We demonstrated that use of simvastatin enhances HuNoV infectivity in the Gn pig model. Thus, its use may also support growth of HuNoV in cell culture. We also verified that the increased infectivity of HuNoV may be associated with the inhibitory effect of statins on innate immunity (IFN-α). This observation might explain the exacerbated HuNoV disease and the related higher mortality described in statin-treated humans 
. Because of the immunosuppressive effects, use of statins have been proposed for immunomodulatory therapy against severe influenza A virus infections in which large amounts of innate (IFN-α) cytokines are involved 
. In addition, we showed that oral treatment with nhIFN-α can curtail early HuNoV fecal shedding in the Gn pig model. Because no HuNoV vaccines are available, use of effective antivirals such as IFN-α should be tested to control multiple genogroups and genotypes of HuNoVs, including the GII.4 variants that have emerged each year 
. Our findings that HuNoV infectivity in Gn pigs can be enhanced by simvastatin treatment or reduced by oral treatment with nhIFN-α, suggest that Gn pigs are a useful model to test efficacy of antivirals against HuNoV.
As a surrogate model for HuNoVs, murine NoV (MNV) infection of mice was useful for investigating the roles of specific immunologic factors such as type I or II IFNs in host defense 
. However, the different pathogenesis of MNV infection with its systemic spread raises concerns about extrapolation of these findings to the HuNoV restricted gastrointestinal infection. A chimpanzee model was recently established to evaluate the efficacy of VLP-derived vaccines against infection with GI or GII HuNoVs 
. Chimpanzees developed serum antibody responses after intravenous injection of GI.1/Norwalk virus or intramuscular injection with Norwalk VLPs and were protected from GI.1/Norwalk virus challenge (but not GII HuNoV infection). The chimpanzee model, however, is compromised by the lack of availability of chimpanzees, and the finding that oral infection of chimpanzees with HuNoVs failed to induce gastroenteric disease comparable to human cases 
, as well as by the intravenous route required for viral challenge. In our study, although HuNoV infection of Gn pigs induced mild enteric disease, Gn pigs were susceptible to oral infection by the GII.4 HS194 strain, which reaffirms the results of our previous studies using a closely related GII.4 HuNoV (HS66 strain) 
and the emerging GII.12 HuNoV (HS206 strain) 
. Fecal HuNoV shedding patterns in Gn pigs, with peak viral titers during an early stage of infection are also typical for other acute enteric viral infections in pigs, such as rotavirus and porcine epidemic diarrhea virus 
. However, how HuNoV shedding in infected Gn pigs is maintained for 2 or 3 weeks after viral inoculation is unclear and requires further investigation.
Cholesterol biosynthesis and metabolism are mainly mediated by hepatic enzymes, such as HMG-CoA reductase 
. Statins act as competitive inhibitors of HMG-CoA reductase and reduce production of cholesterol in the liver. As in humans, our study showed that statins lower serum cholesterol levels in Gn pigs, possibly due to similar cholesterol pathways between swine and humans as reported previously 
. When hepatic cholesterol stores are depleted, the liver increases the expression of LDLR which leads to uptake of LDL from plasma to compensate for the lower cellular cholesterol levels. Several RNA viruses manipulate cholesterol pathways in diverse ways for more efficient viral infection and replication as exemplified for NoVs in comparion to hepatitis C virus (HCV) and coronavirus 
. For example, low cellular cholesterol levels (or high cellular LDLR expression) following statin treatment contributed to increased GI.1/Norwalk virus replication, as verified in an in vitro
replicon system 
. Our study also showed that increased LDLR expression levels in IPEC-J2 cells (a porcine jejunal cell line) treated with simvastatin might similarly contribute to enhanced HuNoV replication in the gastrointestinal tract. Our other ongoing in vitro
studies also found that HuNoV RNA titers in supernatants or lysate samples of IPEC-J2 cells treated with simvastatin were slightly increased in trials using GII.12 HS206 strain compared to those of controls (without simvastatin), but did not differ significantly in cell cultures using the GII.4 HS194 strain. The latter was previously reported 
. The data indicate a positive but inconsistent effect of simvastatin on HuNoV replication in vitro
, possibly depending on the HuNoV strains or different environmental conditions for HuNoV replication in vitro
versus in vivo
. A paper describing more detailed and comprehensive in vitro
cell culture findings is in preparation by Takanashi et al. (unpublished data, 2012). In addition to the cholesterol lowering effects, the inhibitory effects of statins on innate immunity also might influence the immunological and cellular microenvironment for more efficient HuNoV replication. In our study, simvastatin impaired TLR3-mediated innate immunity and inhibited production of IFN-α induced by poly (I:C) in PAMs or intestinal DCs. These observations are similar to the results of an in vitro
study using HEK-293 cells, showing the inhibitory effect of simvastatin on TLR4-mediated immune responses, such as tumor necrosis factor (TNF)-α and interleukin-6 
. Nevertheless, it is notable that observations for Gn pigs and humans infected with HuNoVs 
are contrary to the effect of statins that reduced replication of HCV in replicon-harboring cells 
and a positive correlation between cellular cholesterol levels and entry of coronaviruses 
and of GV/MNV into host cells 
. Further confirmatory data are needed to define the role of the cholesterol pathway in the pathogenesis of HuNoV.
Although simvastatin treatment inhibited IFN-α production, we found that gene expression of IRF3 and NFκB in simvastatin-treated PAMs was increased rather than being decreased. Because activation of NFκB kinase is a shared property among TLRs, including TLR3 
, simvastatin or its cellular byproducts could trigger other TLRs that stimulate NFκB gene expression. Notably, at 24 hours after poly (I:C) treatment, gene expression levels of IRF3 and NFκB in poly (I:C) only-treated cells were reduced remarkably, as compared with other treatments or those at the earlier time-point. This observation could be explained by a cellular negative feedback effect to mediate production of IFN-α. It is also notable that a similar result did not occur in simvastatin + poly (I:C)-treated cells, possibly due to reduced IFN-α levels after simvastatin treatment. At least four families of transcription factors are activated by dsRNA and relate to TLR3: NFκB, IRF-3, c-Jun, and activating transcription factor 2 
. Besides lowered TLR3 expression by macrophages after simvastatin treatment, which was thought to be mainly responsible for reduction of IFN-α production in statin + poly (I:C)-treated cells, other TLR3- or IFN-mediated signaling pathways might be involved in the impaired innate immunity by simvastatin.
Type I IFNs are essential for early viral clearance and development of adaptive immune responses. As a crucial mediator of the innate antiviral immune responses, IFN-α has been an effective antiviral treatment for viral infections, such as HCV and influenza 
. The signal transducer and activator of transcription-1 (STAT-1)-dependent IFN stimulation was essential for controlling murine NoV (MNV) infection. Although MNV did not cause disease in immunocompetent mice, oral MNV infection caused fatal systemic disease in mice lacking either type I and type II interferon receptors or STAT-1, which is critical for IFN signaling 
. Several studies have suggested that repeated oral treatment with nhIFN-α may be effective in treating acute viral gastroenteritis related to coronavirus and rotavirus in domestic pigs 
. Similarly, our study showed that oral administration of nhIFN-α inhibits infection by or replication of HuNoV, as fecal HuNoV shedding is curtailed in the Gn pig model. The nhIFN-α is formulated to be stable at low pH of the stomach. Degradation of nhIFN-α by a variety of intestinal enzymes appears to be slow enough to allow nhIFN-α to reach some IFN-α receptors of cells in mucosal lymphoid tissues of the oral cavity and intestine 
. The therapeutic effectiveness of nhIFN-α might be related to its immunostimulatory effects. Orally delivered nhIFN-α promoted systemic innate immunity by increasing expression levels of innate immunity-related genes, such as IFN-stimulated genes (ISGs) and TNF-α, and phagocytic capacity of phagocytes 
. Further studies with larger numbers of animals are needed to determine the most effective dose and regimen of nhIFN-α to prevent or treat HuNoV infections, and to elucidate the immunological and molecular mechanisms related to the antiviral effects of IFN-α. Based on the effectiveness of a combination of nhIFN-α pre- and post-treatment, the nhIFN-α treatments need to be tested therapeutically in future studies using the Gn pig model or in clinical trials. The mechanisms by which fecal virus shedding recurred and the increased viral RNA titers in nhIFN-treated Gn pigs compared to untreated pigs after nhIFN treatment was discontinued need to be investigated. However, we hypothesize that during the period of nhIFN treatment, IFN signaling pathways might be regulated by a negative feedback in some IFN producing cells in the intestine. Thus, on the termination of treatment such a distinct condition of the IFN system might hinder induction or production of IFN-α in most IFN containing cells or its antiviral activity against HuNoV.
In conclusion, simvastatin treatment increased HuNoV infectivity in the Gn pig model, possibly due to its inhibitory effect on innate immunity as well as its cholesterol lowering effect as reported previously 
. These findings could partially explain the exacerbated HuNoV disease in statin-treated humans 
. Testing of nhIFN-α as an antiviral for HuNoV using the Gn pig model also revealed that IFN-α has potential as a HuNoV antiviral therapy. Development of HuNoV antivirals is important because HuNoVs cause large-scale epidemics with significant mortality in immunocompromised, elderly and young patients. Thus, the Gn pig model for HuNoV will allow testing of new treatment modalities for HuNoV infection and new knowledge on the antiviral mechanisms of innate and adaptive immunity.