H7N9 is a recently identified subtype of influenza A virus that caused a major outbreak in humans in China in 2013. According to the latest data provided by the Chinese Center for Disease Control and Prevention (http://www.moh.gov.cn/zwgk/yqbb3/ejlist.shtml, updated on October 31, 2018), the mortality rate of H7N9 infections in China amounts to 39.7% (611/1536). Thus, H7N9 poses a serious public health threat.
Influenza A viruses comprise a group of antigenically diverse pathogens that cause respiratory tract infections in both animal and human populations (Thornburg et al. 2016). Subtypes of influenza A viruses are characterized by their surface glycoproteins, including hemagglutinin (HA) and neuraminidase (NA). HA consists of a globular head region with a receptor-binding pocket and a conserved stem motif (Liu et al. 2016), and mediates the binding of influenza A virus to host cells (Melikyan et al. 1999). After binding, HA can be divided into three domains: an extracellular domain, a transmembrane domain, and a cytoplasmic tail (CT) (Thornburg et al. 2016). More than a dozen of neutralizing epitopes of HA have been identified to date (Shcherbinin et al. 2016), most of which are on the receptor-binding pocket (Whittle et al. 2011) and the stem motif (Dreyfus et al. 2012). However, H7-subtype inactivated virus is a weak inducer of neutralizing antibody compared to H1N1 and H3N2 (Lee et al. 2015), highlighting the necessity of developing novel strategies in vaccine design against H7N9.
This study aimed to identify a novel antigenic epitope on H7N9 virus HA7. In total, 37 patients with H7N9 infection admitted to Shenzhen Third People's Hospital between December 2012 and March 2015 were included in this study. Plasma and peripheral blood mononuclear cells were separated from fresh blood samples and were stored at -80 ℃.
Full-length HA protein [of strain A/Shanghai/02/2013 (H7N9)] was fragmented into 110 peptides (P1–P110). Each peptide contained 15 amino acids (AAs), with a 10-AA overlap between two neighboring peptides. A library of 108 peptides covering the full-length HA7 was synthesized except for P9 and P18. Plasma samples from 11 H7N9-infected patients were tested for reactivity with the 108 synthetic peptides by peptide microarray ELISA. As shown in Fig. 1A, the OD450 values for P15, P21, P30, P63, P76, P98, and P110 were significantly higher than those for other peptides, suggesting that the patients' plasma strongly reacted with these peptides, which may be candidate epitopes of HA7. Among the seven candidates, P110 (AA sequence: FICVKNGNMRCTICI), the last 11 AAs of which compose the HA7-CT, showed the strongest immunoreaction, whereas the specific affinities of serum antibodies to peptides on extracellular and transmembrane domains were lower. Therefore, the HA7-CT peptide (KNGNMRCTICI) was further investigated in this study. The plasma samples from the 37 H7N9 patients were tested for reactivity with HA7-CT peptide, 40 healthy plasma samples and 10 cord blood samples were used as controls. As shown in Fig. 1B, the OD values in the H7N9 group were significantly higher than those in the control groups. Plasma samples from two H5N6 patients also presented higher affinity than control samples. These findings suggested that the HA7-CT epitope might be specifically recognized by influenza A virus-infected human plasma. Time course analysis of the immune response of plasma of different patients to the HA7-CT epitope revealed that immune reactivity time-dependently increased from day 10 to day 20 post disease onset, peaked on day 20, and then remained constant until day 96 (Fig. 1C). Taken together, these results suggested that the HA7-CT epitope is a potentially antigenic epitope that could be recognized by the plasma of patients infected with influenza A virus and may drive a strong immune response after the onset of virus infection.
Figure 1. Immunogenicity of the cytoplasmic tail (CT) of H7N9 virus hemagglutinin (HA). A Full-length sequence of HA protein [of the strain A/Shanghai/02/2013 (H7N9)] was fragmented into 110 peptides (P1–P110) that were synthesized, except for P9 and P18. The plasma samples collected from H7N9-infected patients (n = 11) were labelled based on the number of patient (before the hyphen) and days post disease onset (after the hyphen). The samples (diluted at 1:100) were incubated with 5 μg/mL peptides pre-coated in the ELISA plate. OD450 values represent immunoreactivity. Peptides with OD450max- > 0.2 were regarded epitopes of HA. B HA7 CT peptide (i.e., P110) was incubated with plasma samples from H7N9- (n = 37) or H5N6-infected (n = 2) patients (collected at 21–42 days post disease onset), or healthy controls (HC; n = 40), or incubated with cord blood (CB; n = 10) samples. ELISA was conducted. C HA7-CT peptide was incubated with plasma samples collected from the H7N9-infected patients at different times post disease onset. ELISA was conducted. D HA7-CT antibody titer (log10OD450) and neutralization titer (log10dilution fold) of plasma samples from H7N9-infected patients (X axis from 1 to 99) and healthy donors (X axis from 100 to 136). Pearson correlation analysis of HA7-CT-specific antibody titer with plasma neutralization titer (E), hemagglutination inhibition titer (F) and antibody-dependent cell-mediated cytotoxicity percentage (G). H The affinity of HA7-CT-specific antibody pulled down from H7N9 patients' total IgG with HA7 extracellular peptide, HA7-CT, or purified virions was evaluated by ELISA. Total IgG of H7N9-infected patients and healthy controls (HC) were used as controls. I Sequence similarity and antigenic index prediction of HA-CT. *P < 0.05; **P < 0.01; ***P < 0.001. HA hemagglutinin, CT cytoplasmic tail, OD optical density, PA the average OD450 values of all peptides to a sample, HC healthy control, CB cord blood, TM transmembrane domain.
HA-CT (Siche et al. 2015) plays an important role in virus replication (Imai et al. 2012), viral infectivity (Siche et al. 2015), and virus entry into host cells (Scolari et al. 2016). Although its functions have been well studied, little is known about the immunogenicity and sub-localization of HA-CT. Previous studies have identified a specific neutralizing antibody against the CT of human immunodeficiency virus glycoprotein 41 (Chen et al. 2015), which led us to hypothesize that HA-CT may function in a similar manner. To investigate the immunogenicity of HA7-CT epitope further, we pulled down CT-specific antibody from total IgG in the plasma samples from five H7N9 patients using biotinylated HA7-CT peptide. We analyzed the relationship between the titer of HA7-CT-specific IgG (measured photometrically) and the total IgG titer in H7N9 patient plasma (measured by virus neutralization assay in H7N9-infected Madin–Darby canine kidney (MDCK) cells or by hemagglutination inhibition (HI) assay). As shown in Fig. 1D, the HA7-CT-specific IgG titer showed a pattern similar to that of the plasma neutralization titer. Further-more, the HA7-CT-specific IgG titer was significantly positively correlated with the plasma neutralization titer (P < 0.0001; Fig. 1E) and the plasma HI titer (P = 0.0303; Fig. 1F), but not with the antibody-dependent cell-mediated cytotoxicity percentage (P = 0.5318; Fig. 1G). The IgG titers of the other candidate epitopes, i.e., P15, P21, P30, P63, P76, and P98 (data not shown), were not correlated with the plasma neutralization titer. These results suggested that HA7-CT possesses an antigenic epitope that is capable of blocking virus infection by neutralization, but not of the antibody-dependent cell-mediated cytotoxic effects.
Because surface protein of the influenza virus is a major antibody target, we next sought to investigate whether HA7-CT is displayed on the H7N9 surface and thus can be recognized by antibodies. We pulled down HA7-CT-specific IgG from total IgG in plasma samples of H7N9 patients and tested its affinity with HA7-CT peptide, HA7 extracellular domain peptide, and inactivated H7N9 virion by ELISA. MDCK cell culture supernatant was used as a blank control. As shown in Fig. 1H, compared with total IgG, HA7-CT-specific IgG showed significantly low binding affinity with the HA7 extracellular domain peptide (P < 0.01), but markedly high binding affinity with the HA7-CT peptide (P < 0.05), indicating that HA7-CT-specific IgG can be specifically recognized by the HA7-CT peptide. On the other hand, there was no difference in binding affinity for inactivated virions between HA7-CT-specific and total IgG, suggesting that HA7-CT-specific and total IgG can bind to whole-virus particles with similar affinity. These results suggested that HA7-CT is displayed on the virus surface, at least occasionally.
The CT of HA consists of 11 AAs. Based on HA and CT sequences of various subtypes of influenza A virus retrieved from the influenza database (https://www.fludb.org) and weblogo3.5 (http://weblogo.threeplusone.com/), respectively, we found that the CT sequence was highly conserved, especially within virus clades. As shown in Fig. 1I, all sequences analyzed shared a CXICI motif at the C terminus. The CT sequences of HAs in group 1 were more similar than those in group 2. Of note, H1, H2, and H5 in clade H1a had identical CT sequences. To explore whether the CTs of different influenza A subtypes commonly possess antigenicity, we analyzed the antigenic indexes of the CTs predicted by the Jameson-Wolf model. The CTs of clades H9, H3, and H7 had significantly higher antigenic index values than those of clade H1a (H1, H2, H5), consistent with our experimental results.
Currently available influenza vaccines are mainly HA- or NA-specific neutralizing antibodies (Gerhard et al. 2006). However, because of the high antigenic variability of HA and NA, it is difficult to produce a universal and effective vaccine to prevent epidemics (Carrat and Flahault 2007; Deng et al. 2015). We found more than 50% sequence similarity in the CT sequences of different HA subtypes (Fig. 1I). The highly conserved CT sequence and high antigenic index scores indicated that HA-CT may serve as a potential antigenic epitope for the development of a broad-spectrum influenza antibody.
Our study had some limitations. First, we used short linear peptides to coat the ELISA plates, and we cannot guarantee consistent coating efficiency. Second, total IgG in the plasma may recognize conformational epitopes on HA, but the linear peptide-based ELISA is incapable of detecting conformational epitope-specific antibodies. Third, we showed that HA7-CT-specific antibody titer was positively correlated with the plasma neutralization and HI titers, but we did not present direct evidence to prove that HA7-CT specific antibody had neutralizing ability. We hypothesize that HA7-CT is a potential neutralizing epitope that could be utilized in universal vaccine design, but further research will be required to verify our results.
Identification of a Novel Universal Potential Epitope on the Cytoplasmic Tail of H7N9 Virus Hemagglutinin
- Received Date: 10 July 2018
- Accepted Date: 05 March 2019
- Published Date: 23 April 2019
Abstract: This study aimed to identify a novel antigenic epitope on H7N9 virus HA7. In total, 37 patients with H7N9 infection admitted to Shenzhen Third People's Hospital between December 2012 and March 2015 were included. A library of 108 peptides covering the full-length HA7 was synthesized except for P9 and P18. Plasma samples from 11 H7N9-infected patients were tested for reactivity with the 108 synthetic peptides by peptide microarray ELISA. Results suggested that the HA7-CT epitope is a potentially antigenic epitope that could be recognized by the plasma of patients infected with influenza A virus and may drive a strong immune response after the onset of virus infection.