Relative thickness of ALI cultures infected with A(H1N1)pdm09 viruses at 8

The relative thickness of the epithelial cell layer of the cultures infected by either of the five viruses was also analyzed at 8 days post-infection. Samples were stained with DAPI to visualize nuclei; antibodies against β-tubulin were applied to detect cilia. As shown in Fig. 17A, a more pronounced reduction of the thickness was observed in ALI cultures of well-differentiated PBEC infected by isolates from 2009/2010; a slightly reduced thickness was detectable in cultures infected with viruses from 2014/2015 compared to mock-infected samples. The thickness of the cultures infected by either of the five viruses was analyzed quantitatively. The result indicated that the difference between the two groups of samples was significant (Fig. 17B)

Taken together, infection of ALI cultures of well-differentiated PBEC by five strains of A(H1N1)pdm09 viruses resulted in (i) virus release over a prolonged period of time;

(ii) the loss of ciliated cells; (iii) a reduction of the thickness of the epithelial cell layer.

These effects are more pronounced after infection by the three viruses from 2009/10 compared to infection by isolates from 2014/2015. These results indicate that the latter viruses are less virulent than the viruses from 2009/10.

Fig. 17 Thickness of well-differentiated PBEC at 8 days post infection

Well-differentiated PBEC cultures were apically infected with either of the five isolates at an infection dose of 10⁴ TCID50/filter. After 8d post infection, samples were subjected to Cy3 labelled antibody against β-tubulin and DAPI to visualize cilia (red) and nuclei (blue). (A) The thickness of well-differentiated PBEC cultures infected with the viruses at 8 days post-infection. The vertical image derived from merged stacks of 5 planes. Scale bar, 50 μm. (B) Quantitative analysis of the thickness of well-differentiated PBEC cultures at 8 days post-infection. The relative thickness (%) of cultures is shown as means ± SEM compared to mock-infected cultures. Nine samples from three independent infections by the viruses were determined, three fields from one sample were evaluated.

B

A

DISCUSSION

1. Ciliostasis of airway epithelial cells facilitates influenza A virus infection

The respiratory tract is exposed to the external environment continuously and thus a target organ for a variety of respiratory pathogens. In order to protect the host from detrimental effects of foreign materials, epithelial cells of the respiratory tract form a functional barrier to prevent infection by viruses and bacteria. Among these functional barriers, the mucociliary clearance system is a first line of defence that microorganisms encounter in the respiratory tract. Mucus-producing cells generate mucins which can entrap foreign substances, and ciliated cells transport the mucus together with entrapped complexes out of the respiratory tract. I have analyzed whether the cilia beating not only has its well-known transport function, but also impedes the viral attachment process. The respiratory pathogen chosen in this project was a swine influenza A virus of the H3N2 subtype which can infect target cells efficiently. The culture system of porcine precision-cut lung slices for differentiated airway epithelial cells was chosen as an ex vivo model to evaluate the function of the cilia because the ciliary activity can be observed easily under the light microscope.

PCLS were infected with H3N2 virus in the absence of ciliary activity for 20 min.

Under these conditions the virus yield was about 2 to 3 times higher compared to samples infected in the presence of ciliary activity. Though this difference was not so dramatic, it may be relevant. In the infection of a host, more infection cycles have to be taken into account; a small effect in one round of infection may be increased in the infection of a multicellular organism.

To interrupt the ciliary activity of the airway epithelial cells, PCLS were treated with 2% NaCl which can induce reversible ciliostasis when the treatment does not take longer than 30min. This concentration of NaCl did not affect the infectivity of influenza viruses. It has to be mentioned that prior to such an analysis, control experiments have to be performed to make sure that the ciliostatic agent does not affect the infectivity of microorganism. In this context, it’s unfortunate that Streptococcus suis was affected by hypertonic salt treatment. The bacteria did not die but they stopped growing. Therefore, it was not possible to analyze whether the cilia beating impedes the attachment process of this pathogen. In the future, it will be

necessary to find suitable reagents to address this question. Pharmacological compounds that induce reversible ciliostasis may help to get around this problem.

2. In vitro studies in porcine differentiated airway epithelial cells reveal an evolution of human-derived A(H1N1)pdm09 viruses towards lower virulence between 2009 and 2015

Pigs are an important host for influenza A viruses and play an important role in the interspecies transmission. They are considered as “mixing vessel” for the generation of reassortant viruses containing gene segments from both a mammalian and an avian parental virus. The latest pandemic event emerged in 2009. The pathogen responsible for this pandemic is a reassortant H1N1 virus that was transmitted from pigs to humans. Comparative analyses of human viruses isolated in the early years after the pandemic are limited (Meunier et al, 2012; Elderfield et al, 2014) and have not been reported for pigs as an infection model. In this study, five different strains of pandemic H1N1 2009 viruses were evaluated in pigs. Among these viruses, two strains were isolated in 2009, one each isolated in 2010, 2014 and 2015. The virus from 2014 was derived from a pig. In parallel to my experiments with cultures of differentiated airway epithelial cells, our collaboration partner Ralf Dürrwald performed infections studies with pigs. Before experimental infection of animals, pigs were treated with antibiotics to minimize the effects of bacteria which can infect airway epithelial cells. The pathogenicity of the viruses isolated in 2014 and 2015 was lower than that of the viruses from 2009 and 2010. The difference was evident from several parameters determined in the experimental infection of pigs: dyspnea, rectal temperature, virus load, and lung lesions. Though the three A(H1N1)pdm09 viruses from 2009/2010 caused disease in infected pigs, a tendency of decreasing virulence (pathogenicity) is detectable already among the early virus isolates. The HA09 and JE09 viruses were isolated in April and June 2009, respectively, and the virulence of the viruses from 2009/2010 in pigs decreased in the order HA09 > JE09 > JE10.

These data indicate that A(H1N1)pdm09 viruses adapt to the new host with reduced pathogenicity in pigs. The decrease was much more pronounced in the 2015 virus and was shared by the virus of 2014 which was isolated from a pig.

The 2009/2010 viruses differed from the 2014/2015 viruses not only by four parameters in the pig experiments, the difference was also evident by three parameters when the viruses were compared with respect to their effect after infection of porcine

air-liquid interface cultures of differentiated airway epithelial cells. Infections by the 2009/2010 viruses were associated with higher amounts of viruses released into the supernatant, a more pronounced loss of ciliated cells and a larger reduction of the epithelial thickness compared to viruses from 2014/2015. Though isolates of 2009/2010 possess stronger virulence in porcine ALI culture than viruses of 2014/2015, the barrier function of cultures infected by either of the five isolates was maintained which was demonstrated by ZO-1 staining and measuring the TEER. The data indicate that a regeneration process has occurred during the infection process which compensated for the loss of specialized cells. The findings observed in vitro are consistent with the clinical symptoms obtained in vivo. Viruses of A(H1N1)pdm09 with stronger virulence in ALI culture caused a more severe dyspnea in pigs which may result from the debris of apoptotic cells induced by the viruses, and from the reduced efficiency of the mucociliary defence system of the respiratory tract. The regeneration function and the retained barrier functions of airway epithelial cells may contribute to the fast recovery of infected pigs.

Influenza viruses can adapt to their reservoir host and form stable lineages, for example, three virus lineages of H1N1, H2N2 and H3N2 subtypes are circulating in human population; three virus lineages of H1N1, H1N2 and H3N2 subtypes are circulating in pigs. Occasionally, influenza A viruses can overcome the species barrier to infect a new host. In order to adapt to the new host, viruses have to modify their properties by mutations (Cotter et al, 2014). Highly pathogenic avian influenza A viruses of the H5N1 and H7N9 subtypes infected humans occasionally, and could cause severe disease, but they could not spread among the human population and they didn’t form stable lineages indicating that they have not adapted to humans sufficiently (Simon et al, 2013; Potter, 2001; Neumann and Kawaoka, 2015;

Dortmans et al, 2013; Yu et al, 2013). The pandemic H1N1 2009 influenza A viruses have formed a stable lineage in the human population and replaced the seasonal H1N1 lineage that was prevalent before the 2009 pandemic. Now the pandemic viruses are co-circulating in humans together with the H3N2 subtype (Bedford et al, 2015). Our study has demonstrated that the adaptation of the pandemic H1N1 influenza A viruses in the human population was accompanied by a decrease of the virulence for pigs.

The early viruses of the 2009 pandemic H1N1 caused disease in infected pigs, which may be related to the fact that they were not yet adapted to the new host.

The five virus isolates analyzed in my thesis differ from each other by a number of amino acid changes. Though JE09 and JE10 differ from HA09 by some point mutations, no one of the mutations was shared by both JE09 and JE10. So it is not possible to predict which amino acids contribute to the reduced virulence of the JE09 and JE10 viruses. These results are consistent with other studies that have analyzed early viruses of the 2009 pandemic in mice and ferrets and that were also unable to assign differences in virulence to specific amino acids (Meunier et al, 2012; Elderfield et al, 2014). Strains HA09, JE09 and JE10 of early isolates differ from strains SC14 and KI15 by 20 amino acid changes that are distributed to different proteins (table 1).

There are seven amino acid changes in the HA protein and two in the NA protein.

Among the proteins of the polymerase complex, three amino acid changes were found in the PA protein and two in the PB2 protein. In the NP protein, there are three amino acid changes. One amino acid change each was found in the M1, NS1 and NS2 proteins. These 20 amino acid changes are shared by 84 isolates of pandemic H1N1 2009 influenza A viruses that have been isolated in different parts of Germany, and the sequences of these isolates have been deposited in the GISAID repository. As none of these amino acid substitutions is known as a virulence marker, these mutations may reflect the adaptation of pandemic H1N1 2009 influenza A viruses to humans. The E374K mutation in the HA protein of isolates of pandemic H1N1 2009 influenza A viruses has been reported not only to affect the stability and functional activity of the HA protein but also to increase the infectivity in ferrets (Cotter et al, 2014). This mutation was also found in the strains SC14 and KI15 which have been shown in our study to have low virulence for pigs. So the E374K mutation may contribute to the adaptation, but not to the virulence. Apart from the HA protein, adaptive mutations of pandemic H1N1 influenza A viruses may also be associated with the NA, NP, PA, PB2, M1 and NS proteins.

Table 1 Overview of mutations distinguishing virulent (HA09, JE09, JE10) from low virulent viruses (SC14, KI15)

As far as the virulence for pigs and for porcine air-liquid interface cultures is concerned, strain SC14 which was isolated from swine was similar to strain KI15 which was isolated from a human patient. But compared to KI15, a relatively large number of amino acid substitutions were found in the proteins of SC14 indicating a longer time of adaptation in pigs. We do not know the background of this isolate:

when the virus was transmitted from a human to a pig, how long time this virus had been staying in pigs.

To investigate the contributions of amino acid substitutions (individual or in combination) to the virulence of influenza A viruses would require a large number of in vivo assays in animals. In this study, the result observed from the infection of porcine air-liquid interface cultures for differentiated airway epithelial cells paralleled the results obtained from experimental infection in pigs. Our data indicate that the reduced virulence of strains SC14 and KI15 in pigs compared to strains HA09, JE09 and JE10 is reflected in the results obtained with ALI cultures: loss of ciliated cells, thickness of epithelial cell layer and released virus into the supernatant. The results of ALI cultures fit the pathogenic effects of viruses observed in patients and infected animals (Weinheimer et al, 2012; Elderfield et al, 2014; Meng et al, 2013). Taken together, these results indicate that ALI cultures of airway epithelial cells provide a powerful tool to characterize influenza A viruses. With this culture system available, it will to easier to determine the importance of mutations (individual or combinations)

HA (H3 numbering) NA PA PB2 NP NS1 NS2 M1 Strains 104 166 188 285 376 453 501 241 369 100 321 330 344 354 425 444 498 205 48 80 HA09 D K S K E S E V N V N I V I V V S N T V JE09 D K S K E S E V N V N I V I V V S N T V JE10 D K S K E S E V N V N I V I V V S N T V SC14 N T T E K N K I K I K V M L I I N S A I KI15 N Q T E K N K I K I K V M L I I N S A I

for virulence properties of influenza A viruses.

Despite the reduced virulence of more recent isolates of H1N1 viruses, there are still influenza virus infections by current viruses that result in severe or even fatal disease.

This phenomenon has to be analyzed in more detail in the future. The ALI cultures of differentiated airway epithelial cells may be a valuable tool to be used for these studies.