Rewiring the Tumor Microenvironment: IGFBP2 at the Nexus of Remodeling and Therapy Resistance?

Abstract

Insulin-like growth factor binding protein 2 (IGFBP2) has emerged as a multifarious and context-dependent oncoprotein that links several mechanisms in the tumor microenvironment (TME) including oncogenic signaling, extracellular matrix (ECM) remodeling, immune evasion, and therapy resistance. Beyond its established role in modulating IGF signaling, IGFBP2 exerts IGF-independent effects via its RGD integrin-binding motif and nuclear localization signal (NLS). By binding integrins (αvβ3, α5β1), IGFBP2 activates focal adhesion kinase (FAK), which then triggers PI3K/AKT signaling and MAPK/ERK signaling pathways, resulting in enhanced proliferation, migration, invasion, and angiogenesis. Nuclear translocation of IGFBP2, mediated by its NLS, enables direct regulation of gene expression, notably by upregulating epithelial-mesenchymal transition (EMT) transcription factors such as ZEB1, SNAI1, and TWIST1, and regulating immune checkpoint molecules. These actions reshape the TME by increasing angiogenesis, stromal stiffness, and tumor invasiveness. IGFBP2 further sustains survival signals under receptor tyrosine kinase inhibition, enhances tumor cell metabolic adaptation to hypoxia, and supports cancer stemness, all of which drive resistance to chemotherapy, radiotherapy, and immunotherapy. Overexpression of IGFBP2 in cancer is linked to increased tumor aggressiveness, unfavorable prognosis, and general resistance to therapy. Importantly, IGFBP2 functions are highly context-dependent, in some epithelial settings, IGFBP2 sequesters IGFs and dampens IGF-IR signaling, resulting suppression of downstream oncogenic pathways. This duality underscores IGFBP2’s nuance and tumor-specific biology. Therapeutic strategies under development specifically target IGFBP2-integrin-mediated signaling and IGFBP2 nuclear activity. There are encouraging results in preclinical studies involving antisense oligonucleotides, monoclonal antibodies, and peptide inhibitors. As both a mediator of oncogenic adaptation and a clinical biomarker, IGFBP2 represents a vulnerability in the TME that may be targeted to improve, personalize, and combine cancer therapies.

1. Introduction

The World Health Organization projects that, by 2030, 13.1 million patients will die of cancer, with 70% occurring in developing countries [1], implicating significant socio-economic and health care challenges. The tumor microenvironment (TME) is increasingly recognized as a potential critical mediator of poor prognosis in many cancers [2]. The TME is a multi-component environment consisting of tumor cells, blood vessels, extracellular matrix (ECM), stromal and fibroblast cells, signaling molecules, and immune cells [2]. The multiple constituents of the TME communicate with each other through an evolving complex circuitry either directly or via soluble factors, including chemokines (e.g., CXCL12), growth factors (e.g., fibroblast growth factors [FGFs]), cytokines, gases (e.g., H₂S and NO), lipids and other metabolites. These communications remodel the TME to support tumor growth, survival and therapy resistance [2–4]. Insulin-like growth factor binding protein 2 (IGFBP2) is a critical component of the TME and overexpression of it contributes to poor outcomes in many cancers [3–5], including pancreatic [6], ovarian [6], breast [6, 7], cervical [6, 7], melanoma [6, 8], glioma [6], leukemia [6], lung [9], liver [10], prostate [6], gastric [11], bladder [12], head and neck [6], colorectal adenoma and carcinoma [13], and synovial sarcoma [14].

IGFBP2 is part of the IGF axis, which includes IGF-I and IGF-II, their corresponding receptors (IGF IR and IGF-IIR), and six high-affinity IGF-binding proteins (IGFBPs 1–6) that regulate IGF actions [15]. IGFBP2 binds IGF-II, IGF-I, and insulin with decreasing affinity [6]. While best known for modulating IGF signaling, new findings highlight IGF-I-independent functions of IGFBP2 in cancer. The arginine-glycine-aspartic acid (RGD) motif mediates integrin-specific interactions (with αvβ3 and α5β1, triggering the focal adhesion kinase (FAK), SRC, phosphoinositide 3-kinase (PI3K)/AKT, and mitogen-activated protein kinase (MAPK/ERK) pathways [6]. These pathways drive cancer proliferation, migration, invasion, and angiogenesis. IGFBP2 also enters the nucleus via its NLS and interacts with transcriptional regulators to regulate gene expression. It upregulates epithelial-mesenchymal transition (EMT) transcription factors (ZEB1, SNAI1, TWIST1), promoting metastasis, and increases immune checkpoint proteins like programmed death-ligand 1 (PD-L1) via EGFR/STAT3, aiding immune evasion [8]. Multiple mechanisms regulate IGFBP2 expression, including transcription, epigenetics, and post-translational control. TME hypoxia activates hypoxia-inducible factor 1-alpha (HIF-1α) to upregulate IGFBP2, and inflammatory signals induce it through STAT3. Epigenetic promoter hypomethylation also elevates IGFBP2 levels in cancer [16].

Importantly, emerging evidence indicates that IGFBP2 functions are highly context-dependent and can even have opposing effects [7]. In some epithelial tissues; e.g. normal breast tissues, secreted IGFBP2 sequesters IGFs and restricts IGF-IR activation, thereby suppressing its classical oncogenic signaling, unlike in other malignancies, such as glioblastoma and prostate cancer where oncogenic effects predominate. This duality reflects the complex interplay between IGFBP2’s structural domains, proteolytic processing, receptor availability, and microenvironmental cues, underscoring the need to interpret IGFBP2 biology within specific tumor contexts. Given these key roles in oncogenic signaling and therapy resistance with potential implications for TME remodeling, IGFBP2 is emerging as a key marker for cancer diagnosis, prognosis, and therapy. This review discusses a possible link between the role of IGFBP2 in remodeling the TME and its potential as a therapeutic target for clinical translation.

2. Overview of IGFBPs

IGFBPs are evolutionarily conserved in vertebrates. In mammals, there are six types, IGFBP1 to IGFBP6 (Fig 2), thought to have been initially formed through two rounds of gene duplication. In contrast, teleost fish (e.g. salmon, zebrafish) possess two copies of each isoform type [17–19], while chickens have only four IGFPBs [19]. IGFBP1 and IGFBP2 both possess RGD binding domains used to execute IGF-I-independent functions. In addition, IGFBP2, unlike IGFBP1, possesses an NLS which facilitates nuclear regulation [6]. Increased expression of IGFBP2 correlates with poor outcomes in a broad range of aggressive tumors. Combined with the extent of its overexpression, its unique oncogenic domains, and its ability to activate multiple protumorigenic pathways, IGFBP2 is a more compelling target than IGFBP1 in cancer models.

Figure 2: Schematic representation of IGFBP family phylogenetic tree.

Successive gene duplication from a single ancestral IGFBP gene has resulted in the modern complement of IGFBPs. Out of eight IGFBPs, two were subsequently lost [19]. The N and C-termini are conserved in all IGFBPs, while the central region is variable across family members. This figure was created in BioRender.

3. Functional Domains of IGFBP2

IGFBP2 is a 36-kDa glycoprotein with a conserved N-terminal IGF-binding domain, a central protease-sensitive region which also contains an NLS, and a C-terminal region with a heparin-binding domain (HBD) and an RGD motif involved in integrin binding [16] (Fig 3).

Figure 3: Domain structure of IGFBP2.

Schematic of the IGFBP2 protein showing its domain architecture and key regulatory elements. IGFBP2 consists of an N-terminal domain (residues 1–98) that includes a high-affinity insulin-like growth factor (IGF)-binding site (gray), a central linker region (residues 99–182), and a C-terminal domain (residues 183–289) involved in ECM interactions. The central linker domain contains a nuclear localization signal (NLS, teal, residues 179–184), which also serves as heparin-binding domain 1 (HBD1), enabling nuclear import and interactions with glycosaminoglycans. The C-terminal domain includes a second heparin-binding domain (HBD2, yellow), enhancing binding to heparan sulfate proteoglycans in the TME. The RGD tripeptide (red, residues 265–267) mediates integrin binding and downstream activation of intracellular signaling cascades such as integrin-linked kinase (ILK)/FAK/AKT. A final highly conserved region (residues 227–244) is rich in lysine and arginine, contributing to both heparin binding and interactions with cell-surface proteoglycans. These motifs enable IGFBP2 pleiotropic function in tumor progression, immune modulation, and angiogenesis [6]. Illustration created in BioRender.

3.1. N-terminal Domain

The N-terminal region of IGFBP2 comprises a conserved, cysteine-rich motif forming six intradomain disulfide bonds, which create a rigid pocket for high-affinity binding of IGF-I and IGF-II. This domain (residues ~1–98) is indispensable for modulating IGF ligand bioavailability and fine-tuning IGF IR activation, thereby affecting proliferation, survival, and differentiation [6].

3.2. Central Linker (Hinge) Region

Spanning residues ~99–183, the central linker or hinge region of IGFBP2 is unstructured and highly sensitive to proteolytic cleavage by matrix metalloproteinases (MMPs) and serine proteases, yielding fragments with modified IGF affinity and novel bioactivities [20]. Cleavage can abrogate IGF binding but exposes cryptic motifs that engage ECM proteins, such as vitronectin and fibronectin, thereby enhancing integrin-mediated adhesion and directional migration independently of IGF [6]. Crucially, this region contains an NLS (residues ~179–184), which permits both full-length and cleaved IGFBP2 fragments to translocate into the nucleus [21]. This cluster is highly basic PKKLRP (Fig. 3, teal; residues 179–184), which enhances nuclear localization [22]. IGFBP2 has a mono-partite NLS sequence (X-K-K/R-X-K/R-X), unlike IGFBP3, 5, and 6, which possess bi-partite NLS sequence [22–24]. IGFBP2’s NLS permits nuclear import via importin-α/β [24]. Nuclear IGFBP2 interacts with STAT3 and β-catenin to regulate gene networks associated with stemness, therapy resistance, and plasticity, particularly in glioblastoma and hematologic malignancies [25].

3.3. C-terminal Domain

The C-terminal domain (residues ~184–289) contains two critical motifs: HBD2 (Fig 3, yellow; residues ~227–244) and the RGD integrin-binding motif (Fig 3, red; residues ~265–267). HBD2 targets IGFBP2 to cell-surface glycosaminoglycans and ECM components, creating spatial concentration hubs that scaffold integrin clustering [26]. The adjacent RGD motif mediates direct binding to integrins including α5β1 and αvβ3, triggering downstream FAK, PI3K/AKT, and MAPK/ERK signaling independent of IGF1R [26]. Engagement of integrins by IGFBP2 promotes EMT, invasiveness, angiogenesis, and drug resistance in multiple models [16]. Notably, proteolytic unmasking of the RGD motif functions as a switch that regulates integrin recruitment. Emerging studies also suggest that phosphorylation or glycosylation within the C-terminal region can fine-tune integrin affinity and downstream signaling specificity, though detailed mechanisms remain to be elucidated [26].

Taken together, the modular architecture of IGFBP2, including its IGF-binding N-terminus, protease-sensitive hinge with nuclear translocation capacity, and integrin-engaging C-terminus, enables highly context-dependent oncogenic functions. The relative contribution of each domain varies by cancer type, proteolytic milieu, and tumor progression stage. In subsequent sections, we will dissect how domain-specific interactions drive malignant phenotypes such as stemness, invasion, and immune evasion, with emphasis on IGFBP2’s ability to modulate the TME.

4. The Tumor Microenvironment: Composition and Components

The TME represents a complex, dynamic, and heterogeneous ecosystem intricately interwoven with a tumor, serving as a critical orchestrator of tumor progression, survival, and resistance to various therapies. It comprises not only cancer cells but a wide variety other cell such as stromal cells, immune cells, and endothelial cells, along with ECM proteins, soluble factors, and metabolites (Fig 4) [27]. The TME exhibits significant variations among different cancer types and within distinct regions of a single tumor. These variations in the TME profoundly affect tumor behavior, including growth, immune evasion, metastasis, and response to therapy [27].

Figure 4: Cellular and molecular crosstalk in the TME.

There is complex interplay between cancer cells and stromal/immune components of the TME. Left panel describes the components of TME, blood flow, nutrients, oxygen gradients, growth factors, secreted molecules, and other components of the TME. The right panel describes different tumor progression processes in the TME. Cancer epithelial cells undergo EMT under the influence of pro-inflammatory cytokines (IL1β, IL6, TNFα), generating migratory mesenchymal-like tumor cells. Upon TGFβ signaling, fibroblasts differentiate into CAFs, which remodel the ECM and support tumor growth. Tumor-derived VEGF drives angiogenesis, resulting in abnormal and leaky vasculature that supplies oxygen and nutrients but also facilitates tumor cell intravasation. Monocytes recruited to the tumor milieu polarize into M2 macrophages in response to IL10, TGFβ, and other cytokines, promoting an immunosuppressive environment, while M1 macrophages are less abundant. Tregs suppress CD8+ cytotoxic T cell activity by secreting IL10, TGFβ, M-CSF, and IL35, thereby dampening antitumor immunity. Collectively, these dynamic cellular and molecular interactions contribute to tumor progression, immune evasion, and metastasis [27, 42, 43]. Illustration created in BioRender.

4.1. Cancer Cells

Tumor cells actively modulate their environment by secreting soluble factors such as cytokines, growth factors, and extracellular enzymes, which help them invade, proliferate, and metastasize. Moreover, cancer cells often alter TME to suppress immune responses, enhance angiogenesis, and increase their metastatic potential. The interactions between cancer cells and other components of the TME can result in feedback loops that continuously promote tumor progression and resistance to therapy [27].

4.2. Stromal Cells

Stromal cells play a pivotal role in shaping TME by providing structural support and modulating cellular behaviors within the tumor. These cells include fibroblasts, endothelial cells, and myofibroblasts, among others. Normal fibroblasts produce and remodel the ECM, but in the TME they are often transformed into cancer associated fibroblasts (CAFs). CAFs are activated by tumor cell signals, including TGF-β and FGFs, which induce them to secrete ECM components and matrix-modifying enzymes [28]. CAFs promote tumor progression through multiple mechanisms including supporting cell proliferation, inducing angiogenesis, and facilitating invasion by remodeling the ECM. They significantly contribute to immune suppression, initiating the process by releasing immunosuppressive cytokines such as interleukin 6 (IL6) and TGF-β, effectively hindering the capacity of the immune system to combat the tumor [28]. CAFs are heterogenous, with distinct phenotypes and characteristics. They are generally classified as myofibroblast, inflammatory, senescent or metabolic CAFs [3], and CAFs have extraordinary functional plasticity across different tumors. For example, in breast cancer tumors, CAF activity in tumor-immune crosstalk inhibits natural killer (NK) cell infiltration and cytotoxic killing through collagen deposition [29]. In contrast, in gastric cancer, antigen-presenting CAFs and macrophages support anti-tumor T cell responses [30].

Endothelial cells form the inner lining of the blood vessels and are central to the process of tumor angiogenesis, a critical step for tumor growth. Tumor angiogenesis provides the tumor with sufficient blood supply to support its increasing metabolic demands and is driven by pro-angiogenic factors including vascular endothelial growth factor (VEGF) and FGFs [31]. In addition, endothelial cells secrete factors that influence tumor progression, invasion, and metastasis [31].

The TME is typically infiltrated by a wide array of immune cells that can have either pro- or anti-tumor effects, depending on their phenotype, activation state, and the local microenvironmental cues (further explored in [3]). Tumor-associated macrophages (TAMs) are one of the most abundant immune cell populations in the TME. There are many subtypes of TAM based on their different origins, (yolk sac-derived tissue-resident macrophages or bone marrow-derived infiltrating macrophages) that can coexist within the same tumors. TAMs typically exhibit an immunosuppressive and pro-tumorigenic phenotype [27] that promotes tumor growth by secreting cytokines (e.g., IL10, TGF-β), matrix-degrading enzymes, and angiogenic factors that enhance tumor cell proliferation, invasion, and metastasis [3, 31]. TAMs can also suppress the activation of cytotoxic T cells, thereby aiding in immune evasion [32].

The TME can contain both cytotoxic T cells (CD8⁺) and helper T cells (CD4⁺), but their functions are often impaired in cancer. Regulatory T cells (Tregs) are frequently enriched in tumors and help maintain an immunosuppressive environment, preventing effective anti-tumor immunity [27]. Additionally, T cells may be functionally exhausted due to chronic exposure to inhibitory signals like PD-1/PD-L1 interactions within the TME [33]. Myeloid-derived suppressor cells (MDSCs) are a heterogeneous population of immune cells that include immature myeloid cells. These cells accumulate in tumors and suppress the activity of T cells and NK cells, thus facilitating immune evasion [27]. MDSCs also promote tumor progression by secreting immunosuppressive cytokines (e.g., IL10, TGF-β) and are a major barrier to effective cancer immunotherapy [29].

4.3. The Extracellular Matrix

The ECM is a complex network of proteins and glycoproteins that provide structural scaffolding to the tissue. In the TME, the ECM not only supports the physical structure of the tumor but also plays a crucial role in regulating cell behavior through biochemical and mechanical signals. Tumor cells frequently alter the composition and organization of the ECM to facilitate their own survival and invasion [34]. Tumor cells secrete MMPs, which degrade ECM components such as collagen, fibronectin, and laminin. This degradation allows tumor cells to invade surrounding tissues and promotes the release of bioactive fragments that can drive further tumor progression. Additionally, ECM stiffness has been shown to influence cell signaling pathways related to migration, survival, and metastasis, with increased stiffness correlating with more aggressive tumor phenotypes [35]. The ECM also influences mechanotransduction, the process by which cells convert mechanical stimuli into biochemical signals. Tumor cells sense and respond to changes in ECM rigidity, which impacts their ability to migrate, proliferate, and metastasize [35]. For example, integrin-mediated adhesion to ECM components such as collagen and fibronectin activates downstream signaling pathways, including those involving FAK, which promotes survival and motility [36, 37].

4.4.1. Soluble Factors

The TME is rich in soluble factors such as cytokines, growth factors, and metabolites, which are critical for regulating the interactions between tumor and stromal cells. Many growth factors are overexpressed in the TME and contribute to tumor progression. VEGF is perhaps the most well-known pro-angiogenic factor, promoting the formation of new blood vessels and improving nutrient and oxygen delivery to the tumor [38]. TGF-β is another major factor in the TME that induces ECM remodeling and immune suppression. Elevated TGF-β levels are associated with the activation of CAFs and the induction of an immunosuppressive environment [10].

Cytokines are pivotal for communication between immune cells and tumor cells in the TME. They can either promote or inhibit inflammation, immune cell infiltration, and tumor growth. For instance, IL6 is a key pro-inflammatory cytokine that promotes tumor progression by enhancing angiogenesis and immune suppression [39]. Conversely, interferons (IFNs) can exert anti-tumor effects by stimulating immune responses and inhibiting tumor cell proliferation [40].

The metabolic environment of the TME is often characterized by hypoxia and acidic conditions resulting from high rates of glycolysis (the Warburg effect). These metabolic changes can affect both tumor cells and stromal cells, influencing their behavior and interactions. For example, HIFs activate genes involved in angiogenesis and glycolysis, thus helping tumors adapt to low-oxygen conditions [41].

5. IGF-Dependent Functions of IGFBP2

IGFBP2, traditionally viewed as a high-affinity carrier of IGF-I and IGF-II, has emerged as a critical regulator of the IGF signaling axis, influencing tumorigenesis through modulation of ligand bioavailability, receptor engagement, and downstream signaling outputs [44] (Fig 5A–B). This regulation is primarily orchestrated through its highly conserved N-terminal domain, which binds IGFs and fine-tunes their spatial and temporal activity across multiple malignancies [6]. By sequestering IGF-I and IGF-II in the extracellular milieu, IGFBP2 protects these ligands from proteolytic degradation, thereby extending their half-life and stabilizing their signaling capacity [45]. This function enables IGFBP2 to modulate the concentration of free IGFs capable of engaging in the IGF-IR. In glioblastoma, elevated IGFBP2 expression promotes pericellular IGF retention, leading to sustained IGF-IR activation that drives proliferation and invasiveness [45]. Conversely, in certain epithelial tumor types, IGFBP2 overabundance may restrict IGF-IR access to its ligands, paradoxically attenuating mitogenic signaling [44, 45]. Adding complexity, evidence suggests that circulating IGFBP2 (e.g. from the blood) may have different effects on glioblastoma cells compared to locally produced IGFBP2 [44, 46].

Figure 5: Different modes of IGFBP action.

(A) Sequestering IGF from its IGF-IR. (B) Facilitating IGF signaling via protease cleavage of the IGFBP-IGF releasing IGF for binding with IGF-IR. (C) IGF-independent function involving integrin binding and subsequent signaling or nuclear translocation of intracellular IGFPB followed by regulation of gene expression[19, 21]. Illustration created in BioRender.

Through its control of ligand availability, IGFBP2 exerts indirect regulation over canonical oncogenic signaling cascades downstream of IGF-IR, notably the PI3K/AKT and MAPK/ERK pathways [47]. In prostate cancer, IGFBP2 overexpression correlates with enhanced AKT phosphorylation and resistance to androgen deprivation therapies [8, 48]. Similarly, in glioblastoma, IGFBP2 potentiates PI3K/AKT signaling and promotes resistance to temozolomide, linking IGFBP2-driven IGF signaling to chemoresistance mechanisms [46, 49]. Though less frequently emphasized, MAPK/ERK activation downstream of IGFBP2-facilitated IGF-IR signaling has also been implicated in cell survival and migration, particularly in melanoma and colorectal cancer models [50].

The oncogenic consequences of IGFBP2 interactions with the IGF axis are highly context-dependent, influenced by tumor type, receptor expression levels, and microenvironmental conditions [6]. In glioblastoma and prostate cancer, IGFBP2 predominantly enhances proliferative and survival signaling, while in other cellular contexts, such as differentiated epithelial tissues, it may sequester IGFs away from IGF-IR, thereby dampening downstream signaling and growth [44]. This duality highlights the complexity of the role of IGFBP2 and the necessity of understanding its activity within specific cellular milieus [16, 51].

In sum, IGFBP2 regulation of IGF ligand dynamics and IGF-IR activity constitutes a foundational mechanism in its oncogenic repertoire. However, a growing body of evidence indicates that IGFBP2 also exerts potent IGF-independent effects. These non-canonical functions, mediated via its HBDs, RGD domain, and NLS, allow IGFBP2 to interact with integrins, remodel the ECM, regulate transcriptional responses, and engage components of the TME [52]. Dissecting these IGF-independent roles is essential, as they often act in parallel with, or independently of, IGF signaling, particularly in invasive, immune-evasive, and drug-resistant cancer phenotypes [53].

6. IGF-Independent Functions of IGFBP2

Beyond its classical role in regulating IGF bioavailability, IGFBP2 exerts several IGF-independent functions that significantly impact cancer progression (Fig 5C). These effects are mediated through structurally conserved motifs such as the C-terminal RGD domain, the centrally located NLS, and the HBD, enabling IGFBP2 to directly engage integrins, translocate into the nucleus, and modulate ECM dynamics. Through these non-canonical functions, IGFBP2 orchestrates oncogenic signaling, promotes EMT, and facilitates immune escape and therapeutic resistance.

The RGD motif in IGFBP2 confers the ability to bind integrins such as αvβ3 and α5β1, initiating downstream signaling cascades independent of IGF-I receptor engagement [44]. This interaction activates FAK and SRC, which in turn stimulate PI3K/AKT and MAPK/ERK pathways, promoting survival, migration, and invasion across several cancer models [6]. In glioblastoma, researchers showed that IGFBP2-integrin engagement facilitates FAK phosphorylation, enhancing tumor cell motility and ECM invasion by using splice acceptor/tv-a glioma mouse system [45]. In melanoma, this axis supports AKT-dependent resistance to anoikis, thereby sustaining metastatic viability [6, 54]. Moreover, in an orthotopic model of pancreatic ductal adenocarcinoma (PDAC), IGFBP2-integrin signaling potentiates EMT, contributing to enhanced plasticity and chemoresistance [6]. Importantly, these interactions circumvent growth factor dependence and sustain tumor aggressiveness even under nutrient-limited conditions, further underscoring the autonomy of IGFBP2-driven integrin signaling in shaping the malignant phenotype.

The IGFBP2 NLS is functional and facilitates its nuclear translocation in multiple tumor types [24]. Once in the nucleus, IGFBP2 modulates oncogenic transcriptional programs by interacting with chromatin remodelers and transcription factors [55]. In glioblastoma and acute myeloid leukemia (AML), nuclear IGFBP2 expression correlates with the upregulation of EMT-associated transcription factors such as TWIST1, ZEB1, and SNAI1, promoting migratory phenotypes and resistance to differentiation [54]. Additionally, nuclear IGFBP2 participates in immune evasion by regulating PD-L1 and cytokine expression, potentially via STAT3 and NF-κB signaling axes [8]. This nuclear functionality contributes to a more plastic, immune-silent, and therapy-resistant tumor state. Such transcriptional reprogramming further establishes IGFBP2 as a critical effector of malignancy, independent of IGF signaling.

Moreover, the C-terminal HBD anchors IGFBP2 to heparin-sulphate proteoglycans and ECM components, facilitating its localization and enhancing integrin-mediated signaling [22, 56]. These IGF-independent pathways have been associated with increased MMP2 and MMP9 expressions, ECM remodeling, Cancer invasion and angiogenesis in several cancer models [45, 57]. Azar et al. demonstrate that overexpression of IGFBP2 in SHEP-BP-2 cells increases the VEGF expression, which resulting in the induction of tumor angiogenesis [58]. Collectively, these IGF-independent mechanisms underscore the role of IGFBP2 as an ECM-interactive signaling platform. Its domain structure enables IGFBP2 to sculpt the TME, support cell invasion, foster immune evasion, and help tumors survive under therapeutic stress, all in the absence of IGF ligands [22, 59].

7. Domain-Specific Roles of IGFBP2 Across Cancer Types

The structural complexity of IGFBP2 underpins its diverse functional roles in tumor progression. The functional reliance on discrete IGFBP2 domains varies among malignancies, reflecting tumor-specific signaling contexts and microenvironmental interactions [6]. Here, we briefly examine how distinct structural elements of IGFBP2 are differentially exploited to drive oncogenesis in various cancer models (summarized in Table 1).

Table 1.

Domain-Specific Mechanisms of IGFBP2 Across Cancer Types

Cancer Type	Key IGFBP2 Domains Involved	Principal Molecular Interactions	Downstream Signaling Pathways	Functional Consequences
Prostate Cancer [6, 60, 61]	N-terminal (IGF-binding), C-terminal (RGD, HBD)	IGF-I/II binding; αvβ3–integrin engagement	IGF-IR→AKT/MAPK; FAK/SRC/NF-κB	Proliferation, ADT resistance, motility
Lung Cancer (NSCLC) [42, 62]	N-terminal, C-terminal (RGD, HBD)	IGF-I/II stabilization; α5β1/αvβ3 integrins	PI3K/AKT, MAPK/ERK, HIF-1α	Hypoxia adaptation, EMT, metastasis
Breast Cancer (TNBC) [6, 67]	C-terminal (RGD, HBD)	Integrin αvβ3/α5β1, TGF-β pathway	FAK/SRC → SMAD2/3 → TGF-β; STAT3	ECM remodeling, EMT, therapy resistance
Colorectal Cancer [68, 72]	Central linker (proteolysis), C-terminal (RGD)	MMP-mediated cleavage; integrin α5β1 binding	FAK/ERK; β-catenin/WNT crosstalk	Invasion, stemness, metastasis
Glioblastoma [45, 55]	C-terminal (RGD, HBD), Central (NLS)	Integrin α5β1/αvβ3 binding, STAT3 interaction, nuclear translocation	FAK → PI3K/AKT → ERK; STAT3 activation	Invasion, hypoxia tolerance, therapy resistance, immune evasion
Melanoma [46, 75]	Central linker (proteolysis), C-terminal (RGD)	ECM interaction; fibroblast-derived IGFBP2	PI3K/AKT, lipid metabolism	Invasion, metastasis (age-related)
Leukemia (AML) [73]	Central (NLS)	Nuclear import via importin-α; STAT3/β-catenin	AKT/PTEN; transcriptional regulation	Stemness, chemoresistance

7.1. Prostate Cancer: Stage-Shift from IGF Modulation to Integrin-Driven Motility

Early prostate tumorigenesis relies on the N-terminal domain of IGFBP2, which stabilizes IGF-I/2 and enhances IGF-IR–PI3K/AKT and MAPK signaling to promote epithelial proliferation and survival [6, 60]. With cancer progression, the C-terminal HBD and RGD motif engage HSPGs and αvβ3 integrins [6], activating ILK/FAK signaling and cytoskeletal remodeling to drive invasion [60]. Notably, IGFBP2-integrin engagement also contributes to the development of resistance to the androgen deprivation therapy (ADT) by activating alternative survival pathways, such as SRC and NF-κB, which allow tumor cells to proliferate under castration conditions [61]. This functional shift from IGF-IR–dependent growth to integrin-mediated motility and therapy resistance mirrors the clinical transition from androgen-sensitive prostate cancer to castration-resistant prostate cancer, highlighting the dynamic and stage-specific utility of IGFBP2’s structural domains in driving prostate cancer progression [61].

7.2. Lung Cancer: Hypoxia-Induced IGF Stabilization and Integrin Activation

In non-small cell lung cancer (NSCLC), HIFs (e.g., HIF-1α) upregulate IGFBP2, enabling its N-terminal domain to maintain IGF ligand stability and promote IGF-IR-driven PI3K/AKT and ERK survival signaling [62]. Concurrently, the C-terminal RGD/HBD region activates αvβ3 and α5β1 integrins under hypoxia, stimulating FAK/SRC and PI3K/AKT pathways that promote EMT, motility, and invasion [62]. IGFBP2-driven integrin activation enables migration along CAF-aligned collagen fibers, supporting dissemination of cancer cells to the brain, liver, and bone [42]. Together, the functional duality of IGFBP2-IGF-IR activation via its N-terminal domain and integrin-mediated migration via its C-terminal region facilitate both survival in hypoxic niches and metastatic dissemination to common metastatic targets in advanced NSCLC.

7.3. Breast Cancer: ECM Engagement and Therapy Resistance

IGFBP2 overexpression has been correlated with poor prognosis and aggressive phenotypes in breast cancer [6]. IGFBP2 binds to α5β1 integrins through its RGD motif, activating FAK and downstream PI3K/AKT signaling, which promotes breast cancer cell proliferation, survival, migration, and invasion [63–65]. In addition, Perks C.M et al demonstrated that, IGFBP2 can suppress the tumor suppressor PTEN, when it is free from IGF-2 [66]. Moreover, recent evidence demonstrates that IGFBP2 promotes resistance to chemotherapy by enhancing DNA-PKcs-mediated DNA double-strand break repair, reducing DNA damage accumulation and increasing tumor cell survival following cytotoxic treatment [67]. Through these IGF-dependent and IGF-independent mechanisms, IGFBP2 contributes to breast cancer aggressiveness, metastasis, and therapeutic resistance, highlighting its potential value as both a prognostic biomarker and therapeutic target. IGFBP2-integrin engagement is also a plausible driver of latent TGF-β activation and SMAD2/3 phosphorylation [43], potentially contributing to a TGF-β/IL10–rich, Treg-recruiting, TIL(low tumor-infiltrating lymphocyte (TIL)) )-suppressive microenvironment, though direct evidence in breast cancer remains to be elucidated.

7.4. Colorectal Cancer: Linker Region Proteolysis and WNT/Integrin Crosstalk

Colorectal cancer exhibits high IGFBP2 levels, characterized by proteolytic cleavage of its linker region by MMPs and serine proteases generating C-terminal fragments that retain the RGD motif [68]. These fragments bind fibronectin/laminin and activate α5β1/αvβ3 integrins, triggering FAK/SRC-PI3K/AKT and ERK signaling to promote cytoskeletal remodeling, EMT, and invasion [69]. Although IGFBP2-integrin activation is known to interface with WNT/β-catenin signaling and IGFBP2 elevates β-catenin in other cancers, direct colorectal cancer evidence for β-catenin stabilization or induction of c-MYC, cyclin D1, or LGR5 by IGFBP2 remains unconfirmed [6, 70].

7.5. Glioblastoma: Integrin Signaling Coupled to Nuclear Reprogramming

In glioblastoma, IGFBP2 is persistently elevated and largely drives malignancy via C-terminal, IGF-independent mechanisms. Its RGD motif activates α5β1–FAK–PI3K/AKT signaling to support survival, hypoxia tolerance, invasion, and therapy resistance [6, 46]. The RGD motif within this region directly interacts with integrin α5β1, leading to activation of FAK and downstream PI3K/AKT signaling. This integrin-mediated cascade promotes cellular survival under hypoxic conditions, supports invasion through brain parenchyma, and contributes to resistance against standard chemoradiotherapy regimens [46]. Beyond its extracellular functions, IGFBP2 also exerts oncogenic effects within the nucleus. The NLS within the linker region mediates nuclear import, where IGFBP2 interacts with transcription factors such as STAT3 to activate programs promoting stemness, immune evasion, and metabolic adaptation [45, 55].

7.6. Melanoma: Age-Driven IGFBP2 Secretion and Metabolic Invasion

Melanoma progression is strongly linked to elevated IGFBP2, particularly in aged microenvironments. Aged fibroblasts secrete high levels of IGFBP2, promoting melanoma invasion, lipid biosynthesis, and metastasis. Neutralization of IGFBP2 suppresses these phenotypes in vivo [71]. IGFBP2 activates PI3K/AKT signaling and metabolic reprogramming, and while integrin-mediated ECM interactions and invadopodium regulation are plausible, direct evidence in melanoma is still limited [72]. Collectively, these findings establish IGFBP2 as a critical factor linking the TME to melanoma aggressiveness, while also pointing to the need for further studies on its potential roles in ECM remodeling and invasion-associated structures.

7.7. Leukemia: Nuclear IGFBP2 and Stemness-Linked Resistance

In AML, IGFBP2 levels correlate with chemoresistance and adverse transcriptional signatures [73]. Genetic depletion of IGFBP2 in an AML mouse model increases apoptosis, reduces dissemination, elevates PTEN, and suppresses AKT signaling[74]. While nuclear IGFBP2 may plausibly regulate stemness and resistance in AML, direct evidence remains lacking [24, 58]. Its PTEN/AKT-linked survival and migration functions nonetheless position IGFBP2 as a strong therapeutic target.

8. IGFBP2 in Tumor Microenvironmental Interactions

8.1. ECM Remodeling & Matrix Stiffness

Through the RGD motif and HBDs, IGFBP2 directly interacts with ECM proteins such as vitronectin and fibronectin, and with integrins [24, 58]. These interactions stimulate integrin clustering, activating FAK, SRC, PI3K/AKT, and ERK/MAPK pathways, central to cellular mechano-transduction and malignancy progression [76]. In glioblastoma, overexpression of IGFBP2 enhances ERK phosphorylation and cell invasiveness via integrin β1 engagement. Knockdown of IGFBP2 or inhibition of β1 integrin significantly reduced ERK activation and invasive behavior in U87 and U251 cell lines [46]. Similarly, in NSCLC, IGFBP2-driven tumor cells show increased FAK phosphorylation and resistance to dasatinib, a dual SRC/FAK inhibitor, reinforcing the importance of integrin-mediated mechanotransduction in IGFBP2-driven tumor aggressiveness [77]. In prostate cancer, IGFBP2 expression is significantly elevated in aggressive tumors and positively correlates with integrin expression and ECM remodeling. Mechanistically, IGFBP2 enhances cell motility and invasion by promoting integrin β1–mediated FAK/ERK activation, contributing to EMT [76]. In glioblastoma, IGFBP2activated FAK/SRC signaling promotes perivascular invasion by reorganizing laminin-rich matrices and enhancing MMP activity [78].

In PDAC, a highly desmoplastic cancer, IGFBP2 is significantly overexpressed in both tumor cells and activated stromal fibroblasts, especially CAFs [79] . IGFBP2 binds integrin α5β1 on pancreatic stellate cells (PSCs), triggering PI3K/Akt/IKKβ activation and subsequent NF-κB–mediated transcriptional changes that promote the myofibroblastic conversion of PSCs into α-smooth muscle actin (α-SMA)-positive CAFs [79]. These CAFs secrete abundant ECM components, particularly collagen I and fibronectin, and express MMP2, MMP9 and tissue inhibitors (TIMPs), which coordinate ECM deposition and remodeling [80]. CAFs also enhance matrix stiffness through lysyl oxidase–driven cross-linking and mechanical alignment of ECM fibers via α5β1-focal adhesion complexes [80, 81]. These events create a mechanically anisotropic environment that promotes directional tumor cell invasion through enhanced mechanical signaling (e.g., FAK, SRC, ERK) in both tumor and stromal compartments. Additionally, CAFs within this stiffened ECM undergo autophagy via αV-integrin–dependent AMPK activation, providing metabolic support to cancer cells [82, 83]. In PDAC and nasopharyngeal carcinoma, IGFBP2-driven stromal signaling further contributes to an immunosuppressive niche through the secretion of TGF-β, IL10, and other STAT3-regulated cytokines, reinforcing fibrogenesis and tumor progression [52, 84–86].

IGFBP2 promotes EMT through activation of the FAK/SRC signaling axis, largely via integrin β1 and αVβ3 engagement. Under 3D culture conditions in low-grade glioma or within stiff extracellular matrices, IGFBP2 expression is further upregulated, and its integrin-binding RGD motif becomes increasingly relevant for anchorage-dependent signaling. These conditions can lead to phosphorylation of FAK (Y397) and downstream SRC activation, which cooperatively enhance ERK and STAT3 signaling, promoting mesenchymal phenotypes characterized by N-cadherin upregulation and E-cadherin loss [6, 87]. Additionally, IGFBP2 stimulates the transcription of matrix-remodeling genes, including LOX, MMP2 and TIMP1, fostering a pro-invasive ECM that reinforces the mesenchymal state [87]. It also sustains nuclear translocation of YAP/TAZ, via EGFR/STAT3 pathway, a hallmark of mechanical activation in high-stiffness microenvironments, thereby linking mechanotransduction to transcriptional control of stemness and survival pathways [88]. These findings suggest that IGFBP2 can contribute to therapy resistance by remodeling TME through its ability to modulate integrin-mediated mechano-transduction and survival signaling under conditions of ECM stiffness or anchorage independence. Although direct measurements of IGFBP2-mediated matrix stiffening remain limited, its consistent activation of integrin-linked mechano-signaling suggests it has the potential to prime tumor cells for stiffness-induced aggressiveness. This is particularly relevant in tumors with desmoplastic stroma such as pancreatic, breast, and ovarian cancers, where ECM rigidity fosters malignancy through integrin β1 and α5β1–dependent FAK/SRC signaling [89, 90]. Beyond its probable role in matrix remodeling and stiffness-induced signaling, IGFBP2 also contributes to the initiation and amplification of fibrotic responses in the TME by driving fibroblast activation and differentiation, hallmarks of desmoplastic cancers [91], but its direct contribution to matrix architecture and stiffness in human tumors remains to be fully explored (summarized in Table 2).

Table 2.

IGFBP2-Mediated Modulation of Tumor Microenvironment Components

TME Component	Mechanistic Role of IGFBP2	Key Pathways Involved	Representative Cancer Models	Functional Outcome
Cancer-Associated Fibroblasts (CAFs) [6]	Promotes fibroblast-to-myofibroblast transition (FMT), collagen deposition	Integrin α5β1 → PI3K/AKT → NF-κB; TGF-β/SMAD2/3	PDAC, Breast, colorectal cancer, Ovarian	ECM stiffening[78], fibrosis, immune exclusion
Endothelial Cells / Angiogenesis [125]	Binds αvβ3/α5β1 integrins on endothelium; enhances VEGFA transcription (nuclear role)	FAK → AKT → HIF-1α → VEGF; ERK	glioblastoma, PDAC, HCC, Ovarian	Angiogenesis, vasculogenic mimicry
Immune Cells (TAMs, MDSCs, Tregs) [126]	Induces STAT3-dependent IL10, PD-L1, IDO expression	IGFBP2–Integrin → STAT3/EGFR/AKT	PDAC, glioblastoma, Melanoma, Prostate	M2 polarization, T cell suppression, immune evasion
ECM and Mechano-transduction [79, 80, 127]	Enhances integrin clustering, FAK activation, MMP secretion	FAK/SRC/ERK/YAP	Ovarian, Breast, PDAC	ECM remodeling, stiffness, invasion
Hypoxia and Metabolism [103, 104]	HIF-1α–induced expression under low O2; sustains glycolysis	HIF-1α, AKT, EGFR/STAT3	NSCLC, glioblastoma, Melanoma	Survival under hypoxia, metabolic adaptation
Cancer Stemness [24, 128]	Nuclear activation of stemness transcription factors	STAT3, β-catenin, SOX2, OLIG2 (hypothetical)	glioblastoma, AML, ovarian	Stem cell maintenance, therapy resistance

8.2. Fibroblast-to-Myofibroblast Transition & Fibrosis Induction

TGF-β signaling and integrin-mediated mechanical activation; two tightly interconnected pathways in fibrotic TMEs, play a pivotal role in fibroblast to myofibroblast transition (FMT) (Fig 6) [89]. TGF-β is secreted as a latent complex (bound to latency associated protein [LAP]) and requires activation through integrins such as α5β1, αvβ3, and αvβ6, which bind to the RGD motif of LAP and apply cytoskeletal tension or recruit MMPs to release bioactive TGF-β [92]. IGFBP2, with its own RGD motif and heparin-binding domains, can localize to these integrin-rich zones of matrix tension and may facilitate the mechanical activation of latent TGF-β, either directly or by enhancing integrin clustering and signal transduction. Once active, TGF-β initiates the differentiation of fibroblasts into α-SMA-positive myofibroblasts, a process central to desmoplasia and stromal stiffening. This transformation contributes significantly to the remodeling of the TME, enhancing its rigidity and potentially impacting tumor progression and therapeutic responses. This differentiation is accompanied by increased synthesis and deposition of collagen I, fibronectin, tenascin-C, and ECM-remodeling integrins (e.g., α5β1, αvβ3, α11β1), creating a feed-forward loop of matrix remodeling and mechano-signaling [93]. Notably, IGFBP2 is upregulated in activated cancer-associated fibroblasts (CAFs) and may amplify this loop by supporting autocrine and paracrine TGF-β signaling, sustaining ECM protein production, and stabilizing IGF-IR or TGF-β receptor engagement on CAF surfaces [79]. Moreover, CAF-derived IGFBP2 may enhance the transcription of pro-fibrotic genes through FAK/ERK/YAP signaling, mimicking or reinforcing the effects of matrix stiffness and TGF-β exposure [89]. This multi-level engagement positions IGFBP2 not just as a passive biomarker of fibrosis but possibly as an active mediator of myofibroblast induction, stromal remodeling, and fibrotic tumor evolution, though further exploration is necessary.

Figure 6. Role of IGFBP2 in cancer-associated fibroblast (CAF)-mediated stromal remodeling.

Schematic representation of the mechanisms by which IGFBP2 regulates cancer-associated fibroblasts (CAFs) and promotes desmoplastic tumor microenvironments. Abbreviations: ECM, extracellular matrix; NF-κB, nuclear factor kappa B; TGF-β, transforming growth factor-beta; PDAC, pancreatic ductal adenocarcinoma. Illustration created in BioRender.

Fibrotic and mechanical mechanisms are particularly prominent in pancreatic, breast, colorectal cancers, and ovarian cancers, where dense desmoplastic reactions are not merely passive histologic features but active contributors to tumor progression, immune exclusion, and therapeutic resistance [42].

In PDAC, for example, cancer-associated fibroblasts (CAFs) are a dominant component of the tumor mass, driving the formation of a rigid, collagen-dense stroma that compresses blood vessels, impedes drug penetration, and fosters immune suppression by excluding cytotoxic T cells [79, 84, 85]. IGFBP2 is enriched in this fibrotic niche and promotes fibroblast activation and ECM production, further stiffening the matrix and amplifying integrin–TGF-β–YAP signaling cascades that support stromal persistence and tumor growth. Similarly, in triple-negative breast cancer (TNBC), high stromal IGFBP2 expression correlates with elevated α-SMA–positive myofibroblast content and is associated with increased tissue stiffness, EMT, and resistance to chemotherapy and immune checkpoint inhibitors [6, 8]. IGFBP2’s role in enhancing integrin β1/FAK signaling under such stiff conditions is thought to reinforce CAF-mediated remodeling and support invasive behavior [94]. In colorectal cancer, stromal expansion and fibrotic matrix remodeling are early events that accompany tumor budding and metastasis, with IGFBP2 linked to both fibroblast reprogramming and metastatic niche conditions. Metastasis-associated fibroblasts (MAFs) in peritoneal metastases of colorectal cancer express elevated levels of IGFBP2 (alongside cytokines like CXCL12 and IL6). These MAFs actively suppress immune responses by dampening macrophage activation and inhibiting T-cell proliferation [95]. Crucially, downregulation of IGFBP2 reverses these immunosuppressive effects, substantiating IGFBP2’s role in fibroblast-mediated immune modulation in colorectal metastasis [96]. In high grade serous ovarian carcinoma, the TME frequently contains a fibrotic stroma enriched with α-SMA–positive cancer-associated fibroblasts (CAFs), which are well recognized as mediators of chemoresistance and disease progression [16, 97]. IGFBP2 is consistently overexpressed in tumor epithelial cells, especially in high grade carcinomas. However, direct evidence for IGFBP2 expression in CAFs or stromal fibroblasts, and for its co-localization with integrin α5β1 or components of TGF-β signaling in the stroma, is currently lacking and warrants further investigation. Emerging evidence suggests that IGFBP2 within a COL11A1–NF-κB–IGFBP2–TGF-β3 signaling cascade reprograms ovarian stromal fibroblasts toward an activated phenotype [91]. This aligns with the broader understanding that CAFs drive matrix remodeling and fibroblast differentiation. However, evidence for a CAF-derived, IGFBP2-enriched exosomal feed-forward loop that amplifies TGF-β signaling and mechanical stiffening remains speculative. Likewise, while stromal periostin (and, by extension, fibrotic gene signatures) associates with disease progression and poor outcomes, further research is required to establish whether stromal IGFBP2 expression independently predicts treatment failure or relapse [91, 98].

Collectively, these findings underscore that IGFBP2 can function as stromal marker. Evidence suggests it is also a functional driver of fibrosis, fostering a mechanoresponsive TME that sustains malignancy, limits therapy access, and predicts poor clinical outcomes across multiple solid tumors, a mechanism that requires further validation in relevant physiological models.

8.3. Immune Microenvironment Modulation by IGFBP2

8.3.1. Immunosuppression via Recruitment of Tregs and MDSCs

The immunomodulatory effects of IGFBP2 reflect a conserved functional program that operates across diverse malignancies, though context-specific differences also exist. Common to multiple tumor types, (e.g. PDAC, ovarian, melanoma, glioblastoma, and prostate cancer) is the consistent activation of STAT3, either directly via IGFBP2-integrin interactions or indirectly through receptor tyrosine kinases such as EGFR and FGFR1 [8, 55]. This activation initiates transcriptional programs that upregulate indoleamine 2,3-dioxygenase (IDO), PD-L1, IL10, and various chemokines, which recruit and activate Tregs, MDSCs, and M2-like TAMs, creating a robustly immune-suppressive microenvironment [43, 52].

In PDAC, for example, IGFBP2 orchestrates immunosuppression through a multi-tiered mechanism involving STAT3-induced IDO expression, chemokine-mediated MDSC recruitment, and FAK–STAT3-driven IL10 production [43, 52]. IGFBP2 expression is enriched within fibrotic, desmoplastic PDAC tumor regions where it interacts with components of the ECM, such as integrins α5β1 and αvβ3, activating FAK and downstream STAT3 signaling [99]. This amplifies immunosuppressive cytokine production, particularly IL10 and TGF-β1, both of which inhibit dendritic cell maturation and further promote Treg differentiation [43, 100]. This establishes a feedback loop whereby stromal remodeling and immune evasion are co-regulated, reinforcing resistance to immune checkpoint inhibitors. In glioblastoma, IGFBP2 promotes immune escape through phosphorylation of FcγRIIB (Fc fragment of IgG receptor IIb) on microglia, resulting in impaired antigen presentation and accumulation of M2-polarized macrophages [43]. In melanoma, nuclear IGFBP2 synergizes with EGFR to activate STAT3-driven PD-L1 expression, enabling tumors to resist CD8⁺ T cell mediated cytotoxicity while paradoxically retaining high T cell infiltration, a profile predictive of favorable responses to PD-1 blockade. This dichotomy underscores the complex role of IGFBP2 wherein its immunosuppressive output may depend on broader tumor immune phenotypes (inflamed vs. non-inflamed) [101, 102].

Notably, across solid tumors such as NSCLC, colorectal, and prostate cancers, IGFBP2 often co-segregates with mesenchymal gene signatures and markers of immune resistance [103, 104]. In NSCLC, IGFBP2-induced STAT3 activation leads to secretion of CXCL1 and CCL5, promoting neutrophil and MDSC infiltration [9]. In prostate cancer, elevated IGFBP2 correlates with PI3K/AKT pathway activation (particularly in the context of PTEN loss) and integrin β1 signaling. Although a direct mechanistic chain from IGFBP2 → integrin activation → AKT/mTOR → PD-L1/IDO up-regulation has yet to be fully demonstrated, this represents a plausible axis by which IGFBP2 could link metabolic reprogramming to immune escape [43, 105].

8.3.2. Immune Checkpoint Upregulation

In addition to fostering a suppressive tumor immune microenvironment through regulatory cell recruitment, IGFBP2 facilitates immune escape by upregulating key immune checkpoint molecules, notably PD-L1, across multiple tumor types [8]. Mechanistically, IGFBP2 modulates PD-L1 expression via activation of canonical oncogenic pathways, including STAT3, EGFR, and AKT, in both tumor cells and immunosuppressive stromal components [6]. In PDAC, IGFBP2 cooperates to activate EGFR–STAT3 signaling, by facilitating EGFR nuclear accumulation, resulting in enhanced transcriptional activity of the IL10 secretion and immunosuppressive macrophage polarization via STAT3. This leads to increased M2-like TAMs, particularly under high tumor burden [43]. These TAMs reduced antigen-presenting capacity and inhibit CD8⁺ T cell activation in a PD-L1-dependent manner, which can contribute to T-cell exhaustion and immune evasion. Beyond PDAC, in melanoma, IGFBP2 binds EGFR at the plasma membrane and facilitates its nuclear import, where it functions as a transcriptional co-activator. Nuclear EGFR partners with STAT3 to induce PD-L1 transcription in tumor cells [101, 104]. Clinical data show that co-expression of IGFBP2 and PD-L1 correlates with poor prognosis; however, patients with this profile paradoxically respond better to anti–PD-1 therapy, possibly due to the presence of a highly infiltrated but suppressed immune milieu [43]. In glioblastoma, IGFBP2 enhances PD-L1 expression in mesenchymal-like tumor cells via STAT3 and AKT activation, contributing to both immune suppression and therapy resistance [106]. Similarly, in NSCLC, IGFBP2 increases PD-L1 expression through integrin–STAT3–CXCl1 signaling pathway, establishing a link between immune evasion and metabolic reprogramming, leading to chemoresistance to gefitinib [62].

In ovarian cancer, similar IGFBP2-driven mechanisms have been postulated. Tumors with elevated IGFBP2 (melanoma) show concurrent upregulation of PD-L1 in both tumor cells and infiltrating MDSCs and TAMs [8]. IGFBP2 enhances nuclear EGFR translocation and STAT3 phosphorylation in both compartments, promoting sustained PD-L1 expression [43]. Importantly, blockade of IGFBP2 or STAT3 reduces PD-L1⁺ MDSCs and improves T cell proliferation in ex vivo ascites models, highlighting a functional link between IGFBP2 and checkpoint-mediated immune suppression [43]. Whether similar mechanisms operate in ovarian cancer, especially within ascites-derived MDSCs or TAMs, remains to be investigated, as EGFR is frequently overexpressed in ovarian carcinomas, resulting in downstream activation of JAK2/STAT3, promoting EMT, motility, and invasive behavior [107]. In high-grade (grade 2 and 3) serous ovarian tumors, activation of EGFR and STAT3 is significantly elevated compared to normal ovarian tissue [107]. Blocking STAT3 using JAK2 inhibitors suppresses STAT3 phosphorylation, EMT markers (e.g., N-cadherin, vimentin), and cell motility [107].

8.3.3. Macrophage/Microglia Polarization

Another key immunosuppressive mechanism orchestrated by IGFBP2 in solid tumors is the functional polarization of TAMs and microglia toward the M2 phenotype, a state associated with immune evasion, stromal remodeling, and tumor progression [52]. Sun et al. describes that IGFBP2 activates STAT3 in PDAC tumor cells, leading to transcriptional upregulation and secretion of IL10, a key cytokine driving immunosuppressive macrophage polarization [43]. This IL10-mediated skewing toward the M2 phenotype enhances Treg recruitment, inhibits cytotoxic lymphocyte function, and supports a tolerogenic immune microenvironment [52]. Notably, TAMs conditioned by IGFBP2-expressing tumor cells show increased expression of ARG1, PD-L1, and CD206, consistent with a pro-tumorigenic M2 profile [108].

In glioblastoma, elevated IGFBP2 expression correlates with increased infiltration of CD163⁺ M2-like TAMs, reduced CD8⁺ T cell activity, and poor overall survival [52, 109]. In central nervous system tumor glioblastoma, IGFBP2’s immunoregulatory functions extend to resident microglia and infiltrating macrophages [6]. High IGFBP2 expression is associated with the mesenchymal transcriptional subtype, a state linked to immune suppression and treatment resistance [109, 110]. IGFBP2 promotes phosphorylation of the inhibitory immune receptor FcγRIIB on myeloid cells, a modification that blunts antigen-presenting capacity and reduces pro-inflammatory signaling. This effect is coupled with a depletion of intratumoral CD8⁺ T cells and CD19⁺ B cells, alongside a striking accumulation of M2-polarized macrophages and Iba1⁺ CD206⁺ microglia, creating a profoundly immunosuppressive niche [109]. Genetic inhibition of IGFBP2 reduces FcγRIIB phosphorylation, reprograms myeloid cells toward a pro-inflammatory (M1-like) state, restores CD8⁺ T cell infiltration, and significantly prolongs survival [109] in a syngeneic mouse glioblastoma model. These findings suggest that IGFBP2 is not only a marker of poor prognosis but also an active driver of myeloid cell reprogramming, supporting its consideration as a therapeutic target in both brain and systemic tumors. In ovarian cancer, TAM polarization similarly contributes to immune evasion and therapy resistance [91, 111]. Tumor-derived IGFBP2 upregulates NF-κB pathways in infiltrating tumor cells or in CAFs, leading to enhanced IL10 and TGF-β3 secretion, which sustains M2-like polarization and suppresses antigen presentation [91]. Tang et al. found an association between ascitic levels of TGF- β and IL10 and the presence of CD163+ M2-like TAMs which is inversely proportionate with relapse-free survival in ovarian cancer [111]. Moreover, with M2-like TAMs supporting tumor angiogenesis and ascitic accumulation, we can postulate a role for IGFBP2 on both immune modulation and malignant effusion in the ovarian TME, but conclusive evidence is still required.

These findings speculatively position IGFBP2 at the intersection of oncogenic signaling, immune regulation, and stromal remodeling. Its pleiotropic effects, ranging from immune checkpoint regulation and immunosuppressive cell recruitment to immune exclusion, and immunosuppressive myeloid polarization highlight its potential as a pan-cancer immunomodulatory target. Therapeutic strategies aimed at disrupting IGFBP2 signaling or its downstream immune effectors (e.g., STAT3 inhibitors, anti-CCL2/CCR2 blockade, anti-PD-L1) may enhance immune checkpoint blockade efficacy, particularly in IGFBP2-high tumors that remain resistant to current immunotherapies. Based on published reports, IGFBP2 may be best described as a potential contributor of immunosuppressive signaling networks, rather than a master orchestrator of immune cell recruitment.

9. Angiogenic and Vascular Mechanisms of IGFBP2

9.1. Extracellular Activation of Endothelial Signaling via Integrins

IGFBP2 promotes tumor angiogenesis through direct engagement of integrin receptors on endothelial and tumor cells [112]. This interaction is primarily mediated by the conserved RGD motif in its C-terminal domain, which facilitates binding to αVβ3 and α5β1 integrins, two receptors critically involved in endothelial activation and migration. Upon binding, IGFBP2 initiates a signaling cascade that culminates in activation of the PI3K/AKT and MAPK/ERK pathways, transcriptional upregulation of pro-angiogenic mediators such as VEGFA and enhanced endothelial sprouting and survival [58]. Mechanistically, the interaction between IGFBP2 and integrins has been shown to be functionally indispensable. In glioblastoma models, mutation of the RGD sequence to a non-binding RGE motif abrogates integrin association and disrupts downstream activation of AKT and ERK, leading to suppression of cell migration, adhesion, and angiogenic signaling [58, 78]. Co-immunoprecipitation assays confirmed that IGFBP2 interacts with integrin α5β1, and RNA interference targeting either IGFBP2 or integrin α5 significantly reduces PI3K/AKT phosphorylation and endothelial-like tube formation [113]. Upon integrin engagement, IGFBP2 activates class IA PI3K, promoting phosphorylation of AKT at Ser473 and Thr308. Activated AKT enhances the transcriptional activity of HIF-1α, which directly binds the VEGFA promoter, leading to increased VEGFA production. Concurrently, AKT promotes endothelial nitric oxide synthase (eNOS) activation and cytoskeletal reorganization, facilitating endothelial cell migration and sprouting [55, 114]. These effects are supported by experimental evidence demonstrating that recombinant IGFBP2 induces rapid AKT and MAPK phosphorylation in human umbilical vein endothelial cells, enhances lamellipodial protrusions in 3D collagen matrices, and promotes endothelial sprouting in vitro [115].

Furthermore, integrin/IGFBP2 interaction activates FAK and downstream SRC family kinases, reinforcing anchorage-dependent survival and cell polarity, key events in endothelial sprouting [72, 115]. In glioblastoma, Shi et al. established that integrin-mediated IGFBP2 signaling activates the ERK/MAPK pathway within 15–30 minutes, driving proliferation and chemoresistance [46]. Transcriptomic analyses of glioma cells expressing integrin-binding-deficient IGFBP2 mutants (RGE) reveal widespread downregulation of angiogenic and motility-related genes, including MMP2, CD44, and VEGFA, further emphasizing the centrality of integrin signaling to IGFBP2 function [116]. Additionally, in hepatocellular carcinoma (HCC), IGFBP2 expression predicts a pro-angiogenic transcriptome and correlates with microvascular invasion [16, 90].

In ovarian cancer, IGFBP2 is markedly overexpressed in high-grade serous and endometrioid carcinomas and has been associated with advanced stage, aggressive tumor phenotype, and reduced survival [117, 118]. Lee et al. and Lancaster et al. reported that elevated IGFBP2 expression distinguished invasive carcinomas from borderline tumors and normal ovarian epithelium [117, 118]. In vitro, IGFBP2 enhances cell motility, invasion, and survival processes reliant on integrin-mediated focal adhesion and cytoskeletal remodeling [21, 119]. While endothelial-specific studies in ovarian cancer remain limited, the observed co-expression of IGFBP2 and VEGF-A, alongside activation of ILK and FAK signaling [117, 119], suggest that IGFBP2 likely contributes to both tumor–endothelial interaction and stromal vascular remodeling.

9.2. Nuclear Regulation of Angiogenesis by IGFBP2

Beyond the role of integrin-mediated interactions in regulating angiogenesis, nuclear IGFBP2 also directly regulates the transcription of angiogenesis-associated genes [24]. In neuroblastoma cells, nuclear IGFBP2 has been shown to bind the VEGFA promoter and enhance its transcription independent of IGF signaling, implicating IGFBP2 as a direct nuclear co-regulator of angiogenic genes [24, 58]. Moreover, chromatin immunoprecipitation assays in glioma have confirmed IGFBP2 occupancy at promoter regions of several genes involved in angiogenesis and invasion [120]. Nuclear IGFBP2 has also been implicated in the modulation of histone acetylation and chromatin accessibility, potentially acting as a transcriptional co-activator in partnership with p300 or STAT3 in hypoxic tumor regions [55].

These nuclear roles are particularly relevant in tumors characterized by hypoxia-driven angiogenesis. In glioblastoma, IGFBP2 expression is enriched in peri-necrotic and pseudo palisading regions, areas with high HIF activity and active neovascularization consistent with regulation by hypoxia-relayed pathways [121]. These findings suggest a dual mechanism by which IGFBP2 coordinates angiogenesis, through extracellular integrin-mediated signaling and direct nuclear transcriptional regulation of angiogenic pathways.

9.3. IGFBP2 and Vasculogenic Mimicry

IGFBP2 also facilitates vasculogenic mimicry (VM), the tumor cell driven formation of perfusable, vessel like channels independent of endothelial cells, allowing them to establish an independent alternative vasculature for their nutrient needs [122]. Mechanistically, IGFBP2 enhances the expression of VE cadherin (CD144) and MMP2 via integrin-dependent activation of FAK/ERK signaling, enabling cytoskeletal rearrangement and ECM remodeling necessary for VM formation [122]. Chiablem et al. also describes that PI3K/AKT/mTOR is one of the key signaling pathways for the formation of VM in tumors [123] and this pathway can be activated by IGFBP2. These pathways facilitate cytoskeletal reorganization and ECM remodeling, both of which are essential for tumor cell plasticity and the formation of vascular-like networks. Silencing IGFBP2 in VM-competent glioma cells led to marked reductions in VM channel formation in 3D Matrigel assays, as well as decreased expression of VE-cadherin and MMP2 [5, 122]. Additionally, IGFBP2 enhances the expression of hypoxia-associated genes such as TWIST1 and SNAI1, further promoting cellular dedifferentiation and trans-differentiation toward a vascular phenotype [124]. Importantly, VM structures in IGFBP2-overexpressing tumors often co-localize with hypoxic and peri-necrotic regions, where classical endothelial angiogenesis is compromised [122]. This spatial association suggests that IGFBP2 may play a compensatory role in maintaining perfusion and nutrient delivery under conditions of vascular insufficiency, facilitating tumor cell survival and growth in severely hypoxic microenvironments [122]. These findings suggest IGFBP2 may act as a key mediator of non-endothelial vascular mimicry, adding to its complexity as a pro-angiogenic regulator.

Taken together, these findings support that IGFBP2 acts as a multifunctional, paracrine angiogenic factor that promotes endothelial activation and vascular sprouting through integrin-mediated PI3K/AKT signaling. Its pro-angiogenic influence is evident across a range of tumor types, including PDAC and ovarian cancer, where it contributes to a pro-invasive and pro-vascular TME.

10. Functional Heterogeneity of IGFBP2: Context Dependent Function

Although IGFBP2 has historically and widely been characterized mostly as an oncogenic factor across various malignancies, increasing evidence also describes its function as context-dependent, exhibiting both tumor-suppressive and oncogenic activities in the same kind of tumor depending on cellular origin, microenvironmental conditions or its immunologic paradigm. This apparent paradox reflects the dual capacity of IGFBP2 to regulate tumor biology through both IGF-dependent ligand sequestration and IGF-independent signaling mechanisms, often with opposing biological consequences.

Numerous studies have linked IGFBP2 overexpression to different cancers for tumor progression, invasion, and poor clinical outcomes [6]. Mechanistically, these oncogenic functions are often attributed to its IGF-independent pathways, including integrin engagement, activation of PI3K/AKT signaling, and transcriptional regulation when it is transported in the nucleus, which is independent of its canonical IGF-binding activity [6].

In contrast, the tumor-suppressive function of IGFBP2 is attributed to its ability to bind IGF-I and IGF-II with high affinity, thereby limiting activation of IGF receptors and suppressing IGF-driven proliferation and survival pathways [57, 59]. This concept is supported by early mechanistic studies demonstrating that IGFBPs act as key regulators of IGF bioavailability and signaling [57, 59]. Experimental evidence from lung cancer models shows that exogenous or secreted IGFBP2 can inhibit tumor cell growth by sequestering IGFs, thereby suppressing mitogenic signaling [16, 21]. In small cell lung cancer, recombinant IGFBP2 supplementation significantly reduces tumor cell proliferation, while neutralization or knockdown of IGFBP2 accelerates growth, further supporting its inhibitory role in IGF-dependent contexts [16, 21].

A particularly important dimension of this opposing effect is the cellular source of IGFBP2. The recent study by Conway et al. [7] provides compelling evidence supporting this distinction. Conway et al. demonstrates that tumor-derived IGFBP2 is generally associated with pro-tumorigenic functions, whereas stromal-derived IGFBP2 may exert protective effects. The authors demonstrated that IGFBP2 secreted by healthy mammary adipocytes acts as an extracellular barrier that inhibits breast cancer invasion by binding and sequestering tumor-derived IGF-II. This adipocyte-derived IGFBP2 effectively suppresses the autocrine IGF-II signaling required for invasive progression, thereby functioning as a tumor-suppressive microenvironmental factor [7]. Importantly, this work suggests that IGFBP2 may play a protective role in normal tissue architecture, and that disruption of this stromal compartment may promote tumor progression.

Further complexity arises from tumor-type-specific and stage-dependent effects, as illustrated by Wang et al. in their NSCLC dataset analysis [9]. In their study, IGFBP2 exhibits opposing prognostic associations depending on histological subtype: high IGFBP2 expression correlates with poor overall survival in adenocarcinoma but paradoxically associates with improved survival in squamous cell carcinoma. These findings provide direct clinical evidence that IGFBP2 cannot be universally classified as either oncogenic or tumor suppressive. Instead, its function varies depending on tumor lineage, molecular context, and possibly distinct microenvironmental interactions. Similarly, mechanistic studies from Yazawa et al. [129] in lung cancer models further support this duality. In small cell lung cancer, IGFBP2 overexpression is linked to neuroendocrine differentiation via transcriptional regulation by NeuroD, yet this overexpression paradoxically suppresses cell proliferation in vitro, despite the aggressive behavior of small cell lung cancer in vivo. This discrepancy suggests that IGFBP2 may act as a growth-modulatory “brake” within highly proliferative tumors, but its suppressive effects may be overridden by other oncogenic drivers, growth factor signaling, or ECM interactions in the in vivo tumor environment. Indeed, Sethi et al. [130] showed membrane-associated IGFBP2 has been shown to promote cancer progression through interactions with the ECM, further illustrating how its functional output depends on subcellular localization and extracellular context.

In conclusion, IGFBP2 should be regarded as a multifunctional and context-dependent mediator with dual oncogenic and tumor-suppressive roles, rather than a purely pro-tumorigenic factor. Investigation into this functional heterogeneity is essential for achieving a balanced and accurate understanding of IGFBP2 biology and for guiding the rational design of future therapeutic strategies.

11. IGFBP2 in Metabolic and Tumor Heterogeneity

Accumulating evidence positions IGFBP2 as a master regulator of metabolic heterogeneity within tumors, integrating growth-factor signalling, nutrient metabolism and stromal crosstalk [97]. Originally characterized as an IGF carrier, IGFBP2 also engages with integrins (e.g., α5β1) and IGF-IR to amplify downstream PI3K–AKT–mTOR and MAPK/ERK pathways, both of which converge on metabolic regulators that control glucose and lipid utilization and promoting anabolic metabolism [45, 105, 119]. Mechanistically, IGFBP2 drives glycolytic reprogramming by upregulating the PKM2/HIF-1α axis [131]. Elevated IGFBP2 interacts with PKM2, prevents its degradation, and promotes its nuclear translocation, where it acts as a co-activator of HIF-1α. This results in enhanced HIF-1α–mediated transcription of glycolytic genes including HK2, LDHA and PFKFB3 in endometrial cancer cells [131]. The shift to aerobic glycolysis provides rapid ATP and anabolic intermediates, acidifies the microenvironment, supports invasion, and selects for clones capable of thriving under hypoxia and nutrient stress [131].

In parallel, IGFBP2 modulates lipid metabolic remodelling via the AKT-mTOR-SREBP1 axis, thereby increasing transcription of lipogenic enzymes such as ACLY, ACC, and FASN and lipid droplet accumulation [71]. In melanoma, IGFBP2 derived from aged fibroblasts increases de novo lipid synthesis and invasion [71]. These lipidogenic reprogramming supports membrane biosynthesis, energy storage, and production of signalling lipids such as prostaglandins that influence motility and intercellular communication [71]. Moreover, adipocyte-secreted IGFBP2 has been shown to modulate tumor energy homeostasis [7] through AKT-dependent mechanisms, suggesting that paracrine IGFBP2 integrates systemic metabolic cues into local tumor nutrient programs. These lipogenic adaptations diversify the metabolic capabilities of cancer subclones, facilitating clonal selection under nutrient or therapeutic stress. These findings illustrate how IGFBP2 fosters lipogenic niches and metabolic plasticity within the TME. Beyond cell-intrinsic effects, IGFBP2 functions as a microenvironmental mediator [71]. In glioma, CAF-derived IGFBP2 fosters macrophage polarization toward an M2, immunosuppressive phenotype via the AKT/STAT3 axis, creating metabolic-immune niches that enable tumor growth and heterogeneity [52, 132]. Clinically, elevated IGFBP2 expression is consistently associated with poor prognosis, therapy resistance and increased metabolic activity across multiple cancers [133, 134]. Collectively, these reports support a model in which IGFBP2 functions as a molecular nexus linking oncogenic signalling, metabolism and stromal/immune communication. Thus, targeting IGFBP2 or its downstream metabolic axes may represent a compelling strategy to disrupt tumor metabolic heterogeneity and improve patient outcomes.

12. Differential Patterning of IGFBP2 signalling: The Next Frontier?

Accumulating evidence places IGFBP2 at the epicenter of three critical biological processes: tumor heterogeneity, metabolic reprogramming, and the TME [6]. Ensuing tumor heterogeneity due to cell extrinsic factors such as hypoxia, the TME, acidosis, and chemotherapy, and cell intrinsic factors such as clonally distinct populations with distinct energy demands arising from genetic diversity, leads to metabolic/bioenergetic reprogramming to survive these multimodal insults [133]. Since IGFBP2 is expressed and secreted by cancer cells, fibroblasts, and endothelial cells [6, 135], it warrants investigation into whether there is differential patterning of IGFBP2 signaling by cancer cells, fibroblasts, and endothelial cells in response to extrinsic and intrinsic factors.

To address these questions, employing experimental strategies such as single-cell RNA sequencing (scRNA-seq) and spatial proteomics will be critical. These methodologies can dissect IGFBP2 signaling pathways at the cellular level, providing insights into its role within heterogeneous TMEs. Single-cell RNA-seq can identify cell-type-specific expression patterns and delineate how IGFBP2-mediated signaling adapts to environmental pressures, while spatial proteomics can reveal the localization and interaction dynamics with other key regulators in different tumor niches.

13. Roles for IGFBP2 Outside of Cancer

While our primary focus here is the role of IGFBP2 in cancer, IGFBP2 does have extensively documented functions in other systems. These non-cancer functions often share cellular mechanisms (e.g. regulation of growth, mesenchymal transition, fibrosis) with those we have discussed above in the context of carcinogenesis and metastasis. Likewise, studies have identified both IGF-dependent and IGF-independent roles for IGFBP2 throughout the body. In depth consideration of this literature is beyond the scope of this review, but much work has focused on the role of IGFBP2 in developmental, inflammatory, metabolic, cardiovascular, and neurological/neurodegenerative disease. Elevated IGFBP2 has been linked to neurological and neurodegenerative disorders, and cerebrospinal IGFBP2 levels are elevated in patients with clinical diagnosis of Alzheimer’s disease (AD) [136–138]. Mechanistically, the negative effects of IGFBP2 and neurodegeneration may be tied to its ability to promote neuroinflammation and oxidative stress in neurons [136].

IGFBP2 levels are metabolically regulated. Circulating IGFBP2 levels are inversely correlated with T2D, obesity, and metabolic syndrome, a protective role for IGFBP2 in metabolic dysfunction, which has been demonstrated in mouse models [139, 140]. In the context of cardiovascular disease, there is evidence that low IGFBP2 levels may be a predictor for atherosclerosis, risk of heart failure, and signs of arterial stiffness [137, 140, 141]. However, elevated IGFBP2 has also been reported in patients with severe acute cardiac events such as myocardial infarction/death, and poor outcomes after mitral valve repair suggesting the pattern is not straightforward [142, 143]. Similarly, IGFBP2 has also been associated with increased all-cause mortality in aging and increases in inflammatory diseases (i.e. sepsis) [144, 145]. This disconnect between protective vs. detrimental effects of IGFBP2 likely has tissue, disease context, disease timing, and cell-type specific underpinnings and parallels many of the context specific effects seen when considering the role of IGFBP2 in cancer.

14. Clinical studies and translational implications

Several clinical studies have corroborated the link between IGFBP2 expression, stromal activation, and adverse clinical outcomes [146]. Retrospective analyses of PDAC and glioblastoma patients show that high IGFBP2 levels in plasma or tumor tissue predict shorter survival, higher tumor grade, and reduced response to immune checkpoint blockade [16, 52, 147]. The combination of IGFBP2 with CA19–9 improves diagnostic accuracy compared to CA19–9 alone, underscoring its potential as a complementary diagnostic biomarker in PDAC [147], although prospective immune checkpoint inhibitor trial data stratified by IGFBP2 are limited.

Exploratory clinical and translational efforts are beginning to evaluate IGFBP2 as a prognostic and predictive biomarker. Preclinical IGFBP2-targeting strategies (neutralizing antibodies, antisense oligonucleotides, and disruption of IGFBP2–integrin signaling) are trying to demonstrate stromal normalization and increased T cell infiltration strategies in animal models. In addition, translational analyses link IGFBP2 to immunosuppressive cytokine programs (IL10/TGF-β) and immune exclusion in relevant tumor types such as glioma, PDAC etc. [52, 104, 116, 120]. Collectively, these findings suggest that IGFBP2 acts as a multifaceted orchestrator of stromal and immune regulation, serving not only as a predictive biomarker of poor prognosis but also as a functional target that may recondition the TME toward immunotherapeutic responsiveness.

One promising approach for direct targeting is IGFBP2-directed immunotherapy. A Phase I clinical trial utilizing a plasmid-based IGFBP2 vaccine in advanced ovarian cancer demonstrated safety and induced robust IGFBP2-specific Th1 immune responses, including increased IFN-γ-producing T cells without expansion of Tregs [148]. Building on these findings, a Phase II clinical trial is currently evaluating an IGFBP2 DNA vaccine in combination with carboplatin for patients with recurrent ovarian cancer (NCT07495124). These studies suggest that IGFBP2 may be effectively targeted as a tumor-associated antigen in humans.

15. IGFBP2 as a Mediator of Therapy Resistance

Emerging evidence positions IGFBP2 as a critical mediator of therapy resistance across multiple cancer types [146].Theoretically,, IGFBP2 can promotes sustained activation of key survival pathways, particularly the PI3K/AKT and MAPK/ERK cascades, even under conditions of receptor tyrosine kinase (RTK) inhibition. For example, in glioblastoma, IGFBP2 facilitates PI3K/AKT reactivation following EGFR inhibition, bypassing the blockade and sustaining downstream proliferative and anti-apoptotic signals [149]. In NSCLC, IGFBP2 overexpression promotes resistance to EGFR inhibitors like gefitinib by activating an alternative pathway through the STAT3/CXCL1 axis [150]. This axis re-establishes survival signaling independent of EGFR, effectively diminishing the efficacy of targeted therapies. IGFBP2 also undermines RTK-targeted therapeutic efficacy by sustaining downstream PI3K/AKT and NF-κB signaling independently of receptor activity, thereby facilitating adaptive resistance to EGFR and PI3K pathway inhibitors [16, 77]. Mechanistically, IGFBP2 interacts with integrins and activates FAK, which in turn stimulates PI3K signaling independent of upstream RTK activity. This survival reinforcement enables continued tumor proliferation and resistance to apoptosis despite therapeutic pressure [77].

Importantly, the nuclear functions of IGFBP2, such as the modulation of gene expression via interactions with STAT3 and β-catenin, further potentiate adaptive resistance mechanisms, particularly under chronic treatment pressure [55]. These nuclear actions suggest that IGFBP2 operates at multiple levels, membrane signaling, cytoplasmic kinase reactivation, and nuclear transcriptional reprogramming, to orchestrate robust, multi-modal therapy resistance. Thus, targeting IGFBP2, either directly or through its downstream effectors, may represent an attractive strategy to overcome resistance and enhance therapeutic responses in multiple cancer contexts.

IGFBP2 directly contributes to therapeutic resistance against EGFR-targeted drugs [62]. In NSCLC, IGFBP2 overexpression leads to resistance to gefitinib, in part through activation of the STAT3/CXCL1 autocrine loop, enhancing survival and inflammatory signaling independent of EGFR [62]. Additionally, in glioma models, IGFBP2 has been shown to protect tumor cells from radiation-induced DNA damage by promoting DNA repair pathway activation and sustaining anti-apoptotic signaling cascades [21, 67, 151]. These dual resistance mechanisms substantially compromise the efficacy of both targeted therapies and conventional radiotherapy.

Beyond kinase reactivation, IGFBP2 also induces mesenchymal transitions, conferring broad-spectrum drug resistance [79, 88]. In glioma and other solid tumors, IGFBP2 promotes the expression of transcription factors such as ZEB1, TWIST1, and SNAI1, driving EMT programs that reduce cellular susceptibility to apoptosis and enhance invasiveness [152, 153]. This mesenchymal reprogramming contributes not only to chemotherapy resistance but also to decreased sensitivity to radiation therapy, as observed in experimental glioblastoma models [149]. Similar mechanisms are at play in PDAC, hepatocellular cancer, and esophageal adenocarcinoma where IGFBP2-induced EMT promotes mesenchymal reprogramming characterized by enhanced motility, invasiveness, and reduced sensitivity to chemotherapy [54, 88]. This transition not only facilitates tumor dissemination but also renders cells intrinsically more resistant to apoptosis and immune clearance, forming a key mechanism by which IGFBP2 sustains therapy resistance across diverse cancer types [146]. A hallmark of solid tumors is hypoxia, which induces a metabolic shift toward glycolysis and alters redox balance [131]. IGFBP2 is upregulated under hypoxic conditions via HIF-1α-mediated transcriptional activation, particularly in glioblastoma and hepatocellular carcinoma [154, 155]. This upregulation promotes metabolic plasticity by sustaining PI3K/AKT signaling and enhancing mitochondrial oxidative phosphorylation, enabling tumor cells to endure nutrient deprivation and oxidative stress [131]. In glioblastoma, IGFBP2 cooperates with STAT3 and β-catenin to activate transcriptional programs linked to therapy resistance, including genes involved in glycolysis, glutaminolysis, and lipid metabolism [55, 152].

Moreover, IGFBP2 supports cancer stem cell maintenance through both IGF-dependent and IGF-I-independent mechanisms [156]. In AML, IGFBP2 sustains stemness by activating PI3K/AKT and STAT3 pathways while repressing differentiation pathways, through IGF-dependent pathway [157]. This persistence of cancer stem cell populations contributes to tumor recurrence and resistance to genotoxic stress [157]. Importantly, these functions are temporally regulated, IGFBP2 expression surges during therapeutic stress and persists during recurrence, highlighting its role as a dynamic responder to environmental and treatment-induced selection pressures [157]. Collectively, the literature suggests that IGFBP2’s ability to simultaneously engage integrin signaling at the membrane, modulate gene transcription in the nucleus, and interface with metabolic and immune networks underscores its unique capacity to orchestrate adaptive resistance. Its pleiotropic functions complicate therapeutic targeting, as inhibition must address both extracellular interactions and nuclear transcriptional activity. This mechanistic complexity, coupled with tumor-specific expression patterns, necessitates the development of precision strategies that disrupt IGFBP2’s context-specific activities without perturbing systemic IGF signaling.

Together, these multifaceted mechanisms underscore IGFBP2 as a central orchestrator of adaptive resistance to both targeted and conventional therapies, highlighting its potential as a therapeutic vulnerability in cancer.

16. Clinical implication of IGFBP2

Given its central role in tumor progression, resistance to therapy, and immune evasion, IGFBP2 is increasingly recognized as a promising therapeutic target [6, 8]. Its functional versatility makes it a uniquely positioned molecule for multi-axis therapeutic intervention [146]. Targeting IGFBP2 may be especially effective in tumors where it acts as a nodal regulator of integrin signaling, transcriptional plasticity, and immune checkpoint gene expression.

Impairing the ability of IGFBP2 to engage αvβ3 and α5β1 integrins via its C-terminal RGD motif may be therapeutically beneficial. For instance, in glioblastoma, deleting αvβ3 integrin by using mice model where β3-integrin-floxed (β3-floxed) mice crossed to 2 endothelial-specific Cre models or mutating the IGFBP2 RGD motif markedly impaired glioma invasion and vascular mimicry [119]. Similarly, in ovarian cancer models, IGFBP2-dependent integrin activation promoted peritoneal dissemination and chemoresistance [88, 119]. Our lab previously described treating ovarian cancer cells with gold nanoparticles (GNP) decreased the expression of IGFBP2, via an autoregulatory feedback loop of IGFBP2/mTOR/PTEN axis [5]. This is likely mediated by GNP binding to the HBD domain of IGFBP2, as previous reports suggest that GNP can binds with HBD containing proteins [5]. Binding of the HBD domain to GNP alters the protein conformation [5] which in turn can regulate the IGFBP2-integrin interactions [5]. Therapeutic inhibition of this axis can be achieved through integrin-blocking antibodies, RGD-based peptides, or small-molecule inhibitors targeting FAK. The anti-αvβ3 antibody etaracizumab, though originally tested in melanoma, may have renewed potential in IGFBP2-high tumors where integrin-FAK signaling is a key driver of malignancy [36, 77]. Moreover, co-targeting FAK and IGFBP2-regulated pathways may be synergistic in preventing metastatic spread. At present, there are no approved therapeutic antibodies or small-molecule drugs that directly target IGFBP2, although the protein remains an active area of translational interest. The most clinically advanced strategy has been IGFBP2-directed vaccination, which has demonstrated safety and immunogenicity in early-phase trials but has not yet yielded definitive efficacy [148]. Preclinical studies strongly support IGFBP2 as a driver of tumor growth, invasion, metabolic adaptation, and therapy resistance across multiple cancer types, providing a rationale for direct neutralization approaches such as monoclonal antibodies [120]. However, development of IGFBP2-targeting biologics faces challenges related to its multifunctional domain architecture, involvement in both IGF-dependent and IGF-independent signaling, and potential systemic effects on the IGF axis. Consequently, most current efforts focus on preclinical antibody development, RNA-based suppression, or indirect targeting of IGFBP2-regulated signaling pathways rather than developing small-molecule inhibitors of IGFBP2 itself [148, 158].

Nuclear IGFBP2 modulates EMT-promoting transcription factors contributing to EMT and tumor plasticity [55, 133, 159]. In glioblastoma stem-like cells, nuclear IGFBP2 drives a transcriptional program that maintains stemness and impairs differentiation, fostering therapy resistance [149]. Furthermore, in malignant melanoma, IGFBP2 activates EGFR–STAT3 signaling, leading to the transcriptional upregulation of PD-L1, which suppresses cytotoxic T cell responses and facilitates immune escape [126]. Given that IGFBP2 harbors a functional NLS and translocates to the nucleus via importin-α/β to regulate oncogenic gene expression [6], therapeutic strategies might include NLS blockade or nuclear import inhibition. Moreover, transcriptional interference via epigenetic modulators (e.g., HDAC inhibitors, BET bromodomain inhibitors) could suppress IGFBP2-driven nuclear programs, although such approaches remain largely conceptual and need further validation [21, 24].

IGFBP2 has emerged as an immune-modulatory molecule by regulating PD-L1 and contributing to immune exclusion within the TME [8]. High IGFBP2 expression correlates with reduced T cell infiltration and inferior response to PD-1/PD-L1 blockades in glioblastoma, melanoma, and NSCLC [8]. In preclinical glioma models, IGFBP2 knockdown promoted CD8⁺ T cell infiltration and increased sensitivity to anti-PD-1 therapy, thereby restoring anti-tumor immunity [8]. These findings suggest that IGFBP2 inhibition may convert “cold” tumors into “hot” immunologically active microenvironments, rendering them more responsive to immune checkpoint blockade. Therefore, combination therapies that co-target IGFBP2 (e.g., via antisense oligonucleotides or monoclonal antibodies) alongside anti–PD-1/PD-L1 or anti–CTLA–4 agents’ nay be promising [160, 161]. The application of IGFBP2-targeted therapy will likely be most effective in biomarker-stratified patients, particularly those with high tumor or serum IGFBP2 levels, EMT-positive gene signatures, or immune-excluded phenotypes.

Despite the growing understanding of IGFBP2’s multifaceted roles in cancer progression, several challenges and opportunities remain in translating these insights into clinical applications. As discussed above, one of the foremost hurdles is the selective targeting of IGFBP2 without disrupting the broader IGF axis, which is important for physiological growth and metabolic functions. As IGFBP2 binds IGF-I and IGF-II with high affinity, global inhibition could inadvertently affect normal tissue homeostasis, necessitating the development of strategies that preserve systemic IGF signaling while neutralizing tumor-promoting functions.

A promising avenue lies in the design of domain-specific inhibitors that selectively block IGFBP2’s IGF-independent functions. For instance, targeting the C-terminal RGD motif to prevent integrin-mediated activation of FAK and PI3K/AKT signaling may inhibit tumor invasion and angiogenesis without altering IGF availability. Small molecules, peptides, or monoclonal antibodies that interfere with these domain-specific interactions could offer precision tools to dismantle the oncogenic network wired around IGFBP2.

Another critical and emerging area is the role of IGFBP2 in cancer stem cell biology and metabolic adaptation. IGFBP2 has been shown to maintain stemness phenotypes in glioblastoma and prostate cancer by modulating PI3K/AKT and STAT3 pathways, while also enhancing oxidative phosphorylation and fatty acid metabolism [7]. These functions suggest that IGFBP2 is not merely a downstream effector but may act as a metabolic gatekeeper that sustains cancer stem cell pools under therapeutic stress or nutrient limitation. Moreover, circulating IGFBP2 levels have shown potential as non-invasive biomarkers for cancer diagnosis and progression monitoring [162]. As IGFBP2 can be detected in plasma and cerebrospinal fluid [101, 136], its dynamic regulation during treatment could provide insights into therapeutic efficacy, minimal residual disease, or early relapse. However, further clinical validation is required to determine its predictive power across tumor types and therapeutic contexts. Collectively, these directions underscore the necessity for integrative studies combining structural biology, functional genomics, and translational research to unlock the full clinical potential of IGFBP2 as both a therapeutic target and a biomarker.

17. Conclusion and Future Perspectives:

IGFBP2 is a multifaceted protein whose dysregulation contributes to multiple hallmarks of cancer, including sustained proliferative signaling, invasion, metastasis, and therapeutic resistance [6]. Through IGF-dependent and IGF-independent mechanisms, IGFBP2 exerts a broad oncogenic influence [6]. Its overexpression in diverse malignancies, correlation with poor clinical outcomes, and capacity to drive resistance to standard therapies underscore its potential as a therapeutic target and biomarker. Importantly, IGFBP2 can function in a context-dependent manner and can have biologically opposing effects. This duality reflects its complex interplay and underscoring the needs of interpret IGFBP2 biology within specific tumor context. However, the clinical translation of IGFBP2-targeted approaches remains limited by the challenge of selective inhibition without perturbing physiological IGF signaling. Unlike upstream growth factors (e.g., IGF-I/2) or RTKs that often exhibit redundancy and compensation in signaling, IGFBP2 functions as a nodal effector that integrates diverse oncogenic and microenvironmental cues [8, 16]. Through its IGF-independent interactions with integrins, nuclear transcription factors, and ECM components, IGFBP2 orchestrates a coordinated program of invasion, immune evasion, and resistance to therapy [119]. Therefore, IGFBP2 represents a convergent vulnerability point within the TME, offering an opportunity for broad-spectrum therapeutic disruption.

A particularly attractive strategy lies in domain-specific inhibition of IGFBP2’s functional motifs. The C-terminal RGD motif mediates binding to integrin receptors such as αvβ3 and α5β1, activating FAK and downstream PI3K/AKT and MAPK/ERK pathways. These signaling axes support survival, motility, angiogenesis, and resistance to therapy, particularly in glioblastoma, melanoma, and pancreatic cancers [37, 119]. In preclinical glioma models, mutation or inhibition of the RGD motif impairs perivascular invasion, vascular mimicry, and radioresistance [36]. Monoclonal antibodies or peptidomimetic inhibitors targeting the RGD-integrin interface have demonstrated efficacy in disrupting this axis, with potential to enhance responses to chemoradiation and anti-angiogenic agents [163, 164].

Simultaneously, the central linker region of IGFBP2, which harbors a nuclear localization signal (NLS), facilitates nuclear translocation and subsequent modulation of transcriptional programs governing EMT, immune checkpoint regulation (e.g., PD-L1), and stemness [6, 43]. Pharmacologic disruption of nuclear IGFBP2, either via NLS blockade, nuclear import inhibitors, or epigenetic modulators, may suppress EMT-related gene expression, reduce immune evasion, and sensitize tumors to immune checkpoint blockade. This strategy could be particularly effective in tumors with mesenchymal phenotypes or immune-excluded profiles, where nuclear IGFBP2 is functionally dominant. Together, these findings underscore the clinical value of IGFBP2 not only as a prognostic marker of poor outcome but also as a reflection of biologically aggressive, treatment-refractory disease. Its consistent association with mesenchymal phenotypes, immune evasion, and integrin/FAK pathway activation suggests that IGFBP2 levels may also guide therapeutic targeting and risk-adapted treatment strategies.

In colorectal cancer, integrin signaling is known to interface with core oncogenic networks, including the WNT/β-catenin pathway, a central driver of tumor initiation and progression due to frequent APC or CTNNB1 mutations. Moreover, integrin engagement and ECM remodeling have been shown to activate downstream mechano-transduction pathways such as YAP/TAZ, which can cooperate with β-catenin to sustain stemness and resistance programs [70, 165]. Evidence from breast cancer models suggests that IGFBP2 can enhance β-catenin nuclear accumulation via FAK and IGF-IR signaling. These findings point to a possible convergence of IGFBP2-integrin signaling with β-catenin transcriptional activity in colorectal cancer, potentially amplifying the expression of key target genes such as MYC, CCND1, and LGR5 that promote proliferation and stem cell maintenance. Future research should aim to define whether IGFBP2-derived fragments retain integrin-binding capabilities in colorectal cancer and whether they contribute to β-catenin stabilization or YAP/TAZ activation. Investigating this crosstalk may uncover novel therapeutic opportunities in colorectal cancer subtypes characterized by WNT or integrin pathway dependency.

From a therapeutic standpoint, targeting IGFBP2 offers multiple avenues for clinical exploration. Combining IGFBP2 inhibition (via neutralizing antibodies, antisense oligonucleotides, or peptide antagonists) with immune checkpoint blockade could help overcome immunosuppressive resistance by reactivating cytotoxic T cells and reversing M2 macrophage polarization. Similarly, co-targeting IGFBP2 and pro-fibrotic pathways, such as TGF-β or integrin–FAK signaling, may normalize the desmoplastic stroma, enhance drug penetration, and sensitize tumors to chemotherapy or immunotherapy. Future studies should also investigate IGFBP2 as a circulating biomarker for stratifying patients most likely to benefit from such combination regimens. Indirect targeting strategies focus on downstream signaling pathways activated by IGFBP2, many of which are already amenable to clinical intervention. IGFBP2 promotes tumor progression by activating integrin-mediated signaling, particularly the integrin–FAK–PI3K/AKT axis, which regulates invasion, angiogenesis, and survival. Experimental studies demonstrate that IGFBP2 engages integrin receptors, such as α5β1, to activate FAK and STAT3 signaling, thereby enhancing tumor growth and immune evasion [166]. These findings provide a mechanistic rationale for targeting IGFBP2-driven tumors with FAK inhibitors or integrin-targeting therapies, several of which are currently under clinical evaluation in solid tumors. Immune checkpoint regulation represents another significant therapeutic axis. IGFBP2 upregulates PD-L1 expression via EGFR–STAT3 signaling, which promotes immune escape and resistance to T cell–mediated cytotoxicity [8]. Therefore, targeting IGFBP2 in combination with PD-1/PD-L1 blockade constitutes a rational therapeutic strategy. Experimental inhibition of IGFBP2 has been shown to reduce immunosuppression, increase CD8⁺ T cell infiltration, and enhance anti-tumor immunity in preclinical glioblastoma models [8, 109, 167]. Collectively, these putative mechanisms position IGFBP2 as a candidate master regulator of immune tolerance in epithelial malignancies. However, direct evidence for many of these interactions in ovarian cancer remains limited. Future studies are warranted to delineate the immunological consequences of IGFBP2 activity in the ascites microenvironment and to determine whether therapeutic targeting of IGFBP2 could enhance antitumor immunity or improve responses to immunotherapy in ovarian cancer. Ultimately, advancing IGFBP2 research will require integrated preclinical and clinical approaches, combining molecular mapping, single-cell and spatial transcriptomics, and functional genomics, to define its mechanistic and therapeutic landscape. By dissecting the multi-level interplay between IGFBP2 and the TME, we may reveal a tractable vulnerability that transforms this “door to hell” into a promising gateway for precision cancer therapy.

Figure 1. IGFBP2: Multifaceted Roles in Cancer Progression.

Schematic illustration of the established functions of IGFBP2 effects across major cellular and extracellular compartments in cancer. The central panel depicts the domain organization of IGFBP2. IGFBP2 promotes tumor progression through coordinated effects on four interconnected compartments. (1) Tumor cells, (2) Extracellular matrix (ECM), (3) Cancer-associated fibroblasts (CAFs), and (4) Immune compartment. Collectively, based on current evidence IGFBP2 acts as a central mediator of tumor progression by coordinating tumor-intrinsic signaling, stromal remodeling, and immune suppression. Illustration created in BioRender.

Figure 7. Proposed roles of IGFBP2 in tumor progression.

Schematic illustration of the multifaceted functions of insulin-like growth factor binding protein 2 (IGFBP2) in the tumor microenvironment (TME). (1) Hypoxia and metabolic niche, (2) IGFBP2-driven TME remodeling and, (3) Endothelial/vascular compartment. Our hypothesis is these actions promote angiogenesis, endothelial sprouting, leaky vasculature, and vasculogenic mimicry. Dashed arrows indicate putative or inferred mechanisms based on available evidence in other cancers. Illustration created in BioRender.

Data availability

No data was used for the research described in the article.