In the cell nucleus, RNA polymerase II (Pol II), in association with the cellular factor DSIF, transcribes cellular genomic DNA into diverse types of RNA, including messenger RNA (mRNA). During early Pol II transcription, pausing occurs and a succession of capping enzymes (RNGTT, RNMT, and CMTR1) synthesize the cap1 at the 5′ end of the nascent RNA. Once the cap is complete, the pause is released and processive transcription elongation begins. The cap is essential for RNA stability, intra-nuclear transport or export and, for mRNAs, to be translated into proteins by ribosomes. In an influenza-infected cell nucleus, FluPol, bound to viral genomic RNA (vRNA), competes for the completed cap, to which it binds via its cap-binding domain, and associates with the DSIF-POL II nascent RNA complex, stalling cellular transcription. The intrinsic endonuclease activity of FluPol then cleaves the nascent RNA around 13 nucleotides downstream from the cap, a process known as cap-snatching. The resulting capped RNA fragment is redirected to the active site of FluPol, hybridizes to the 3′ end of the vRNA and primes initiation of viral mRNA synthesis. At the end of transcription, FluPol adds a poly(A) tail to the viral mRNA via a stuttering mechanism [1]. The capped and polyadenylated mRNA is thus recognized as bona fide cellular mRNA and exported through nuclear pores to the cytoplasm where it encounters ribosomes, which translate the viral mRNA into viral protein. For each viral mRNA synthesized by FluPol, one cellular RNA is necessarily destroyed (Figure 1).
To visualize cap-snatching by cryo-EM, the minimal active complex was reconstituted in vitro [2]. This comprises vRNA-bound FluPol, mammalian Pol II phosphorylated with TFIIH CDK-activating kinase and bound to a 46-mer dsDNA, DSIF and a 35-mer nascent capped RNA. A FluPol point mutant with impaired endonuclease activity was used, allowing trapping of the pre-cleavage state, and then, upon increasing the magnesium concentration, the post-cleavage state. 3D reconstructions were made of both states to about 3 Å resolution (Figure 2). The structures show that the endonuclease domain of FluPol binds to the KOWx-4 domain of DSIF and the cap-binding domain to the RPB1, RPB3, and RPB11 subunits of Pol II. The binding is accompanied by significant reorientation of several DSIF and Pol II domains. The capped RNA emerges from the Pol II RNA exit channel and traverses the endonuclease active site, with the 5′ cap being bound in the FluPol cap-binding site. In addition, FluPol binds to the serine 5 phosphorylated C-terminal domain (CTD) of Pol II, as previously described [3]. Upon activation of the endonuclease, RNA cleavage occurs but structurally very little changes except that the 3’ end of the capped primer reorients into the FluPol active site to initiate viral transcription. Complementary virological experiments show that disruptive mutations in the FluPol interfaces with Pol II or DSIF impair viral replication. This suggests that drugs targeting these interfaces could be developed, but this is probably difficult due to the multivalent nature of the interaction. What happens to the complex as viral transcription proceeds remains to be investigated, but at some point, dissociation of FluPol and recycling of Pol II is expected.
S. Cusack (EMBL)
[1] Wandzik J, Kouba T, Karuppasamy M, Pflug A et al. (2020). Cell, 181, 877-893.
[2] Rotsch AH, Li D, Dupont M, Krischuns T et al. (2026). Nature, 652, 1281-1288.
[3] Lukarska M, Fournier G, Pflug A, Resa-Infante P et al. (2016). Nature, 541, 117-121
Figure 1: Schematic of how cap-snatching from cellular Pol II by FluPol leads to synthesis of a viral mRNA at the cost of a host cell RNA.
Figure 2: Structures of the (a) pre- and (b) post-cleavage cap-snatching complexes as
determined by single-particle cryo-EM.