It has been previously reported that anti-prM mAbs provided cross-protection against all four DENV serotypes [40, 55]. However, the potential importance of these prM-containing particles and prM antibodies in DENV infection pathogenesis has been ignored for a long time as numerous functional studies have reported that fully immature particles are noninfectious . Interestingly, recent studies on human and mouse anti-prM mAbs [24–32] suggest that prM-specific mAbs have a significant role to enhance infection of standard DENV and imDENV particles. However, there have been few attempts to locate the epitopes of prM ptotein. To gain a deeper understanding of the antigenic structures of prM and their functions in human immune response to DENV, we identified the epitope of prM mAb 4D10 and investigated the ability of mAb 4D10 and antibody against epitope peptide PL10 to mediate ADE infection of standard DENV1-4 and imDENV particles.
In this study, we generated and characterized a DENV serocomplex cross-reactive prM mAb 4D10. Then, we successfully mapped the epitope of 4D10 to amino acid residues 14 to 18 of DENV1-4 prM protein using phage display technology. The epitope peptide showed conformity with one region (amino acid residues 12 to 26) predicted by bioinformatics analysis. Consequently, the epitope peptide (13IVSRQEKGKS22) was synthesized for further study. We confirmed that PL10 was a DENV serocomplex cross-reactive epitope peptide and showed to be highly immunogenic in Balb/c mice. Also, PL10 could successfully distinguish DENV serotypes from other flaviviruses in immunized mice sera. The high degree of antibody cross-reactivity among different flaviviruses has been a diagnostic challenge to distinguish various flaviviral infections, and this limitation is apparent for members of DENV serotypes [57, 58]. It has been previously reported that prM-specific antibodies could be applied as a diagnostic marker to distinguish previous infection of DENV from JEV . Thus, it is remarkable that the DENV-specific epitope in prM has great potential to improve DENV serological diagnostic tests. Furthermore, PL10 could successfully recat with DENV2-infected patient sera but not with sera of healthy donors, suggesting that the epitope peptide PL10 could possibly be used as a serologic reagent in the diagnosis of DENV-infected patients.
The control peptide PH10 (3LTTRGGEPHM12) may be the possible epitope region of prM protein predicted by bioinformatics analysis, but the antibody titer of PH10 was not high enough. For synthetic peptides to serve as effective immunogens, they must comprise potential antigenic sites to promote B cell interaction .
Immature particles produced in furin-deficient LoVo cells have very high levels (94%) of prM-containing particles. Interestingly, both mammalian cells (BHK-21 or Vero) and insect cells (C6/36) infected with DENV release as many as 30% prM- containing immature particles [42, 60] suggesting that cleavage of prM to M is not very effective. Therefore, cells infected with DENV release a heterogeneous mixture of not only fully mature(containing M) and immature (containing prM) but also partially mature virus particles (containing prM and M) [42, 61, 62]. Previous studies have demonstrated that mature particles and partially mature particles were infectious whereas immature particles were virtually non-infectious [24, 27, 42, 56]. The maturation state of virus particles can influence the neutralizing and enhancing capacity of antibodies direct against DENV surface proteins [24, 27, 63]. We detected the specific infectivity of the LoVo-released virus particles and found that the infectious properties of imDENV2 was 10,000-fold lower compared to that of C6/36-cultured standard virus preparations. This agrees with previous results [27, 42] and proves that immature virus is virtually non-infectious.
Antibodies induced by DENV infection may have dual roles: obstruct infection through neutralization activity or enhance viral infection via ADE activity. Consistent with prior studies [24–27, 31, 41, 42], the mAb 4D10 and antibody against epitope peptide PL10 described in the present study showed broad cross-reactivity and poor neutralizing activity with the four standard DENV serotypes and imDENV but significantly enhanced the infectious properties. These results suggested 4D10 and anti-PL10 sera were infection-enhancing antibodies and PL10 was infection-enhancing epitope. We found mAb 4D10 and antibody against PL10 showed different neutralizing against different virus strains, suggesting the existence of structural differences in the epitope region. The mechanism of virus neutralization and ADE in the presence of antibody against prM is still elusive. Consistent with these results, during protection assay in vivo, our data clearly suggested the epitope peptide PL10 indeed elicit enhancing antibodies and promote DENV replication. The partial neutralization of antibodies against prM to standard dengue viruses implies that some infectious particles within the virus preparation are partially mature (containing a mixture of prM and M) and also indicates that prM antibodies have the capacity to block the infectivity of partially mature particles. Meanwhile, partial cleavage of prM from the viral surface reduces available antigens for neutralization activity. The cross-reactive among four DENV serotypes, together with partial cleavage of prM, makes dengue viruses susceptible to ADE by antibody against prM [24, 56]. It was recently shown that anti-prM antibodies could render essentially non-infectious imDENV particles highly infectious. The prM antibodies bind to the virion surface prM antigens and facilitate efficient binding and cell entry of virus-antibody complexes into Fc receptor-bearing cells following which the endosomal furin clears prM into M and renders immature particles infectious [24, 27]. Taken together, our results support the notion that antibodies against prM can enhance infectivity of prM-containing immature and partially mature DENV particles due to an interaction with Fc receptor expressed on immune cells.
As shown in Figure 8A and B, at low dilution of antibody, enhancing and neutralizing activities are mixed and enhancement is suboptimal, so little or no infection enhancement is found. Then, enhanced viral growth occurs at a higher dilution. At some dilution of antibody, optimal viral infections occur and peak enhancement is observed. At a still higher dilution, the concentration of infectious antibody–virus complexes is not great enough to elicit the system response and the infection enhancement is gradually lost . The peak infection enhancement also need a large number of virus receptors on FcR-bearing cells, the efficient cell entry of virus, the viability of virus in the cytosol, and capability to accomplish all steps to achieve assembly and final release of virus particles.
Since recent studies found that DENV particles released from infected cells contained as many as 30% prM particles, the infectious potential of immature particles may have significant implications for understanding of the dengue pathogenesis. In the early stages of a primary infection, immature particles fail to enter host cells in the absence of antibodies, and therefore are of minor importance in disease development. On the other hand, prM-specific antibody response will activate the infectivity of fully immature particle upon secondary infection, and increase the number of infectious particles.
The epitope recognized by our own anti-prM antibody was located in amino acid residuals 14–18 of the prM protein and was different from the published sequence recognized by other anti-prM mAb 2H2 (mapped to amino acid residuals 40–49) and 70-21 (mapped to amino acid residuals 53–67) [40, 41]. Previous studies have shown that 2H2 provided cross-protection against all four DENV serotypes [40, 55]. However, many studies demonstrated that 2H2 could enhance the infectivity of standard DENVV and imDENV [27, 65, 66]. Also, antibody 70–21 as well as many other prM mAbs has been reported to enhance DENV infectivity [24, 26, 27, 31]. Our results support that anti-prM antibodies could enhance infectious properties of DENV and prM epitopes could be not protective but infection enhancing. We propose that the length of epitope sequence has an important role to mediate ADE infection. For long epitope peptide sequences, they may contain two or more epitopes, which may be immunodominant or cryptic. These findings suggest that antigenic structures of prM and their functions are complicated and not well studied.
Most current dengue vaccines contain native dengue prM, it may be important to consider better vaccine approaches that eliminate ADE activities induced by infection-enhancing epitopes on prM during vaccine design . Vaccine candidates that eliminate pathogenic infection-enhancing epitopes may thus become increasingly important. Most importantly, identification of the epitopes on prM protein will provide new insights for further understanding of humoral immune responses to DENV at the epitope level. However, to our knowledge, there have been few reports of epitopes mapping to aim at prM protein. Here, we report a novel infection-enhancing epitope on dengue prM, the findings from our study may have significant implications for future vaccine design and facilitate understanding the pathogenesis of DENV infection.