Atlas One · Molecular / Signaling · Scale L02 · Pleiotropic Cytokine

TGF-β — Transforming Growth Factor Beta

A pleiotropic cytokine superfamily whose most abundant member, TGF-β1, orchestrates wound healing, fibrosis, immune suppression, epithelial-mesenchymal transition, and paradoxically acts as both a tumor suppressor in early carcinogenesis and a tumor promoter in advanced cancer — driven primarily through SMAD2/3 canonical signaling and multiple non-SMAD pathways.

Gene: TGFB1 (chr 19q13.2) Scale: L02 Molecular / Signaling System: Multi-system Type: Cytokine / growth factor Reviewed: 2026-06-05

TGF-β Superfamily and Structure

The TGF-β superfamily is the largest family of secreted signaling proteins in metazoans, comprising 33 members in humans. Members share a conserved C-terminal "cystine knot" growth factor domain (~110 amino acids, 7 invariant cysteines) and signal through heteromeric serine/threonine kinase receptor complexes. Major branches:

SubfamilyMembersKey Functions
TGF-βTGF-β1, TGF-β2, TGF-β3Fibrosis, immune suppression, wound healing, EMT; all three signal via ALK5/TβRII → SMAD2/3. TGF-β1 most abundant and widely expressed; TGF-β3 important in palatogenesis (cleft palate when absent); TGF-β2 in eye/cardiac development
BMPs (bone morphogenetic proteins)BMP2, BMP4, BMP5–9, BMP15 (20 members)Osteogenesis, cartilage formation, neural patterning (BMP4 — dorso-ventral axis), vascular regulation (BMP9/10 — hepatic ligands for ALK1/ENG); signal via ALK1/2/3/6 → SMAD1/5/8
ActivinsActivin A, B, AB, C, EFSH secretion (gonadotropin regulation), erythropoiesis, wound healing, inflammatory modulation; signal via ALK4 → SMAD2/3
GDFs (growth and differentiation factors)GDF5/6/7 (joint/tendon formation), GDF8/myostatin (muscle mass suppressor), GDF11 (aging), GDF15 (stress cytokine, cachexia)Musculoskeletal development; GDF8/myostatin inhibits skeletal muscle growth via ActRIIB → ALK4/5 → SMAD2/3; GDF15 elevated in cancer cachexia, heart failure, metformin treatment
AMH / MISAnti-Müllerian hormoneRegression of Müllerian ducts in male fetal development; clinical marker for ovarian reserve
NodalNodalLeft-right body axis determination, mesoderm induction; signals via ALK4/7 + EGF-CFC co-receptor → SMAD2/3

TGF-β1 Protein Structure and Latency

TGF-β1 is synthesized as a 390 amino acid precursor (pre-pro-TGF-β1). The signal peptide is cleaved, yielding pro-TGF-β1; furin in the trans-Golgi cleaves the prodomain (latency-associated peptide, LAP) from the growth factor domain, but the two remain noncovalently associated as the small latent complex (SLC) — a disulfide-linked homodimer of mature TGF-β1 with its LAP. LAP wraps around the growth factor domain and sterically blocks receptor binding. The SLC is additionally disulfide-linked to LTBP-1, -3, or -4 (latent TGF-β binding proteins) to form the large latent complex (LLC), which is secreted and sequestered in the ECM via LTBP's fibronectin/fibrillin-1 binding domains.

Latent TGF-β activation — the step that determines when and where TGF-β signals — occurs via multiple mechanisms:

Canonical SMAD Signaling

The canonical TGF-β pathway is defined by the activation of receptor-regulated SMADs (R-SMADs) following receptor kinase activation:

Receptor Complex Assembly and Kinase Activation

Active TGF-β1 homodimer binds the TGF-β type II receptor (TβRII) — a constitutively active serine/threonine kinase. TβRII-TGF-β1 complex recruits and transphosphorylates the TGF-β type I receptor (TβRI/ALK5) in the GS domain (Gly-Ser rich juxtamembrane region) at Ser/Thr residues. Active ALK5 phosphorylates R-SMADs.

Different type I receptors (ALKs) specify downstream SMAD pathway:

SMAD Activation and Nuclear Signaling

ALK5 phosphorylates SMAD2 at C-terminal SxS motif (Ser465/Ser467) and SMAD3 (Ser423/Ser425). Phosphorylated SMAD2/3 dissociate from the receptor → form heterotrimeric complexes with the co-SMAD SMAD4 → nuclear import via importin-β3/α → binding to SMAD binding elements (SBE; CAGACA consensus) in target gene promoters → recruitment of co-activators (p300/CBP, PCAF histone acetyltransferases) or co-repressors (HDAC, Ski/SnoN) depending on cellular context.

Key TGF-β/SMAD target genes in fibrosis: COL1A1 (collagen Iα1), COL3A1, ACTA2 (α-SMA), CTGF/CCN2, FN1 (fibronectin), SERPINE1 (PAI-1), MMP-2/9 (paradoxically both induced and later suppressed by SMAD).

Key TGF-β/SMAD target genes in growth arrest: CDKN2B/p15 (CDK4/6 inhibitor) and CDKN1A/p21 (CDK2 inhibitor) — both strongly induced by SMAD3 → G1 cell cycle arrest. This is the tumor suppressor function of TGF-β.

Inhibitory SMADs — Negative Feedback

SMAD7 is the primary inhibitor of TGF-β/SMAD signaling. SMAD7 is a transcriptional target of TGF-β/SMAD3 itself — creating a negative feedback loop. SMAD7 acts by: (1) competing with R-SMADs for receptor interaction, preventing phosphorylation; (2) recruiting SMURF1 and SMURF2 (E3 ubiquitin ligases) to TβRI → ubiquitin-mediated receptor degradation; (3) recruiting PP1c phosphatase to dephosphorylate TβRI. In IPF and fibrotic disease, SMAD7 expression is reduced, removing this brake and sustaining TGF-β signaling. SMAD6 preferentially inhibits BMP/SMAD1/5/8 signaling via analogous mechanisms.

SMAD paradox in cancer: SMAD4 is a tumor suppressor gene mutated or deleted in ~50% of pancreatic ductal adenocarcinomas (PDACs) and ~10–15% of colorectal cancers. Loss of SMAD4 abolishes TGF-β growth arrest (loss of p15/p21 induction) while preserving or even enhancing non-SMAD pro-tumorigenic signaling — explaining why TGF-β becomes a tumor promoter rather than suppressor in cancers with SMAD4 loss.

Non-SMAD Signaling Pathways

TGF-β activates multiple non-SMAD pathways, primarily through TRAF6 (TNFR-associated factor 6) recruited to the receptor complex and through direct activation of kinases by TβRI/TβRII:

PathwayMechanismFunctional Outcome
TRAF6/TAK1 → NF-κBTβRI binds TRAF6 → TRAF6 Lys63-polyubiquitylation → TAK1 (MAP3K7) activation → IKK complex → IκBα phosphorylation/degradation → NF-κB nuclear translocationPro-inflammatory gene expression; survival signals; context-dependent induction of IL-6, TNF-α, MMP-9. Relevant in epithelial cells and macrophages where TGF-β drives inflammatory phenotypes despite SMAD-mediated growth arrest
TAK1 → p38 MAPK / JNKTAK1 → MKK3/6 → p38; TAK1 → MKK4/7 → JNKp38: stress-responsive fibroblast activation, EMT (α-SMA induction), apoptosis in epithelial cells. JNK: EMT transcription factor (Snail, Twist) induction; TGF-β-induced apoptosis via Bim/Bad phosphorylation. p38 and JNK often cooperate with SMAD3 to maximize EMT gene responses
PI3K/Akt/mTORTβRI/II directly phosphorylates and activates PI3K regulatory subunit p85 → Akt → mTORC1. Also indirect: TβRI → TRAF6 → PI3KCell survival (anti-apoptotic via BAD phosphorylation, Mcl-1 upregulation); cell migration and invasion; mTORC1-dependent protein synthesis for fibrotic ECM production. PI3K/Akt cooperates with SMAD2/3 in EMT and resistance to TGF-β-induced apoptosis in cancer cells
Rho/ROCKTGF-β → RhoA GTP loading (via direct receptor interaction or SMAD-independent mechanisms) → ROCK1/2 → LIM kinase → cofilin phosphorylation (inactivation) → F-actin polymerization; also ROCK → MLC phosphorylation → increased cell contractilityStress fiber formation (α-SMA incorporation), cell contraction (myofibroblast phenotype), cell migration and invasion (EMT), mechanosensing amplification. ROCK inhibitors (fasudil, Y-27632) reduce TGF-β-driven fibrosis in animal models
Src/ERKTβRII contains a Src-homology 2 (SH2) binding motif → direct Src recruitment and activation → Grb2/Sos → Ras → Raf → MEK → ERK1/2Cell proliferation signals (paradoxically co-exist with SMAD-mediated growth arrest in non-epithelial cells); synergy with EGFR signaling in cancer; regulation of SMAD nucleo-cytoplasmic shuttling via ERK-mediated SMAD linker phosphorylation (at Thr179/Ser204/Ser208 of SMAD3) → attenuated transcriptional activity
β-catenin/Wnt crosstalkTGF-β/SMAD3 complex can directly bind β-catenin in the nucleus → cooperative gene activation; SMAD3 also stabilizes β-catenin via inhibition of GSK-3β (APC/Axin/GSK-3β destruction complex)Fibroblast-to-myofibroblast transition; AT2 cell fate decisions in lung injury; Wnt/β-catenin and TGF-β cooperatively drive fibrosis and EMT in multiple organs. IPF patient lungs show co-activation of Wnt and TGF-β targets

Pleiotropic Biological Functions

Anti-Proliferative / Tumor Suppressor Activity

In normal epithelial cells, TGF-β induces G1 cell cycle arrest via SMAD3-dependent transcriptional induction of p15/CDKN2B (inhibits CDK4/6 → prevents pRb phosphorylation → E2F remains bound to pRb → no S-phase entry) and p21/CDKN1A (inhibits CDK2 → reinforces G1 arrest). Simultaneously, SMAD3 represses c-Myc and Id proteins that drive proliferation. This makes TGF-β a potent inhibitor of epithelial cell proliferation — the basis for its tumor-suppressor role in early carcinogenesis.

Epithelial-Mesenchymal Transition (EMT)

TGF-β is the canonical inducer of EMT — the process by which epithelial cells lose apical-basal polarity, disrupt cell-cell junctions (E-cadherin/CDH1 downregulation), and acquire a mesenchymal gene expression program (N-cadherin/CDH2, vimentin/VIM, fibronectin/FN1, α-SMA/ACTA2). EMT transcription factors induced by TGF-β/SMAD3 + non-SMAD pathways:

Immune Suppression and Treg Induction

TGF-β is the dominant cytokine for peripheral immune tolerance:

Wound Healing and Fibrosis

TGF-β1 (released by platelets in the initial injury response) orchestrates wound healing: fibroblast recruitment → myofibroblast differentiation (α-SMA, collagen production, wound contraction) → VEGF induction (angiogenesis) → provisional matrix remodeling. In normal healing, TGF-β signaling is self-limited by SMAD7 feedback and myofibroblast apoptosis as the wound closes. In fibrosis, persistent TGF-β1 signaling — from ongoing injury, ER stress, senescent cells, or autocrine loops — prevents myofibroblast apoptosis and drives progressive ECM accumulation (see fibrosis.html).

TGF-β Paradox in Cancer

TGF-β exhibits a striking contextual duality in cancer — tumor suppressor in early carcinogenesis vs. tumor promoter in advanced cancer:

Tumor Suppressor Phase (Early)

In normal epithelium and early-stage cancer with intact SMAD4, TGF-β enforces G1 arrest (p15/p21 induction), induces apoptosis, and suppresses oncogene-driven proliferation. SMAD4 mutations (loss of heterozygosity 18q21.2) occur in ~55% of PDACs and ~15% of CRCs — inactivating the SMAD4 co-SMAD disrupts all canonical TGF-β growth-arrest signals. Inactivating mutations in TβRII (microsatellite instability-driven frameshift mutations in HNPCC/Lynch syndrome colorectal cancers) similarly abolish growth arrest without eliminating pro-tumorigenic TGF-β responses.

Tumor Promoter Phase (Advanced)

In advanced tumors with KRAS mutations, PI3K activation, or SMAD4 loss, TGF-β no longer arrests growth but retains and may amplify pro-tumorigenic functions:

TGF-β as an immunotherapy resistance mechanism: TGF-β in the TME suppresses CD8+ T cell effector function and prevents T cell infiltration into "immune-excluded" tumors — a major mechanism of resistance to PD-1/PD-L1 checkpoint blockade. Bintrafusp alfa (bifunctional anti-PD-L1/TGF-β trap) was designed to simultaneously block PD-L1 and neutralize TGF-β at the tumor surface. JAVELIN Bladder 100 and NSCLC trials showed this approach is feasible but efficacy improvements over anti-PD-L1 monotherapy have been modest, suggesting complex biology.

Therapeutic Targeting of TGF-β

AgentClass / TargetMechanismClinical Context
Galunisertib (LY2157299)Small molecule; ALK5/TβRI kinase inhibitorATP-competitive inhibition of ALK5 kinase → blocks SMAD2/3 phosphorylation → abolishes canonical TGF-β signalingHepatocellular carcinoma (HCC): phase 2 (RECOURSE) — OS improvement in subset. Glioblastoma: phase 2 with lomustine — OS improvement. NSCLC + pembrolizumab combination trials. Major concern: cardiac toxicity (hypertrophy) with continuous dosing → intermittent 2-weeks-on/2-off schedule used in clinical trials.
Vactosertib (EW-7197)Small molecule; ALK5/ALK4 kinase inhibitorDual inhibitor of TGF-β and activin signaling; highly selective and orally bioavailablePhase 1/2 trials in colorectal cancer, cholangiocarcinoma, and fibrodysplasia ossificans progressiva (FOP — ALK2 gain-of-function → heterotopic ossification)
Fresolimumab (GC1008)Human anti-TGF-β1/2/3 pan-neutralizing monoclonal antibodyNeutralizes all three TGF-β isoforms; pan-blockade; long half-life IV administrationIPF: phase 2 (Lonzano 2011): small trial showed some benefit on HRCT scores; safety concerns (skin lesions, neoplasms). NSCLC with focal radiation (SBRT): pilot trial showed improved immune activation (abscopal-like effects). Not approved; limited by systemic side effects.
Bintrafusp alfa (M7824)Bifunctional fusion protein: anti-PD-L1 IgG + TGF-β RII extracellular domain (TGF-β trap)PD-L1 arm blocks checkpoint → disinhibits T cells; TGF-β trap arm sequesters tumor-microenvironment TGF-β → relieves immunosuppression and EMT. Designed to deliver TGF-β blockade locally at PD-L1+ tumor sitesJAVELIN Bladder 100 phase 3 — did not improve OS vs avelumab. NSCLC first-line phase 3 (INTR@PID Lung 037) — did not improve PFS vs pembrolizumab. Biliary tract cancer: phase 2 showed activity. Research ongoing in HPV+ cancers (cervical, HNSCC) where TGF-β immunosuppression is dominant.
PirfenidoneSmall molecule; indirect TGF-β suppressionInhibits TGF-β-stimulated collagen synthesis, myofibroblast differentiation, and PDGF-mediated fibroblast proliferation; also anti-TNF-α and anti-oxidant; mechanism imprecisely definedIPF: approved 2014 (see fibrosis.html). Cardiac fibrosis post-MI: PIROUETTE trial (phase 2 RCT, n=116): pirfenidone 2403 mg/day × 52 weeks → reduction in extracellular volume fraction on CMR (−0.44% vs +0.47% placebo, p=0.048). Signals of efficacy in HFpEF-related fibrosis.
Lorecivivint (SM04690)CLK2/DYRK1A kinase inhibitor; indirect Wnt/TGF-β modulatorInhibits CLK2 → reduced SMAD2/3 phosphorylation; also activates β-catenin cartilage anabolic effects; cartilage-protective via dual Wnt-activating and TGF-β-suppressing actionsOsteoarthritis: phase 2 (STRIDES-OA) intra-articular injection — reduced cartilage loss and pain vs saline at 1 year. Phase 3 (STRIDES 3) ongoing.

Connections

References

  1. Massagué J. TGFβ signalling in context. Nat Rev Mol Cell Biol. 2012;13(10):616–30. doi:10.1038/nrm3434
  2. Derynck R, Turley SJ, Akhurst RJ. TGFβ biology in cancer progression and immunotherapy. Nat Rev Clin Oncol. 2021;18(1):9–34. doi:10.1038/s41571-020-0403-1
  3. Hinz B. Myofibroblasts. Exp Eye Res. 2016;142:56–70. doi:10.1016/j.exer.2015.07.009
  4. Travis MA, Sheppard D. TGF-β activation and function in immunity. Annu Rev Immunol. 2014;32:51–82. doi:10.1146/annurev-immunol-032713-120257
  5. Bierie B, Moses HL. Tumour microenvironment: TGFβ: the molecular Jekyll and Hyde of cancer. Nat Rev Cancer. 2006;6(7):506–20. doi:10.1038/nrc1926
  6. Akhurst RJ, Hata A. Targeting the TGFβ signalling pathway in disease. Nat Rev Drug Discov. 2012;11(10):790–811. doi:10.1038/nrd3810
  7. Heldin CH, Moustakas A. Signaling receptors for TGF-β family members. Cold Spring Harb Perspect Biol. 2016;8(8):a022053. doi:10.1101/cshperspect.a022053
  8. Kamato D, Burch ML, Piva TJ, et al. Transforming growth factor-β signalling: role and consequences of Smad linker region phosphorylation. Cell Signal. 2013;25(10):2017–24. doi:10.1016/j.cellsig.2013.06.001