Genetic Analysis And Receptor Studies On SARS And SARS-like Coronaviruses

A new identified coronavirus has been demonstrated as the causative agent ofsevere acute respiratory syndrome (SARS) that emerged in 2002-2003, designatedSARS coronavirus (SARS-CoV). Shortly after the genomic determination ofSARS-CoV, the origin of SARS becomes a focus for many virologists andepidemiologists. Previously, SARS-like coronaviruses (SL-CoVs) were found in batRhinolophus by our group, demonstrating bats are a natural reservoir of SARS-likecoronaviruses. On the basis of our previous researches, the following work wasundertaken and several crucial findings were obtained:1. Two full-length sequences of bat SL-CoVs were determined, one fromRhinolophus ferrumequinum (designated SL-CoV Rf1) and another from Rhinolophusmacrotis (designated Rm1). Comparative analysis showed that bat SL-CoVs Rf1 andRm1 have similar genomic organization as that of SARS-CoV and, the similaritiesamong all bat SL-CoVs and between bat SL-CoVs and SARS-CoV were 87-92%.Like SARS-CoV, all bat SL-CoVs had similar transcription regulatory sequences(TRS) and s2m motif as well as the existences of stem-loop and pseudo-knotstructures understream of nucleocapsid (N) gene. The variable regions on the genomeof bat SL-CoVs were mainly concentrated in the genes of spike (S), ORF3a andORF8. The ORF8 was only found in the genomes of all animal SARS-likecoronaviruses and early human SARS-CoVs. Genetic analysis showed that batSL-CoVs were clustered with human and palm civet SARS-CoVs in groupâ…¡coronaviruses, although they were distantly related to the recognized groupâ…¡coronaviruses. Together with human SARS-CoV and SL-CoVs from bats and palmcivets, we called these viruses SARS-cluster or group 2b coronaviruses. Phylogenetictrees based on different genes or ORFs were largely the same as that based on genomes. The incongruent trees from different genes suggested the existence ofrecombinant events between coronaviruses. However, no significant recombinationwas detected by recombinant analysis and also no recombinant hotspot was found.For several genes, negative selection pressures between bat SL-CoVs and human orpalm civet SARS-CoVs were detected suggesting that bat SL-CoVs had residedindependently in bats for a long time and were continuously changing in nature.Although the progenitor of SARS-CoV was not found in bats so far, a high diversityof coronviruses circulated in bats, suggesting bats are natural reservoir of all knowncoronaviruses.2. There are two deletions found in bat SL-CoVs within the correspondingreceptor-binding motif (RBM) of SARS-CoV S protein by sequence analysis,implying the receptor of bat SL-CoVs is different from that of SARS-CoV. To verifythis hypothesis, we constructed the pseudotyped virus with SL-CoV Rp3 S proteinand the infectivity of the pseudovirions was detected. The results showed thatHIV/Rp3-S could not infected cells stably expressing ACE2 protein of human, palmcivet or bat R. pearsonii. Also SARS-CoV could not use R. pearsonii ACE2 (RpACE2)as receptors.3. To map the minimal receptor-binding domain (RBD) on SARS-CoV S, a serialof chimeric spikes with different BJ01-S1 fragments on the backbone of Rp3 S wereconstructed. The expression in cells, incorporation and infectivity of the chimericspikes were investigated. On the lever of pseudotyped virus, the minimal RBD ofSARS-CoV S protein was defined from 310 to 518aa. Meanwhile, the chimeric spikeshad varying infectivity on permissive cells.