Effects of co-infection of pandemic influenza and seasonal strains in the ferret model
To establish whether the current pandemic H1N1 virus has a biological advantage over the seasonal H1N1 and/or H3N2 strain(s) and whether reassortment between the pandemic and seasonal strains is rapidly favored in vivo
, we performed co-infection studies in ferrets. The ferret is a well-established model for human influenza infection, pathogenesis and transmission studies 
. Ferrets infected with either seasonal H1N1 (A/Brisbane/59/07) [BR/59] or seasonal H3N2 (A/Brisbane/10/07) [BR/10] virus supported replication and transmission to both direct and respiratory droplet contacts (Fig 1B, D and data not shown). All ferrets shed virus for 5-7 days and seroconverted (Table 1 and data not shown). However, A/California/04/09 (H1N1) [Ca/04], an early representative isolate of the pandemic strain, replicated to higher titers, caused more severe clinical signs, and transmitted to direct and respiratory droplet contact ferrets more efficiently when compared to either seasonal strain (Fig 1A). Tissue tropism studies indicated that both seasonal strains and the pandemic strain were isolated at comparable titers in the nasal turbinate however virus isolated from the trachea of either seasonal strain was on average two folds lower than the pandemic H1N1. Only one ferret infected with BR/59 had detectable virus in the lungs, albeit to titers 3 logs lower than the pandemic virus (Fig 2A). In summary, Ca/04 was isolated in all respiratory tissues sampled and was present at higher titers than either seasonal H1N1 or H3N2 strain, which were only consistently isolated in the nasal turbinate and trachea (Fig 2A). The Ca/04 virus also caused more congestion leading to bronchopneumonia and infiltration of inflammatory cells compared to the seasonal strains (Fig 2B-I).
Figure 1. Influenza respiratory droplet transmission in single and co-infection studies.
Ferrets were inoculated with 106 TCID50 of the corresponding virus combination. Direct contact ferrets and respiratory droplet contact ferrets were introduced at 24 h.p.i. Black and white bars represent individual ferrets. Respiratory droplet contacts shown: (A) A/California/04/09 [Ca/04] H1N1 virus, (B) A/Brisbane/59/07 [BR/59] H1N1, (C) co-infected with BR/59 and Ca/04, (D) A/Brisbane/10/07 [BR/10] H3N2, (E) or co-infected with BR/10 and Ca/04 (E). In (E), day 10 p.c., the respiratory contact from the group represented by the black bars died, as noted by a star in the bar graph. Post-mortem examination indicated intestinal hemorrhage and virus was isolated from respiratory tissues. Titers are expressed as log10 values of TCID50/ml with the limit of detection at 0.699 log10TCID50/ml. Viruses were grown and titrated in MDCK cells as previously described 
. All experiments with live virus were performed in an animal biosafety level-3+ containment facility.
Figure 2. Growth and pathology of pandemic and seasonal influenza viruses in the ferret.
(A) Ferrets were infected with 106 TCID50 of either Ca/04, BR/59, BR/10 or nasal washes from the last day of respiratory droplet contact shedding in the co-infected groups (passed once in MDCKs to achieve a titer of 106 TCID50). Tissues were collected 4 dpi and samples were taken for histopathology and to determine titers and tropism. Nasal turbinate, trachea, lung, intestine, liver, spleen, heart and olfactory bulb were homogenized and titrated in MDCKs. 10% (w/v) of the homogenates were used to determine viral titers in MDCK cells. Abbreviations on figure include: NT, nasal turbinate. Asterisk indicates when only one out of two ferrets was positive for virus. For histopathology studies tissues were paraffin-embedded and stained for hematoxylin and eosin. PBS (B,C), Ca/04 (D,E), BR/59 (F,G), BR/10 (H,I), BR/59:Ca/04 [H1/H1] (J,K), BR/10:Ca/04 [H3/H1] (L,M). Lung (B,D,F,H,J,L) and trachea (C,E,G,I,K,M) samples are shown.
Table 1. HI titers pre- and post-infection of single and co-infection transmission studies
To determine the possibility of co-infection of pandemic and seasonal influenza viruses, ferrets were either co-infected with Ca/04 H1N1 and the seasonal BR/59 H1N1 virus [BR/59:Ca/04] or co-infected with Ca/04 and the seasonal BR/10 H3N2 virus [BR/10:Ca/04]. Successful co-infection was confirmed through PCRs performed directly on nasal washes collected 1-day post-infection (p.i.) with virus-specific primers (Fig. 3C,F), as well as through seroconversion against both viruses in inoculated ferrets (Table 1). Transmission to direct and respiratory droplet contact ferrets was observed in both co-infections (Fig 1C, E and data not shown). In each case, co-infection led to an intermediate transmission phenotype where virus transmitted to direct and respiratory droplet contacts earlier than the seasonal strain and within a day of the pandemic strain. In addition, ferrets shed viral titers higher than the seasonal strain but comparable to the pandemic strain. Clinical signs including onset of sneezing and body temperatures were similar in the co-infected groups; however, ferrets from the BR/59:Ca/04 [H3/H1] group presented with diarrhea and weight loss and one respiratory droplet contact had to be euthanized while still shedding virus (Fig 1E). Tissues collected from the euthanized ferret indicated that virus was present in the nasal turbinate, trachea, lung and olfactory bulb (data not shown).
Figure 3. RTPCR for co-infected BR/59:Ca/04 (H1:H1) and BR10:Ca/04 (H3:H1) respiratory droplet contacts.
Nasal washes from co-infected ferrets (infected, direct contact, respiratory droplet contact) were collected days 1-11 p.i. Nasal washes from respiratory droplet contacts first and last day of shedding were analyzed by PCR using primers specific for BR/59, BR/10, and Ca/04. Data for the last day of shedding from each respiratory droplet contact is shown for the BR/59:Ca/04 (H1:H1) pair day 8 post-contact (A, B) and the BR10:Ca/04 (H3:H1) pair days 7 and 10 post-contact (D, E). Nasal wash from ferrets directly inoculated with BR/59:Ca/04 (H1:H1) and BR10:Ca/04 (H3:H1) virus combinations were analyzed 1 dpi to confirm efficient replication and shedding of both viruses (C, F, respectively).
denotes non-specific binding confirmed by sequencing the PCR product to eliminate suspicion of reassortment.
No evidence of reassortment between pandemic influenza H1N1 strain and seasonal influenza strains
To establish the virus population(s) selected for respiratory droplet transmission following co-infection in both BR/59:Ca/04 [H1/H1] and BR/10:Ca/04 [H3/H1] groups, viral RNA was purified and RTPCRs were performed directly from the nasal washes collected from the respiratory droplet contacts. We selected the first and last days of viral shedding as the most fit virus population to a) transmit to respiratory droplets, b) to survive transmission, and c) to be passed on in respiratory droplet form. We performed RTPCRs on purified viral RNA from nasal washes collected from respiratory droplet contacts and/or stock virus used in our single Ca/04, BR/59, and BR/10 infection studies, as controls for virus-specific primers. In addition, virus in nasal washes from co-infected and single infected ferrets in the respiratory droplet contact groups were amplified once in cell culture (P1 MDCK) to determine which type of virus population was prevalent. Viruses isolated from the first and last days of shedding from respiratory droplet contacts co-infected with either BR/59:Ca/04 (H1/H1) or BR/10:Ca/04 (H3/H1) show a single virus population, a full Ca/04 virus. The data was consistent regardless of whether we analyzed RNA obtained directly from nasal washes or P1 MDCK passage virus (last day shown, Fig 3A, B, D, E and data not shown.) In contrast, directly inoculated ferrets with either the BR/59:Ca/04 (H1/H1) or BR/10:Ca/04 (H3/H1) virus combinations showed both virus populations at 1 dpi, from either nasal washes or P1 MDCK grown virus (Fig 3C and F). HI analysis of the BR/59:Ca/04 (H1/H1) ferret sera indicated that only inoculated ferrets had an antibody response to the seasonal BR/59 while all ferrets had strong responses to Ca/04, suggesting that the Ca/04 out-competes the seasonal H1N1 virus, most likely at direct transmission (Table 1). Interestingly, although a majority of the ferret sera from the BR/10:Ca/04 (H3/H1) group responded to both BR/10 and Ca/04 (Table 1), only Ca/04 was isolated in all nasal washes and organs tested (Fig 3D and E, organ data not shown). To further determine the nature of the predominant virus population in respiratory droplet contacts in the co-infected groups, the nasal washes from the respiratory droplet contact ferrets, collected on the last day of shedding (P1 MDCKs), were used to infect a new set of ferrets and determine which virus population(s) prevailed (Fig 2A). Nasal washes and respiratory tissues collected from this new set of ferrets revealed the presence of high virus titers comparable to Ca/04 single infection (Fig 2A, nasal wash not shown). Interestingly, PCR analysis of tissue homogenates (nasal turbinate, trachea, lung, and intestine) from this set of ferrets revealed the presence of Ca/04 and no evidence of BR/59 or BR/10 (not shown). In addition, histopathology of lungs and trachea from both co-infected groups show similar bronchopneumonia and inflammatory infiltration like Ca/04 single infection (Fig. 2J-M).
These studies emphasize the notion that the pandemic strain has growth and transmissibility advantages over the seasonal strains. More importantly, no evidence of reassortment was detected in PCRs performed on direct nasal washes or nasal washes passed once in MDCK cells in either of the respiratory droplet contacts from either co-infection group (Fig. 3). It should be noted that the pre-infection HI titers for both the H1 and H3 seasonal strains were minimal, although detectable, in all ferrets (Table 1). Thus, we cannot exclude the possibility that the relative transmissible advantage of the Ca/04 H1N1 is merely a consequence of prior immunity in ferrets against the seasonal influenza strains. Our data, however, clearly showed that both the BR/59 (H1N1) and BR/10 (H3N2) replicated and transmitted well in single virus infections. It remains to be determined whether the minimal antibody levels in ferrets against the seasonal flu strains contributed to the Ca/04’s ability to out-compete them. However, we must note that our studies reflect the most likely scenario in the human population as it is expected that most people carry antibodies against H1 and H3 seasonal strains. This is particularly true in those within the 15-40 age group that appear to have shown more exacerbated disease with the pandemic strain. Nevertheless, these observations are consistent with the notion that the emergence of novel pandemic strains in humans may be favored, not only because of lack of immunity against the new virus but also because of the presence of immunity against extant seasonal strains.