Since 2000, genotype 4 HEV has replaced genotype 1 as the dominant cause of hepatitis E in China [16–18]. Recently, Zhu et al.  identified a novel virus belonging to subtype 4i, the same subtype to which SAAS-FX17 has been assigned. In view of the close relationship between SAAS-FX17 and strain E067-SIJ05C, derived from an acute hepatitis E patient in Japan who had traveled to Shanghai before the onset of disease symptoms, it appears HEV strains belonging to this subtype may undergo zoonotic transmission.
Zhang et al.  previously suggested that the 5' UTR of the viral genome may play an important role in replication and/or translation. Other researchers reported that the 5' UTR and a conserved 58 nt region within ORF1 were likely to fold into conserved stem-loop and hairpin structures, which were postulated to be important for HEV RNA replication . In the present study, a unique nt substitution, C23T, was identified within the putative stem-loop structure of the 5' UTR, which resulted in a change in the size of the second loop. Previous studies have shown that nt variations in the central portion of the 5' UTR may influence the severity of type A hepatitis . However, although a potential virulence determinant for genotype 4 HEV, the impact of C23T on HEV disease symptoms remains to be established.
The 3' UTR and an adjacent region of the HEV genome form a putative stem-loop structure that affects the binding of recombinant viral RNA dependent RNA polymerase (RdRp) and initiation of RNA synthesis . Graff et al.  showed that a seemingly minor change caused by a mutation at nt 7106, which eliminated one putative base pair within the stem-loop, significantly inhibited RNA replication, and the magnitude of virus replication could be the reason for the severity difference of HEV. Therefore, based on these previous research findings, a single nt difference (T3A) identified between the 3' UTR sequences of HEV genotypes 3 and 4 also represents a putative virulence determinant.
HEV ORF1 encodes a large nonstructural protein with several putative functional motifs . Our data derived from sequence comparisons of this region in HEV genotypes 3 and 4 revealed 12 specific aa substitutions, 11 of which occurred in the protease motif. However, since no functional activity relating to disease severity has so far been attributed to this enzyme, these substitutions are unlikely to have a role as virulence determinants. The remaining substitution was located in the RdRp motif, which is essential for genomic RNA replication. In addition, two aa variations were recorded in the conserved initiation motifs of the HVRs of genotype 3 and 4 strains. Virus attenuation that accompanied complete deletion of this region of ORF1 led to the suggestion that the HVR played a biological role in HEV pathogenesis . Therefore, we propose that these two variations together with the substitution in the RdRp motif are potential candidates for virulence variation in type 4 HEV.
ORF2 encodes the viral capsid protein, including a signal peptide (aa 1–22) involved in the translocation of the protein from the endoplasmic reticulum , and an arginine-rich domain (aa 23–111) that may be involved in RNA encapsidation . Córdoba et al.  recently verified that mutations within the latter domain contributed to virus attenuation. Our data comparing HEV genotypes 3 and 4 identified six specific aa substitutions within this region, five of which (aa residues 66–70) were continuous. This entire sequence motif may represent a single putative virulence determinant , implying the possible existence of two such determinants in the entire arginine-rich region. Three structural domains have been defined within the C-terminus of the HEV capsid: S (residues 118–313), P1 (residues 314–453) and P2 (residues 454–606), which function in forming the capsid shell, binding of the virus to host cell receptors, and antigenicity, respectively [29, 30]. Single or multiple variations in the aa sequences of the capsid or envelope proteins resulted in attenuated viral phenotypes [31, 32] and, more recently, Córdoba et al.  verified that HEV attenuation was also linked to mutations in the P1 domain. Our study revealed four, three and four specific substitutions in the S, P1 and P2 domains, respectively of genotype 4 HEV. Each might constitute individual putative virulence determinants although the proximate substitutions at positions 146–147 could represent a single influencing factor.
ORF3 protein is essential for virion release from HEV infected cells . However, it remains unclear if differences in the length of the ORF3 regions of HEV genotypes 3 and 4, or specific aa variations in the encoded proteins, influence the severity of the respective clinical symptoms.
The Junction Region (JR) denotes the genome segment between the stop codon of ORF1 and the putative initiation codon of ORF2, in which a bicistronic subgenomic mRNA encodes both ORF2 and ORF3 proteins of HEV . Cao et al.  demonstrated that nt mutations or a mutation in the stem-loop structure formed within the JR significantly inhibited HEV replication. Furthermore, Shukla et al.  reported that the distance between the initiation codons of ORF2 and ORF3 affected initiation preferences. Therefore, nt mutations and distance variation between the ORF2 and ORF3 initiation codons within the JR of genotypes 3 and 4 may constitute strong candidates for determinants of disease severity. Although four nt insertions at sites 10, 30, 31 and 32 of the JR were identified in this study, the contiguous insertions at positions 30–32 may represent a single putative virulence determinant.