Binding of MYC to nascent RNA suppresses innate immune signaling by R-loop-derived RNA-DNA hybrids [ChIP_Rx]

In response to perturbed transcription elongation, the MYC oncoprotein multimerizes and undergoes a phase transition. The underlying mechanisms and their function are unknown. Here, we demonstrate that MYC globally relocalizes from its canonical positions on DNA to nascent RNA in response to the accumulation of intronic RNA. Binding of MYC to RNA induces MYC to form multimers that concentrate the nuclear exosome - a 3'-5' RNA exonuclease - and its targeting complexes around double-stranded RNA and R-loops. MYC harbors four RNA-binding regions (RBRI–IV), which overlap with evolutionarily conserved regions. Of these, RBRIII promotes MYC multimerization and is necessary for recruiting the exosome to R-loops. While RBRIII is dispensable for MYC-dependent transcriptional activation and pancreatic tumor cell proliferation in culture, it is indispensable for sustaining tumor growth in vivo. Via RBRIII, MYC suppresses the accumulation of R-loop-derived RNA-DNA hybrids and prevents them from binding to the pattern recognition receptor TLR3 and activating the innate immune kinase TBK1. Our data demonstrate that MYC has mechanistically distinct functions in transcription and RNA metabolism, and that its phase transition is an RNA-driven stress response that suppresses the accumulation of immunogenic RNA-DNA hybrids. Overall design: One aim of the study was to investigate how treatment with CBL0137 and PlaB affects the DNA binding of MYC. To investigate this, ChIP-Rx using MYC Y69 antibody was performed in U2OSMYC-Tet-On cells treated with doxycycline to induce oncogenic MYC levels and CBL0137 (5 µM, 2 h), PlaB (1 µM, 4 h) or DMSO as control. To further evaluate the behavior of RNAPII and RNAPIIpS2 under these treatment conditions, RNAPII and RNAPIIpS2 ChIP-Rx experiments were additionally performed. 10% murine NIH 3T3 cells were added to each sample as a spike-in. Additionally, a comparison was made of the DNA binding of WTMYC and RBRIIImutMYC in U2OS-AID and KPC cells. For the U2OS-AID experiments, cells were treated with auxin (500 µM, 24 hours) and doxycycline to deplete the endogenous MYC and express either WT or RBRIIImutMYC. The binding behavior was examined in the absence or presence of CBL0137 (5 µM, 2 h). For KPC cells, samples were treated for 48 h with doxycycline to simultaneously deplete endogenous, murine MYC and express WT or RBRIIImut MYC. For U2OS-AID cells, HA-antibody was used for ChIP-Rx. 10% murine NIH 3T3 cells were added to each sample as a spike-in, but not used for further analysis. For murine KPC cells, ChIP-Rx was performed with MYC antibody (Y69), and samples contain 10% of human spike-in. EtOH condition or IgG served as pulldown controls.