Hepatitis E causes a self-limiting illness that lasts a few weeks in most patients. Most (>95%) infections are asymptomatic but the disease can be icteric or fulminant. The clinical presentation in developing and developed countries is quite similar [15]. An initial incubation period of 2–6 weeks is followed by symptoms of hepatitis, including fever and nausea, and then by abdominal pain, vomiting, anorexia, malaise, and hepatomegaly. About 60% of patients become jaundiced [16]. Mortality rates during an outbreak can vary from 0.5% to 4.0% of symptomatic infections.

The virus affects mainly young adult males (15–30 years) in developing countries, but pregnant women are particularly vulnerable. The mortality rate may reach 30% during the third trimester of pregnancy [17]. Pregnant women die of obstetric complications such as hemorrhage or eclampsia. Fulminant liver failure can also occur. Stillbirths are common, as is vertical transmission to infants which leads to increased neonatal morbidity and mortality [18]. One study in India found that HEV-related and non-HEV-related acute liver failure during pregnancy had similar mortality rates, although HEV-related acute liver failure was more common during pregnancy [19].

Patients in developed countries who become infected with HEV are usually middle-aged or elderly men (>55 years). Pregnant women do not seem to suffer from severe HEV infections. Patients with underlying liver disease have a poor prognosis in both developing and developed countries [20,21].

HEV3 and HEV4 can persist in immunocompromised patients, including solid organ transplant recipients [22,23,24] and those with human immunodeficiency virus (HIV) infections having a low T CD4+ count (<200/mm³) [25,26,27] or hematological disease [28,29,30,31]. To date, there has been no report of patients with an HEV1 infection suffering from chronic hepatitis E. Chronic HEV infection is defined by persistent HEV replication for more than three months [32]. A chronic HEV infection can lead to chronic hepatitis and progress rapidly to cirrhosis [24,33]. Some of these patients may die from decompensated cirrhosis. Retransplantation of liver transplant recipients infected with HEV usually leads to reinfection of the transplanted graft.

Extra-hepatic manifestations can also occur in patients with acute or chronic HEV infection. These include a range of neurological symptoms and impaired kidney function associated with cryoglobulinemia.

Neurological disorders are the most widely documented, with descriptions of Guillain–Barré syndrome, neuralgic amyotrophy, and encephalitis/meningoencephalitis/myositis [34]. In the Netherlands, 5% of patients with Guillain-Barré syndrome have an associated acute HEV infection [35]. Acute hepatitis E was also found in 10% of patients with neuralgic amyotrophy from the United Kingdom and the Netherlands [36]. The pathophysiology of HEV-associated neurological injury remains uncertain. The immune response triggered by an HEV infection may play a role. It seems likely that Guillain-Barré syndrome and neuralgic amyotrophy are immune mediated. Another possibility is that HEV is directly neurotropic; a kidney recipient with a chronic HEV infection was found to have pyramidal syndrome. Analysis of HEV RNA from the cerebrospinal fluid of this patient showed that the variants differed from those in the serum, at the same time, suggesting the presence of neurotropic variants [37], and possibly HEV replication in the central nervous system. A recent study demonstrated that M03.13 oligodendrocytic cells are permissive and support the HEV lifecycle [38]. Human mesodermal and neuroprogenitor cells derived from pluripotent stem cells support HEV replication only when transfected with subgenomic replicon [39].

Kidney injuries and impaired renal function have been reported during acute HEV3 infections, without any clear explanation. Kidney biopsies from patients with biological glomerular abnormalities revealed histological features of membranoproliferative glomerulonephritis (MPGN) (Figure 3), membranous glomerulonephritis and a relapse of immunoglobulin A nephropathy [40]. Cryoglobulins were detected in the serum in most cases, but only one case of HEV-induced cryoglobulinemic crescentic MPGN in an immunocompetent male has been documented [41]. Anti-HEV IgG, anti-HEV IgM, and HEV RNA were all detected in both the serum and the cryoprecipitate. Renal function improved after HEV clearance [41]. The kidney MPGN disease in patients with a hepatitis C virus (HCV) infection is linked to the deposition of immune complexes formed from the HCV antigen, anti-HCV IgG antibodies, and a rheumatoid factor in the glomerulus [42]. A similar mechanism could be at work in HEV infections. Both HEV antigen and RNA were detected recently in the urine of patients chronically infected with HEV [43]. However, there is still no evidence that HEV is directly nephrotoxic or that it can replicate in renal cells.

Lastly, hematological manifestations, such as aplastic anemia and severe thrombocytopenia, can occur. Acute pancreatitis has also been described [15].

Apolipoprotein E (ApoE) plays an important role in the transport of lipids in the plasma. It may also be involved in the propagation and release of viruses like HCV [47]. Analysis of the proteomes of acutely infected pigs revealed that the ApoE was upregulated, suggesting that it is involved in HEV pathogenesis [48]. ApoE isoforms ε3 and ε4 are significantly associated with protection against HEV infection in American non-Hispanic blacks [49]. Several hypotheses have been formulated to explain this protective role. ApoE could inhibit virus binding by competing with heparan sulfate proteoglycans [50]; ApoE may be in the lipid membrane associated with HEV virions in the blood and could be essential for virus entry [51]; or, lastly, ApoE may modulate the immune response to HEV by regulating T lymphocyte activation and proliferation [52]. It is also possible that this association reflects differences in lifestyle, especially the consumption of pork meat [53].

A (G/A) polymorphism at position 308 in the promotor region of the tumor necrosis factor alpha (TNF-α) is associated with susceptibility to HEV infection. In vitro, the 308A allele results in a seven fold higher TNF-α production [54]. Single nucleotide polymorphisms in the promotor of TNF-α (1031 T/C) and in the promotor of interferon gamma (IFN-γ +874 T/A) seem to contribute to the severity of the disease [55]. The T allele at IFN-γ +874, which is associated with a higher IFN-γ production, was overrepresented in symptomatic cases [55].

Microarray analyses of the intrahepatic transcriptome in serial liver biopsies obtained from chimpanzees infected with HEV or HCV suggest that HEV is more susceptible than HCV to the innate immunity induced by interferon alpha (IFN-α) [56]. However, HEV has developed mechanisms to suppress IFN-α signaling. In vitro studies on A549 human lung epithelial cells [57] and Huh7 hepatocarcinoma cells [58] indicate that IFN-induced phosphorylation of signal transducer and activator of transcription STAT1 can be inhibited by the ORF3 protein, leading to a down regulation of two key antiviral proteins, dsRNA-activated protein kinase and 2′,5′-oligoadenylate synthetase. Nan et al., working on HEK293T cells, showed that ORF3 protein enhanced type I IFN production by interacting directly with the pattern recognition receptor (PRR) retinoic acid-inducible gene I (RIG-I) [59]. In the same cells, they found that ORF1 protein inhibited RIG-I signaling and prevented interferon beta (IFN-β) induction by de-ubiquitination of RIG-I and tank binding kinase 1 [60]. The inhibitory effect seems to be rather small, making only a minor contribution to HEV’s resistance to IFN. However, the gene silencing of the key component of the Janus kinase (JAK)-STAT cascade of the IFN signaling, including JAK1, STAT1, and interferon regulatory factor 9, stimulated HEV infection/replication, indicating that the IFN cascade can restrict HEV infection [58].

Analyses of gene/protein expression in A549 cells infected with HEV showed a robust induction of inflammatory cytokines/chemokines, such as interleukin (IL)-6, IL-8, TNF-α and RANTES (regulated on activation, normal T cell expressed and secreted). HEV infection also led to the activation of both nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) and IFN regulatory factor 3 (IRF3), two transcription factors activated in innate immune signaling pathways [61].

NK and natural killer T (NKT) cells could also play a major role in the innate immune response to HEV. The proportion of CD4+ cells in the peripheral blood of patients with acute hepatitis E is greater than in the blood of healthy controls, but that of CD8+ cells is unchanged. In vitro stimulation with ORF2 protein do not show expansion of HEV ORF2-specific CD4+/CD69+ cells producing helper T cell type 1 (IFN-γ and TNF-α) or helper T cell type 2 (IL-4) cytokines. Conversely, peripheral blood mononuclear cells (PBMCs) from patients acutely infected with HEV have higher IFN-γ concentrations than do PBMCs from controls when stimulated in vitro with ORF2 protein. Thus, CD4+ cells producing IFN-γ could be NKT [62]. The proportions and activation status of the peripheral NK and NKT cells are reversibly altered during acute hepatitis E, with fewer NK (CD3−/CD56+) and NKT (CD3+/CD56+) cells among PBMCs [63]. However, the proportion of activated NK cells in patients with acute hepatitis E is much greater than in healthy controls. The apparent depletion of total NK and NKT cells among PBMCs could reflect the preferential accumulation of these cells in the liver of infected patients [63]. An immunohistological study of liver biopsies from HEV-infected acute liver failure patients showed that their counts of CD56+ cells were significantly higher than in biopsies from patients infected with HAV, HBV or HCV [64].

A serological anti-HEV response is generally detected in patients at the time of onset of illness. Anti-HEV IgMs are detected in the early phase of clinical illness, and can persist for several months. Anti-HEV IgG appears shortly after the IgM response and can last several years (Figure 4). Cross protection is possible due to the existence of only one serotype [65]. The capsid protein contains several neutralizing epitopes [66]. Anti-HEV antibodies can also be induced by vaccination. The only available vaccine, Hecolin^® (Xiamen Innovax Biotech, Xiamen, China), is composed of a truncated HEV capsid protein, p239, that confers protection against hepatitis E infection for up to 4.5 years [67,68].

The risk of HEV reinfection is unclear. Studies on humans and primates indicate that anti-HEV IgG antibodies are protective [69,70,71]. Although the minimum protective concentration of antibodies has not been determined, a vaccine study suggests that an antibody concentration of 2.5 World Health Organization (WHO) units/mL is protective [68]. Solid organ transplant recipients can be reinfected when the antibody concentration is below 7 WHO units/mL [72].

Several studies have provided evidence that effector T cells are activated during acute hepatitis E, with infected patients having more CD8+ cells than healthy controls [73]. The proportions of PBMCs producing IFN-γ in response to stimulation with recombinant ORF2 or ORF3 proteins were also higher in patients than in healthy controls [73]. Using immunohistochemistry in liver biopsies from patients with HEV-induced acute liver failure, Prabhu et al. demonstrated the infiltration of activated CD8+ T cells in liver [64].

An increased expression of CD11a integrin in naïve CD45RA+ T cells, as well as overexpression of CCR5 and CCR9, was also reported. The presence of an expanded CD45RA+ CD11a high subpopulation during the early phase of acute infection suggests enhanced recruitment of these cells from the periphery to the target tissue. Thus, CD45RA+ CD11a high CCR5+ cells are involved in the pathogenesis of HEV infection [74]. Patients with acute hepatitis E were found to have a greater proportion of CD4+ CD25+ forkhead box P3+ (FoxP3+) regulatory T cells and higher concentrations of IL-10, a signature cytokine of regulatory T cells, than healthy controls. This suggests that these cells are important in the acute phase of the disease, although their exact role needs further investigation [75]. Lastly, specific T cell immunity seems also to confer a cross-protection against HEV1 and HEV3 [76], which could protect patients exposed to genotype 3 when they travel to areas where HEV1 is endemic.

The reasons why a hepatitis E infection becomes fulminant are still obscure. Viral factors may be important, as suggested by the observation that HEV4 infections tend to be more severe than those of other strains [78]. However, a recent study re-examined the published evidence for an association between fulminant hepatic failure (FHF) and HEV genotypes and concluded that host factors rather than virus genotype, variants, or specific aa substitutions are responsible for the development of fulminant hepatitis [79].

There is little agreement about the mechanism underlying FHF. Saravanabalaji et al. reported that stimulating both Th1 and Th2 type immune responses could play a role in liver failure. Patients with FHF were found to have higher anti-HEV IgM and IgG titers than those with self-limiting infections [80], and PBMCs from patients with FHF were found to produce higher IFN-γ, TNF-α, IL-2, and IL-10 concentrations after stimulation with ORF2 peptides than PBMCs from controls. In contrast, Srivastava et al. reported less marked antiviral cellular immune responses and heightened humoral antiviral responses in patients with fulminant hepatitis E than in patients with uncomplicated infection and control patients [81]. The heightened humoral response was associated with a more severe HEV disease in both studies.

The situation in peripheral blood may not reflect what occurs at the site of infection. CD4+ T cells are more frequent in the livers of patients with fulminant hepatic failure caused by an HEV infection [64] and CD8+ T cells have been shown to infiltrate the liver of patients with fulminant hepatitis E [64,82]. Thus, cytotoxic CD8+ T cells could be particularly important in the pathogenesis of fulminant hepatitis.

The cause of elevated maternal mortality (30%, with most deaths occurring in the third trimester) of pregnant women infected with HEV1 living in developing countries has been the subject of many studies, but it is still unclear. HEV genotype could explain, at least in part, the poorer outcome in pregnant women since HEV3 is not particularly deadly for pregnant women [83,84]. No data are available for HEV4 infection during pregnancy.

Pregnancy is associated with changes in sex hormone levels and the immune system that are designed to protect the fetus from the maternal immune system. A shift from a Th1-dominated immune response to a Th-2 dominated one,“Th2 bias”, may help protect the fetus by suppressing macrophage activation [85]. Pal et al. confirmed the existence of a Th2 bias in pregnant women infected with HEV, but its implication for the severity of a hepatitis E infection is unknown [86]. Women with acute liver failure (ALF) presented a reduced expression of toll-like receptor (TLR) 3/TLR7/TLR9, a type of PRR that plays a key role in the innate immune system, and have weaker phagocytic macrophages than women with acute viral hepatitis E [87]. However, the phagocytic capacity of the monocytes of the two groups was essentially the same. Impaired monocyte-macrophage function in pregnant women with ALF could contribute to an inadequate innate immune response, and hence to the development and severity of ALF. Kumar et al. reported that high concentrations of cytokines (TNF-α, IL-6, IFN-γ and TGF-β1) may also be associated with an adverse pregnancy outcome [88].

An increased incidence of FHF was reported in pregnant women with the progesterone receptor gene mutations PROGINS haplotype [89]. PROGINS carriers with HEV infection showed reduced expression of progesterone receptor and progesterone-induced blocking factor (PIBF). PIBF exerts its anti-abortive activity by inhibiting NK cells and influencing both the humoral and cellular immune responses [90,91]. Other host factors such as nutritional status or differences in major histocompatibility complex may also influence the immune response of pregnant women to an HEV infection [92]. This may be why an HEV infection is benign in pregnant women in Egypt although it is caused by HEV1 [17].

Lastly, pregnancy-associated hormones can also contribute to a poor outcome. The concentrations of estrogen, progesterone, and β-human chorionic gonadotrophin in HEV-positive pregnant FHF women are higher than in HEV-negative pregnant FHF women or controls [93]. An in vitro study showed that serum from pregnant women, especially those in the third trimester, enhanced the replication of HEV by inhibiting estrogen receptor and type I IFN expression [94]. While some studies have found that the high HEV RNA concentrations in HEV-infected pregnant women were associated with a poor outcome [89,95]; another study by Saravanabalaji et. al, did not confirm these results, with only 1/14 pregnant women having detectable HEV RNA [80].

Non-human primates, including Rhesus and Cynomolgus macaques and chimpanzees, are all primary models for studying the clinical course of infections by the four major genotypes of HEV. However, most of these studies in non-human primates have used intravenous inoculation with HEV because oral inoculation requires much higher doses of virus [46].

Several animal species are naturally susceptible to HEV3 and HEV4. The first non-human strain of HEV was found in pigs [111]. Pigs have been very useful for studying HEV cross-species infections, replication, and pathogenesis [44,112]. They can be used to study chronic HEV infections when co-infected with porcine reproductive and respiratory syndrome virus [113].

Rabbits also provide an interesting model for studying HEV pathogenesis. The rabbit strain of HEV can infect both pigs and macaques, while, conversely, rabbits can be infected with HEV4 [114,115]. Another advantage is that rabbits can be used to simulate the high mortality rate associated with pregnancy in humans [116]. Rabbits can even be used as a model of chronic infection with HEV3 and extra-hepatic manifestations [117].

Ferrets are also a potentially interesting species as they support persistent infections with ferret HEV [118]. Whether ferrets are susceptible to infection with other zoonotic strains remains to be seen. Mongolian gerbils were also successfully infected with swine HEV [119]. Rats are not a very suitable model for human HEV infection, since HEV1, HEV2, HEV3, HEV4 strains do not infect rats [120,121]. Conversely, the rat HEV strain does not infect Rhesus monkeys [121]. Mice with humanized livers have recently been established for studying chronic HEV infections [122,123,124]. They seem to be a valuable tool for exploring HEV replication and evaluating the efficacy of antiviral molecules.

Lastly, chickens have been used to identify extra-hepatic sites of avian HEV replication [125]. The impact of deletion of the PPR on HEV infectivity was also studied with avian HEV strains [126,127].

HEV3 extracted from the feces of a patient with an acute HEV infection has been grown on PLC/PRF/5 (hepatoma) and A549 (lung adenocarcinoma) cells. This initial propagation of an HEV strain was successful because of the high HEV titer in the inoculum (2.0 x 10⁷ copies/mL) [128]. Another fecal suspension of HEV4, this time from a Japanese patient with fulminant hepatitis E, was used to establish a culture system on PLC/PRF/5 and A549 cells [129]. More recently, the HEV3 Kernow strain purified from the feces of a chronically infected patient was also replicated efficiently on HepG2/C3A (hepatoma) cells [108]. Adaptation of the Kernow strain to in vitro growth selected a recombinant virus containing an insertion of 174 ribonucleotides (58 amino acids) from a gene encoding the human S17 ribosomal protein. Other strains with insertions of human S19 ribosomal protein (117 nt) or human inter-α trypsin inhibitor (75 nt) have also been grown efficiently on HepG2C3A cells [106,109]. HEV1, HEV3, and HEV4 strains from serum samples can also replicate efficiently in PLC/PRF/5 and A549 cells. Replication depends on the HEV RNA concentration in the inoculum, but is not influenced by any anti-HEV antibodies in serum [7]. It is important to distinguish between viruses obtained from the feces and serum because the presence of lipid modifies the mechanism by which virus enters the cell [130].

Cells of the human hepatoma-derived cell line HepaRG and the porcine embryonic stem cell-derived cell line PCM19 have been found to support the replication of an HEV3 strain isolated from swine feces. The morphological and functional properties of these two cell lines are similar to those of primary hepatocytes [131]. Lastly, hepatocytes derived from pluripotent stem cells have been shown to support the complete replication cycle of HEV. This is an attractive in vitro model system (non-cancerous hepatocytes) for studying HEV replication [39].