The neurotrophin-3 (NT-3) receptor tropomyosin receptor kinase C (TrkC/NTRK3) has been described as a dependence receptor and, as such, triggers apoptosis in the absence of its ligand NT-3. This proapoptotic activity has been proposed to confer a tumor suppressor activity to this classic tyrosine kinase receptor (RTK). By investigating interacting partners that might facilitate TrkC-induced cell death, we have identified the basic helix-loop-helix (bHLH) transcription factor Hey1 and importin-α3 (karyopherin alpha 4 [KPNA4]) as direct interactors of TrkC intracellular domain, and we show that Hey1 is required for TrkC-induced apoptosis. We propose here that the cleaved proapoptotic portion of TrkC intracellular domain (called TrkC killer-fragment [TrkC-KF]) is translocated to the nucleus by importins and interacts there with Hey1. We also demonstrate that Hey1 and TrkC-KF transcriptionally silence mouse double minute 2 homolog (MDM2), thus contributing to p53 stabilization. p53 transcriptionally regulates the expression of TrkC-KF cytoplasmic and mitochondrial interactors cofactor of breast cancer 1 (COBRA1) and B cell lymphoma 2-associated X (BAX), which will subsequently trigger the intrinsic pathway of apoptosis. Of interest, TrkC was proposed to constrain tumor progression in neuroblastoma (NB), and we demonstrate in an avian model that TrkC tumor suppressor activity requires Hey1 and p53.

The memory CD8 T-cell pool must select for clones that bind immunodominant epitopes with high affinity to efficiently counter reinfection. At the same time, it must retain a level of clonal diversity to allow recognition of pathogens with mutated epitopes. How the level of diversity within the memory pool is controlled is unclear, especially in the context of a selective drive for antigen affinity. We find that preservation of clones that bind the activating antigen with low affinity depends on expression of the transcription factor Eomes in the first days after antigen encounter. Eomes is induced at low activating signal strength and directly drives transcription of the prosurvival protein Bcl-2. At higher signal intensity, T-bet is induced which suppresses Bcl-2 and causes a relative survival advantage for cells of low affinity. Clones activated with high-affinity antigen form memory largely independent of Eomes and have a proliferative advantage over clones that bind the same antigen with low affinity. This causes high-affinity clones to prevail in the memory pool, despite their relative survival deficit. Genetic or therapeutic targeting of the Eomes/Bcl-2 axis reduces the clonal diversity of the memory pool, which diminishes its ability to respond to pathogens carrying mutations in immunodominant epitopes. Thus, we demonstrate on a molecular level how sufficient diversity of the memory pool is established in an environment of affinity-based selection.