Coronaviruses, with the largest known viral RNA genome, infect birds and mammals, including humans, and cause diseases ranging from subclinical to lethal. Cis-acting RNA signals critical for genome replication, and for a necessary RNA-dependent RNA polymerase template switching step responsible for subgenomic mRNA synthesis, are only partially defined as are their protein transacting partners. We have identified four 5'-proximal higher-order cis-acting structures within a contiguous 421-nt region, the only region missing by comparison to the shortest non-replicating subgenomic mRNA, that are required for replication of a bovine coronavirus defective interfering RNA. The required structures, which are homologous in the mouse hepatitis coronavirus, are stem-loops III and IV within the genomic 5'untranslated region and stem-loops V and VI within the adjacent nsp1 coding region. We hypothesize one or more of the stem-loops in the (+) strand are needed at three steps in virus replication. First, we postulate one or more are needed, perhaps through cyclization, to inhibit translation and initiate (-)-strand synthesis for genome replication. Second, we postulate stem-loop IV is needed for the critical template switching step for subgenomic mRNA synthesis during (-)-strand elongation. Third, we postulate stem-loop III, possibly in a role mechanistically linked to replication, functions as an encapsidation signal during genomic (+)-strand synthesis. (1) With a well-established reverse genetics system for the closely-related mouse hepatitis coronavirus and site-directed mutagenesis, we will (a) test the importance of each of the four stem-loops in vivo for (i) genome replication and (ii) subgenomic RNA synthesis (template switching), and (b) characterize suppressor mutations in revertant viruses in order to identify and then prove putative mechanistic interactions in vivo. (2) With affinity pull-down and mass spectrometry we will identify the putative viral proteins of 55, 38, and 22 kDa that UV-crosslink to stem-loop III and nine abundant cellular proteins that UV-crosslink to stem-loops III-VI in order to characterize potential components of the replication/transcription complex. (3) With N, the putative 55 kDa stem-loop III-binding protein purified from E. coli and from infected cells, we will determine what role stem-loop III and its flanking sequences play in encapsidation. These results will characterize cis- and trans-acting signals in a critical 421-nt 5'-proximal genome region required for essential features of coronavirus infection.
The human coronaviruses OC43 (a cause of human respiratory disease) and 4408 (a cause of human gastroenteric disease) are thought to have arisen from the bovine enteric coronavirus, a close relative of the mouse hepatitis coronavirus, and the human SARS coronavirus is thought to have arisen from a bat enteric coronavirus. All of these (except for the bat virus which has not yet been cultured) have been shown to share common genome replication signals (that is, some of the replication signals in the bovine coronavirus and in the SARS-coronavirus genome function for replication in the mouse hepatitis coronavirus genome, and so on). Shared signals among coronaviruses for genome replication and for subgenomic messenger RNA synthesis are likely to contribute to virus recombination and the cross-species selection of new coronavirus pathogens. Knowing in detail what molecular interactions lead to coronavirus replication and recombination, the focus of this proposal, are essential for the development of stable and safe recombinant vaccines and the design of anti-viral drugs.
