5' cap sequestration is required for sensing of unspliced HIV-1 RNA by MDA5

Heterogenous transcription start site (TSS) usage dictates the structure and function of unspliced HIV-1 RNAs (usRNA). We and others have previously reported that expression and Rev/CRM1-mediated nuclear export of HIV-1 usRNA in macrophages activates MDA5, MAVS, and innate immune signaling cascades. In this study, we reveal that MDA5 sensing of viral usRNA is strictly determined by TSS, 5 leader structure, and RNA function. We show that cap-sequestered HIV-1 usRNAs (cap1G) destined for viral genome packaging are specifically targeted by MDA5, while translation-destined (cap3G) usRNAs are remarkably immunologically silent. Using mutant viruses which express usRNA with altered 5 cap-exposed leader structure, or inclusion of a retroviral constitutive transport element which drives mRNA-like NXF1-dependent nuclear export of viral usRNA, we show that cap exposure and nuclear export pathway choice are major determinants of both lentiviral RNA immunogenicity and function. In total, we identify innate immune system evasion as a possible rationale for the universal conservation of heterogenous TSS usage among ancestral and extant HIV-1 isolates and shed light on how MDA5 fundamentally discriminates between self and non-self RNAs. SIGNIFICANCEInnate immune activation is critical to both the process of initial infection establishment and ongoing chronic inflammation in HIV-1 infection. While MDA5 has been identified as the sensor which detects unspliced HIV-1 RNA produced in infected cells, it remains unclear how HIV-1 unspliced RNAs, which are generated by cellular transcriptional processes, are recognized as non-self. Here, we reveal that HIV-1 RNA function determines MDA5-driven immunogenicity. We show that only unspliced RNAs which traffic to membrane-associated viral assembly sites are immunogenic, while unspliced RNAs which are ribosomally translated to produce viral proteins are immunologically silent. These findings not only advance our knowledge of how the human innate immune system recognizes HIV-1 unspliced RNAs as foreign but also provide a rationale for the selective advantage to generate two pools of unspliced RNAs during HIV-1 replication.