Transcriptomics of host-virus interactions: immune responses to avian influenza virus in chicken

Abstract

Upon entry of the respiratory tract avian influenza virus (AIV) triggers early immune responses in the host that are aimed to prevent or in case of already established infection control this infection. Although much research is performed to elucidate the course of events that follow after AIV infection, the interactions between the virus and the host at molecular level at an early stage of infection are unclear. More insight in the mechanisms underlying innate and adaptive immune responses leading to pathogenesis or elimination of the AIV may contribute to a better understanding of AIV induced pathogenesis and new concepts for vaccine development. In this thesis we unravelled several aspects of early host responses after primary AIV infection and viral challenge in immunised chickens at host transcriptional level. The use of genome-wide microarray analysis in combination with more classical techniques allowed us to elucidate possible correlates of protection or pathogenesis at cellular and transcriptional level. This last chapter summarises the findings in this thesis. Respiratory epithelial cells are the first target of AIV infection and their initial response will affect the development of the host immune response. In mammals early host responses to AIV infection have been investigated using respiratory epithelial cell lines and primary epithelial cell cultures. Since chicken respiratory epithelial cell lines or primary cell cultures were not available at the time, we decided to explore the use of tracheal organ cultures (TOC) for this purpose as described in chapter 2. Although H9N2 AIV was able to infect and replicate in TOC, virus specific gene expression profiles were masked by wound healing responses that are induced by preparation of TOC independent of the virus infection. Therefore only a small overlap was found in host response genes between infected TOC and in vivo infected trachea. We were able to keep TOC in culture for a longer time period, but histopathological changes occurred which would also affect host responses at transcriptional level. This implies that although TOC is a suitable model for culturing of virus and lectin or virus binding studies, it is not suitable for measuring early immune responses upon viral infection at transcriptional level. Upon entry of the respiratory tract AIV encounters several barriers that are able to block the virus from entering epithelial cells. One of these barriers are collectins which in mammals can bind and neutralize a wide range of pathogens including influenza A virus. Mammalian collectins have been implicated to play an important role in the early defense against influenza A virus infection, but for chicken collectins this is not yet clarified. In chapter 3 recombinant chicken Lung Lectin (cLL) was used to characterize structural and functional properties of this collectin. Purified recombinant cLL has lectin activity, but failed to neutralise influenza virus infectivity of the human isolates, while for an avian virus isolate neutralisation was seen once. On the other hand recombinant cLL proved to have haemagglutination inhibiting activity to at least one human influenza virus strain. How expression of chicken collectins is affected by influenza virus infection in the chicken is unknown and this was analyzed in chapter 4. Collectin mRNA expression was down regulated in lung and up regulated in trachea after AIV inoculation, indicating tissue specific expression. We also detected that the effect of AIV inoculation on collectin mRNA expression was age specific in both the trachea and lung and viral RNA expression only correlated with collectin mRNA expression in the lung of 1-wk-old but not in 4-wk-old birds. Without knowing the exact function of chicken collectins in innate defenses, it is difficult to relate the observed changes in gene expression to a biological effect. However, these findings indicate that in lung collectins may play a role in limiting respiratory infection in neonatal chickens. Before we could start with the in vivo analysis of early host responses to AIV infection in the respiratory tract at the molecular level, there was a simple question unanswered in literature. It was described that differences in influenza virus deposition in the lung of macaque had an affect on host responses at transcriptional level. Whether this uneven virus deposition occurred in the chicken was unknown. In chapter 5 we investigated whether differences in anatomy and airflow in the avian respiratory tract affected deposition of virus and subsequent host responses. Although the upper trachea contained more viral RNA than the lower trachea there were no significant differences in gene expression. Lung was divided into 4 segments according to airflow and anatomy, which proved to affect virus deposition in that lung segments containing the larger airways and the bifurcations to the secondary bronchi contained highest viral RNA levels. In lung differences in viral RNA distribution enhanced the differences in gene expression that were already seen in noninfected birds. Host responses shared by trachea, lung L1 (cranial) and L4 (caudal) have been previously described as common response to pathogens in mouse, rat, macaque and human indicating that these are common responses independent of the amount of viral RNA or the type of respiratory inflammatory disease. These common responses involved chemokine activity and inflammatory responses correlating to massive KUL-01+ macrophage influxes. However, in trachea and L1 more genes were expressed due to infection and larger KUL-01+, CD4+ and CD8α+ cell influxes were found compared to L4. These findings suggest that an unequal deposition of pathogens throughout an organ will induce localised responses and sampling at specific sites will affect the outcome of the study. The development of the avian immune system starts early during embryogenesis and reaches maturity several weeks after hatching. It is unknown to what extent and through which mechanisms the host responses to AIV infection are affected by age. For replication and transcription of the influenza virus genome, the virus uses both viral and cellular host factors. Whether age affects expression of these host factors and thereby AIV replication is also unknown. In chapter 6 the effect of age on early host responses to AIV and on host factors affecting replication in the respiratory tract were analyzed. Gene expression between 1- and 4-wk-old birds was compared in PBS inoculated control birds and in H9N2 AIV inoculated birds early after inoculation. When comparing 1- and 4-wk-old control birds, most genes were expressed at a higher rate in 4-wk-old birds, while genes related to innate responses and development of the respiratory immune system were expressed at a higher rate in 1-wk-old birds. These differences in gene expression between the age groups related to differences in tissue development and maturation of the immune system. In AIV inoculated birds gene expression was most affected at 16 h.p.i. in 1-wk-old birds, and at 16 and 24 h.p.i. in 4-wk-old birds. Furthermore, in 1-wk-old birds less genes were affected by AIV inoculation than in 4-wk-old birds and which might be due to age-dependent reduced functionality of APC, T cells and NK cells. Expression of cellular host factors that block virus replication by interacting with viral factors was independent of age or tissue for most host factors. These data show that differences in development are reflected in gene expression and suggest that the strength of host responses at transcriptional level may be a key factor in age-dependent susceptibility to infection, and the cellular host factors involved in virus replication are not. To gain more insight in the mechanisms leading to protection after vaccination with different adjuvants, w/o, Al(OH)3 and CpG, we investigated correlates of protection after AIV challenge, described in chapter 7. H9N2+w/o and H9N2+Al(OH)3 vaccinated birds were protected based on low viral RNA expression. We found that protection after AIV challenge correlated to a lower number of genes induced after challenge and gene expression patterns with a low amplitude of change. The gene expression profile of protected birds showed that expression was especially up regulated at 1 d.p.c., while in unprotected birds higher and prolonged gene expression was found. Protected birds had smaller cellular influxes en high neutralizing antibody titers. These findings suggest that lack of immune activation is the most important correlate of protection after challenge with AIV most likely due to high neutralizing antibody titers resulting in lower viral RNA expression. Early host responses to respiratory virus infections are complex processes providing the first defense against AIV infection and are influenced by various factors as discussed in this thesis. The research into host responses is influenced by the sampling method, because airflow and anatomy affect virus distribution in the lung enhancing differences in host responses between different segments in the lung already seen in uninfected birds. This points out the importance of sampling approach which has to be taken into account when investigating in vivo responses to respiratory virus infections in large organs. Host responses themselves are directly affected by the age and adjuvants. Prolonged and enhanced responses relate to the maturation of the respiratory immune system and may be key factors in age-dependent susceptibility to infection and induction of protective responses. A high neutralizing antibody titer is a know correlate of a protective response. This correlate of protection relates to lower gene expression levels and less cellular influxes indicating that a lack of immune activation is the most important correlate of protection after AIV challenge. The findings presented in this thesis enhance our knowledge on the course of events that follow early after AIV infection leading to pathogenesis or elimination of AIV.