Hepatitis E-related adverse pregnancy outcomes and their prevention by hepatitis E vaccine in a rabbit model

ABSTRACT Hepatitis E virus (HEV) can lead to high mortality during pregnancy. This study was to investigate the adverse pregnancy outcomes caused by different HEV genotypes and their prevention by HEV 239 vaccine in rabbits. Forty-two female rabbits were randomly and equally divided into 7 groups (A-G). HEV 239 vaccine and a placebo were administered to groups E (10 μg×2), F (5 μg×2) and G (1 mL of PBS×2) before copulation. After pregnancy, 1 mL of 1.5×106 copies/mL rabbit HEV3 was inoculated to groups A, E, F and G, swine HEV4/human HEV3 to groups B/C, and group D was a negative control. Anti-HEV antibody, HEV RNA, and alanine aminotransferase (ALT)/aspartate aminotransferase (AST) levels were monitored. Pregnant rabbits infected by HEV manifested HEV infection symptoms including fecal virus shedding, ALT/AST elevation, and histopathological changes, and adverse pregnancy outcomes. Immunized pregnant rabbits in groups E and F showed no HEV infection symptoms and adverse outcomes. The newborn rabbits delivered by pregnant rabbits with/without immunization showed without/with HEV infection symptoms. This study demonstrated that multiple genotypes of HEV infection can cause adverse outcomes and HEV 239 vaccine can prevent HEV-related adverse outcomes in pregnant rabbits.


Introduction
Hepatitis E virus (HEV) is the main cause of acute viral hepatitis in developing countries, which has aroused wide public concern [1][2][3]. Each year, there are approximately 20 million people infected with HEV worldwide, causing 70,000 deaths [4]. HEV, a singlestranded positive RNA virus, spreads through mainly the fecal-oral route and has been recognized as a zoonotic pathogen [5]. HEV can be divided into 8 genotypes (HEV1-8), among which HEV1-4 are closely related to human infection. HEV1 and HEV2 can only infect humans, while HEV3 and HEV4 are zoonotic [1][2][3].
HEV often causes acute self-limited disease, with a mortality of approximately 0.2-1% [3]. However, in pregnant women, the mortality rate of HEV infection can reach as high as 25% [6][7][8][9]. HEV infection in pregnant women can also lead to adverse pregnancy outcomes including miscarriage, stillbirths, fulminant hepatic failure and membrane rupture [10]. However, to date, the mechanism of the high mortality rate of HEV infection in pregnant women remains unclear. HEV-related adverse pregnancy outcomes are caused by mainly HEV1 [10], and whether other genotypes of HEV can lead to adverse outcomes during pregnancy is still unclear. In addition to HEV genotypes, the pathogenesis of HEV-related adverse pregnancy outcomes may be related to changes in cellular immunity and hormones during pregnancy [10], but that possibility warrants further investigation. Currently, there is no recommended drug for HEV-infected pregnant women therefore, an effective prevention strategy is of high priority for this population. Although the HEV 239 vaccine (Hecolin®), which is a bacterially expressed recombinant peptide corresponding to amino acid residues 368-606 of ORF2 of HEV1, has been commercialized in China and can successfully prevent HEV infection for years [11], it still lacks studies on its efficacy for adverse pregnancy outcomes. Previously, we successfully established a HEV-infected pregnant rabbit model that can simulate the HEV-related high mortality rate observed in pregnant women [12]. Therefore, this study aimed to investigate the pathogenicity of different HEV genotypes in pregnant rabbits and the efficacy of the HEV 239 vaccine in preventing adverse outcomes and vertical transmission of HEV infection in pregnant rabbits.

Vaccines
The HEV 239 vaccine (Hecolin; Xiamen Innovax Biotech, Xiamen, China) is a 26 kDa recombinant polypeptide expressed by the Escherichia coli system derived from the 368-606 aa segment of the HEV1 (GenBank D11092) ORF2 protein. It is currently the only commercially available HEV vaccine globally. To date, the HEV 239 vaccine is only available in the Chinese market and has been approved for use in people older than 16 years, which is indicated for vaccinating individuals at high risk of HEV infection [11].

Viruses
The rabbit HEV strain (CHN-BJ-R14, genotype 3, GenBank JQ768461), swine HEV strain (CHN-SD-SW2, genotype 4, GenBank KP284140) and human HEV strain (CHN-SH-W, genotype 3, GenBank MF996356) used in this study were all isolated from feces. Inocula were prepared as described previously [12]. The viral load of each HEV strain was adjusted to 1.5×10 6 copies/mL by using a one-step real-time quantitative PCR assay [12] Experimental design All selected female rabbits (n = 42) were randomly divided into 7 groups (A-G), with 6 rabbits per group ( Supplementary Fig. 1). Before copulation, rabbits in groups E and F were inoculated intramuscularly with a 10 and 5 μg dose of HEV 239 vaccine on weeks 0 and 4 respectively, while rabbits in group G as a vaccine control group were inoculated with a sterile PBS solution. On week 7, all female rabbits (n = 42) copulated with healthy male rabbits. After their pregnancy was confirmed [12], rabbits in groups A, E, F, and G were inoculated intravenously with 1 mL of 1.5×10 6 copies/mL rabbit HEV strain (rHEV3). Rabbits in groups B and C were inoculated intravenously with 1 mL of 1.5×10 6 copies/mL swine HEV4 strain (sHEV4) and human HEV3 strain (hHEV3), respectively, and rabbits in group D were inoculated intravenously with a rabbit feces suspension without HEV infection as a negative control.

Sample collection and processing
After virus inoculation, serum and fecal specimens from pregnant rabbits were collected weekly for serum ALT and AST level, fecal HEV RNA load, and serum estrogen and progesterone level testing. Three weeks after newborn rabbits were born, their serum and fecal samples were collected weekly for serum ALT and AST level and fecal HEV RNA load testing. Livers, ovaries, or placentas were collected immediately from dead HEV-infected pregnant rabbits, euthanized HEV-infected pregnant rabbits, dead newborn rabbits and euthanized HEV-infected young rabbits. Each tissue was stored at −80°C for detection of HEV RNA positive and negative strands and HEV viral load. In addition, the tissue specimens were also processed for histopathology, immunofluorescence and immunohistochemistry (IHC) by being immediately fixed in 10% neutral buffered formalin [12].

Detection of pregnancy-related hormones
To confirm the female rabbits' pregnancy, their serum (day 0, day 10, day 17 and day 24 after pregnancy) and postpartum serum were all tested for progesterone and estrogen by the chemiluminescent enzyme immunoassay method on a Siemens Immulite 2000 Immunoassay System according to the manufacturer's instructions [12].

Detection of ALT and AST concentrations
All serum samples were tested immediately for ALT and AST concentrations using standard methods on a Hitachi Automatic Clinical Analyzer 7180. When the ALT or AST level became more than 2 times the preinfection baseline value, the rabbit was considered to have liver inflammation [13].

Evaluation of anti-HEV antibodies by ELISA and quantitative detection
All rabbit serum samples were detected by ELISA for total anti-HEV antibodies according to the manufacturer's instructions (Wantai, Biopharmaceutical, Beijing, China). When the S/CO value > 1, it was considered to be positive [14]. The geometric mean antibody titre (GMT) and quantitation of anti-HEV antibodies were performed as previously described [15].

Extraction and detection of HEV RNA
Total RNA in serum and fecal suspensions was extracted according to the RNAgents Total RNA Isolation System Kit (Promega, Madison, USA) instructions, and HEV RNA fragments were amplified by RT-nPCR [16]. RT-nPCR was carried out to detect HEV RNA as previously described [16]. Tissues with detectable positive-stranded HEV RNA were then assayed for negative-sense HEV RNA by RT-nPCR with the same 2 sets of universal primers, which used the external forward primer for cDNA synthesis and the amplification conditions were the same as those used in the detection of positive-sense HEV RNA [16].

Real-time fluorescence quantitative PCR
According to the literature [17], real-time fluorescence quantitative PCR (RT-qPCR) in this study was performed by using the Taqman probe detection method and QuantiTect® Probe RT-PCR Kit (Qiagen, Hilden, Germany).

Statistical analysis
Statistical analysis was performed using the SPSS PASW Statistics v18.0 statistical software package (SPSS, Inc., USA, http://www.ibm.com/cn/). Data were compared using Student's t-test or chi-squared test. A P-value of <0.05 was considered significant.

Pathogenicity of HEVs infection in pregnant rabbits
To analyze and compare the results of different genotypes of HEV (HEVs) infection in pregnant rabbits, we inoculated pregnant rabbits in groups A, B and C with rabbit HEV3 (rHEV3), swine HEV4 (sHEV4), and human HEV3 (hHEV3) respectively after the pregnancy was confirmed. Group D was set up as a negative control group without HEV infection.
Different durations of fecal virus shedding were observed in groups A-C. Rabbits in group A (rHEV3) began fecal virus shedding at 1 week post inoculation (wpi) that lasted for an average of 13.5 ± 0.4 weeks; rabbits in group B (sHEV4) began shedding at 2wpi that lasted for 10.2 ± 0.5 weeks on average; and rabbits in group C (hHEV3) began shedding at 2wpi that lasted for 4.8 ± 0.7 weeks on average. The durations of fecal virus shedding were statistically different among the three groups, with group A shedding (13.5 ± 0.4 w) being significantly longer than that of group B (10.2 ± 0.5 w) and group C (4.8 ± 0.7 w) (P < 0.05) ( Table 2 and Figure 1). The sequences of HEV detected in group A, B and C feces were all in accordance with the inoculation viruses.

Characteristics of histopathology and immunofluorescence of HEVs infection
To clarify the histopathological changes in pregnant rabbits' tissues, we performed HE staining on rabbit livers and other organs ( Figure 2). In group A, hepatic hemorrhage was found in livers of pregnant rabbits, with a large number of red blood cells in hepatic sinusoids, extensive edema, degeneration and necrosis of hepatocytes and inflammatory cell infiltration; necrosis and degeneration of hepatocytes with inflammatory cell infiltration in liver tissue of pregnant rabbits of groups B and C were also observed, but the degree of inflammation was milder than that of group A; and no obvious pathological change in the liver tissue was observed in group D. In addition, the ovary tissue of group A showed slight inflammatory cell infiltration, but groups B, C and D showed no obvious pathological changes in ovary tissues.
To assess the replication efficiency of different genotypes of HEV in the liver, ovary and placenta, the expression of HEV ORF2 and ORF3 was detected by immunofluorescence staining. HEV ORF2 and ORF3 antigens were positive in the liver, ovary, and placenta of rabbits in groups A, B, and C but not in the organs of group D (Figure 2).

The estrogen levels and cellular immunity in HEV-infected pregnant rabbits
By regression analysis of serum estrogen levels and fecal viral load in group A after virus inoculation, a positive correlation between serum estradiol levels and fecal viral load in pregnant rabbits (r 2 = 0.4339) was observed, which indicated that estrogen levels were correlated with HEV infection and the higher the estrogen level, the higher the fecal viral load in pregnant rabbits was (Supplementary Fig. 2A).
with HEV infection in the liver, ovary, and placenta through IHC. IHC experiments on CD4 + and CD8 + T cells showed that the signals of CD4 + T cells and CD8 + T cells in the liver, ovary and placenta of pregnant rabbits infected with rHEV3 were all positive, and the expression of CD8 + T cells was higher than that of CD4 + T cells, but pregnant rabbits of group D showed low but similar levels of CD4+ T cells and CD8+ T cells in liver, ovary and placenta (Supplementary Fig. 2B).

HEV infection in newborn rabbits and vertical transmission
By analyzing HEV infection in surviving newborn rabbits delivered by pregnant rabbits infected with different genotypes of HEV, 50% (2/4) of surviving newborn rabbits of group A had fecal virus shedding after birth, 25% (1/4) were positive for anti-HEV antibody without HEV infection and 25% (1/4) did not have HEV infection; in surviving newborn rabbits of group B, 20% (1/5) had viremia after birth and 80% (4/5) did not have HEV infection; and in group C's surviving newborn rabbits, 100% (7/7) were without HEV infection. The HEV infection rate of surviving newborn rabbits in group A was significantly higher than that of surviving newborn rabbits in group C (P < 0.05) (Supplementary Table 1).
One newborn rabbit of group A (Aa) shed virus in its feces for 12 weeks, and the other (Ab) shed for 23 weeks. Both ALT and AST levels of these two rabbits significantly increased more than 2 times the baseline value and seroconversion of anti-HEV antibodies was also observed ( Figure 3A and 3B). The third newborn rabbit (Ac) maintained a low level of anti-HEV antibodies that decreased over time ( Figure 3C).
The liver tissues of euthanized newborn rabbits were analyzed by histopathology and immunofluorescence. Structural disorder, hepatocyte necrosis and inflammatory cell infiltration were observed in the liver tissue of Aa and extensive hepatocyte edema, balloon-like degeneration and hepatocyte necrosis were observed in the liver tissue of Ab. Immunofluorescence analysis showed positive signals for HEV ORF2 and ORF3 in the liver tissues of Aa and Ab ( Figure 3D).

Antibody responses after vaccination
Different doses of the HEV 239 vaccine were administered to pregnant rabbits to evaluate the protective efficacy. Four weeks after the first immunization in groups E and F, the serum anti-HEV antibody-positive rates of rabbits inoculated with the 10 and 5 μg dose HEV 239 vaccine were 83.3% and 66.7%, respectively. After the whole immunization process, all vaccinated rabbits were positive for anti-HEV antibodies, whereas the rabbits of the placebo group, G, were all negative for anti-HEV antibodies (Table 3).
To evaluate the dynamics of anti-HEV antibodies induced by different doses of the HEV 239 vaccine throughout the whole immunization process, the GMT was measured weekly. The GMT induced by the 10 and 5 μg dose vaccines gradually increased after the first immunization and increased rapidly after the second immunization ( Figure 4A).
Three weeks after the whole immunization process, the anti-HEV antibody concentrations of rabbits' serum were quantified and the average anti-HEV antibody concentrations were compared among the groups. The results showed that the average anti-HEV antibody concentrations between group E and group F were not significantly different (P > 0.05) and that the average anti-HEV antibody concentrations in groups E and F were higher than that in group G (P < 0.05) ( Figure 4B).  2×0 0/6(0) 0/6(0) 0/6(0) a A week after the first immunization. b Before the second immunization. c A week after the second immunization.

Protective efficacy of the HEV 239 vaccine in pregnant rabbits
After pregnancy was confirmed, rabbits in groups E, F and G were inoculated with the rHEV3 virus on the 10th day post copulation. None of the rabbits in groups E and F developed HEV infection symptoms, such as viremia, fecal virus shedding and elevation of ALT and AST levels, with only one rabbit in group F having a stillbirth. In contrast, all rabbits in group G began seroconversion to anti-HEV antibodies and fecal virus shedding at 2 wpi. Adverse pregnancy outcomes, including miscarriage, stillbirth, and death, occurred at 4 wpi, indicating that HEV can successfully infect rabbits without HEV 239 vaccination. Statistical analysis showed that compared with placebo administration, two doses of the 10 μg HEV 239 vaccine not only protected pregnant rabbits from rabbit HEV infection but also prevented the occurrence of adverse pregnancy outcomes (P < 0.05). Two doses of the 5 μg HEV 239 vaccines could protect pregnant rabbits from rabbit HEV infection, with only one case of stillbirth occurring, which also had a significant difference compared with placebo administration (P < 0.05) ( Table 4).
Serum anti-HEV antibody levels of newborn rabbits in groups E and F were followed. From the 3rd to 8th week after birth, the levels of anti-HEV antibodies continued to decrease and then remained at the same level, while the anti-HEV antibody levels in the pregnant rabbits of groups E and F decreased slightly and remained at the same level. The initial level of anti-HEV antibody in newborn rabbits in group E was higher than that in group F but remained at the same level eventually ( Figure 4C).

Discussion
Although studies have suggested that mainly HEV1 infection is associated with high mortality in pregnant women [10], there is limited research to investigate the influence of other HEV genotypes. In this study, the incidences of adverse pregnancy outcomes were not significantly different among the three groups of rabbits infected with rHEV3, sHEV4 and hHEV3, indicating that different genotype HEVs with the same viral load can all cause adverse pregnancy outcomes. This study is the first time that rabbits have been able to be infected by hHEV3, which may be due to the use of different subgenotypes compared with other studies [20] or because that pregnant rabbits are more prone than nonpregnant rabbits to hHEV3 infection.  However, the manifestations of different HEV genotypes infection, including fecal virus shedding duration, the anti-HEV S/CO levels and pathological changes, were different among the groups, and rabbits infected with rHEV3 showed more obvious symptoms than the other groups. The vertical transmission of HEV has always attracted much attention [21][22][23], and several studies have shown that HEV can replicate in the human placenta [11,24]. In this study, the positive and negative strands of HEV RNA were detected in placentas and immunofluorescence results showed that both HEV ORF2 and ORF3 signals were positive in the placenta, suggesting that HEV may replicate in the placenta, possibly leading to vertical transmission. We also found that the ovaries of HEV-infected pregnant rabbits were positive for HEV ORF2 and ORF3, indicating that HEV may also replicate in rabbit ovaries. Two of the surviving newborn rabbits (Aa and Ab) delivered by HEVinfected pregnant rabbits showed symptoms of HEV infection, suggesting possible vertical transmission.
The alteration in hormone secretion during pregnancy may be associated with hepatitis E-related poor pregnancy outcomes, and estrogen has been proven to be able to promote HEV replication in vitro [25]. Our results showed that the estrogen level was positively correlated with fecal viral load, suggesting that estrogen may be able to promote HEV replication in pregnant rabbits. The decrease in the CD4/CD8 ratio in blood may be related to high mortality and adverse outcomes during pregnancy caused by HEV infection [26,27]. In this study, in accordance with the lowered CD4/CD8 cell ratio in blood samples [26,27], the IHC results of tissues showed that compared with the negative control rabbits, pregnant rabbits infected with HEV were positive for both CD4 + and CD8 + T cells in the liver, ovary, and placenta and that the proportion of CD8 + T cells was higher than that of CD4 + T cells.
In this study, similar to previous studies [28,29], our results showed that the HEV 239 vaccine can effectively prevent rabbit HEV infection in pregnant rabbits. For the first time, our results showed that the HEV 239 vaccine can also prevent the occurrence of adverse pregnancy outcomes in pregnant rabbits. Different doses of vaccination in pregnant rabbits could produce anti-HEV antibodies without significant differences in immunization efficacy. We also noted that anti-HEV antibodies could be detected in newborn rabbits delivered by immunized pregnant rabbits, which may be related to breastfeeding of female rabbits. The anti-HEV antibody persisted in newborn rabbits, and different doses of the HEV 239 vaccine had the same efficacy on newborn rabbits.
In summary, our results showed that different genotypes of HEV can infect pregnant rabbits and lead to adverse pregnancy outcomes and vertical transmission. The HEV 239 vaccine can effectively protect pregnant rabbits from rabbit HEV infection and adverse pregnancy outcomes. Although the animal numbers in the present study are limited, our results still suggest that to avoid hepatitis E-related adverse pregnancy outcomes and vertical transmission, childbearing-age women should be immunized with the HEV vaccine before pregnancy.