Pulmonary fibrosis,
mapped as a system.

Scope. This primer explains the biology shared across fibrosing interstitial lung diseases, using idiopathic pulmonary fibrosis (IPF) as the best-studied model. It aims to cover the major, well-supported mediator families and named pathways—not literally every molecule ever reported in a cell or animal experiment. The same mediator can be protective during acute repair yet harmful when persistent, and findings from bleomycin-treated mice do not automatically establish causality in human IPF.

Use. Educational and research reference; not medical advice or a treatment guide. Updated 19 October 2026.

1. The one-paragraph model

In a susceptible, usually ageing lung, repeated injury to alveolar epithelial cells—especially type 2 alveolar epithelial (AT2) cells—produces incomplete regeneration rather than clean repair. Injured epithelial cells release alarmins, cytokines, growth factors and lipids; recruit and reprogramme macrophages and other immune cells; and activate resident mesenchymal cells. Fibroblasts expand, migrate into injured regions and become contractile, apoptosis-resistant myofibroblasts that deposit and remodel extracellular matrix (ECM). The stiffened, cross-linked matrix then activates integrins, focal adhesions, Rho–ROCK, MRTF and YAP/TAZ, mechanically releases latent TGF-β, and further stabilizes the myofibroblast state. Thus fibrosis can become a feed-forward circuit even after the initiating insult has faded.[1–6]

System flow: susceptibility and ageing → epithelial injury → organelle stress and danger signals → immune and stromal reprogramming → myofibroblasts and extracellular matrix → stiffness and mechanotransduction → further latent TGF-β activation. Endogenous resolution brakes oppose this loop but may be insufficient.

2. Normal repair versus fibrosis

Human single-cell and spatial studies refine this model: fibrotic niches place myofibroblasts near aberrant KRT17-positive basaloid epithelium and SPP1-positive macrophages, while persistent transitional AT2→AT1 states indicate failed epithelial differentiation rather than a simple conversion of epithelial cells into all scar-forming fibroblasts.[4,5,7,8]

3. The principal cellular actors

Alveolar and airway epithelium

• AT1 cells cover the gas-exchange surface and are highly vulnerable to injury.
• AT2 cells make surfactant and serve as alveolar progenitors. Telomere dysfunction, surfactant-protein misfolding, ER stress, mitochondrial dysfunction, impaired autophagy/mitophagy and senescence reduce their regenerative capacity.
• Transitional AT2→AT1 cells (often KRT8/CLDN4/SFN positive) are normal temporary repair intermediates; persistence is associated with fibrosis.
• Aberrant basaloid/bronchiolized cells (commonly KRT17 positive, sometimes KRT5 negative) express epithelial, developmental, senescence and mesenchymal-associated programmes and occupy fibrotic niches.
• Injured epithelial cells supply TGF-β, PDGF, CTGF/CCN2, IL-1, IL-6, IL-11, IL-33, TNF, CCL2, CXCL8, CXCL12, WNT ligands, SHH, LPA and extracellular vesicles, while losing antifibrotic epithelial signals such as PGE2.

Fibroblasts, myofibroblasts and mural cells

• Resident fibroblast lineages are the main source of scar ECM. Disease states include CTHRC1-high, HAS1-high, SPINT2-high, inflammatory and invasive fibroblast programmes; nomenclature varies across datasets.
• Myofibroblasts express ACTA2/α-SMA, generate traction, contract matrix and produce collagens, fibronectin and matricellular proteins.
• Pericytes and other mesenchymal cells can contribute in experimental and human disease. The quantitative contribution of circulating fibrocytes is probably smaller and more context-dependent.
• Full epithelial-to-mesenchymal transition is not considered the dominant source of fibroblasts in human IPF; epithelial cells more clearly influence fibrosis through maladaptive states and paracrine signalling.

Monocytes and macrophages

• Resident alveolar macrophages and recruited CCR2-positive monocytes acquire disease-associated states rather than fitting neatly into an M1/M2 binary.
• SPP1/osteopontin-high, TREM2, GPNMB, CD9, MERTK and lipid-handling programmes are recurrent in fibrotic macrophages.
• Key outputs include TGF-β, PDGF, IL-1β, TNF, IL-6, IL-10, IL-13, CCL2, CCL18, SPP1, galectin-3, amphiregulin, MMPs and lipid mediators. Effects depend on timing, origin and niche.

Other immune cells

• Neutrophils: ROS, elastase, MMP-9 and extracellular traps can intensify epithelial and endothelial injury; their exact role varies by fibrotic ILD.
• CD4 T cells and ILC2s: type-2 signals IL-4/IL-13 support profibrotic macrophage and fibroblast programmes; IL-17A can amplify inflammation and fibrosis in some contexts.
• Regulatory T cells and IL-10: may restrain inflammation early but can have stage- and disease-specific effects.
• B cells/plasma cells: autoantibodies, antigen presentation and lymphoid aggregates are especially relevant in autoimmune fibrotic ILD; less clearly causal in IPF.
• Mast cells: histamine, tryptase, chymase, TGF-β and IL-13 can activate fibroblasts and remodel matrix.

Endothelium and vasculature

• Barrier loss, endothelial apoptosis/senescence, capillary rarefaction, angiogenic imbalance and endothelial-to-mesenchymal-like programmes alter repair.
• VEGF is spatially and temporally context-dependent: adequate signalling supports capillary maintenance, while disordered angiogenesis and vascular remodelling may be pathogenic.
• Endothelial S1P receptor 1, nitric oxide, endothelin-1, coagulation and complement connect vascular injury to fibroblast activation.

4. Master mediator and pathway atlas

Evidence labels are deliberately qualitative:

• Core: convergent human tissue, genetic, spatial/omics and functional evidence.
• Supported: consistent human association plus mechanistic evidence, but context or causal rank remains unsettled.
• Emerging/preclinical: biologically plausible and often strong in models, but not yet a universal human disease driver.

4.1 TGF-β: the central convergence node

Key idea: total TGF-β abundance is less informative than where, when and by which integrin/protease it becomes active. Broad systemic blockade risks disrupting immunity, epithelial homeostasis and tumour suppression.

4.2 Growth factors, peptide mediators and receptor kinases

IL-11 is a useful lesson in evidence ranking: strong genetic and antibody results in several mouse systems and human expression data support involvement, but independent work disputes the claim that it is the universal master effector downstream of TGF-β.[9,10]

4.3 Cytokines, chemokines and immune–stromal signals

4.4 Innate danger sensing, complement and coagulation

• DAMPs: extracellular ATP, HMGB1, mitochondrial/nuclear DNA, oxidized phospholipids, urate, S100 proteins, fragmented hyaluronan and ECM neo-epitopes.
• Pattern-recognition receptors: TLR2/4/9, RAGE, scavenger receptors and C-type lectins activate NF-κB, MAPKs and interferon pathways. TLR effects depend strongly on ligand, compartment and timing.
• ATP–P2X7: potassium efflux activates NLRP3; ATP is degraded by CD39/CD73 to adenosine, which can become profibrotic through A2B receptors.
• cGAS–STING–TBK1–IRF3: senses cytosolic DNA from mitochondrial/nuclear damage. Acute activation can aid host defence, while chronic activation can reinforce senescence, type-I interferon/NF-κB signalling and fibrosis.[11]
• Complement: C3a–C3aR and C5a–C5aR1/2 recruit and activate myeloid cells and interact with coagulation and TGF-β pathways.
• Coagulation: tissue factor–factor VIIa, factor Xa and thrombin–PAR1 signalling increase permeability and fibroblast activation. PAI-1/SERPINE1 suppresses fibrinolysis and is also a senescence/YAP-TGF-β target.
• NETs and proteases: neutrophil extracellular traps, elastase and cathepsins can propagate epithelial/endothelial injury and expose new DAMPs; importance differs across IPF, autoimmune ILD and post-injury fibrosis.

4.5 Developmental pathways reactivated in adult repair

4.6 Mechanotransduction and the self-sustaining scar

1. Fibroblasts deposit collagens I/III/V, fibronectin and other ECM.
2. LOX/LOXL enzymes and transglutaminase 2 cross-link matrix, raising stiffness.
3. Integrins cluster and activate FAK/SRC → RhoA–ROCK → actomyosin tension.
4. Actin dynamics release MRTF-A/MKL1 to cooperate with SRF; stiffness and tension drive nuclear YAP/TAZ–TEAD.
5. Contractile cells pull on αv integrin-bound latent TGF-β, releasing active ligand.
6. TRPV4, PIEZO1 and other mechanosensitive channels alter Ca²⁺ and cytoskeletal signalling.
7. Fibroblasts migrate up stiffness gradients (durotaxis) and create still more stiff matrix.

This loop explains how fibrosis may progress with modest classical inflammation and why a decellularized IPF matrix can impose a profibrotic phenotype on otherwise normal fibroblasts.[2,3,6]

4.7 ECM components and remodelling enzymes

4.8 Lipid and purinergic mediators

• LPA: autotaxin/ENPP2 converts lysophosphatidylcholine to lysophosphatidic acid. LPA1 and other LPA receptors promote epithelial leak, fibroblast migration, survival and contraction.
• S1P: produced by sphingosine kinases and signals through S1PR1–5. Effects are receptor- and cell-specific: fibroblast/epithelial SPHK1–S1P signalling can be profibrotic, whereas endothelial S1PR1 helps preserve barrier function.
• Eicosanoids: increased leukotriene tone (LTB4, LTC4/LTD4/LTE4; 5-LOX pathway) tends to be profibrotic. PGE2–EP2/EP4–cAMP is a major endogenous brake on fibroblast proliferation, collagen synthesis and myofibroblast differentiation; IPF fibroblasts may produce less PGE2 or become resistant to it.
• Pro-resolving lipids: lipoxins, resolvins, protectins and maresins restrain inflammation and promote efferocytosis; pulmonary-fibrosis evidence is promising but less mature.
• Oxidized phospholipids and cholesterol derivatives can stress AT2 cells and reprogramme macrophages.
• Adenosine: sustained extracellular adenosine, generated by CD39/CD73, can signal through A2B receptors to promote IL-6, fibroblast activation and remodelling; acute adenosine can still be tissue-protective.
• Extracellular vesicles transport lipids, proteins, mitochondrial material and miRNAs between epithelial, immune and mesenchymal cells.[12]

4.9 Oxidative, organelle, ageing and cell-death pathways

4.10 Metabolic reprogramming

• Activated fibroblasts increase glycolysis (GLUT1, PFKFB3), glutaminolysis (GLS1 → α-ketoglutarate), proline/glycine supply, one-carbon metabolism and mitochondrial substrate use to support collagen synthesis.
• mTORC1–ATF4 coordinates amino-acid acquisition and matrix-protein translation; AMPK generally opposes anabolic activation.
• Succinate, lactate and altered NAD redox can stabilize HIF signalling, change chromatin and acidify the niche; lactate can facilitate TGF-β activation.
• AT2 surfactant and macrophage lipid handling are disrupted; altered ABCA3, fatty-acid oxidation, cholesterol efflux and oxidized lipids link epithelial stress to immune states.
• Macrophage and fibroblast metabolism is state-specific; “glycolytic = inflammatory” and “oxidative = reparative” are oversimplifications.

4.11 Epigenetic and RNA regulation

• DNA methylation, histone acetylation/methylation, chromatin accessibility and SWI/SNF complexes stabilize abnormal epithelial and fibroblast states.
• Commonly studied profibrotic miRNAs include miR-21 (amplifies TGF-β/SMAD by suppressing inhibitory nodes) and context-dependent miR-199 family members.
• Common antifibrotic or homeostatic miRNAs include miR-29 (represses collagen/ECM genes), let-7d, miR-200 and miR-326 in selected systems.
• Long non-coding and circular RNAs—including MALAT1 and many disease-specific candidates—can tune TGF-β, WNT and fibroblast programmes, but many have limited replication or uncertain human causality.
• Extracellular-vesicle RNA enables intercellular transfer. A statistically changed RNA is not automatically a disease mediator; cell-specific perturbation and human validation matter.

5. Predisposition: what lowers the threshold for maladaptive repair

These are not soluble mediators, but they determine why the same injury heals in one person and scars in another.

• MUC5B promoter variant rs35705950: strongest common IPF risk allele; affects distal-airway mucociliary biology and has a complex association with survival.
• Telomere maintenance: TERT, TERC, RTEL1, PARN, DKC1, TINF2 and related genes; short telomeres promote stem/progenitor failure and constrain immune/epithelial renewal.
• Surfactant/proteostasis: SFTPC, SFTPA1/2, ABCA3 and NKX2-1 variants can cause epithelial ER stress and familial disease.
• Other susceptibility loci include DSP, DPP9, TOLLIP, AKAP13, FAM13A and immune/barrier genes.
• Ageing supplies convergent pressure through telomere attrition, mitochondrial dysfunction, senescence, clonal haematopoiesis, epigenetic drift and reduced proteostasis.
• Exposures—cigarette smoke, metal/wood dusts, silica/asbestos, air pollution, aspiration/reflux, drugs, radiation, viral injury and recurrent microinjury—vary by disease and interact with genotype.

6. The most important reinforcing loops

1. Epithelial stress loop: telomere/proteostasis/mitochondrial failure → AT2 senescence or death → incomplete regeneration → more exposed and mechanically stressed epithelium.
2. TGF-β–matrix loop: TGF-β → myofibroblast ECM → stiffness/traction → αv-integrin activation of more latent TGF-β.
3. Macrophage–fibroblast loop: CCL2/CSF1 and injury recruit monocytes → SPP1/galectin-3/TGF-β/PDGF outputs → fibroblast activation → ECM and chemokines that retain disease-associated macrophages.
4. Senescence–SASP loop: DNA/organelle damage → p16/p21 arrest → IL-6/IL-11/TGF-β/chemokine/protease secretome → paracrine senescence and immune remodelling.
5. Lipid loop: surfactant and oxidized-lipid disruption → macrophage reprogramming and epithelial stress → autotaxin–LPA/S1P/leukotriene signalling → fibroblast activation and more tissue damage.
6. Coagulation–matrix loop: barrier damage and tissue factor → thrombin/PAR signalling + fibrin → fibroblast activation, while PAI-1 suppresses fibrinolysis.
7. Mechanical bronchiolization loop: regional stretch/stiffness and failed alveolar repair favour aberrant basal/airway epithelial states that signal to subjacent mesenchyme.

7. Endogenous brakes and resolution biology

Fibrosis reflects both excessive drivers and failed brakes.

• PGE2–EP2/EP4–cAMP/PKA suppresses fibroblast proliferation, collagen and contractility.
• BMP signalling, HGF/MET and selected FGF signals support differentiation and epithelial repair.
• NRF2, glutathione, catalase and mitochondrial quality control limit oxidative injury.
• AMPK, PTEN and Hippo kinases restrain mTOR/AKT and YAP/TAZ activity.
• SMAD7, inhibitory phosphatases and receptor turnover restrain TGF-β.
• Lipoxins/resolvins/maresins, IL-10 in appropriate contexts, efferocytosis and macrophage state transitions promote inflammatory resolution.
• Myofibroblast apoptosis, deactivation or senescence followed by immune clearance is essential; persistence is pathological.
• Balanced MMP/TIMP, uPA/tPA–plasmin and matrix phagocytosis permit scar remodelling. Matrix degradation can also release bioactive fragments, so more proteolysis is not always beneficial.

8. How to read mechanistic claims

Use this evidence ladder:

1. Human genetics or Mendelian disease that points to a causal cell process.
2. Human spatial tissue evidence showing ligand, receptor and responding cells in the same niche.
3. Longitudinal human biomarker or functional evidence, ideally linked to progression.
4. Perturbation in primary human cells, organoids or precision-cut lung slices.
5. Multiple animal models, including treatment after fibrosis is established—not only prevention in bleomycin injury.
6. Single cell-line or association-only studies, which are hypothesis-generating.

Three recurring cautions:

• Timing: blocking inflammation before fibrosis is not the same as reversing established scar.
• Cell specificity: one receptor may be pathogenic in fibroblasts and protective in endothelium.
• Model specificity: bleomycin fibrosis is partly resolving and inflammation-heavy; human IPF is spatially heterogeneous, chronic and age-associated.

9. Compact alphabetical mediator index

This index maps frequently encountered names to their dominant module. It is a navigation aid, not a claim that every item is equally causal.

10. Basic techniques used to study pulmonary fibrosis

No single assay demonstrates “fibrosis.” A convincing study combines tissue architecture, matrix quantity, cell-state evidence and lung function, then shows that the proposed mediator changes these outcomes in the expected cell type and time window.

10.1 Human material and clinical phenotyping

Minimum metadata include diagnosis and pattern, age, sex, smoking/exposure history, disease duration, medications, physiology, sampling site, tissue handling time and comorbidities. Fibrotic lungs are patchy, so spatially matched and multi-region sampling is valuable.

10.2 Histology and tissue imaging

• H&E: overall architecture, inflammation, epithelial injury and fibroblastic foci.
• Masson's trichrome: collagen-rich scar; convenient but semi-quantitative and stain-batch sensitive.
• Picrosirius red with polarized light: fibrillar collagen abundance and organization; colour alone should not be treated as exact collagen type.
• Second-harmonic generation microscopy: label-free imaging of fibrillar collagen, alignment and fibre morphology.
• Immunohistochemistry/immunofluorescence: localizes proteins such as α-SMA, COL1A1, fibronectin, KRT8/17, SFTPC, SPP1, CD68, pSMAD2/3, YAP/TAZ and Ki-67.
• RNAscope/in-situ hybridization: maps low-abundance transcripts and ligand–receptor pairs with cellular context.
• Multiplex IF, imaging mass cytometry or cyclic imaging: measures many markers in the same section and defines cellular neighbourhoods.
• Micro-CT/confocal 3D imaging and tissue clearing: resolves airway/alveolar remodelling and collagen networks in three dimensions.

Use blinded scoring, prespecified regions, whole-slide or systematic random sampling, batch-balanced staining and automated segmentation checked by a human. Marker co-expression supports a cell state but rarely proves lineage or mechanism.

10.3 Measuring collagen and ECM

Best practice is to pair one spatial method (histology/SHG) with one biochemical method (hydroxyproline or ECM proteomics) and one functional/mechanical measure.

10.4 Cell culture and perturbation systems

• Primary human lung fibroblasts: test TGF-β, PDGF, matrix stiffness, migration, contraction and ECM production. Record donor, passage and anatomical region; stiff plastic itself activates fibroblasts.
• Primary AT2 cells and air–liquid interface cultures: study surfactant biology, injury, differentiation and epithelial–mesenchymal signalling. Phenotype is rapidly lost in conventional 2D culture.
• Induced pluripotent stem cell-derived AT2 cells: useful for patient variants and scalable epithelial models; maturity may differ from adult AT2 cells.
• Macrophage cultures: BAL-derived, monocyte-derived or iPSC-derived macrophages allow mediator and efferocytosis studies; they are not interchangeable.
• Co-culture: epithelial–fibroblast, macrophage–fibroblast and tri-culture systems distinguish contact-dependent from soluble crosstalk. Transwells separate these mechanisms.
• 3D organoids/alveolospheres: measure epithelial progenitor capacity and fibroblast support; matrix formulation strongly shapes results.
• Precision-cut lung slices (PCLS): preserve multicellular architecture and ECM for perturbation in human or animal tissue; viability and diffusion limit experiment duration.
• Lung-on-chip / microfluidics: add cyclic stretch, air–liquid interface, perfusion and immune-cell recruitment; powerful but less standardized.
• Decellularized healthy versus IPF matrix: directly tests whether ECM composition/mechanics imposes disease phenotypes.

Common perturbations include recombinant ligands, neutralizing antibodies, receptor inhibitors, siRNA/ASO, CRISPR knockout or CRISPRi/a, inducible overexpression and conditioned-medium transfer. Use vehicle, non-targeting reagent, isotype and rescue controls; verify target engagement before interpreting phenotype.

10.5 Core functional assays

10.6 Molecular and biochemical methods

• RT-qPCR/digital PCR: targeted expression with stable, validated reference genes.
• Western blot, ELISA, Luminex/Olink/SomaScan: protein abundance and multiplexed soluble mediators. Confirm antibody specificity and account for matrix-bound versus soluble pools.
• Flow cytometry/CyTOF: quantitative cell populations and phospho-signalling. Enzymatic tissue digestion can destroy epitopes or selectively lose fragile cells.
• Reporter assays: SMAD, NF-κB, STAT, WNT/TCF, YAP/TEAD and antioxidant response activity.
• Co-immunoprecipitation/proximity assays: protein complexes and receptor interactions.
• Chromatin methods: ATAC-seq, ChIP-seq/CUT&RUN/CUT&Tag and DNA methylation profiling reveal regulatory state.
• Proteomics, phosphoproteomics, secretomics and metabolomics/lipidomics: identify pathway activity beyond RNA; batch, ion suppression and annotation uncertainty require careful QC.
• Stable-isotope tracing: follows glucose, glutamine, acetate or proline into energy pathways and collagen precursors.

10.7 Transcriptomics, single-cell and spatial approaches

• Bulk RNA-seq: robust whole-tissue or sorted-population expression, but averages cell composition and state.
• Single-cell RNA-seq: resolves cell types/states and candidate trajectories; dissociation stress, dropout and sample integration can create artefacts.
• Single-nucleus RNA-seq: works well with frozen or fibrotic tissue and reduces dissociation bias, but misses much cytoplasmic RNA.
• Spatial transcriptomics: maps expression to fibrotic niches; spot-based systems mix cells, while imaging-based systems trade gene number against resolution.
• scATAC-seq and multiome: link chromatin accessibility to cell-state RNA.
• CITE-seq / spatial proteomics: combine RNA with surface or tissue protein measurements.
• Ligand–receptor inference: CellPhoneDB, NicheNet and similar tools generate hypotheses; co-expression is not proof of signalling and must be validated by perturbation.
• Pseudotime/RNA velocity: suggest state transitions; they do not establish lineage. Genetic lineage tracing is needed in animals, and clonal or temporal evidence is needed in humans.

Essential design principles are multiple biological donors, balanced batches, donor-aware statistics, prespecified QC, ambient-RNA/doublet handling, replication in an independent cohort and spatial or protein validation of headline states.

10.8 Genetics and causal testing

• GWAS and rare-variant sequencing identify susceptibility loci and pathways; colocalization and fine mapping are needed to nominate causal genes.
• eQTL/pQTL and Mendelian randomization connect variants to expression, proteins and outcomes, subject to instrument assumptions.
• CRISPR screens can identify regulators in fibroblasts or epithelial cells; hits require individual validation.
• Conditional mouse alleles and lineage tracing localize gene function and cell origins. Inducible systems allow separation of initiation from progression.
• Organoids or iPSC lines carrying patient variants, paired with isogenic corrected controls, provide strong genotype-to-cell-phenotype tests.

10.9 Animal and ex-vivo models

For translational antifibrotic claims, administer an intervention after measurable fibrosis is established, not only prophylactically. Report randomization, blinding, exclusions, both sexes where appropriate, dose/exposure, mortality, histology, collagen and physiology. A change in inflammatory cells at day 7 is not proof of reversal of established fibrosis.

10.10 Lung mechanics and matrix biophysics

• Atomic force microscopy (AFM): local tissue or hydrogel stiffness maps.
• Rheometry: bulk storage/loss modulus of tissue or engineered matrices.
• Traction-force microscopy and micropost arrays: forces exerted by individual cells.
• Tunable hydrogels: separate stiffness from ligand composition; control pore size and viscoelasticity.
• Cyclic stretch systems: model breathing-associated mechanical load.
• Optical tweezers/magnetic beads: receptor- or adhesion-scale force responses.
• Pressure–volume loops, compliance and forced-oscillation measurements: whole-lung mechanical function in animals.

Elastic modulus alone is insufficient: viscoelasticity, fibre alignment, cross-linking, ligand density and stress relaxation also control cell behaviour.

10.11 A practical minimum experimental workflow

For testing whether mediator X drives fibrosis:

1. Locate it in human disease: demonstrate X and its receptor in the same fibrotic niche using spatial RNA/protein methods; compare multiple donors and non-fibrotic controls.
2. Establish association with state or progression: use donor-aware bulk/single-cell data, BAL or blood only as appropriate, and avoid treating correlation as causation.
3. Perturb primary human cells: inhibit and activate X in the relevant epithelial, macrophage or fibroblast system; confirm target engagement and perform rescue.
4. Measure multiple axes: cell state, signalling, deposited/cross-linked ECM and a functional phenotype such as force, barrier or differentiation.
5. Validate in multicellular human tissue: co-culture, organoid or PCLS to test cell–cell and matrix context.
6. Test in more than one in-vivo context: use therapeutic timing after fibrosis begins and include physiology plus histological/biochemical matrix endpoints.
7. Challenge the claim: test cell specificity, stage dependence, sex/age, off-target effects and whether benefit reflects reduced initial injury rather than true antifibrotic action.

11. Selected sources and further reading

1. Lederer DJ, Martinez FJ. Idiopathic Pulmonary Fibrosis. N Engl J Med. 2018;378:1811–1823. DOI
2. Herrera J, Henke CA, Bitterman PB. Extracellular matrix as a driver of progressive fibrosis. J Clin Invest. 2018;128:45–53. DOI
3. Tschumperlin DJ et al. Mechanosensing and fibrosis. J Clin Invest. 2018;128:74–84. DOI
4. Pardo A, Selman M. The interplay of the genetic architecture, aging, and environmental factors in the pathogenesis of idiopathic pulmonary fibrosis. Am J Respir Cell Mol Biol. 2021;64:163–172. PubMed
5. Vancheri C et al. Idiopathic Pulmonary Fibrosis: From Common Microscopy to Single-Cell Biology and Precision Medicine. Am J Respir Crit Care Med. 2024;209:1074–1085. Article
6. Upagupta C et al. Matrix abnormalities in pulmonary fibrosis. Eur Respir Rev. 2018;27:180033. DOI
7. Habermann AC et al. Single-cell RNA sequencing reveals profibrotic roles of distinct epithelial and mesenchymal lineages in pulmonary fibrosis. Sci Adv. 2020;6:eaba1972. DOI
8. Mayr CH et al. Spatial transcriptomic characterization of pathologic niches in IPF. Sci Adv. 2024;10:eadl5473. PubMed
9. Cook SA, Schafer S. Hiding in plain sight: interleukin-11 emerges as a master regulator of fibrosis, tissue integrity, and stromal inflammation. Annu Rev Med. 2020;71:263–276. DOI
10. Widjaja AA et al. / counterpoint literature summarized in IL11-mediated stromal cell activation may not be the master regulator of pro-fibrotic signaling downstream of TGFβ. Front Immunol. 2024;15:1293883. Article
11. Ma R et al. The role of cGAS-STING signaling in pulmonary fibrosis and its therapeutic potential. Front Immunol. 2023;14:1273248. PMC
12. Burgy O et al. Extracellular lipids in the lung and their role in pulmonary fibrosis. Cells. 2022;11:1209. PMC
13. Wynn TA. Integrating mechanisms of pulmonary fibrosis. J Exp Med. 2011;208:1339–1350. PMC
14. Sgalla G et al. Idiopathic pulmonary fibrosis: pathogenesis and management. Respir Res. 2018;19:32. DOI
15. Kamiya M et al. Immune mechanisms in fibrotic interstitial lung disease. Cell. 2024;187:3506–3530. DOI
16. Mei Q et al. Idiopathic Pulmonary Fibrosis: An Update on Pathogenesis. Front Pharmacol. 2022;12:797292. PMC

Bottom line

The shortest accurate description is: susceptible epithelium fails to regenerate, immune and stromal niches lock into maladaptive repair, and a mechanically active ECM makes the response self-perpetuating. TGF-β is the central convergence signal, but fibrosis emerges from its coupling to epithelial stress, macrophage states, developmental pathways, cell metabolism, coagulation, lipids and matrix mechanics—not from any single mediator acting alone.