Feline infectious peritonitis (FIP), one of the most important lethal infections of cats, is caused by feline infectious peritonitis virus (FIPV), the high-virulence biotype of feline coronaviruses (FCoVs). FIPVs are suggested to emerge from feline enteric coronaviruses (FECVs) by acquiring mutations in specific genes in the course of persistent infections. Although numerous studies identified mutations predicted to be responsible for the FECV-FIPV biotype switch, the presumed roles of specific genetic changes in FIP pathogenesis have not been confirmed experimentally. Reverse genetics systems established previously for serotype I and the less common serotype II FCoVs were based on cell culture-adapted FIPV strains which, however, were shown to be unsuitable for FIP pathogenesis studies in vivo. To date, systems to produce and manipulate recombinant serotype I field viruses have not been developed, mainly because these viruses cannot be grown in vitro. Here, we report the first reverse genetics system based on a serotype I FECV field isolate that is suitable to produce high-titer stocks of recombinant FECVs. We demonstrate that these recombinant viruses cause productive persistent infections in cats that are similar to what is observed in natural infections. The system provides an excellent tool for studying FCoVs that do not grow in standard cell culture systems and will greatly facilitate studies into the molecular pathogenesis of FIP. Importantly, the system could also be adapted for studies of other RNA viruses with large genomes whose production and characterization in vivo are currently hampered by the lack of in vitro propagation systems.

Feline coronaviruses encode five accessory proteins with largely elusive functions. Here, one of these proteins, called 7b (206 residues), was investigated using a reverse genetic approach established for feline infectious peritonitis virus (FIPV) strain 79-1146. Recombinant FIPVs (rFPIVs) expressing mutant and/or FLAG-tagged forms of 7b were generated and used to investigate the expression, processing, glycosylation, localization and trafficking of the 7b protein in rFIPV-infected cells, focusing on a previously predicted ER retention signal, KTEL, at the C-terminus of 7b. The study revealed that 7b is N-terminally processed by a cellular signalase. The cleavage site, 17-Ala|Thr-18, was unambiguously identified by N-terminal sequence analysis of a 7b processing product purified from rFIPV-infected cells. Based on this information, rFIPVs expressing FLAG-tagged 7b proteins were generated and the effects of substitutions in the C-terminal 202KTEL206 sequence were investigated. The data show that (i) 7b localizes to and is retained in the medial- and/or trans-Golgi compartment, (ii) the C-terminal KTEL sequence acts as a Golgi [rather than an endoplasmic reticulum (ER)] retention signal, (iii) minor changes in the KTEL motif (such as KTE, KTEV, or the addition of a C-terminal tag) abolish Golgi retention, resulting in relocalization and secretion of 7b, and (iv) a KTEL-to-KDEL replacement causes retention of 7b in the ER of rFIPV-infected feline cells. Taken together, this study provides interesting new insights into an efficient Golgi retention signal that controls the cellular localization and trafficking of the FIPV 7b protein in virus-infected feline cells.