Membrane receptors serve as the initiating nodes of cancer signaling networks by sensing extracellular and microenvironmental cues and transmitting signals intracellularly to drive tumor cell proliferation, migration, survival and other malignant behaviors. Major receptor classes include receptor tyrosine kinases [e.g., EGFR, human EGFR 2 (HER2)], G protein-coupled receptors, integrins, immune checkpoint receptors [e.g., programmed cell death 1 (PD-1)/programmed death ligand 1 (PD-L1)] and cytokine receptors (e.g., TNF and Wnt receptors). Upon ligand binding, these receptors activate downstream cascades, particularly the MAPK, PI3K/Akt and JAK/STAT pathways, thereby promoting tumorigenesis and disease progression (9–12). Aberrant EGFR and HER2 activation is frequently observed across multiple cancers (13,14), while increased PD-1/PD-L1 signaling contributes to tumor immune evasion (15). As central regulatory nodes, membrane receptors are primary targets in precision oncology, with targeted inhibitors and monoclonal antibodies demonstrating substantial clinical benefit across diverse malignancies (16).

Signal transducers act as critical intermediates between membrane receptors and downstream effectors; they integrate, amplify and diversify signaling inputs to dictate cellular responses to extracellular stimuli (17). These molecules fall into four major categories.

G proteins are categorized as heterotrimeric G proteins, which couple to G protein-coupled receptors, and small GTPases, including members of the Ras and Rho families (18). Heterotrimeric G proteins regulate intracellular second messengers such as cAMP, inositol 1,4,5-trisphosphate and diacylglycerol (19), thereby modulating cellular metabolism, ion channel activity and gene expression (20–23). Small GTPases are central nodes in cancer signaling: Ras proteins act as key signaling hubs, where oncogenic mutations often trigger constitutive MAPK/ERK and PI3K/Akt activation, promoting proliferation, survival and metastasis (24,25). By contrast, Rho family proteins primarily regulate cytoskeletal dynamics and cell motility, contributing to cancer invasion and metastatic potential (26).

Kinase cascades, particularly the MAPK and PI3K/Akt/mTOR pathways, amplify and diversify signaling through sequential phosphorylation events (27,28). These pathways regulate key biological processes, with MAPK being primarily involved in proliferation, differentiation and stress responses (29,30), and PI3K/Akt in cell survival, metabolism and therapeutic resistance (31–33). Their multi-layered structure and extensive branching not only confer flexibility in signal processing but also foster tumor heterogeneity and treatment resistance when persistently activated.

Signal integration proteins organize signaling pathways in space and time, including scaffold proteins such as receptor for activated C kinase 1, A-kinase anchoring proteins and IQ motif-containing GTPase-activating protein 1, as well as adaptor proteins such as growth factor receptor-bound protein 2 (Grb2), Src homology 2 domain-containing transforming protein and Grb2-associated binder 1/2 (34–36). By assembling signaling complexes, these molecules enhance signal transduction efficiency and facilitate pathway crosstalk. Certain proteins, such as β-arrestin and Grb2 (Table I), function as both scaffolds and adaptors to coordinate signaling (37,38). Dysregulated proteins disrupt signaling homeostasis, triggering malignancy. Representative scaffold and adaptor proteins involved in signaling integration, together with their associated pathways and tumor-related functions, are summarized in Table I (39–48).

Regulatory enzymes and negative regulators critically balance signaling. These include phosphatases such as phosphatase and tensin homolog (PTEN), as well as ubiquitin ligases and SUMOylation enzymes (49,50). These molecules limit persistent pathway activation and preserve network homeostasis. Loss or inactivation of PTEN leads to uncontrolled amplification of oncogenic signaling and represents a key event in cancer progression (51,52).

Downstream effectors are the final mediators that execute the biological consequences of signaling activation. They include transcription factors, cell cycle and apoptosis regulators, metabolic enzymes, cytoskeletal proteins and immune modulators. They collectively control processes such as proliferation, apoptosis, differentiation, metabolism, migration and immune responses (7). Specifically, transcription factors including NF-κB, c-Myc and β-catenin regulate tumor-associated gene expression (53,54), while proteins such as Bcl-2 family members, caspases and p53 control apoptosis and cell cycle progression (55–57). Metabolic regulators such as mTOR and HIF-1α support cellular adaptation (58), while Rho GTPases and Fascin mediate cytoskeletal remodeling and invasion (59,60). Additionally, PD-L1 plays a pivotal role in tumor immune evasion (61). Dysregulated effectors generate malignant phenotypes, drive therapeutic resistance and often yield poor clinical outcomes (62,63). These downstream consequences arise not from isolated signaling events but from the coordinated activation of oncogenic pathways operating within interconnected networks.

The PI3K/Akt/mTOR pathway centrally regulates cancer cell proliferation, survival, metabolic reprogramming and therapeutic resistance. Activated downstream of receptor tyrosine kinases and G protein-coupled receptors, sequential activation of PI3K, Akt and mTOR promotes cell growth while suppressing apoptosis (64–68). Aberrant activation of this pathway is frequent in cancer and is often driven by PTEN loss or activating mutations in phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit α (PIK3CA), resulting in constitutive output that supports tumor progression and resistance to therapy (69–71). Its clinical relevance is particularly notable in breast cancer, endometrial cancer and glioblastoma, where PI3K alterations are prevalent. Given its central role, the PI3K/Akt/mTOR pathway has become a major target in precision oncology and its mechanistic inhibition may relieve negative feedback on upstream receptor tyrosine kinases, thereby triggering compensatory activation of parallel signaling cascades, particularly MAPK/ERK (72). This adaptive reprogramming sustains tumor survival, limits the durability of single-pathway inhibition and drives therapeutic resistance (73).

The MAPK/ERK cascade, organized around the Ras-Raf-MEK-ERK kinase axis, regulates multiple biological processes, including proliferation, differentiation, apoptosis and migration (74,75). Oncogenic alterations, particularly mutations in Ras or BRAF, frequently trigger constitutive MAPK/ERK signaling, thereby driving tumor growth, metastasis and therapeutic resistance; accordingly, the clinical importance of this pathway is especially evident in melanoma and colorectal cancer, where Ras or BRAF mutations often function as primary oncogenic drivers (76–81).

Notably, the MAPK/ERK and PI3K/Akt/mTOR pathways are intimately interconnected. Their extensive crosstalk coordinates control of proliferation, survival and stress responses. In numerous tumor contexts, inhibiting one pathway can induce compensatory activation of the other, thereby maintaining downstream signaling output and reducing therapeutic efficacy (82). This reciprocal regulation is a key mechanism of treatment resistance, providing a strong rationale for pharmacological combination therapies targeting both axes. Often, this compensatory response is mediated by the relief of ERK-dependent negative feedback on upstream nodes, which restores signaling after targeted inhibition (83).

The Wnt/β-catenin pathway is a vital regulator of cancer cell stemness, self-renewal and metastatic potential. Upon signaling, β-catenin accumulates in the cytoplasm and translocates to the nucleus to drive the expression of genes involved in proliferation and epithelial-mesenchymal transition (EMT) (84–88). Aberrant Wnt activation has been implicated in colorectal cancer and other tumor contexts, where it contributes to tumor initiation, progression and the maintenance of stem-like and aggressive phenotypes (89,90). Beyond its tumor-intrinsic effects, the Wnt/β-catenin axis modulates the tumor microenvironment and immune responses, further supporting tumor progression (91,92). Mechanistically, stabilized nuclear β-catenin promotes the transcription of EMT-related genes, such as Snail and c-Myc, enhancing tumor invasion, metastasis and therapeutic resistance (93). These features position the pathway as a key candidate for combination-based therapeutic strategies.

In addition to the major pathways described, several other signaling cascades regulate cancer progression, including the JAK/STAT, Notch, Hippo/Yes-associated protein (YAP), NF-κB and Hedgehog pathways. These regulate diverse processes, such as inflammation, immune modulation, cell fate determination and stemness, in a context-dependent manner. For instance, JAK/STAT and NF-κB signaling link inflammation, immune responses to therapeutic resistance (94–98), whereas Notch and Hippo/YAP primarily influence cell fate decisions and cellular plasticity (99,100). Aberrant Hedgehog signaling further promotes tumor proliferation and metastasis, particularly in cancers reactivating developmental programs (101–103). Importantly, these pathways are integrated through extensive crosstalk and feedback regulation, forming a dynamic network with context-dependent outputs. Such interactions, mediated by shared signaling intermediates and transcriptional programs, coordinate regulation of inflammation, immune evasion and tumor cell plasticity (104).

Signaling events are tightly regulated spatiotemporally, with their activation, propagation and termination restricted within specific cellular compartments and temporal windows. For example, receptor tyrosine kinases frequently localize to membrane microdomains or lipid rafts for activation (106), while the subcellular localization and nucleo-cytoplasmic shuttling of proteins such as NF-κB, β-catenin and STAT critically influence downstream gene expression and cellular phenotypes. Furthermore, signaling pathways may display transient or sustained activation patterns, as exemplified by the MAPK/ERK cascade (107–110), triggering distinct cellular outcomes ranging from proliferation to differentiation or apoptosis.

A defining feature of cancer signaling networks is their capacity for dynamic adaptation under stress. In response to environmental changes or therapeutic pressure, cancer cells can reprogram signaling activity to sustain survival. Inhibiting dominant pathways, such as PI3K/Akt, often triggers compensatory activation of alternative cascades, including MAPK, Wnt or JAK/STAT (111–113). This adaptive reprogramming may arise from altered protein expression, pathway switching or regulatory mutations (114), representing a central mechanism for drug resistance, metastasis and tumor heterogeneity (115,116). In addition, resistance emerges from pathway reactivation, activation of parallel signaling routes or microenvironment-mediated survival signals (117).

Fig. 1 outlines the signaling hierarchy and directionality from membrane receptors to downstream effectors, while Fig. 2 illustrates the mechanistic basis of pathway crosstalk, therapeutic inhibition and adaptive responses under treatment pressure. Specifically, inhibiting dominant hubs such as PI3K/Akt/mTOR can induce compensatory activation of parallel routes, notably MAPK/ERK and Wnt/β-catenin, maintaining downstream signaling and promoting therapeutic resistance. These adaptive responses often stem from pathway reactivation, bypass signaling or microenvironment-derived cues such as cytokines and growth factors, which collectively shield signaling activity from therapeutic inhibition (118).

Cancer signaling networks possess substantial topological complexity, incorporating branched pathways, positive and negative feedback loops, feedforward regulation and redundant signaling routes. Key nodes such as Ras, Akt and β-catenin serve as central hubs that integrate and distribute signals across multiple axes (8,119). While this complex architecture enhances cellular adaptability under stress, it also enables compensatory signaling and therapeutic escape (120).

Certain signaling circuits exhibit oscillatory or bistable behaviors in response to external stimuli, as seen in the p53-MDM2 and NF-κB-IκB regulatory loops (121,122). These dynamic patterns trigger switch-like responses or periodic signaling activity, enabling flexible transitions between cellular states. Such behaviors drive tumor heterogeneity, cellular adaptability and the development of therapeutic resistance.

Signaling activation is closely linked to epigenetic regulation, including chromatin remodeling, histone modification and DNA methylation (123,124). These processes can encode prior signaling events, generating a form of epigenetic ‘memory’ that stabilizes tumor phenotypes even after the initial stimulus dissipates (125).

Simultaneously, cancer signaling networks continuously engage with the tumor microenvironment (104). Immune cells, fibroblasts and other stromal components can modulate pathway activity, while tumor cells reciprocally remodel their surroundings. These bidirectional interactions further promote tumor progression, immune evasion and adaptive responses to therapy (126). Tumor microenvironment-mediated resistance is also an important cause of treatment failure. In addition to tumor cells themselves, fibroblasts, immune cells, endothelial cells, extracellular matrix components, hypoxia and soluble cytokines can all influence pathway activity during therapy (117,127). For instance, cancer-associated fibroblasts may release hepatocyte growth factor, which activates MET signaling and downstream PI3K/Akt and MAPK pathways, thereby reducing the efficacy of EGFR tyrosine kinase inhibitors (128). Cytokines such as IL-6 and TGF-β can also activate JAK/STAT3, SMAD or EMT-related signaling, promoting tumor cell survival, immune escape and drug resistance (95,97,129). Furthermore, extracellular matrix remodeling may strengthen integrin-FAK/Src signaling (130), while hypoxia can induce HIF-1α-dependent programs related to angiogenesis, metabolic adaptation and treatment resistance (131). These findings suggest that effective treatment may need to target both tumor-intrinsic oncogenic pathways and protective signals from the surrounding microenvironment.

The biological relevance of cancer signaling networks lies in their coordinated ability to shape malignant phenotypes and remodel the tumor microenvironment. Dysregulation of these networks drives key features of cancer, including uncontrolled proliferation, resistance to apoptosis, invasion, metastasis and metabolic reprogramming (132,133). Beyond these tumor-intrinsic effects, these networks reconfigure surrounding stromal and immune components through cytokines, exosomes and other mediators (134).

For example, persistent PI3K/Akt/mTOR and MAPK/ERK activation promotes tumor growth and therapeutic resistance (135–137), while Wnt/β-catenin signaling enhances stemness and metastatic potential (138,139). These alterations can upregulate immunosuppressive molecules, notably PD-L1 (140), and reprogram tumor-associated stromal cells to reinforce immune evasion, angiogenesis and tumor aggressiveness (141). Notably, although many signaling principles are shared across cancers, the phenotypic and therapeutic consequences of dysregulation are highly context-dependent, shaped by lineage, genomic background and the microenvironment. These differences ultimately regulate the efficacy, durability and resistance patterns of targeted therapies.

From a clinical perspective, the significance of cancer signaling networks lies in their coordinated behavior, rather than in isolated oncogenic drivers. Traditional approaches targeting single signaling nodes such as EGFR, PI3K or BRAF have yielded substantial clinical benefits in selected patient populations (142,143). Representative examples include EGFR tyrosine kinase inhibitors for EGFR-mutant non-small cell lung cancer, BRAF/MEK inhibitors for BRAF-mutant melanoma and colorectal cancer, PI3K pathway inhibitors for biomarker-selected tumors and immune checkpoint inhibitors for tumors with specific immune-related biomarkers. In the case of PI3K pathway inhibitors, patients are usually selected according to molecular alterations in the PI3K/Akt/mTOR axis, such as activating PIK3CA mutations, PTEN loss, or other evidence of pathway activation (144). For example, PI3K inhibitors have been used in hormone receptor-positive, HER2-negative, PIK3CA-mutated advanced breast cancer (145). For immune checkpoint therapy, ‘high immune regulatory signaling’ mainly refers to clinically used biomarkers, including increased PD-L1 expression, microsatellite instability-high status, mismatch repair deficiency, high tumor mutational burden or an inflamed tumor microenvironment (146). However, these biomarkers are not perfect predictors, and their clinical value may differ among tumor types and treatment settings (147).

However, the clinical benefit of single-target inhibition is often limited by the adaptive nature of cancer signaling networks (148). In many tumor contexts, suppression of a dominant node fails to fully extinguish oncogenic output, as parallel pathways are reactivated or newly engaged. A representative example is the reciprocal crosstalk between the PI3K/Akt/mTOR and MAPK/ERK axes, where inhibition of one axis can relieve negative feedback or trigger bypass signaling through the other, sustaining proliferation and survival (149,150). Furthermore, intratumoral heterogeneity and microenvironment-mediated survival signals may promote treatment resistance by allowing resistant cell states to persist or emerge under therapeutic pressure. These observations underscore that cancer signaling networks function as integrated systems rather than isolated pathways.

Accordingly, a network-based therapeutic framework has emerged, guiding strategies by pathway interactions and dynamic network behavior. Key approaches include targeting dominant oncogenic drivers such as EGFR, PI3K or BRAF; co-inhibition of parallel pathways to prevent compensatory signaling, notably combined targeting of PI3K and MAPK; vertical inhibition within signaling cascades to ensure sustained pathway suppression; and integration of targeted therapy with immunotherapy or conventional treatments to modulate both tumor-intrinsic and microenvironmental signaling (151). From a translational and pharmacological perspective, these strategies aim not only to suppress primary oncogenic drivers, but to preempt or overcome the pathway reactivation, bypass signaling and adaptive reprogramming that frequently emerge during treatment.

Despite these advances, major challenges remain, including intratumoral heterogeneity, acquired resistance and adaptive network responses (152). Addressing these challenges will require therapeutic strategies accounting for the dynamic and context-dependent signaling networks. A deeper understanding of network-level regulation and its interaction with the tumor microenvironment is warranted for advancing precision oncology with improved long-term clinical outcomes. Mechanistically, resistance to targeted therapies often arises from feedback reactivation of inhibited pathways or compensatory activation of parallel signaling networks, underscoring the need for rational combination therapies based on network-level interactions (153).