Investigations on the immunopathology of enterovirus 71 4

CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS 6.2.6 Fatal-causing strains induced greater cytotoxicity in vitro To elucidate the cellular mechanisms underlying the increased virulence of S10 and S41 in AG129 mice, a plaque formation assay was first conducted to evaluate the cytotoxicity of each EV71 strain in vitro. Human RD cells were infected with the four EV71 virus strains for a period of nine days before staining with crystal violet. The area of plaques was then calculated. Results indicated that S10 and S41 formed larger plaques as compared to C2 and MS strains (Figure 6.6A and B). Since plaques are areas of clearing induced by infection, it appeared that S10 and S41 infection were more effective at infecting and lysing RD cells as compared to the other two strains. To further investigate the mechanisms that account for the difference in the ability to trigger cell death, a cell viability assay was performed. Minimal cellular cytotoxicity was observed for all cells infected with either EV71 strains for the first 24 hours PI(Fig. 6.7A) At 48 hours PI, some cellular cytotoxicity was observed for the cells infected with MS (88%) and C2 (75%) Strains, whereas it was more pronouncedfor the cells infected with S10- (50%) and S41- (51%). At 72 hours PI, the difference between infection of RD cells with fatal and nonfatal strains was further amplified by an additional decrease in cell viability for S10- (41%) and S41- (47%) infected cells. One possible hypothesis that may explain the greater cellular cytotoxicity observed with S10and S41-infected cells is that these two strains could have a greater replication rate in RD cells thereby leading to greater cell death over time, which would also correlate with the 180 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS greater plaque area observed (Fig. 6.6). Figure 6.7B shows the viral yield of infectious particles recovered from the culture supernatant for the four different strains measured at various time points by plaque assay. A one-way ANOVA revealed that there was no significant difference between the virus production rate of the four virus strains in RD cells at the P < 0.05 level. Hence, greater cell death caused by S10 and S41 cannot be explained by a higher rate of virion production, suggesting that other viral virulence mechanisms might play a role in the induction of host cell death. 181 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS A B Figure 6.6 In vitro plaque formation assay. (A) Plaques formed in RD cell monolayer after infection with non-infected control, MS Strain, C2 Strain, S10 and S41 at a MOI of 0.025 for a period of nine day. Red arrows indicate cell boundary separating two plaques. (B) Areas of plaques (mm2) were measured using grid paper and tabulated. The observations are representative of two independent experiments. Data were expressed as mean ± SEM. 182 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS Figure 6.7 In vitro analysis of EV71 strain virulence. RD cells were infected with the four different strains of EV71 at an MOI of 0.025. (A) Following incubation with XTT for hours, absorbance at 450 nm was then measured at the indicated time points. Cell viability at each time point was expressed as a percentage of the absorbance measured for the noninfected cells. (B) Quantitative analysis of virus yield were done by plaque assays on the culture supernatants obtained from various time points PI. Data are expressed as log fold change over time. Results are representative of two independent experiments. 183 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS 6.2.7 Comparative genomic analysis of EV71 strains To identify the specific sequences of more virulent EV71 strains, multiple sequence alignment of the complete EV71 protein-coding sequences were performed according to genomic sequence acquired from GenBank. The ORF of EV71 encodes a polyprotein of 2,193 amino acids. Analysis of the amino acid sequences indicated that differences between fatal and non-fatal causing strains were not limited to one of the specific viral proteins, but a total of amino acid residue differences in VP3, 2A, 2C, 3A, 3C, 3D regions (Table 6.2). In VP3 a proline is found at position 329, in the fatal strains whereas a leucine is present in the non-fatal strains. In 2A protein, an aspartic acid in fatal strains versus an asparagine in non-fatal strains is found at position 919. Another difference was observed in protein 3A, at position 1502 with a valine in the fatal strains and a methionine in the non-fatal strains. Most of the amino acid differences were located in the 3C region, at positions 1627, 1706 and 1730 with isoleucine, valine and methionine respectively in the fatal strains; and threonine, isoleucine and glutamic acid in the non-fatal strains. Two more differences were identified in the 3D polymerase region. Cysteine and alanine were present at positions 1844 and 2167, respectively for the fatal strains, while histidine and threonine were sequenced at these respective positions for the non-fatal strains instead. 184 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS Table 6.2 Amino acid substitutions in fatal and non-fatal causing EV71 clinical isolates. Position Region Fatal Non-Fatal Protein function 329 VP3 P L Capsid Protein 919 2A D N Protease 1502 3A V M Replication Complex 1627 3C I T Protease 1706 3C V I 1730 3C M E 1844 3D C H 2167 3D A T RNA Polymerase 185 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS 6.3 Discussion For EV71, the overall virulence phenotype is probably best measured by the ability of a virus strain to cause neurologic complications in its human host following infection by the oralfecal route (Weng et al, 2010). This phenotype reflects sequential steps in infection including replication in the gastrointestinal tract; invasion of the CNS and intrinsic neurovirulence (Pallansch MA, 2001; Shieh et al, 2001; Yang et al, 2009). A number of clinical studies have demonstrated that these parameters may vary independently for different EV71 strains, as a wide range of clinical manifestations and outcomes have been observed in EV71 infected patients (Shimizu et al, 1999; Bible et al, 2007). What defines the clinical outcome is poorly understood, although several hypotheses of risk factors of both host and virus origin were proposed to be involved in the wide-ranging clinical phenotypes displayed by infected individuals. In the present study, we have shown that infection with different strains of EV71 resulted in different clinical outcomes and manifestations in AG129 mice. Severe disease in S10- and S41-infected mice was correlated with the presence of lesions in the CNS and severe, unresolved myositis in the skeletal muscles. Mice infected with neural and severe muscle lesion-inducing strains exhibited limb paralysis and died as early as day PI. Further tropism analysis of the infected animals showed that fatal-causing strains were highly neurotropic, while the less virulent, non-fatal EV71 strains were transiently present in most organs systemically but did not reach the host’s brain throughout the infection. Thus, limb paralysis observed in S41- and S10-infected mice before death was likely to be caused by the distinct, stereotypic spread of the virus to the brain, resulting in CNS damage that impaired motor function although the contribution of myositis could not be entirely excluded. Similar observations were also made in several other animal models including mouse model using 186 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS mouse-adapted strains of EV71 and cynomolgus monkeys (Wang et al, 2004; Arita et al, 2005). In these mouse models of EV71 encephalomyelitis and the AG129 mouse model reproted herein, the distribution of virus infiltration and inflammation was similar to that seen in human encephalomyelities, suggesting that motors pathways are importatnt for viral transmission into and within the CNS (Wang et al, 2004; Ong et al, 2008). Based on these observations, we hypothesized that neurovirulence is possibly the predominant virulence factor to cause limb paralysis and death in EV71 infection. Our hypothesis is further supported by the observations with C2-infected mice. Despite high quantities of virus infectious particles found in the muscle and spine of C2-infected animals, only mild inflammation was observed in the skeletal muscle during late stage of infection (day 12 post-infection). Consequently, the animals remained asymptomatic throughout the infection suggesting that mice infected with C2 virus did not experience any severe disease causing pathology throughout the course of infection. While we could not entirely rule out the contribution of myositis to EV71-induced disease manifestations, our results suggest that the spread of EV71 from the CNS through the neuronal pathway, rather than from the muscle through the haematogenous pathway, is likely to be the primary reason for the neurological manifestations and for the morbidity and mortality of AG129 mice. This discrepancy in viral tropisms between the different EV71 strains remains to be further investigated but we speculate that the ability to bind to a specific receptor at the surface of neuronal cells could be different between EV71 strains. As EV71 infections for our study were carried out in the immunocompromised AG129 mouse model, caution must be taken when interpreting immune responses. Despite the lack of an interferon response, AG129 mice were reported to be able to mount a normal antibody 187 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS response although with a bias towards the IgG1 subclass (van den Broek et al, 1995). Nonetheless, analyses here encompasses the comparison of immune responses between mice infected with the different EV71 strains instead of determining the absolute immune responses. In general, no significant difference in adaptive immune responses was evoked between animals infected with different strains of EV71. Despite a diminished CD19+ B cell response in mice infected with the non-fatal strains, the downstream systemic IgG and IgM responses were not in concordance in these mice. IgG and IgM responses were in fact greater in non-fatal strains. We therefore concluded that the adaptive immune response may not play a role in determining clinical outcomes observed in mice infected with the different strains. In contrast, the levels of pro-inflammatory cytokine, IL-6 showed a greater potential in discriminating severe from non-severe clinical outcomes in the EV71 infected AG129 mice. Previously, high levels of IL-6, IL-10, IFN-! and TNF-" in the serum and cerebral spinal fluid from EV71-infected patients have been consistently reported (Lin et al, 2002; Wang et al, 2003; 2007). IL-6 and TNF-" levels were elevated in patients with encephalitis and pulmonary edema (Lin et al, 2002). In addition, exogenous IL-6 and IFN-! treatment was reported to induce mild pulmonary edema in EV71-infected mice (Huang et al, 2011), whereas the abrogation of endogenous IL-6 functions improved survival rates and clinical scores of S41-infected neonates (Chapter 3). Collectively, these results support that IL-6 represent a critical host factor that plays a role in the virulence of EV71 infection in both human and mice. Besides enhanced mouse lethality possibly via increased neurotropism and IL-6 production, our data from plaque formation assays demonstrated that S10 and S41 exhibited heightened virulence in vitro compared to the other two strains. A cell viability assay on RD cells also 188 CHAPTER INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS showed that S10 and S41 caused a greater cytopathic effect, resulting in much greater cell death over time. This correlated with the amount of damage observed in the CNS and the limb muscles for both S10- and S41-infected mice. Greater virulence appeared to depend on the ability of the virus to cause cytopathic damage. However, the virus infectious particle yield of the four strains were similar, thus indicating that increased cellular death was not the result of more infectious progeny being produced and released into the supernatant. The ability of EV71 to initiate apoptosis in cultured cells has been widely documented. Apoptosis is likely induced by several non-structural proteins, such as 2A and 3C, that resulted in cleavage of eIF4G and stimulation factor-64 (CstF-64) complex, respectively (Kuo et al, 2002; Weng et al, 2009). One explanation to the phenomenon mentioned earlier is that akin to PV, cell destruction is not required for release of infectious EV71 virions and that interaction between virus and host factors alone may be sufficient to trigger the cytopathic effect by initiating apoptosis (Morrison et al, 1994). Interestingly, we have also identified differences in the 2Apro and 3Cpro coding regions between fatal and non-fatal strains. A mutation in these two proteases in fatal strains could therefore result in better affinity and interaction with host factors such as eIF4G, leading to increased apoptosis of the infected cells. Mutations in other structural and non-structural proteins apart from 2A and 3C regions might also change viral infectivity and virulence. Multiple sequence alignments of the virus strains used in this study revealed that amino acid differences were also seen in the VP3, 3A and 3D regions. The viral capsid VP3 is responsible for receptor binding (Plevka et al, 2012), while 3A and 3D both play critical roles in the replication and life cycle of the virus (Solomon et al, 2010). Molecular genetic approaches such as reverse engineering and site- 189 APPENDIX IV VII APPENDIX IV VIII APPENDIX IV IX APPENDIX IV X APPENDIX IV XI APPENDIX IV XII APPENDIX IV XIII APPENDIX IV XIV APPENDIX IV XV APPENDIX IV XVI APPENDIX IV XVII APPENDIX IV XVIII APPENDIX IV XIX APPENDIX IV XX APPENDIX IV XXI APPENDIX IV XXII [...]... should therefore be focused on elucidating the mechanisms of the interactions of EV71 with DCs, the relative contributions and significance of various cellular receptors as well as the signaling pathways that contribute to the functional defects in DC maturation Further understanding of DC’s role in this aspect should shed light on EV71 pathogenesis, which would in turn facilitate the development of more... determined upon confirming their susceptibility to the virus In addition, comparison of virus yield and replication efficacy between EV71-infected immunocytes would provide key insights into the putative cellular source of EV71 reservoir and hence, contribute to the understanding of EV71-induced immunopathogenesis 198 CHAPTER 7 CONCLUSION 7.3 Development of a novel mouse model for EV71 infection The lack of. .. effective intervention against EV71 infection and might aid in the development of novel strategies for the design of vaccines against EV71 infection In addition, we have shown that EV71-infected BMDCs successfully induced the upregulation of chemokine receptor CCR7 that resulted in enhanced the migration capability of 196 CHAPTER 7 CONCLUSION the immunocytes in vivo Enteroviruses use the enteric tract... would limit the time permissive for the experimental study of EV71 Understanding the mechanisms of virulence of EV71 is crucial not only to develop improved treatment options for clinical care, but also as a predictor of the severity of disease In this 190 CHAPTER 6 INVESTIGATIONS ON EV71 VIRULENT DETERMINANTS study, we have developed a robust platform for the in vivo characterization of EV71 pathogenesis,... made from EV71-infected patients suggested that cytokine storm may contribute primarily to the neuronal dysfunction, which represents the often-lethal stage of EV71 disease (Lin et al, 2002; Wang et al, 2006) To address the salient question in EV71 research, the study herein aimed to address the hypothesis and was the first to confirm the detrimental role of a pleiotropic cytokine, IL-6 in EV71 immunopathogenesis... settings 193 CHAPTER 7 CONCLUSION Second, we have shown that IL-6 plays a fundamental role in the pathophysiology of EV71 infection We showed in the neonate mouse model for EV71 infection that sustained high levels of IL-6 induced upon viral infection are detrimental to the host, leading to severe tissue damage, and eventually death Furthermore, we demonstrated that the administration of antiIL-6 neutralizing... viruses in the gut mucosa Elucidating the interactions of EV71 with DCs in vivo will be vital for uncovering the contribution of DCs to the pathogenesis of EV71 197 CHAPTER 7 CONCLUSION Finally, similar approaches employed in this study should be extended to various target cells of EV71, including T lymphocyte, monocyte, macrophage and B lymphocyte The proinflammatory cytokine and chemokine responses of immune... administration of IL-13 alone failed to do so (Huang et al, 2011) Holistic approaches such as combination therapy that target the signaling cascades as a whole might thus be potentially more efficacious than anti-IL-6 treatment alone due to the simultaneous inhibition of multiple effectors 1 94 CHAPTER 7 CONCLUSION One limitation of the present study is the use of neonatal mice as animal model IL-6 functions mainly... 2011) Thus the development of a transgenic mouse model that 200 CHAPTER 7 CONCLUSION expresses known human EV71 receptors in their native tissue distribution patterns would be ideally suited for the advancement of EV71-related research 201 CHAPTER 7 CONCLUSION 7 .4 Investigations on EV71 virulent determinants in the AG129 mouse model The molecular and genetic factors specifically associated with EV71 virulence... the Taxonomy of Viruses International Union of Microbiological Societies Virology Division Fearon, D.T & Locksley, R.M (1996) The instructive role of innate immunity in the acquired immune response Science, 272, 50–53 Foo, D., Alonso, S., Phoon, M., Ramachandran, N., Chow, V & Poh, C (2007) Identification of neutralizing linear epitopes from the VP1 capsid protein of Enterovirus 71 using synthetic . spread of viruses in the gut mucosa. Elucidating the interactions of EV71 with DCs in vivo will be vital for uncovering the contribution of DCs to the pathogenesis of EV71. CHAPTER 7 CONCLUSION. CONCLUSION 195 One limitation of the present study is the use of neonatal mice as animal model. IL-6 functions mainly as the inducer of acute-phase response and T- and B-cell stimulation including CD40, CD80 and MHC class II. Future studies should therefore be focused on elucidating the mechanisms of the interactions of EV71 with DCs, the relative contributions and significance of various