In normal adult tissues stem cells are undifferentiated cells that reside in a proper niche, where they are protected and can exert their functions. Stem cells show the ability of self-renewal and differentiation in adult cell tissue, maintaining tissue homeostasis. Stem cell niche can be identified in several tissues such as the crypts of the intestine, the bone marrow, the liver or lung tissues (11). After tissue injury, the niche transmits activation signals such as adhesion molecules, matrix proteins, oxygen, growth factors or cytokines to the stem cells for tissue regeneration. These signals are factors that allow cell-cell interactions between stem cells and neighboring differentiated cells (12).

In normal tissues stem cells remain in an undifferentiated state throughout adult life. Stem cells reach a first stage by becoming transient amplifying cells and highly proliferative cells then they asymmetrically divide and finally reach the last stage of differentiated cells, that leads them to build up and support tissues (13). In both stem and differentiated cells, the potential accumulation of intracellular pathways mutations can lead to a tumor-type phenotype (14). When a critical mutation threshold is reached, cells become CSCs, changing to a more aggressive behavior (15). CSCs, as normal stem cells, have the ability of self-renewal and differentiation, while they create a niche that proliferates independently of the surrounding tissue. These characteristics contribute to tumor initiation, growth and maintenance (13). Among the most deregulated intracellular pathways, wingless-related integration site (WNT)/beta-catenin, NOTCH and Sonic Hedgehog emerge, as they promote self-renewal and tissue morphogenesis (16). In addition, cellular growth, migration, differentiation and epithelial-mesenchymal transition (EMT) are regulated by phosphatidylinositol 3-kinase/AKT serine/threonine kinase 1/Phosphatase and tensin homolog (PI3K/AKT1/PTEN) axis, one of the majors signaling pathways in CSCs (17). TGF, SMAD, peroxisome proliferator-activated receptor (PPAR), mitogen-activated protein kinases (MAPK) and Janus kinase/signal transducers and activators of transcription (JAK-STAT) are often deregulated in CSCs (18). CSCs are not only involved in the process of tumor initiation, growth and maintenance, but also in metastasis (19). CSCs represent the most aggressive tumor subpopulation able to spread and form metastases even at distant sites. One of the key requisites for successful metastasis formation is stemness. Indeed, depletion of various stemness markers such as cluster of differentiation 44 (CD44) in breast CSCs (20) or octamer-binding transcription factor 4 (OCT4) and SRY-Box Transcription Factor 2 (SOX2) in colon CSCs, prevented tumor metastasis and tumor growth (21). Beyond stemness markers, several studies have been focused on the identification of cell-surface markers specifically expressed in the subpopulation of CSCs endowed with metastatic potential such as CD44v6, a CD44 variant isoform, in colon CSCs (22). A broad and extensive description of CSCs hallmarks and the methodologies used to characterize the CSC state is reported in (23). In the next paragraphs, we will briefly introduce how CSCs evade the immune system and resist conventional therapies.

Immunosurveillance is a set of immune-system related processes aimed at controlling the development of normal cells and detecting cancer cells. The innate and adaptive cells of the immune system respond to stress conditions, caused by tumor development, mainly by upregulating natural killer (NK) activator ligands and stimulating a more specific T lymphocyte response against cancer cells (24). NK cells are innate immune cells that recognize cells lacking major histocompatibility class I complex (MHC-I) and exert potent cytolytic activity releasing perforin and granzyme against transformed cells (24). NK cells mediate the tumor killing also triggering apoptotic pathways in tumor cells through the production of TNF-alpha or via direct cell–cell contact through activation of the Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL) and FAS ligand (FASL) pathways (24). Otherwise, T cells are the main component of the adaptive immunity that orchestrate a protective effector immune response, indeed, a high level of T cell infiltration in tumors is associated with a favorable prognosis in cancer patients (24). CD8+ T and CD4+ T helper 1 cells are the most prominent anti-tumor T cells, instead, through the exocytosis of perforin and granzyme containing granules, the former, and secretion of high amounts of proinflammatory cytokines, such as interleukin-2 (IL-2), TNF-alpha, and interferon-gamma (IFNG), the latter, promote T cell priming activation, cytotoxic T lymphocytes (CTL) cytotoxicity, but also, the anti-tumoral activity of macrophages and NK cells (24).

T and NK cells destructive effect on cancer cells is regulated even by TAMs, by increasing the number of active NKs, upregulating inhibitory T cell receptors programmed cell death protein 1 (PD-1) and Cytotoxic T lymphocyte associated protein 4 (CTLA-4), releasing factors such as TRAIL and inducing apoptosis in cancer cells (25, 26). However, during inflammation, TAMs can directly inhibit the proliferation of CD8+ T cell lymphocytes by regulating their metabolism or recruiting regulatory T cells (Tregs) (27). TAMs can also inhibit dendritic cell (DC) maturation and the secretion of IL-12 by DCs (28). TAMs and Tregs boost an immune-tolerant TME by secretion of molecules such as interleukin-10 (IL-10), TGFB1, and prostaglandins (28). Indeed, poor prognosis and reduced overall survival in oncological patients is correlated with high-grade TAMs (28). Tumor cells can evade the immune system by using different strategies like losing surface antigens that prevent recognition by cytotoxic T cells or downregulating cell surface NK activators, becoming invisible to detection by NK cells (28). However, the immune system can self-contribute to tumor development and progression, orchestrating an immunosuppressive inflammatory TME (24). This process is called “cancer immunoediting” and proceeds through three phases: elimination, equilibrium and escape (29). During the first phase the cytotoxic immune cells such as NK and CD8+ T cells kill transformed cells, although rare tumor subclones can survive (30, 31). These tumor subclones may enter the second phase where their growth is limited and stalled over time (30, 31). The steady pressure from the adaptive immune system and the genetic instability of cancer cells can make tumor subclones escaping immunosurveillance (30, 31). Cancer cells start proliferating unconditionally and adopt many features to escape from the immune system like downregulation of the antigen presentation machinery or inducting inhibitory immune checkpoint molecules (32). Moreover, cancer cells remodel the vasculature and extracellular matrix and supports cancer progression as well as therapy resistance (30, 31). This process can entail decreased IFN-gamma secretion by T cells, loss of antigen presentation and epigenetic changes (33).

Within the tumor CSCs control the immune system and regulate the composition of TME through the release of cytokines, chemokines, growth factors, metabolites and hormones playing an immunomodulatory role (34). CSCs develop different immunosuppressive strategies that promote tumor maintenance and growth. Downregulation of MHC-I complexes and activation of immune molecules such as cluster of differentiation 80 (CD80), human leukocyte antigen (HLA) and major Histocompatibility Complex Class I chain-related protein A/B (MICA/MICB), renders CSCs more resistant to cytotoxic effects exerted by CTL (35). Of note, the degree of tumor progression in the CSC niche has been attributed to a reduced CD8+ T cell infiltration and to an increase in TAMs (35). Moreover, CSCs interact through human leukocyte antigen G (HLA-G) with killer cell immunoglobulin like receptor, two Ig domains and long cytoplasmic tail 4 (KIR2DL4) and killer cell lectin like receptor C1 (KLRC1) to suppress NK activity (34). CSCs further drive recruitment and polarization Treg cells by secretion of factors like Chemokine (C-C motif) ligand 1 (CCL1), IL-2, interleukin-8 (IL-8), IL-10 and Transforming growth factor-beta-1 (TGFB1) (34). Moreover, Tregs produce TGFB1 and interleukin-17 (IL-17) to promote CSCs properties toward tumor progression and invasion (34). CSCs immune evasion properties are influenced by humoral factors: TGFB1, a cytokine that induces immune suppression, EMT and stemness; IL-6, secreted by TAMs, that induces and maintains CSCs, signal transducer and activator of transcription 3 (STAT3), a transcription factor required for the maintenance of pluripotency in stem cells or Chemokine (C-C motif) ligand 20 (CCL20) and its receptor that recruits Tregs to promote tumor progression enhanced by immune evasion (34). CTLA-4 and PD-1/programmed death-ligand 1 (PD-L1) represent two of the major immune checkpoints (34). Immunosuppressive myeloid cells, including macrophages and monocytic myeloid-derived suppressor cells (MDSCs) represent an additional layer of regulation of T cell activity and partially depend on secretion of factors like CSF1, CCL2, Chemokine (C-C motif) ligand 5 (CCL5), TGFB1 and prostaglandin E2 (PGE2), by CSCs (34). Collectively, all these interactions reshape the tumor microenvironment and create a habitat where immune cells support and are suppressed by CSCs (34).

Conventional therapies developed for cancer treatment are based on the following approaches such as chemotherapy, radiation therapy and surgical excision (36, 37). Chemotherapy is the most widely used and effective treatment for cancer; however, cancer cells as well as CSCs often elaborate simultaneous resistance to many drugs, even if they are structurally and functionally quite different (36, 37). This phenomenon is called multidrug resistance (MDR) or multifactorial pleiotropic drug resistance (36, 37). Many in vivo and in vitro studies demonstrated that administering chemotherapeutic drugs led to an enrichment in CSCs (36, 37). Drug resistance is caused by regular administration of chemotherapy drugs, that are dose- or time-dependent. The multiple mechanisms underlying MDR can be listed as follows: increased drug efflux and reduced drug uptake, efficient DNA repair mechanisms, reduced presence of reactive oxygen species (ROS), apoptosis evasion, hypoxia, vasculogenic mimicry (VM) activation, increased autophagy and decreased ferroptosis (38).

Mechanisms responsible for therapy resistance in CSCs are summarized in Figure 2 .

Several studies demonstrated that many chemotherapeutic agents in clinical use are susceptible to ATP-binding cassette transporters-mediated efflux (ABC), such as microtubule-targeting, alkaloids, taxanes, topoisomerase inhibitors, DNA-damaging anthracyclines and tyrosine kinase inhibitors (38, 39). This subfamily of transporters is mainly localized in human tissues of the brain, lung, breast, kidneys, liver, ovaries, prostate, placenta and pancreas (40).

CSCs express higher levels of MDR transporters than cancer cells or healthy cells (41). ABCB1, ABCG2 and ABCB5 are overexpressed respectively in ovarian CSCs (41), breast CSCs (42) and malignant melanoma initiating cells (MMIC) (43). Inhibitors of the ABC transporters are currently used in clinical settings, although side-effects and high toxicity have been reported in patients (44).

Efficient DNA repair mechanisms in CSCs are thought to be a major contributing factor in counteracting treatment-induced DNA damage (45). Efficient DNA damage repair system and the CSC long permanence in a quiescent G0 phase greatly reduce potential exogenous and endogenous DNA damage that could occur during DNA replication (45). Evidence demonstrates that DNA damage response (DDR) sensor proteins are upregulated in CSCs rather than tumor bulk cancer cells in monolayer cultures, thus conferring radio and chemotherapy resistance (46). Enhanced expression of DNA polymerase nu (POLN) contributes to chemoresistance in ovarian stem cells (47). Thus, cytotoxicity by chemotherapeutic drugs or radiotherapy-induced can be attenuated in CSCs based on an efficient DNA damage repair system (26).

CSCs show low intracellular levels of ROS, a group of highly reactive molecules, containing oxygen, that can promote DNA damage and influence the DDR machinery (48). Therefore, CSCs can dampen the entity of exogenous DNA damage induced by conventional therapy by expressing low levels of ROS, which production is mainly determined by the slow division rate of CSCs (48). Lower levels of ROS in CSCs result crucial in maintaining a stem cell-like phenotype, along with conferring resistance to radiation therapy and/or chemotherapy (49).

The ability of CSCs to metastasize and reach other organs should be reduced as cells undergo programmed cell death or apoptosis, where they lose contact with their extracellular matrix or neighboring cells (“anoikis”) (50). However, CSCs were reported to be anoikis resistant (50). Indeed, CSCs endowed with metastatic potential evade anoikis mechanism, therefore surviving and promoting the formation of metastatic lesions at a distant site (51). Notably, co-culturing CSCs with non-CSCs conferred anoikis resistance to non-stem cells in breast cancer (51). CSC-like cells protected non-stem cells from anoikis and promoted tumor growth (51).

Oxygen is necessary for metabolism and cellular energy production. In many tumors, oxygen levels are usually between 0% and 2% compared to normal physiological levels that can reach up to 9% and therefore the high metabolic demand requires the activation of hypoxia-inducible factors (HIFs) (52, 53). HIFs are heterodimers consisting of two subunits a and b that can translocate in the nucleus and interact with specific sequences leading to activation or repression of gene expression (52, 53). There are three different genes encoding for HIF subunits: hypoxia inducible factor 1 subunit alpha (HIF1A), hypoxia inducible factor 2 subunit alpha (HIF2A), and hypoxia inducible factor 3 subunit alpha (HIF3A) (52, 53). All three heterodimerize with the hypoxia inducible factor 1 subunit beta (HIF1B) subunit and are subject to posttranslational regulation that is dependent on oxygen levels in the environment (52–54). HIF1A and HIF2A through the upregulation of regulators such as SOX2, Nanog homeobox (NANOG), OCT4, KLF Transcription Factor 4 (KLF4), and the transcription factor MYC proto-oncogene protein (MYC), have been shown to promote stemness and CSC phenotype (55). Upregulation of HIF-1 induces the expression of genes involved in angiogenesis, cell survival, and metabolism, conferring a selective advantage to CSCs (56). It has been demonstrated that breast cancer cells lines, MCF-7 and MDA-MB-231, display increased subpopulations of tumor cells with stem-like characteristics (56). Hypoxia is a hallmark of the CSCs environment that is essential for CSCs development, maintenance, tumor growth and resistance to therapy (57). Evidence suggests that the hypoxic niche in colon cancer protects CSCs from chemotherapy (58). Moreover, in ovarian cancer stem cell lines, SK-OV-3 and HO-8910, it has been demonstrated that chemotherapy treatment, under hypoxia conditions, induced CSC-like properties (59). The mechanisms through which hypoxia exerts its function are complex but can be summarized in shifting the metabolism toward aerobic glycolysis, reduced expression of pro-apoptotic factors, dysregulation of ROS and redox mechanisms, increasing genomic instability and aberrant cell cycling (48).

Vascularization plays an important role during carcinogenesis and metastasis. VM can provide sufficient blood supply for tumor growth, independently of endothelial cells (60, 61). VM is a process of blood vessel formation that cancer cells and CSCs employ to increase the blood supply of angiogenesis (60, 61). It is a mimicry process whereby malignant cells mimic the function of endothelial cells to form blood vessels by reshaping the extracellular matrix (60, 61). CSC VM has been observed in many tumors such as breast cancer and melanoma (62). Evidence shows that vasculogenic mimicry is mostly present at the early stages of tumor development when blood supply is most needed, as the tumor grows where the vessels created by endothelial cells are established (63). Studies also show that the early stage of CSC serves as tumor vasculogenic stem/progenitor cells that can differentiate into tumor vasculogenic endothelial cells (64, 65). New vessel formation, and particularly VM, makes the eradication of the tumor even more complex and unsuccessful, giving the tumor the ability to metastasize (64, 65).

Autophagy is a catabolic process that degrades and recycles cellular components and exhibits both protective and destructive roles in the TME under physiological stress conditions such as nutrient deprivation and hypoxia (66). The activation of autophagy may lead to an arrest of tumor development, but at the same time it can support CSC self-renewal and resistance to therapy (66). In CSCs, autophagy contributes to maintain self-renewal and proliferation properties, avoiding senescence (67). Evidence suggests that autophagy is involved in mechanisms that mediate resistance to therapy, in renal carcinoma and breast cancer (68, 69). Experiments carried out to inhibit autophagy have shown increasing sensitivity to radio- and chemotherapy in nasopharyngeal and breast CSCs, respectively (70, 71). In addition, the upregulation of signaling pathways mediating autophagy, such as SOX2- beta-catenin/BECLIN1, determines resistance to chemotherapy (72).

Iron is an essential cofactor for several metabolic reactions and contributes to the formation of ROS (73). Ferroptosis can be defined as a form of iron-catalyzed necrosis and occurs through the intracellular accumulation of ROS, induced by lipid peroxidation (74). Current studies demonstrate that during tumor development the levels of iron and its transporters increase in CSCs compared to cancer cells (74). Although iron accumulation promotes ferroptosis, CSCs maintain a balance that prevents toxic lipid peroxidation (75). Chemotherapeutic drugs generate ROS that can induce oxidative damage and apoptosis (75). However, CSC ability to control ferroptosis reduces the harmful effect of ROS species conferring chemotherapeutic resistance (75). Inducing high levels of ferroptosis is indeed currently used as an innovative approach to revert chemotherapy resistance, specifically in the CSC population (76).

CAFs are highly heterogeneous stromal cells which represent the major modifiers of TME by the synthesis of soluble factors that promote tumor progression, stemness and angiogenesis in several cancers including prostate, gastric and non-small cell lung cancer (79–81). CAFs also contribute to tumor immune evasion both directly and indirectly. Different studies prove that CAFs are associated with T cells impairment, preventing their activation by secretion of C-X-C motif chemokine ligand 12 (CXCL12) and TGFB1 (82, 83). The primary role of CAFs is the establishment and apposition of the extracellular matrix (ECM) (84). ECM composes the scaffold for tissues and organs and facilitates cells crosstalk, both in healthy and malignant conditions. Jachetti et al. demonstrated that ECM proteins inhibit T cell proliferation and effector function (85). In addition, ECM can improve drug resistance by acting as a physical barrier. Besides, it has been shown that collagen, one of the most abundant proteins in ECM, can promotes stemness through the activation of an integrin/PI3K/AKT1/SNAIL signaling pathway (86).

Mesenchymal stromal cells (MSCs) are a substantial component of TME, recruited and re-educated by tumor cells in order to sustain tumorigenesis (87). Indeed, tumor associated-MSCs are crucial promoters of cancer hallmarks. It is shown that IL-6 produced by MSCs increases endothelin 1 (ET-1) expression in colorectal cancer (CRC) cells, resulting in the activation of AKT1 and ERK in endothelial cells which lead to tumor neo-angiogenesis enhancement (88). Several studies demonstrated that MSCs contribute also to tumor invasiveness and progression by regulating EMT regulators, like Twist, Snail and Zinc finger E-box binding homeobox 1 (ZEB1) (89–92). Finally, MSCs interact and suppress TME-embedded immune cells, either directly or through the release of factors like TGFB1, IL-2 and IL-10 (93) and, moreover, play a crucial role in enhancing stemness of cancer cells. Indeed, in physiological conditions, MSCs shape and support tissues and promote stemness features of the stem cell niches. Similarly, MSCs interact and promote CSC stemness in tumors via soluble factors (52).

Although immune cells should prevent and resolve tumor progression, they act as promoters of cancer development under the pressure of TME signalosome (94). MDSCs are regulators of immune homeostasis (95). Cancer cells exploit MDSCs activity to escape immune surveillance, indeed MDSCs are commonly present in TME for their capability to facilitate tumor progression by establishing immune-suppressive conditions in different ways (96). ROS, IL-10 and TGFB1 produced by MDSCs negatively regulates CD8+ T cells activity against cancer cells (96). Moreover, MDCSs up-regulate PD-L1 expression, resulting in suppression of the immune response against tumors (97). MDSCs also regulate indirectly the immune response exacerbating TME by factors essential for T lymphocytes functions, such as L-arginine, which is crucial for T cells proliferation and activity (98, 99). MDSCs promote CSC stemness by miRNAs able to trigger CSCs stemness program (100).

Tregs are spontaneously attracted by immunosuppressive cytokines produced by tumor and tumor-associated cells (101). As well as MDSCs, Tregs promote tumors immune evasion by releasing cytokines able to suppress the activation of the immune response effectors (102, 103). Recent evidence suggests that Tregs are important regulators of CSCs stemness. Indeed, in several cancers, Tregs promote stemness-related pathways (104), facilitate EMT (105) and angiogenesis (101).

Tumor cells and TME not only re-educate and exploit MDSCs and Tregs but also induce depletion of tumor killing activity exerted by immune response effectors cells, NK cells and lymphocytic cells (106, 107). Although in the early stages of tumorigenesis NK cells are lethal for tumor cells, they slowly exhausted their killing function under the pressure of TME factors (108). Indeed, TGF-beta produced by CSCs, MDSCs and Tregs, impairs NK cells cytotoxicity, inhibits the release of IFNG and reduces the expression levels of killer cell lectin like receptor K1 (KLRK1) receptor in several tumors (109–111). Also, TME hypoxia conditions inhibit NK cells by downregulating expression of NKp46, NKp30, NKp44, KLRK1, perforin (PRF1), and granzyme B (GZMB) (112). Finally, lactate produced by tumor cells leads to the acidification of TME which induces apoptosis of NK cells (113). The same conditions which inhibit NK cells affect also lymphocytic cells activity, the most potent immune weapons against tumor cells (114, 115). Besides, downregulation of MHC-I, along with the up-regulation of immune checkpoints, (i.e. PD-L1) allows tumor cells to ensure themselves immune evasion (116, 117). Among the immune cells present in the TME, a focus on TAMs and their hallmarks will be provided in the next paragraphs.

Recently, CSCs-TME interplay gained interest in cancer research as a potential therapeutic target against tumors. TME promotes a stem-like state in CSCs supporting their self-renewal, survival, and therapeutic resistance through different molecular mechanisms (118). CAFs, the most represented cells in TME, release cytokines like IL-6, able to sustain the expression of stemness-related genes like SOX2, NANOG and OCT4 in CSCs (119). On the other hand, CSCs drive TME immunosuppressive polarization and persistence (35). CSCs can regulate immune system activity through the release of immunosuppressive secretome (i.e. IL-10, TGFB1) showing a more efficient capability to recruit immune cells with pro-tumoral activity (Tregs, MDSCs and especially TAMs) which sustain CSCs stemness by releasing factors like platelet-derived growth factor (PDGF), IL-8, CXCL12 (120, 121).

A key hallmark of cancer is the ability to engage in an intricate communication with tumor cells and by activating proliferative signaling programs (138). TAMs positively support the cancer cell-cycle state by secreting various growth factors and cytokines. Among TAMs released factors, IL-6, IL-10 and IL-8 foster signaling pathways directly involved in stimulating cancer cell proliferation and tumor growth (139–142).

TAMs predominantly display an immunosuppressive M2 state in TME (130). M2 TAMs unbalance the immune surveillance role of T cells and favor the promotion of cancer cells escape from the immune system (130). More in detail, TAMs produce a plethora of molecules, such as TGFB1, IL-10, and PGE2 that act on T cells, disrupting the anti-tumoral role both CD4+ and CD8+ subtypes, and increasing the recruitment of Tregs, that enhance the pro-tumoral immune depletion (130, 143).

Chemoresistance is another cancer hallmark. TAMs play a crucial role in the acquisition of a cancer chemoresistant phenotype, through the secretion of inflammatory cytokines such as IL-6 and TNF-alpha which activate pro-survival programs (144, 145). Moreover, TAMs can enhance the efflux of chemotherapeutic drugs from cancer cells, reducing their therapeutic efficacy (146).

TAMs promote neo-angiogenesis, vital for both tumor growth and metastasis. Accordingly, TAMs release pro-angiogenic factors such as vascular endothelial growth factor (VEGF) and fibroblast growth factor (FGF), which drive the activation of new blood vessels formation signaling pathways (147, 148). The neo-angiogenesis not only is essential to feed cancer cells with nutrients and oxygen, but also it is critical for tumor mass growth and to guide metastatic spreading process (147).

The pro-invasiveness and pro-metastatic role of TAMs is well-documented in literature (5, 149). Accordingly, TAMs can induce EMT in tumor cells toward a more mesenchymal phenotype, enhancing their more malignant invasive phenotype (149). Furthermore, TAMs can secrete matrix metalloproteinases through which they digest the extracellular matrix components, allowing cancer cells to reach the surrounding tissues (150).

In summary, TAMs significantly impact multiple hallmarks of cancer. Through their roles in sustaining proliferative signaling, immunosuppression, chemoresistance angiogenesis and metastasis, TAMs represent a critical player in all stages of cancer progression, from early to late ones.

One of the major drawbacks in counteracting cancer spread and resistance consists of the capacity of CSCs to migrate into secondary sites and avoid immune surveillance (2). Given their plastic behavior, their self-renewal capacity and treatment resistance, CSCs can foster metastasis formation from the primary tumor environment by disseminating into further districts, establishing the metastatic niche (151). In this context, a model that explains the architecture of the niche has been proposed by Lyden et al., in which the CSCs by migrating, reach a permissive and suitable microenvironment, the pre-metastatic niche, and by becoming disseminated tumor cells (DTCs), they can colonize and proliferate (152) through direct competition with normal stem cells for the niche occupation and establishment (153).

The metastatic niche characteristics vary depending on the specific components considered: the interacting cell types, ECM proteins, survival and self-renewal signals, but mostly the secondary site locations, that can either sustain and foster the metastatic niche, or set a hostile environment for the DTCs (154). The DTCs have to face several issues when colonizing a secondary site, including the lack of growth and extracellular matrix remodeling factors, that can hamper their survival and proliferation, thus adjusting into the new niche and metastasize (155). The disseminated cells will shape their surroundings to build a supportive metastatic niche and exploit the functions of both CSCs and metastatic stromal cells (2). However, studies analyzing human colorectal cancer samples displayed that metastatic occurrence arises from primary tumor cells, that are resistant to chemotherapy and might stay quiescent for a prolonged time (156).

Notwithstanding, little evidence investigating the genetic profiling of the tissue-derived and metastatic CSCs emerged, in the consideration of the metastasis signature mutations occurrence at the level of the primary tumor. This process can represent the first tool of selection in the CSCs population to direct a pool toward migration and extravasation into secondary sites (151). With this premises, the most accredited option relies on the fact that metastasis-driving alterations are present within the heterogeneous CSCs profile, and their expression selects the DTCs that will acquire a plastic and resistant profile (157).

Nevertheless, further evaluations need to be carried out, especially in the context of EMT pathways, and stem-like features involved in both the primary tumoral site and in the metastatic environment, highlighting similarities and differences among CSCs and DTCs (158).

Historically, the metastatic niche has been described as a cell-enriched environment constituted mostly by immune and stromal cells which secrete proteins and factors that sustain growth and self-renewal of CSCs, that consequentially stimulate the activation of angiogenic pathways aimed to the promotion of tumor invasion and metastatization (154, 159).

CSCs form the primary tumor can favor the diffusion of pro-tumorigenic and proangiogenic factors such as VEGF-A, TGFB1, TNF-alpha and lysyl oxidase (LOX) that induce the expression of S100A (a Ca2+ binding protein involved in endothelial remodeling) in the metastatic area (2). In the metastatic site the vasculature system boosts the recruitment of metastatic cancer stem cells (MetCSCs) by producing fibronectin and vascular endothelial cell adhesion molecule (VCAM). Consistently, it has been observed that the CCL2-CCR2 (C-C chemokine receptor type 2) axis promotes the establishment of inflammatory monocytes to the metastatic site, where they will transition into metastatic-associated macrophages (MAMs) and will enhance the extravasation and survival of metastasis-resident cancer cells (160).

Another important lead of metastasis formation is depicted by TAMs. Principally, TAMs promote tumor cell invasion and dissemination, and through their ability to release cytokines and factors that support growth and ECM-shaping (MMP-2, MMP-9), milk fat globule-EGF factor 8 (MFGE8), IL-6 are correlated with tumor progression and metastasis (161). TAMs derive from circulating Ly6C+CCR2+ inflammatory monocytes that are produced by hematopoietic stem cells (HSCs) in the bone marrow, that, when interacting with tumor tissue, are addressed toward a more cancerous-like profile (161, 162).

TAMs and CSCs crosstalk has been widely described in the last years, investigating whether their interaction may be direct or indirect, and which may be the effects on CSCs in the primary tumor, including chemoresistance, differentiation and proliferation (163). TAMs are essential in supporting metastasis establishment once CSCs migration has occurred. More specifically, studies conducted on lung and liver metastatic murine models, showed how inhibiting TAM recruitment in metastatic niches resulted in a reduced burden, indicating its paramount role in the onset and maintenance of metastasis by supporting both extravasation and intravasation in secondary sites of CSCs (164, 165).

One accredited metastasis hypothesis linking the role of TAMs in facilitating CSCs metastatization relies on the ability of metastatic cells to occupy niches in which are present CSCs (166). More in detail, it is thought that TAMs and CSCs derive from cell hybrids and set metastasis in further sites (166). The theory was proposed by John Pawelek in 2006, and he explained that myeloid and tumoral cells could perform a genomic hybridization (167). TAMs due to their migratory ability and the tissue-repair feature could transport the CSCs spheroids through either bloodstream or lymphatic circulation, and permit a favorable environment for metastatic initiation (168). Within the metastatic microenvironment, TAMs play a crucial role in shaping the behavior of CSCs, especially regarding tumor advancement and the colonization of cancer cells at secondary sites (169). A crucial aspect of TAMs is their role in promoting EMT, which is a vital process in the morphological alterations of cancer cells and contributes to the enhancement of their malignant traits (170). In triple-negative breast cancer, CCL2 secreted by TAMs activates AKT signaling pathways, resulting in heightened beta-catenin activity in CSCs (171). This pathway is essential for facilitating EMT and sustaining the properties of CSCs within the TME (172). In oral squamous cell carcinoma (OSCC), high levels of TAM-derived IL-6, promote EMT and enhance the expression of genes associated with stemness, via the IL-6/STAT3/thrombospondin 1 (THBS1) signaling pathway (173).

To sustain CSCs in pancreatic ductal adenocarcinoma (PDAC), TAMs utilize a critical mechanism involving the interferon-stimulated gene 15 (ISG15) signaling pathway (174). By releasing the ISG15, TAMs enhance the self-renewal, invasive potential and tumorigenic capabilities of CSCs (175). Among the several ways in which TAMs support CSCs behavior, the creation of an immunosuppressive microenvironment exerts a key function. Within the TME, TAMs predominantly display a M2 phenotype, which is known for its role in promoting immunosuppression (176). This phenotype fosters a protective microenvironment that shields CSCs from immune system attacks. By releasing immunosuppressive cytokines like IL-10 and TGFB1, TAMs effectively suppress the function of cytotoxic T cells and other immune cells, allowing CSCs to remain undetected and avoid destruction (177). TAMs influence the growth of CSCs through both direct contact and secretory mechanisms. In highly metastatic breast cancer, CSCs express hyaluronan synthase 2 (HAS2), which is crucial for creating a pro-metastatic microenvironment (177). This expression facilitates interactions between CSCs and TAMs, leading to the secretion of platelet-derived growth factor-B subunits (PDGFB) by TAMs (177).

PDGFB subsequently stimulates bone stromal cells to secrete fibroblast growth factors 7 and 9 (FGF7 and FGF9), which support CSC proliferation and survival (178). Moreover, in breast cancer, the EMT enhances the expression of cluster of differentiation 90 (CD90) and ephrin type-a receptor 4 (EPHA4), facilitating direct physical interactions between CSCs and TAMs through the binding with their respective receptors. When the EPHA4 receptor on carcinoma cells is activated, it triggers the sarcoma SRC proto-oncogene, non-receptor tyrosine kinase (SRC) and nuclear factor- kappa B (NF-kappa-B) signaling pathways. This activation leads to NF-kappa-B in CSCs induction of the secretion of various cytokines that help maintaining the stem cell state (179).

By preserving the stem-like properties of CSCs and boosting their migratory and invasive abilities, TAMs facilitate the detachment of CSCs from the primary tumor, enabling the formation of secondary tumors in distant organs (149).

Once malignant cells escape from the primary tumor, they intravasate and disseminate through the lymphatic and/or circulatory system, eventually establishing secondary tumors at distant sites. Research into lung metastasis reveals that when tumor cells reach their target location, they form micro-clots in conjunction with platelets, resulting in their entrapment within the blood vessels of the target tissue (180). Once arrested, the tumor cells secrete CCL2, which creates a gradient that attracts Ly6C monocytes (181). These recruited monocytes undergo differentiation into MAMs, which play a pivotal role in facilitating the extravasation of tumor cells by releasing VEGF, a factor known to enhance vascular permeability (182). Under the influence of CSF1, the primary lineage regulator for most macrophage populations, MAMs support the survival of tumor cells and contribute to their sustained growth through processes related to angiogenesis (183).

Recent studies conducted on CRC evidenced the interaction of CRC cells and TAM. Of note, a paramount interaction between CRC cells and M2 macrophages in the promotion of colorectal liver metastasis (CRLM) emerged (184). To date, CRLM is mediated by interactions between tumor cells and the TME in the liver and is considered one of the most common secondary liver cancers (185). Nevertheless, the mechanisms involved in the cancer cell-derived activation of M2 macrophages need further investigations in both CRC and CRLM. Notably, exosomes derived from tumors can polarize macrophages toward a M2 cellular profile, which in turn promotes metastasis (186, 187). Zhao et al, demonstrated that exosomes derived from CRC cells displayed a role in inducing M2 polarization through the secretion of microRNA-934 (miR-934) and the downregulation of PTEN expression, and activation of PI3K/AKT signaling cascade. Finally, miR-394 activated polarized M2 macrophages which promoted CRLM through C-X-C motif chemokine ligand 5 and 13 (CXCL5)/(CXCL13)/NF-kappa-B/p65/miR-394 positive feedback mechanism (188).

Another study (189) conducted on glioblastoma investigated the role of glioblastoma stem cells (GSCs) and TAMs in tumor progression and metastatic potential. The authors screened GSCs factors that could polarize macrophages, and they evaluated a potential group of proteins produced by GSCs with the ability of behaving as TAMs chemoattractant (189). Periostin (POSTN) emerged as a valuable factor expressed by the stem cells (189). It plays a role in the PI3K/AKT and WNT signaling pathways, which are involved in tumorigenesis (190, 191). In particular, evidence highlighted that CSCs profited from the POSTN-induced WNT augmented signaling, supporting a favorable metastatic colonization in breast cancer setting (192). Additionally, when silencing POSTN, TAM density was sensibly reduced, thus reinforcing the idea that GSCs can recruit TAMs and foster tumor growth by secreting POSTN. Consistently, GSCs established in the tumoral area, where they exploited the surrounding microenvironment by attracting TAMs from the peripheral circulation to set a more beneficial space for the reciprocal survival and growth of the resident populations and enhancing the metastatic CSCs potential. These observations need further investigation and open new scenarios regarding the involvement of TAMs, the role of CSCs, and their complex interplay in affecting the metastatic niche.

TAMs represent a cellular immune system subpopulation directly involved in the tumor formation and progression through the activation of several pro-tumoral signaling pathways within the CSCs, thus providing the creation of a tumor niche necessary for CSCs survival and expansion. Cellular matrix elements represent critical components for the tumor niche structure maintenance which help the direct crosstalk between CSCs and the surrounding cells, including TAMs. The intricate bi-directional communication between TAMs and CSCs is increasingly recognized as a critical factor in tumor biology. This interaction is underscored by a growing list of factors, ligands/receptors, shown in Table 1 , derived from both TAMs and CSCs that are implicated in the mutual co-dependent maintenance of CSC stemness and the supportive actions of TAMs. The complex network of signaling molecules and pathways involved in this crosstalk not only influences tumor progression but also impacts therapeutic resistance, making it a focal point for cancer research.

In breast cancer stem cells (BCSCs) the overexpression of the HAS2 is implicated in the new synthesis of hyaluronic acid, a major polysaccharide component of the ECM which drives the physical interaction to TAMs, via CD44 receptor expressed on their surface. The hyaluronic acid/CD44 interaction stimulates TAMs to release the growth factor PDGFB, which induces CSC self-renewal (178, 193). In addition, TAMs/CSCs in vitro co-culture confirmed the oncogenic role of hyaluronic acid-expressing CSCs/CD44-TAMs interaction in the activation of different signaling pathways such as PI3K–Eukaryotic Translation Initiation Factor 4E Binding Protein 1 (EIF4EBP1)–SOX2, implicated in CSCs pool maintenance (194, 195). Interestingly, it has been found that BCSCs cooperate directly with TAMs through cluster of differentiation 11b (CD11b) and CD90 binding. This anchoring stimulates EPHA4 receptor-mediated induction of both the NF-kappa-B and SRC signaling pathways ensuring CSCs pool stemness state (196). Similarly, in a triple negative breast cancer (TNBC) model, the butyrophilin subfamily member A3 (BTN3A3) receptor enhances cancer stemness markers (i.e. NANOG, OCT4, SOX2) via juxtacrine interaction with its ligand, liver and lymph node sinusoidal endothelial cell C-type lectin (LSECtin), a transmembrane protein expressed on TAMs surface (197). Furthermore, CSCs engage a juxtacrine signaling pathway with the TAMs via GPI-anchored protein CD90/CD11b. Specifically, CSCs express the membrane GPI-anchored protein CD90 and EPHA4. Mechanistically, CD90 creates a bridge to bind the integrin CD11b on TAM surface, whereas the receptor EPHA4 interacts with its ligand, Ephrin, expressed by TAMs, inducing the expression of both SRC and NF-kappa-B driving tumor progression and metastatic dissemination (196). In pancreatic cancer, immunomodulatory cationic antimicrobial peptide 18/LL-37 (hCAP-18/LL-37) on TAM, anchors the formyl peptide receptor 2 (FPR2) and the P2X purinoceptor 7 receptor (P2X7R) expressed on pancreatic cancer cells, which lead to the activation of stemness genes (i.e. KLF4, SOX2, OCT3/4 and NANOG) driving CSC self-renewal, invasion, tumorigenicity (198).

Different studies showed that an indirect paracrine interaction between TAMs and CSCs, driven by a plethora of inflammatory molecules including cytokines, chemokines, growth factors, was also crucial in the determination of CSCs fate and behavior. Particularly, IL-6 is one of the most representative pro-inflammatory cytokines in the context of TME. It is critically upregulated in many tumors, underlying the strong correlation between inflammatory stimuli and tumor progression by affecting multiple cancer signaling pathways (199). IL-6 derived from TAMs induces the proliferation of CD44+ Human Hepatocellular Carcinoma Stem Cells (HHCSCs) via STAT3 pathway induction (139). In addition, TAMs secrete high levels of IL-6 increasing stemness markers (i.e. SOX2, OCT3/4 and NANOG) and consequently CSCs expansion in breast cancer cells via STAT3 pathway supporting tumor cells migration and angiogenesis (125).

Paracrine communication mechanisms between TAM and CSCs are driven by several molecules. TAMs can enhance the CSC-like phenotype via TGFB1, which induces EMT program activation in a hepatocellular carcinoma (HCC) (120). Similarly, TAMs induce stemness, EMT and chemoresistance in HCC by realizing TNF-alpha via the WNT/β-catenin axis (200). It has been discovered that TAMs can produce Chemokine (C-C motif) ligand 18 (CCL18). In squamous cell carcinoma of the head and neck model (SCCHN), CCL18 produced by TAMs regulates metastasis through the activation of EMT program and cancer stemness (201). TAMs releasing CCL2 is correlated with worse prognosis in breast cancer. Particularly, TAM-produced CCL2 in the context of breast cancer microenvironment activates AKT/beta-catenin signaling resulting in EMT and CSC properties in TNBC (172).

Exosomes derived from TAMs have shown unrevealed aspects about the role of TAMs in the support of cancer progression. Specifically, it has been found that annexin A3 (ANXA3)-loaded exosomes derived from TAMs impaired ferroptosis process in laryngeal cancer cells supporting lymphatic metastasis. More in detail, ANXA3 in exosomes regulates negatively the ubiquitination of activating transcription factor 2 (ATF2), a transcription factor that induces ChaC Glutathione Specific Gamma-Glutamylcyclotransferase 1 (CHAC1) expression, thus blocking ferroptosis in lung squamous cell carcinoma (LSCC) cells (202). Moreover, CD163+ TAMs release exosomes that are absorbed by epithelial ovarian cancer cells (EOCCs) (203, 204).

During the tumor progression, TAMs can create an immunosuppressive TME facilitating the immune escape of CSCs. The creation of an immunosuppressive milieu depends on a fine balance between the inhibition of pro-inflammatory immune cells and the activation of immunosuppressive TAMs-dependent counterparts. Accordingly, TAMs promote the upregulation of cluster of differentiation 47 (CD47) ligand on different cancers stem cells (including pancreatic, HCC and leukemia), which interacts to signal-regulatory protein alpha (SIRPA) on immune cells inhibiting phagocytic process (205–207). Parallelly, TAMs can also inhibit the adaptive immune system. Particularly, TAMs boost both inhibitor immune checkpoints expression PD-1 and its ligand PD-L1 in T cells and CSCs, respectively (208). The concomitant expression of PD-L1 and PD-1 impedes the cytotoxicity in T-cells (208).

Overall, some evidence showed how TAM-derived factors and TAM-CSCs physical interactions drive the activation of a great number of pathways in CSCs that are responsible of the maintenance of stemness in different cancer histotypes. These stemness-related hallmark pathways include Sonic hedgehog (SHH), STAT3, NOTCH, PI3K/AKT, WNT/beta-catenin, and NANOG (18). Particularly, TAMs induce STAT3 pathway regulating the expression of stemness genes, via NF-kappa-B activation, in CSCs in different malignancies including breast cancer, liver cancer, prostate cancer, pancreatic cancer and colon cancer (139, 209–213). TAMs activate WNT/beta-catenin and SHH pathways, in CSCs, by leading transcriptional activation of stemness related genes in liver cancer, prostate cancer and lymphoma after secreting TNF-alpha, CCL5, pleiotrophin respectively (200, 211, 214). Furthermore, TAMs support cancer stemness through the direct activation of SHH pathway or through the induction of stemness-related alternative pathways (196, 213, 215–217). Specifically, TAMs sustain stemness via direct activation of SHH pathway in colon cancer (213), meanwhile SHH alternative signaling pathways are TAM-induced in pancreatic cancer (TGFB1/SMAD2/SMAD3/NANOG pathway) (215), in liver cancer (via the NOTCH pathway) (216), breast cancer (via the SRC Proto-Oncogene, Non-Receptor Tyrosine Kinase (SRC) pathway) (196), and in glioma via extracellular regulated kinase 1/2 (ERK1/2) pathway (217).

Innovative studies indicate that the complex communication between CSCs and TAMs has a critical pivotal role in the acquisition of a chemoresistant phenotype refractory to anticancer therapies. In OSCC TAMs influence positively the formation of CSC-like cells, via the induction of stemness markers of the SOX2, OCT4, and NANOG genes, leading to a strong reduction of the percentage of apoptosis in OSCC, supporting cell migration and chemoresistance to vincristine (218). Similarly, TAMs release Pleiotrophin (PTN), which interacts with the protein tyrosine phosphatase receptor type Z1 (PTPRZ1) receptor on the surface of CSCs, in OSCC model. The ligand/receptor interaction activates the FYN proto-oncogene (FYN)-AKT pathway, sustaining both the expression of stemness characteristics in CSCs and chemoresistance in tumor cells (219). Furthermore, MFGE8 in cooperation with IL-6, from TAMs induces both STAT3 and SHH signaling pathways in non-small cell lung cancer stem cells (NSCLCCSCs) leading to chemoresistance (213). Despite a growing body of research that has elucidated various molecular mechanisms underlying the interactions between TAMs and CSCs, significant gaps in our understanding remain. The intricate crosstalk between these two cellular populations is a complex phenomenon that has not yet been fully characterized.

An overview of the most significant mechanisms of indirect and direct interaction between TAMs and CSCs are shown in Figure 5 .

scRNA-seq has emerged as a pivotal tool for dissecting the heterogeneity of TAMs within the TME. This technology allows the analysis of gene expression at the resolution of individual cells, providing unprecedented insights into the distinct subpopulations of TAMs and their functional states. Numerous studies have highlighted significant variations in the transcriptional profiles of TAMs across different tumor types, underscoring their role in modulating the immunosuppressive landscape of the TME (133). The application of scRNA-seq in cancer research has revealed the coexistence of TAMs with pro-tumor (M2-like) and anti-tumor (M1-like) phenotypes within tumors. This duality emphasizes the functional plasticity of TAMs in cancer progression. Recent advancements have utilized scRNA-seq to trace the developmental trajectories of TAMs, identifying key signaling pathways that regulate their polarization and function. Valdes-Mora et al. demonstrated the utility of high-throughput scRNA-seq for analyzing thousands of tumor cells, including TAMs, revealing transcriptional programs associated with different TAM states, further elucidating their roles within the TME (254).

To fully leverage scRNA-seq for TAM characterization, several bioinformatics methodologies are employed (252):

Quality Control and Filtering: Tools such as Seurat (255) and Scanpy (https://scanpy.readthedocs.io/en/stable/) are commonly used to filter low-quality cells based on the number of detected genes and mitochondrial content.

Normalization: Normalization of scRNA-seq data is essential for accurate downstream analysis. Techniques like Scran or DESeq2 (256) provide effective approaches for normalization.

Clustering: To identify distinct cell populations, clustering algorithms like Louvain or Leiden are employed, allowing for robust community detection in high-dimensional datasets (257).

Differential Expression Analysis: To uncover differences in gene expression across TAM subpopulations, tools such as edgeR (258), DESeq2 (256), or MAST (259) are frequently used, depending on the analysis framework.

scRNA-seq has provided groundbreaking insights into the transcriptional diversity and functional heterogeneity of TAMs across various cancer types (260). This approach has enabled the identification of distinct TAM subtypes, each contributing differently to tumor immunity and progression. Specifically, studies in breast cancer have delineated M1-like and M2-like TAM populations, revealing their unique roles in promoting or inhibiting tumor growth (261, 262). This emerging knowledge is crucial for the development of targeted therapies aimed at reprogramming TAMs to a more anti-tumor state, offering new avenues for therapeutic intervention in cancer. scRNA-seq represents a transformative approach in TAM research, providing high-resolution profiling of individual TAMs and enabling the identification of diverse subpopulations based on their gene expression profiles. By combining scRNA-seq with advanced bioinformatics tools, researchers can uncover the full spectrum of TAM heterogeneity and its implications for cancer progression and therapy. This method excels at revealing transcriptional heterogeneity and elucidating the cellular and molecular mechanisms underlying TAM function within the TME. However, it requires tissue dissociation, which disrupts the spatial organization of the tumor microenvironment and leads to a loss of spatial information. This limitation prevents a direct understanding of TAM interactions within their native tissue context, which is crucial for fully characterizing TAM functionality in relation to the TME.

While scRNA-seq has provided significant insights into the heterogeneity of TAMs, it lacks spatial resolution, which is crucial for understanding their interactions within the TME. Spatial transcriptomics bridges this gap by integrating gene expression data with spatial information, allowing for the precise mapping of TAM distribution and organization within tumor tissues. This spatial context is essential for capturing the complexity of TAM interactions with other cell types and their influence on tumor progression. Spatial transcriptomics has been extensively applied to study the spatial dynamics of TAMs across various cancer types. It has been demonstrated that TAMs located within the tumor stroma and at invasive tumor margins exhibit distinct gene expression profiles and functional states, which play a pivotal role in driving tumor progression and metastasis (263). By integrating spatial transcriptomics with scRNA-seq data, researchers can gain a more comprehensive understanding of TAM trajectories and their interactions with other cells in the TME. Recent developments in bioinformatics have facilitated the analysis of spatial transcriptomics data, from preprocessing to functional interpretation.

These advancements include:

Data Preprocessing: The initial steps of spatial transcriptomics analysis involve generating a gene expression matrix along with spatial coordinates. Tools such as Space Ranger (10X Genomics), SAW (Stereo-seq), and starfish (ISS/ISH) are widely used for data preprocessing, depending on the platform and methodology employed (264).

Comprehensive Analysis Tools: Seurat (255) and Scanpy (https://scanpy.readthedocs.io/en/stable/) are versatile tools frequently used for both scRNA-seq and spatial transcriptomics analysis. These platforms offer functionalities for filtering, normalization, and various downstream analyses. For more specialized spatial transcriptomics tasks, Giotto (265), STUtility (266), and Squidpy (267) provide extended capabilities, including advanced spatial analyses (264).

Dimensionality Reduction and Clustering: Techniques such as principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), and uniform manifold approximation and projection (UMAP) are widely employed for dimensionality reduction (268). In addition, spatial-specific algorithms like BayesSpace, SC-MEB, SpaGCN, and STAGATE leverage spatial information to enhance clustering accuracy and identify spatial features (264).

Deconvolution and Cell Typing: Since sequencing data often represents aggregate signals from multiple cell types, deconvolution techniques are required to resolve individual cell-type contributions. Tools such as RCTD and SPOTlight facilitate accurate cell type identification in spatial datasets by leveraging scRNA-seq data for reference (269).

-Spatial Data Integration: Integrating scRNA-seq data with spatial transcriptomics enables comprehensive spatial characterization of gene expression patterns. Regression-based models and deep learning approaches are commonly used to reconstruct missing spatial features and enhance gene expression data (263).

Functional Analysis and Visualization: Tools like Seurat and Scanpy provide robust visualization capabilities for spatial transcriptomics data, while specialized platforms like Giotto and Squidpy allow for more detailed analyses of cellular interactions, spatial neighborhood graphs, and trajectory inference (264).

The integration of spatial transcriptomics with advanced bioinformatics tools has significantly enhanced our understanding of the spatial organization of TAMs within the TME. These methods enable the comprehensive analysis of spatial gene expression patterns, providing valuable insights into the architecture of tumor tissues and the interactions between TAMs and other cell types. These advancements are critical for identifying novel diagnostic markers and therapeutic targets, furthering our ability to design effective cancer therapies (264).

Spatial transcriptomics complements scRNA-seq by offering spatial context to gene expression data, allowing researchers to visualize TAM localization and their interactions with other cell types in situ. This approach has been utilized to map TAM heterogeneity across lung cancer subtypes, revealing distinct macrophage compositions that correlate with specific tumor characteristics (270). By preserving tissue architecture, spatial transcriptomics facilitates a comprehensive analysis of cellular communication and the organization of the TME, which is essential for elucidating the functional roles of TAMs in cancer progression. However, compared to scRNA-seq, spatial transcriptomics typically offers reduced sensitivity and lower coverage, particularly for detecting genes expressed at low levels, which can limit the depth of transcriptomic insights.

scRNA-seq has become an essential tool for studying cellular heterogeneity within tumors, enabling the characterization of distinct cell populations, including TAMs. By applying trajectory inference methods to scRNA-seq data, researchers can reconstruct the developmental pathways of individual cells based on their gene expression profiles, providing critical insights into cellular differentiation and function within the TME.

Pseudotime analysis is a widely used approach to order cells along a developmental trajectory, providing insights into their differentiation states. In the context of TAMs, pseudotime analysis has been employed to reveal the dynamic transitions of these cells as they interact with tumor cells and other components of the TME. Wang et al. elucidated the TAMs transition from pro-inflammatory to immunosuppressive phenotypes during breast cancer progression, demonstrating the utility of pseudotime analysis in understanding TAM functional changes over time (271).

Trajectory inference tools such as Monocle (272) and Slingshot (273) are commonly used to identify genes that are differentially expressed along inferred cellular trajectories. These techniques have proven valuable in uncovering key molecular pathways involved in TAM function and tumor progression. Yang et al. discovered that TAMs regulate BCSCs through a paracrine signaling pathway involving epidermal growth factor receptor (EGFR), STAT3, and SOX2, highlighting the relevance of trajectory analysis in elucidating cell-cell interactions within the TME (274).

Trajectory analysis aims to reconstruct the differentiation pathways and developmental trajectories of cells over time, offering a temporal perspective on how TAMs transition between distinct functional states. Applying trajectory inference to TAM scRNA-seq data has enabled researchers to discern how these macrophages evolve in response to tumor signals and alterations in the TME. Saelens et al. utilized trajectory analysis to gain insights into the temporal dynamics of TAM polarization during cancer progression, identifying key transitions from pro-inflammatory to immunosuppressive states (275).

The interactions between TAMs and CSCs are critical for driving tumor progression and metastasis. TAMs secrete various factors that enhance CSC properties, promoting tumor growth and resistance to therapy. Valdes-Mora et al. showed that TAM-derived cytokines, such as IL-6 and IL-10, help maintain the stemness of CSCs in breast cancer, underscoring the importance of TAM-CSC crosstalk in the TME (254).

In a glioblastoma study the authors used Monocle to trace the differentiation trajectories of TAMs, revealing changes in their functional states in response to tumor-derived signals (276). Furthermore, the integration of spatial transcriptomics with scRNA-seq allows for a more nuanced understanding of TAM interactions with other immune and tumor cells, providing a spatial and temporal view of TAM dynamics within the TME (277).

Advanced bioinformatics techniques have significantly enhanced the ability to identify critical signaling pathways that regulate the functions of TAMs and their impact on cancer progression. Jin et al. utilized spatial transcriptomics to uncover spatially regulated biomarkers and signaling pathways within TAM populations, providing valuable insights into their roles and functional states within the TME (264). These approaches have also elucidated specific signaling pathways that govern TAM polarization and their pro- or anti-tumoral activities. The CCL2-CCR2 signaling axis has been shown to play a pivotal role in recruiting and polarizing TAMs toward a pro-tumorigenic phenotype (278). Importantly, the inhibition of this pathway holds therapeutic potential by reprogramming TAMs toward an anti-tumor phenotype, thereby enhancing the effectiveness of cancer treatments (278). This underscores the significance of targeting TAM-specific pathways in therapeutic strategies aimed at modulating the TME.

TAMs influence a wide array of signaling pathways within the TME, directly interacting with tumor cells and other TME components to drive cancer progression. Pathways such as WNT, NOTCH, and TGF-beta, which are crucial for maintaining CSC properties and promoting tumor aggressiveness, are modulated by TAM activity (279). These pathways are critical for the survival and function of CSCs, further supporting the tumor’s growth and metastasis.

Recent studies have leveraged computational models to simulate the effects of TAMs on tumor growth, shedding light on the importance of cell-cell communication in the TME. Zhao et al. demonstrated that TAMs secrete cytokines and chemokines that enhance CSC survival and drive tumor progression (138). These computational insights highlight the complex interactions within the TME that are essential for tumor evolution. Emerging bioinformatics tools such as CellChat (https://github.com/sqjin/CellChat) or CellPhoneDB (280) have proven effective in analyzing intercellular communication networks, providing a deeper understanding of ligand-receptor interactions that regulate TAM and CSC dynamics. Through the application of such tools, researchers have been able to map intricate communication networks between TAMs and other cells in the TME. These findings offer opportunities to identify novel therapeutic targets aimed at reprogramming TAMs toward an anti-tumor phenotype, potentially improving cancer treatment outcomes (281). Overall, integrating bioinformatics approaches with experimental data has been pivotal in uncovering the signaling pathways that govern TAM activity. These insights offer potential strategies for therapeutic interventions aimed at altering TAM function and modulating the TME to halt tumor progression. An overview of the most important discoveries made about the CSCs-TAMs axis by advanced bioinformatic technologies in different cancer histotypes is summarized in Table 5 .