CDK4/6 inhibition induces a DNA damage-independent senescence-associated secretory phenotype driven by delayed activation of NF-κB

Cellular senescence protects against aberrant cancer proliferation by enforcing stable cell cycle arrest. The best-studied triggers of senescence involve DNA damage signaling, while DNA damage-independent senescence is less understood. In the cancer context, senescence has recently been shown to be induced by CDK4/6 inhibitors (CDK4/6i), a class of targeted therapies currently applied to breast cancer and sarcoma, among other tumors. The senescence-associated secretory phenotype (SASP) can exert either pro- or anti-tumorigenic effects depending on its composition, making it a critical determinant of therapy outcome. The dependence of the SASP on the senescence trigger has been a controversial question, which we investigate here, particularly the involvement of NF-{kappa}B signaling. Using a multi-omics approach, we profiled the temporal dynamics of transcriptional and epigenetic state rewiring during therapy-induced senescence in CDK4/6i-treated liposarcoma cells over 28 days; we compared this to the effects of the DNA damaging chemotherapeutic, doxorubicin. Doxorubicin triggers an early and robust activation of p53 and NF-{kappa}B, as expected. Surprisingly, a delayed NF-{kappa}B-driven SASP signature also emerges in CDK4/6i-induced senescence after a 10-day exposure, in the absence of additional DNA damage. We replicated this effect in breast cancer cells over similar time courses. Inhibition of NF-{kappa}B blunts pro-inflammatory SASP expression without reversing cell cycle arrest in both treatment contexts, showing that NF-{kappa}B is a conserved SASP regulator, regardless of trigger and cancer type. Instead, the nature of the trigger regulates differential dynamics of the NF-{kappa}B response and additional SASP gene signatures, including a CDK4/6i-specific gene module driven by TEAD factors. Overall, our 28-day multi-omics time course reveals previously unknown dynamics of the SASP and demonstrates that both common and trigger-specific components of therapy-induced SASP can be separately tuned and uncoupled from cell cycle arrest, with implications for the rational design of adjuvant therapies.