H7N9 less efficient by droplet - but occurred in over 30% of ferrets.

Belser and colleagues from Tumpey's group describe their findings after putting two H7N9 viruses (A/Anhui/1/2013 and A/Shanghai/1/2013) into mice and ferrets. Some highlights from the Nature paper released yesterday... 

• Mice inoculated intranasally showed greater disease with human than with avian H7N9 virus 
• Ferrets inoculated intranasally (106 plaque-forming units) did not generally show signs of severe disease, although 2/8 were euthanased because of severe lethargy. 
• H7N9 was detected mainly in nasal turbinate, tracheal and lung tissues of ferrets but also in brain, olfactory bulb, intestinal tract and rectum. Spleen, kidney and liver were virus-negative. H3N2 was found in nasal tissue but not elsewhere. 
• Direct contact between ferrets was an efficient method of transmission (100% of naive ferrets placed in a cage with inoculated ferrets were infected and developed antibodies) but ferrets housed in neighbouring cages with perforated walls were less frequently infected (33% were virus positive and seroconverted). 100% of H3N2 were positive by both modes of transmission. 
• Bird and "human" H7N9 viruses grew better at 37°C (human lower respiratory tract temperature) than 33°C (human upper respiratory tract temperature) in polarized Calu-3 (derived from human bronchial epithelium) cells. They still grew at both temperatures though. 
• H7N9 produced a lot more virus in Calu-3 cells by 24-hours than seasonal H3N2 (human) influenza virus 
• The MDCK cell line required exogenous trypsin to be added for formation of H3N2 and H7N9 plaques 
• H3N2 grew well at either temperature 
• While ocular inoculation of mice with H7N9 or H3N2 did not produce consistent virus replication at the site, virus did grow in the nose at days 3 and 6. Suggest the tear ducts could be a path for self-inoculation as is seen with rhinoviruses. 
• Limited droplet transmission supports the lack of sustained human-to-human transmission seen in China. 
• Shanghai/1 preferred the more avian α2,3-sialosides (sialic acids that end a carbohydrate group which sticks out of a cell's glycoprotein or glcyolipid) whereas Anhui/1 did not favour either, having a mixed α2,3/α2,6-sialoside preference 

In another paper received a little later but released yesterday by Nature, Watanabe and colleagues from Kawaoka's group find the same increase pathogenicity in mice, same proportion of droplet transmissions, using the same human H7N9 strains. Some differences though... 

• The authors also included macaques as a non-human primate and pigs, quails and chickens
• This paper found A/Anhui/1/2013 replicated in the lower respiratory tract (1 or more of tracheal, bronchial or lung tissue) of intranasally infected chickens and quail and also in the upper respiratory tracts of miniature pigs, ferrets and macaques 
• Macaques had a lot of severe lung inflammation starting as early as 3-days after inoculation. 
• In MDCK cell culture (with trypsin?), the human H7N9s grew as efficiently at 33°C as H1N1pdm09 (A/California/4/2009). What caused that difference or did Watanabe not compare both temperatures? IN bronchial epithelial cells however, they took 24-hour longer at 33°C to get to the same viral titres as at 37°C, The H1N1pdm09 reached higher titres than H7N9 at the cooler temperature but was as efficient at the warmer temps. 
• Antibody studies revealed no H7N9 immunity among 500 people in Japan 
• Evaluation of anti-influenza antivirals in mice; Favipiravir out-performed Oseltamivir, Zanamivir and Laninamivir by limiting weight-loss (a sign of illness in mice - they go off their food) and limiting viral loads. 
• Anhui/1, Shanghai/1 and Hangzhou/1 H7N9 HA proteins all preferred α2,6-linked sialosides (human-like); Anhui/1 and Hangzhou/1 most strongly with a special liking for the sialosides covering bronchial epithelial cells. The Shanghai/1 strain bound equally well to α2,3- or α2,3-sialosides. This seems quite different to Belser et al. Technical issue - perhaps to do with the preparation of samples? 

These results are not that different from those previously described by Zhu and colleagues in May, covered earlier on VDU.

If 33% of naive humans challenged with H7N9 might became infected, we might have an explanation for the sporadic nature and low number (considering the potential for exposure) of H7N9 cases in China.

Its also worth noting that even if 33% of humans were infected, and the current PFC remained true (32%) that predicts around 10% of cases dying. Of course its not that simple, and if models and gut feelings can be trusted, there are likely to be many more cases of mild or asymptomatic disease out there that affect the PFC.