HSP-70 and/or HSP-90, in the TEX cargo as well as the ready uptake and processing of TEX by DCs have motivated a number of investigators to use leukemic exosomes as components of anti-tumor vaccines.76 TEX isolated from supernatants of K562 cell line induced strong anti-leukemia immune responses in vitro and in vivo in mice.77 Shen et al.78 isolated HSP-70+ TEX from supernatants of a human acute promyelocytic leukemia cells (NB4) and co-incubated them with DCs. These DCs were more effective in inducing expansion of NB4-specific cytotoxic T lymphocytes than DCs not treated with TEX. In other studies, leukemic exosomes carrying molecules necessary for antigen processing and presentation were shown to elicit cytotoxic T lymphocyte responses in vitro and in vivo. These exosomes also induced Th1-type responses via activation of B cells.79 Thus, considerable evidence, largely from mouse leukemia models, supports the role of LAA-carrying TEX in activation of anti-leukemia responses and argues in favor of their utilization as vaccine adjuvants and components of therapeutic vaccines.
This dichotomy of TEX, equipped to mediate suppression as well as stimulation of immune cells, suggests that the microenvironment in which TEX operate provides the context for their activity. In the immunosuppressive context of the tumor, TEX exercise immune suppression. Removed from this context, TEX become immunostimulatory. The intercellular communication system in the TME utilizing TEX might be subverted to meet the need of the tumor; however, when the balance changes in favor of anti-tumor immunity, for example, as a result of immunotherapy, TEX might convert to facilitate anti-tumor immunity.
TEX mediate drug resistance
The acquisition of tumor cell resistance to chemotherapeutic drugs is a major barrier in therapy of hematological malignancies. Among various known mechanisms of drug resistance, TEX have attracted special attention, because of their ability to horizontally transfer drug resistance between tumor cells. TEX produced by adriamycin- and docetaxel-resistant breast cancer cell lines were shown to transfer resistance to sensitive tumor cells by means of intercellular transfer of specific miRNAs, including miR-100, miR-222, miR-30a and miR-17.80 Docetaxel resistance was also studied in other solid tumors, where it was conferred via transfer by TEX of multi-drug resistance protein-1 (MDR-1/P-gp), a P-glycoprotein transporter that is overexpressed in drug-resistant tumors.23 Cisplatin-resistant tumors produce TEX, which are enriched in other transporter proteins such as MDR-2, ATP-7A or ATP-7B.81 TEX may also contribute to chemoresistance by drug expulsion, a process that was initially described as ‘garbage shuttles’ and that consists of concentrating and packaging drugs into exosomes, followed by export of drug-laden exosomes from the resistant cell.81 Studies of apoptosis-resistant and -sensitive primary AML cells showed that TEX secreted by the former carried proteins that antagonized apoptosis, protecting leukemic blasts from drug-induced death.82 In MM, exosomes obtained from patients’ BMSCs induced tumor cell resistance to bortezomib by activating several survival pathways, including c-Jun N-terminal kinase, p38, p53 and Akt.83 Interestingly, in vivo studies in melanoma-bearing mice showed that resistance to cisplatin mediated by TEX could be reduced by the combined use of proton pump inhibitors and low pH.84 Understanding of mechanisms used by TEX to transfer drug resistance is necessary for the development of novel strategies to successfully treat cancer.
TEX interfere with immunotherapies
The ability of TEX induce apoptosis of activated CD8+ T cells in cancer has been previously reported.55 As the major goal of immunotherapies is to restore anti-tumor functions of effector CD8+ T cells, TEX carrying death ligands, and interacting with T cells via, for example, FAS/FasL pathway, are likely to contribute to the demise of activated effector T cells. Thus, TEX may abrogate beneficial effects of anti-tumor vaccines or other immunostimulatory regimens. TEX, which carry HER2 or other TAAs, can interfere with Ab-based therapies and reduce binding of these Abs to tumor cells, as has been shown for trastuzumab.85 Sequestration by TEX of Abs with the IgG1 isotype also interferes with antibody-dependent cellular cytotoxicity mediated by immune cells, one of the major mechanisms of therapeutic efficacy of many anti-cancer Abs. In an animal model of B-cell lymphoma, TEX were shown to bind complement, and thus protect tumor cells from complement-dependent cytolysis.86 Adoptively transferred activated NK cells and engineered CAR T cells or vaccine-induced CD8+ T effector cells are likely to encounter circulating TEX bearing an immunosuppressive cargo in patients with advanced malignancies.87 These effector cells will promptly undergo functional paralysis or apoptosis. In our hands, adoptive transfer of NK-92 cells to relapsed AML patients did not provide significant therapeutic benefits, presumably because anti-leukemia activity of the transferred cells was silenced by pre-existing immunosuppressive TEX.88 Emerging data suggest that TEX play a key role in regulating the sensitivity of malignant cells to immunotherapies by downregulating anti-leukemia functions of immune cells.
TEX as biomarkers of leukemia progression
TEX are currently viewed as tumor cell surrogates or ‘liquid biopsies’ and as a promising non-invasive metrics of immune dysfunction in cancer. Used as biomarkers of immune cell dysfunction and of tumor progression, TEX might emerge as the most informative non-invasive predictors of cancer outcome or response to therapy. In AML, plasma exosomes could serve as biomarkers of leukemic relapse, as discussed above, or as biomarkers for MRD.89 The potential of TEX for non-invasive monitoring of disease progression in cancer patients has been recently reviewed.90 Validation of TEX as immune biomarkers and/or biomarkers of tumor progression in prospective clinical trials is urgently needed. However, before such validation can be undertaken, many unsolved issues concerning TEX have to be resolved, including methodologies for their isolation from body fluids. The concept that TEX serve as surrogates of leukemic blasts can only be validated by side-by-side comparisons of TEX and leukemic blasts. Major challenges that remain are related to in vivo relevance of TEX in progression of hematologic malignancies and in leukemic relapse.
TEX as therapeutic targets
Since TEX are involved in blocking functions of anti-leukemia effector cells and driving tumor escape, their silencing or inhibition of TEX production is a desirable objective. It seems rational to aim for selective silencing of those TEX that promote leukemic transformation and proliferation of leukemic cells. Table 5 lists some of the relevant strategies that are currently being investigated. The strategies for specific silencing of TEX are likely to require careful and detailed mechanistic studies, because of a danger that indiscriminate blocking of all exosomes might interfere with the physiological intercellular communication system, which is necessary for sustaining life. Thus, while ameliorating pathological effects of TEX is highly desirable, it must be accomplished without interference with physiological activities of exosomes.
TEX are currently a subject of intensive investigations fueled by their emerging significance as regulatory elements of numerous cellular functions and as cancer biomarkers. Their unique molecular and genetic characteristics and abilities to alter functions of various normal and malignant recipient cells suggest that TEX play major regulatory roles in health and disease. In hematological cancers, TEX seem to be involved in all aspects of malignant transformation, promoting survival and resistance of blasts and suppressing anti-tumor activities of immune cells. The potential of TEX as biomarkers of malignant progression or responses to therapy are of special interest. TEX interference with immune therapies encourages the development of novel strategies for silencing their pro-tumor functions. The future progress in therapy of hematological malignancies largely depends on the understanding of how leukemic cells use TEX to promote their growth and progression.
This work was supported in part by National Cancer Institute Grants R01 CA168628 to TLW and R21 CA205644 to TLW and MB.