An Arginine-Rich Motif in the ORF2 capsid protein regulates the hepatitis E virus lifecycle and interactions with the host cell

Hepatitis E virus (HEV) infection is the most common cause of acute viral hepatitis worldwide. Hepatitis E is usually asymptomatic and self-limiting but it can become chronic in immunocompromised patients and is associated with increased fulminant hepatic failure and mortality rates in pregnant women. HEV genome encodes three proteins including the ORF2 protein that is the viral capsid protein. Interestingly, HEV produces 3 isoforms of the ORF2 capsid protein which are partitioned in different subcellular compartments and perform distinct functions in the HEV lifecycle. Notably, the infectious ORF2 (ORF2i) protein is the structural component of virions, whereas the genome-free secreted and glycosylated ORF2 proteins likely act as a humoral immune decoy. Here, by using a series of ORF2 capsid protein mutants expressed in the infectious genotype 3 p6 HEV strain as well as chimeras between ORF2 and the CD4 glycoprotein, we demonstrated how an Arginine-Rich Motif (ARM) located in the ORF2 N-terminal region controls the fate and functions of ORF2 isoforms. We showed that the ARM controls ORF2 nuclear translocation likely to promote regulation of host antiviral responses. This motif also regulates the dual topology and functionality of ORF2 signal peptide, leading to the production of either cytosolic infectious ORF2i or reticular non-infectious glycosylated ORF2 forms. It serves as maturation site of glycosylated ORF2 by furin, and promotes ORF2-host cell membrane interactions. The identification of ORF2 ARM as a unique central regulator of the HEV lifecycle uncovers how viruses settle strategies to condense their genetic information and hijack cellular processes.

Reviewer #2: In their manuscript entitled "An Arginine-Rich Motif in the ORF2 Capsid Protein Regulates the Hepatitis E Virus Lifecycle and Interactions with the Host Cell" Hervouet et al. use an HEV cell culture model to elucidate the role of ARM on multiple steps of HEV propagation. Utilizing a series of ORF2 mutants expressed either in the p6 HEV strain or as ORF2 capsid protein they delineate a crucial role of the ARM as functional nuclear localization signal which seems to be important for viral production. With the help of specific inhibitors and co-localization studies, the authors identified a CRM1-dependent egress from the nucleus and 3 important nuclear export signals within the ORF2 protein. The main body of the manuscript focusses on the maturation and addressing of ORF2. The authors found evidence for a role of a furin-dependent secretion pathway in the maturation process. With the generation and evaluation of additional mutants including the signal and/or the encoding sequence of CD4 by state-of the art techniques the authors delineate a tight interplay of the ARM with the respective signal peptide for addressing, topology and egress of the capsid proteins. Bringing this data set together the authors finally delineate a model for the generation of the infectious or non-infectious forms of the HEV particles. The manuscript is well written, the experiments are appropriate and on a technically high level. Following on the previous reports from the group this manuscript sheds light into the so far under investigated HEV life-cycle and the generation of the different capsid forms. Overall, this manuscript describes several novelties in the field and sets the ground for several future scientific perspectives concerning HEV.

Part II -Major Issues: Key Experiments Required for Acceptance
Reviewer #1: 1) Subcellular localization of ORF2 does not change obviously during infection condition ( Fig  1D) compared with electroporation ( Fig 1A-C), suggesting that it may be an artifact in electroporation condition due to ORF2 overexpression. RESPONSE: This reviewer probably did not notice that this point is addressed in the revised manuscript. As mentioned in page 6 line 126 and illustrated in the figure below, infected Huh-7.5 cells displayed ORF2 populations with different N/C ratio. On one hand, a population displayed a nuclear localization which decreased overtime (in red in the figure below). This ORF2 population displayed a same profile than that of electroporated cells (Fig 1A, B, and C). On the other hand, one or two other populations with higher N/C ratio were observed because at 6 days post-infection, a new round of infection occurs and therefore some cells are newly infected and display a N/C ratio as in early time points postinfection.
To corroborate this observation, we performed a statistic analysis. We used the gaussian mixture model which looks for a combination of normal distributions and homogeneous clusters. The distribution of nuclear fluorescence intensities at day 6 followed a normal distribution (pvalue of the shapiro test = 0.13, same conclusion at 48 hours and 72 hours post-infection). We used the Rmixmod::mixmodCluster function and tested the distribution of fluorescence intensities from 1 to 4 clusters and the default given models (mixmodGaussianModel Finally, we performed an additional control in which we treated HEV-p6 infected Huh-7.5 cells with nuclear export inhibitors. As shown in the new S8A Fig, the N/C ratio of ORF2 fluorescence in treated infected cells increased significantly, as was observed for the treated electroporated cells. These results indicate that, in HEV-infected cells, ORF2 is no longer exported from the nucleus as compared to untreated cells, upon treatment with nuclear export inhibitors. Therefore, the ORF2 protein undergoes the same nuclear import/export pathway in electroporated and infected cells. 2) Fig 2D, there are multiple rounds of infection already, so the phenotype could be due the alterations of ORF3. This concern is not been addressed. According to the result in S7 Fig, the direct conclusion is that deletion of ORF3 don't impair ORF2 nuclear localization. Our question is how to ensure that the fewer infectious particles produced is indeed caused by ARM mutation rather than ORF3 mutation (ORF2 and ORF3 ORFs are largely overlapped)? To answer this question, I think the possible strategies are as follows: (1) When generating ARM mutation to impair ORF2 nuclear localization in the contents of genomic sequence, please make sure the ORF3 amino acids unchanged. (2) utilize the trans-complementation system to uncouple the expression of ORF2 and ORF3 protein, then investigate the ARM mutation on virus production.
RESPONSE: As mentioned previously, since ORF3 and ORF2 are overlapping, we do agree with the Reviewer 1 that mutations or deletions in ORF2 can cause deleterious effect on expression/functionality of ORF3 protein. Therefore, we monitored ORF3 expression (Fig 2C  and 3D) and showed that only DSP and DSP1 deletions were deleterious for ORF3 whereas ARM mutations did not affect ORF3 expression ( Fig 2C). As this protein plays an essential role in particle secretion, ORF3 expression was taken into account for the interpretation of extracellular titers (Fig 2D and 3F).
The ORF3 protein is a small protein with a viroporin activity [1] that has been shown to play an essential role in particle egress but that is not required for particle assembly [2][3][4][5][6]. Therefore, regardless of ORF3 expression/functionality/localization, intracellular infectious titers are representative of HEV particle assembly. As shown in Fig 2D, ARM mutations were totally lethal for assembly of intracellular progeny whereas the ORF3-deficient DSP mutant still produced intracellular particles. Therefore, we are convinced that observed phenotypes are not related to ORF3.