e The host protease TMPRSS2 plays an essential role in proteolytic activation of the influenza A virus (IAV) hemagglutinin (HA) protein possessing a monobasic cleavage site. However, after passages in TMPRSS2 knockout mice, an H3N2 subtype IAV began to undergo cleavage activation of HA, showing high virulence in the mice due to the loss of an oligosaccharide at position 8 in the HA stalk region. Thus, the H3N2 IAV acquired cleavability by an alternative HA activation mechanism/protease(s).
Influenza A virus (IAV), a member of the Orthomyxoviridae family, affects and kills many humans worldwide. The viral hemagglutinin (HA) glycoprotein, which is responsible for receptor binding and subsequent membrane fusion, is synthesized as the inactive precursor, HA 0 , and cleaved by a host cell protease(s) into HA 1 and HA 2 subunits. The cleavage is essential for HA to mediate membrane fusion.Using mouse models, three research groups recently demonstrated that a type II transmembrane serine protease, TMPRSS2, expressed in the airway epithelium, is critically important for HA cleavage in vivo (1-3). Mice lacking TMPRSS2 expression (TMPRSS2 knockout [tmprss2 KO ] mice) are highly tolerant of challenge infection by low-pathogenic (LP) IAVs possessing a monobasic cleavage site in HA. Our group (1) and Hatesuer et al.(2) showed that TMPRSS2 is essential for proteolytic activation of both H1N1 and H3N2 subtype IAVs. However, Tarnow et al. (3) reported that TMPRSS2 is not critical for activation of H3N2 IAV. Since the three groups used different H3N2 IAV strains, the discrepancy likely arose from possible variations in HA among the H3N2 IAV strains. The present study reveals a molecular determinant that modulates the TMPRSS2 dependency of H3N2 IAV in mouse lungs.A mouse-adapted human H3N2 IAV strain, MA-A/Guizhou-X (H3 WT ) (1), was used as a reference wild-type (WT) virus. The virus was a reassortant H3N2 virus of A/Guizhou/54/89 (H3N2), a Chinese human isolate from 1989, and A/Puerto Rico/8/34 (H1N1) (4). The H3WT virus was passaged 10 times in tmprss2 KO mice, and a tmprss2 KO mouse-adapted MA-A/Guizhou-X virus ( H3 p10 ) was obtained.The parental H3 WT and H3 p10 viruses were inoculated intranasally into tmprss2 KO mice (n Ï 5). After challenge infection with 3.0 Ï« 10 3 PFU of H3 WT , tmprss2 KO mice showed no body weight loss, while all TMPRSS2 Ï©/Ï© wild-type mice (tmprss2 WT ) showed severe body weight loss (Fig. 1A). Conversely, both tmprss2 KO and tmprss2 WT mice infected with 1.7 Ï« 10 2 PFU of H3 p10 showed severe body weight loss and required euthanasia on day 6 postinfection (p.i.) (Fig. 1B). To compare the susceptibility between the two mouse strains in detail, both mouse strains were also infected with 1.7 Ï« 10, 1.7, and 0.17 PFU of H3 p10 ( Fig. 1C to E). The H3 p10 infections exhibited pathogenicity in both tmprss2 KO and tmprss2 WT mice, although the pathological changes were slightly less severe in tmprss2 KO mice than in tmprss2 WT mice. The 50% lethal doses of H3 p10 for tmprss2 WT and tmprss2 KO mice w...