Apoptosis and cellular senescence operate as anti-tumor safeguard mechanisms. Unlike apoptotic cells, senescent cells remain viable, and, hence, may crosstalk to other cells in their vicinity over extended periods of time. In fact, cells that entered oncogene-induced senescence or anticancer therapy-induced senescence (TIS) present with a senescence-associated secretory phenotype (SASP), a massive production of secretable factors, which reportedly reinforces senescence through an intracellular mechanism. Utilizing the Eμ-myc transgenic mouse lymphoma model, we provide evidence for an outcome-relevant paracrine, DNA damage-independent secondary senescence program (SecS) in vitro and in vivo.
Apoptosis-blocked (bcl2-infected) lymphoma cells from different genetic backgrounds were treated with the DNA-damaging anticancer agent adriamycin in vitro or the alkylating agent cyclophosphamide upon lymphoma formation in mice in vivo. TIS and SecS was detected based on senescence-associated b-galactosidase activity (SA-b-gal), Ki67 staining and BrdU incorporation. The secretome of senescent cells was analyzed by proteomics, gene expression and protein arrays. Overall and progression free survival in mice and patients was assessed by Kaplan-Meier analysis.
Transcriptome and secretome analyses followed by functional studies found extracellular matrix proteins, especially small leucine-rich proteoglycans (SLRP), but not NF-kB-dependent cytokines and chemokines, to induce SecS in proliferating lymphoma cells in a paracrine fashion, and linked a “high secretor” status to stronger SecS induction. Dissecting senescence-mediating pathways in recipient cells by biochemical, genetic and pharmacological means unveiled an essential role for the LDL receptor-related protein 1 (LRP1), a receptor for SLRP and other SASP components, through the cell-cycle inhibitor p21CIP1 in SecS. Accordingly, mice harboring TIS-capable but genetically SecS-defective lymphomas (e.g. lacking LRP1 or p21CIP1 expression) experienced inferior long-term outcome to therapy. Not only the recipient cell-based LRP1 status but also the genetically and biologically distinct donor cell-based secretor gene signature stratified outcome in mice. Strikingly, humanized versions of both classifiers were predictive in a large cohort of diffuse large B-cell lymphoma (DLBCL) patients, where they identified – although composed of different gene sets – largely overlapping patient subgroups with superior prognosis, again suggesting SecS as the critical underlying treatment effector principle. Our study highlights the predictive power of senescence for treatment outcome in DLBCL, and provides functional examples (which will be discussed at the meeting) for SASP-related non-genotoxic pro-senescent therapies.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.