SMAD2 and SMAD3 are intracellular proteins that relay signals from the cell surface to the nucleus, acting as the central messengers of the TGF-β signaling pathway. In healthy tissue, they help keep cell growth in check and trigger programmed cell death when needed. In cancer, the same proteins can flip roles and start promoting tumor invasion, immune evasion, and drug resistance. That dual personality makes SMAD2/3 among the most studied and most frustrating targets in cancer biology.
How the Canonical Pathway Works
The story starts at the cell membrane. When TGF-β (transforming growth factor-beta) binds to its receptor, the receptor’s kinase domain activates and directly phosphorylates SMAD2 and SMAD3 at a cluster of serine residues near their tail end. For SMAD3, phosphorylation at this C-terminal SSVS motif is required both for blocking cell proliferation and for switching on target genes like PAI-1, a key regulator of tissue remodeling.1PubMed. Transforming growth factor beta-induced phosphorylation of Smad3 is required for growth inhibition and transcriptional induction in epithelial cells Once phosphorylated, SMAD2 and SMAD3 pair up with a shared partner called SMAD4. The resulting complex moves into the nucleus, where it lands on DNA and turns genes on or off.2PubMed Central. TGF-beta receptor-mediated signalling through Smad2, Smad3 and Smad4
This shuttling between the cytoplasm and the nucleus is not a one-way trip. SMAD2 and SMAD3 continuously cycle in and out of the nucleus, essentially “sampling” whether the receptor is still active. If the TGF-β signal fades, the SMADs lose their phosphorylation and drift back to the cytoplasm, returning the cell to its resting state.3PubMed. Nucleocytoplasmic shuttling of Smads 2, 3, and 4 permits sensing of TGF-beta receptor activity Meanwhile, SMAD4’s own nuclear-export signal gets masked when it is bound to phosphorylated SMAD2 or SMAD3, which keeps the complex locked in the nucleus for as long as the signal persists.4PubMed Central. Regulation of intracellular dynamics of Smad4 by its leucine-rich nuclear export signal The whole system is elegantly self-correcting: the pathway ramps up fast and shuts down fast, giving cells a real-time readout of how much TGF-β is in their environment.
Why SMAD2 and SMAD3 Are Not Interchangeable
Despite being close cousins at the protein level, SMAD2 and SMAD3 behave differently in important ways. The most well-characterized distinction involves DNA binding. The predominant form of SMAD2 carries an extra segment, encoded by exon 3, that sits right in front of its DNA-binding domain and physically blocks direct contact with DNA.5Journal of Biological Chemistry. The DNA Binding Activities of Smad2 and Smad3 Are Regulated by Coactivator-mediated Acetylation SMAD3, which lacks that insert, can grab onto DNA on its own. A naturally occurring splice variant of SMAD2 that skips exon 3 regains the ability to bind DNA and plays a role during embryonic development, but the full-length form needs help from co-activator proteins to associate with target genes.
Recent structural work has added nuance to this picture. The exon 3 insert in SMAD2 is not simply an on/off switch; its three-dimensional shape can shift, and in certain conformations SMAD2 can interact with DNA after all, particularly in the context of transcription factor partners like FOXH1.6PubMed Central. Structural basis for distinct roles of SMAD2 and SMAD3 in FOXH1 pioneer-directed TGF-β signaling So the textbook rule that “SMAD3 binds DNA and SMAD2 does not” is an oversimplification. Both proteins regulate transcription, but they do it through partly different mechanisms and, as a result, partly different gene targets.
The functional divergence shows up dramatically in animal models. Mice that completely lack SMAD2 die during embryonic development, while mice that lack SMAD3 survive to adulthood and, when fed a high-fat diet, are actually protected against obesity. Conditional knockouts zeroing in on fat tissue found that SMAD3 loss selectively reduced subcutaneous fat without affecting the deeper visceral fat pads, whereas removing SMAD2 in the same tissue had little effect on body weight regardless of diet.7PubMed. SMAD2 and SMAD3 differentially regulate adiposity and the growth of subcutaneous white adipose tissue These are strikingly different outcomes from two proteins that receive the same upstream signal.
Tumor Suppression and Its Collapse
In normal epithelial tissue, TGF-β signaling through SMAD2/3 acts as a brake on uncontrolled growth. It slows division, promotes programmed cell death, and helps maintain tissue architecture. Work in prostate epithelial cells illustrates how the two SMADs divide this labor. In stem-cell-like prostate cells, knocking down SMAD2 alone was enough to block TGF-β-induced cell death. In more differentiated cells from the same lineage, SMAD3 was the dominant mediator of apoptosis, while SMAD2 only became relevant when both were silenced together.8PubMed Central. Critical Role of Smad2 in Tumor Suppression and TGF-β-induced Apoptosis of Prostate Epithelial Cells The implication is that SMAD2 may be especially important for killing off undifferentiated or precancerous cells, while SMAD3 handles the same job in cells that have already committed to a particular tissue identity.
When mutations knock out these proteins, the tumor-suppressive brake fails. A large-scale screen of sporadic colorectal cancers found SMAD4 mutations in roughly 9% of tumors, SMAD3 mutations in about 4%, and SMAD2 mutations in about 3%. Both SMAD4 and SMAD3 showed a pattern of two genetic hits per tumor consistent with their acting as classic tumor suppressors: lose one copy, then lose the other, and the growth brake is gone.9PubMed. SMAD2, SMAD3 and SMAD4 mutations in colorectal cancer Most of the identified mutations were predicted to either destabilize the protein or prevent it from forming the SMAD complexes needed for signaling, meaning even a mutation that does not destroy the protein entirely can cripple the pathway.
The Switch to Tumor Promotion
The paradox at the heart of TGF-β biology is that the same pathway that suppresses early tumors often accelerates advanced ones. A big part of that switch involves epithelial-to-mesenchymal transition (EMT), the process by which stationary epithelial cells acquire the ability to move, invade surrounding tissue, and eventually seed metastases. SMAD2/3 are central players here. In colorectal cancer cells exposed to TGF-β from tumor-associated immune cells called macrophages, the SMADs activate a transcription factor named Snail that reprograms cells toward a migratory, invasive state.10Cancer Research and Treatment. Tumor-Associated Macrophages Derived TGF-β‒Induced Epithelial to Mesenchymal Transition in Colorectal Cancer Cells through Smad2,3-4/Snail Signaling Pathway
The EMT-promoting power of SMAD2/3 is not limited to TGF-β itself. In breast cancer cells, epidermal growth factor (EGF), which signals through an entirely different receptor, can phosphorylate SMAD2/3 and trigger EMT through the same Snail-dependent route. Activated SMAD2/3 upregulated mesenchymal markers, suppressed E-cadherin (a protein that glues epithelial cells together), and increased the cells’ ability to migrate and invade.11PubMed Central. EGF induces epithelial-mesenchymal transition through phospho-Smad2/3-Snail signaling pathway in breast cancer cells This means the cancer-promoting side of SMAD2/3 can be engaged even when TGF-β levels are low, as long as other growth factors converge on the same proteins.
Non-TGF-β Signals That Feed Into SMAD2/3
TGF-β is the best-known activator of SMAD2/3, but it is not the only one. Activin-A, a member of the same superfamily, phosphorylates both SMAD2 and SMAD3 through its own set of receptors. In breast cancer, activin-A signaling is ramped up in more advanced tumors, and higher levels of phosphorylated SMAD2 and SMAD3 are found in tumor tissue compared to normal breast. The resulting EMT, migration, and invasion appear to depend particularly on SMAD3.12npj Breast Cancer. Activin-A signaling promotes epithelial–mesenchymal transition, invasion, and metastatic growth of breast cancer
Nodal, another TGF-β superfamily member best known for its role in embryonic development, also works through the SMAD2/3 pathway. In breast cancer cells, Nodal signaling through SMAD2/3 upregulated markers of stemness and boosted the cells’ ability to form tumors and grow as floating spheres in culture, both hallmarks of cancer stem cell behavior.13PubMed Central. Nodal signaling activates the Smad2/3 pathway to regulate stem cell-like properties in breast cancer cells The convergence of multiple ligands onto SMAD2/3 means that blocking TGF-β alone may not be enough to shut down SMAD-driven tumor programs if activin or Nodal are providing backup activation.
Linker Phosphorylation and Pathway Crosstalk
The C-terminal phosphorylation by TGF-β receptors is the canonical way to activate SMAD2/3, but a second layer of regulation happens in the middle “linker” region of these proteins. MAP kinases, which are switched on by a wide variety of growth factors and stress signals, phosphorylate the SMAD linker. The consequences are messy: linker phosphorylation can block the SMADs from entering the nucleus, effectively shutting down TGF-β signaling and potentially pushing cells toward uncontrolled growth.14PubMed Central. Smad linker region phosphorylation in the regulation of extracellular matrix synthesis But in some contexts, linker-phosphorylated SMADs do enter the nucleus, where they drive expression of extracellular matrix components that remodel the tissue around the tumor.
To restore full SMAD activity, phosphatases called small C-terminal domain phosphatases (SCPs) strip the linker phosphorylation away, allowing the SMADs to respond fully to TGF-β. Removing these phosphorylation marks boosted TGF-β-driven gene activation, confirming that linker phosphorylation acts as a dimmer switch on SMAD output.15Journal of Biological Chemistry. Mechanisms of Signal Transduction Small C-terminal Domain Phosphatases Dephosphorylate the Regulatory Linker Regions of Smad2 and Smad3 to Enhance Transforming Growth Factor-β Signaling The interplay between C-terminal activation and linker inhibition turns SMAD2/3 into integrators of multiple incoming signals, not just TGF-β messengers.
SMAD2/3 also talk to the Hippo pathway, another signaling system that controls organ size and cell contact inhibition. In keratinocytes and colon cancer cells, TGF-β stimulation triggered the formation of physical complexes between SMAD2/3 and YAP/TAZ, key effectors of Hippo signaling. When cells were sparse, these complexes accumulated in the nucleus; when cells were packed tightly together, the complexes shifted to the cytoplasm.16PubMed. Crosstalk between Hippo and TGFβ: Subcellular Localization of YAP/TAZ/Smad Complexes This mechanism links mechanical crowding cues to SMAD-driven gene expression and may help explain why the same TGF-β signal produces different outcomes depending on how densely cells are packed.
Shaping the Tumor Microenvironment
Cancer cells do not operate in isolation. Fibroblasts, blood vessels, and immune cells in the surrounding stroma all influence tumor behavior, and SMAD2/3 are deeply involved in shaping that microenvironment. In lung adenocarcinoma, tumor-associated fibroblasts with high SMAD3 levels migrate more aggressively through collagen matrices, not because they chew through the collagen faster but because SMAD3 primes their internal migratory machinery.17British Journal of Cancer. 3D collagen migration patterns reveal a SMAD3-dependent and TGF-β1-independent mechanism of recruitment for tumour-associated fibroblasts in lung adenocarcinoma These fibroblasts are recruited into the tumor stroma and become part of the supportive scaffold that helps tumors thrive.
The balance between SMAD2 and SMAD3 in fibroblasts also dictates how well tumors build their own blood supply. Fibroblasts engineered to have high SMAD3 and low SMAD2 (mimicking the pattern seen in lung adenocarcinoma) secreted factors that boosted endothelial cell migration and blood-vessel formation. High SMAD3 drove increased production of VEGF-A, a powerful pro-angiogenic molecule, along with TIMP-1. The reverse pattern, low SMAD3 with high SMAD2, produced the weakest blood-vessel-promoting signals, resembling the pattern seen in squamous cell lung carcinoma.18Cell Death & Disease. Antagonistic SMAD2/3 control of TIMP-1, VEGF-A, and hypoxia signaling in myofibroblasts shapes histotype-specific angiogenesis in lung cancer In other words, the SMAD2-to-SMAD3 ratio in the stroma can influence whether a tumor is well-supplied with blood vessels or relatively starved.
Immune Evasion Through SMAD3
One of the most clinically significant findings about SMAD3 in recent years involves its connection to immune checkpoint molecules. TGF-β1 directly enhances expression of PD-1, the receptor on T cells that, when engaged, tells the immune cell to stand down. This upregulation works through SMAD3 and is independent of SMAD2. In mouse tumor models, tumor-infiltrating T cells showed SMAD3-dependent increases in PD-1, providing a direct link between SMAD3 activity and the ability of tumors to hide from the immune system.19PubMed Central. TGF-β1-mediated Smad3 enhances PD-1 expression on antigen-specific T cells in cancer This finding has implications for immunotherapy: patients whose tumors have high TGF-β/SMAD3 activity may show less response to PD-1-blocking drugs, and combining TGF-β pathway inhibitors with checkpoint immunotherapy is an active area of clinical investigation.
Metabolic Reprogramming Under Low Oxygen
Tumors frequently outgrow their blood supply, creating zones of low oxygen. Under these hypoxic conditions, a transcription factor called HIF-1α accumulates and redirects cellular metabolism toward glycolysis, the less efficient but oxygen-independent way of generating energy. HIF-1α physically binds to the same domain on phosphorylated SMAD3 that normally interacts with SMAD4, essentially hijacking the SMAD3 complex and redirecting it toward glycolytic gene targets in non-small cell lung cancer cells.20PubMed Central. HIF-1α switches the functionality of TGF-β signaling via changing the partners of smads to drive glucose metabolic reprogramming in non-small cell lung cancer This partner-swapping mechanism shows how the same SMAD3 protein can be co-opted for entirely different jobs depending on the cellular environment. In well-oxygenated tissue, SMAD3 partners with SMAD4 to regulate growth. In a hypoxic tumor, it pairs with HIF-1α to fuel the metabolic rewiring that helps cancer cells survive.
Epigenetic Control of SMAD Activity
Whether SMAD2/3 can actually turn on their target genes once they arrive in the nucleus depends heavily on the chromatin landscape they encounter. SMAD2 recruits the enzyme p300, which acetylates specific positions on histone H3 (lysines 9 and 18), opening up the chromatin around target genes. SMAD2 also pulls in Brg1, the motor subunit of a chromatin-remodeling complex, in a TGF-β-dependent manner. Both of these co-factors are required for SMAD2-dependent genes to be turned on.21PubMed Central. Smads orchestrate specific histone modifications and chromatin remodeling to activate transcription
The practical upshot is that two cells receiving the exact same TGF-β dose can respond completely differently if their chromatin is marked differently. In breast tumor-initiating cells, regions with activating histone marks allowed SMAD3-dependent gene expression to proceed, while regions silenced by DNA methylation or repressive histone marks blocked it entirely.22Cell Reports. Epigenome-Directed Regulation of TGF-β Responses in Breast Tumor-Initiating Cells This helps explain why TGF-β can be tumor-suppressive in some cell populations and tumor-promoting in others within the same patient: the epigenetic state of the cell determines which SMAD target genes are accessible and which are locked away.
Non-Coding RNA and the SMAD3 Connection
The regulation of SMAD3 expression itself is subject to control by non-coding RNAs, a layer of biology that has attracted increasing attention. In bladder cancer, a long non-coding RNA called LINC02470 acts as a molecular sponge for microRNA-143-3p. Under normal circumstances, miR-143-3p would suppress SMAD3 translation, keeping its levels in check. When LINC02470 mops up the microRNA, SMAD3 protein levels rise, EMT is activated, and the cancer cells become more aggressive, with enhanced migration, invasion, and the ability to form tumors in mice.23PubMed Central. Long Noncoding RNA LINC02470 Sponges MicroRNA-143-3p and Enhances SMAD3-Mediated Epithelial-to-Mesenchymal Transition to Promote the Aggressive Properties of Bladder Cancer This “competing endogenous RNA” mechanism represents yet another way that SMAD3 levels can be dysregulated in tumors without any mutation in the SMAD3 gene itself.
Drug Resistance and Cancer Stem Cells
Sustained SMAD3 activation is emerging as a driver of resistance to targeted therapies. In HER2-positive breast cancer, continuous TGF-β/SMAD3 signaling induced resistance to anti-HER2 drugs including trastuzumab and promoted cancer stem cell traits. Nuclear SMAD3 levels in tumor tissue inversely correlated with sensitivity to trastuzumab-based treatment. A small-molecule SMAD3 inhibitor called SIS3 not only prevented resistance from developing but also resensitized already-resistant cells to trastuzumab.24PubMed. A small-molecule inhibitor of SMAD3 attenuates resistance to anti-HER2 drugs in HER2-positive breast cancer cells
A related mechanism operates in liver cancer, where cyclin D1 overexpression drives stemness by directly interacting with and activating SMAD2/3 and SMAD4. The resulting signaling loop maintains liver cancer stem cell self-renewal and chemoresistance. Blocking SMAD activity with a small-molecule inhibitor forced these stem-like cells to differentiate, making them sensitive to conventional chemotherapy again.25PubMed Central. Smad inhibitor induces CSC differentiation for effective chemosensitization in cyclin D1- and TGF-β/Smad-regulated liver cancer stem cell-like cells Together, these findings paint a picture in which SMAD-driven stem cell programs are not just a curiosity but a practical barrier to treatment.
SMAD Phosphorylation as a Biomarker
Given the dual nature of SMAD2/3 signaling, measuring their activation status in patient tumors is an appealing way to predict outcomes. But the reality is complicated, because the same phosphorylation event can mean different things depending on where in the tissue it occurs. In non-small cell lung cancer, SMAD2 linker phosphorylation in tumor cells was associated with an increased risk of death or relapse. But the same modification in immune cells within the tumor stroma was associated with better outcomes.26British Journal of Cancer. SMAD2 linker phosphorylation impacts overall survival, proliferation, TGFβ1-dependent gene expression and pluripotency-related proteins in NSCLC That spatial dependence means a simple bulk measurement of phospho-SMAD2 in a biopsy could wash out the signal entirely; you need to know which cells are phosphorylated to draw any useful conclusion.
In early-stage breast cancer, C-terminal phosphorylated SMAD2 has been explored as a prognostic marker. Tumors with intermediate levels of phospho-SMAD2 had a higher risk of recurrence compared to tumors with high levels, consistent with the idea that robust TGF-β signaling still plays a tumor-suppressive role in early disease.27PubMed Central. Tumoral pSMAD2 as a prognostic biomarker in early-stage breast cancer: insights from the randomized SweBCG91RT trial As cancers progress and the pathway shifts toward pro-tumorigenic functions, the interpretation of phospho-SMAD levels flips accordingly. Any useful biomarker strategy will have to account for cancer stage, phosphorylation site (C-terminal versus linker), and the specific cell type being measured.
Therapeutic Targeting of the Pathway
The most straightforward way to block SMAD2/3 activation is to inhibit the kinase that phosphorylates them, the TGF-β type I receptor (also called ALK5). ALK5 is an attractive drug target because it sits at the top of the pathway and is highly specific to TGF-β signaling. Several potent, selective ALK5 inhibitors have been developed that fit into the receptor’s ATP-binding pocket.28PubMed. Tgf-beta type I receptor (Alk5) kinase inhibitors in oncology The prototype compound SB-431542, widely used in laboratory research, blocks TGF-β-induced transcription, gene expression, apoptosis, and growth suppression.29PubMed Central. A specific inhibitor of TGF-beta receptor kinase, SB-431542, as a potent antitumor agent for human cancers
The challenge is obvious: if TGF-β/SMAD signaling is tumor-suppressive in early cancers and tumor-promoting in late ones, a drug that blocks the whole pathway risks removing the brake in patients whose tumors have not yet made the switch. The emerging focus on SMAD3-specific inhibitors like SIS3, along with strategies that target the pathway’s pro-tumorigenic outputs (EMT, immune evasion, angiogenesis) rather than the receptor itself, reflects an attempt to thread that needle. Combination approaches that pair ALK5 inhibitors with immune checkpoint blockade are particularly interesting, given the direct role of SMAD3 in upregulating PD-1 on tumor-infiltrating T cells. Whether these combinations can deliver durable benefits without unacceptable toxicity is one of the defining open questions in TGF-β therapeutics.