Pandemic Swine H1N1 Influenza Viruses with Almost Undetectable Neuraminidase Activity Are Not Transmitted via Aerosols in Ferrets and Are Inhibited by Human Mucus but Not Swine Mucus

Abstract

A balance between the functions of the influenza virus surface proteins hemagglutinin (HA) and neuraminidase (NA) is thought to be important for the transmission of viruses between humans. Here we describe two pandemic H1N1 viruses, A/swine/Virginia/1814-1/2012 and A/swine/Virginia/1814-2/2012 (pH1N1low-1 and -2, respectively), that were isolated from swine symptomatic for influenza. The enzymatic activity of the NA of these viruses was almost undetectable, while the HA binding affinity for α2,6 sialic acids was greater than that of the highly homologous pH1N1 viruses A/swine/Pennsylvania/2436/2012 and A/swine/Minnesota/2499/2012 (pH1N1-1 and -2), which exhibited better-balanced HA and NA activities. The in vitro growth kinetics of pH1N1low and pH1N1 viruses were similar, but aerosol transmission of pH1N1low-1 was abrogated and transmission via direct contact in ferrets was significantly impaired compared to pH1N1-1, which transmitted by direct and aerosol contact. In normal human bronchial epithelial cells, pH1N1low-1 was significantly inhibited by mucus but pH1N1-1 was not. In Madin-Darby canine kidney cell cultures overlaid with human or swine mucus, human mucus inhibited pH1N1low-1 but swine mucus did not. These data show that the interaction between viruses and mucus may be an important factor in viral transmissibility and could be a barrier for interspecies transmission between humans and swine for influenza viruses.

Importance: A balance between the functions of the influenza virus surface proteins hemagglutinin (HA) and neuraminidase (NA) is thought to be important for transmission of viruses from swine to humans. Here we show that a swine virus with extremely functionally mismatched HA and NAs (pH1N1low-1) cannot transmit via aerosol in ferrets, while another highly homologous virus with HA and NAs that are better matched functionally (pH1N1-1) can transmit via aerosol. These viruses show similar growth kinetics in Madin-Darby canine kidney (MDCK) cells, but pH1N1low-1 is significantly inhibited by mucus in normal human bronchial epithelial cells whereas pH1N1-1 is not. Further, human mucus could inhibit these viruses, but swine mucus could not. These data show that the interaction between viruses and mucus may be an important factor in viral transmissibility and could be a species barrier between humans and swine for influenza viruses.

FIG 1

pH1N1_low viruses showed almost undetectable neuraminidase activity, but no change in viral neuraminidase protein content was detected. (A and B) The mean neuraminidase (NA) activity of pH1N1_low viruses was significantly less than the mean NA activity of pH1N1 viruses and was almost undetectable, as measured with MUNANA (P < 0.01 at a 0.5 dilution, n = 4) (A) and fetuin (P < 0.001 at a 0.5 dilution, n = 4) (B) as the substrates. (B) The mean NA activity of pH1N1_low viruses was also significantly less than the mean NA activity of other pandemic H1N1 viruses isolated in swine during the period 2009 to 2011 in the United States (pH1N1 viruses, P < 0.001 at a 0.5 dilution, n = 25) (refer to Table 1 for strain information). (C) The enzyme kinetics of the NA in pH1N1_low-1 and pH1N1-1 were also reduced compared to that in A/Brisbane/59/2007 (H1N1), as measured using MUNANA as a substrate (V_max = 0.03, 0.10, and 0.47, respectively; K_m = 413.3, 33.31, and 51.98, respectively; n = 2). (D) Viral growth kinetics in Madin-Darby canine kidney (MDCK) cells showed that the replication kinetics of pH1N1_low viruses (red and purple) were similar to the replication kinetics of pH1N1 viruses (blue and green) (n = 3). (E) Separation of viral proteins by SDS-PAGE did not reveal any sizeable changes in protein content between these viruses, as measured using densitometry. M, marker; lane 1, pH1N1_low-1; lane 2, pH1N1_low-2; lane 3, pH1N1-1; lane 4, pH1N1-2. (F) Comparing the intensity of the NA bands normalized to the nucleoprotein (NP) bands of each virus in panel E also did not reveal any appreciable differences between pH1N1_low and pH1N1 viruses. **, P < 0.01; ***, P < 0.001.

FIG 2

Exogenous neuraminidase increases the size of plaques formed by pH1N1-1 and pH1N1_low-1 in Madin-Darby canine kidney cells. Exogenous neuraminidase did not increase the number of plaques formed by pH1N1-1 (A and B) or pH1N1_low-1 (C and D) but significantly increased the mean size of plaques formed by both pH1N1-1 and pH1N1_low-1 (P < 0.001). (E) The mean size of plaques formed by pH1N1_low-1 was significantly smaller than the mean size of plaques formed by pH1N1-1 (P < 0.001). The difference between the mean size of plaques formed by pH1N1-1 and the mean size of plaques formed by pH1N1_low-1 in the presence of exogenous neuraminidase was also statistically significant (P < 0.001). The mean size of plaques formed by pH1N1_low-1 increased 203.3% in the presence of exogenous neuraminidase compared to a 145% increase in the mean size of plaques formed by pH1N1-1, indicating that exogenous neuraminidase had a greater effect on plaque formation by pH1N1_low-1 than on plaque formation by pH1N1-1. Exogenous neuraminidase increased the mean size of plaques formed by pH1N1_low-1, but the mean size of the plaques was still significantly smaller than the mean size of the plaques formed by pH1N1-1 (P < 0.05). *, P < 0.05; ***, P < 0.001.

FIG 3

pH1N1_low-1 is significantly less sensitive to oseltamivir than were pH1N1-1 or A/Brisbane/59/2007 (H1N1). Oseltamivir resistance assays using Madin-Darby canine kidney (MDCK) cells showed that pH1N1_low-1 was significantly less sensitive to oseltamivir than were both pH1N1-1 (P < 0.01) and A/Brisbane/59/2007 (P < 0.01), which showed similar sensitivities (mean EC₅₀ [μM] = 33.94 ± 19.59, 0.26 ± 0.09, and 0.11 ± 0.02, respectively). SD, standard deviation.

FIG 4

pH1N1_low viruses show increased binding affinity of hemagglutinin to sialic acids. Binding affinity to α2,3 sialic acids was almost undetectable for all viruses (A), while binding affinity of pH1N1_low viruses to α2,6 sialic acids was greater than that of pH1N1 viruses (B and C). N.C, negative control, allantoic fluid from uninfected eggs. **, P < 0.01; *, P < 0.05 (calculated by comparing the mean of pH1N1-1 and -2 to the mean of pH1N1_low-1 and -2 at each dilution of sialylglycopolymer [n = 2]).

FIG 5

Transmissibility of pH1N1_low-1 is significantly attenuated compared to that of pH1N1-1 in ferrets. Viral titers, expressed as log₁₀ 50% tissue culture infectious doses (TCID₅₀), were measured in ferret nasal washes on days 3, 5, 7, 9, and 11 postinfection with pH1N1-1 (A and C) or pH1N1_low-1 (B and D). (A and B) All inoculated (I) ferrets shed virus, but the duration of shedding was reduced in pH1N1_low-1-infected ferrets. (C and D) Transmission of pH1N1_low-1 to direct-contact (DC) ferrets was delayed compared to transmission to pH1N1-1-infected ferrets, and transmission of pH1N1_low-1 to aerosol-contact (AC) ferrets was abrogated. Each bar represents a viral titer obtained from a single ferret. +, ferret died 7 dpi showing pathology typical of influenza virus infection. Asterisks denote significantly greater mean titers in samples taken from ferrets infected with pH1N1-1 than in those infected with pH1N1_low-1 at the indicated day postinfection. *, P < 0.05; ***, P < 0.001.

FIG 6

pH1N1_low-1 replication in normal human bronchial epithelial cells is inhibited by mucus. (C and E) Viral titers, expressed as log₁₀ 50% tissue culture infectious doses (TCID₅₀), in washed and unwashed normal human bronchial epithelial (NHBE) well infected with pH1N1-1 and A/Brisbane/59/2007 were similar at all time points (dashed and solid lines). (A, C, and E) The mean viral titer in washed wells infected with pH1N1_low-1 at 24 h postinfection (hpi) was significantly less than that in washed wells infected with pH1N1-1 and A/Brisbane/59/2007 at 24 hpi (P < 0.01 and P < 0.001, respectively, n = 3) (asterisks shown in panels C and E at 24 hpi). (A) Virus was not detected in unwashed wells 2 and 3 infected with pH1N1_low-1, and well 1 was also negative at 24 hpi (solid lines). Further, the viral titer at 48 hpi in well 1 was approximately 2-fold less than the viral titers in washed wells infected with pH1N1_low-1. The differences between the mean viral titers in washed and unwashed wells infected with pH1N1_low-1 were statistically significant at 24 and 48 hpi (P < 0.001 and P < 0.05, respectively). (B, D, and F) Mucus content was similar in all washed wells at each time point (dashed lines). (B, D, F, and G) The differences between the mean mucus contents of washed and unwashed wells at time zero were statistically significant in all experiments (P < 0.001, n = 3). (G) Mucus content in uninfected unwashed wells remained stable throughout the experiment, while mucus content in uninfected washed wells recovered to levels comparable to the amounts in the unwashed wells within 24 h. (D and F) In wells infected with pH1N1-1 and A/Brisbane/59/2007, mucus content in unwashed wells was similar to that in washed wells at 48 hpi and later time points. (B) However, in wells infected with pH1N1_low-1, the mean mucus content in unwashed wells was significantly greater than that in washed wells at 0, 48, 72, and 96 hpi (P < 0.01 at 48 hpi and P < 0.05 at subsequent time points, n = 3). (A and B) Further, mucus content in pH1N1_low-1-infected well 3 decreased coincidentally with increasing viral titer. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

FIG 7

Neuraminidase activity is important early in infection. Washed (A, C, E, and G) and unwashed (B, D, F, and H) normal human bronchial epithelial (NHBE) cells were infected at an MOI of 1 with the respective viruses, incubated for 8 h at 37°C, and then fixed and immunolabeled using monoclonal antibodies against β-tubulin (red), which stains bundles of cilia on the apical surface of the cells, and influenza virus nucleoprotein (green). Scale bar = 50 μM. Counts of infected cells, as determined by positive influenza virus nucleoprotein staining, did not show significant differences between pH1N1-1 and pH1N1_low-1 viruses in washed wells (I), although the numbers of cells infected by these viruses were much lower than the numbers of those infected by A/Brisbane/59/2007. (J) In unwashed wells, the number of cells infected by pH1N1-1 was significantly greater than the number of those infected with pH1N1_low-1 (P < 0.05, n = 3). (I and J) Exogenous neuraminidase (NA) increased the number of cells infected by pH1N1_low-1, but this was not statistically significant. (K) The difference between washed and unwashed cells in the numbers of infected cells was greater in pH1N1_low-1-infected and pH1N1-1-infected wells. Numbers of cells infected by A/Brisbane/59/2007 were too great to count; therefore, they are not included in panels I, J, and K. Bars show the means ± standard errors of the means (SEM) in panels I and J. *, P < 0.05.

FIG 8

Infection is inhibited by human mucus but not swine mucus. Madin-Darby canine kidney (MDCK) cell monolayers were overlaid with swine mucus (B, E, and H), human mucus (C, F, and I), or no mucus (A, D, and G) prior to addition of pH1N1_low-1, pH1N1-1, or A/Brisbane/59/2007 (H1N1) at an MOI of 1. Cultures were fixed 7 h postinfection and stained for influenza virus nucleoprotein (green) and cell nuclei with DAPI (blue). Human mucus greatly decreased the number of cells infected by all viruses, while cultures overlaid with swine mucus appeared similar to cultures to which mucus was not added. Scale bar = 10 μM. Images are representative of the results of 3 replicates of each experimental condition.

FIG 9

Human mucus significantly inhibits viral infectivity, while swine mucus does not. Madin-Darby canine kidney (MDCK) cell monolayers were overlaid with swine mucus, human mucus, or no mucus prior to addition of pH1N1_low-1 (A), pH1N1-1 (B), or A/Brisbane/59/2007 (H1N1) (C) at an MOI of 1. Cultures were fixed 7 h postinfection, and both infected and uninfected cells were counted in five representative fields in each well. Data are presented as mean percentages of cells infected. *, P < 0.05; **, P < 0.01; ***, P < 0.001.