For many years mechanistic studies were hampered by lack of a robust cell culture system. Efforts to grow norovirus in a number of well-established cell lines, including primary kidney cell lines, primary intestinal cell lines, and colon carcinoma cell lines, failed [24]. Despite the established tropism of murine norovirus (MNV) for innate immune cells, efforts to replicate human norovirus in the same cell types derived from peripheral blood mononuclear cells (PBMCs) also failed [25]. It has been reported that human noroviruses are capable of replicating in human B cells [26, 27]. Attempts to obtain sustained norovirus replication in B cells using unfiltered, unprocessed stool as inoculum showed that bacterial surface expressed human blood group antigens (HBGA) are important factors for successful virus replication [26]. Additionally, the addition of HBGA-expressing Enterobacter cloacae to the cell culture could restore the infectivity of filtered human norovirus positive stool filtrate, whereas a non-HBGA-expressing bacterium could not [26]. BJAB and Raji B cell lines initially showed promise for replicating human norovirus however these studies have been shown to be difficult to reproduce [26, 27].

Several replicon models that stably express human norovirus RNA have been developed as tools to facilitate studying the immune response to norovirus infection [28–30]. A number of animal models have also been utilized to study human norovirus including non-human primates [31, 32], gnotobiotic pigs [33, 34] and humanized mice [35]. Each of these models had limitations including low levels of virus replication. Recently, zebrafish larvae have been reported as a robust model for human norovirus infection [36]. However, this model does not necessarily represent cells in the human gut and the ensuing innate immune responses after a norovirus infection. Despite the above-mentioned efforts, what remains unclear are the antiviral proteins that specifically restrict virus replication as well as the molecules that recognize norovirus to initiate the antiviral signal pathways. Until recently, much of our understanding of the molecular mechanisms involved in norovirus pathogenesis and immune response has been derived from transformed cell lines and infection with human norovirus surrogate viruses such as feline calicivirus, porcine calicivirus, MNV and Tulane virus [37–39] (Fig. 1).

In recent years, technical advances in the culture of primary human intestinal epithelial cells using intestinal 3D organoid cultures [40] along with availability of advanced molecular technologies, have revolutionized approaches to study the immune response in norovirus infection. Techniques such as gene manipulation, bulk and single-cell RNA sequencing (scRNA-seq), proteomics [41–43], as well as gut-on-a-chip technology, which mimic the intestinal physiological environment [40–42], have the potential to be applied to human norovirus studies in order to move the field forward (Fig. 2).

HIEs and organoid cultures [40, 44–47] have created platforms to study the cellular processes and signalling pathways involved in restricting replication of enteric viruses including human norovirus [48, 49]. Human intestinal organoids (HIO) and HIE are three-dimensional (3D) cultures containing multiple intestinal cell types that are derived from Lgr5+ intestinal stem cells [44] (Fig. 2). HIO contain a mesenchymal niche and are derived from embryonic or pluripotent stem cells (iPSCs) [50], whereas HIE are derived from adult stem cells isolated from intestinal biopsies [40, 51, 52]. These stem cells are propagated in a 3D format supported by Matrigel which allows assembly of the cells into organoid structures that retain cellular composition and physiological functions of the intestinal epithelium [53]. Additionally, adult stem cells are intrinsically programmed with their location-specific function [54], and the differentiated cells that are derived from these stem cells retain an immune profile akin to that of the cells in the corresponding intestinal segment [55]. For replication of human norovirus infection in vitro, these 3D HIE cultures are dissociated and plated as monolayers which are then utilized in a wide variety of studies [56] (Fig. 2).

Successful norovirus replication in vitro using monolayers of HIEs was first reported in 2016 [57] and later confirmed by several other laboratories [58–62]. The cell types found in enteroid cultures include enterocytes, goblet, enteroendocrine, and Paneth cells [53]. Human norovirus potentially replicates in multiple cells types including enterocytes and enteroendocrine cells (EECs) [63, 64]. Presence of human norovirus in enterocytes was first discovered by histological comparison of tissue biopsies from infected and uninfected immunocompromised transplant patients which showed presence of the major capsid protein VP1 in enterocytes from infected individuals [63]. Interestingly, the VP1 expression was also detected in other cells types including, macrophages, T cells and dendritic cells, however, non-structural proteins RdRp and VPg were detected along with VP1 only in enterocytes [63]. Recently human norovirus has been shown to replicate in enteroendocrine epithelial cells (EEC) [64]. Immunohistochemical staining of tissue from the jejunum and ileum of a paediatric intestinal transplant recipient with severe gastroenteritis showed the presence of human norovirus VP1 protein in EEC. Confocal fluorescence microscopy showing colocalization of positive and negative sense human norovirus RNA with the EEC marker (chromogranin A -CgA), confirmed active norovirus replication in this cell type in vivo [64, 65].

The use of commercial media has further optimized human norovirus replication in HIE yielding higher levels of virus replication compared to home-made conditioned media [61]. However, not all norovirus strains can replicate in HIEs [57, 58] with success rate of samples with high viral load as low as 20% [58]. GII.4 viruses demonstrate higher replication levels compared to other genotypes such as GII.3 [57, 58]. A potential explanation for this difference could be that norovirus strains respond differently to the antiviral mechanisms employed by the host cell to restrict virus replication. Other components of the complex intestinal environment, such as the intestinal microbiome and M cells, may also play a role in the strain-specific differences in virus replication [61]. The absence of these components potentially represent major drawbacks of the HIE system.

Human norovirus replication is enhanced or depends on the inclusion of bile acids in the cell culture media [57, 66]. However, the requirement for bile is strain dependent as inclusion is critical for replication of GI.1, GII.1, GII.3, GII.6, and GII.17 strains whereas GII.4 virus replication occurs without supplementation, but is enhanced by bile [61]. This breakthrough has cleared the way for other lines of research including investigation of the antiviral mechanisms that restrict virus replication.

Several research groups have begun exploring whether cells derived from HIE and HIO are amendable to genetic manipulation. Using CRISPR/Cas9 technology, a knockout cell line for the FUT2 gene was created [16], FUT2 encodes an enzyme that affects HBGA expression in intestinal epithelial cells and susceptibility to human norovirus infection [17]. This FUT2 knockout cell line demonstrated diminished replication of GII.4, GII.17, and GI.1 viruses. Further, norovirus replication was shown in secretor-negative J4 cells by knocking in the FUT2 gene, thereby demonstrating that FUT2 expression is necessary and sufficient for norovirus replication in HIEs. While the role of FUT2 has been established epidemiologically, these knockout HIE cell lines provided the genetic basis for this observation [16].

In another study which explored the role of interferon signalling in norovirus infection, lentiviral vectors were used to express proteins that antagonize interferon signalling thereby generating intestinal organoid lines incapable of interferon signalling [59]. Specifically, lentiviral vectors were used to express bovine viral diarrhoea virus NPro or parainfluenza virus type 5 (PIV5) V proteins. BVDF Npro blocks IFN production by degrading interferon regulatory factor 3 (IRF3) whereas the PIV5 V protein compromises IFN production and signalling by targeting key molecules such as STAT1, melanoma differentiation-associated protein 5 (MDA5), and LGP2 for degradation [59]. This study demonstrated that enteroid cells are robust enough for transfection and have the potential to be genetically modified to create cells that reliably sustain norovirus replication. Importantly, it showed that the enteroid cells can be modified to increase virus yield by disrupting the interferon signalling pathway, as demonstrated by a 33-fold increase in norovirus GII.3 replication in BVDF Npro-expressing cells compared to control (nontransduced) cells, and a six-fold increase in PIV5 V protein-expressing cells compared to control cells.

While the receptor for human norovirus is yet to be found, CD300lf (Fig. 1a) has been identified as the primary receptor for MNV [19, 67–69]. Following viral entry, virus particles are recognized by molecular sensors on the plasma membrane, in endosomes, and in the cytosol which trigger induction of an antiviral or inflammatory response [70] (Fig. 1b). The specific sensors that recognize human norovirus have yet to be identified; however, a recent study showed that Toll-like receptors (TLR) 2 and 5 are activated by norovirus virus-like particles (VLPs) [71] (Fig. 1b). Using a TLR2-transfected HEK293 responder cell line, the authors demonstrated that norovirus VLPs can attach to TLR2. Using a TLR5 expressing cell line with an NF-kB-luciferase cassette, they also demonstrated that norovirus VLPs attachment to TLR5 can induce NF-κB driven inflammatory signalling (Fig. 1b, c). These results suggest that TLR 2 and 5 may be involved in recognition of human norovirus leading to induction of an inflammatory response upon infection. Whether this observation can be recapitulated in a physiologically relevant system such as enterocytes derived from HIE requires additional studies.

An early study utilizing 293FT cells transfected with stool-isolated human norovirus RNA showed that while these cells were capable of replicating norovirus RNA, a robust type I interferon response is not induced [72], which was in contrast with the prominent role of type I interferon in the restriction of MNV replication in macrophages and dendritic cells [73–75]. It has now been demonstrated that human noroviruses indeed induce a robust innate immune response chiefly orchestrated by type I and type III interferon [59, 76, 77] (Fig. 1d). Studies with MNV have further dissected this pathway to reveal pivotal roles for transcription factors STAT1 and 2 as well as interferon regulatory factors (IRF) 1, 3 and 7 [59, 78, 79] (Fig. 1b, e). MDA5, another molecular sensor that recognizes single stranded RNA, is also thought to be involved [80, 81]. Using the Norwalk replicon system, it was shown that MDA5 activation by human norovirus RNA results in activation of the JAK-STAT pathway which leads to production of interferon, a cytokine response that induces an antiviral state in infected cells and surrounding uninfected cells [81]. Fig. 1 summarizes what is currently known regarding the antiviral response to norovirus infection. Although some factors involved in the antiviral response to MNV are included, the immune response to MNV infection was out of the scope of this review and has been covered extensively elsewhere [76, 82].

Interferon (IFN) is a major component of the antiviral response that is induced upon norovirus infection [75]. To investigate this, HIE cells that had been treated with exogenous IFN (type I IFN [IFNα1 and IFNβ1] or type III IFN [IFNλ1, IFNλ2, and IFNλ3]) were infected with norovirus GII.3 or GII.4. Both strains showed reduction in replication suggesting that GII.4 and GII.3 norovirus strains are sensitive to IFN [77]. Consistent with this finding, when enteroid-derived IFN-receptor-knockout cell lines were infected with GII.3 and GII.4 strains, both strains showed higher levels of replication compared to infection in wild-type cells [77]. However, GII.3 virus replication was rescued to a greater extent than GII.4, suggesting that GII.3 infected cells are more susceptible to IFN restriction [77]. This was further confirmed using transcriptome analysis which demonstrated that human norovirus elicits a predominantly type III IFN response, and that GII.3 strains induced a more robust IFN-stimulated gene response compared to GII.4 strains [77].

Using a specific Janus kinase 1 (JAK1)/JAK2 inhibitor Ruxolitinib (Rux) to disrupt IFN signalling downstream of the IFN-receptor prior to infection of duodenal IECs with GII.3 or GII.4 strains of norovirus resulted in an increase in GII.4 virus replication. This further demonstrated the importance of IFN signalling in restricting virus replication in HIE [59]. Altogether, these studies clearly highlight the benefit of using HIE in understanding the role of IFN in the antiviral response against human norovirus.

The IFN signalling pathway is a cytokine-based response that results in restriction of virus growth in infected cells and upregulation of antiviral genes in surrounding uninfected cells. IFN secreted from virus-infected cells functions in an autocrine and paracrine manner to engage the IFN-receptors on the cell surface and activate JAK kinases and phosphorylation of STAT1/2 which facilitate upregulation of hundreds of interferon stimulated genes (ISG) (Fig. 1e, f). These ISGs encode effectors of the antiviral response, which antagonize virus replication [83, 84](Fig. 1f).

Little is known about specific antiviral genes that restrict norovirus replication in enteroid/organoid-derived cells. To investigate this, monolayers from two organoid-derived cell lines (terminal ileum organoids) were infected with GII.4 viruses [59] and using RNA-sequencing, 162 genes were found to be differentially regulated in one cell line, and 70 genes were differentially regulated in another cell line [59]. A majority of these were ISGs, demonstrating that human norovirus induces a robust ISG response. The highly upregulated genes included IFI44L, OAS2, OASL, MX-1 and ISG15 which have shown antiviral activity against several viruses including, Zika virus, respiratory syncytial virus, and influenza [85–89].

In another study, transcriptome analysis of two enteroid cell lines using RNA-sequencing also demonstrated a robust transcriptional response 72 h after infection. Additionally, this study found that diverse type I (IFN β) and type III (IFN λ) IFN-driven responses were induced [90]. The use of HIE/HIO has significantly advanced the identification of potential antiviral genes that control norovirus infection. However, much work remains to be done to fully understand the molecular mechanisms by which the antiviral proteins restrict virus replication.