Caffeic Acid Silences Senescent Cell Inflammation by Disabling a Key Protein

Idiopathic pulmonary fibrosis (IPF) is a progressive, age-associated lung disease with limited treatment options. The two FDA-approved drugs, pirfenidone and nintedanib, slow progression but carry significant side effects. Emerging evidence implicates senescent cells and their secretion of pro-inflammatory factors — collectively termed the senescence-associated secretory phenotype (SASP) — as key drivers of IPF pathology. Targeting these cells with either senolytics (which kill them) or senomorphics (which silence their inflammatory output) has gained traction, but senomorphics with clearly defined molecular targets remain scarce.

To address this gap, the authors screened a library of natural products and food-derived compounds for the ability to suppress three canonical SASP cytokines — TNF-α, IL-6, and IL-1β — in cisplatin-induced senescent A549 lung cancer cells. Caffeic acid (CA), a hydroxycinnamic acid found in coffee, olive oil, fruits, and traditional Chinese medicine, emerged as the top hit, reducing all three cytokines to near-baseline levels at 10 µM and exhibiting a dose-dependent effect starting at 3 µM. Crucially, CA did not affect senescent cell viability even at 150 µM, confirming its senomorphic — not senolytic — profile.

To uncover CA's mechanism, the team applied competitive activity-based protein profiling (ABPP), exploiting CA's α,β-unsaturated carbonyl (Michael acceptor) structure to detect covalent binding to cysteine residues across the proteome. Annexin A5 (ANXA5) Cys316 emerged as the most reactive site, with a competitive ratio of 36-fold over control. Multiple orthogonal assays validated this interaction: fluorescence competition with IAA probes, microscale thermophoresis (binding affinity Kd = 3.68 µM), cellular thermal shift assay (CETSA), and a CA-functionalized chemical probe that labeled wild-type but not C316A-mutant ANXA5. Molecular docking confirmed a plausible binding pose at the Cys316 pocket. The catechol moiety — not oxidized dopaquinone — was identified as the active pharmacophore for covalent engagement.

Functionally, CA binding caused proteasomal degradation of ANXA5, leading to deactivation of PKCθ (a kinase whose activity depends on ANXA5 scaffolding) and downstream suppression of NF-κB signaling. siRNA knockdown of ANXA5 phenocopied CA's SASP-suppressing effects, and ANXA5 overexpression partially rescued the phenotype, confirming on-target activity. In a bleomycin-induced pulmonary fibrosis mouse model, CA administration significantly reduced inflammatory cytokines in bronchoalveolar lavage fluid and serum, lessened collagen deposition, improved histological lung architecture, and enhanced physical performance metrics. These in vivo findings corroborated the cellular mechanistic data.

The study positions CA as a well-tolerated, naturally sourced senomorphic and identifies ANXA5 as a previously unrecognized druggable node in SASP regulation. While exciting, the work is primarily conducted in mice and cancer-derived cell lines, and translation to human IPF will require clinical validation. Nonetheless, ANXA5 emerges as a compelling target for senescence-driven aging diseases.